Issues

Economic
• it

Fourth Edition
AGR
II^JWmJRAL. ECONOMIC

Gerald M. Meier
 iiiiinjpr
Gianni ni FDN Library

ling Issues in
lomic Development
Leading Issues in
Economic Development

FOURTH EDITION

GERALD M. MEIER
Stanford University

New York Oxford

OXFORD UNIVERSITY PRESS
1984
Copyright © 1964, 1970, 1976, 1984 by Oxford University Press, Inc.

Library of Congress Cataloging in Publication Data

Main tntry under title:
Leading issues in economic development.
Includes index.
1. Economic development. I. Meier, Gerald M.
HD82.L3273 1984 338.9 83-19372
ISBN 0-19-503415-5 (pbk.)

Printing (last digit): 987654321

Printed in the United States of America
 G1ANNIN1

For Daniel
PREFACE TO THE FOURTH EDITION

The overriding objective of this fourth edition of Leading Issues in Economic Development
remains the same as stated in the first edition — to present a new kind of course book that will
bring some order to the diffuse subject of economic development without sacrificing a variety
of viewpoints and different perspectives. The evolution of the subject confirms my belief that
it can best be studied in a manner different from the usual approaches of either a textbook
or a book of readings. Development practitioners must operate in an imperfect, second-best
world, and development economists cannot yet claim that their subject constitutes a self-con-
tained discipline. I believe that the student's most sensible introduction to economic devel-
opment is through a study of several "leading issues" that are at present a major pre-
oc upation ofboth development economists and development practitioners.
This new edition reflects the current development mood of stock-taking and reassessment.
It focuses on an appraisal of development experience and the reinterpretation of development
thought in order to highlight future policy options. After four decades of struggling against
poverty, the international development community is asking what has gone right? What has
gone wrong? And still we must be concerned with future efforts to lessen the pain of poverty
suffered by two-thirds of mankind. But we must also put the future of the development move-
ment in perspective by building on the lessons of development experience. From that experi-
ence, there should emerge fresh thinking on the leading issues in economic development.
I have chosen to concentrate on a relatively few issues that are now of central concern to
development professionals. On each of these strategic issues, I have brought to bear a variety
of materials that should be read together. To provide additional direction and cohesion, I have
written a substantive commentary through a series of connecting text Notes. In some in-
stances,have
I introduced a Note to treat a topic more concisely than could be done through
separate readings; in other instances, a Note is designed to cover a topic that is not yet ade-
quately treated in the literature, or to tie one issue with another.
Even more than in earlier editions, I want to emphasize the importance of the interrelat-
edness of the readings on each issue — taken as a set — rather than any one particular reading.
Each selection acquires added significance through its contextual position, and the materials
are enhanced by their very combination. This is especially true for the interrelations among
the various materials that deal with analysis, policy implications, and the lessons of develop-
ment experience. The major focus is on the analytical core of the subject. At the same time,
the analysis is made as policy-oriented as possible. And the materials can be readily supple-
mented bycountry cases and empirical studies.
The changes in this edition are extensive. I have introduced new chapters. I have also ex-
tensively modified all the earlier chapters, replacing many of the older selections with new
selections that emphasize more current topics. I have tried to strengthen the entire book by
eliminating the merely descriptive or wordy selections of previous editions and by including
new selections that will deepen the student's understanding of fundamental economic prin-
ciples and of empirical relationships. I have added more of my own Notes and revised the
Notes contained in previous editions.
viii PREFACE TO THE FOURTH EDITION

This edition also introduces Comments and Exhibits. The Comments are intended primar-
ily to elaborate on selections, to raise questions, or to suggest additional readings on a topic
that is not adequately covered in this book. The Exhibits present charts and tables that give
more empirical content to the subject.

I wish to express my appreciation to the authors and publishers who have granted permis-
sion to use excerpts from articles, books, and other publications for which American or foreign
copyrights exist. Specific acknowledgment is given with each selection. Some parts of the
original versions of the selected materials have been (silently) omitted out of consideration
for relevancy and to avoid repetition. In some cases, tables and diagrams have been renum-
bered; inother cases, footnotes have been omitted, while others have been renumbered.
As in previous editions, many of the revisions in this edition have been inspired by my
students at Stanford University and by lecture audiences in developing countries. Their chal-
lenging questions and incisive observations have meant much to me. I hope this volume will
prove of value to another generation of students concerned with the future of the developing
world. This hope has sustained me through the laborious process of preparing this new edition.
My thanks are also due to Timothy Quey, Sridhar Moorthy, and Srikant Datar who re-
lieved me of many library chores. And a very special acknowledgment is due Pat Sharp who
has taken the roughest of copy and, with remarkable good humor and admirable efficiency,
has guided the manuscript from handwritten scrawls through the word processor to the
printer. Without the speedy imprint of her skills, this new edition would have been postponed
indefinitely.
Finally, I wish to acknowledge the assistance I received while a Visiting Professor at the
Economic Development Institute of the World Bank as well as the advice of Bruce F. John-
ston, Walter P. Falcon, Derek Healey, and Paul P. Streeten — and the entire profession of
development economists whose writings have contributed to this volume.

January 1984 G.M.M.

Stanford, California
CONTENTS

I. THE DEVELOPMENT RECORD l

LA. OBJECTIVES OF DEVELOPMENT— NOTE 5
LB. INDICATORS OF DEVELOPMENT 15
I.B.I. Measuring Development — Ahluwalia and Chenery 15
I.B.2. Dimensions of Poverty— World Bank 20
, I.B.3. A Basic Needs Yardstick — Hicks and Streeten 26
I.B.4. The Inverted-U Hypothesis— Fields 30
I.B.5. Growth-Distribution Tradeoffs — Chenery 36
I.C. COUNTRY PERFORMANCE 45
I.C.I. Twenty-Five Years of Development — Morawetz 45
I.C.2. Korea — Bergsman 56
I.C.3. Brazil— Knight and Moran 62
I.C.4. Sub-Saharan Africa— World Bank 66
I.C.5. China— Perkins 69
I.C.6. Poverty Elimination and Entitlements — Sen 78
I.D. LEADING ISSUES— NOTE 85

II. DEVELOPMENT PERSPECTIVES 89

II.A. HISTORICAL PERSPECTIVES 91
II.A.l. Rostow and Marx — Note 91
H.A.2. Empty Economic Stages? — Fishlow 94
II.A.3. Economic Backwardness in Historical Perspective — Gerschenkron 101
II.A.4. Future Development in Historical Perspective — Note 104
II.B. ANALYTICAL PERSPECTIVES 111
II.B.l. Contrasting Perceptions — Streeten 111
II.B. 2. Structuralist Approach to Development Policy — Chenery 118
II.B. 3. Changes in Development Theory — Ranis 123
II.B. 4. Development Modeling: State of the Art — Tinbergen 129
II.C. RADICAL PERSPECTIVES 133
II.C.l. A Marxist View— Gurley 133
II.C.2. Dependency Theories of Underdevelopment — Palma 1 39
II.D. ECONOMICS FOR DEVELOPMENT— NOTE 145
 CONTENTS

III. SURPLUS LABOR AND DUALISTIC DEVELOPMENT 149

HI. A. THE LABOR SURPLUS ECONOMY 153
III.A.l. Labor Surplus on the Land — Nurkse 153
III.A.2. The Concept of "Disguised Unemployment" — Viner 155
III.A.3. Choice of Techniques — Sen 159
III.B. DUAL SECTOR MODELS 162
III.B.l. Intersectoral Relationships in a Dual Economy — Note 162
III.B.2. Technological Dualism— Note 167
III.B.3. Dualism and Wage Differentials— Myint 170
III.C. RURAL-URBAN MIGRATION 176
III.C.l. A Model of Rural-Urban Migration— Todaro 176
III.C.2. The Informal Sector— ILO 183
HID. GROWTH WITHOUT EMPLOYMENT— NOTE 191
HIE. MEASUREMENT OF THE "UNEMPLOYMENT EQUIVALENT"—
NOTE 196
III.F. EMPLOYMENT STRATEGIES 198
III.F.l. Output and Employment Trade-offs — Stewart and Streeten 198
III.F.2. Technology and Employment— Stewart 207
III.F.3. Policy Implications— Note 210

IV. MOBILIZING DOMESTIC RESOURCES 215

IV. A. ROLE OF CAPITAL 219
IV.A.l. Capital Accumulation and Development — ECAFE 219
IV.A.2. Criticisms of the Capital-Output Ratio— Note 222
IV. A. 3. The Place of Capital in Economic Progress — Cairncross 225
IV.B. SOURCES OF CAPITAL FORMATION— NOTE 234
IV.C. FINANCIAL DEEPENING 240
IV.C.l. Financial Structure and Economic Development — Gurley and Shaw 240
IV.C. 2. Organized and Unorganized Money Markets — Tun Wai 246
IV.C. 3. Demand-Following or Supply-Leading Finance — Patrick 251
IV.C.4. Financial Repression and Liberalization — McKinnon 254
IV.D. INFLATION AND ITS EFFECTS 259
IV.D.l. Effects of Inflation— Lewis 259
IV.D. 2. Inflation and Development Policy — Johnson 263
IV.D. 3. Monetarism and Structuralism in Latin America — Campos 268
IV.D.4. Causes and Consequences of Inflation — Killick 274

V. MOBILIZING FOREIGN RESOURCES 281

V.A. PUBLIC FINANCIAL AID 283
V.A.I. Calculation of Capital Requirements — Hawkins 283
V.A.2. Growth-and-Debt Stages— De Vries 286
V.A.3. Adverse Repercussions of Aid— Bauer and Yamey 293
xi CONTENTS

V.A.4. A Renewed Aid Mandate — Lewis 298

V.A.5. Improving the Quality of Aid — Note 302
V.B. EXTERNAL DEBT AND COUNTRY RISK ANALYSIS— NOTE 310
V.C. THE INTERNATIONAL MONETARY SYSTEM AND
DEVELOPMENT FIN ANCE— NOTE 3 15
V.D. PRIVATE FOREIGN INVESTMENT 322
V.D.I. Benefits and Costs of Private Foreign Investment — Note 322
V.D.2. Transnational Enterprises and Linkages — Lall 332
V.D. 3. Multinational Enterprises in Developing Countries — Note 339
V.E. TRANSFER OF TECHNOLOGY 345
V.E.I. Appropriate Technology — Stewart 345
V.E.2. Capital-Stretching Innovations — Ranis 349
V.E. 3. Factor Intensity— Pack 352

VI. INDUSTRIALIZATION STRATEGY 357

VIA. PATHS TO INDUSTRIALIZATION 361
VI.A.l. The Theory of the "Big Push" — Rosenstein-Rodan 361
VI.A.2. Critique of the "Big Push" Argument— Ellis 365
VI. A. 3. Linkage Effects and Industrialization — Hirschman 368
VI.A.4. The Case for Balanced Growth— Nurkse 373
VI.A.5. Balanced versus Unbalanced Growth — Streeten 377
VLB. LESSONS OF RECENT HISTORY 381
VI.B.l. The Industrial Achievement — Hughes 381
VLB. 2. Industrialization via Import Substitution — Note 388
VLB. 3. Industrialization via Export Substitution — Note 394
VLB.4. Interactions Between Industrialization and Exports — Chenery 397
VLC. AGRICULTURE-INDUSTRY INTERACTIONS 406
VI.C.l. Industrialization and Agricultural Development — Bacha 406
VI.C.2. Complementarity of Industry and Agriculture — Lewis 413
VLC. 3. Overcoming the Weakness of Agriculture — Lewis 418

VII. AGRICULTURAL STRATEGY 423

VILA. IMPORTANCE OF AGRICULTURE 427
VII.A.l. Agriculture's Contributions to Development — Note 427
VII.A.2. Strategies for Transferring Agricultural Surplus — Lee 431
VILA.3. Urban Bias— Lipton 437
VII.B. FOODGRAIN REVOLUTION 441
VII.B.l. The Green Revolution: Seven Generalizations — Ruttan 441
VII.B. 2. Generations of Problems — Falcon 446
VII.C. DESIGNING AN AGRICULTURAL STRATEGY 453
VII.C.l. The Case for a Unimodal Strategy — Johnston 453
VII.C. 2. A Three-Pronged Strategy for Rural Development — Johnston and
Clark 461
 CONTENTS

VII.C.3 The New Rural-Based Development Strategy — Mellor 464

VII.C.4 Land Reform — Warriner 468
VII.C.5 Price Incentives — Brown 471
VII.C.6 Food Prices — Timmer 476

VIII. TRADE STRATEGY 485

VIII.A. INTERNATIONAL TRADE AND INTERNATIONAL
INEQUALITY 489
VIII.A. 1. Trade as an "Engine of Growth"— Note 489
VIII.A.2. Patterns of Trade and Development— Nurkse 493
VIII.A.3. Trade as a Mechanism of International Inequality — Myrdal 498
VIII.A.4. Conditions of Export-Led Development — Note 503
VIII.B. INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT
STRATEGIES 510
VIII.B. 1. Trade Policy as an Input to Development — Krueger 510
VIII.B. 2. Phases in Exchange Control Regimes — Bhagwati 514
VIII. B.3. Inward-Oriented Strategies — Balassa 516
VIII.B. 4. Trends and Prospects for Export of Manufactures — Note 522
VIII.B. 5. Export Promotion: Lessons of Experience — Note 538
VIII.B. 6. The Slowing Down of the Engine of Growth — Lewis 546
VIII.C. TRADE IN PRIMARY PRODUCTS 553
VIII.C.l. Techniques of Price Raising — Maizels 553
VIII. C.2. International Resource Bargaining — Note 555
VIII.D. REGIONAL INTEGRATION AND DEVELOPMENT— NOTE 561

IX. HUMAN-RESOURCE DEVELOPMENT 567

IX.A. POPULATION 569
IX.A.l. Population and Poverty— Note 569
IX.A. 2. The Demographic Momentum — Berelson 572
IX.A. 3. Fertility Determinants— Birdsall 576
IX.A.4. Economics of Population Quality — Schultz 580
IX.B. HEALTH AND NUTRITION 584
IX.B.l. Intercountry Variations in Health — Golladay and Liese 584
IX.B.2. Nutrition Objectives— Berg 591
IX.B.3. Nutrition Actions— Berg 597
IX.C. EDUCATION 610
IX.C.l. Investment in Human Capital — Note 610
IX.C. 2. Educational Investment — Musgrave 614
IX.C. 3. Approaches to Human-Resource Development — Harbison 619
IX.C. 4. Education and Employment — Edwards and Todaro 626
IX.C. 5. Development as a Generalized Process of Capital Accumulation —
Johnson 631
xiii CONTENTS

X. PROJECT APPRAISAL 637

X.A. GENERAL RATIONALE OF PROJECT APPRAISAL— NOTE 639
X.B. BENEFITS AND COSTS 643
X.B.I. Plans and Projects— Little 643
X.B.2. Measures of Profitability— UNIDO 646
X.B.3. Social Cost-Benefit Analysis — Little and Mirrlees 652
X.C. SHADOW PRICING 658
X.C.I. Distortions in Factor Markets — Lai 658
X.C. 2. Derivation of Shadow Prices — Helmers 671
X.C.3. Guide to Little-Mirrlees— Baldwin 675
X.C.4. Foreign Exchange Rate— UNIDO 681
X.C.5. Opportunity Cost of Capital— UNIDO 685
X.C.6. Social Prices — Squire and van der Tak 688
X.D. OBJECTIVES AND RELATIVE WEIGHTS 692
X.D. 1 . Objectives of Planning— Sen 692
X.D.2. Choice of Technology as a Problem of "Second Best" — Sen 694
X.D. 3. Social Importance of Employment Creation — UNIDO 702

XI. DEVELOPMENT PLANNING AND POLICYMAKING 707

XI.A. PLANNING AND THE MARKET 709
XI.A.l. The Flaw in the Mechanism of Market Forces — Rosenstein-Rodan 709
XI.A.2. The Market Mechanism as an Instrument of Development — Johnson 711
XI.A.3. The Fragmented Economy — McKinnon 716
XI.A.4. Planning and Cost-Benefit Analysis — Kornai 718
XI.B. POLICY MODELS 723
XI.B.l. Planning Models— Chenery 723
XI.B.2. A Model of Development Alternatives — Chenery and Bruno 726
XI.B.3. Policy Instruments and Development Alternatives — Chenery 732
XI.C. PLANNING EXPERIENCE 744
XI.C.l. The Crisis in Planning— Killick 744
XI.C.2. Planned Development in India — Bhagwati 747
XI.C. 3. The Future of Development Planning — Note 753
XI.D. INTERNATIONAL POLICYMAKING AND THE NIEO— NOTE 758

INDEX 763
ACRONYMS

ASEAN Association of South East Asian Nations

C-20 Committee of Twenty
DAC Development Advisory Committee of OECD
ECA Economic Commission for Africa
ECLA Economic Commission for Latin America
ECOSOC Economic and Social Council
FAO Food and Agriculture Organization
G-24 Group of Twenty-four
GAB General Arrangements to Borrow
GATT General Agreement on Tariffs and Trade
GNP Gross national product
IBRD International Bank for Reconstruction and Development
IDA International Development Association
IDB Inter-American Development Bank
IFC International Finance Corporation
ILO International Labour Organisation
IMF International Monetary Fund
ITO International Trade Organization
LDCs Less developed countries
MDCs More developed countries
MTN Multilateral trade negotiations
NICs Newly industrializing countries
NIEO New International Economic Order
ODA Official Development Assistance
OECD Organization for Economic Co-operation and Development
OPEC Organization of Petroleum Exporting Countries
SDRs Special drawing rights
UN United Nations
UNCTAD United Nations Conference on Trade and Development
UNDP United Nations Development Program
UNESCO United Nations Educational, Scientific and Cultural Organization
WFC World Food Council
WFP World Food Program
WHO World Health Organization
Leading Issues in
Economic Development
CHAPTER I

The Development
Record

We are concerned in this book with the future standard of living for 70 percent of the world's
population — the poverty-ridden peoples in the less developed countries (LDCs) of Asia, Af-
rica, and Latin America. Two centuries after the Industrial Revolution, most of the world is
still poor. The need to accelerate the development of the poor countries remains, in the words
of the recent President of the World Bank, Robert McNamara, "the central historical event
of our times."
We examine the major analytical and policy issues raised by the challenges to accelerate
economic growth rates, eradicate absolute poverty, reduce inequality, and create more pro-
ductive employment opportunities. These challenges now require the international community
to reassert the development priority and reshape development policies.
The World Bank emphasizes that the failure to achieve a minimum level of income above
the "poverty line" has kept some 40 percent of the people in the LDCs in the condition of
"absolute poverty" — a condition of life so degraded by disease, illiteracy, malnutrition, and
squalor that it denies its victims the basic human necessities. The persistence of absolute pov-
erty, despite respectable achievements in rates of growth in gross national product (GNP), is
now of more concern than that of relative poverty, or a "widening gap" between the rich and
poor countries. With the increase in the world's population, the number of people in absolute
poverty below minimum standards of nutrition, health, shelter, and education has grown and
is now of the order of 800 million.
A major theme of this book is the emphasis on discovering and promoting those positive
forces that raise real income per head, improve employment opportunities, and achieve a
wider distribution of the gains from development.
These forces of development are not, however, readily identifiable. After disappointments
in the record of development, economists can no longer rely on simple causal relationships
that emphasize development planning, capital accumulation, and foreign aid. The forces of
development are too complex, subtle, and insufficiently known to yield to any simple formula.
 2 THE DEVELOPMENT RECORD

One way of recognizing that a wide range of development forces must be identified is to
realize at the outset that the development process encompasses more than the economic theory
of production. Empirical studies of the sources of growth in output in a number of countries
have demonstrated that much of the increase in aggregate production over a long period can-
not be explained by an increase in only the standard physical inputs of the factors of produc-
tion. A large part of the increase in total output remains to be attributed to some "unexplained
residual factor" in the economy's aggregate production function. In historical studies of sev-
eral presently advanced countries, the "residual factor" — the unisolated source of growth —
is left to account for 50 percent or more of the increase in total output. This residual has been
called a "measure of our ignorance" of the determinants that create a rise in productivity —
the complex of little understood forces that cause output per unit of utilized resources to rise.1
A fundamental question is: What are the other sources of growth that can be identified in
the catchall of the "residual," and introduced as inputs that are amenable to policy promo-
tion? Another theme of this book, therefore, is the need to proceed beyond a narrow economic
theory of production to a broader interpretation that requires attention to "efficiency" in the
utilization of inputs as well as a wider understanding of an economy's "learning process."
An additional theme is that national development must now occur within the international
context of coexisting rich and poor countries. Capital, technology, knowledge, values, and
institutions are more readily transferable from rich to poor countries than ever before. There
may, however, be drawbacks as well as advantages in being a latecomer to development, and
the late-developing country is increasingly aware of the need for "appropriate" transfers.
There is now a more imaginative search for means by which to raise the social benefit-cost
ratio of these transfers to the developing countries.
At the same time, the greater interdependence between the more developed countries
(MDCs) and the less developed countries (LDCs) gives the LDCs new negotiating strategies
for designing trade, investment, and international monetary arrangements. The LDCs have
assumed an enhanced role in shaping the emergence of a new international economic order
to replace the old Bretton Woods system. The contentions of the United Nations Conference
on Trade and Development (UNCTAD) and the criticisms of the General Agreement on
Tariffs and Trade (GATT) and the International Monetary Fund (IMF) must receive serious
attention because the future of national development will be closely related to reforms in
international trade and monetary arrangements — a new Bretton Woods II.
The time has also come for a reassessment of development planning and policymaking.
Many have become disenchanted with centralized, detailed planning as it has been practiced
in some LDCs. But even for the "lighter" type of framework planning, indicative planning,
sectoral planning, the simpler monetary and fiscal policies, or only project evaluation — even
for these less ambitious policies, the quality of policymaking must be raised. More efficient
public-sector management is of the essence. The analysis of what policies are needed, the
choice of policy instruments, and the methods of implementation must all be improved. Inti-
mately related to improvement in policymaking is the role to be assigned to the market price
system and decentralized decision making. The removal of price distortions and the use of the
market as an instrument of policy are aspects of this theme of improving the quality of de-
velopment policymaking.
To begin the pursuit of these several themes, this first chapter attempts to reach a clearer
understanding of the meaning of economic development. The objectives of development are
examined and certain misconceptions of development clarified in section LA. The selections
in section LB. then consider different measurements of economic development and various

'Moses Abramovitz, "Resource and Output Trends in the United States Since 1970," American Economic
Review, May 1956, pp. 11-14.
 3 THE DEVELOPMENT RECORD

dimensions of poverty. Some of these selections reflect the shift from viewing poverty as in-
adequacy ofmeans — for example, inadequate income — to viewing poverty as unsatisfactory
results, such as low expectation of life and low levels of literacy. The recent emphasis on
"basic human needs" is considered in selection LB. 3. Other selections explore the question of
who benefits from development. Particular attention is given to the inequality of the "in-
verted-U hypothesis" (I.B.4) and the possible tradeoff between growth and equity (I.B.5).
Turning to an examination of country performance over the past three or four decades,
section LC. outlines some successful development strategies; summarizes some of the empir-
ical evidence on growth, absolute poverty, and distribution (I.C.I); and evaluates the devel-
opment performance of a number of countries — South Korea (I.C.2), Brazil (I.C.3), Sub-
Saharan Africa (I.C.4), and China (I.C.5).
Our emphasis is on the disappointments in the development record: the low growth perfor-
mance of the poorest countries, the unanticipated high rate of population growth, the weak-
ness of the agricultural sector, the huge number in absolute poverty, and the pervasive un-
derutilization of labor.
Considering these disappointments in the development record, the final Note (section I.D)
in the chapter outlines the strategic policy issues that now confront the developing countries —
issues to be analyzed in subsequent chapters. Unless these issues are resolved, the develop-
ment effort will remain stalled, and the policy changes necessary to regain the development
momentum and allow the LDCs to reach their potential will not be forthcoming. Such a loss
will be all the more tragic when it is realized that forecasts for the year 2000 indicate that
there will be four times as many people in the developing world than in the industrialized
world. And with each year that the development effort stagnates, the severity of future prob-
lems only intensifies.
 I.A. OBJECTIVES OF
DEVELOPMENT-NOTE

Dissatisfaction with the results of the devel- and restrictions on the inflow of private for-
opment effort over the past three decades has eign capital. The emphasis on national inde-
led to a refocusing of strategy to meet the fu- pendence through "inward-looking" policies
ture problems of development. The growth of and the advocacy of policies to avoid "foreign
GNP is no longer regarded as the main ob- domination" become part of an ideology that
jective orindex of development — but no sin- might be called the "economics of discon-
gle criterion can be readily substituted. To tent." But the "economics of discontent"
the dismay of the purist, but not to the sur- should not be confused with the economics
prise ofthe development practitioner, it is dif- for development. National independence and
ficult to give one precise meaning to "eco- the process of national consolidation may be
nomic development." Perhaps it is easier to called, as it is in India, "emotional integra-
say what "economic development" is not. tion";1 but this is not to say that this noneco-
At the outset it should be recognized that nomic objective also contributes to economic
economic development is not equivalent to development. "Inward-looking" policies are
the total development of a society: it is only most likely to run counter to economic devel-
a part — or one dimension — of general devel- opment, as is argued more fully in Chapter
VIII.
opment. We usually focus on the nation-state
as the unit of development, but "national de- Another aspect of the economics of discon-
velopment" isa term that encompasses, at a tent is the poor country's protest against
minimum, social and political development, being a primary-producing country. Indus-
as well as economic development, in the trialization tends to be viewed as a superior
building of national identity. Depending on way of life; rich countries are believed to be
the orientation of one's discipline, it is also rich because they are industrialized; and poor
possible to consider other types of develop- countries are believed to be poor because
ment— for example, legal or administrative. they are primary-producing. Whether an in-
The interrelationships among these various dustrial society enjoys a superior way of life
types of development are extremely impor- is, however, a noneconomic question. The rel-
tant. A major question implicit in our entire evant economic question for a poor country is
subject is how sociocultural and political de- whether agricultural development or indus-
velopment contribute to economic develop- trial development is now the appropriate
ment and are, in turn, determined by it. It
strategy for accelerating the country's eco-
will be apparent that much more interdisci- nomic development.
plinary study is needed to determine how eco- Economic development is not to be equated
nomic and noneconomic forces interact. simply with industrialization for several rea-
It is also necessary to caution against sons. First, the concentration of a large per-
equating economic development with either centage ofproduction in the primary sector is
"economic independence" or "industrializa- in itself not a cause of poverty: the cause is
tion." As a result of their colonial history and the low productivity in agriculture. A poor
newly acquired political independence, many country's high ratio of agricultural popula-
poor countries have expressed discontent tion to total population is also more appro-
with their "dependence" on export markets priately viewed as a consequence, rather than
and foreign capital. Such "dependence" is a cause, of poverty. Whenever the agricul-
often interpreted as synonymous with "for-
eign domination" or "exploitation" — to be 'Gunnar Myrdal, Asian Drama, New York, 1968,
avoided now by import-substitution policies
pp. 53, 722-3.
 6 THE DEVELOPMENT RECORD

tural population is poor, the nonagricultural improved techniques of production — in our

population serving the agricultural popula- growing control over nature. It is also likely
tion tends to be relatively small in size and to appear in the development of institutions
also at a low standard of living. When the and a change in attitudes and values.
rural sector is prosperous, the nonrural sector If we turn from what economic develop-
tends to be large and also prosperous.2 ment is not, and attempt to consider its
Second, progress in industrialization is meaning more directly, we immediately en-
highly dependent on agricultural develop- counter ambiguities because the ideal of eco-
ment. Without the necessary support of im- nomic development tends to be associated
provements inprimary production, the poli- with different policy goals. The phenomena
cies of industrialization will be severely that one chooses to denote as "economic de-
handicapped. In Chapter VII we examine the velopment" are very much a matter of what
essential ways in which the rate of industrial- one values as the economy's policy goals. And
ization depends on surmounting the limita- any definition of development inevitably be-
tions of the agricultural bottleneck. comes a "persuasive definition," implying
Third, economic development is much that development — as so defined — is a desir-
more than the simple acquisition of indus- able objective.
tries. Itmay be defined as nothing less than Although requiring careful interpretation,
the "upward movement of the entire social perhaps the definition that would now gain
system";3 or it may be interpreted as the at- widest approval is one that defines economic
tainment ofa number of "ideals of moderni- development as the process whereby the real
zation," such as a rise in productivity, social per capita income of a country increases over
and economic equalization, modern knowl- a long period of time — subject to the stipu-
edge, improved institutions and attitudes, lations that the number below an "absolute
and a rationally coordinated system of policy poverty line"ofdoes not does
increase, and thatmore
the
measures that can remove the host of unde- distribution income not become
sirable conditions in the social system that unequal. We emphasize process because this
have perpetuated a state of under- implies the operation of certain forces in an
development.4 interconnected and causal fashion. In the fol-
These views also imply that economic de- lowing chapters we examine the process of
velopment involves something more than eco- economic development as a form of progres-
nomic growth. Development is taken to mean sive action — a working-out of certain princi-
growth plus change; there are essential qual- pal forces that reveal the inner structure or
itative dimensions in the development process "logic" of an economy's development. To in-
that may be absent in the growth or expan- terpret development in terms of a process in-
sion of an economy through a simple widen- volving causal relationships should prove
ing process. This qualitative difference is es- more meaningful than merely to identify de-
pecially likely to appear in the improved velopment with a set of conditions or catalog
performance of the factors of production and of characteristics. If our interest in the de-
velopment ofa poor country arises from our
desire to remove mass poverty, then we
2For an elaboration of primary production as only
should also emphasize as the primary goal a
an associative — rather than causative — characteristic rise in per capita real income rather than
of poverty, see S. Kuznets, Economic Change, New
York, 1953, pp. 222-5; J. Viner, International Trade simply an increase in the economy's real na-
and Economic Development, Oxford, 1953, p. 50; G.
tional income, uncorrected for population
M. Meier and R. E. Baldwin, Economic Develop- change. For, if the criterion were only an in-
ment, New York, 1957, pp. 315-16. crease in real national income, then it would
3Myrdal, Asian Drama, p. 1 869. be possible for aggregate output to rise with-
4C. E. Black, The Dynamics of Modernization, out a per capita improvement in living stan-
New York, 1966, pp. 55-60. dards. Population growth may surpass the
 7 OBJECTIVES OF DEVELOPMENT— NOTE

growth of national output or run parallel with might still rise even though there had been no
it; the result would be falling, or at best con- change in the indigenous economy. Judgment
stant, per capita income, and we would not on the distribution of income is thus an inte-
consider this as economic development. gral part of the development problem.
We also stress a long period of time be- A few of the many possible other subgoals
cause what is significant from the standpoint may be the specification of a minimum level
of development is a sustained increase in real of consumption, a certain composition to the
income — not simply a short-period rise such consumption stream, a maximum level of un-
as occurs during the upswing of the business employment that will be tolerated, avoidance
cycle. The underlying upward trend over dec- of marked disparities in the prosperity and
ades— at least two or three decades — is the growth of different regions within a country,
stronger indication of development. From diversification of the economy, and the at-
this standpoint, a five-year development plan tainment ofthe "ideals of modernization."
is only the start of the development process, Owing to this variety of policy objectives,
and it remains to be seen whether there is the the emphasis on various dimensions of eco-
power to sustain the process so that per cap- nomic development will vary at different
ita real income continues to rise over the times and in different countries. We should,
longer period. There is a vital distinction be- therefore, beware of interpreting economic
tween initiating development and the more development as meaning economic progress
difficult task of sustaining development over or an increase in economic welfare. An in-
the longer run. crease inreal per capita income is not by it-
Although the increase in real income per self a sufficient condition. Per capita real in-
head can be adopted as the primary goal, it come is only a partial index of economic
has also become common to interpret eco- welfare because a judgment regarding eco-
nomic development in terms of a number of nomic welfare will also involve a value judg-
subgoals or particular categories of the over- ment on the desirability of a particular dis-
all primary objective. Thus, a certain distri- tribution ofincome. All observers would not,
bution of income is another policy objective. therefore, definitely say that economic wel-
A diminution in economic inequality is a gen- fare has increased even if per capita income
erally stated objective of development plans. has risen, unless the resultant distribution of
Most students of development would also un- income is also considered desirable.
doubtedly qualify the primary goal by re- Economic welfare poses not only the ques-
quiring that the absolute number of people tion of distributive justice. There are also the
below a minimum level of real income should prior questions of what is the composition of
also diminish at the same time that real per the total ouput that is giving rise to an in-
capita income rises. Otherwise it is conceiv- crease inper capita real income, and how this
able that if there is population growth, the output is being valued. Whether a larger
numbers of those living below a poverty line total output corresponds to individual pref-
may actually have grown while there has erences— let alone the more difficult test of
been a rise in the average income of the pop- collective choice — depends as much on what
ulation as a whole. When a dual economy ex- is produced and its quality as it does on the
ists— with a division between the modern quantity produced. The valuation of the out-
money economy and the traditional indige- put may also be biased insofar as it is valued
nous economy — it is also possible for all of by market prices. These prices become the
the increase in total income to occur in the equivalent of weights, but they have been af-
modern economy,5 and income per head fected bythe distribution of income: with a
different distribution of income, prices would
be different, and both the composition and
5The problem of dualistic development is discussed
value of the national output would also be dif-
fully in Chapter III. ferent. Market prices will also have limited
 8 THE DEVELOPMENT RECORD

value insofar as they do not reflect external such an approach because in many countries
diseconomies or social costs. it has become only too painfully apparent
If such considerations of the composition that despite growth in aggregate output there
of aggregate output, its valuation, and its dis- can still be a larger number of people below
tribution make it difficult to equate economic the poverty line, rising unemployment, and
development with economic welfare, it is all greater income inequality. The quality of de-
the more unreasonable to claim that eco- velopment iscompletely masked if the poli-
nomic development means an increase in so- cymaker does not pierce the aggregate meas-
cial welfare in general. Economic welfare is ure of GNP and consider its composition and
but a part of social welfare, and even if in the distribution.
course of a country's development all the con- Development economists no longer worship
ditions necessary to promote economic wel- at the altar of GNP, but concentrate more di-
fare have been satisfied, this need not also rectly on the quality of the development pro-
mean that social welfare has been promoted. cess. In the words of an official for the gov-
For the process of development has a pro- ernment of Pakistan, "the problem of
found impact on social institutions, habits, development must be defined as a selective
and beliefs, and it is likely to introduce a attack on the worst forms of poverty. Devel-
number of sources of tension and discord. opment goals must be defined in terms of pro-
Some aspects of human welfare might suffer gressive reduction and eventual elimination
if relations that were once personal become of malnutrition, disease, illiteracy, squalor,
impersonal, the structure and functions of unemployment, and inequalities. We were
the family change, the stability in one's way taught to take care of our GNP because it
of life is disrupted, and the support and as- would take care of poverty. Let us reverse
surance oftraditional values disappear. Ten- this and take care of poverty because it will
sions also arise when the inequalities in in- take care of the GNP. In other words, let us
come distribution, both among individuals worry about the content of GNP even more
and among regions in the developing country,
than its rate of increase."6
tend to increase; when development creates Years before the start of the United Na-
"open unemployment" as well as employ- tions' first Development Decade, Professor
ment; and when the pressures of excessive ur- Viner had offered a similar warning that
banization occur. In a fundamental sense, should have received more attention. He
discords arise from the contrasts between the stated that:
modern and the backward — from the super- While the supplementing of data as to economic
imposing ofmodern functions on traditional
institutions. aggregates by per capita averages provides addi-
tional and often essential information . . . even this
In sum, even though it is conventional to does not suffice for some purposes. Let us suppose,
begin with an increase in per capita real in- for instance, that a country which has embarked
come as the best available overall index of on a programme of economic development engages
economic development, we abstain from la- in periodic stock-taking of its progress, and finds
beling this an increase in economic welfare, not only that aggregate wealth, aggregate income,
let alone social welfare, without additional total population, total production, are all increas-
considerations of various subgoals and ex- ing, but that per capita wealth, income, produc-
plicit recognition of the value judgments re- tion, are also all increasing. All of these are fa-
vourable indices, but even in combination do they
garding at least the composition, valuation,
and distribution of the expanded output. The suffice to show that there has been "economic
student of development must adopt such a
cautious approach as a result of the strictures
of welfare economics and the need for clarity 6Mahbub ul Haq, "Employment and Income Dis-
on value premises in social research. tribution inthe 1970's: A New Perspective," Paki-
stan Economic and Social Review, June-December
Even more, the policymaker must adopt 1971, p. 6.
 9 OBJECTIVES OF DEVELOPMENT— NOTE

progress," an increase in economic "welfare," ulation. Although the average per capita income of
rather than retrogression? the Third World has increased by 50 percent since
Suppose that someone should argue that the one 1960, this growth has been very unequally distrib-
great economic evil is the prevalence of a great uted among countries, regions within countries,
mass of crushing poverty, and that it is a paradox and socio-economic groups. Paradoxically, while
to claim that a country is achieving economic growth policies have succeeded beyond the expec-
progress as long as the absolute extent of such pov- tations ofthe first Development Decade, the very
erty prevailing in that country has not lessened or idea of aggregate growth as a social objective has
has even increased? Such a country, nevertheless,
increasingly been called into question.8
might be able to meet all the tests of economic de-
velopment which I have just enumerated. If its
population has undergone substantial increase, the This questioning has become common
numbers of those living at the margin of subsis- among development policymakers. If there
tence or below, illiterate, diseased, undernour- have been misconceptions during the past
ished, may have grown steadily consistently with a
rise in the average income of the population as a decades of development, the future of devel-
whole. . . . opment calls for a basic reconsideration of
Were I to insist, however, that the reduction of the meaning of development and a funda-
mass poverty be made a crucial test of the reali- mental redirection of development policy. In-
zation ofeconomic development, I would be sepa- stead of settling for any aggregate, or even
rating myself from the whole body of literature in
per capita, index of "development," many
this field. In all the literature on economic devel- now advocate that direct attention be given
opment have
I seen, I have not found a single in- to the achievement of "first things first" — to
stance where statistical data in terms of aggre- the achievement of better nourishment, bet-
gates and of averages have not been treated as
ter health, better education, better living con-
providing adequate tests of the degree of achieve- ditions, and better conditions of employment
ment of economic development.7 for the low-end poverty groups in the poor
Long after Viner's warning, Hollis Che- countries of the world. Instead of seeking
nery introduced the World Bank's influential "development" as an end, we might better
study on Redistribution with Growth with view it as a means — as an instrumental pro-
this statement:
cess for overcoming persistent poverty, ab-
It is now clear that more than a decade of rapid sorbing the surplus labor, and diminishing in-
growth in underdeveloped countries has been of equality. To help illuminate this process is
little or no benefit to perhaps a third of their pop- the intention of this book.

7Jacob Viner, International Trade and Economic 8Hollis Chenery et al., Redistribution with Growth,
Development, Oxford, 1953, pp. 99-100. London, 1974, p. xiii.

Comment
The term "economic development" is subject to various interpretations. The relationship
between development and welfare was considered early by W. A. Lewis in The Theory of
Economic Growth (1955), in an appendix. The extension of the meaning of development be-
yond GNP or GNP per capita is advocated by Dudley Seers; see his paper "What Are We
Trying to Measure," Journal of Development Studies (April 1972); and "The Meaning of
Development" in David Lehmann (ed.), Development Theory (1979).
Seers states:

The questions to ask about a country's development are: What has been happening to poverty? What
has been happening to unemployment? What has been happening to inequality? If all three of these
central problems have been growing worse, it would be strange to call the result "development," even if
per capita income had soared. This applies, of course, to the future too. A "plan" which conveys no
targets for reducing poverty, unemployment and inequality can hardly be considered a "development
plan." [Lehmann, ed., Development Theory, p. 12]
 10 THE DEVELOPMENT RECORD

Indicators of infant mortality, life expectancy, and basic literacy have been used as the
components of a composite "Physical Quality of Life Index" (PQLI) that measures perfor-
mance in meeting the most basic needs of people. This composite indicator of poverty is de-
signed to measure results, not inputs such as income. The rationale for the PQLI is explained
in detail by Morris David Morris, Measuring the Condition of the World's Poor (1979).
For the less developed countries, "economic development" involves a process of emerging
from poverty. Concepts of poverty, however, are subject to various interpretations. Normally,
both absolute and relative deprivation are essential ingredients of a common understanding
of poverty. Absolute deprivation relates to the denial of basic needs (see LB. 3). Relative dep-
rivation relates to interpersonal gaps in the income distribution within the poor country and
also to the international gaps in standards of living. Nonincome factors captured by the PQLI
are important, but so are income and consumption statistics and distribution — sensitive meth-
ods of aggregation by which to obtain an overall poverty index. For example, see Amartya
Sen, "Levels of Poverty Policy and Change," World Bank Staff Working Paper No. 401 (July
1980).
Intercountry comparisons of levels of income are often misleading when such comparisons
are made by converting the incomes of the various countries into a common currency, say the
U.S. dollar, through the use of official exchange rates. These nominal exchange rates do not
reflect the relative purchasing powers of different currencies, and thus errors are introduced
into the comparisons. Studies by Irving Kravis and his associates have attempted to adjust
international comparisons for the real purchasing power parities of national currencies. See
I. B. Kravis, A. Heston, and R. Summers, World Product and Income: International Com-
parisons ofReal Gross Domestic Product (3 volumes: 1975, 1978, 1982); "Real GDP per
Capita for More Than One Hundred Countries," Economic Journal (June 1978); and "In-
ternational Comparison of Real Product and its Composition: 1950-1977," Review of Income
and Wealth (March 1980).

EXHIBIT 1.1. The Level of GNP and Population of Developing Countries in 1980 and the Growth of Real
GNP and GNP per Capita, 1970-80°

1970-80
Real Growth Rate
1980 (.%)
GNP/ GNP/ 1980
GNP
-0.3
Countries
us$
CAP Population
CAP GNP
Million
(millions)
-0.1
LLDCs
Afghanistan 170 15.94 1.3
120 90.20 3.9
2.6
Bangladesh 11,490
Benin 300 3.53 0.6 5,000
1.30 -3.1
3.5
Bhutan 80 -5.0 2.1 US$*
110
1,120
780
Botswana 910 0.80
11.7
-2.1 14.3
Burundi 200 4.10 0.9 3.0 90
300 0.32 880
Cape Verde 730
300 2.29 3.8 5.9
Central African Rep.
Chad 120 4.46 0.5 2.7
Comoros 300 0.35 1.6 500
100
Ethiopia 140 31.47 0.8 160
Gambia 250 0.60 1.6 3.0 4,080
4.7
 11 OBJECTIVES OF DEVELOPMENT— NOTE

EXHIBIT 1. 1. (Continued)

1970-80
Real Growth Rate
1980 (%)
GNP/ GNP/ 1980
GNP
CAP Million
US$ Population
Countries (millions) CAP GNP
-0.5
-2.0
Guinea 290 5.43
0.79 0.1
0.3 3.1
1.9 140
Guinea-Bissau 160 4.0 1,480
Haiti 270 5.10 2.3
110 3.43 1,400
360US$*
Laos -0.2
Lesotho 390 1.34 10.0
-0.2
7.4 570
40
Malawi 230 5.95
3.2 6.1
Maldives 260 0.15 4.5 1,490
0.2
1.8 3.2
Mali 190 6.94
140 14.29 2.1 1,420
Nepal -1.6
Niger 330 5.32 2,030
200 -4.4
0.6 2.7
Rwanda 5.10 1,810
380 3.91 -1.4
0.8 3.4 1,050
Somalia
470 0.4 3.7
3.1
Sudan 18.37 1,520
260 7,640
Tanzania 18.14
0.2 3.7 4,860
280 13.20 10,230
Uganda 190
Upper Volta 5.73
850 0.8
-4.3
4.1 130
Western Samoa 0.16 1,140
Yemen 460 5.81 -3.1
3.5
6.5
430 1.90 5.4
8.9 4,010
Yemen, Dem. 6.4
Totals and averages 210 276.33 2.9 1,000
67,350
0.2
-4. 4
LICs-LLDCs
Angola 470 7.08
570 -1.5
1.1 4.1 2,880
Bolivia 5.57 4.6
Burma 180 33.31 2.4 5,790
480 0.35 -2.3 170
5,580
Djihouti 4.8 7.0
3.5 25,300
Egypt 580 39.77
El Salvador 590 4.54 1.8
2.8
460 -0.7 170
3,330
Equatorial Guinea 0.36
420 11.68 5.1
Ghana 15,360
560 0.6 0.7
4.0
Honduras 3.69
India 240 673.21 1.3 173,310
2,420
-1.9 7.0
3.4
Indonesia 420 146.24 4.5 64,420
730
110 -1.2
0.3
Kampuchea 6.40 -1.1
Kenya 420 15.87
Liberia 1.86 -2.1
3.6 6,850
520 0.1 5.7
3.4
0.6 1,000
Madagascar 350 8.71
1.63 1.6 10
3,210
510
Mauritania 320
230 0.05 2.3
Mayotte 5.2
Mozambique 270 10.47 1.6 4.8
Pakistan 300 82.17 25,050
2,810
St. Helena 440 - - -
0.01
 12 THE DEVELOPMENT RECORD

EXHIBIT 1. 1. (Continued)

80
Real Growth Rate
1980 1970-:
(%)
GNP/ GNP/
-0.6 1980
GNP
CAP -0.9 Million
US$ Population
Countries (millions) CAP GNP
-1.4
St. Vincent 520 0.11 1.4 60
Sao Tome and Principe 490 0.11 0.3 60
450 5.66 2.2
1.6
Senegal 970
270 1.2 2,600
Sierra Leone 3.47 130
460 0.23
Solomon Islands (Br.) 1.2 US$*
4.7
Sri Lanka 270 14.82
-3.0
1.5 4.7 4,100
Togo 410 2.48 2.8
Tokelau Islands 550 0.00 -2.8
-0.4 - -1,010
-1.1 50
Tonga 480 0.10
0.9 2.4
Turks and Caicos
Islands 470 0.01 -2.2 - -
-2.6 -
Tuvalu 550 0.01 -
Vanuatu (N. Hebrides) 520 0.12 2.5 60
Vietnam, Soc. Rep. 190 54.05 2.3 10,110
220 0.5
Zaire 28.29
560 0.5
Zambia 5,340
5.77 0.4
Totals and averages 290
1,168.20 -1.6
1.1 3,560
366,920
-2.0 4.6
3.9
China 290 976.74 2.8 283,250

MICs
630 0.01 1.5 - -
Anguilla
Antigua 0.08 100
1,270 0.2
1.1 800
Bahamas 3,300 0.24
Bahrain 0.42 1.1 8.3
5,560 0.25 2,260
Barbados 3.0 3.5 800
3,040 -1.9
Belize 0.15 160
1,080
Bermuda 11,230 0.06 2.0
3.7 3.3 660
5.7
Brunei 11,890 0.22 11.4
5.7
Cameroon 670 8.44 2,620
3.1 5.4
3.5 50
Cayman Islands 2,890 0.02 5,840
Chile 11.10 2.1 27,170
2,160 -3.6
1.2 3.9
6.0
Colombia 26.67 32,520
730
1,180 -1.1
2.6
Congo 1.54 -3.6
0.1
2.6 2.5 20
Cook Islands 1,090 0.02 1,160
Costa Rica 730 2.21 2.6 5.1
1,730 4,850
Cuba 9.86 0.2
1.9
0.62 2.2 7,180
Cyprus 3,580
Dominica 620 0.08 50
2,100
1.140 -0.5
Dominican Republic 5.44 3.7
Falkland Islands 0.00 5.8 10
6,480
3,800 6.7
5.8
Fiji 0.63
1,850 3.5 5.5
Gibraltar 0.03 4.4 150
1,190
5,080 3.3 1.1 80
Grenada 690 0.11
 13 OBJECTIVES OF DEVELOPMENT— NOTE

EXHIBIT 1. 1. (Continued)

Real 1970-80
Growth Rate
1980
GNP/ GNP/ (%) 1980

GNP
CAP Million
US$ Population
Countries (millions) CAP GNP
-0.2
4.6
Guadeloupe 0.33
3,870 7.01 2.9 -3.0
4.7
Guatemala 1,270
1,110 -4.5 5.8
Guiana, French 0.06 1.0 180
7,810
2,880 3.2 560
Guyana 690 0.79
0.8
Israel 3.88 4.5 US$*
4,500 6.6
Ivory Coast 8.64 1.7
0.9 2,030
Jamaica 1,150 10,170
1,030 2.19
4.3 40
2,400
Jordan 8.0
1,010
730 3.24
Kiribati 0.06 1.8
0.1 3,380
Lebanon 2.66 6.8
1,890 6.0
12.9
Macao 0.31 15.5
2,030 5,400
640
Malaysia 5.5 7.9
13.44 22,700
1,670
Malta 0.34 10.3 10.9
3,470 0.33 4.3 4.1
Martinique 1,240
4,640 0.96 4.1 -0.4
Mauritius
1,060 -3.5 5.6 1,510
970
Montserrat 0.01 2.8 20
1,910
860 20.18 3.6
-2.8
2.6 5.9 40
Morocco 17,690
Nauru 6.6
5,300 0.01 6.6
Netherlands Antilles 0.26 2.9
4,290 1.4
New Caledonia 0.15 -2.5 1,160
7,000
Nicaragua 720 2.67 1,050
Nigeria 84.73 4.7 7.3 1,050
1,010 7.2 0.7 91,990
Niue Island 0.00 - -
1,100
Oman 0.89 3.5
Pacific Islands (U.S.)
4,380
920 0.13
-6.8 120
4,550
Panama 1.83 2.1 4.4
1,730
780 2.6
Papua New Guinea 3.01 3,190
0.3
Paraguay 3.06 8.8 2,430
1,340 5.7 3.0
Peru 930 17.63 4,320
720 47.88 0.2 18,420
Philippines 4.1
3.6
1.5 35,920
Polynesia, French 0.15 6.4 970
6,530 2.0
Reunion 0.53
3,830 0.2
2,010
St. Kitts-Nevis 920 0.05
850 2.4
2.0 3.2 110
St. Lucia 0.12 50
St. Pierre and Miquelon
- -4.2 10
1,630 0.01 120
Seychelles 6.9
1,770 0.07 4.4
Surinam 0.35 4.9
2,840 5.5 400
Swaziland 680 0.56 1,040
8.98 4.4 6.1
Syria 3.4 8.2 12,390
Thailand 1,340 46.45 4.1
670 6.6 32,200
Trinidad and Tobago 1.17 5.2
4,370 6.35 38
5.9
Tunisia
1,310 8.2 5,440
8,440
 4 THE DEVELOPMENT RECORD

EXHIBIT 1. 1. (Continued)

Real 1970-80
Growth Rate
1980
GNP/ GNP/ (%) 1980
GNP
CAP -2.6 Million
US$ Population
Countries (millions) CAP
-2.0 GNP
-0.6
2.5
Turkey 45.36 5.1
1,460 59,110
Uruguay 2.92 2.8 3.2
2,820 -
Virgin Islands (Br.) 0.01 30
9,740
2,100 -
Wallis and Futuna 990 0.01 10
Zimbabwe 630 7.40 2.8 US$*
Totals and averages 415.35 3.0 5.6 476,250
4,710
1,150
NICs 2.5
Argentina 27.74 142,030
2,390 118.67 0.9 8.6
Brazil 230,450
2,050 9.33 6.3
Greece 4.1 4.7
4,520 42,210
Hong Kong 5.11 6.3 9.1
4,210 21,760
Korea, Rep. 38.45 6.1 6.1
1,520 2.3 57,670
Mexico 67.46 162,650
2,130 2.3 3.6
Portugal 9.84 7.4 5.4 23,750
2,350 2.39
Singapore 4,480 10,380
37.38 2.6 8.9 207,250
Spain 5,340 7.4 3.7
Taiwan 17.65 9.5
2,160 40,210
Yugoslavia 22.33 6.2 71,930
2,620 5.2
3.8 5.8
Totals and averages 356.33 1,010,280
2,540 -0.1
OPEC 2.5
Algeria 18.92 4.5 5.9
1,920 39,730
Ecuador 8.36 -2.9 8.0 10,890
1,220
Gabon 0.66 2.3 3.6
4,440
Iran 38.13 2.9 83,480
Iraq 1,940 3,410
9.4 13.2
3,020 13.07 39,040
Kuwait 22,840 1.35 2.0 8.3
1.1 30,650
Libya 2.98
8,640 0.23 2.7 30,540
Qatar 26,080 10.4
Saudi Arabia 8.96 13.6
11,260 6,650
117,580
United Arab Emirates 0.89 8.7
3.1
30,070 16.4
4.8 27,060
Venezuela 14.93 7.3
3,630 108.48 1.4
3.8 60,720
Totals and averages 3,760 449,750
Total LDC and averages 750 3.2 5.5
3,301.43 2,653,800
Source: OECD, Development Co-Operation 1982 Review, Paris, 1982, pp. 254-6.
Note: LICs: Low-income countries (per capita income in 1980 under $600).
LLDCs: Least-developed countries.
MICs: Middle-income countries (per capita income in 1980 exceeding $600).
NICs: Newly industrializing countries.
"World Bank and Secretariat estimates.
*At current prices and exchange rates, not comparable to data in other columns.
 LB. INDICATORS OF
DEVELOPMENT

I.B.1. Measuring Development*

So long as economists were willing to assume fined as a weighted sum of the growth of in-
the possibility of unrestricted transfers come in all groups:
among income groups, they found no conflict
G = W,gi + w2g2 + w3g3
in principle between the objectives of distri- + VV4g4 + W5g5 [1]
bution and growth. Once it is recognized that
large-scale transfers of income are politically where G is an index of the growth of total so-
unlikely in developing countries, however, it cial welfare and w, is the weight assigned to
becomes necessary to evaluate the results of
any development policy in terms of the ben- A summary measure of this type enables
efits itproduces for different socio-economic to seti.1development targets and monitor de-
usgroup
groups. While this idea has been accepted in velopment performance not simply in terms
the recent literature of project evaluation, it of growth of GNP but in terms of the distri-
has found little reflection in the methodology butional pattern of income growth. The
of macroeconomic planning and policy weights for each income class reflect the so-
formulation. cial premium on generating growth at each
An index of economic performance reflect- income level; they may be set according to
ing these objectives can be developed as fol- the degree of distributional emphasis desired.
lows. Assume a division of society into N As the weight on a particular quintile is
socio-economic groups, defined by their as- raised, our index of the increase in social wel-
sets, income levels, and economic functions. fare reflects to a greater extent the growth of
For purposes of policy analysis, it is neces- income in that group. Thus if we were only
sary to distinguish several poverty groups concerned with the poorest quintile we would
such as small farmers, landless laborers, set vv5 = 1 and all other wt = 0, so that
urban underemployed, and others according growth in welfare would be measured only by
to the similarity of their responses to policy gs. This approach is closely related to the
measures. In order to illustrate the problem more formal approach to welfare choices
of evaluation, we will classify merely by in- using explicit social welfare functions to
come size into ordinally ranked percentile measure improvements in welfare.
groups. In these terms the commonly used index of
Assuming a division by income level into performance — the growth of GNP — is a spe-
quintiles, the rate of growth of income of cial case in which the weights on the growth
each group, gh can be taken to measure the of income of each quintile are simply the in-
increase of its social welfare over the speci- come share of each quintile in total income.
fied period. The rate of increase in welfare of The shortcomings of such an index can be
the society as a whole can therefore be de- seen from the following income shares for the

*From Montek S. Ahluwalia and Hollis B. Che- 'This measure can be applied either to the income
nery, "The Economic Framework," in Redistribution or the consumption of each group. When applied to
with Growth, joint study by World Bank's Develop- income, the weight assigned should take account of
ment Research Center and Institute of Development the contribution made by each group to the financing
Studies at the University of Sussex, Oxford Univer- of investment and government expenditure. For long-
sity Press, 1974, pp. 38-43, 48-9. Reprinted by run simulations of policy an index based on consump-
permission. tion only is preferable.

15
 16 THE DEVELOPMENT RECORD

different quintiles, which are typical for contrast to the GNP measure, the equal
underdeveloped countries: weights index gives the same weight to a one-
percent increase in income in the lowest quin-
tile as it does to a one-percent increase in the
Quintiles 1 2 3 4 5 Total highest quintile. In this case a dollar of ad-
ditional income in the lowest quintile is val-
Share in 7% ued at eleven times a dollar of additional in-
Total Income 53% 22% 13% 100%
5% come in the highest quintile.
Many individuals (and some countries)
may wish to define social objectives almost
The combined share of the top 40 percent exclusively in terms of income growth of the
of the population amounts to about three- lowest groups, placing little value upon
quarters of the total GNP. Thus the rate of growth in the upper-income groups beyond
growth of GNP measures essentially the in- its contribution to national savings and in-
come growth of the upper 40 percent and is vestment. The welfare implications of such a
not much affected by what happens to the in- "poverty weighted" index are stronger than
come of the remaining 60 percent of the those underlying either the rate of growth of
population. GNP or the "equal weights" index, since it
An alternative welfare principle that has would be a welfare function based primarily
considerable appeal is to give equal social on the lower-income groups.
value to a one-percent increase in income for Weighted indexes of the sort discussed
any member of society. On this principle, the above provide a very different evaluation of
weights in equation [1] should be propor- performance in many countries than is ob-
tional to the number of people in each group tained from conventional measures. This can
and would therefore be equal for each quin- be illustrated using data for the fourteen de-
tile. Thus a one-percent increase in income in veloping countries for which we have obser-
the lowest quintile would have the same vations at two points in time. The numerical
weight in the overall performance measure as results presented here are subject to the lim-
a one-percent increase in income for any itations ofthe data and are essentially "illus-
other quintile, even though the absolute in- trative" to show the potential usefulness of
crement involved is much smaller for the low- weighted growth indexes in evaluating per-
est quintile than for the others. formance. They are not presented as defini-
When we use the growth of GNP as an tive assessments of country experience.
index of performance, we implicitly assume Table 1 shows the difference in estimates
that a dollar of additional income creates the of welfare increments based on three differ-
same additional social welfare regardless of ent weighting systems: shares of each quintile
the income level of the recipient.2 Given the in GNP (giving the rate of growth of GNP);
typical income shares of the different quin- equal weights for each quintile; and "poverty
tiles, itfollows that a one-percent growth in weights" of 0.6 for the lowest 40 percent, 0.3
income in the top quintile is given almost for the next 40 percent, and 0.1 for the top
eleven times the weight of a one-percent 20 percent.3 The following differences can be
growth in the lowest quintile (in the preced- observed among the countries when compar-
ing example) because it requires an absolute ing GNP growth with the other two indexes.
increment which is eleven times as great. In (i) In four countries (Panama, Brazil,

2This statement has to be qualified to allow for the

higher savings of the upper-income recipients and 3In terms of weights for unit increments, these
their greater contribution to future growth. In a more weights imply that a dollar of income accruing to the
complete analysis, the increase in social welfare can bottom 40 percent is worth 33 times a dollar accruing
be measured by the weighted growth of consumption to the top 20 percent, instead of 11 times as with
rather than income. equal weights.
 17 INDICATORS OF DEVELOPMENT

TABLE 1. Income Distribution and Growth

1. Income Growth II. Annual Increase in Welfare

Poverty
(B) (C) III. Initial
Middle Lowest (A)
GNP
Period Upper
20% 40% Weights Gini
Coefficient
Country 40% Weights Weights
Equal
7.8 9.3 9.0
Korea 1964-70 10.6 9.3 9.0 .34
.48
Panama 1960-69 8.8 9.2 3.2 8.2 5.6
Brazil 4.8 5.2 6.9 6.7
7.0 .56
1960-70 8.4 6.6 7.6 .56
Mexico 8.0 7.0 5.7 5.4
1963-69 .55
Taiwan 4.5 9.1 12.1 6.9
1953-61 6.8 10.4
7.0 .52
Venezuela 7.9 4.1 9.4 4.2
1962-70 7.3 3.7
7.0 6.4 4.7
Colombia 1964-70 5.6 6.2 6.8
7.1 .57
El Salvador 4.1 10.5 5.3 6.2 .53
1961-69 4.6 .50
4.9 5.0 5.5 6.7
Philippines 1961-71 6.4 5.4 .59
Peru 7.5 3.2 5.2 5.4
1961-71 4.7 5.4 7.2 .45
Sri Lanka 1963-70 3.1 6.2 8.3 5.0 4.6 .33
4.3 6.4
Yugoslavia 1963-68 4.9 5.0 4.8 4.7 4.0 .40
4.5
India 1954-64 5.1 3.9 4.1
3.9

Note: The rates of growth of income in each income group were calculated as follows: income shares were applied to GNP
(constant prices) to obtain the income of each group in each year. The growth rate is the annual compound growth rate
estimated from the two endpoint income estimates for each income group. GNP series are from the World Bank data files.
Equal weights imply a weight of 0.2, 0.4, and 0.4 for the three income groups while poverty weights are calculated by giving
weights of 0.1, 0.3, and 0.6, respectively.

Mexico, and Venezuela), performance is eral the extent to which a weighted index of
worse when measured by weighted indexes. growth diverges from the growth rate of
In these countries the data show that relative GNP and the direction of divergence are
income distribution worsened over the period measures of the extent and direction in which
considered; i.e., growth was disproportion- growth is distributionally biased.
ately concentrated in the upper-income It is important to note that the proposed
groups. The indexes giving greater weight to index is a measure of the increase in welfare
the growth of income in lower-income groups rather than of total welfare. Increasing
are therefore lower than the rate of growth of equality is indicated by a weighted growth
GNP. rate in excess of GNP growth and increasing
(ii) In four countries (Colombia, El Sal- inequality by the opposite difference. The
vador, Sri Lanka, and Taiwan), the weighted measure cannot be used to compare perfor-
indexes are higher than GNP growth. In mance among countries without allowing for
these countries the data show that distribu- their initial distribution of income. For ex-
tion improved over the period; i.e., the growth ample, Table 1 shows inequality increasing in
of incomes in lower-income groups was faster both Brazil and India, but in India the start-
than that in higher-income groups. ing point was a relatively equal distribution
(iii) In five countries (Korea, the Philip- in 1954 while in Brazil distribution was al-
pines, Yugoslavia, Peru, and India) the use of ready quite unequal in 1960.
weighted indexes does not alter the GNP We recognize that the use of such indexes
measurement of growth to any great extent. for evaluation of performance will be severely
In these cases the data show that distribution limited by the initial lack of accurate data.
remained largely unchanged and all income For the present they are perhaps more valu-
classes grew at about the same rate. In gen- able as analytical devices to be used in rede-
 18 THE DEVELOPMENT RECORD

fining the objectives of development strategy. the levels of employment and remuneration
They help to clarify both the statement of of the factors of production, usually grouped
distributional objectives and the limits of ac- into capital and labor. They differ mainly in
ceptable trade-offs between growth and in- their assumptions about market behavior and
come distribution. In particular, the use of the way in which wages and product prices
welfare indexes emphasizes the importance are determined. Neoclassical theory assumes
of increasing the rates of growth of income in competitive equilibrium in all markets and
the poverty groups instead of focusing on the thus derives factor returns from pure produc-
static picture of income inequality. tion relationships and demand patterns, given
factor supply conditions. At the other ex-
treme, the classical and Marxist wage theory
TOWARD A THEORY OF that forms the basis for most dual economy
DISTRIBUTION AND GROWTH models assumes relatively fixed real wages
with all surplus value appropriated by the
The preceding section suggests that the ob- owners of capital.
jective ofdistributive justice is more usefully The inadequacy of existing theories for our
conceived of as accelerating the development purposes arises less from the lack of consen-
of the poorer groups in society rather than in sus as to the determinants of the functional
terms of relative shares of income. As a way division of income than from the omission of
of implementing this approach, we can visu- other aspects of the problem. The available
alize the role of the state as using available evidence on the nature of poverty in under-
policy instruments (including the allocation developed countries shows that half of the
of investment in physical and human capital) poor are self-employed and do not enter the
so as to maximize a welfare function of the
wage economy. Most wage-earners are al-
type just described. State intervention of this ready in the middle-income groups, so that
sort requires both an analysis of the deter- policies affecting the split between wages and
minants of income in poverty groups and of profits mainly concern the upper end of the
the linkages between the incomes of different distribution.
groups. The fact of income linkages is crucial The principal element that is missing from
for any analysis of distributional problems existing theories is an explicit treatment of
since they impose important constraints on the distribution of the various forms of as-
policy. Thus tax-financed transfers from the sets.4 A more general statement would rec-
rich to the poor may raise the income of the ognize that the income of any household is
poor but, if they reduce savings and capital derived from a variety of assets: land, pri-
accumulation by the rich, they may in time vately owned capital, access to public capital
lead to lower income in the poorer groups. An goods, and human capital embodying varying
analysis of these interactions requires an in- degrees of skills.
tegration ofgrowth and distribution theory. We distinguish four basic approaches to
the problem of raising the welfare of the low-
income groups: (i) Maximizing GNP growth
Distribution of Income and Capital through raising savings and allocating re-
sources more efficiently, with benefits to all
Existing theories of income distribution groups in society, (ii) Redirecting investment
are of only limited value in establishing an to poverty groups in the form of education,
analytical framework for comprehensive gov- access to credit, public facilities, and so on.
ernmental action because they are somewhat
narrowly focused on the functional distribu-
tion of income between labor and capital. Recognition of the importance of asset distribution
is common to neoclassical theorists as well as to
Most theories conceive the central problem of Marxists, but it has not been incorporated in empiri-
income distribution as the determination of cal models.
 19 INDICATORS OF DEVELOPMENT

(iii) Redistributing income (or consumption) welfare index in the short run, but they have
to poverty groups through the fiscal system or too high a cost in terms of foregone invest-
through direct allocation of consumer goods, ment to be viable on a large scale over an ex-
(iv) A transfer of existing assets to poverty tended period. Nevertheless, some direct con-
groups, as in land reform. In most countries sumption supplements for specific target
some elements of each of these approaches groups (child nutrition, maternal health ser-
will be applicable, depending on the initial vices) are a necessary supplement to an in-
economic and social structure. vestment-oriented strategy, since they are the
The advantages and limitations of each only way to alleviate some types of absolute
strategy will vary with the circumstances of poverty.
each country, and an assessment of these (iv) Political resistance to policies of asset
considerations is necessarily a matter of de- redistribution makes this approach unlikely
tailed study. Nevertheless, it is useful to con- to succeed on any large scale in most coun-
sider some broad characteristics of each tries. However, in areas such as land owner-
strategy in a relatively pure form. The gen- ship and security of tenure, some degree of
eral conclusions from the analysis can be asset redistribution is an essential part of any
summarized as follows: program to make the rural poor more pro-
(i) Maximizing the growth of GNP in- ductive. Beyond this essential minimum, a
volves some measures that benefit all groups, vigorous policy of investment reallocation in
as well as others — such as favoring high-sav- a rapidly growing economy may well be a
ings groups through lower income taxes or more effective way of increasing the produc-
wage-restraint policies — in which there is a tive capacity of the poor than redistribution
conflict with distributional objectives. Be- from the existing stock of assets, which is
cause of the relatively weak income linkages likely to have a high cost in social and polit-
between the poverty groups and the rest of ical disruption.
the economy, their growth tends to lag until (v) In the longer term, population policy
the expansion of employment creates a short- can have an important influence on both the
age of unskilled labor and hence an upward distribution of incomes and the level of con-
pressure on wage rates. Although — as sug- sumption inthe poverty groups. Our simula-
gested byTable 1 — the poor may be better tions show a tendency of income distribution
off even in this case than with slower GNP to worsen with population growth above 2.5
growth, the welfare effects of a maximal percent per year, while with more optimistic
growth strategy can almost always be im- assumptions it tends to improve. There is
proved byadding transfers. considerable demographic evidence that in-
(ii) As compared to maximal growth of vestments inthe health, education, and eco-
GNP, increased investment in the physical nomic growth of the poverty groups may also
contribute to a reduction of fertility and
and human "assets of the poverty groups is
likely to require some sacrifice of output in hence indirectly to better income dis-
the short run because returns on ipvestment tribution.
in human capital take longer to develop. Particular emphasis should be given to di-
Even so, the welfare index will be higher be- recting public investment to raise the produc-
cause these investments lead to income tive capacity and incomes of the poor. There
growth in target groups which have higher is a strong analogy between this strategy and
welfare weights. While this strategy has a an international strategy of assisting invest-
short-run cost to the upper-income groups, in ment in poor countries. In both cases trans-
the longer run they may even benefit from fers of resources that increase productive ca-
the "trickle up" effects'of greater productiv- pacity and lead to greater self-support in the
ity and purchasing power of the poor. future are more efficient as well as more at-
(iii) General transfers of income in support tractive todonors than continuing subsidies
of consumption can also raise the weighted for consumption.
 20 THE DEVELOPMENT RECORD

I.B.2. Dimensions of Poverty

It is difficult to measure the extent of poverty. per 1,000 — 20 times that in the industrial-
To begin with, absolute poverty means more ized countries.
than low income. It also means malnutrition, Nor is there any serious disagreement
poor health and lack of education — and not about who the poor are. Half of the people in
all of the poor are equally badly off in all re- absolute poverty live in South Asia, mainly in
spects. There is also room for disagreement India and Bangladesh. A sixth live in East
about where to draw the line between the and Southeast Asia, mainly in Indonesia. An-
poor and the rest, and about the correct way other sixth are in Sub-Saharan Africa. The
to calculate and compare incomes and living rest— about 100 million people — are divided
standards at different times and in different among Latin America, North Africa and the
places. Middle East. With the partial exception of
To compound these difficulties, the data Latin America (where about 40 percent are
are inadequate. Household surveys, if they in the towns) the poor are primarily rural
exist, sometimes underrepresent the poor. dwellers, overwhelmingly dependent on agri-
Very few follow the fortunes of individuals culture— the majority of them landless (or
and families through time, or disaggregate nearly landless) laborers. Some minority
the household to examine the well-being of groups — for example, the Indians in Latin
women, children and the elderly. Nor is di- America and the scheduled castes in India —
rect observation necessarily a reliable basis are also overrepresented among the poor.
for generalization, especially in the country- And there is a tendency for absolute poverty
side, where many of the poor are beyond the in particular places, families and social
gaze of the casual visitor to villages and rural groups to persist from generation to
development projects — away from the roads, generation.
away from the markets and project sites, or
on the outskirts of the villages. THREE DECADES OF
Despite all this, no one seriously doubts POVERTY REDUCTION
that a very large number of people are ex-
tremely poor. Taking as the cutoff a level of In aggregate, however, considerable prog-
income based on detailed studies of poverty ress has been made in reducing the incidence
in India, the number of people in absolute of poverty over the past 30 years (see Figure
poverty in developing countries (excluding 1). Progress would have been greater still but
China and other centrally planned econo- for the dramatic growth of population, which
mies) isestimated at around 780 million. In has doubled the number of people in the de-
1975, about 600 million adults in developing veloping world since 1950 and has begun to
countries were illiterate; and only two-fifths slow down — though as yet slightly — only
of the children in these countries currently since the mid-1960s.
complete more than three years of primary Since 1950 income per person in the de-
school. In 1978, 550 million people lived in veloping world has doubled. But in low-in-
countries where the average life expectancy come countries, the average increase has
was less than 50 years, 400 million in coun- been half that, and in both low- and middle-
tries where the average annual death rate of income countries the incomes of the poor
children aged one to four was more than 20 have grown more slowly than the average.
The proportion of people in absolute poverty
in developing countries as a group is esti-
•From World Bank Development Report, 1981,
mated to have fallen during the past two dec-
Oxford University Press, 1981, pp. 33-6. Reprinted ades (though probably not in Sub-Saharan
by permission. Africa in the 1970s). But because population
Income
1950 1960 1980 Industrialized countries
GNP per Person" (1980 dollars)
Industrialized countries
3,841 5,197 9,684
Middle-income countries 625 802
1,521 Industrialized Ml(id|e_jn<:ome
mtncs countries
Low-income countries 164 174 245 1950 Middle-income Average annual growth. 1950-80
countries
Average annual growth (percent) 1950-60 1960-80 1980
3.2
-
3.0
Industrialized countries 3.1
jw-inco
U
Middle-income countries 2.5 3.3 Low-inco 3.1
Low-income countries 0.6 countries 1.3
1.7 1980
"Excludes all centrally planned economies. L i
0 1.000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000
GNP per person ( 1 980 dollars)

\ = 50 million people
Developed
Health
Life Expectancy at Birth (years)
Increase,
1950 fftttfftttttttti
countries"
1978
1950 1960 1978
1950-78
Industrialized 73.5 Middle-income
7.5
countries 66.0 69.4
1950 fftttfttti
Middle-income 9.1 countries*
countries 51.9 54.0 61.0
Low-income 49.9 Low-income
countries 35.2 41.9 14.7
Centrally planned 7.6 1950
countriesc
tttfttttffttfi
62.3 67.1 69.9
economies'7

'Includes Bulgaria, Czechoslovakia, German DR, Hungary,

Poland, Romania, USSR. 20 30 40 50 60
^Includes Albania, Cuba, North Korea, Mongolia.
^Excludes China. Life expectancy at birth (years)

f = 50 million people
Education 1950
Adult Literacy Rate (percent)
1960 1975
Literate tttfi
1950
71
Industrialized countries 95
•97
48
9-7 99 Illiterate mmttti
Middle-income countries 38 1975
22 98
Low-income countries 29
Centrally planned economies
54
99
Literate mttttttmi
Illiterate tttttftttfti
'Excludes centrally planned economies. i i
20 40 60 80
Adults in all developing countries (percent)"

ft = 100 million people

Population
Average Annual Percentage Growth 1980
tttttttttti Developed countries"

Industrialized countries
1950-60 1960-70
1.2 1.0
1970-80
0.7
2000 tttttftttfti
Middle-income countries 2.4 2.5 2.5 1980 IfTIMMttTttl Middle-income countries
1.9 2.5 2.3
tftfttttftttffti
Low-income countries
1.9 1.7 1.3 2000
Centrally planned economies

"Includes
Poland,
Bulgaria, Czechoslovakia, German DR, Hungary,
Romania, USSR. 1980 ftttttttftttttttttttttt Low-income countries'7
Includes
includes
Albania, Cuba, North Korea, Mongolia,
China.
2000
tttfttttttttttfttttttttttftttttti
Total population (billions)

FIGURE 1. Three decades against poverty.

21
 22 THE DEVELOPMENT RECORD

has grown, the number of people in absolute Venezuela

poverty has increased.
Argentina
There has also been progress in education. South Korea
The proportion of adults in developing coun- Colombia
tries who are literate is estimated to have in-
creased over the past three decades from
about 30 percent to more than 50 percent;
the proportion of children of primary-school
age enrolled in school rose from 47 percent in Proportion of
1960 to 64 percent in 1977. These advances population above
absolute income
have been shared by most countries and re-
ligions, including those that initially were poverty line, 1975"

furthest behind, such as Sub-Saharan Africa.

But the quality of schooling remains low in
many countries; and because of population
growth, there has been an increase of about
100 million in the absolute number of illit-
erate adults since 1950.
The most striking advances against pov-
erty have been in health. Average life expec-
tancy in middle-income developing countries
has risen nine years over the past three dec-
ades. In low-income countries, the increase
has been even greater — 15 years. But even • Ivory Coast
though infant mortality rates (which are a Life
major determinant of life expectancy) have • Yemen AR expectancy, 1978
fallen substantially in developing countries I I I I
since 1950, there now are so many more chil-
dren born that the absolute number of infant
deaths probably has not declined. Burma
Chile
Another way of viewing the progress of the 100 Lanka
Thailand/ / /
Urug / Trinidad & Tobago
past three decades is to compare the devel- 90
oping with the industrialized countries. The
gap in income per person between them has
widened, even in proportional terms (though
in the case of the middle-income countries
only slightly). But the gaps in education and
health have narrowed — by 15 percentage
noints in adult literacy and five years in life
expectancy.

500 1,000 1,500 2,000 2,500 3,000

Poverty and Growth
GNP per person (dollars)6
'Absolute income poverty line is income of 45th
Most poor people live in poor countries. percentile in India.
Whether absolute poverty is measured by low ''Refers to year specified in each section of figure,
expressed in the prices of that year.
income, low life expectancy or illiteracy,
there is a strong correlation between the ex- FIGURE 2. National income and national poverty.
tent of poverty in a country and its GNP per
 23 INDICATORS OF DEVELOPMENT

person (see Figure 2). This suggests that the relevant data are sparse and unreliable, how-
solution to poverty is economic growth. There ever, itremains a matter of dispute how con-
is a great deal of truth in this proposition, but sistently growth over comparatively short pe-
it needs to be carefully qualified. riods has reduced the proportion of the
First, comparing countries, the relation be- population in absolute poverty.
tween the extent of their absolute poverty What is clear is that different countries
and the level of GNP per person is (as the have had different experiences. The propor-
dispersion of points in Figure 2 shows) far tion below the absolute poverty line appar-
from perfect. Because of differences in in- ently has not fallen in some slow-growing
come distribution, the proportion of the pop- countries (including rural India betwen 1956
ulation below the poverty line in 1975 was and 1974) or in some periods in faster-grow-
more than twice as high in Colombia as in ing countries. But it appears to have fallen
South Korea, even though the average in- markedly over the past 25 years in several
comes of the two countries were close. Sri fast-growing countries (including Thailand
Lanka is a low-income country, yet the life and Yugoslavia) and in some slower-growing
expectancy of its people approaches that of ones (including Costa Rica and Sri Lanka).
the industrialized countries. Some middle-in- The association between economic growth
come countries, such as Morocco and the and improvements in education and health
Ivory Coast, have literacy rates below those has also been imperfect.
of the average low-income country. Third, the connection between economic
Second, looking at changes over time growth and poverty reduction goes both
within particular countries, the connection ways. Few would dispute that the health, ed-
between growth and poverty reduction over ucation and well-being of the mass of people
periods of a decade or two appears inexact. in industrialized countries are a cause, as
There is general agreement that growth, in well as a result, of national prosperity. Simi-
the very long term, eliminates most absolute larly, people who are unskilled and sick make
poverty; but also that some people may (at little contribution to a country's economic
least temporarily) be impoverished by devel- growth. Development strategies that bypass
opment— as when a tenant farmer is dis- large numbers of people may not be the most
placed by his landlord's tractor or a shoe- effective way for developing countries to raise
maker by mass-produced shoes. Because their long-run growth rates.

Comment
There is not yet a refined theory to explain the determinants of inequality in relative income
shares. Besides Kuznets' evidence, some partial hypotheses and empirical evidence are pre-
sented byI. Adelman and C. T. Morris, Economic Growth and Social Equity in Developing
Countries (1973); Felix Paukert, "Income Distribution at Different Levels of Development,"
International Labour Review (August-September 1973); Shail Jain, Size Distribution of In-
come: A Compilation of Data (1975); A. K. Sen, On Economic Inequality (1973); W. R.
Cline, "Income Distribution and Development," Journal of Development Economics (Feb-
ruary 1975); Jeffrey Williamson, "Regional Inequality and the Process of National Devel-
opment: ADescription of the Patterns," Economic Development and Cultural Change (1965).
The empirical basis for statements about trends in income distribution is still weak. For the
debate over income distribution in Latin America, see A. Fishlow, "Brazilian Size Distribu-
tion of Income," American Economic Review (May 1972); Richard Webb, "Government Pol-
icy and the Distribution of Income in Peru, 1963-1973" (1974); Joel Bergsman, "Income
Distribution and Poverty in Mexico," World Bank Staff Working Paper No. 395 (June 1980).
 24 THE DEVELOPMENT RECORD

• Taiwan (53-61)

Bulgaria (57-62)

Iran (60-68) • ^ Korea (64-7

Sri Lanka (63-70)

9 -
Colombia (64-70)
^ 7 - Canada (61-65) • • Mexico (63-68)

. Brazil (60-70)
•El Salvador (61-69)
Philippines (61-71)
' Yugoslavia (63-68)
• Venezuela (62-70)
5 -
Peru (61-71) Panama (60-69)
(54-64)
3 -

12 3 4 5 6 7 8 10
Growth rate of GNP
1~
EXHIBIT 1.2. Growth and the Lowest 40 Percent

Countries above the 45 degree line are countries in which the income share of the lowest
40 percent increased over the period so that the estimated rate of growth of income for this
group is higher than for the economy as a whole. Countries below the 45 degree line are
countries in which the relative income shares of the lowest 40 percent declined.
Comment
The Gini coefficient is a single measure of relative poverty and the most frequently en-
countered instudies of income distribution. It is based on a curve fitted to percentile shares
(see Table 1), which was developed by Lorenz and, not surprisingly, named after him — the
Lorenz curve. Figure 1 illustrates this curve and a Gini coefficient that flows from it.
The vertical axis measures the percentage of income going to income recipients, who are
arrayed in percentiles on the horizontal axis. Income recipients are ordered from the poorest
to the richest, moving from left to right. Thus in Figure 1, OX percent of the population (the
poorest group) receives a percent of the income, and so on, giving the Lorenz curve, L. Com-
plete equality would occur only if a percent of the population receive a percent of the income,
as indicated by the curve of complete equality, E. The curve of perfect inequality is OGH,
with a right angle at G. This curve represents the case where one person has 100 percent of
the income.
The shaded area in the figure, enclosed by the theoretical line of equality, £, and the ob-
served Lorenz curve, L, is known as the concentration area (or area of inequality). The Gini
coefficient is the ratio of this area to the total area under the line of equality. The simplest
computation of the Gini proceeds by taking the sum of the areas under all trapezoids, such
 25 INDICATORS OF DEVELOPMENT

as WXYZ, and subtracting this from the area under E to give the concentration area. The
required ratio then follows. As a measure of income concentration, the Gini coefficient ranges
from 0 to 1 — the larger the coefficient, the greater the inequality. Thus 0 represents perfect
equality and 1 represents perfect inequality.
For problems of data analysis, see Shail Jain, Size Distribution of Income (1975) and
Nanak Kakwani, Income Inequality and Poverty (1980).
For an illuminating application of the methodology of Gini coefficient analysis, see John C.
H. Fei, Gustav Ranis, and Shirley W. Y. Kuo, Growth with Equity: The Taiwan Case (1979).

TABLE 1. Percentile Distributions for Households

in Brazil, 1970

Percentage
Population
Decile of Income

First 1.2
Second 1.8
Third
Fourth 2.7
4.6
3.5
Fifth
Sixth 6.0
Seventh 7.9
Eighth 10.8
Ninth 16.0
Tenth 45.5
100.0

Source: Shail Jain, Size Distribution of Income (1975).

100%

Curve of
complete
inequality

X Y

Percentage of income recipients

FIGURE 1. The Lorenz curve.

 26 THE DEVELOPMENT RECORD

I.B.3. A Basic Needs Yardstick*

The disappointment with GNP per head and with serious health problems. We are back
its growth has led to greater emphasis on em- with the problem of "regrettable necessities,"
ployment and redistribution. But it was soon which should not be counted as final goods or
seen, on the one hand, that unemployment in as social achievements. Moreover, the num-
the sense in which the term is used in the de- ber of doctors does not measure the distri-
veloped countries was not the problem in the bution ofthese doctors and medical services,
developing countries and that, on the other or the degree of their specialization. Re-
hand, redistribution from growth yielded sources may be deployed in inefficient ways,
only very meager results. Furthermore, it is failing to benefit the poor. Measures such as
clear that mass poverty can coexist with a infant mortality and life expectancy, how-
high degree of equality, and reductions in ab- ever, indicate the degree to which basic needs
solute poverty are consistent with increases in have been fulfilled, rather than the resources
inequality. The concern has shifted to the expended. Likewise, literacy measures the ef-
eradication of absolute poverty, particularly fectivenes ofthe educational system, and is,
by concentrating on basic human needs. in principle, a better indicator than the num-
Meeting these needs in nutrition, education, ber of students enrolled or the student/
health, and shelter may be achieved by var- teacher ratio. In general, output measures
ious combinations of growth, redistribution are better indicators of the level of welfare
of assets and income, and restructuring of and basic needs achievement. Moreover,
production. It is the composition of produc- most outputs are also inputs. Health, educa-
tion and its beneficiaries, rather than indices tion, and even nutrition are valued not only
of total production or of income distribution, in their own right, but also because they raise
that have become the principal concern. This the productivity of present and future work-
new focus on meeting basic human needs re- ers, though higher productivity is valued be-
quires an indicator or a set of indicators, cause itcontributes to a better life.
therefore, by which deprivation can be Input measures, such as doctors or hospital
judged and measured, and policies directed beds per 1000 or enrollment rates in schools,
at its alleviation and eradication can be ini- on the other hand, also have their uses. They
tiated and monitored. may reflect government intention, commit-
Whether social indicators should reflect in- ment, and efforts to provide public services.
puts or results depends on their purpose. For For purposes of assessing policies and moni-
performance testing there is something to be toring performance, both sets of indicators
said for the approach of choosing indices that are necessary. Input measures are useful in-
measure results or outputs, since these are dicators ofresources devoted to certain ob-
closer to what we are trying to achieve. Fur- jectives (though these can be misdirected).
thermore, measures of inputs can introduce To the extent to which we know how to link
biases toward certain patterns of meeting
inputs to results, i.e., have a "production
needs which may not be universal. For in- function," we can trace the connections be-
stance, a country with fairly acceptable tween means and ends. Even where we do not
health standards should not be encouraged to
have knowledge of a "production function"
acquire the same number of doctors as one (e.g., linking expenditure on family planning
to a decline in the fertility rate), the combi-
nation ofinput and output measures presents
*From Norman Hicks and Paul Streeten, "Indica- the raw material for research into the causal
tors of Development: The Search for a Basic Needs links between the two, particularly since, in a
Yardstick," World Development, Vol. 7, 1979, pp. social system of interdependent variables, so
568, 571-2, 577-9; Paul Streeten et al., First Things
First, Oxford University Press, 1981, pp. 35-7. Re- many outputs are also inputs. In addition,
printed bypermission. where output measures cannot be readily
 27 INDICATORS OF DEVELOPMENT

found, it might be necessary to fall back on was widening or narrowing. They would be
measures of inputs as useful proxies. useful in understanding which countries were
The current discussion of basic needs ori- meeting their basic needs, and how their pol-
ented development focuses on the alleviation icies are related to the growth of output,
of poverty through a variety of measures trade, investment, infrastructure, etc. Not
other than merely redistribution of incremen- enough is being done internationally to im-
tal output. Such a focus supplements atten- prove the capacity of developing countries to
tion to how much is being produced, by at- identify, collect, and issue better primary
tention to what is being produced, in what data on a regular, systematic and comparable
ways, for whom, and with what impact. Ob- basis. Such data collection can be costly, but
viously, the rapid growth of output will still a substantially greater effort seems to be
be important to the alleviation of poverty,
justified.
The problem of selecting the appropriate
and GNP per head remains an important fig-
ure. What is required, in addition, are some index in each field is best taken up by tech-
indicators of the composition and beneficiar- nical experts in each sector. To give an indi-
ies of GNP, which would supplement the cation, however, of the indicators which
GNP data, not replace them. The basic needs might be included, the following have been
approach, therefore, can be the instrument identified as a preliminary set:
for giving the necessary focus to work on so-
cial indicators.
As a first step, it might be useful to define Health: Life expectancy at birth
the best indicator for each basic need. At Education: Literacy
present, the essential basic needs are consid- Primary school enrollment
ered to cover six areas: nutrition, basic edu- (as percent of population
cation, health, sanitation, water supply and aged 5-14)
housing, and related infrastructure. This list Food: Calorie supply per head or
is not exhaustive, nor do all needs listed have calorie supply as a percent
the same status. It is intended to be illustra- of requirements
tive. A limited set of core indicators covering Water supply: Infant mortality (per thou-
sand births)
these areas would be a useful device for con-
centrating efforts. Once defined, however, Percent of population with
this set could then serve as a call for the col- access to potable water
lection ofmore adequate, standardized, com- Sanitation: Infant mortality (per thou-
sand births)
parable international statistics, and thus help
focus data gathering efforts on only the most Percent of population with
important indicators. It is not clear that, be- access to sanitation facilities
cause there are six basic needs, there need be Housing: None
only six core indicators. It may be that more
than one indicator will be necessary to mea-
sure adequately progress in any one area of The core indicators identified here attempt
basic needs. Nevertheless, the basic needs to follow the philosophy of the paper in
concept serves as a useful device for integrat- stressing measures of results, rather than in-
ing efforts of data gathering and analysis. puts. Infant mortality is assumed to be a good
Once defined, these core basic needs indi- indicator of the availability of sanitation and
cators would have the potential for important clean water facilities because of the suscep-
policy analysis relating, for instance, to inter- tibility of infants to water-borne diseases.
national comparisons of performance and rel- Furthermore, data on infant mortality are
ative aid levels. Such indicators would be a generally more readily available than data on
more useful guide to the relative "gap" be- access to water. While literacy is a good gen-
tween rich and poor countries, and offer a dif- eral measure of progress in education, the
ferent view on the speed with which this gap percent of the relevant age group enrolled in
 28 THE DEVELOPMENT RECORD

primary school is included to measure coun- would be a good single measure of basic
try effort. Input measures have also been needs. In a sense, life expectancy is a kind of
identified for water supply and sanitation as weighted "composite" of progress in meeting
supplementary measures. It has not been pos- physiological basic needs. It has the advan-
sible, however, to identify a satisfactory mea- tage of capturing the impact on individuals,
sure of housing needs. The only readily avail- not only of non-market factors but also of in-
able indicator is people per room, but this come net of taxes, transfer payments, and so-
really does not capture much of the quality cial services, without raising all the difficul-
of housing, only the number of rooms, which ties of income per head measures, such as the
in turn is a very rough index of crowding. appropriate unit (individual, household, or
Ideally, these indicators should be supple- family), the appropriate magnitude (capital,
mented bydata about distribution of calories consumption, income), the appropriate set of
per head, etc. prices (market prices, international prices),
If an acceptable system of weights could what to value as final goods, and what as
be developed, it might be possible to combine costs, etc. . . .
the core indicators into a composite basic
needs index. The chances of an acceptable
system of weights being developed, however, INCOME APPROACH VERSUS
are extremely small. Despite considerable re- BASIC NEEDS
search on composite indices, no one has come
close to developing a rational weighting sys- The income approach recommends mea-
tem. It is difficult even to suggest directions sures that raise the real incomes of the poor
for further research. (Some may question the by making them more productive, so that the
desirability of such a composite index, even if purchasing power of their earnings, together
it could be constructed.) with the yield of their subsistence production,
Instead of attempting to develop a compos- enables then to acquire the basic needs bas-
ite index of basic needs, a useful alternative ket. There can be no doubt that efforts to
may be to narrow the range of indicators make the poor more productive and their ac-
from six to one or two, which correlate highly tivities more remunerative are central to all
with basic needs development. This approach poverty-oriented development strategies.
would serve the need of those who desire a And some features of the basic needs ap-
single number for making quick judgments proach were contained in the earlier ap-
on social performance, without introducing proaches. The basic needs approach in the
the problems of weighted composite indices. narrow sense, however, regards the income-
The prospects for doing this are considerably orientation of earlier approaches as incom-
enhanced by the fact that many of the so- plete and partial, for seven reasons.
called "basic needs" are, in fact, inputs 1. Some basic needs can be satisfied only,
rather than ultimate goals. Certainly nutri- or more effectively, through public services
tion, water supply, and sanitation are valued (education, health, water, sanitation),
because they improve the health status of the through subsidized goods and services, or
population. To a more limited extent, this is through transfer payments. These services
also true of housing and education. All of call for progressive taxation, for indirect tax-
these can be considered to be inputs into the ation of luxury goods, for ensuring that the
health poor have access to the services, and for a
valued "production
for reasons other function." They influence
than their may be
system of checks against abuse. The provi-
on health status, but a high association be- sion of public services is, of course, not a dis-
tween the various core indicators can be tinctive feature of the basic needs approach.
traced to their impact on health. Therefore, But the approach is distinguished by its em-
it could be argued that some measure of phasis on investigating why these services
health, such as life expectancy at birth, have so often failed to reach the groups for
 29 INDICATORS OF DEVELOPMENT

whom they were intended, or were claimed to of their needs met than do males. In many
be intended, and why they have often rein- societies women also carry the heaviest work
forced inequalities in the distribution of pri- load, so that it cannot be argued that food is
vate income. By redesigning these services, distributed according to effort.
the basic needs approach ensures that they do 5. A substantial proportion of the destitute
reach the poor. are sick, disabled, aged, or orphaned; they
2. There is some evidence that consumers may be members of households or they may
(both poor and rich) are not always efficient, not. Their needs can be met only through
especially in optimizing nutrition and health, transfer payments or public services since, by
and especially in the case of subsistence definition, they are incapable of earning. This
farmers who become cash earners. Addi- group has been neglected by the income and
tional cash income is sometimes spent on productivity approach to poverty alleviation
food of lower nutritional value than that pre- and employment creation. Of course, the
viously consumed (as when polished rice is problems of implementation are particularly
substituted for coarse grains, or rice for difficult. Even some quite affluent societies
wheat) or on items other than food. have not been successful in eradicating the
3. The manner in which additional income poverty of their handicapped, and societies
is earned may affect nutrition adversely. Fe- with very meager resources have a much
male employment may reduce breast feeding more difficult task.
and therefore the nutrition of babies, even 6. The income approach has paid a good
though the mother's income has risen. More deal of attention to the choice of technique
profitable cash crops may replace "inferior" but has neglected to provide for appropriate
and cheaper crops, such as millets, that are products. Many developing societies import
grown for home use; or dairy farming, though or produce domestically oversophisticated
it creates employment, may divert land from products that meet excessive needs trans-
cheaper but more nutritious maize. The ferred from relatively high-income, high-sav-
human energy costs of producing a cash crop ing economies. This has frustrated the pur-
that replaces subsistence agriculture may be suit of a basic needs approach by catering to
so great in relation to wages that the depen- the demand of a small section of the popula-
dent members of the family are systemati- tion or by preempting an excessive slice of the
cally deprived of adequate nutrition. In such low incomes of the poor. An essential feature
a situation more food would mean lower lev- of the basic needs approach is to choose ap-
els of nutrition. Hydroelectric dams and ir- propriate final products and produce them by
rigation or drainage schemes, while raising appropriate techniques, thereby giving rise to
incomes, can contribute to the spread of more jobs and a more even income distribu-
water-borne diseases, such as malaria, on- tion, which in turn generates the demand for
chocerciasis, and schistosomiasis. In some these products. This goal cannot necessarily
cases, the extra costs of preventing these dis- be fully achieved by a redistribution of in-
eases are more than offset by the additional come and reliance on market responses
returns from the project. But in other cases, (though foreign trade is not ruled out).
the fate of the victims has no bearing on the 7. As already mentioned, the income ap-
project. proach neglects the importance of nonmate-
Both reasons 2 and 3 raise difficult and rial needs, both in their own right and as in-
controversial questions about free choice and struments of meeting some material needs
society's right to intervene, and about effec- more effectively, at lower cost, and in a
tive methods of aiding choice and strength- shorter period. This point becomes particu-
ening and reaching the weak. larly relevant if the nonsatisfaction of non-
4. There is maldistribution within house- material needs (such as participation) in-
holds, as well as between households; women creases the difficulty of meeting basic needs
and children tend to have a lower proportion more than that of achieving income growth.
 30 THE DEVELOPMENT RECORD

Comment
The literature of basic needs is now extensive. Most notable is Paul Streeten et al., First
Things First (1981), which contains an extensive bibliography. Other studies are the Inter-
national Labour Organization's The Basic Needs Approach to Development (1977); Paul
Streeten and Shahid Javed Burki, "Basic Needs: Some Issues," World Development (March
1978); Paul Streeten, "Basic Needs: Premises and Promises," Journal of Policy Modeling
(January 1979); Norman Hicks, "Growth vs. Basic Needs: Is There a Trade-Off?" World
Development (November/December 1979); Frances Stewart, "Country Experience in Provid-
ing for Basic Needs," Finance & Development (December 1979); Shairid Javed Burki and
Mahbub ul Haq, "Meeting Basic Needs: An Overview," World Development (February
1981); Jorge Garcia-Bouza, A Basic Needs Analytical Bibliography (OECD Development
Centre, 1980).
For critiques of the basic needs approach, see T. N. Srinivasan, "Development, Poverty and
Basic Human Needs: Some Issues," Food Research Institute Studies (1977) and Sidney Dell,
"Basic Needs or Comprehensive Development" CEPAL Review (1978). Dell argues that
there are difficulties of concept, of measurement, and of implementation in using "basic
needs" as an operational tool for planning. Three issues are raised: (1) a trade-off between
higher rates of growth and basic needs; (2) the rate at which any particular level of basic
consumption should be approached over time; (3) whether the basic needs strategy implies a
pattern of labor-intensive rural development instead of large-scale industrial growth.
Comment
The correlations between GNP per head, indicators of basic needs, and the components of
the PQLI are generally high, but by no means perfect. For individual countries, there are
deviations in the rankings that are worth pursuing for insights into policy issues. For a dis-
cussion ofthese correlations — and why per capita GNP cannot serve as a surrogate for the
other indicators of poverty — see Morris David Morris, Measuring the Condition of the
World's Poor (1979), pp. 52-6; Norman Hicks and Paul Streeten, "Indicators of Develop-
ment: The Search for a Basic Needs Yardstick," World Development (June 1979), pp. 572-
5; and A. K. Sen, "Levels of Poverty: Policy and Change," World Bank Staff Working Paper
No. 401 (1980).

I.B.4. The Inverted-U

Hypothesis*

The pathbreaking work in this area is that of oping countries. Some of his data are given in
Kuznets (1955). ' Measuring inequality by Table 1 . Kuznets explains the difference be-
the income shares of various quintiles, Kuz- tween less-developed and developed countries
nets compared India, Ceylon, Puerto Rico, this way:
the - United
, Kingdom, . and the,. United
. x. States
, , The■ former
, .
have no M middle
.... „ classes:
. 4U
there • a
is
and observed greater inequality in the devel- sharp contrast between
the preponderant propor.
♦From Gary S. Fields, Poverty, Inequality and De- tion of the Population whose average income is
velopment, Cambridge University Press, 1980, pp. well below the generally low countrywide average,
60-7, 70-1. Reprinted by permission. and a small top group with a very large relative
'Simon Kuznets, "Economic Growth and Income income excess. The developed countries, on the
Inequality," American Economic Review, March other hand, are characterized by a much more
1955, pp. 1-28. gradual rise from low to high income shares, with
 31 INDICATORS OF DEVELOPMENT

TABLE 1. Relative Inequality in Various Countries in about 1950

Income Share (%) Ratio

Poorest Richest Richest 20%

Country/Date 60% 20% Poorest 60%

India (1949-50) 28 55 1.96

Ceylon (1950) 30 50 1.67
Puerto Rico (1948) 24 56 2.33
34 44 1.29
United States (1950)
United Kingdom (1947) 36 45 1.25

Source: Kuznets, 1955, pp. 201.

substantial groups receiving more than the high The pattern of greater relative income in-
countrywide income average, and the top groups equality in less-developed countries than in
securing smaller shares than the comparable or- the developed countries was confirmed for
dinal groups in underdeveloped countries. [1955, eighteen countries in a subsequent paper by
p. 22]
Kuznets (1963).4 His findings include:
Viewed from this perspective, the variations
1 . The shares of the upper income groups are dis-
in inequality reflect real differences across
tinctly larger in the less-developed countries
countries in participation in the "advanced" (LDCs) than in the more-developed countries
or "modern" sectors of the economy. (DCs).
A few years later, Kravis (1960) presented 2. Although the shares of the lowest-income
evidence on patterns of inequality in eleven groups in some LDCs are lower than in some
DCs, the differences are much narrower than
countries.2 Taking the United States as a
standard for comparison, he found higher in- for the shares of the upper-income groups.
3. It follows from (1) and (2) that the shape of the
equality than that of the United States in income distribution curve is different in LDCs
Italy, Puerto Rico, El Salvador, and Ceylon; than in DCs.
comparable inequality to that of the United
States in Great Britain, Japan, and Canada; These findings are depicted in Figure 1.
and less inequality than in the United States Finding (1) is illustrated in the upper right
in Denmark, the Netherlands, and Israel. corner; finding (2) in the lower left. At either
Thus, like Kuznets, Kravis found that less- end of the income distribution, the Lorenz
developed countries had relatively greater in- curve for the typical LDC lies below that for
equality than did more-developed countries. the typical DC. On the likely assumption that
At about that same time, Oshima (1962) the two curves do not intersect in the middle,
suggested that inequality increases up to and it would follow that the Gini coefficient of in-
through the semideveloped stage and declines equality would be greater in LDCs than in
once a country is fully developed.3 But this DCs. The data show an average Gini coeffi-
speculation was not tested, owing to data cient of .44 for the LDCs covered by the
limitations.
data;
On .37
thatforevidence,
the DCs.5Kuznets was led to the

2Irving B. Kravis, "International Differences in the view that the level of economic development
Distribution of Income," Review of Economics and
Statistics, November 1960, pp. 408-16. 4Simon Kuznets, "Quantitative Aspects of the Eco-
nomic Growth of Nations: VIII, Distribution of In-
3Harry Oshima, "The International Comparison of
Size Distribution of Family Incomes with Special come by Size," Economic Development and Cultural
Reference to Asia," Review of Economics and Statis- Change, January 1963, pt. 2, pp. 1-80.
5Ibid., p. 17.
tics, November 1962, pp. 439-45.
 32 THE DEVELOPMENT RECORD

Income
share of
richest
XLDC

y%

Percent of

population
Richest

FIGURE 1. Patterns of relative inequality in developed countries (DC) and less-developed countries (LDC).
y%
(as measured by gross national product per The key question that is raised again and again is
capita) is a major determinant of the extent whether or not the beginnings of rapid growth in
of income inequality in a country. The spe- the developing economy must necessarily be asso-
cific nature of that dependence has come to ciated with a worsening distribution of income
[meaning here increasing relative inequality, not
be known as the inverted-U hypothesis, which
states that relative income inequality rises absolute immiserization] . . . . The careful exami-
during the early stages of development, nation of even one successful counter-example to
reaches a peak, and then declines in the later any such "historical necessity" should prove useful
in its own right. But beyond that, a fuller under-
stages. Much research effort has gone into at- standing ofsome of the underlying causal relation-
tempts to confirm or refute the inverted-U ships .. . will hopefully provide us with some policy
pattern. relevant conclusions concerning the precise condi-
One line of research has regarded the hy- tions under which "things do not have to get worse
pothesis dynamically, testing it by measuring before they get better."
relative inequality at several points in the The other way of viewing the inverted-U
course of a given country's economic devel- hypothesis is as a cross-sectional statement,
opment. Kuznets himself assumed that LDCs whereby countries at intermediate stages of
had greater inequality in their earlier stages development are thought to have more un-
of development, because everyone was equal distributions of income than do richer
thought to be more or less the same, that is, or poorer countries. Such a pattern is gener-
equally poor. No data were availabe to test
ally supported by Kuznets' studies from the
this speculation. Even today, for only a hand- 1950s and 1960s. More recently, several new
ful of less-developed countries are there data and more comprehensive data sets have been
on the distribution of income at several points compiled, which give further credence to the
in time. Economists have had to perform in- inverted-U pattern. Some of the major stud-
depth analyses of individual countries to test ies are reviewed here.
the inverted-U hypothesis. This motivation In the late 1960s and early 1970s, Adel-
appears explicitly in a number of studies of man and Morris gathered new data for forty-
growth and distribution, for instance, in the three developing countries. Their 1973 book
work of Fei, Ranis, and Kuo (1978, p. 17) on presented considerable evidence on the cor-
Taiwan.6 They write: relates of relative income inequality.7 By
6John C. H. Fei, Gustav Ranis, and S. W. Y. Kuo, 7Irma Adelman and Cynthia Taft Morris, Eco-
"Growth and the Family Distribution of Income by nomic Growth and Social Equity in Developing
Factor Components," Quarterly Journal of Econom- Countries, Stanford, Calif.: Stanford University
ics, February 1968, pp. 17-53. Press, 1973.
 33 INDICATORS OF DEVELOPMENT

means of analysis of variance, they found six opment reveal two patterns: higher inequality
factors to be important in explaining varia- on the average in the less-developed countries
tions in relative income inequality, one of than in the developed countries; and within
which was the level of economic develop- the less-developed countries, lower average
ment. We will return to the correlates of in- inequality in the very poorest countries than
equality later in this section. in the less poor ones. Let us now explore fur-
A short while later, Paukert (1973) refined ther the relationship between income in-
Adelman and Morris's estimates.8 He dis- equality ina country and the level of its eco-
carded information that he thought to be par- nomic development.
ticularly unreliable, added some new coun-
tries where good data had recently become
available, and presented summary informa-
tion on the size distribution of income in fifty-
ON THE INEVITABILITY
six countries. The Gini coefficients and gross OF THE INVERTED U
domestic product per capita are presented in
The evidence on relative inequality pat-
Table 2. For each of several alternative rela-
tive inequality measures, Paukert found that terns presented in the last section was dis-
turbing to many development economists.
inequality begins at a comparatively low The prevailing view as recently as 1975 was
level, reaches a peak in countries with per
that income distribution must get worse be-
capita incomes of $301 -$500, and then di- fore it gets better. Considerable pessimism
minishes at higher incomes. Thus the in- was expressed over the supposed trade-off be-
verted-U pattern is reconfirmed. tween growth and income equality. Some
Most recently, new intercountry evidence
were even skeptical about the welfare impli-
has been reported by Chenery and Syrquin9 cations ofthe development process itself.
and Ahluwalia10 at the World Bank, and by There are two immediate problems with
Lydall11 for the International Labour Office. this inference. One is that the conclusion is
Using updated data compiled by Jain12 for based on cross-section data rather than on
sixty-two countries, these authors also found analysis of historical trends over time. For
the inverted-U pattern in the cross-sectional this reason, Adelman and Morris, for exam-
studies.
ple, in the introduction to their book, use
In summary, the cross-section data on in-
come inequality at different stages of devel- words like "preliminary," "exploratory," and
"tentative" to describe their efforts. But hav-
ing voiced these words of caution, they and
8F. Paukert, "Income Distribution at Different others like them proceed to conclusions about
Levels of Development: A Survey of Evidence," In- the process of economic development by
ternational Labour Review, August-September
1973, pp. 97-125. viewing countries at different stages of devel-
9Hollis Chenery and Moises Syrquin, Patterns of opment. Making inference from cross-section
Development, 1950-1970, New York: Oxford Uni- data about dynamic growth processes is a
versity Press, 1975. well-established practice pioneered at Har-
10Montek Ahluwalia, "Dimensions of the Prob- vard and practiced by many over the last two
lem," in Hollis B. Chenery et al. (eds.), Redistribu- decades. The maintained assumption of such
tion with Growth, New York: Oxford University
analyses is that currently developing coun-
Press, 1974, pp. 3-37; "Income Distribution and De- tries will follow much the same patterns in
velopment: Some Stylized Facts," American Eco-
nomic Review, May 1976a, pp. 128-35; "Inequality, their development experiences as are found in
Poverty and Development," Journal of Development the cross section. This assumption requires a
Economics, Vol. 3, 1976b, pp. 307-42. leap of faith that many, myself included, are
1 'Harold B. Lydall, "Income Distribution during hesitant to make.
the Process of Development," International Labour A second problem with the inverted U is
Office, World Employment Programme Research
Working Paper No. 52, February 1977. that we are dealing with averages among
,2S. Jain, Size Distribution of Income: A Compi- groups of countries and not, for the most
lation ofData, Washington, D.C.: World Bank, 1975. part, with the information on individual
 34 THE DEVELOPMENT RECORD

TABLE 2. Gini Coefficient of Inequality and Gross Domestic Product per Capita

GDP per GDP per

Head in Head in
1965 Gini 1965
Country and Level of Gini Country and Level of
GDP per Head Ratio (U.S. $) GDP per Head Ratio (U.S. $)
Jamaica (1958) .56 465
Under $100 .30
Chad (1958) .35 68 Surinam (1962) 424
73 .55 440
Dahomey (1959) .42 Lebanon (1955-60)
Niger (1960) .34 81 Barbados (1951-2) .45
.44 368
.51 74 Chile (1958) 486
Nigeria (1959) 441
.40 Mexico (1963) .53
Sudan (1969) .48 490
.54 97
61 Panama (1969)
Tanzania (1964)
.35 Group average .494 426.9
Burma (1958) 95
.33 64
India (1956-7) $501-$1,000
.53 92
Madagascar (1960)
.419 78.3 Republic of South
Group average Africa (1965) .58 521
.42 782
$101-$200 Argentina (1961)
Morocco (1965) .50 180 Trinidad and Tobago
704
Senegal (1960) .56 192 .44
.56 (1957-8) (1962)
Venezuela .42
Sierra Leone (1968) 142 .38
.53 187 Greece (1957) 904
591
Tunisia (1971)
Bolivia (1968) .53 132 Japan (1962) .39 838
.438 723.3
Ceylon (Sri Lanka) Group average
.44 140
(1963)
101 $l,001-$2,000 .30
Pakistan (1963-4) .37 Israel (1957)
.26
South Korea (1966) 107 United Kingdom
.468 147.6 1,243
Group average (1964) .38
Netherlands (1962) .42 1,590
$201-$300
.36 278 Federal Republic of 1,400
Malaya (1957-8) .46
Fiji (1968) 295 Germany (1964) .45
213 .50
.43 France (1962) 1,667
Ivory Coast (1959)
.48 Finland (1962) .46
Zambia (1959) 1,732
207 Italy (1948) .40
Brazil (1960) .54 207 1,568
Ecuador (1968) .38 202 Puerto Rico (1963) .44 1,011
El Salvador (1965) .53 249 Norway (1963) .35 1,101
Peru (1961) .61 Australia (1966-7) .30
237
285 1,717
.60 Group average .401 1,823
Iraq (1956)
.48 240 1,485.2
Philippines (1961)
.62 275 $2,001 and Above
Colombia (1964) Denmark (1963) .37
Group average .499 244.4 .39
Sweden (1963)
2,078
$301-$550 United States (1969) .34 2,406
.365
Gabon (1960) .64 368 Group average 3,233
Costa Rica (1969) .50 360 2,572.3

Source: Paukert, 1973, Table 6.

countries themselves. Figure 2 presents Pau- relative inequality within countries grouped
kert's data. In Figure 2, the individual data by gross domestic product per capita than be-
are indicated by asterisks and averages for tween them. That is, level of income is only
each income class by large dots. Even casual an imprecise predictor of income inequality
inspection suggests much more variation in in a country.
 35 INDICATORS OF DEVELOPMENT ~~T=| 1.
00

538.50 1,176.50 1,814.50 2,452.50 3,090.50

219.50 857.50 1,495.50 2,133.50 2,771.50
1.00 h I I I I I I I I I I I I I
0.90 0.90

0.80 0.80

0.70 0.70

~ 0.60
H 0.60
£ 0.50 0.50

0.40 H 0.40
0.30
| 0.30

0.20
0.20 -
0.10
0.10 -
0.0 I I I I I I 1 I I I 1 I I I I I I I I -I 0.0
60.00 698.00 1,336.00 1,974.00 2,612.00 3,250.00
379.00 1,017.00 1,655.00 2,293.00 2,931.00
GDP per capita

FIGURE 2. Gini coefficient and gross domestic product per capita, 56 'countries. Computed from data in
Paukert, 1973, pp. 114-15.

Before regarding the inverted-U pattern as means,* very simply, that the inverted U is
inevitable, even in the cross section, we need avoidable.
to know how we.ll the inverted U fits the data. In the perspective of the overall economic
To explore this question, let us work directly development literature, it should not be too
with the individual country data rather than surprising to find that income level fails to
with the grouped data. By means of multiple provide a satisfactory explanation for the de-
regression analysis, we may determine (1) gree of inequality in a country. These results
whether an inverted U is the appropriate are consistent with the writings of many lead-
characterization of the relationship between ing development economists (e.g., Fei and
inequality and level of income, and (2) Ranis, 1964;13 Kuznets, 1966;14 and Adel-
whether any particular pattern of inequality man and Morris, 1973)15 who have been say-
change over time is inevitable. On both ac- ing that the income distribution is deter-
counts, the evidence suggests that income mined as much or more by the type of
distribution need not get worse before it gets economic development and the policies fol-
better. lowed in a given country as by the level of
These empirical findings need to be evalu- development. One can hope, therefore, that
ated carefully. The evidence does show the appropriate public policy can be designed so
highest relative inequality in the group of as to avoid a deterioration in the relative dis-
lower-middle-income countries than else- tribution ofincome and to effect an improve-
where. However, this pattern is far from in- ment in the economic status of the poor.
evitable. Concerning the inevitability issue
(the view that income distribution must get ,3J. C. H. Fei and G. Ranis, Development of the
Labor Surplus Economy, Homewood, 111.: Irwin.
worse before it gets better), we should note 1964.
how little of the variance in relative inequal- 14Simon Kuznets, Modern Economic Growth, New
ity isexplained by income level. At most one- Haven: Yale University Press, 1966.
fourth of the intercountry variation in in- ,5Adelman and Morris, Economic Growth and So-
equality isexplained by income level. This cial Equity in Developing Countries, 1973.
 36 THE DEVELOPMENT RECORD

24

22
£O 20
.5
\ Sri Lanka Yugoslavia .
e 18 # Spam
8
S. 16 1 India * South Korea ^^
o o
I Argentina .x^ Ph
E 14
o
— \ Philippines >V^ yC. Rise in
<*-I 12 — >^ ^^^Costa
Turkey Rica y^ ^\^ average incomes
o / of bottom 40 percent
£ 10 — w , -™ • • Venezuela for each 1 percent rise in
Malaysia • Mexico GNP per person
" 8 — Honduras •Peru
• ./ . Brazil
1 1 1 1 1 1 1 1 1 1 1
0 1000 2000 3000 1 00 500 1000 1500 2000 25(
GNP per person (1978 dollars) GNP per person (1978 dollars)

EXHIBIT 1.3. Income of Poorest Groups

This pattern is described by the Kuznets curve, which demonstrates that the incomes of the
poorest 40 percent of the population normally grow more slowly than the average until income
per person reaches a range of $700 or $900. Beyond this range, the incomes of poorer groups
tend to grow faster than the average. Thus the distribution of income typically is less unequal
in developed countries than in developing countries.

I.B.5. Growth-Distribution
Tradeoffs*

Concepts of progress in most developing often led to political tensions. An alternative

countries are heavily conditioned by their co- view of progress focuses more on achieving
lonial past. Many of them express their ob- an equitable society and reducing poverty,
jectives interms of "catching up" with the with growth regarded as a necessary but by
advanced industrial societies and pattern no means sufficient condition.
their economies on this model. This tendency The postwar experience of relatively rapid
is reinforced by political objectives in coun- growth in developing countries has provided
tries that wish to acquire military power and a rich body of data on this set of relations
influence. that is only now being analyzed. Since there
One drawback to the emphasis on growth is relatively little established theory to guide
is that its benefits have usually been concen- this analysis, the collection of data and the
trated on the modern sectors of the economy, formulation of hypotheses have gone hand in
and increasing inequality of incomes has hand. Although substantial progress has been
made in understanding the economic forces
at work, the results to date are largely spec-
♦From Hollis B. Chenery, "Poverty and Progress — ulative and fall considerably short of the
Choices for the Developing World," Finance & De-
velopment, June 1980, pp. 26-30. Reprinted by needs of policymakers. This article explores
permission. some of the implications of the prevalent
 37 INDICATORS OF DEVELOPMENT

views of progress in developing countries in Several of these East Asian countries pro-
the light of the information available on the vide modern approximations to the earlier
results.
idea of progress as a process in which "good
things come in clusters." Unlike most devel-
CONCEPTS OF PROGRESS oping countries, the benefits of growth have
been widely distributed in Japan, Taiwan,
Catching Up Singapore, and Korea, and the incomes of the
poor have grown almost as fast as those of the
The material success of the industrialized rich. Postwar governments have been growth
West has been a powerful incentive to the minded and authoritarian but not very re-
rest of the world to adopt elements of West- pressive, and these countries have ranked
ern experience that are conducive to accel- high on most indicators of social progress. In
erated growth. The success of countries with more typical cases growth has been achieved
different historical backgrounds and eco- at the expense of increasing the concentra-
nomic and political systems has served to tion of wealth and income, however, and the
reinforce this objective. poor have benefited much less.
The concept of catching up with the indus-
trial leaders is a product of the Industrial Equity
Revolution and its outward spread from
Western Europe. This concept both provides Although the more equitable sharing of in-
a goal for social action and suggests a means come features prominently among the politi-
by which this goal can be achieved. The tech- cal objectives of virtually all governments, it
nology and forms of economic organization is taken much less seriously in practice than
created by the advanced Western countries is the objective of rapid growth. Even though
have provided the means for accelerated widespread government intervention in pro-
growth for countries in all parts of the world. duction and income distribution is justified
Nations following this model have differed largely on the grounds of reducing poverty, in
primarily in their choice of the economic and fact most studies show that on balance the ef-
social elements to be incorporated in their fects of government revenue collection and
societies. expenditure in developing countries favor the
The prototype of a successful process of upper-income groups rather than the poor.
catching up is Japan, whose economic struc- A few developing countries have, however,
ture and income level in 1910 were not sig- gone beyond the endorsement of equitable
nificantly different from those of the poor growth and have adopted policies designed to
countries today. Econometric estimates of achieve it. Notable examples include the Peo-
the sources of Japanese growth suggest that ple's Republic of China, Cuba, India, Israel,
the process of borrowing technology from Sri Lanka, Tanzania, and Yugoslavia. Al-
more advanced countries is now virtually though their social goals vary with the form
completed and that Japan is likely to attain and extent of government control of the econ-
the income level of the United States by 1990 omy, there is a common emphasis on provid-
(Jorgensen and Nishimizu, 1978). ing a minimum level of income to the poorest
The Japanese example has had a powerful groups. In the more extreme socialist formu-
effect on Taiwan, Korea, Singapore, Thai- lations, greater equality is considered a goal
land, and other countries of East Asia. All of in itself, even if it is achieved with an adverse
these economies are now growing consider- impact on efficiency — that is, lowering the
ably faster than those of the advanced coun- incomes of the rich rather than raising those
tries and some may be able to complete the of the poor.
transformation from a state of underdevel- A pioneering attempt to reconcile the ob-
opment to one of maturity in less than the 60 jectives ofgrowth and the alleviation of pov-
years taken by Japan. erty in an operational framework was made
 38 THE DEVELOPMENT RECORD

in 1962 by the Perspective Planning Division the attainment of given levels of nutrition,
of the Indian Planning Commission (Srini- education, shelter, or industry, are mislead-
vasan and Bardhan, 1974). This approach ing because they ignore the need to achieve a
was based on a formulation in which the rate balance among the several dimensions of so-
of poverty reduction in India was determined cial progress.
by the growth of the national income, while The economist's answer to this problem is
the extent of redistribution considered feasi- to replace a set of separate objectives by a so-
ble was based on the experience of other cial welfare function that defines the goal of
countries. This approach has been refined in a society in utilitarian terms as the increase
the concept of "redistribution with growth" in a weighted average of income or consump-
(Ahluwalia and Chenery, 1974), which tion of its members over time. Although the
forms the basis of the comparative analysis in national income is one such average, the typ-
the following section. If the idea of a feasible ical income distribution gives a weight of
limit to the redistribution that can be over 50 percent to the rich (the top 20 per-
achieved with a given set of institutions is ac- cent) and less than 5 percent to the poor (the
cepted, the conflict between growth and dis- bottom 20 percent). If the growth of aggre-
tribution isreduced. gate national income is used as a goal, it
A further refinement in the concept of pov- therefore implies giving 10 to 20 times as
erty alleviation has been achieved by shifting much weight to a 1 percent increase in the
from the use of income as a measure of pov- incomes of the rich as to a 1 percent increase
erty to physical estimates of the inputs re- in those of the poor (Ahluwalia and Chenery,
1974).
quired to achieve minimum standards of nu-
trition, health, shelter, education, and other In principle, any set of weights could be
essentials. These indicators of basic needs applied to the income or consumption of dif-
provide a way of evaluating the effectiveness ferent groups to remedy this bias. One pos-
of any set of policies designed to reduce pov- sibility isto give equal weight to a given per-
erty (Streeten, 1979). The "basic needs" ap- centage increase in the income of each
proach focuses particularly on the distribu- member of society, which is the equivalent of
tion of education, health, and other public weighting by the population in each group. A
services as a necessary element of policies de- more extreme welfare function, which corre-
signed to raise productivity and to alleviate sponds to the announced goals of a few so-
poverty. This is an area in which some of the cialist societies, concentrates entirely on rais-
more effective socialist societies, such as the ing the incomes of the poor and gives social
People's Republic of China, showed marked value to increasing other incomes only to the
improvement. extent that they contribute to this objective.
Although there is no scientific way to de-
Formulating Social Objectives termine the appropriate welfare function for
any given society, the concept is useful in
The social goals of developing countries — bringing out potential conflicts in the idea of
and of international bodies representing progress and in deriving alternative measures
them — tend to be stated in political terms of performance. It will be used for this pur-
that confuse ends and means and ignore the pose in the following section.
different dimensions of progress. For exam-
ple, the goal of catching up with more ad- EXPERIENCE WITH
vanced countries is a poor proxy for improv- DISTRIBUTION
ing welfare because it often leads to emphasis
on heavy industry and other policies that con- Perceptions of the nature of progress have
centrate growth in the modern sectors of the evolved considerably as a result of the varied
economy. Similarly, many of the goals an- experience of the postwar period. Many of
nounced by international agencies, such as the early postcolonial governments set forth
 39 INDICATORS OF DEVELOPMENT

optimistic objectives that now seem highly Korea distribute the benefits of moderniza-
oversimplified. However, there has also been tion more widely. A number of other factors,
a notable willingness to learn from experi- such as the greater demand for skilled than
ence in countries with varying ideologies. Eq- for unskilled labor and the concentration of
uity-oriented countries such as the People's public expenditure in urban areas, also con-
Republic of China, Cuba, Sri Lanka, and tribute to growing inequality in many
Tanzania have found it necessary to give countries.
greater attention to economic efficiency and My present concern is with the broader as-
growth, while some of the leading exponents pects of the relations between growth and
of rapid growth — Brazil, Mexico, Thailand, distribution. How universal is the tendency
Turkey — are now taking poverty alleviation toward less equal distribution in developing
more seriously. countries? Does it lead to an absolute decline
Although scholarly interest in these rela- in welfare for some groups? What kinds of
tions has expanded rapidly in recent years, policies have served to offset these tenden-
the statistical measures needed to test and re- cies? Is social conflict an inevitable concom-
fine hypotheses are only now becoming avail- itant ofeconomic advance? Although none of
able. Twenty-five years ago, Simon Kuznets these questions can be answered with great
addressed the question: "Does inequality in confidence, the average relationships and the
the distribution of income increase or de- variety of individual experience can be
crease inthe course of a country's economic brought out by combining the available
growth?" Although his answer was based on cross-country and time-series evidence for
evidence for only a handful of countries and the postwar period.
was labeled "perhaps 5 percent empirical in- The average relationship between rising
formation and 95 percent speculation," it has income and its distribution is best shown by
provided the starting point for empirical estimates of the Kuznets curve from data for
work in this field (Kuznets, 1955). Kuznets all countries having comparable measures in
hypothesized that the distribution of income some recent period (Ahluwalia, Carter, and
tends to worsen in the early phases of devel- Chenery, 1979). Although the variation in
opment and to improve thereafter. This "U- income shares was computed separately for
shaped curve" hypothesis has been subse- each qiiintile, the general phenomenon is de-
quently verified in several cross-country stud- picted in Figure 1 by considering only two
ies based on samples of 50 or 60 developing groups: the rich (upper 40 percent) and the
countries (Ahluwalia, 1976). poor (lower 60 percent). As national income
There are several reasons for the earnings rises from the lowest observed level to that of
of middle-income and upper-income groups the middle-income countries, the share re-
to rise more rapidly than those of the poor in ceived bythe poor declines on average from
the early stages of growth. Development in- 32 percent to 23 percent of the total. In a hy-
volves a shift of population from the slow pothetical country following this average re-
growing agricultural sector to the higher-in- lationship, 80percent of the increase in in-
come, more rapidly growing modern sector. come would go to the top 40 percent of
In this process inequality is first accentuated recipients.
by more rapid population growth in rural The relationship between the income
areas and ultimately reduced by rising wages growth of different groups and that of the
produced by more rapid absorption of labor whole society can be brought out more
in the modern sector (Frank and Webb, clearly by expressing it in terms of the per
1977). The more capital-intensive type of de- capita income of each group. This is done in
velopment strategy followed by Mexico or Figure 1, which plots the per capita income
Brazil absorbs less labor and produces of the poor against that of the rich. Since the
greater concentration of income, while the income level "F" of the society is a weighted
more labor-intensive forms of Taiwan and
average of the two groups "0 and 6" ( Y =
 3,000

2,500 :

2,000

o o

1,500

1,000 -

500

500 1,000
Per capita income of bottom 60 percent 1,500
(1970 dollars)

FIGURE 1 . The Kuznets curve with country observations.

40
 41 INDICATORS OF DEVELOPMENT

AY a + .6Yb), the downward sloping straight growth of the per capita income of the poor
lines define given levels of per capita income. to the national average in the last column.
Points on these lines indicate different distri- This information, together with less com-
butions, and a growth process with a constant plete data on other countries, provides a basis
distribution is represented by a straight line for describing the following patterns of
through the origin, as in the case of Yugo- growth and distribution observed in the de-
slavia. A line deviating toward the vertical veloping world:
axis indicates growing inequality, as in the
case of Mexico or Brazil. Growing equality is • Growth-oriented pattern, illustrated by
Brazil and Mexico.
shown by Sri Lanka and Taiwan.
The Kuznets curve shown in Figure 1 con- • Equity-oriented, low growth, illustrated by
Sri Lanka.
sists of two segments: a phase of worsening
distribution up to an income level of about • Rapid growth with equity, illustrated by
$800 (of constant purchasing power) and a Taiwan, Yugoslavia, and Korea.
phase of improving distribution thereafter. In These cases illustrate the main types of de-
the first phase the per capita income of the viation from the average pattern that can be
rich grows from about $300 to $1,600 while observed in the 12 countries of Table 1;
that of the poor increases from about $100 to India, Turkey, the Philippines, and Colombia
$300. For the poorest 20 percent, the rate of follow the average relations of the Kuznets
growth is considerably less. Since an increase curve.
in national income of this magnitude may These examples suggest the following ob-
take 40 or 50 years even with the relatively servations on the relationship between in-
rapid growth rates recently experienced in come growth and social welfare in developing
developing countries, in the typical country countries. First, a small group of countries
the very poor cannot look forward to an an- has achieved rapid growth with considerable
nual increase of much more than 1 percent — equity. In addition to Taiwan, Korea, and
even though the economy is growing at two Yugoslavia, this group includes Israel, Sin-
or three times that rate. Furthermore, there
gapore, and perhaps the People's Republic of
is nothing automatic about the improvement China. The policies underlying this success-
in distribution above $800, as shown by Mex- ful performance vary from primary reliance
ico and Brazil. on market forces in Taiwan, Korea, and Sin-
gapore to substantial income transfers and
other forms of intervention in Yugoslavia and
Tradeoff between Growth and Equity
Israel. Second, substantial tradeoffs between
Although acceptable time-series data are growth and equity are illustrated by the other
only available for a dozen or so countries, cases. Although Sri Lanka has grown much
they indicate a considerable variation around less rapidly than Mexico or Brazil, the poor
this average relation. Table 1 gives selected have done considerably better in the former
measures of overall growth and of the share case. Cuba presents an even more extreme
going to the lower 60 percent for countries tradeoff, since the welfare of the poor has
having observations for a decade or more.
risen despite a continuous fall in the nation's
They are divided into three groups according per capita income since 1960 (Seers, 1974).
to the share of the increment in income going Only in the few cases where economic
to the poor. The five good performers show growth has been both rapid and fairly equi-
over 30 percent of the increment going to the tably distributed is it possible to make un-
bottom 60 percent, while the three poor per- ambiguous comparisons among countries —
formers show less than 20 percent. Whether or among different development strategies
distribution is getting better or worse is in- for a single country. In other cases it is nec-
dicated bycomparing these increments to the essary to define some properties of a social
initial distribution and by the ratio of the welfare function to make such comparisons.
 c_ ° ° rf 00 o
- O n ^ >0
oE c<3 ^ -^ <N © —« r- o «*■
© ~ ~ © © ~
°* 5 A*
CO cu

so rn »^ m <n
^ Tt m
rj <N
VO - N «ri ^
00
o Cm •rf os _<
r-' </S

I2
m m vp ^
«n
-h <n m <*
SO <N © r-> <N
O ^t fN 06 en <N c-i <-n en
in en m' <N

LT)
«t
SO
so w>
O
00 ©

£ -^
«/-> ^
m rn
O SO
q
0 0 e m 00
2 S m m cr | i-h 0 <N CN CM
0
c 0Q so
0
<N
3 U u OS
r4 <N
C3 i~ O m ^ (N OO ON
O
Ui
JO _C00
b <->
1) 06 © ON
O^3 m m m
(/I
V cu
rt 2
.J
ed
ed <*
c
u On
0 © 00
q 00 ON
on r-^ Tt ; os r-;
In
O 6CM
Cm en
so en
»/S cn en <N
r-' t* en m Csl <N (N

Os

o ON

— o ^ 10 <N
O- rf — ■ O r- —< '
03
en en ~* c^l <N
«n Tf r~-
O Cm ^r
£ 4) -<*
i-
ON

r»

I
en tr>
<N Tt f*
*
Tf
>/~> r- ^
00 00 00 <N O
v~, m m <N m
00
fS
■<l- SO
r--
SO
"*■
l^
Tf ON
1^1 <N
v-> wS
so en
so <N ON CO
en
m m so
00
0
oc
00
O
r_<
00
00
<N p
sO
so
(N cm 4 0
IT)
3CO
5ci
l^N CNl
<N

i
tr> ON en
m 00
r- «-> <N
vs
O
"d"
OO ^
0 0 <N m
«/->
00
CNl
0 m m OO m 00 u
« >< «*■
CN>

"**
-d- r-
ON
ON
w On ON
ON

»=. P £ - ^
^ m 1 ^
r- \o <*^ so so ^ <* J, S
CT~ oI ^
r- r-
tj- ^7O
I on <5 r- on
tJ- ^^ On we*
1 —
v. On ^<
5 <* ed w ^[ 00 ON ^ S O
m r-
r «d J-4 o

.2 -Js ^ o a, R
8I
5J 1133 *£
r
^ a-
§ H >- cvd ^ U O -^
r-
 43 INDICATORS OF DEVELOPMENT

To take two extreme cases from Table 1 , the fiable costs of eliminating poverty by the year
incomes of the poor have grown nearly four 2000 (Streeten and Burki, 1978). Since
times as fast over a decade in Sri Lanka as in
three-fourths of the world's poor live in very
Brazil, while the opposite is true of the in- poor countries, however, the annual cost of
comes of the rich. Since the latter receive eliminating poverty in these countries is more
greater weight in the national income, per meaningfully stated as equal to about 1 5 per-
capita income has grown 50 percent faster in cent of their gross national product (GNP),
Brazil; conversely a population-weighted even if expenditures could be designed to
index of welfare increases 50 percent faster serve only the target groups. In the light of
for Sri Lanka. Even this limited sample the distributional experience outlined in the
therefore demonstrates that judgments about previous section, the problem is seen to be
economic progress cannot be separated from vastly more difficult.
social and ethical postulates. Some of the principal constraints to a more
realistic attempt to reduce global poverty
include:
REDUCING WORLD POVERTY
Attempts to extend the concept of material 1. The multiple objectives of nation-states,
progress to a global scale run up against more among which the alleviation of poverty is
acute problems of equity than the national is- usually subordinated to a variety of na-
tionalistic goals.
sues described above. Although most govern-
ments recognize their national income as one 2. The limited scope for resource transfers in
dimension of national welfare, no one has the existing international economic order.
Official development assistance from the
suggested that global income has much rele- industrialized countries has declined from
vance to an assessment of global welfare. In-
0.50 percent of their GNP in 1960 to 0.35
stead political and economic efforts of inter-
national institutions are increasingly focused percent or less since 1970. Transfers from
on the reduction of poverty and other aspects the Organization of Petroleum Exporting
of equity as objectives that command the Countries (OPEC), while substantial, do
support of people of widely varying political not offset the negative effects of higher oil
views. prices on the growth of the oil importing
In recent years considerable efforts have developing countries.
been made to establish measures of poverty 3. Rapid growth of population, which will
double in the next 35 years even though
based on standards of nutrition, health, shel- the rate has started to decline.
ter, education, and other essentials. Conser-
vative estimates set the proportion of the What are the possibilities of more rapid
world's population that falls below a poverty progress in the face of these and other con-
line based on such minimum standards at be- straints? In an attempt to compare ap-
tween 20 and 25 percent. Although this pro- proaches to poverty alleviation, Ahluwalia,
portion has declined somewhat in the past 30 Carter, and Chenery have simulated income
years, theulationoverall growth and the numbers of absolute poor
has meant increase
that thein absolute
the world's pop-
number over the next 20 years for a large sample of
of people below this poverty line has contin- developing countries (Ahluwalia, Carter, and
ued to grow and is currently of the order of Chenery, 1979, tables 3 and 9). If the trends
800 million.
of the past 20 years — a period of relatively
In technical terms the reduction or even
rapid growth of income— continue, the num-
elimination of world poverty seems decep- ber of absolute poor in 2000 would be at
tively easy. If resources could be shifted to about the same level as in 1960. This repre-
satisfying the needs of poverty groups effi- sents rapid progress in one sense, since the
ciently, iwould
t only require a reallocation of proportion of the poor would fall from 50 per-
2 to 3 percent of the world's output per cent to 20 percent of the population of devel-
annum from 1980 onward to meet the identi- oping countries. However, since this result
 44 THE DEVELOPMENT RECORD

would be achieved only by a reduction in ab- nomic cooperation. It may even accentuate
solute poverty in middle-income countries them.
that offsets the rising numbers in the very
poor countries, it is not a long-term solution. References
The reduction in poverty will have to come M. S. Ahluwalia, "Inequality, Poverty and
from one of three sources: improved distri- Development," Journal of Development
bution, accelerated growth, or a more rapid Economics, Vol. 3, September 1976, pp.
decline in population growth. Improved dis- 307-42.
tribution isparticularly important in many M. S. Ahluwalia, N. Carter, and H. Che-
middle-income countries, such as those in nery, "Growth and Poverty in Developing
Latin America (where income is quite un- Countries," Ch. 1 1 in H. Chenery, Struc-
equally distributed), but some acceleration of tural Change and Development Policy, Ox-
growth is essential in the poor countries of ford, Oxford University Press, 1979.
Africa and South Asia. Although there are M. S. Ahluwalia and H. Chenery, "The Eco-
some short-term tradeoffs between growth nomic Framework," in H. Chenery et al.
and distribution, in the longer term it is more Redistribution with Growth, Oxford, Ox-
likely that all three types of policy will be ford University Press, 1974.
mutually reinforcing. Even within restrictive J. Edelman and H. Chenery, "Aid and In-
limits to capital transfers, the industrial come Distribution," in The New Interna-
countries can considerably improve the out- tional Economic Order: The North-South
come by giving greater priority to poverty al- Debate, J. N. Bhagwati, (ed.), Cambridge,
leviation inallocating aid among countries Mass., M.I.T. Press, 1977.
(Edelman and Chenery, 1977). C. R. Frank and R. C. Webb (eds.), Income
These projections lead to the conclusion Distribution and Growth in the Less Devel-
that, although the elimination of poverty is oped Countries, Princeton, N.J., Princeton
much more difficult than is sometimes sug- University Press, 1977, chapter 2.
gested, itremains a plausible goal for inter- D. Jorgensen and M. Nishimizu, "U.S. and
national policy. One of the principal means to Japanese Economic Growth, 1952-74: An
this end would be accomplished if the ten- International Comparison," Economic
dency of the poor to lag behind the higher- Journal, Vol. 88, December 1978, pp. 707-
income groups in the process of development 26.
could be eliminated. There is increasing ac- Simon Kuznets, "Economic Growth and In-
ceptance ofthe idea that international efforts come Inequality," American Economic Re-
should be more directly focused on reducing view, Vol. 45, March 1955, pp. 1-28.
poverty in order to offset this tendency of the Pitambar Pant, "Perspective of Development
international system. Enough examples of (1961-1976): Implications of Planning for
how this result can be accomplished have a Minimum Level of Living," reprinted in
been cited in economic systems ranging from T. N. Srinivasan and P. K. Bardhan, Pov-
socialist to free enterprise to suggest that it is erty and Income Distribution in India, Sta-
a feasible objective. tistical Publishing Society, Calcutta,
This conclusion leaves several fundamental India, 1974.
issues unresolved. To what extent should pov- Dudley Seers, "Cuba" in Chenery et al., Re-
erty alleviation replace the principle of self- distribution with Growth, Oxford, Oxford
help as a guide to international action? To University Press, 1974.
achieve this objective, will it not be necessary Paul Streeten, "Basic Needs: Premises and
to establish enforceable standards of perfor- Promises," Journal of Policy Modeling 1,
mance to assure that the benefits actually
1979,Streeten
Paul pp. 136-46. and Shahid Javed Burki,
reach the poverty groups? The new emphasis
on poverty alleviation does not resolve these "Basic Needs: Some Issues," World Devel-
old dilemmas in the field of international eco- opment, Vol. 6, March 1978, pp. 411-21.
 I.C. COUNTRY PERFORMANCE
I.C. 1 . Twenty-Five Years of
Development*

Any attempt to evaluate the post- 1945 devel- hopes. After all, the industrialized countries,
opment experience needs first to recognize in their long period of economic growth un-
the change in objectives over this period. In paral eled inhuman history, had managed to
recent years, there has been a sharp shift in increase per capita income by only about 2
emphases. It is now considered that maximi- percent per annum. There was no reason to
zation of the gross national product (GNP) expect that the so-called "underdeveloped
per capita is too narrow an objective; aims re- countries," many of which had experienced
lated to the reduction of poverty also need to no growth for millenia, would do any better.
be considered, such as improving income dis- The GNP per capita of the developing
tribution, increasing employment, and fulfill- countries grew at an average of 3.4 percent
ing basic needs. per annum during 1950-75, or 3.0 percent if
the People's Republic of China is excluded
GROWTH IN GNP PER CAPITA (see Table 1). This was faster than either the
developed or the developing nations had
Up to 1950 there had been little serious grown in any comparable period prior to
thought about the growth and development 1950, and exceeded both offical goals and pri-
prospects of what were then called the "back- vate expectations.
ward areas." The few people who had given This high average rate masks a wide diver-
the matter some thought did not have high sity of performance. On the one hand, nine
countries, with a combined 1975 population
of about 930 million people, grew at an av-
erage rate of 4.2 percent or better for the full
*From David Morawetz, "Twenty-five Years of
Economic Development," Finance & Development, period (Table 2), while a second group of
September 1977, pp. 10-13. Reprinted by permission. nine countries with 220 million people grew

TABLE 1. GNP per Capita and Its Growth Rate, by Regions, 1950-75

GNP per Capita

In 1974 Dollars Growth Rate

Population
1975 1950-75
Region (in millions) 1950 1975 (in percent
per annum)

Africa 384 170 308 2.4

4:2
820 113 320
China, People's Rep. of
East Asia 312 130 3.9
2.6
495 341
Latin America 304 1-7 ,
460 944
Middle East 81 5.2
132
1,660
South Asia 830 85
160 375 3.4
Developing countries" 2,732
654 3.2
Developed countries* 5,238
2,378
aIf the People's Republic of China is excluded, the line for developing countries reads: 1,912, 187, 400, 3.0
*A11 OECD members except Greece, Portugal, Spain, and Turkey.

45
 46 THE DEVELOPMENT RECORD

TABLE 2. GNP per Capita and Its Growth Rate, Selected Countries, 1950-75

GNP per Capita

. In 1974 Dollars Growth

Rate

Population
1975 1950-75
Country 1950 1975 (in annum)
per percent
(in millions)

Eight most populous countries

820 113 320 4.2
China, People's Rep. of
India 610 95 139 1.5
2.0
103 169
Indonesia 132
Brazil 79 373 927
307 103
Bangladesh 75 3.7
2.6
150
Nigeria -0.6*
131
287
Pakistan 69
Mexico 60 562
'- 1,092 2.7
34 7.4
Nine fastest-growing count riesc
Libya 2
Iraq 283 105a
11 4,675 5.9
16 224 1,180 5.3
China, Rep. of 817
Korea 146 504 5.1
-
Iran 34 5.1 1.6
'; 384
470
Hong Kong 43 1,321 -1.0
5.0
Jamaica 2 376 786' 6a
81,584 4.7-0.5
Israel 1,185 4.5-0.3
113
1,090 3.2*
820 320
3,287 4.2
China, People's Rep. of
Nine slowest-growing countries™ t
Rwanda 4 119 81
Burundi 4 * 195 91
117
Upper Vo!ta 6 99
Madagascar 9 180
87
Central African Empire 2 '202
272 212
290 0.2
Bolivia 6 244
Chile 11 596 700 0.7
Ghana 10 427 0.7
354
Honduras 3 0.7
322 0.7
Source: World Bank.
"1960.
*1960-75.
Countries with population of 1 million or more.
rfReal growth rates (and hence estimated 1950 GNP per capita figures) for oil exporting countries depend heavily on the
choice of base years in calculation of constant price national accounts. For example, in 1950 prices, Libya's 1950 GNP per
capita was probably less than $100.
'Excluding Cambodia (growth rate, — 1.4 percent), Vietnam (0.5 percent), and the Lao People's Democratic Republic (0.3
percent).
 47 COUNTRY PERFORMANCE

at between 3 and 4 percent. On the other fare weight attached to its recipient. The wel-
hand, the large, poor countries of South Asia fare weights would presumably be heavier for
and many countries in Africa, with a total of the poor than for the rich, and may also vary
some 1 . 1 billion people, grew in per capita in- by location, occupation, or other criteria. Un-
come terms by less than 2 percent per annum fortunately, since there is seldom agreement
between 1950 and 1975. That is to say, al- on the precise welfare weights to be applied
though itis true that for some 33 percent of to different income recipients, the ideal can-
the people of the developing world per capita not usually be calculated in practice. For this
income roughly trebled during the past reason, interest has recently returned to var-
twenty-five years, it is also true that for an- ious second-best, partial welfare indicators,
other 40 percent the increase during the same which may be used in conjunction with data
period in measured per capita income has on average per capita income to examine the
been only one or two dollars per year. extent to which growth has improved the eco-
The increasing disparity between richer nomic condition of the poor. Some of these
and poorer developing countries can be seen supplementary indicators relate to employ-
from two different perspectives. At the re- ment, others to income distribution, and a
gional level, in 1950 the average per capita third set to fulfillment of basic needs.
income of the richest regions (Latin America Most people in developing countries work
and the Middle East) was five or six times within the family or are self-employed in ag-
that of the poorest (South Asia); whereas by riculture, services, and informal industry,
1975 the multiple for the Middle East over where the concept of a "job" is much less
South Asia had risen to 13, while that for clear than in the formal sector. Furthermore,
Latin America was seven. At the individual in the absence of unemployment compensa-
country level, the ranking of 80 developing tion, only the relatively well-off can afford to
countries by GNP per capita remained re- be openly unemployed. Therefore, in most
markably stable between 1950 and 1975, developing countries, the employment prob-
while the absolute disparity between the rich- lem expresses itself more in terms of under-
est and the poorest developing countries in- employment— working too few hours or with
creased by a factor of about three. Note, excessively low productivity — than in open
though, that the relationship between initial unemployment.
regional per capita income and subsequent What has been happening to rates of un-
regional growth rate is by no means consis- deremployment indeveloping countries over
tent: Latin America, initially rich, grew rel- time? We do not know, and, what is more, it
atively slowly, while the People's Republic of may not be possible to know. A meaningful
China and East Asia, initially poor, grew "percentage of underemployment" is by def-
quickly (Table 1). Similarly, at the country inition almost impossible to estimate.
level, the correlation between the initial GNP The only data maintained over a period
per capita and the subsequent growth rate is which are available are on unemployment in
weak. a small number of countries for the period
between 1960 and 1974. These data are de-
POVERTY REDUCTION ficient for two principal reasons. They do not
measure the concept that is needed, which is
A major problem with using growth in av- unemployment plus underemployment. And
erage per capita income as the sole index of even what they do measure, they do not al-
development is that it grants equal impor- ways measure accurately or consistently. Fpr
tance to each extra dollar of income, whether example, Indian labor force surveys often
it is earned by a rich or a poor person. place urban open unemployment at around 3
Ideally, one would like to measure growth in percent. But after making a series of simple
weighted per capita income, where each and apparently quite reasonable adjustments,
extra dollar is multiplied by the social wel- David Turnham concludes in The Employ-
 48 THE DEVELOPMENT RECORD

merit Problem of the Less Developed Coun- experienced an absolute worsening in their
tries (OECD, 1971), that the true rate may situation over time? Unfortunately, the avail-
be closer to 6 to 9 percent. able data do not allow an unambiguous an-
For what they are worth, the data which swer to this question. Among the largest de-
are available for nine countries in Latin
veloping countries, only in the People's
America, three countries in East Asia, and Republic of China and possibly Mexico do
one nation each in Africa and the Middle the poor seem to have clearly become better
East show no clear trend toward a worsening off. In Bangladesh, Brazil, India, Indonesia,
of open unemployment. In most of these and Pakistan the matter is in dispute, with
countries, open unemployment fluctuated alternate sets of evidence — and sometimes
around a fairly constant trend, while in even a single set of evidence — yielding differ-
Korea and the Republic of China it clearly ent conclusions to different investigators.
declined. But note that no appropriate data
are available for the poor, heavily populated BASIC NEEDS
South Asian countries, and that nothing at
all can be said with certainty about the trend In terms of poverty reduction, even better
in underemployment. incomes are only a means to an end. The ul-
timate goal is access of the poor to the goods
INCOMES OF THE POOR and services required to fulfill their basic
needs: food, health care, education, and the
Increasing employment is not an objective like.
in itself. Rather, what is wanted is to raise The amount of food available per capita
the incomes of the poor, both relative to those seems to have increased relatively little in de-
of the rich and in absolute terms. veloping countries since 1950. The problem
In quite a large number of countries the does not seem to be basically one of supply.
share of the poorest people in GNP seems to The existing "calorie deficit" (representing
have either increased or at least remained the level below minimum nutritional needs),
fairly constant over time. This group includes 25 million tons, represents no more than 2
fast-growing, market-oriented countries like percent of current world production of food-
Iran, Israel, Korea, the Republic of China, grains. The main problem seems to be one of
and Singapore, as well as countries which effective demand — how to get purchasing
have consciously concerned themselves with power to the undernourished. Distribution is
income distribution like the People's Repub- an obstacle, too. Even when large supplies of
lic of China and the slower-growing Sri food are available in one area, distant and
Lanka. It also includes Costa Rica, El Sal- scattered rural communities often do not re-
vador, and possibly Colombia. By contrast, in ceive their share.
a second, also quite large, group, including Health standards, however, in the devel-
Argentina, Brazil, India, Mexico, Panama, oping countries have improved significantly
Peru, the Philippines, and Puerto Rico, the over the past twenty-five years, if life expec-
snare of the poorest people seems to have de- tancy istaken as the best single indicator of
clined over time. national health levels. Life expectancy, the
There seems to be no clear relationship be- average life span of people in the developing
tween the rate of economic growth and either countries, has increased in the past two dec-
(1) the degree of income inequality at a point ades by as much as it rose in a century in the
in time, or (2) the trend of inequality over industrialized nations. Today, average life
time. Fast growers include both equal and expectancy in the developing countries stands
unequal societies which have been growing at 50 years, a level attained in Western Eu-
more, and less, unequal. The same is true also rope only at the beginning of this century.
of slow growers. The increases mainly reflect a sharp drop in
Are there countries in which the poor have infant mortality rates, and provide one of the
 49 COUNTRY PERFORMANCE

strongest available indications that real stan- lect its own socioeconomic system from the
dards of living have risen on a broad front in full menu of choices? Or do large, poor coun-
the developing world since 1950. The in- tries which wish to eradicate poverty neces-
creases inlife expectancy were due more to sarily have to follow a route like that of the
these general improvements in living condi- People's Republic of China? What about
tions than to more closely defined medical small trade-dependent economies: Do they
improvements. Nevertheless, there was also have any realistic option other than to follow
considerable progress in medical services es- the Republic of China-Korea route? A num-
pecial y incontrol of communicable diseases. ber of smaller countries — Burma, Cuba,
By 1975 smallpox and plague had been vir- North Korea, Sri Lanka, and Tanzania, to
tually eradicated, while malaria — which name a few — have tried to follow basic
seemed to have been controlled by 1966 but needs-oriented paths; to what extent have
subsequently rebounded — and cholera un- they succeeded? In general, is it possible to
doubtedly kill fewer people today than they follow a strategy oriented toward income
did in 1950. equality and toward the fulfillment of basic
In education, too, significant progress has needs without closing off the economy from
been achieved. Between 1950 and 1970 the
foreign trade as the People's Republic of
number of pupils in primary schools in devel- China and some of the smaller socialist coun-
oping countries trebled, reaching 200 million. tries have done? If the economy is not closed
During the same period, the number of stu- off, how can talented people be prevented
dents in secondary and tertiary education in- from leaving (other than by distorting the do-
creased sixfold, reaching 42 million and 6 mestic income distribution) and how can the
million, respectively. The increases in atten- generation and adaptation of technology be
dance rates are equally remarkable if viewed fostered?
in percentage terms, and were widely shared Some countries seem to have been success-
throughout Africa, Asia, and Latin America. ful in leaving day-to-day decision making to
The proportion of developing country adults lower-level local management, in spite of
who are literate, which stood at about 40 per- strongly centralized political and economic
cent in 1960, had risen to 50 percent by 1970, systems. By contrast, in a number of coun-
although, again, there are disparities among tries in which the central government got it-
individual countries. self involved in detailed decision making at a
micro level — whether by production- and im-
MANY ROADS, MANY QUESTIONS port-licensing ian market-oriented system or
by detailed planning and programming in a
What can be learned from the experience socialized one — the result was often an ini-
of the past twenty-five years and what are tiative-stifling combination of red tape, bu-
some of the questions suggested by it? reaucratic delays, and arbitrary decisions.
Clearly, there is more than one feasible route Whatever the underlying political philoso-
to equitable growth and development. Some phy, adecentralized system of decision mak-
countries have succeeded by pursuing mar- ing relies heavily for success on the response
ket-oriented, outward-looking strategies, re- of the local-level functionaries to centrally
lying on entrepreneurial skills (Hong Kong, determined incentives. In a market-oriented
Korea, the Republic of China) or physical re- system these functionaries are the entrepre-
sources (oil exporting countries) as the keys neurs and managers; in a socialized one they
to growth. By contrast, the People's Republic are the local officials. Many developing coun-
of China has followed a socialist, inward- tries (the striking exceptions being some Af-
looking strategy based on considerable natu- rican countries) seem to have an abundant
ral resources, ideology, and highly effective supply of petty entrepreneurs — small traders
social organization. and "informal sector" producers — but only
Does this mean that each country can se- in a few cases are there significant numbers
 50 THE DEVELOPMENT RECORD

of larger-scale operators. What are the fac- is the underlying structure of the economy;
tors that determine the growth of medium- once growth is taking place, it seems to be dif-
scale to large-scale entrepreneurship? And ficult to effectively redistribute income
what kind of policy intervention can help to through the use of "marginal" instruments
foster it? Although something is already such as taxation and public employment.
known on the subject, more work might also These combined observations have poten-
profitably be done on the ways in which so- tially powerful implications: in particular, if
cialized systems have succeeded (and failed) equality is to be a short- to medium-term
in the parallel task of motivating local offi- goal, it simply may not be possible to "grow
cials to take initiatives and make socially ap- first and redistribute later." Rather, it may
propriate decisions. be necessary to tackle asset redistribution as
a first priority by whatever means are at
hand.
GROWTH AND POVERTY
Although it is true that some of the poorest POLITICS AND EXPECTATIONS
people in some developing countries may
have become worse off in absolute terms The historical experience suggests that po-
since 1950, the blame for this state of affairs litical stability of whatever ilk may be an im-
can hardly be laid on economic growth. portant and underrated determinant of eco-
Countries in which many poor people have, nomic growth. Most of the countries which
or may have, become worse off include at grew fastest during the period were stable,
least as many fast growers as slow growers. many of the slowest growers conspicuously
For poor, heavily populated nations like were not. Bolivia provides an interesting ex-
Bangladesh, India, Indonesia, and Pakistan, ample in this regard. To oversimplify the
only long-term, sustained growth of per cap- case: up to 1970 Bolivia had had 184 govern-
ita income, equitably distributed, offers the ments in 146 years, and its growth rate for
majority of the people any hope of economic 1950-70 was one of the slowest in the devel-
advancement. oping world. Then a period of unusual polit-
Many of the countries which experienced ical stability began — and after a short period
rapid, equitably distributed growth between the growth rate accelerated significantly.
1950 and 1975 began the period with rela- There also seems to be a tendency for ex-
tively equal asset and income distributions; pectations to rise in step with development
many of those that experienced rapid, ineq- performance. "Upper-limit" projections have
uitably distributed growth began with
been surpassed, "almost unmeetable chal-
sharply unequal distributions. This suggests lenges" have been met; yet everywhere there
that the initial distribution of assets and in- is dissatisfaction. It might be useful to bear
comes may be an important determinant of in mind that first-round successes (health im-
the trend in inequality. Such a hypothesis provements, the Green Revolution) very
makes some intuitive sense. People who own often lead to second-round problems (popu-
assets — whether physical or human capi- lation explosion, worsening of rural income
tal— are best placed to profit once growth be- distribution), and that even in the best of
gins. Furthermore, both historical and simu- cases, development is a long, slow process
lative evidence suggest that the most measured in generations rather than decades.
powerful determinant of income distribution
Comment
By viewing development as a set of interrelated changes in the structure of an economy,
several studies have attempted to measure these structural changes and to provide an empir-
ical basis for models of development. Outstanding quantitative analyses of changes in the
economic structure of LDCs are: Hollis Chenery and Moises Syrquin, Patterns of Develop-
 51 COUNTRY PERFORMANCE

ment, 1950-1970 (1975) and Hollis Chenery, Structural Change and Development Policy
(1970). These studies try to establish testable links between empirically derived development
patterns and the deductive results of development theory.
Intercountry and intertemporal comparisons of economic structure are expected to yield
more or less regular or "normal" patterns of changes in economic structure as per capita
income increases. Chenery and Syrquin examine 10 types of structural characteristics that
provide the dependent variables of the statistical analysis: investment, government revenue,
education, domestic demand, production, foreign trade, labor allocation, urbanization, de-
mographic transition, and income distribution. Ordinary least squares regressions specify all
the structural characteristics as a function of per capita GNP, population, net resource inflow,
and a time trend. The regression results are used to present the normal variation in the dif-
ferent measures of economic structure with changes in per capita income between $100 and
$1,000 (at 1964 prices). Most of the processes can be described by S-shaped curves, having
asymptotes at low and high levels of income. The periods of more rapid change, however,
occur at different levels of income for different processes. Cross-section patterns are also com-
pared with time series experience.
Comment
The statistical record of country performance can be documented from World Bank reports
for individual developing countries; World Bank's World Development Report (annual);
United Nations Economic Surveys of the regional commissions; Annual U.N. Yearbooks on
National Accounts, Demography, Trade; and the Annual Report and World Economic Out-
look of the International Monetary Fund.
The statistics should be interpreted in light of Simon Kuznets, "Problems in Comparing
Recent Growth Rates for Developed and Less Developed Countries," Economic Development
and Cultural Change (January 1972), and the references to Kravis and his associates (cited
on p. 10).

EXHIBIT 1.4. Basic Indicators

Average
Annual
Rate of Inflation
GNP per Capita
(percent)
Average
Life
Literacy
Adult
Area Annual Expectancy
Growth at Birth
Population (thousands
(millions) of square Dollars
1980
Mid-1981 kilometers) 1981 (percent) 70 81 (percent) 1981
1960-81
1960- (years)
1970-
11.2 m
Low-income economies 2,210.5 f 31,020 t 270 ^ 2.9 w
China and India 12,849 t
— — 52 w
58*
1,681.5 t 280 w -2.2
Other low-income 529.0 t 240 w 3.3 m 56 h 61 w
1 Kampuchea, Dem. — 18,171 t — —3.*H> 3.8
11.6
— m — —45
50 w
181 43
2 Bhutan 1.3 80 0.1 — mc
— —
43
3 Lao, PDR 3.5 47 80 —0.8 h' —4.63.5 — 15
44
237
4 Chad 4.5 7.4
1,284 110
5 Bangladesh 140 0.3 26 46
48
90.7 144 15.7 4<)K'
45
6 Ethiopia 32.0 1.222 140 2.1
3.7 4.1
9.3 15
141 150 1.4 7.7
7 Nepal 15.0
190 0.0 2.7 66
8 Burma 34.1 677 10.7 19 45
9 Afghanistan 16.3 648 — —1.4 11.9 5.0 20
10 54
10 Mali 1.3 5.0
6.9 1,240 190 37
9.7
 52 THE DEVELOPMENT RECORD

EXHIBIT 1.4. {Continued)

Average
Annual
Rate of Inflation
GNP per Capita
(percent)
Average
Literacy Life
Area Annual Adult Expectancy
-0.1
Growth at Birth
Population (thousands 70
(millions) of square Dollars -0.6
kilometers) 1981 (percent) 1980
Mid-1981 81 (percent) 1981
1960-81
1960- (years)
11 Malawi 6.2 118 200 2.4 10.3
1970- 25 44
210 2.7
12 Zaire 29.8 29.9 35.3 55 50
2,345 41.2 45
13 Uganda 13.0 236
28 220 52 48
4.2 230 3.2
2.8 11.6 25
14 Burundi -1.1
0.2
6.3 274 240 2.4 1.3 9.5 5 46
44
15 Upper Volta
16 Rwanda 5.3 26 250 13.1
7.1 13.4 50
17 India 260 1.7
1.4 8.1 36 52
690.2
3,288 12.6
18 Somalia 4.4 280 4.5 60
638
19 Tanzania 19.1 945 280 1.9 1.8 11.9 79 52
20 Viet Nam 55.7 330 — — — — 39
63
43
21 China 991.3 300 — — 87 67
5.6 9,561
246 300 5.0
0.2 1.5 4.6 20
22 Guinea 4.0 85
69
23 Haiti 5.1 28 300 0.5
2.5 10.0 23 54
24 Sri Lanka 15.0 66 300 -0.5 1.8 13.1 69
25 Benin 3.6 113 320 -1.6
0.6 1.9 9.4 28 43
50

26 Central African Rep. 2.4 623

72 320 4.1 12.6 48
27 Sierra Leone 3.6 320 0.4
0.4 — 12.2 15 47
45
33
10
28 Madagascar 9.0 -0.3 3.2 10.6 50
587 330
29 Niger 330 12.2
5.7 1,267
30 Pakistan 84.5 804 350 -1.1
2.8 2.1 13.1 24 50
3.3
31 Mozambique 12.5 802 — — — — 33 —
32 Sudan 19.2 380 15.9 32 47
2,506 2.5 3.7
1.3 8.9 18 48
33 Togo 2.7 380 7.6
34 Ghana 11.8 57
239 400 —
36.4
Middle-income economies 3.0 m 13.1 m 54
1,128.4 t 41,108 t 3.7 w
506.5 t 1,500 w 3.0 m
Oil exporters 15,036 t 3.8 w 13.8 m 58 w 60 wh
1,250 w 57
Oil importers 621.9 t 26,072 t 3.7
-0.3w 3.0 m 13.0 m 72 w
1,670 w 63 w
Lower middle-income 663.7 t 19,302 t 850 w 3.4 w 2.8 m 11.1 m 59 w 57 w
420 65 h-
35 Kenya 17.4 583 2.9 1.6 7.9
10.2 47 56
36 Senegal 196 430 1.7 9.0 10 44
43
5.9 460
37 Mauritania 1.6 1.5 2.1 17 44
46
38 Yemen Arab Rep.
7.3 1,031
195
460
5.5 — 15.6 21
39 Yemen, PDR 2.0 333
460 — — — 40

40 Liberia 111 520 1.2 1.9 25

1.9
41 Indonesia 149.5 530 4.1
7.0
— 8.9
20.5 62
1,919 2.7 54
52
42 Lesotho 1.4 540 10.5
30 63 54
51
43 Bolivia 600 1.9 3.5 23.0
5.7 52
44 Honduras 1,099
112 600 1.1 2.9 9.1 60 59
3.8
45 Zambia 5.8 753 600 0.0 7.6 8.4 44 51
2.6
46 Egypt 43.3 650 44
1,001 3.5 11.1
47 El Salvador 650 1.5 63
4.7 770 0.5 10.8 57
48 Thailand 48.0 21
514 4.6 1.8 10.0 86 63
49.6 300 790 62
75 63
49 Philippines 2.8
5.8 13.1
 53 COUNTRY PERFORMANCE

EXHIBIT 1.4. (Continued)

Average Annual
GNP per Capita Rate of Inflation

(percent)
Average
Life
Area Literacy
Adult Expectancy
Annual
Growth at Birth
Population (thousands 70
(millions) of square Dollars
Mid-1981 kilometers) 1981 (percent) 81 1980
(percent) 1981
1960-81 42
(years)
1960- 1970-
50 Angola 7.8 8.6
3.1 462
1,247 840 28
51 Papua New Guinea 2.5 4.0 32 51
52 Morocco 20.9 447 860 2.0 8.2
2.8 130 860 2.4
0.6 1.8 14.2 90
53 Nicaragua 4.0 14.2
87.6 924 870 3.5 57
49
54 Nigeria
57
7.2 391 870 34
55 Zimbabwe 1.0 1.3 10.1 55
10.6 —69 73
56 Cameroon 8.7 475 880 2.8 4.2 50
57 Cuba 9.7 115 — —1.0 — — 95
1.7 342 2.6 11.8 — 60
58 Congo, People's Rep. 7.5 1,110 5.9 —
59 Guatemala 109 0.3 59
1,140 10.4
60 Peru 17.0 1.0
4.3 10.4 34.3 80 58
1,285 90 71
61 Ecuador 8.6 284 1,170 6.1 14.1 81 62
1,180 4.0
62 Jamaica 2.2 11 0.8 16.8
1,180
63 Ivory Coast 8.5 322 2.3 2.8 13.0 35 47
5.6 49 1,200 3.3 2.1 9.1
64 Dominican Rep. 1,260
65 Mongolia 1.7 — — — — —67 62
66 Colombia 26.4 1,565 11.9 81 63
1,139 1,380 3.2 22.4 73
64
67 Tunisia 6.5 164 4.8 3.6 62
68 Costa Rica 2.3 51 1,420 3.0 1.9 8.2
15.9 90
69 Korea, Dem. Rep. 18.7 121 —
1,430 — — — — 61
66
45.5 781 3.5 5.6 60 62
70 Turkey 32.7
1,540 70
71 Syrian Arab Rep. 9.3 185 2.6 12.0 58 65
72 Jordan 3.4 98 1,570 —3.8 —- — 62
1,620 0.5
73 Paraguay 3.1 407 3.5 3.1 12.4 65
1,630 -0.3
464.7 t 3.0 m 18.6 m 84 w
76
93
Upper middle-income 21,806 t 65 w
2,490 w 58
74 Korea, Rep. of 38.9 98 6.9
4.2 h> 17.5 19.8 66
75 Iran, Islamic Rep. of 40.1 —
1,700 — 20.1 50
76 Iraq 13.5
1,648
435 — — 1.7 —7.4 —
85 71
77 Malaysia 14.2 330 4.3 7.6 60 65
1,840 57
78 Panama 1.9 77 3.1 1.6
1,910 76
79 Lebanon 2.7 10 — — 14.6 — 66
19.6 1.4 13.4 35 56
80 Algeria 5.1
81 Brazil 120.5 2,382 2,140 3.2 2.7
46.1 42.1
2,220 78 71
72
82 Mexico 71.2 8,512 3.8 3.5 19.1 83 66
1,973 2,250 4.8 64
83 Portugal 9.8 92 17.0 93
2,520 3.0
84 Argentina 28.2 1.9 134.2
21.4
85 Chile 11.3 2,767
757 2,560
33.0 164.6 — 71
71
86 South Africa 29.5 1221 2,560 0.7
2.3 3.0 12.8 — 63
2,770 12.6 68
87 Yugoslavia 22.5 256 5.0 85
176 2,790 19.4
88 Uruguay 2.9 1.6 51.1 60.2 94
2,820
89 Venezuela 15.4 912 2.4 1.3 12.5 82 73
68
4,220 — 75
90 Greece 132 5.4 3.2 14.8 74
9.7 4,420
91 Hong Kong 5.2 1 2.4 90
18.4
92 Israel 4.0 21 5,100 6.9
3.6 6.2 45.5 —
5,160
 54

EXHIBIT 1.4. (Continued)

Average Annual
GNP per Capita Rate of Inflation

Average (percent)

Literacy Life
Expectancy
Area Annual Adult
Growth at Birth
Population (thousands 70
(millions) of square Dollars
Mid-1981 kilometers) 1981 (percent) 1980
(percent)
1960-81 81 1981
72
(years)
7.4 1970-
93 Singapore 2.4 1 1.1
1960- 5.2 83
1.2 5 5,240 3.2 95 72
94 Trinidad and Tobago
5,670 2.9 18.7
High-income -0.4
oil exporters 15.0 t 13,460 * 6.2*
— 18.2 m 32* 57*
4,012 t
95 Libya 3.1 4.7
7.8 5.2 17.3 — 57
70
96 Saudi Arabia 9.3
1,760 8,450
12,600
— 24.3 25 55
97 Kuwait 1.5 2,150
20,900
— 18.2 60
98 United Arab Emirates
1.1 18
84 24,660 — — — 56 63
Industrial market
9.9 m
economies 719.5 t 3.4* 4.3 m 99* 73
75*
30,935 / 11,120*
74
99 Ireland 70 5.2 14.2 98
3.4
100 Spain 38.0 505 5,230 3.1
4.2 8.2 16.0 —
101 Italy 56.2 301 5,640 3.6 4.4 98 74
6,960 15.7 74
102 New Zealand 3.3 269 1.5 3.6
4.1 12.9 99 74
245 7,700
103 United Kingdom 56.0 2.1 99
9,110 14.4
73
104 Japan 117.6 372 10,080 6.3
4.0 5.1 7.4 99 77
75
7.6 6.1 99
105 Austria 10,210
4.8 84 3.6 3.7 12.0 100
106 Finland 337 10,680
6.0 99
107 Australia 14.9 11,080 2.5 3.1 11.5 100 74
108 Canada 24.2 7,687 3.3 3.1 9.3 76
75
11,400 7.6
9,976
109 Netherlands 14.2 41 5.4 7.3 99 76
11,790 3.1
110 Belgium 9.9 11,920 3.8 3.6 99 73
31 99
111 France 54.0 12,190 4.2 9.9
7.2 99 75
547 3.8
112 United States 229.8 43 12,820 2.3 2.9 75
113 Denmark 9,363 2.6 10.0 99 73
5.1 13,120
6.4 76
1 14 Germany, Fed. Rep. 61.7 249 3.2 5.0 99
13,450 3.2
1 15 Norway 4.1 14,060 3.5 4.4
4.3 8.8 76
1 16 Sweden 8.3 450
324 2.6 99 77
41 14,870 4.8
10.0 99
117 Switzerland 17,430 99
6.4 1.9 4.4
-0.2
East European
nonmarket economies 380.8 t 23,422 t
— — — — 70
72*
99*
118 Albania 2.8 29 — — — —
119 Hungary 10.7 93 5.0 — 2.9 99 71
73
120 Romania 22.5 238 2,100 8.2 — 98 71
73
121 Bulgaria
111 —
2,540 — — — —
8.9
122 Poland 35.9 313 — — — — 98

123. USSR 268.0 22,402

— — — 100 72
73
124 Czechoslovakia 15.3 128 — — — — — 72
125 German Dem. Rep. 16.7 108 — — — — —

Source: World Bank, World Development Report, 1983, pp. 148-9.

"Indicates a totai (/)•
^Indicates a weighted average (w).
indicates a median value (m).
 55 COUNTRY PERFORMANCE

EXHIBIT 1.5. Population, GNP, and GNP per Capita: Shares, Relationships, ;and Growth, 1955- 1980

Per Capita
GNP in
Current Prices Per Capita GNP
Share in World Share in World as Percentage in Constant
Populi ition of U.S. GNP 1980 Dollars

1953 1980 1955 1980

Country group 1980 1955 1955 1980

73.6 GNPa21.5 4.5 730

All developing countries 68.1 20.7 340
4.8 6.4 160
Low-income 44.7 47.1 8.1 2.7 2.2
2.5 260
160 290
China 22.0 22.2 4.7 2.5 3.2
1.6 170
India 14.4 15.2 2.2 2.3 2.2 260
190
Other 8.3 9.7 1.2 2.1 140
12.6 0.7 1.7 700
Middle-income 23.4 26.5 16.7 8.1 13.7
Major exporters of 1,580
manufactures 7.1 7.3 5.1 7.7 22.9
4.0 10.7 600
Other oil importers 6.7 8.0 3.8 10.9 1,050 2,650
9.6 11.2 5.0 8.4 9.7 500
Oil exporters 3.7 1,260
0.3 0.1 5.8
8.1 1,120
High-income oil exporters 0.2 95.8 11,080
1.4 4,900
Industrial nonmarket
8.6
economies 12.4 10.7 12.4 10.4 25.0 940
Industrial market 2,880
economies 19.3 15.4 70.6 64.8 54.7 91.8 10,610
Europe 9.2 6.5 26.6 27.9 43.2 92.8 4,940
11.0 77.9 4,640 10,720
Japan 3.3 2.6 9.5
2.4
United States 6.0 40.3 23.7 100.0 100.0 1,600 9,010
11,560
5.1 14.9 7,030
World 100.0 100.0 100.0 100.0 21.7
1,320 2,510
Source: World Development Report, 1982.
"Evaluated at current prices and exchange rates.

EXHIBIT 1.6. Growth of Population, GNP, and GNP per Capita (annual growth rates in -percentages)
1979(P)

Indicator 1975 1976 1978

1960-70 1970-75 1977

Population
Developing countries 2.3
Low income: Asia and Pacific 2.4 2.3 2.2 2.2 2.1 2.1
Africa south of Sahara 2.5 2.7 2.7 2.8
2.7 2.7 2.9
Middle income: Africa south of 2.6 2.8 2.8 2.8 2.9 2.7
Sahara 2.7
2.8
North Africa and 2.6 2.5 2.7 2.9 2.9
Middle East 2.9
2.5
Asia and Pacific 2.8 2.5 2.5 2.1 2.2
2.8 2.4
Latin America and
2.7 2.7 2.7 2.7 2.7
Caribbean 2.7 1.6 1.6 1.6
1.6
Europe 1.6 1.6
1.4
Total 2.4 2.4 2.3 2.3 2.5
2.4 2.4 3.5
Capital surplus oil exporters 3.0 3.3 3.5 3.2
1.0 0.9 0.8 3.5
0.5 0.6 3.6
0.6 0.5
Industrialized countries
1.8 1.5
Centrally planned economies 1.3 1.2 1.1 1.2
1.4
 56 THE DEVELOPMENT RECORD

EXHIBIT 1.6. (Continued) 1979(P)

Indicator -0.6 1976

1975 1977 1978
1960-70 1970-75

Gross National Product (GNP)

Developing countries 7.6 4.0 5.8 4.0
Low income: Asia and Pacific 3.9 6.2
2.9 2.8
3.6
Africa south of Sahara 4.2 3.4 2.5
4.8 5.6 2.4 0.1
Middle income: Africa south of 1.3 5.2 3.2 4.7
Sahara 8.0 7.3 7.2 4.0
North Africa and 4.9 6.2 11.6
Middle East 5.0 8.3
7.9
Asia and Pacific 10.7 4.3 9.8 7.7
5.6
7.7
Latin America and 5.6 7.1 3.2
-0.8 4.9 4.8
Caribbean
Europe 6.9 6.2 2.1 3.5 3.8
Total 5.6 3.6 5.3 4.9 3.4
5.0 3.7
5.0
6.1 7.3 -0.1
Capital surplus oil exporters 9.9 9.3 12.2 10.2 5.6
-2.7
Industrialized countries 3.3 5.3
-0.3
4.0 3.5
5.1
— -3.2 -0.2
3.7 3.7
Centrally planned economies 6.0 4.8 3.2
6.1 -1.4 5.4
Gross National Product (GNP) per capita
Developing countries
Low income: Asia and Pacific 1.5 1.3 5.3 1.8 4.0 3.6
Africa south of Sahara 1.8 0.2 1.7
0.7 2.3
Middle income: Africa south of 2.2 0.3 1.9
2.7
Sahara 4.3
North Africa and 2.2 3.6 8.6 4.4 1.1
Middle East 5.2
7.6
Asia and Pacific 5.0 5.1 2.4 8.1 8.
2.8 4.3 5.4 5.4
Latin America and 0.5
-1.6 2.1 1.5 2.1 2.8
Caribbean 2.0
Europe 4.5
2.2
2.5 1.9
2.6
5.4 0.5
Total 3.6 1.2 2.9 2.1
2.5
3.1
Capital surplus oil exporters 5.8 6.6 3.8 2.2
6.7 8.4
Industrialized countries 4.0 4.7 4.2 3.0 3.0
— 2.4 3.1
Centrally planned economies 4.5 4.5 2.6 2.0
3.7
Source: World Bank.
P = Preliminary.
"Excludes Iran.

I.C.2. Korea*
4.3°
Focusing on Korea, we see that her rapid reasons. Korea's separation from Japan and
growth dates from about 1963. Few observers the North-South partition after World War
at that time were optimistic about Korea's II left South Korea a poor agricultural half-
economic prospects, and for apparently good economy. The North had about half the man-
ufacturing sector and almost all the good
mines, and more importantly almost all the
*From Joel Bergsman, "Growth & Equity in Semi- electric power generation capacity. After five
Industrialized Countries," World Bank Staff Work-
ing Paper No. 351, 1979, pp. 22-32. Reprinted by years of independence came the Korean War,
permission. which lasted from 1950 to 1953, and devas-
 57 COUNTRY PERFORMANCE

tated the South. Sixty percent of the culti- suggested. Widespread education was vigor-
vated land was laid waste, most of the limited ously promoted by United States advisors
industrial capacity was destroyed, and over after World War II, and continued strong in
one-quarter of the population were homeless the 1953-63 decade — the literacy rate rose
refugees. Per capita income in 1953 was from 30 to over 80 percent,4 and school at-
about $196 (1976 prices), and manufactur- tendance at all levels increased dramatically.
ing was only 9 percent of GNP.1 By 1965, Korea had achieved a level of
Over the next ten years, from 1953 to human resources above the level expected for
1963, Korean policy aimed at reconstruction, a country with three times its GNP per cap-
defense, and increased private consumption. ita (Harbison and Myers, pp. 3 Iff; cited in
U.S. aid was heavily relied on; from 1953 Adelman and Robinson p. 41).
through 1958 aid averaged 15 percent of A new military government took office in
GNP.2 Import-substituting industrialization 1961. During its first year in office the strat-
in nondurable consumer goods led manufac- egy of Korea's growth was changed to ex-
turing growth of 1 1 percent per year, while port-oriented industrialization, a path it has
agriculture was somewhat neglected and followed ever since.5 The government chose
grew at about 2.5 percent per year. GNP not to carry import substitution back to in-
grew at 4.6 percent per year during 1954-58, termediate goods, capital goods, and con-
but during 1959-62 growth slowed to 3.4 sumer durables, but rather to promote con-
percent per year3 as the import substitution tinued growth of output of consumer
possibilities in consumer nondurable goods nondurables, through export markets. While
were exhausted, agriculture did poorly, and the motives of decision makers can never be
economic and political instability further re- completely known, participants in the process
duced investment and growth. The industrial suggest that the most important factor in this
growth that did occur in the 1950s did, never- turnaround was a need to earn foreign ex-
theless, further increase Korea's stock of both change. Exports had averaged a miniscule 2
physical and human capital. percent of GNP during the 1950s, and the
Looking back we can see that there were foreign aid that was paying for the rest of im-
some important bright spots in this mostly ports (11 percent of GDP) was clearly not
bleak picture of Korea in the early 1960s. Be- going to keep flowing forever. Opportunities
hind the shattered economy and the low-in- for more import substitution were seen as un-
come level was a society that has an egalitar- profitable, and there were no natural re-
ian ethos and situation, and was open to sources to export. So the chosen alternative
advancement by merit, with a population was to export manufactures.
that was relatively well educated and The turnaround in trade incentives was not
achievement oriented. Two land reforms (in sharp and immediate. Some measures to pro-
1947 and 1950) and the destruction wrought mote exports had already been taken in the
by the war of 1950-53 had left almost all Ko- late 1950s. In the 1961-64 period, progress
reans equally poor by 1953. This poverty was, was far from continuous. Devaluations were
however, the result of the preceding chaotic neutralized by inflation. Import liberalization
events; in human resources Korea was far was partly reversed in 1963. Special export
richer than the abnormally low-income level subsidy schemes were installed, modified,
and replaced by others. But in 1964 and 1965
'Edward S. Mason, Dwight H. Perkins, Kwang an effective and long-lasting package was im-
Suk Kum, David C. Cole, and Mahn Je Kim, The
plemented bythe newly elected government
Economic and Social Modernization of Korea: 1945-
1975, draft, 1978, chapter IV, p. 2. 4Irma Adelman and Sherman Robinson, Income
2David Cole and Princeton Lyman, Korean Devel- Distribution Policy in Developing Countries: A Case
opment: The Interplay of Politics and Economics, Study of Korea, Stanford University Press, 1978, p.
Harvard University Press, 1971, p. 165. 41.
3Mason et al., table IV-1. 5Mason et al., chapter IV, p 6.
 58 THE DEVELOPMENT RECORD

of President Park Chung Hee- A sharp deval- cent in 1977. Manufactured exports rose 51
uation was supported by restrictive monetary percent per year (at current prices), from 1
and fiscal policies, and subsequently inflation percent of value added in manufacturing to
was compensated for by devaluations. Quan- 96 percent in the same period. In fact, the in-
titative restrictions on imports were eased, al- crease in manufactured exports accounted
though tariffs rose. Producers for export directly for 25 percent of the increase of
could import raw materials free of duty, were manufacturing output during 1960-66, and
exempted from all indirect taxes and paid re- for 46 percent during 1966-73. For all ex-
duced rates of income tax on profits earned ports, the increase was 1 3 percent of the in-
by export, received low-interest loans, and crease in GNP during 1960-66 and 35 per-
were granted other subsidies as well. Korean cent during 1966-73.
producers had easy access to purchased in- Using input-output linkages to estimate di-
puts at world market prices and quality, and rect and indirect effects combined, Kubo and
received world market prices for their export Lewis (1978)7 estimated that import substi-
sales. At the same time, reduction in protec- tution accounted for 21 percent of Korea's
tion in the domestic market meant that prof- growth in total gross production during
its were higher for exports than for domestic 1955-63, and none during subsequent pe-
sales. This package continues in force, with riods. Export growth accounted for an esti-
some modifications, today. mated 10 percent in 1955-63, 22 percent in
Three other important sets of measures 1963-70, and 56 percent in 1970-73. In the
complemented the new trade policies. First, last period export growth was even more im-
interest rates were raised dramatically. They portant than growth in domestic demand,
reached about 1 0 to 20 percent in real terms which is very unusual. Thus Korea is an ex-
during the first few years6 and remain higher treme example of export led growth. And this
than in most other semi-industrialized coun- growth was also amazing: GDP grew by 9.3
tries (SICs). This sharp rise in interest rates percent per year from 1960 through 1977.
was followed by increased flows of personal Agricultural output grew at 4.7 percent per
savings through financial intermediaries, pro- year, and industry at over 17 percent per
viding an important source of noninflationary year. Aided by a decline in population growth
financing for the boom that was underway. (from 3 percent per year around 1960 to
Second, the President and the government about 1.7 percent in the mid-1970s), per cap-
were strongly committed to economic ita GDP in 1977 prices rose from $244 in
growth, perhaps in large part to compete with 1960 to $820 U.S. dollars in 1977— an aver-
the Korean Democratic Republic to the age annual increase of 7.4 percent.
North. As an illustration of this commitment,
Since the early 1960s the government's
the President personally took part in the ex- stable and strong commitment to economic
port promotion drive. He held monthly meet- growth and its pragmatic competence have
ings to review achievements, to revise targets been important contributors to growth. Nei-
^upwards; actual exports almost always ex- ther conflicting goals nor inept management
ceeded targets), and to direct the government have been allowed to restrain actions needed
to solve problems and ease the path for for growth. Decisions are taken quickly, and
exporters.
The final, and crucial, part of the success if they don't work they are changed. Private-
sector activities are regulated in the first in-
story was the well-educated labor force that stance by generalized price incentives, but
was available to work at relatively low wages.
The results were amazing. Exports rose particular firms that don't perform are said
to find that they run into trouble with tax ex-
from 3.3 percent of GNP in 1960 to 48 per-
7Y. Kubo and J. Lewis, "Sources of Industrial
Growth in Three Asian Countries," draft, September
sMason et al., chapter VIII, p. 47. 19, 1978.
 59 COUNTRY PERFORMANCE

aminers or banks. (Financial intermediaries but still place Korea among the more egali-
in Korea are controlled by the government.) tarian SICs.
Policies and targets are set by the govern- The overall low degree of inequality was
ment, but private businessmen are consulted the result of low inequality for both rural and
both before and after the decisions are taken urban incomes separately and a relatively
and, as already noted, the decision is changed small difference between rural and urban in-
if the desired results are not following. Many comes. These were in part the result of delib-
public-sector enterprises yield a profit (as erate policy designed to achieve such a goal
they typically do not in other SICs) because (land reform, maintaining agriculture's
they are supposed to and because when they terms of trade), and in part a result of certain
don't their management is replaced. Many initial conditions combined with the type of
observers think that an important factor for growth that took place. For Korea there was
private-sector growth (which has, after all, no trade-off between growth and equity; to
been the overwhelming source of GDP the contrary, the same conditions and policies
growth) in Korea has been the knowledge that produced the growth also maintained the
that government is committed to help, and egalitarian distributions.
the fact that the government does help to Our simplified interpretation of the me-
achieve increased output and increased ex- chanics ofthat process is shown in Figure 1 .
ports, at or very near to internationally com- We can start from the results on the bottom
petitive costs. of the figure and trace the mechanism back
But while a commitment to growth by a to its proximate causes. The equality within
stable and competent government was impor- the rural sector is easy to understand. The
tant, the crucial aspect seems to have been two land reforms of 1947 and 1949 gave land
the combination of policies and comparative to virtually every rural family and estab-
advantage, together with favorable condi- lished a ceiling of approximately 3 hectares
tions inworld markets. Korea's policies to ex- on holdings. The result was that almost every
port manufactures and to keep her labor family had a very small farm. This equality
cheap were the perfect complements to her was maintained by: (a) continued enforce-
lack of natural resources, modest internal ment of the 3 hectare ceiling; (b) the slow-
market, but well-educated and willing labor down of population growth; and (c) the rapid
force and her talents for management and absorption of labor by industry — 10.7 per-
organization. cent per year from 1963 to 1975, one of the
Korea's record of rapid income growth is highest rates ever recorded in any country.
all the more interesting because of her si- These factors prevented both a reconcentra-
multaneous near-elimination of absolute pov- tion of holdings and the reformation of a
erty. Thus Korea's income distribution is large group of landless agricultural workers.
among the more equal in the developing The reasons for equality within the urban
world — although it apparently has been get- sector are more complex. Before the analysis,
ting less equal in the 1970s. The poorest 40 a qualification as to fact is in order: the urban
percent of all households received an esti- surveys on which this analysis is mostly based
mated 16.9 percent of the national total in do not cover families with incomes above
1976, and the ratio of the top 20 percent about $5,000 U.S. dollars per year.8 Analysis
share to the bottom 20 percent share was of car ownership, income tax returns, and
only 8 to 1 . These were among the most egal- casual observation suggest that a small but
itarian values measured for any LDC. Data growing percentage of urban families has
from 1963 to 1970 show at most a small in- been receiving very high incomes. Even so,
crease in inequality — remarkable in view of the estimates of overall inequality through
the rapid and highly capitalistic nature of the early 1970s are low compared to other
growth during that time. The latest data
(1976) show sharper increases in inequality, "Mason et al., chapter XII, p. 3.
 T3 (U
o S w<« 2
•°
o so — I sits. •- c-a 2
_ <*- w JS

i c « "

2.° *

>>
c ° 5 3

M O°° ««
UE -c
™
o

>
a
<Ji

3 O
c

ill
eo +- °
2 E 3 rt
D. c
o
!/)
3
P >
£
o

4)

x>

£ ■O O
U

(50 3
ca 3

E-ac
OS

99f 60

£ .s ^

£ 5 S
&o<*2 >>
„ 3 a
«a pw cir
'3 M

Z .3 p -ago
o 2 3
3 J23 OS
T3
Pi
-2/ X! m

60
 61 COUNTRY PERFORMANCE

countries. More recently, estimated inequal- the other extreme such as Mexico with a
ity has increased. ratio around 0.4 (1975 survey) or Brazil
The low degree of urban poverty — the rel- around 0.3. 9 Korea's small urban-rural in-
atively high share of income received by the come differential is also reflected in her very
urban poor— had its proximate cause in the low ratio of value added per worker in indus-
very rapid absorption of labor in higher pro- try to value added per worker in agriculture.
ductivity sectors and occupations. Mining This aspect of Korea's relative equality has
and manufacturing absorbed 38 percent of three proximate causes. First and foremost,
the total employment increase during 1963— the rapid absorption of labor in modern
75. Workers did not have to become landless urban activities that has prevented the for-
rural laborers, or urban shoe-shine boys, car mation of a large pool of poverty-stricken
guarders, etc., in any significant number be- landless rural laborers. Second, government
cause better jobs were available. This good purchase of rice and barley at high prices, the
performance on the creation of productive rapid growth in demand for vegetables and
jobs was in turn the result of rapid growth in other high-value cash crops in urban markets
output, absence of bias against labor-inten- (caused by income gains), and the rapid shift
sive products or processes, and a highly edu- of labor out of agriculture (which is the other
cated and achievement-oriented population, side of the rapid labor absorption in industry)
willing to work very long hours under highly
disciplined conditions. Korea has no effective have kept farm prices in line with farmers'
costs — whereas in many other LDCs farm
union pressures, and no effective minimum output prices have fallen relative to input
wage. Real wages have risen slightly faster
prices. Third, Korea's farmers are mostly lit-
than value added per worker for the economy erate and achievement oriented, and with
as a whole — 7 and 6.1 percent per year, re- government help they have enormously high
spectively. Inmanufacturing, real wages rose productivity per hectare, comparable to
about 9 percent per year. But at least until Japan. Korea has no important backward
the late 1970s wages remained low compared rural regions, as are found in all of the SICs
to productivity, and the absence of market re- that still have a significant amount of abso-
strictions resulted in less of the urban dual- lute poverty.
ism so common in many other countries Note that it was the same policies and so-
(from Brazil and Mexico to the United cial characteristics that promoted manufac-
States!), where those workers who can get tured exports and efficient growth elsewhere
jobs in oligopolistic, high productivity, and/ in the economy that also promoted rapid
or unionized sectors do very well while others labor absorption and hence helped to main-
subsist in what has come to be called the tain the egalitarian income distribution.
"secondary" labor market, characterized by Like almost all other SICs, Korea has not
high turnover, low earnings, and little chance engaged in much redistribution of income
for advancement.
through taxation. What equality there is re-
The third element of Korea's egalitarian sults mainly from equality in before-tax
distribution was the relatively small gap be- income.
tween urban and rural average incomes. The
ratios of average rural to urban family in- 9Carlos Geraldo Langoni, Distribuicao de Renda e
comes of 0.6 to 0.8 in Korea (except in bad Desenvolvimento Economico do Brasil, Editora Ex-
crop years) compare favorably to countries at pressao e Cultura, 1973, tables A. 1.10 and A. 1.1 2.

Comment
Much attention has been given to the "success stories" of the Asian foursome — Taiwan,
South Korea, Singapore, and Hong Kong. Excellent studies are Walter Galenson (ed.), Eco-
nomic Growth and Structural Change in Taiwan (1979); John C. H. Fei, Gustav Ranis, and
Shirley W. Y. Kuo, Growth with Equity: The Taiwan Case (1979); Larry E Westphal, "The
 62 THE DEVELOPMENT RECORD

Republic of Korea's Experience with Export-Led Industrial Development," World Develop-

ment (March 1978); Larry E. Westphal, "The Private Sector as Principal Engine of Devel-
opment: Korea," Finance & Development (June 1982); D. C. Cole and P. N. Lyman, Korean
Development: The Interplay of Politics and Economics (1971); I. Adelman and S. Robinson,
Income Distribution Policy in Developing Countries: A Case Study of Korea (1978); Edward
S. Mason et al., The Economic and Social Modernization of the Republic of Korea, 1945-75,
(1979); Bela Balassa, The Newly Industrializing Countries in the World Economy (1981),
essays 15-18.

I.C.3. Brazil*

Up to now, Brazil's extraordinary economic nomic performance has been its lagging pri-
progress — which led to real growth rates of mary energy production, which has grown at
8.5 percent a year between 1965 and 1980 — a rate one fourth less than that of energy con-
has sprung from a strategy to build an econ- sumption between 1967 and 1980 (see Table
omy modeled on those of the Western indus- 1). The shortfall has been made up by in-
trial nations. This approach which has been creasing imports of petroleum and, to a much
followed by many other middle-income coun- lesser extent, coal. Since 1973, petroleum-
tries, usually with less success, has brought fueled growth has been maintained only at
material well-being to a considerable portion the cost of massive foreign borrowing and ac-
of the population. But serious social problems celerating inflation. By 1980, petroleum im-
persist. Malnutrition, high infant mortality, ports, accounting for 83 percent of domestic
and lack of access to basic public services are petroleum consumption, absorbed almost 50
still experienced by a large proportion of percent of total export earnings. Servicing
Brazilians. the foreign debt required another 60 percent.
Meanwhile domestic savings rates fell in both
PATTERN OF GROWTH the private and public sectors. Incentives to
save were eroded as interest rates increas-
Despite Brazil's rapid rate of population ingly lagged behind inflation over the past
increase, per capita gross domestic product five years. Thus Brazil has become highly de-
(GDP) growth over the past 15 years has pendent on continued capital inflows and pe-
been impressive, averaging 5.7 percent. With troleum imports.
a gross national product (GNP) equivalent to The style of Brazilian growth was an im-
over US$230 billion (roughly $1,940 per cap- mediate outcome of the major national goal
ita) in 1980, Brazil's economy is now the to build in Brazil a "modern" economy. De-
world's tenth largest, roughly on a par with velopment projects often utilized the most
that of the People's Republic of China or up-to-date technologies as soon as they could
Canada. Brazil is now the third largest agri- be brought to Brazil. Exchange rate policy,
cultural exporter and the tenth largest pro- fiscal incentives, and subsidized credit re-
ducer of automotive vehicles. duced the cost of capital goods, while the use
A hint of vulnerability in the country's eco- of labor was taxed. With its large domestic
market, Brazil was able to follow a strat-
*From Peter T. Knight and Ricardo Moran,
"Bringing the Poor into the Growth Process: The egy of import-substituting industrialization
Case of Brazil," Finance & Development, December
longer and more successfully than most other
1981, pp. 22-4. Reprinted by permission. countries that have taken this route and has
 63 COUNTRY PERFORMANCE

TABLE 1. Growth of the Brazilian Economy, 1965-80

Annual Average
Rate of Change

1965 (in percent)

1980
1965-80
2.6
Population (in millions) 119
100
81
Real GDP (1970 prices) 338 8.5
100
Real GDP per capita 231
100 7339 5.7
General price index, domestic supply 33.2
100
Total exports (1970 prices) 390 9.5
432
Total imports (1970 prices) 100 10.2
Agricultural production 100
100 172
400 3.7
Industrial production 9.7
185 1165 13.0
Automotive vehicle production (in thousands)

In Million Metric
Tons Petroleum

Equivalent
1967 1967-80
40
Primary energy production 5.2
Primary energy consumption 51 122
77 6.8
Primary energy imports 11 44 11.0
1980°

Sources: Fundacao Getiilio Vargas, Conjuntura Econdmica, December 1978, and February 1980; Fundacao Instituto Brasi-
leiro de Geografia e Estatistica, Censo Demografico do Brasil 1980, Resultados Preliminares; Instituto Brasileiro de Geografia
e Estatistica, Anuario Estatistico do Brasil, 1967; Ministry of Mines and Energy, National Energy Balance, 1978.
°1980 figures are estimates based on projections for 1980 in National Energy Balance, 1978, and adjusted using 1980 trade
data.

even been able to expand rapidly its exports DISTRIBUTION; EMPLOYMENT

of manufactures.
But this development strategy has required The available evidence suggests that to
increasing amounts of investment to produce some degree growth benefited the poor. Gains
an additional unit of output. The incremental in absolute income levels appear to have been
capital output ratio averaged 2.9 over the pe- widespread. The distribution of consumer du-
riod 1965-70 and 3.0 for 1970-75. But by rables illustrates the extent of trickle-down,
1975-79 it had risen to an average of 3.9. which has been facilitated by a rapid expan-
Plant expansion often involved a high degree sion of consumer credit. In 1976 almost 76
of automation. Capital-intensive and high- percent of Brazilian homes had a radio, 47
technology industries — such as petrochemi- percent had a television, and 42 percent had
cals, aircraft, and nuclear power — absorbed a refrigerator. Even among the poorest third
large amounts of capital and led to rapid in- of the rural households, located mainly in the
creases inthe demand for skilled labor that
rural Northeast, Brazil's poorest region, 52
helped raise the salaries of skilled technical percent had a radio (although very few had
and managerial personnel to levels at or either a television or refrigerator).
above those paid in much richer countries. But the gains of the rich have been im-
They have not, however, created as many measurably larger than those of the poor, re-
jobs for the unskilled. flecting persistently high levels of income in-
 64 THE DEVELOPMENT RECORD

equality. Available estimates suggest that, in its income class — the quality and distri-
for the last 20 years or so, the richest 10 per- bution ofthese services are uneven. Residents
cent of families have been receiving over 50 of the Southeast are generally much better
percent of the income, while the poorest 40 supplied than those of other regions — partic-
percent have received well under 10 percent. ularly of the poverty-ridden Northeast;
Not only is the income received by the poor- urban areas are much better supplied than
est 40 percent of families a very small share rural areas.
of the total but also the absolute amounts of To accelerate improvement in the eco-
income received by the households are mea- nomic welfare of the neglected groups in Bra-
ger. Data from the 1978 National Household zil will require a concerted effort by the au-
Sample Survey show that in 1978 about 43 thorities to extend basic services and to
percent of Brazilian families had a total in- redistribute incomes. The surest way for the
come (including income in kind) equivalent authorities to redistribute incomes is to make
to less than about $216 per month in today's more jobs available tc the poor and under-
prices (or about $47 monthly per family skilled. Rapid population growth during the
member). Such low incomes received by so past decades assures that the potential labor
many households largely explain why life ex- force will continue to grow between 2.5 and
pectancy and infant mortality in Brazil are 3.0 percent until the end of the century. If the
comparable to those in countries with much trend toward increasing labor force partici-
lower average incomes. pation rates continues, the rate of job crea-
The large differences in the incomes tion will have to increase even faster.
earned by different population groups are re- Over the period 1960-78, rapid economic
flected incorrespondingly large differences in growth expanded the number of jobs paying
the amenities available to them in different
more than one 1970 "minimum wage" (or
parts of the country. In 1970, the latest year about $76 per month in today's prices) al-
for which sufficiently detailed data are avail- most twice as fast as the working age popu-
able, life expectancy at birth in the indus- lation but roughly 30 percent slower than
trialized southern regions of Brazil was GDP. By 1978 such jobs were held by about
roughly comparable to that in many devel- half the employed labor force compared with
oped countries; the average life expectancy of one-third in 1960. Jobs paying more than one
the lowest income group in the urban areas minimum wage held by low-skilled workers
of the central northeast states, however, at 40 (those with four or less years of schooling),
years, was about the same as Ethiopia's av- however, increased about one-third as fast as
erage— a country with a per capita GNP those held by workers with five or more years
about one-eighth of Brazil's. While average of formal education, 40 percent as fast as
life expectancy in the Northeast increased 26 GDP, and only slightly faster than the work-
percent between 1960 and 1976, faster than ing age population. Between 1973 and 1976,
for Brazil as a whole, a Northeastern child total employment of adults (aged 20 and
born in the latter year could still expect to above) outside the agricultural sector in-
live eight years less than the average Brazil- creased at 4.4 percent per year while GDP
ian child. Data on health, nutrition, water growth averaged 8.1 percent. All this infor-
supply, sanitation, and education tell a simi-
lar story, though rates of improvement have mation suggests that, unless Brazil's eco-
nomic development strategy is altered, mini-
been quite fast in some areas, particularly mum growth rates of GDP on the order of 5
water supply. (See Table 2.) Despite substan- to 7 percent will be necessary in the 1980s to
tial progress over the past two decades and avoid growing social tension due to inade-
despite high numbers of physicians, hospital quate creation of jobs paying more than the
beds, educational opportunities, and so on, minimum wage, a rather low standard that is
relative to the population — which allow Bra- not enough even to feed a family of five ade-
zil to compare favorably with other countries quately except in a few rural areas of Brazil
 65 COUNTRY PERFORMANCE

TABLE 2. Selected Social Indicators, 1960 and 1978

1960 1978

Brazil Northeast Brazil" Northeast

Years

52 42 61 53
Life expectancy at birth6

Deaths of infants aged 0-12 months per thousand live births

123 169 89 122
Infant mortality'

In percent of total population

Population with access to general

water networks 21 5 52 28
Population living in homes with any 75
sanitary device (latrine, septic tank, 49
flush toilet, and so on) 24
40 76 42
Adult literacy rate 60 56

In aged 7-14 years

percent of
population
Students in grades 1-8 36 79
57 94

In percent of all children aged 0-17 years

Children with second or third degree

n.a. 2! 30
malnutrition*"

Sources: Knight, Moran, et al., 1979; Fundacao Instituto Brasileiro de Geogralia e Estatistica, Pesquisa Nacional por Amostra
de Domicilios 1978; Instituto Brasileiro de Geografia e Estatistica, Censo Demografico do Brasil, 1960; and United States
National Academy of Sciences Committee on Population and Demography, Panel on Brazil, Preliminary Report, 1979.
"Brazil data for 1978 excludes the rural areas of the North and Center West regions, roughly 5 percent ot total population.
^Estimates centered on the end of 1957 and of 1973, derived from data in the 1960 demographic census and 1976 National
Household Sample Survey, the latter excluding rural areas of the North and Center West regions.
c75 percent or less of normal body weight.

(given prevailing dietary habits), much less increases employment generated per unit of
meet other needs. new investment, reduces imports per unit of
Brazil's serious balance of payments prob- additional output, increases the proportion
lems, aless favorable international environ- of children from low-income families obtain-
ment, rising energy costs, and the need to ing the higher skills that more than four
bring down the rate of inflation suggest that years of schooling provide, and raises the pro-
the sustainable rate of GDP growth in the ductivity and well-being of the poor through
1980s will be in the lower half of this range. improving their health. Since decreased in-
Unexpected difficulties could reduce it fur- fant mortality, higher educational status of
ther. If Brazil is to continue or accelerate its mothers, and increased participation of
social progress under these difficult condi- women in the labor force are known to reduce
tions, adevelopment strategy is required that fertility, such a strategy could also help slow
 66 THE DEVELOPMENT RECORD

population growth even in the absence of an come and employment while fighting infla-
official family planning program. tion and achieving equilibrium in the balance
of payments. The idea is to alter the profiles
A NEW STRATEGY of investment and consumption, to slow im-
ports and energy use, while creating more
Many elements of such a strategy are con- jobs and giving more people the opportunity
tained in Brazil's Third National Develop- to do them. The essence of the strategy is to
ment Plan for the years 1980-85, major ob- accelerate the expansion of basic public ser-
jectives of which are to improve income vices (through the state sector) and of basic
distribution and accelerate the growth of in- wage goods (through the private sector).

I.C.4. Sub-Saharan Africa*

During the past two decades economic devel- of income per capita. And by the end of the
opment has been slow in most of the coun- 1970s, economic crises were battering even
tries of Sub-Saharan Africa. When, in the high-growth countries like Kenya, Malawi,
mid-1970s, the world economy experienced and Ivory Coast — where per capita GNP
inflation and recession, nowhere did the crisis growth had averaged an annual 2.7 percent
hit with greater impact than in this region. between 1960 and 1979 — compelling them to
The picture is not uniformly bleak. There design programs, supported by the World
are signs of progress throughout the conti- Bank, to restructure their economies. Output
nent. Vastly more Africans are in schools, per person rose more slowly in Sub-Saharan
and most are living longer. Roads, ports, and Africa than in any other part of the world,
new cities have been built and new industries particularly in the 1970s, and it rose more
developed. Technical and managerial posi- slowly in the 1970s than in the 1960s (see
tions, formerly occupied by foreigners, are Table 1).
now held by Africans. Of the 45 countries in The tragedy of this slow growth in the Af-
the region, nine posted annual growth rates rican setting is that incomes are so low and
of over 2.5 percent per capita between 1960 access to basic services so limited. Per capita
and 1979. income was $329 in 1979 (excluding Ni-
But for most African countries, and for a geria) and $411 when Nigeria is included.
majority of the African population, the re- Death rates are the highest in the world and
cord isgrim and it is no exaggeration to talk life expectancy is the lowest (47 years). Fif-
of crisis. Slow overall economic growth, slug- teen to twenty percent of the children die by
gish agricultural performance coupled with their first birthday, and only 25 percent of
rapid rates of population increase, and bal- the population have access to safe water. Of
ance-of-payments and fiscal crises — these the 30 countries classified by the United Na-
are dramatic indicators of economic trouble. tions Conference on Trade and Development
Between 1960 and 1979, per capita income (UNCTAD) as the poorest in the world, 20
in 19 countries grew by less than 1 percent are African. Of the 36 countries listed in the
per year, while during the last decade, 15 World Bank's World Development Report
countries recorded a negative rate of growth 1981 as "low income" (a per capita income
of less than $370), almost two-thirds are
African.
*From World Bank, Accelerated Development in
The economic crisis is especially evident in
Sub-Saharan Africa, 1981, pp. 2-6. Reprinted by agriculture, and is reflected in output figures.
permission. Export crop production stagnated over the
 67 COUNTRY PERFORMANCE

TABLE 1. Sub-Saharan Africa and the World: Basic Data

GNP per Capita

Average Annual Death Rate
Growth Rate Adult
Per Capita Growth Literacy Life of Children
1970-79 (percent) Expectancy
(percent) -3.5
Population -0.9 Rate at Birth thousand)
-4.5
(millions) -1.1 Volume of Aged 1-4
-2.8 (percent)
Countries Mid-1979 1960-70 1970-79 2.e8 Exports
-0.3 cult-ur 1976 1979
(years) 1979

1.3 0.8 Agri- -3.5 28 46

47 25
Sub-Saharan Africa 343.9
1.6 -0.5 -0.4
Low-income 187.1 26
82.6 0.1 4.2 n.a. 49
Nigeria
-1.5 27(per
Other middle- 22
income 74.2 1.9 -3.1 50 22
1.5 1.5 34
890.5 0.0 0.6 36 15
South Asia"
52
All developing 3,245.2 3.5 58 11
0.1 10
Low-income 2,260.2 1.8 0.1 57
50 11
Middle-income 985.0 3.9 0.6 1.9 72 57
All industrialized 671.2 4.1 0.2 5.2 99 74
61 1

Source: World Bank data files.

2.7*
*
"Bhutan, Bangladesh, Nepal, Burma, India, Sri Lanka, and1.6Pakistan.
*1970-80. 2 *
. 8
2.5*

past two decades. A 20 percent increase in rent account deficits in the region as a whole
production registered during the 1960s was rose from a modest $1.5 billion in 1970 to $8
wiped out by a decline of similar proportions billion in 1980. External indebtedness
in the 1970s. Consequently, Africa's share of climbed from $6 billion to $32 billion be-
the world market dwindled. As for food tween 1970 and 1979, and debt service in-
crops, while data are uncertain, they leave no creased from 6 to 1 2 percent of export earn-
doubt about general tendencies. Total food ings in the same period. Foreign exchange
production rose by 1 .5 percent per year in the reserves, which were comfortable in 1970,
1970s, down from 2 percent in the previous fell sharply. In 1970, reserves could cover
decade. But since population was rising rap- only two months' imports and by 1980 re-
idly— by an annual average of 2.5 percent in serves had fallen even lower. Fiscal pressures
the 1960s and 2.7 percent in the 1970s— food also intensified in many countries, as indi-
production per person was stagnant in the cated bydeclining real budgetary allocations
first decade and actually declined in the next. for supplies and maintenance, growing im-
Imports of food grains (wheat, rice, and balances between salary and nonsalary
maize) soared — by 9 percent per year since spending, and difficulties in financing local
the early 1960s — reinforcing food depen- and recurrent costs of externally funded de-
dency. Food aid also increased substantially. velopment projects.
Since 70 to 90 percent of the population The crises that evolved in much of the re-
earns its income from agriculture, the drop in gion are particularly disturbing since, during
production in this sector spelled a real income the period from 1960 to 1974, world trade
loss for many of the poorest. and the world economy in general expanded
The deterioration in agriculture and other rapidly, and many less-developed countries
internal and global factors led to widespread elsewhere experienced relatively high growth
balance-of-payment crises in the 1970s. Cur- rates. Now, against a backdrop of global eco-
 68 THE DEVELOPMENT RECORD

nomic recession, the outlook for all less-de- tors hostile to development, and rapidly
veloped nations — but especially for the Sub- growing population.
Saharan region — is grim. Although cyclical Growth was also affected by a set of exter-
factors may push prices of some African ex- nal factors — notably adverse trends in the in-
ports up from their low levels of the recent ternational economy, particularly since 1974.
past, mounting energy costs, slow growth in These include "stagflation" in the industrial-
the industrial countries (which translates into ized countries, higher energy prices, the
diminished markets for the developing relatively slow growth of trade in primary
world), and reduced growth of international products, and — for copper and iron-ore ex-
trade (factors that have plagued the global porters— adverse terms of trade.
economy for the last half decade) will make The internal "structural" problems and
renewed African growth difficult. the external factors impeding African eco-
In sum, past trends in African economic nomic growth have been exacerbated by do-
performance and continued global recession mestic policy inadequacies, of which three
together explain the pessimistic projections are critical. First, trade and exchange-rate
for African development in the 1980s. The policies have overprotected industry, held
World Development Report 1981, under its back agriculture, and absorbed much admin-
most optimistic set of assumptions about the istrative capacity. Second, too little attention
expansion of the world economy, forecasts has been paid to administrative constraints in
virtually no growth in per capita income for mobilizing and managing resources for de-
the continent in this decade1; under less fa- velopment; given the widespread weakness of
vorable assumptions, a negative rate of planning, decision making, and management
growth (—1.0 percent per year) is projected capacities, public sectors frequently become
for the poorest nations in the region. overextended. Third, there has been a consis-
These prospects and their political, social, tent bias against agriculture in price, tax, and
and economic implications are not acceptable exchange-rate policies.
either to the countries concerned or to the in-
ternational community. There is an urgent
need to understand what has gone wrong and NEW PRIORITIES AND
what must be done — by African governments ADJUSTMENTS IN POLICY
themselves and the concerned international
community — to assure a better future for Af- A reordering of postindependence priori-
rica's people. ties is essential if economic growth is to ac-
celerate. During the past two decades most
African governments rightly focused on po-
litical consolidation, on the laying down of
SOURCES OF LAGGING GROWTH basic infrastructure (much of it tied to the
goal of political integration), and on the de-
Africa's disappointing economic perfor- velopment of human resources. Relatively
mance during the past two decades reflects,
less attention was paid to production. Now it
in part, internal constraints based on "struc-
tural" factors that evolved from historical is essential to give production a higher prior-
circumstances or from the physical environ- ity— without neglecting these other goals.
ment. These include underdeveloped human Without a faster rate of production increase,
resources, the economic disruption that ac- other objectives cannot be achieved, nor can
companied decolonization and postcolonial past achievements be sustained. Three major
consolidation, climatic and geographic fac- policy actions are central to any growth-ori-
ented program: (1) more suitable trade and
exchange-rate policies; (2) increased effi-
1World Bank, World Development Report 1981, ciency of resource use in the public sector;
New York, Oxford University Press, 1981, table 1.1. and (3) improvement in agricultural policies.
 69 COUNTRY PERFORMANCE

EXHIBIT 1.7. Indexes for 18 Low-Income African Countries"

1960-70 -3.3
1970-77 -5.3
1977-80
2.5 -2.6
Population growth rate 2.8
4.1 2.7
2.9 2.9
GDP: growth rate p.a.6 16
18 16 16
Investment as percent of GDPC
12 11
Savings as percent of GDPC
6.2 2.3
Exports: growth rate p.a.*
Imports: growth rate p. a.6 6.7
2.7
70 5.3 7.2
110
Resource gap as percent of GDPC 100
Export price indexc
89
79 100 162
Import price indexc
Terms of trade indexc 100 68
1.8 3.0
9.0
Current account deficit as percent of GDPC
6.4 5.4
18.2
Debt service as percent of exports'7 4.3
Incremental capital — output ratio 6.0
5.9
249 923
Current account deficit'7 (million $ current prices) 2,258
Gross disbursement of term loansc (million $ current 423
prices)
Gross disbursement of concessional loansc (million $ 3,024
current prices) 241 939
220 ,561
569 1,672
Debt service0 (million $ current prices)
1,351
Source: World Bank.

"Benin, Burundi, Central African Republic, Chad, Ethiopia, Guinea, Kenya, Lesotho, Madagascar, Malawi, Mali, Niger,
Rwanda, Somalia, Tanzania, Uganda, Upper Volta, Zaire.
*In constant prices.
Tor terminal year.

I.C.5. China*

Three broad categories of reasons help ex- ing in certain key areas. In most nations, in
plain why China has accomplished what it contrast, it is the rich or the more educated
has. First, the Chinese Communist Revolu- and organized "middle class" that dominates
tion was fought in the name of and to a large decision making, and it is reasonable to pre-
extent by the poor. Revolutions, of course, sume that giving the poor a major share of
have been fought in the name of many causes power would make a difference. In fact, sev-
only to see the causes betrayed when the rev- eral key features of the Chinese experience,
olution succeeded. But in China a number of in rural areas at least, can be understood only
concrete steps were taken, particularly in in these terms.
rural areas, to ensure that the interests of the To say that the Chinese put the poor in
poor majority would dominate decision mak- command does not imply that all organiza-
tions in China were run by people of poor
*From Dwight H. Perkins, "The Central Features peasant origins. The key point is that coop-
of China's Economic Development," in Robert F. eratives and communes were run by such
Dernberger (ed.), China's Development Experience
in Comparative Perspective, Harvard University people. The Politburo and other centers of
Press, 1980, pp. 121-33, 138-9, 146-7. Reprinted by power, in contrast, were manned by people of
permission. diverse origins. Even these higher-level power
 70 THE DEVELOPMENT RECORD

centers, however, saw the former poor as the China's economy is socialist to a degree not
main base of their political support. found in many other nations in the less de-
Second, China has a particular authoritar- veloped or non-Communist world. Private
ian political system. Just being authoritarian ownership is confined in the main to personal
does not distinguish China's politics from household effects, much rural housing (and a
that of a few other less developed countries. little urban housing), individual savings de-
But China is an authoritarian nation with a posits, and the right to farm small private
difference. The degree of political control plots equivalent to about 5 to 7 percent of the
over daily activities in China, whether eco- total arable land. The land in these private
nomic or social, is far more thoroughgoing plots is not privately owned; only the rights to
than in all but a handful of other less devel- farm some small piece of land can be said to
oped nations. The economic aspects of this be private.
phenomenon show how development is af- Industry, even a firm with only a few work-
fected. China has been able to institute a sys- ers, is owned by the state or collectively by
tem of centralized planning more complete rural communes and brigades. Much the
than that in any other less developed nation same is true of commerce, banking, and
except possibly North Korea. It is not that transport. There are small private rural mar-
other nations have no interest in instituting a kets where farmers sometimes trade the pro-
similar system; but most know that if they duce of their private plots, and individuals
tried, they would bring their economies may contract to do repairs on private homes
grinding to a halt. The Chinese economy has in their spare time. But these activities are a
slowed down, but it has not ground to a halt, tiny fraction of the total spectrum of industry
and it is worth attempting to understand and commerce in China. In agriculture the
why. Part of the reason, it appears, is that private sector is larger than in industry be-
China's people had acquired enough experi- cause it has so far proved impossible effi-
ence to make the new system work. They ac- ciently to collectivize much vegetable and
quired this experience over decades and even hog raising, household handicrafts, and the
centuries; it was not an instantaneous addi- like. Still, three-quarters or more of farm
tion in 1949. output is handled collectively.
Third, like all nations, China has its own Only a relatively small proportion of
peculiar resource and factor endowment. Chinese industry is controlled soley by cen-
Much of this endowment is a reflection of tral planners in Peking. Even large modern
China's great size in land mass and people. plants producing precision products for na-
Equally important is that crops can be grown tional markets are placed formally under
on only a small proportion of China's surface. provincial and city planning authorities or
China, in a very fundamental way, is a land- under joint province-central control. Smaller
short nation. plants and those producing for more localized
Elements other than these three, of course, markets are placed under planning authori-
have also played some role in shaping China's ties at the district and county level. . . .
economic development. Marxist-Leninist ide- At all levels, county, province, and center,
ology isan important example. But the term planner control is exercised directly and not
"putting the poor in command" covers some through manipulation of the market. Each
of the essential ground of that ideology and enterprise receives a series of targets (gross
does so with a simple clarity that cannot be value output, cost, employment, profits, and
achieved by reference to the many works of so forth) from the planners at their level. The
Marx and Lenin, not to mention Mao Tse- process involves three stages and negotiation
tung. between planners and enterprise managers.
There are many features of China's econ- Initially the planners propose broad targets,
omy that distinguish that economy from oth- and the factory replies with what it thinks it
ers, but several stand out. To begin with, can do to meet these targets. The center (or
 71 COUNTRY PERFORMANCE

TABLE 1 . China's Gross Domestic Product, Selected Years (in billions

of 1957 yuan)

Industrial Value Agricultural

Value Added (A) Services (S)
Year Added (M + )
GDP

1952 73.85
20.79 35.59 17.47
1957 38.16 44.72 21.80 104.68
1962 50.15 35.30 22.84 108.29
1965 72.60 49.10 28.94 150.64
1970 106.79 59.96 37.15 203.90
1974 152.80 67.09 46.35 266.24

TABLE 2. Sectoral Shares in GDP

Services (S)
Year Industry (M + ) Agriculture (A) GDP

1952 .282 .482 1.000

.237 1.000
1957 .365 .427 .208
1962 .463 .326 .211 1.000
1965 .482 .326 .192 1.000
1970 .524 .294 .182 1.000
1974 .574 .252 .174 1.000

province or county) then attempts to achieve Union. But we know that China's economy
consistency between what each enterprise has grown and that the most complicated sec-
can do and sends the final targets back to the tor in both engineering and planning terms,
enterprise. The enterprise is then expected to namely the modern producer goods industry,
guide its actions to surpass those targets. has grown much more rapidly than the other
Agriculture is not subject to as rigorous sectors.
control by plans as industry. Plans exist, of Estimates of Chinese GDP, the shares of
course, and they influence which crops are agriculture, manufacturing, and services and
planted where. But the degree of control is growth rates are presented in Tables 1 to 3.
much looser than in industry, and plans are Different assumptions would lead to slightly
often backed by general controls such as different results; but all the data in these tab-
price manipulation rather than physical tar- les (except services) were derived from offi-
gets enforceable by law. Because agricultural
production teams are cooperatives whose
members divide their net income among TABLE 3. Chinese Growth Rates, 1957 Prices
themselves, prices have a significant influ- (in percentages)
ence on decisions. Planning in agriculture, as Industry
a result, is much less complex than that in Period 7.2
GDP
Agriculture
industry.
1952-1957 12.9 4.7
A second key feature of China's economic 1957-1965 1.2
4.1 4.7
development is that its highly complex sys- 8.4
8.0
1965-1970
tem works reasonably well. We do not yet 2.8 6.2
1970-1974 9.4
9.5 6.9
have enough data to say whether China's 1952-1974 3.4
economic system is more or less efficient than 1957-1974 8.5 6.2
5.6
that of, say, India, Japan, or the Soviet 2.4
 72 THE DEVELOPMENT RECORD

cial Chinese publications or releases, and sharply in the 1960s; despite this rise there is
there is an increasing degree of consensus little doubt that China's investment rate
among people working on the Chinese econ- (gross domestic capital formation as a per-
omy that the official data, if used with care, centage ofGDP) has increased to a very high
are the best available. level. Precise estimates are not available, but
indirect methods of estimation indicate that
Several distinguishing features of China's
economy are immediately apparent from the China's current rate of gross domestic capital
estimates in Tables 1 to 3. First, China's formation has reached or passed 25 percent
gross domestic product has grown at a rate of of GDP and is still rising.
about 6 percent a year for over two decades. Even more dramatic has been China's abil-
Different estimation procedures might bring ity to mobilize its labor force for develop-
this rate down to 5 percent, but in either case ment. Although there is a continuing debate
the rate is well above the percentage increase in the economics literature whether "surplus
in China's population, which is about 2 per- labor" exists anywhere in the world, few
cent and falling. Not many nations have done would dispute that the adult population of
as well for such a sustained period. A number most less developed nations is not fully em-
have grown at faster rates in recent years, but ployed throughout the year. Within most cit-
only a few have matched China's longevity. ies in the less developed world there are large
China's GDP, as a result, has risen to numbers of overtly unemployed. In rural
nearly four times the level of 1952 when the areas unemployment is disguised and sea-
post- 1949 development effort began. Per cap- sonal, but extensive.
ita GDP has more than doubled. In China, in contrast, both urban and rural
Another aspect of this development in populations are hard at work throughout the
China has been the degree to which growth entire year. Much of this activity is in occu-
has been concentrated in the industrial sec- pations of very low productivity. Peasants
tor. The share of industry (and transport) has carrying dirt from a nearby mountainside to
risen steadily from about a quarter of GDP build a dam or a small addition to the culti-
to over half. Agriculture, in contrast, has vated acreage are doing back-breaking labor
managed to stay only a bit ahead of popula- in exchange for often only modest increases
tion growth, and its share has declined stead- in farm output. Many rural people, one sus-
ily as a result. In the course of economic pects, would choose leisure time if they were
growth the share of industry always rises and free to do so. Whether voluntarily or not,
that of agriculture falls, but the pace at China has succeeded in mobilizing billions of
which these changes have occurred in China man-days of labor from a labor pool that in
is more pronounced than in most other devel- most less developed nations would have re-
oping nations. Within industry, the producer mained idle.
goods sector has grown more rapidly than the China has also fully mobilized its agricul-
consumer goods sector. Machine building, in tural land resources. There is no area in
particular, has increased at a very rapid rate. China comparable to parts of Latin America
A third distinguishing characteristic of where land that could be used to grow crops
Chinese development has been the ability of is slated for low productivity activities such
the nation to mobilize fully its resources for as grazing. . . .
growth. The rising share of the producer A fourth characteristic of Chinese eco-
goods sector is one aspect of this phenome- nomic development, the one that currently
non.- As national product has increased, receives the most attention, is the Chinese ef-
China has succeeded in holding the rise in fort to narrow the inequalities in income in
personal consumption to modest levels. Gov- both urban and rural areas. Great attention,
ernment consumption, mainly in the form of however, has not led to much effort to mea-
military expenditures necessitated by the sure the degree of inequality that exists in
1960 break with the Soviet Union, rose China today. The one major work on the sub-
 73 COUNTRY PERFORMANCE

TABLE 4. Changes in the Distribution of Rural Income, 1930s to 1952

-2.8
Farm Income Total Income -7.0
Percentag -2.6
Percentage Share -4.4 s Share Change
Change
-7.6
Income Category 1930s 1952 in Share
-2.5 1930s 1952 in Share

26.0 21.6 24.4 21.6

Top 10%
Top 20% 42.7 35.1 42.0 35.0
2nd 20% 23.8 21.3 23.9 21.3
3rd 20% 16.3 17.5 14.9 17.4
14.8 + 1.2 15.0 + 2.5
4th 20% 11.4 13.2
11.3 + 3.4 + 1.8
Bottom 20% 5.8 + 5.5 6.0 11.3
Bottom 10% 1.8 5.1 + 5.3
2.5
+ 3.3 5.1 + 2.6
Doctoral
Source: C. Robert Roll,"The Distribution of Rural Income in China: A Comparison of the 1930's and the 1950's,
dissertation, Harvard University, 1974, p. 76.

ject, on rural incomes only, is that of C. Rob- equality in rural areas. Income within the
ert Roll. His estimates, presented in Table 4, team is determined by the number of work
reveal that the land reform of the early 1950s points earned by family members; thus a
during which landlordism was effectively family with more and better workers has a
eliminated from rural China led to a substan- higher income. Far more important sources
tial redistribution of income. The incomes of of inequality are the remaining differences
the top 20 percent of the population were re- between teams in the amount and the quality
duced by nearly one-fifth, and those of the of their land. Land quality, of course, is de-
bottom 40 percent were raised by roughly 50 termined not only by the properties of the soil
percent. but also by proximity to roads, electricity,
The redistribution of Chinese income did urban markets, and the like.
not stop in 1952, but was given another boost How large are these remaining sources of
in the cooperative movement of 1955-56. A*. inequality, and have they been reduced over
that time all the peasants in a given co-op the years? Unfortunately, only speculation is
(which was frequently coterminus with a vil- possible because no serious quantitative
lage) pooled their land, thus eliminating in- study of the subject has yet been made. As
come inequalities due to differences in the Roll demonstrates, there is little doubt that
size and quality of individual landholdings remaining between-region differences in in-
within the cooperative. In 1958-59 coopera- come are substantial.
tives were pooled into communes, thereby The situation in urban areas is somewhat
temporarily eliminating inequality caused by similar. A major redistribution of income and
differences in landholdings between cooper- assets took place in the 1950s through the so-
atives within a single commune. The transfer cialization ofindustry and commerce and the
down to the production team as the basic ac- elimination of most property incomes. The
counting unit, however, brought the situation closed nature of Chinese society, notably the
back to that of 1957. In fact, the average pro- lack of free emigration, meant that China
duction team today is much smaller than the could keep highly skilled personnel (other
old cooperative, but significant amounts of than those who had already fled) without
income are generated at the brigade and paying salaries competitive with those in the
commune level as well. West. The eight-grade wage system deter-
There are two other major sources of in- mined the incomes of most workers and em-
 74 THE DEVELOPMENT RECORD

ployees. The eight-grade system was similar the easiest to document. Estimates of China's
throughout the country and between regions balance of trade are presented in Table 5.
except for minor variations due to cost of liv- The balance of trade is an imperfect guide to
ing differentials, hardship allowances, and capital flows into and out of a country, but in
the like.
China's case a considerable body of addi-
There have been few changes in the nature tional information supports the view that
of this system since 1956. The roughly three- China's balance of trade reflects in a rough
to-one differential between grades 1 and 8 way the direction and magnitude of Chinese
has been altered little if at all. Greater external capital movements.
worker equality since 1956, therefore, could Beginning modestly in 1973 and in a major
only have come about through promotions way in 1974, China borrowed abroad on an
that significantly changed the distribution of intermediate-term basis (up to five years) to
workers in the grades, but we know next to finance a rapid expansion in imports of new
nothing about any such changes. There has plants in chemicals, steel, and other key sec-
also been some rise in the free services pro- tors. Up to the mid-1970s it is fair to say that
vided to the workers (health, housing, and so China depended on foreign financial aid on
forth), and these tend to be provided on the
basis of need.
Technicians and cadres are paid according TABLE 5. China's Foreign Trade, 1950-1974
-340 (in
millions of U.S. dollars) -1
to scales other than the eight-grade system. 40
Such wages can reach levels ten times that of -215
Year Trade Balance
-230
Exports Imports
the bottom of the eight-grade system. There
is some evidence of creeping equality in these 1950 620 590 -23085
780
incomes, however. The wages of older tech- 1951 -70
1952 875 1,120
nicians are not being lowered, but as they re-
1953 1,015
tire, they are replaced by younger people who
are paid much lower salaries. If this process 1954 1,040 1,255 -5
1955 1,060 1,290
continues, one suspects that within a decade
1956 1,375 1,660 150
or two few people will be earning more than 1,635 1,485 175
1957 115
twice the top of the eight-grade system. 1958 1,615 1,440 170
Finally, there is the difference in income 1959 1,940 1,825
between urban and rural areas. Work by 1960 2,230 2,060
Roll, however, suggests that this differential 1961 1,960 2,030 35
may not have been very great in the 1950s 1962 1,525 1,490 375
when allowance is made for differences in the 1963 1,525 1,150 370
1964 1,570 1,200 280
cost of living. Whatever the precise magni- -190
190-80
1965 1,750 1,470
tude of the disparity, rising farm purchase 2,035
1966 1,845 175
prices have contributed to an increase in av- 1967 2,210 2,035
erage rural incomes while average wages 1968 1,945 125
1,950
have changed little. The greater provision of 1969 1,945 1,820 -200
975
urban services, on the other hand, may tend 1970 2,030 1,830
to widen the gap, although recent efforts to 1971
2,050 2,240 110
promote rural health work in the opposite 1972 2,415 2,305 250
direction. 1973 3,085 2,835
4.975
4,895
The fifth and final distinguishing charac-
teristic ofChinese economic development to 6,515 7,490
be discussed here is the Chinese effort to min- Source: Nai-Ruenn Chen, "China's Foreign Trade, 1950-
74," China: A Reassessment of the Economy, Washington,
" Congress,
imize the nation's dependence on external as- 1974U.S.
D.C., Joint Economic Committee, 1975, p.
645."Preliminary.
sistance and foreign trade. China's limited
dependence on external financial assistance is
 75 COUNTRY PERFORMANCE

either concessionary or commercial terms to From the beginning the People's Republic
a lesser degree than any other less developed has cut all nonessentials from its import bill.
country that has achieved a sustained period In the 1950s over 90 percent of all imports
of economic growth. Nations such as Burma were producer goods, and these provided
may also have eschewed foreign assistance much of the key equipment for the first and
for long periods, but they have failed to grow second five-year plans. In the 1960s the share
as well. In per capita terms, aid to China of grain in imports rose to significant levels,
never passed $2 in total and averaged under but the purpose of these imports was to re-
$0.50 a year throughout the early 1950s and lieve pressure on and hence encourage the
zero thereafter. grain surplus areas to produce. In the 1970s,
An analysis of the degree of China's de- with the accelerated investment program in
pendence on foreign trade is far more com- industry, there has been a renewed surge of
plex than an appraisal of the role of external complete plant imports.
financial assistance. The appropriate meth- There is no question that China in 1976 is
odology would involve estimating China's far less dependent on imported producer
GNP in the absence of foreign trade or some
goods than it was in the 1950s. China's ma-
substantial part of that trade. Data limita- chine-building capacity now provides most
tions make such an analysis impossible. One domestic requirements, whereas in the early
is left, therefore, with the task of attempting 1950s virtually all machinery had to be im-
to judge whether China's ratio of foreign ported. Yet it is equally clear that where cer-
trade to GNP is unusually low and whether tain types of advanced technology are in-
Chinese imports appear to consist of items es- volved (petro-chemicals, and so forth) China
sential to the proper functioning of the econ- must still import to expand capacity. In many
omy. Even these inadequate measures can be respects, therefore, China's current depen-
dealt with only superficially here. dence on imports is more like that of nations
Estimating China's ratio of foreign trade with much higher per capita incomes than
(exports plus imports) to GDP is difficult be- the "typical" less developed country. By rap-
cause the Chinese trade data available out- idly expanding its producer goods sector,
side China are calculated in U.S. dollars China can now continue to expand in many
while Chinese GDP figures are in yuan. directions with domestic equipment. Few if
Either the former must be converted into any other nations with per capita incomes
yuan or the latter into dollars, but at what below $300 (or $1000) can make that state-
exchange rate? At the official 1974 exchange ment. On the other hand, China is far more
rate which floated around 1.95 yuan to the dependent on imports of high technology
dollar, the 1974 trade ratio was 11 percent. than the industrially most advanced nations:
Similar procedures for 1970 lead to an esti- the United States, the Soviet Union, Japan,
mate for that year of only 6 percent. But it is West Germany.
unclear whether the higher percentage in Thus it is incorrect to say that China's eco-
1974 reflects a real increase in the share of nomic development has proceeded solely on
trade or simply a rise in the prices of traded the basis of China's own efforts. It is right to
goods that was more rapid than the revalua- say that China has provided a far higher pro-
tion of the yuan relative to the dollar. Until portion of its own development investments
an updated purchasing power parity compar- (in both financial and physical terms) than
ison is made, we shall have to defer judg- most other less developed nations in the
ment. For purposes of this analysis a ratio of world today.
10 percent is used as indicative of China's A low foreign trade ratio and a high in-
"typical" trade level. vestment rate also go a long way toward ex-
Have the commodities imported by China plaining China's overall development strat-
played a key role in China's economic devel- egy. In brief, a nation that cannot rely
opment? Here the answer is a clear yes. heavily on imports is forced to follow some-
 76 THE DEVELOPMENT RECORD

thing approximating a (supply-side) bal- hectare, particularly in areas suitable for

anced growth strategy. To begin with, there rice, are comparable to the highest yields
must be balance between agriculture and in- achieved anywhere in the world, including
dustry. Single-minded pursuit of growth in Japan. Further breakthroughs depend cru-
one sector to the exclusion of the other cially on scientific advances such as the de-
quickly generates import requirements far velopment of new plant varieties and on
beyond the nation's capacity to pay. Within large-scale investment in water conservancy
industry as well there must be a degree of in north China. Food output can be expanded
balance. If China obtained all its mid-1970s by increases in inputs of labor and items such
steel needs of 25 million plus tons from im- as chemical fertilizer, but the return to these
ports, that item alone would cost China inputs is falling and in many cases is already
around U.S. $6-$7 billion, an amount equal very low. Expansion on a major scale will
to its entire import bill in 1974. come instead through methods that take time
Thus the principal policy options open to (scientific breakthroughs) and are extremely
Chinese planners were between a high and a expensive
River). (for example, control of the Yellow
low rate of investment and, within the con-
sumption sector, between government If the analysis that lies behind these state-
(mainly military) and private consumer ments is correct, then China's rapid indus-
goods. Given these decisions and assuming trial development and much slower develop-
some degree of consumer sovereignty over ment in agriculture, one of the key
private consumption goods, the remaining characteristics of China's post- 1949 eco-
options available to planners were severely nomic performance, can be explained in part
limited. If one removes the assumption of by China's factor endowment. In the absence
consumer sovereignty over the consumer of a severe shortage of land in both of the
goods mix, the range of options increases senses mentioned before, current high rates
slightly although much less so in a poor coun- of investment in Chinese agriculture would
try than in a rich one. Chinese planners, for have achieved much more dramatic results.
example, could raise the rate of growth in ag- Whether Chinese planners would have chos-
riculture from say 2.5 percent to 3.5 percent en to achieve higher agricultural growth
a year, but they could not let it fall much rates or alternatively to shift investment
below 2.5 percent for long without causing funds in agriculture to other sectors cannot
major economic and political difficulties. be known. But with a different factor endow-
Saying that Chinese planners' options are ment they would at least have had that
limited given a low trade ratio and a high chance.
rate of investment is not the same thing as China's revolution created political condi-
saying that China's development strategy tions that made possible major shifts in eco-
will be identical to that of any other nation nomic policy. With the poor in command in
with comparable trade and investment ratios. rural areas, cooperatives and communes
China's factor endowment also plays an im- could be formed with little disruption of the
portant role, particularly in determining the economy, and these collective units could in
relative performances of industry and agri- turn mobilize labor and resources to begin
culture. The most significant part of China's transforming the countryside. In the cities
factor endowment is its shortage of cultivable revolutionary political change also eased the
land. China is a land-short nation in two processes of replacing privately controlled in-
basic ways. First, there has been little or no dustry and commerce with state-owned firms
net expansion in China's cultivated acreage and of introducing rationing of necessities.
since the 1950s largely because the costs of But China's revolution was only part of the
developing currently uncultivated land are
reason for the post- 1949 government's ability
prohibitive. Second, Chinese grain yields per to choose from a wide range of economic pol-
 77 COUNTRY PERFORMANCE

icy options. The political triumph of the poor On the other hand, China's size and land-
short factor endowment dictated certain
had little to do with China's ability to intro-
duce a centralized system of planning and strategies by limiting other kinds of options.
management on the Soviet model. Of central A major shift in investment funds toward ag-
importance was that China had the personnel riculture inthe early 1960s managed to keep
to administer such a system. Those people farm output growing faster than population,
were available because of a Chinese heritage but not fast enough to alter sectoral devel-
of emphasis on education and experience opment with a pronounced industrial pro-
with a complex commercial economy. ducer goods emphasis. Similarly, limited de-
China's authoritarian heritage, reinforced by pendence on foreign trade has been turned
the power of the Chinese Communist party, into a political virtue, but China's size has
also opened key options to policymakers, no- more to do with this degree of independence
tably in the area of increased taxes and the than any set of deliberate policies.
potential for a high rate of investment.

Comment
In considering various country studies, we should ask whether the success stories (or fail-
ures) are the result of favorable (or unfavorable) initial conditions, internal factors, or exter-
nal factors. It is also instructive to ask whether the specific institutions and programs of the
success stories are transferable to other developing countries.
For discussion of the relevance of China's experience for other countries, see Robert F.
Dernberger and Francoise Le Gall, "Is the Chinese Model Transferable?" in Robert F. Dern-
berger (ed.), China's Development Experience in Comparative Perspective (1980). Also in-
structive are D. B. Keesing, "Economic Lessons from China," Journal of Development Eco-
nomics (March 1975); Joint Economic Committee, Congress of the United States, Chinese
Economy Post-Mao: A Compendium of Papers (1978); Doak Barnett, China's Economy in
Global Perspective (1981); Chu-yuan Cheng, China's Economic Development (1982).

EXHIBIT 1.8. Poverty and Human Development in China

Annual Life
Net Primary Expectancy
Population Birth
GNP per Growth Rate Adult
Literacy School Enrollment
at!
Person (percentage) (percentage)
Country and Country (dollars) (percentage)
Group 1979 1975 or 1977 (years)1979
1970-79 1976

China 260 1.9 [36]

Sri Lanka 230 85 62 66
1.7 [55] 64
India 190 2.1 36 52
Indonesia 370 2.3 62 66
64 [38] 53
1950*
[35]
Low-income countries 210 2.3 56
72 [37]
Middle-income countries 2.4 39 71 [48] 61
1,420 51
Industrial countries 99 74
9,440 0.7
94
Source: World Bank Development Report, 1981, p. 101. 66* 67
"Most 1950 data are estimated.
*1979. 93*
 78 THE DEVELOPMENT RECORD

I.C.6. Poverty Elimination and

Entitlements*

Each economic and political system produces factors as, say, guaranteeing better terms of
a set of entitlement relations governing who trade for poor peasants, or ensuring employ-
can have what in that system.1 Ultimately ment at a living wage, or providing social se-
poverty removal is a matter of entitlement curity protections.
raising. There are both aggregative and dis- The method of removing poverty used in
tributive influences on the determination of South Korea and Taiwan is one of guaran-
entitlements, and while balanced growth teeing employment at a tolerable wage, and
might work through the former, institutional this has been possible by a very fast expan-
arrangements affecting the "division of the sion of these economies using labor-absorb-
ing production processes. In contrast, the
cake" work through the latter. The precise
characterization of entitlement relations will, level of unemployment in Sri Lanka is high,
of course, vary with the nature of the econ- and the wages, especially of estate workers,
omy and society,2 but some distinguished quite low. There the guarantee of basic enti-
cases are easy to identify. tlements has not come through the market,
For a market economy, the determining but outside it, in the form of a direct right
variables of entitlements can be broadly split against the state.
into (i) ownership vector (e.g., the land, cap- The constrast is not, of course, a pure one.
ital or labour power which a person owns), Some social security provisions do exist in
and (ii) exchange entitlement mapping (e.g., Taiwan, and to a lesser extent in South
for each ownership bundle the set of alter- Korea. Similarly, in Sri Lanka even the pre-
native bundles of commodities any one of tax "gross" distribution of income is not as
which the person can acquire through pro- unequal as in some developing countries. But
duction or trade). The ownership vector for a the crucial and distinctive features of entitle-
particular group may be enhanced either ment enhancement in the two types of econ-
through overall increase (e.g., capital accu- omies are different. In Taiwan and South
mulation), or through asset redistribution Korea the social services only rather margin-
(e.g., land reform). These influences on enti- ally improve what is provided by a buoyant
tlements through ownership changes are labor market. Sri Lanka's rather "middling"
rather more palpable and tend to be more performance in market-based entitlements is
often discussed. Entitlement raising through vastly improved by public policy of distribu-
improving the exchange entitlement mapping tion and social service.
is a bit more complex to analyze, but may be More importantly, there is an important
no less important. It includes such diverse similarity between the two strategies, which
superficially might look poles apart. Ulti-
mately, poverty removal must come to grips
*A. K. Sen, "Levels of Poverty: Policy and with the issue of entitlement guarantees. The
Change," World Bank Staff Report No. 401, July two strategies differ in the means of achiev-
1980, pp. 53-65. Reprinted by permission.
ing this guarantee. While one relies on the
'A. K. Sen, "Real National Income," Review of
Economic Studies, Vol. 43, 1976; "Starvation and successfully fostered growth and the dyna-
Exchange Entitlement: A General Approach and Its mism of the encouraged labour market, the
Application to the Great Bengal Famine," Cam- other gives the government a more direct role
bridge Journal of Economics, Vol. 1, 1977; "Fam- as a provider of provisions. The similarities
ines," World Development, Vol. 8, 1980a; and Pov-
erty and Famines: An Essay on Entitlement and can be seen more clearly if we look at the dis-
Deprivation, Clarendon Press, Oxford, 1981. tributional pattern through which these
means work to guarantee entitlement.
A 2Sen,
and B.1980b, chapters 1 and 5-10 and appendixes
It is not easy to obtain internationally com-
 79 COUNTRY PERFORMANCE

TABLE 1 . Comparisons of Relative Equalities of Income Distribution

Relative Relative Relative Relative

Share of Share of Share of Share of Borda
Country Bottom 20% Bottom 40% Bottom 60% Bottom 80% Score

Taiwan 4.0
8.7 21.9 38.5 60.8
1971
Sri Lanka 7.5 56.6
19.2 10.0
1969-70 34.7
Yugoslavia 6.5 18.4 11.0
1973 36.0 60.0

India 17.0
6.7 17.2 31.5 51.1
1964-65
South Korea
5.7 16.9 32.3 18.0
1976 54.7

Argentina 4.4 14.1 28.2 49.7 25.5

1970
Chile
4.4 13.4 27.2 48.6 27.5
1968
Philippines 11.9 25.1 46.1 34.0
1970-71 3.7
Costa Rica 12.0 45.2
3.3 25.3 37.5
1971
Turkey 43.4 42.0
3.4 11.4 23.9
1973
Mexico 45.5
2.9 10.3 23.5 45.5
1977
Venezuela 10.3 46.0 45.5
3.0 23.2
1970
Malaysia 43.5 46.5
3.3 10.6 22.8
1970
Peru 7.0 59.5
1.9 18.0 39.0
1972
Honduras 60.0
2.3 7.3 15.3 32.2
1967
Brazil 7.0
2.0 16.4 60.5
1972 33.4

Source: Calculated from table 24 of the World Development Report, 1979.

parable income distribution data, but to get The World Development Report 1979 gives
some broad idea Table 1 presents income dis- the data in the form of shares of each succes-
tributional data for 16 countries taken from sive quintile. For ready comparison of pa-
the World Development Report 1979. (These rameters ofinequality, these have been pre-
are all the countries with income per head sented here in the form of cumulative
less than $3,000 for which such data are aggregates, i.e., the shares of the bottom 20
given.) The incomes in question are of "total percent, bottom 40 percent, bottom 60 per-
disposable household income," net of taxes. cent and bottom 80 percent, respectively.
 80

The most satisfactory method of compari- Poverty removal and related features, in-
son with these data is the so-called "Lorenz cluding longevity enhancement, are ulti-
dominance." Country A Lorenz-dominates mately dependent on a wide distribution of ef-
country B in terms of equality of income dis- fective entitlements, and this — for any given
tribution ifcompared with country B, coun- level of per capita income — would tend to be
try A has at least as high an income share for reflected in the low level of inequality in the
each x percent poorest population, and a distribution of income. At this immediate
strictly higher share for some x percent poor- level of explanation, in terms of entitlements
est population. When Lorenz dominance and income distributions, Yugoslavia, Sri
holds between two distributions of income, Lanka, Taiwan and South Korea have much
many interpretable conclusions can be in common despite their widely different po-
drawn, and in general this may be regarded litical systems and economic strategies. At
as the most satisfactory method of comparing the remoter level the explanations differ; Tai-
the income distribution of two countries. On wan and South Korea rely on high employ-
the other hand, this is a demanding criterion, ment at a reasonable wage as an instrument,
and very often neither country Lorenz-dom- while Sri Lanka leans on security outside the
inates the other in a pairwise comparison. market mechanism. At a still remoter level,
Lorenz domination yields a partial order- the former pair of countries have profited
ing of "clear" rankings, but given its incom- from the strategy of export expansion, while
pleteness, some less demanding method of Sri Lanka has not used such a strategy, nor
ranking has to be chosen to supplement Lor- is Yugoslav planning especially export ori-
enz comparison if completeness is sought. A ented. Learning from the experience of Tai-
simple method of doing this — surprisingly wan or Korea need not take the form of
little used in empirical comparisons — is to seeing virtues in export expansion irrespec-
follow Borda's method of rank-order scoring, tive of the nature of the country, and indeed
giving points equal to the rank value of each the remoter we move in the chain of expla-
country in each criterion of comparative nation, the more conditional, in general, the
ranking. This produces a complete ordering explanation becomes.
based on all the criteria taken together in The specific nature of the economic prob-
terms of lowness of the sum of ranks (Borda lems faced by these countries to which their
scores). economic strategies responded has often
Both the Lorenz and Borda methods of tended to be ignored. But on these specifici-
ranking the 16 countries in Table 1 have been ties, the remoter explanations may substan-
used, and the results are represented in the tially depend. Consider, for example, the
form of two Hasse diagrams (with descend- strategy of relatively free trade with absent
ing lines representing superiority of the or refunded import duties, used by Taiwan.
higher placed country vis-a-vis the lower). The necessity of such a policy for export ex-
The partial ordering of Lorenz dominance pansion in the Taiwanese context has been
given by Hasse Figure 1 is subsumed by the explained admirably by Maurice Scott
complete ordering of Borda ranking in Hasse
Figure 2. Interestingly enough, the countries Taiwan has very few mineral resources and strictly
which were commended for their perfor- (1979):3agricultural resources. It could only pros-
limited
mance in poverty removal and longevity ex- per, and has in fact prospered, by exporting what
tension come very much at the top of these effectively are labor and capital services. These
rankings. Countries which were separated services, however, have mainly taken the form of
out for disappointing performance in terms of adding value to imported materials rather than
poverty removal commensurate with their in- being services pure and simple (as are tourism,
come, viz., Brazil, Peru and Mexico, also fig- 3M. Scott, "Foreign Trade," in W. Galenson (ed.),
ure here towards the bottom of the two Economic Growth and Structural Change in Taiwan,
rankings. Ithaca, Cornell University Press, 1979.
 81 COUNTRY PERFORMANCE

Taiwan
1971

Yugoslavia
1973

South Korea
1976

Costa Rica
1971

Malaysia
1970

Brazil
1972

FIGURE 1 . Hasse diagram showing partial order of Lorenz dominance.

banking, transportation, and the like). Thus the all potential wages, fuel costs, or payment for
most important single condition for an export other bought-in services (advertising and other
"take-off" was access by manufacturers to im- selling costs, rent of buildings and land, and so
ported materials, components, and capital equip- forth), and a much smaller tax, say 10 percent,
ment at world prices. For a typical manufacturer would make many exports totally unprofitable.
in one of Taiwan's export industries . . . materials Under the circumstances the compelling
account for around 65 to 70 percent of the selling
value of output, capital altogether receives a fur- advantages of Taiwan's policy regarding im-
ther 5to 12 percent of selling value, and, of this, ports are clear enough, but these may or may
over half can be attributed to costs of plant ma- not be shared by other countries with differ-
chinery, vehicles, and stocks of materials. Hence, ent economic characteristics (e.g., larger
the total percentage of selling value represented by economies, greater mineral resources, less
the cost of materials or equipment might be skill of the kind needed for export expansion
around 70 to 75 percent or even more for a typical
in the present world markets, etc.). Similar
exporter. If his exports are sold at world prices,
without benefit of any subsidy, he must be able to discriminating analysis can be done with
buy these materials and equipment at or close to other instruments, e.g., high savings rate, so-
world prices if the export is to be profitable. A 40 cial welfare delivery systems, and there is
percent tax on materials and equipment would need for caution in reading lessons from suc-
wipe out, not merely all potential profit, but also cess stories.
 82 THE DEVELOPMENT RECORD

Taiwan
1971
The ways of achieving widespread entitle-
ments differ, and while different economic
strategies have things in common both in
Sri Lanka terms of ultimate consequences as well as in
1969-70 underlying causation, the precise combina-
tion of instruments and policies varies tre-
Yugoslavia mendously with the conditions in different
1973
economies. In a somewhat stylized form Fig-
ure 3brings out such contrasts. Roughly path
India
a tries to capture the causal mechanism in
1964-65
South Korea and Taiwan; path c that in Sri
Lanka; and paths b and c that in Yugoslavia.
South Korea
1976 It is common to all these strategies that low
poverty in the form of high longevity and low
proportion of population below the poverty
Argentina
1970 line are brought about by a widespread dis-
tribution ofentitlements, but beyond this im-
Chile
mediate link there are sharp differences in
1968 the way such entitlement delivery is made.
There is similarity also in the role ultimately
Philippines played by calculated and determined public
1970-71 policy. But while the Korean-Taiwanese
causal mechanism goes through industrial
Costa Rica expansion with exports expanding much
1971 faster, employment-oriented fast growth,
achievements in total income enhancement
Turkey and in income distribution, the Sri Lankan
1973
path goes through social welfare pro-
grammes, public distribution systems,
Mexico and Venezuela achievements in income distribution and in
1977 1970
non-income advantages (e.g., in people hav-
\/ ing better health when the income subsidy
Malaysia comes in the form of food, medical provi-
1970
sions, etc.). The Yugoslav experience fits nei-
ther since it is somewhat similar to the Ko-
Peru
1972
rean-Taiwanese strategy of high growth with
good distribution, but differs from it in not
having export expansion noticeably faster
Honduras than the growth of national product and in
1967
having a more developed social security
Brazil system.
1972 It is only natural that "lessons" will be
sought from the experience of those countries
FIGURE 2. Hasse diagram showing complete order which have been unusually successful in pov-
of Borda scores of relative equality. erty elimination. But the lessons to be drawn
have to concentrate on understanding the
functional roles of various instruments rather
than on blind imitation of the instruments
themselves. Other countries can learn much
from the experience of success stories if the
 Sea

<D

■« z
*-> o .c

&l

ts
expor

ith exp
o o *S
<*-

out
is s
— . CO

ial
a * &
C/l
g£

th
£ £ E

dusgtrr
oo o.

faster

l a
trri
dusg
uch

mnugi
owi

u
3
cr
u

Sou
th

CO o
60
.2 >. £ .3
>>
o
c
EC.
i
o co C
£ O „
O <u C
J X) ^ j
o -r co
V

11
e co *

^ 2
.3 a>>
1 1 1
a -o o

83
 84 THE DEVELOPMENT RECORD

causal mechanisms are fully studied and re- same applies to the possibility of delivering
moter causes distinguished from the less re- provisions through social welfare programs.
mote ones. To what extent, say, fast growth Neither of these circumstantial conditionali-
is possible with export industries taking the ties renders the experience of the successful
lead, is a question that can be answered for countries unproductive for policy formula-
any particular country only by looking at the tion, but blind imitation may well have
details of its economic circumstances. The counter-productive consequences.
Comment
Numerous forecasts have been made for the prospects of developing countries in the
decades ahead. For the decade of the 1980s, most assessments point to lower growth rates,
greater numbers of poor, and scarcer and more expensive food and energy. Together with
much higher populations, the problems of development are expected to become more severe.
For a range of forecasts, see the World Bank's World Development Report 1981; The
Global 2000 Report to the President of the United States; Wassily Leontief , The Future of
the World Economy (1977); Interfutures: A Study of the World Economy sponsored by
OECD; and the Brandt Commission's North/ South: A Program for Survival (1980).
Some of these reports are summarized by Shahid Javed Burki in "The Prospects for the
Developing World: A Review of Recent Forecasts," Finance & Development (March 1981).
Table 1 shows World Bank projections for various country groups in a low and high range
of growth possibilities.

TABLE 1. Past and Projected Growth of GDP, 1960-1995 {average annual percentage change)

1985-95
High
Country Group 1960-73 1973-80 1980-82 Central
1982-85 Low

All developing countries 6.0 1.9 4.7 5.5

4.7 4.4
Low-income 4.9 6.2
4.1 4.5 4.5
Asia 4.6 5.3
5.4
Africa 3.5 1.4 0.5 2.9 3.3 3.9
Middle-income 2.7
4.5
Oil importers 6.3 1.2 4.4
7.0 5.2 5.7
Oil exporters 4.0 5.3 6.9
5.8
3.7 1.7 5.7
Industrial countries 5.1 2.5 0.4 2.5 5.0
3.0
3.7
Source: World Bank, World Development Report, 1983, p. 27.
 I. D. LEADING ISSUES-NOTE

From the preceding discussion of the devel- development are to be located mainly within
opment record, three overriding questions the economies of the LDCs or whether exter-
emerge: (1) What have been the constraints nal obstacles to development are to be
on the potential rates of development in the blamed. There are also differing conclusions
LDCs? (2) Why have the actual rates of de- on whether the obstacles are best removed by
velopment been below the potential rates in promotion of the market mechanism or by
most of these countries? and (3) What poli- comprehensive central planning.
cies in a strategy of development can now Behind the disappointments in the statis-
raise growth rates, eradicate absolute pov- tical record lie five major constraints that
erty, reduce inequality, and provide more have limited the attainable rate of develop-
productive employment opportunities? These ment in most of the less developed world. One
questions will dominate much of the analysis major constraint stems from the low level of
throughout this book. savings. The inability to mobilize sufficient
The disappointments that have been expe- domestic resources, or to supplement domes-
rienced in the development record now call tic resources with external resources, has
for correction of the errors of omission and continued to inhibit development. An in-
commission in development thought and for crease in the ratio of savings to national in-
a reappraisal of policy options. Contrary to come isstill imperative in most of the LDCs.
the optimistic views of development econo- It is, however, disturbing that the rate of in-
mists in the 1950s and 1960s, the actual de- crease of capital formation has slowed down
velopment performance has been disappoint- in many countries, and it is indeed question-
ing in several respects. Although some able whether in many LDCs the domestic
countries have grown markedly, the poorest savings rate and capital inflow are now suffi-
of the poor countries have grown the least. ciently high to sustain a satisfactory rate of
Population growth has been higher than ex- development. Problems associated with the
pected because of the fall in mortality rates savings constraint are examined in Chapters
and a lag in the decline of fertility rates. As IV and V.
emphasized above, the benefits of growth Another bottleneck relates to the agricul-
have not trickled down, and despite high tural sector. Agriculture constitutes a large
growth rates in per capita income, the num- share of the GDP of developing countries,
bers in absolute poverty have increased in and agricultural commodities account for a
many poor countries. So too have rates of un- considerable part of the value of their total
employment and underemployment risen in exports. Agricultural development is there-
several countries. The problem of inequality fore essential for economic growth, accumu-
in the distribution of income has also inten- lating capital through savings and taxation,
sified insome countries. Furthermore, the ag- and earning foreign exchange. It must also
ricultural sector has lagged, or even stag- support and complement industrial develop-
nated, in many countries. The transfer of ment by providing food for a growing non-
resources from rich to poor countries has agricultural labor force, raw materials for in-
diminished. dustrial production, and a growing home
Many of the readings in subsequent chap- market for domestic manufacturers. The de-
ters consider whether the development pro- velopment record has demonstrated a close
cess is now encountering increasingly association between the agricultural perfor-
severe — or even intractable — difficulties. It mance and the overall growth rate of devel-
will become apparent that there is a diver-
gence of views on whether the limitations to oping countries: in the "high-growth" devel-
oping countries the average rates of increase

85
 86 THE DEVELOPMENT RECORD

in agricultural production have been consid- dividends and profits on private direct foreign
erably higher than in the slower-growth investment have grown, so that in many
countries. countries the import surplus that can be sup-
For the majority of developing countries, ported by external financial resources has
however, agriculture has been a problem sec- also diminished. The oil crisis has had severe
tor. A major restraint on the development balance-of-payments repercussions for many
rate has been the slow growth of their agri- LDCs. Except for some oil-exporting coun-
cultural output in general and of food crops tries and a very few other exceptional devel-
in particular. In almost one-half of the devel- oping countries, the foreign exchange con-
oping countries, food production per head straint has persisted as a severe limitation on
was lower at the end of the 1970s than a dec- a country's development program by making
ade earlier. In many LDCs, the demand for it impossible to fulfill import requirements.
food has grown faster than food production, The problems of mobilizing external re-
and an increasing number of LDCs have be- sources are discussed in Chapter V; the re-
come net importers of food. The neglect of lated notion of a "widening trade gap" — and
the agricultural sector has been pervasive, its policy implications with respect to import
and it underlies problems created by the pop- substitution and export promotion — is con-
ulation explosion, unemployment and under- sidered inChapter VIII.
employment, inequality, and absolute pov- Another major constraint is connected
erty. In Chapter VII, we stress the important with the need for human-resource develop-
role of agricultural improvements in the de- ment. The policy issues relating to the need
velopment process, and we examine the inter- for social development are now critical. It has
actions between the industrial and agricul- become increasingly apparent that an im-
tural sectors.
provement inthe quality of human life can-
As distressing as the lagging performance not be simply awaited as an ultimate objec-
of the agricultural sector, and not unrelated tive of development but must instead be
to it, is the lag in the export sector. The for- viewed as a necessary instrument of devel-
eign exchange constraint — or the deteriora- opment. Even more necessary than an in-
tion in the capacity to import — is an acute crease inthe quantity of productive factors is
limitation on the size of development plans. the need to improve the quality of people as
The ratio of value of exports from LDCs to economic agents. If development is growth
total world exports has declined. At the same plus change, and change is social and cul-
time, development programs have stimulated tural as well as economic, then the qualitative
the demand for imports. It is also contended dimensions of development become ex-
that for structural reasons the demand for
tremely significant in terms of human-re-
imports rises more rapidly than national source development. Without such change,
product and also exceeds the rate of increase the process of development will not become
in exports. The import demand by LDCs has self-sustaining.
thus risen more than their capacity to import This qualitative change requires a greater
based on export earnings. And since the early emphasis on investment in human capital and
1970s, the stagnation of the world economy on measures to modify social and cultural
and the slowing down of world trade have values. Recognizing that the problem of con-
contrasted with the buoyancy of overseas trolling population growth has now reached
markets during the 1950s and 1960s. A new serious dimensions through most of the
export pessimism has set in. Compounding underdeveloped world, many would also
the difficulties, the net inflow of foreign cap- argue that population control policies are es-
ital has slowed down, debt servicing of am- sential tofoster a rise in the standard of liv-
ortization and interest on public foreign debt ing in many poor countries. Unless this is
has risen markedly, and income payments of done, it will be all the more difficult to im-
 87 LEADING ISSUES— NOTE

prove the quality of the population, and the tradeoffs that involve growth, employment,
potential for development will not be and equity. Is there a conflict between in-
realized. creasing output and increasing employment?
Finally, another lesson of experience is the Given that capital is a scarce resource, is the
need to overcome organizational, staffing, type of production that economizes on the use
and logistical bottlenecks if public policy of capital per unit of output consistent with
measures are to be effectively implemented. maximizing employment? Or does output
The quality of national economic manage- have to be sacrificed in providing more em-
ment has become of crucial importance. On ployment? The conflict lies not so much with
the basis of the success stories and failures in the relationship between current output and
development experience, it can be said that a current employment, but with the relation-
poor country is poor because of inappropriate ship between current employment and the
policies. In developing countries, the pursuit growth rate of employment and output in the
of appropriate policies, stability of govern- future.
ment, strong commitment to development Another troublesome tradeoff is between
goals, and capacity for efficient public man- output and equity. Does social justice have to
agement are essential even if the other con- be sacrificed for greater output? Or is it pos-
straints on the attainable rate of development sible to achieve what might be termed "effi-
are relaxed. Only then can the actual rate of cient equity" — that is, an improvement in
development approach more closely the po- equity together with an increase in output?
tentially attainable rate. Does this also hold in intertemporal terms?
Although we do not minimize the signifi- In other words, is it possible to have both
cance of noneconomic factors in the process greater equity and greater output in the pres-
of modernization, we emphasize in this book ent period and in future periods as well? Or
the strategic economic policy issues con- is it more probable that a country can have
nected with a relaxation of the savings, agri- greater equity and greater output in the fu-
cultural, foreign exchange, human-resource, ture only if it endures greater inequality in
and organizational constraints. the present?
These constraints — multiple in number Various perspectives on the development
and both internal and external to the devel- record and development theories, to which we
oping country's economy — aggravate the now turn, may begin to illuminate these
problem of development. The problem is all issues.
the more complicated by the question of
CHAPTER II

Development
Perspectives

Against the background of the development record, we may now consider the interplay be-
tween experience and changes in hypotheses about the development process. In the early post-
war period of development planning, a development theory was only too often misread as if
it provided a recipe for successful development, and access to the teachings of modern eco-
nomics was thought to bestow some new magical quality. Since the disillusioning experience
of postwar development efforts, however, a reaction has set in against such a naive approach.
In the LDCs, modern economics is losing its mystique and policymakers are seeking more
relevant analyses. If we emphasize the constraints on development, we must ask just what is
the relevance of economic theory and history in understanding how these constraints might
be relaxed.
It might be thought that we must depart radically from "Western economics" and intro-
duce a completely different system of economic thought in order to reach a fuller understand-
ing of development problems. After all, economic analysis has matured in the more advanced
industrial nations, and the problems that have preoccupied economists' attention in these
countries might be considered to have little pertinence for less developed economies. In this
vein, one could argue for new approaches in economic theory to fit the problems and interests
of poor countries.
The evolution of development economics has not, however, proceeded in this direction. In-
stead of becoming a new subdiscipline, quite distinct from traditional economics, theoretical
foundations of development economics have increasingly relied on the fundamental principles
of mainstream economics— the neoclassical economics of resource allocation, international
trade theory, and the macroeconomics of monetary and fiscal policies. But the application of
these principles and the derivation of policy proposals have had to change as the conditions
of developing countries altered.
Changes in the perception of the development process can be traced through the selections
in this chapter.

89
 90 DEVELOPMENT PERSPECTIVES

If we are to appreciate fully the variety, complexity, and pervasiveness of development

problems, we must be aware of their historical dimension. Historical perspective is one of the
best safeguards against taking a superficial view of these problems. Section II. A, therefore,
considers the relevance of historical experience in contributing to our understanding of the
sources of development. It does so, in part, by focusing on Professor Rostow's provocative
application of a stage approach to the course of development. As a grand historical thesis,
Rostow's analysis of stages of growth had especially strong appeal during the 1960s in the
less developed countries. But it received substantial criticism, and some of these critiques
deserve equal attention for their concern with those historical factors neglected by Rostow.
Basic to Rostow's original analysis, in Stages of Economic Growth, was his sketch of a
dynamic theory of production that emphasized the composition of investment and the growth
of particular sectors in the economy. This theory of production allowed Rostow to identify
certain "leading sectors," the growth of which is thought to be instrumental in propelling the
economy forward. Rostow also indicated that a sequence of optimum patterns of investment
can be postulated from a set of optimum sectoral paths determined by the level of income and
population, by technology, by the quality of entrepreneurship, and by the empirical fact that
deceleration is the normal optimum path of each sector. The actual course of investment,
however, generally differs from these optima inasmuch as they are influenced not only by
private choices, but also by the politics of governments and the impact of wars. Nonetheless,
Rostow believes that, at any period of time, leading sectors can be identified, and the changing
sequence of leading sectors plays an important role in Rostow's stages of growth. The se-
quence of stages suggests, in turn, that a succession of strategic choices is open to societies,
and that political and social decisions about the allocation of resources are made in terms
beyond the usual market processes. Of Rostow's five stages of growth, the most relevant for
poor countries at present are the first three — the "traditional society," the emergence of the
"preconditions for take-off," and the "take-off."
Other selections in section II. A criticize Rostow's stage theory and suggest different view-
points for attempting to draw some conclusions regarding the correspondence and variation
between past and present efforts at development. The Note on "Future Development in His-
torical Perspective" (II.A.4) places special emphasis on some of the differences between past
and present cases of development — differences that give rise to problems that will receive
extensive consideration in subsequent chapters.
The changes in development thought over the past two or three decades are brought out by
the selections in section II. B. A general survey of contrasting perceptions of development is
offered by Streeten (II.B.l) who indicates that from an initial concentration on growth, de-
velopment economists moved to stressing the creation of employment, and then to the redis-
tribution ofbenefits to the poor, and more recently to an emphasis on basic human needs. The
evolution of the methodology of structural analysis is discussed by Chenery (II. B. 2). Some of
the economist's advances in understanding of the development process are summarized by
Ranis (II. B. 3). Turning to the more specific practice of development modeling, Tinbergen
summarizes the state of the art (II. B. 4).
In section II. C, we consider some alternative perceptions outside the mainstream. Among
alternative analyses, the Marxist view (II.C.l) and "dependencia" school of thought (II. C. 2)
are most prominent.
The chapter ends with a Note that examines the relevance and realism of an "Economics
for Development" to be used by the development practitioner (II. D).
 II.A. HISTORICAL
PERSPECTIVES
II.A. 1 . Rostow and Marx-Note

It has always been tempting to search for traditional society; the preconditions for
regularities in history, and many writers have take-off; the take-off; the drive to maturity;
adopted a uni-directional view of develop- the age of high mass-consumption.2
ment in terms of some pattern of stages. As The "take-off" is meant to be the central
summarized by Professor Kuznets, "a stage notion in Rostow's schema, and it has re-
theory of long-term economic change implies: ceived the most critical attention. The take-
(1) distinct time segments, characterized by off is interpreted as "a decisive transition in
different sources and patterns of economic a society's history" — a period "when the
change; (2) a specific succession of these seg- scale of productive economic activity reaches
ments, so that b cannot occur before a, or c a critical level and produces changes which
before b\ and (3) a common matrix, in that lead to a massive and progressive structural
the successive segments are stages in one transformation in economies and the societies
broad process — usually one of development of which they are a part, better viewed as
and growth rather than of devolution and
changes in kind than merely in degree." The
shrinkage. Stage theory is most closely asso- take-off is defined "as requiring all three of
ciated with a uni-directional rather than
the following related conditions":
cyclic view of history. In the cyclic view the
1 . a rise in the rate of productive investment
stages are recurrent; in a uni-directional
view, a stage materializes, runs its course, from, say, 5% or less to over 10% of na-
tional income (or net national product);
and never recurs. Even in the process of de-
volution and decline, the return to a level ex- 2. the development of one or more substan-
tial manufacturing sectors, with a high
perienced previously is not viewed as a recur- rate of growth;
rence of the earlier stage."1 3. the existence or quick emergence of a po-
The central question raised by Kuznets is: litical, social, and institutional framework
How can such a simple design be a summary which exploits the impulses to expansion
description or analytic classification of a vast
in the modern sector and the potential ex-
and diverse field of historical change suffi-
ciently plausible to warrant the formulation ternal economy effects of the take-off and
and persistence of many variants? gives to growth an on-going character.3
At one extreme, Adam Smith referred to Of the earlier proponents of stages, only
the sequence of hunting, pastoral, agricul- Marx commands Rostow's explicit attention.
tural, commercial, and manufacturing Indeed, Rostow presents his analysis as an al-
stages. At the other, Karl Marx related He- ternative toMarx's theory of modern history.
gel's thesis, antithesis, and synthesis to the Describing his system as "A Non-Commu-
Marxian stages of feudalism, capitalism, and nist Manifesto," Rostow poses his five stages
socialism. Most recently, Professor Walt of growth against Marx's stages of feudal-
Rostow attempted to generalize "the sweep
of modern economic history" in a set of 2W. W. Rostow, "The Stages of Economic
stages of growth, designated as follows: the
Growth," Economic History Review, August 1959;
The Stages of Economic Growth, Cambridge, 1960;
'Simon Kuznets, "Notes on Stage of Economic The Economics of Take-Off Into Sustained Growth,
London, 1963.
Growth as a System Determinant," in Alexander
Eckstein (ed.), Comparison of Economic Systems, 3Rostow, The Stages of Economic Growth, pp. 36-
Berkeley, 1971, p. 243. 40.

91
 92 DEVELOPMENT PERSPECTIVES

ism, bourgeois capitalism, socialism, and economy — and whether it is capitalist or

communism. not — it is the total procedure by which
We can recognize some broad similarities choices are made."7 Marx's assumption that
between Rostow's analysis and Marx's se- a society's decisions are merely a function
quence. Both are audacious attempts to in- of who owns property is therefore rejected as
terpret the evolution of whole societies, pri- inaccurate; instead it is maintained that "one
marily from an economic perspective; both must look directly at the full mechanism of
are "explorations of the problems and con- choice among alternative policies, including
sequences for whole societies of building the political process — and indeed, the social
compound interest into their habits and insti- and religious processes — as independent
tutions";4 and both recognize that economic arenas for making decisions and choices."8
change has social, political, and cultural The implications of this broader view of
consequences. human motivation become especially signifi-
From other viewpoints, however, there are cant when Rostow's interpretation of post-
fundamental differences. The basic Marxian traditional societies is contrasted with
problems of class conflicts, exploitation, and Marx's account of the post-feudal phase.
inherent stresses within the capitalist process Thus, Rostow concludes that his account of
find no place in Rostow's analysis. Nor does the break-up of traditional societies is
Rostow reduce the complexities of man to a based on the convergence of motives of private
single economic dimension. Rostow recog- profit in the modern sectors with a new sense of
nizes that, in terms of human motivation, affronted nationhood. And other forces play their
many of the most profound economic part as well, for example the simple perception
changes must be viewed as the consequence that children need not die so young or live their
of noneconomic human motives and aspira- lives in illiteracy: a sense of enlarged human hori-
tions. Instead of limiting human behavior to zons, independent of both profit and national dig-
simply an act of maximization, Rostow inter- nity. And when independence or modern nation-
hood is at last attained, there is no simple,
prets net human behavior "as an act of bal- automatic switch to a dominance of the profit mo-
ancing alternative and often conflicting
tive and economic and social progress. On the con-
human objectives in the face of the range of trary there is a searching choice and problem of
choices balance among the three directions policy might
Rostow men allows perceive
for theto be open tofacets
different them."5of
go: external assertion; the further concentration of
human beings, and interprets the total per- power in the centre as opposed to the regions; and
formance ofsocieties as an act of balance in
economic growth.9
the patterns of choice made by individuals
within the framework permitted by the This approach may have more immediate
relevance for the problems now confronting
changing setting of society. Rostow insists
that although his "stages-of-growth are an many underdeveloped countries than Marx's
economic way of looking at whole societies, narrower view that political behavior is de-
pendent on economic advantage, and that the
they in no sense imply that the worlds of pol-
itics, social organization, and of culture are a decisions of capitalist societies are made sim-
mere superstructure built upon and derived ply in terms of the free-market mechanism
and private advantage.
uniquely from the economy."6 On the con- Moreover, as Rostow observes, the Marx-
trary, what most concerns Rostow is how so- ian sequence suffers by basing its categories
cieties go about making their choices and
on only one historical case: the case of the
balances: "the central phenomenon of the British take-off and drive to maturity. Ros-
world of post-traditional societies is not the 8Ibid.
tow reminds us that Marx had presented his
4Ibid.,p. 148. 7Ibid., p. 150.
5Ibid.,p. 149.
6Ibid., p. 2. 9Ibid., p. 152.
 93 HISTORICAL PERSPECTIVES

whole system before any other society except patterns and seen from many perspectives.10
Britain had experienced the take-off , and in- What matters, therefore, is how suggestive
stead of revising his categories so as to be and useful Rostow's pattern is in providing
more applicable to other cases, Marx merely answers to our questions as we attempt to
generalized and projected his interpretation make sense out of the past and make the fu-
of the British case. A concentration on the ture more predictable. This comes down to
British case, however, misses the variety of
the question of the adequacy of Rostow's pat-
experience in the evolution of different soci- tern in helping us isolate the strategic factors
eties, and makes the Marxian analysis of the which make for change, especially those fac-
"march of history" unduly rigid and artifi- tors that constitute the necessary and suffi-
cial. Iffor no other reason than that it draws cient conditions for determining the transi-
upon a far wider range of historical knowl- tion of an economy from a preceding stage to
edge, and is thereby more comprehensive and a succeeding stage.
less doctrinaire, Rostow's analysis can claim In this respect, Rostow's efforts are more
to be a superior alternative to the Marxian substantial than those by other proponents of
sequence. stages. Recognizing how important the
Nonetheless, if Rostow's thesis is to assert search for strategic factors is, Rostow adopts
with a high degree of generality that it is able an approach that is more analytical and re-
to trace a structure of history in the form of lated to a wider range of issues than any of
a sequence of stages, then it must also answer the approaches of his predecessors. His ar-
a number of criticisms that have commonly gument abounds with terms such as "forces,"
been levied against stage-theorists. "Stage- "process," "net result," "inner logic" — all in-
making" approaches are misleading when dicative ofhis desire to present an analytical,
they succumb to a linear conception of his- not merely a descriptive, set of stages. Ac-
tory and imply that all economies tend to cording to Rostow, the "analytic backbone"
pass through the same series of stages. Al- of his argument is "rooted in a dynamic the-
though aparticular sequence may correspond ory of production," and he believes that his
broadly to the historical experience of some set of stages reveals a "succession of strategic
economies, no single sequence fits the history choices" that confronts a country as it moves
of all countries. To maintain that every econ- forward through the development process.
omy always follows the same course of devel- On this basis, perhaps the most illumination
opment with a common past and the same fu- can be gained from Rostow's analysis by in-
ture is to overschematize the complex forces terpreting each stage as posing a particular
of development, and to give the sequence of type of problem, so that the sequence of
stages a generality that is unwarranted. A stages is equivalent to a series of problems
country may attain a later stage of develop- that confronts a country in the course of its
ment without first having passed through an development. Rostow's ultimate objective,
earlier stage, as stages may be skipped, and however, has been to present through his set
different types of economies do not have to of stages a theory about economic growth
succeed or evolve from one another. The se- and a means of uncovering both the uniform-
quence isalso blurred inasmuch as frequently ities and the uniqueness of each nation's ex-
the stages are not mutually exclusive, and perience. To be in a better position to judge
characteristics of earlier stages often become how successful Rostow has actually been in
mixed with characteristics of later stages. fulfilling these claims, we should now con-
Anyone who attempts to impose upon eco- sider the following appraisal of his analysis.
nomic history a one-way course of economic
evolution is bound to be challenged, since it 10Although Rostow gives little attention to the
problem, his analysis raises many questions related to
is difficult to accept one unique schema as the
basic social theory. In this connection, it is illuminat-
only real framework in which the facts truly ing to consult Isaiah Berlin, Historical Inevitability,
lie; the same facts can be arranged in many London, 1954, especially sections 2, 8.
 94 DEVELOPMENT PERSPECTIVES

II. A. 2. Empty Economic

Stages?*

Professor Rostow's bold generalisations or- consists of a recognisable discontinuity in the

dering the process of economic growth are no observed growth of income per capita, which
less controversial now than they were in 1960 under these conditions means a sharp in-
and of no less interest. Scholars and planners crease inthe savings rate:
continue to wrestle with the same basic prob- It is . . . useful to regard as a necessary but not
lem of economic development, but in a con- sufficient condition for the take-off the fact that
text considerably influenced by Rostow's the proportion of net investment to national in-
writings. The appearance of this volume af- come (or net national product) rises from, say, 5%
fords us all a valuable opportunity to re-ex- to over 10%, definitely outstripping the likely pop-
amine and rethink the issues that have ulation pressure (since under the assumed take-off
evolved in the debate of almost a decade upon circumstances the capital output ratio is low) and
the concept of take-off.1 yielding a distinct rise in real output per capita.2
The theories, of course, are not mutually
THE TWO THEORIES OF TAKE-OFF exclusive. But neither are they ever inte-
grated into a single framework. The discus-
Despite frequent clarifications and restate- sion of interindustrial relationships takes for
ments byRostow of the concept of take-off in granted the availability of resources and em-
the course of his two essays and liberal par- phasises instead the power of the demand
ticipation inthe discussion, its underlying nexus in stimulating growth among comple-
duality is not resolved. The implicit recogni- mentary activities. Its purpose is the exposi-
tion that there are indeed two theories of
tion of a mechanism for transmitting auton-
take-off is perhaps one of the most fruitful omous impulses affecting one sector to the
contributions made by the Conference. At economy as a whole. Not only output but also
one level the take-off is a sectoral, non-linear, technology is involved. The final objective is
threshold notion. This is the realm of the
a description of how the industrial matrix —
leading sector, with its forward, backward
its quantities and coefficients — undergoes a
and spreading effects breathing regular in- decisive and discontinuous transformation.
novation into the heretofore slumbering cor- The aggregate model takes as its starting-
pus economicum. It is the domain of certain point, on the other hand, the constraint of
crucial industries, such as those of coal, iron
savings, and hence investment, without which
and engineering. Economic growth is viewed growth cannot attain regularity at high rates.
as a constant struggle among a succession of Within an environment where the induce-
activities first accelerating and decelerating. ment to invest is the limiting factor, rather
The other level is highly aggregative. Here than savings, the sectoral approach seems
the setting is that of the familiar Harrod- much the more fruitful. The propensity to
Domar model in which the rate of growth of
save is dependent not so much upon con-
income is the product of an average propen- sumer decisions as those by entrepreneurs,
sity to save and the inverse of a reasonably
and the elasticity of resources and automatic-
stable capital-output ratio. Take-off then ity of technological advance is high. Some
underdeveloped countries may indeed fit such
*From Albert Fishlow, "Empty Economic a mould; Professor Hirschman's writings
Stages?" Economic Journal, Vol. 75, No. 297, stress this theme, and in this volume Profes-
March 1965, pp. 112-16, 120-25. Reprinted by
permission. sor Leibenstein also argues that such a shift
'A review of W. W. Rostow, ed., The Economics
of Take-off into Sustained Growth, Proceedings of a
Conference held by the International Economic As- 2W. W. Rostow, The Stages of Economic Growth,
sociation, London, Macmillan & Co. Ltd., 1963. Cambridge, 1960, p. 37.
 95 HISTORICAL PERSPECTIVES

in emphasis to investment rather than con- tory, the observed regularity of the ratio
centration upon savings is desirable. might suffice and causality be regarded more
Yet Rostow fails to elaborate the condi- lightly. In a world where take-off has become
tions under which one approach rather than a goal, and policies are at issue, there can be
the other yields a more cogent explanation of no such facile resolution, particularly when
how growth took root, or how the two are re- some of the careful qualifications stated by
lated. This explains why Professor Solow is Rostow are lost in the midst of a widening
troubled by the absence of any coherent the- audience. Professor Cairncross' essay speaks
ory and why Rostow's reply to him is unsat- cogently and quite usefully to these matters
isfactory. Byshifting between them in almost in more detail, and it is unnecessary to elab-
random fashion Rostow does the concept of orate further here.
take-off little service. And by not introducing Thus, at best the regularity hypothesised
resource constraints and a more adequate
discussion of technological diffusion into his by Rostow — the sharp increase in the saving's
rate and accelerated growth of per capita in-
sectoral theory he deprives it of a clarity and come within the span of a generation — pro-
consistency that the aggregate model does vides us with entree to the problems of tran-
enjoy. Furthermore, the data required to es- sition rather than its solution. Unfortunately,
tablish the acceleration in income per capita however, the hard facts of developmental ex-
and savings proportions are easier to come by perience cast considerable doubt upon such a
than the detailed analyses of individual coun- convenient and universal order. Drawing
tries, and Rostow confines himself with infre- upon data for Japan, Sweden, Germany,
quent exception to such evidence. Thus, while Britain, Canada and the United States, Pro-
he frequently protests the neglect of his sec- fessor Kuznets forcefully denies that such
toral approach, the blame resides not only on discontinuity is common to all. Rather he
his critics. sketches a picture of economies initally
blessed with respectable savings proportions
and growth rates, and experiencing gradual
THE USEFULNESS OF THE
increases in both well beyond what have been
AGGREGATE APPROACH
denoted as the take-off decades. Sometimes,
The result, then, is that it is the aggregate in his zeal, his conclusions overreach his own
notion of take-off that has become most evidence. For Germany and Japan the avail-
bruited about. How well does it perform? able series begin during the supposed take-
Not very, I believe. Despite its elegant sim- off; hence the relevant comparison with the
plicity, the conception of growth as a resul- earlier magnitudes is excluded. Similarly, for
tant of the savings rate and the aggregate Britain the discontinuity in the annual rate of
capital-output is singularly inappropriate for change of aggregate product understates the
a society in transformation. Even if formally break in the per capita series after 1785.
correct, it obscures and misleads concerning Moreover, even such a rise in the growth rate
the fundamental forces at work. The national of total output as from 0.9% prior to 1770
and 1.5% from 1770 to 1880 should not be
capital-output ratio lacks behavioural con-
tent. Its constancy subsumes complementary dismissed too lightly. Finally, more recently
changes in other inputs, such as labour and compiled data for countries such as Italy and
entrepreneurial skills, not to mention tech- Denmark appear to fit the original Rostow
nological progress. It excludes the important specification very well indeed.3
category of human capital. Its stability is These amendments to Kuznets' findings
consistent with wide-ranging shifts in indus- are not to argue for the validity of the Rostow
trial composition and investments in which
we precisely are interested. And as pointed
out above, the savings propensity itself can 3Reported in Simon Kuznets, "Long Term Trends
hardly be regarded as subject to autonomous in Capital Formation Proportions," Economic Devel-
opment and Cultural Change, Vol. IX, No. 4, Part II,
change. Were the question merely one of his- 1961.
 96 DEVELOPMENT PERSPECTIVES

conjecture. Still other countries, like Nor- Nor was it simply exploitation of mineral re-
way, do not conform at all, and recently com- sources: manufacturing expanded at the sal-
piled United States evidence stands in even utary rate of 5% per year as well. Again, in
greater contradiction than that available to Argentina the net national captial formation
Kuznets in 1960. The increase in the share of proportion averaged some 20% in the years
output devoted to domestic capital formation 1900-15, with annual growth of product of
from the 1830s on appears to be substantially almost 3%. Much social overhead capital had
more modest than that observed in other been accumulated, and in Argentina, at least,
economies: the gross rate starts at 15% in much human capital as well.5 Many of these
1834-43, changes little during the ante hel- same features are common to Chile. Yet ret-
ium period and increases (in part due to rel- rospectively, the anticipated prize of sus-
ative price changes) to about 25% in the post- tained growth did not materialise. Must we
war decades.4 There is obviously a variety of then agree that take-off did not occur since
national experiences to be reckoned with. "the subsequent periods of stagnation or re-
Some economies may have executed the req- lapse ... are clear" (p. 27)? Why not take-
uisite saltus in aggregate terms, while others off, followed by structural difficulties that left
pursued a more gradual course. (Appropri- such economies modernised only in certain,
ately enough, Kenneth Berrill finds the same but not all, respects?
diversity in the dependence upon foreign cap- The unnecessarily narrow Rostow view is
ital in the course of development.) Since both imposed by the logic of a stage theory with-
groups attained the same conversion to eco- out an endogenous propulsive force. Take-off
nomic growth rooted in a modern technology, accordingly must be defined not in terms of
the Rostow hypothesis must be adjudged of the extent and nature of the inner transfor-
limited merit.
mation, but the later accomplishment of con-
Note also the additional and unfortunate
tinuing flight. For only then, precisely by ref-
rigidity imposed by enmeshing the take-off erence to the self-sustained character of
within a full-blown stage theory of economic subsequent growth, can we be certain take-
evolution. Like most nineteenth-century con- off has occurred. In turn, the preconditions
ceptions ofprogress, the path is orderly and have been satisfied, since they are specified
monotonic: societies move through each of only as those permitting take-off to occur.
the stages in succession. But unfortunately The minimum social overhead capital, the
excluded by such a hierarchy are many of the rise of a large enough entrepreneurial class,
most interesting cases — economies that en- the prior sufficient growth in agricultural
tered well into the transition phase but did productivity, are all ratified. The other, and
not succeed. Some of the nations of Latin
America seem to have been so beset. At the necessary, way of articulating the three cen-
tral stages of the global theory is to specify
end of the nineteenth and beginning of the the conditions for moving from one to the
twentieth century, the pace of development other independently of later events; this in
quickened in that part of the world as it did turn means a complete theory of social
in Russia and Italy, among others. In Mex- change, including non-economic factors.
ico, between 1895 and 1910, one set of per Rostow, despite even more than customary
capita income estimates places growth at obeisance to the latter, never successfully in-
1.6% annually, or faster than Britain and tegrates them into his analysis, precisely for
possibly even Russia during their take-offs. the reason that the linkage he imposes runs
in the opposite temporal direction.

^hese are the estimates of Robert E. Gallman and

5For Mexico, see Mexico, Cincuenta Ahos de Rev-
will appear in his "Gross National Product in the olution, Mexico City, 1960, Vol. I, p. 585; for the Ar-
United
Studies States, 1834-1909,"
in Income and Wealth.to appear in Vol. 29 of gentine data, see Kuznets, "Long Term Trends in
Capital Formation Proportions."
 97 HISTORICAL PERSPECTIVES

The consequence is a non-operational con- 5. Industrial, rather than agricultural,

cept. To talk knowingly of countries in the growth affords the initial basis for sustained
world to-day as in the pre-conditions phase or development.
the midst of take-off is sheer guesswork, un- Now, to be sure, these propositions are by
related to the theory itself. This is why it is no means proven. Indeed, one can set out
possible (so far) to be right with the assess- many concrete instances in which the Rostow
ment of Mexico but wrong with Argentina. examples purporting to establish one or more
In neither instances are objective criteria in- of the points is grossly in error. Not the least
voked. The consequence of this counter-tem- of these deal with the supposed importance of
poral linkage is to rob the take-off phase at the railroad in triggering take-off. In the
the same time of its attractiveness as a deci- United States, for example, the role played
sive period — possibly one of many — in the by the railroad before 1 860 in developing for
growth of a society and to reduce it to a me- the first time a machinery industry is clearly
chanical construct whose identification mis-stated. The reason locomotives and other
within a specific set of years is more to be equipment were produced domestically so
worried about than an elaboration of its soon after the introduction of the railway was
content. . . . that machinery firms were already in exis-
tence to serve the textile and steamboat in-
terest. Virtually all the locomotive firms grew
THE SECTORAL VARIANT out of such origins. Nor did their specialised
production dominate the industry anything
OF TAKE-OFF
like the mistaken horse-power estimates Ros-
The aggregate approach to take-off has not tow cites. Where the unique significance of
fared well. It is unable to deliver the univer- the railway resides with respect to engineer-
sal, endogenous theory of development it ing inputs is in its substantial maintenance
promises. But such deserved criticism should requirements. Widely dispersed repair shops
not obscure the many insights take-off has to training pools of local talent undoubtedly
offer at the sectoral level. There its great were a significant factor in technological
strength is an explicitly partial orientation diffusion.
emphasising process, and so subject to empir- What is more, the entire emphasis upon
ical test. Rostow's various views may be sum- the industrial consequences following from
marized by the following interrelated the backward linkage of railroadisation as
propositions: being crucial in the United States misses the
1. Successful industrialisation is unbal- point. At the time when pig-iron output first
anced in the sense that a single, or limited underwent apparent rapid increase in the
number of industries, is the source from 1840s iron rail demand was an insignificant
which an initial acceleration ramifies through proportion of the whole. Coal, too, was a
the economy. There is a consequent discon- minor need for railroad operations in an era
tinuity inproduction of manufactures. of wood-burning engines. Rather what made
2. Such leading sectors have three paths of the rapid projection of the railroad in the
influence upon the economy: forward, lateral 1850s so consequential was the impetus it
and backward linkages, of which the latter gave to agricultural expansion. In this and
route has historically predominated. other instances Rostow tends to understate
3. Certain industries have played the role the role of the market, actual and antici-
of leading sector in a number of different pated, in favour of derived demand nexuses.
countries, notably the railroad. While the latter may to-day be more potent,
4. The development of certain subsidiary we should not re-write history in its image.
activities — coal, iron and machinery — is a Modernisation may proceed impelled by rel-
good index of the extent of industrialisation ative prices as well as by government, and by
and the probability of its continuation. way of forward linkages as well as backward.
 98 DEVELOPMENT PERSPECTIVES

It also follows that yet another proposition considerably improves upon the earlier ver-
closer study may amend and qualify is the sion, which tended to stress the observed rate
extent to which commercial agriculture may of growth of the leading sector as a measure
serve as a leading sector. Professor Bulhoes of significance. The legitimate criticisms of
treats the case of the state of Sao Paulo in Kuznets, that leadership represents a combi-
Brazil, where he outlines a sequence leading nation of internal growth, weight within the
from expanding coffee production to social industrial matrix and the magnitude of other
overhead investment to industrial diversifi- effects, thus are fully met. . . .
cation. The evolutionary process in this in- With the forward, backward and lateral
stance contradicts both the Rostow and Ger-
linkages spelled out in this fashion, Rostow's
schenkron emphasis. For new countries, leading sectors now become virtually the his-
drawing extensively upon immigration and torical analogue of Hirschman's strategy for
capital imports, and with large land re- contemporary unbalanced growth. Certain
sources, this may be the natural pattern. One activities have capacity for greater influence
may cite the United States, Canada, Aus- than others because they have more dis-
tralia and New Zealand. Where such a pro- persed, or larger, demands upon manufac-
cess is arrested, as in Brazil and Argentina, tures as a whole, or some critical subset;
it may cause sustained growth to be much other sectors pass along their output not to
more difficult to achieve than might other- consumers directly but to other industries for
wise be the case. Ungrasped opportunities further processing, and so present a potential
may exact a continuing toll as a disadvantage force from the side of supply; finally, some in-
of forwardness. dustries, byvirtue of location, or training of
On the other hand, Professor Boserup's a skilled labour force, or demonstration ef-
comprehensive essay defends what has al- fects, operate beyond interindustrial relation-
most come to be the conventional view that ships. Leading sectors presumably possess
industrialisation proceeds despite agriculture some critical combination of all three, the
rather than as a consequence of a progressive first two of which, at least, are subject to
primary sector. "In the normal course of well-defined empirical investigation.
events, new things do not originate in the ag- There is much to be done to extract the po-
ricultural milieu" (p. 205). At best, rising ag- tential gains of this approach. Forward link-
ricultural productivity serves as a precondi- ages, which are more behavioural in their in-
tion for take-off, and some type of ducements than the technically determined
institutional reform is required to raise sup- backward linkages, are yet to be suitably de-
ply elasticities. Within the European and fined. Various proposed measures within the
South Asian context of sedentary agricul- input-output framework, such as the propor-
ture, one can find little fault with these con- tion of sales to industrial purchasers versus
clusions; the principal caveat is the applica- final demand or more detailed summaries of
bility of the results to more recently settled the inverse matrix of technological coeffi-
lands. cients to rank the absorptive consequences of
These limitations are of no matter. Much an initial expansion in output, do not fully
more detailed research is required before the render the market dimensions of this supply
central tenets of the hypothesis can be fairly influence.6 The extent to which backward
evaluated and suitably qualified. That the
sectoral variant of take-off both permits and
structures such research is testimony to its 6Albert Hirschman, The Strategy of Economic De-
usefulness. Indeed, Rostow himself in this velopment, New Haven, 1958, pp. 106-7, uses the
volume has restated the mechanism in such a sales ratio measure. P. N. Rasmussen, Studies in
way as to further its application. By making Inter-Sectoral Relations, Copenhagen, 1956, devel-
more explicit the type of linkages by which ops an index of sensitivity of dispersion that "ex-
presses the extent to which the system of industries
leading sectors transmit their impulses he draws upon industry no. /" (p. 135).
 99 HISTORICAL PERSPECTIVES

linkages have been emphasised may reflect cability of many of the initial Rostow propo-
the better analytical basis for measuring de- sitions. Take-off becomes a sectoral joint the-
rived demands as much as their differential ory of output change and technological
importance. diffusion.
Nor can we ignore the very real data lim- So interpreted, the leading sector becomes
itations. Complete, disaggregated industrial not only the agent of triggering quantity re-
matrices are not now available, and perhaps sponse inother industries but also of dissem-
never can be, far distant into the past. None inating modern techniques. It therefore is
the less, in most actual instances relatively likely to be a new industry itself using ad-
few, but significant, interactions are involved. vanced technology, a description consistent
However crude the first historical applica- with the broad contours of historical indus-
tions of the leading sector concept inevitably trialisations. Corresponding to the disconti-
turn out to be, they promise to add a consid- nuity in industrial production is a threshold
erable sophistication to our studies of past in- level of acceptance of novel methods. Here is
dustrialisations, successes and failures. The where the thrust of non-economic influences
input-output apparatus so mobilised distin- should be directed: which social circum-
guishes clearly between the magnitude of di- stances are barriers and aids to technological
rect and indirect effects, and settles the re- change. Once the level of resistance has been
current question raised at this Conference overcome, diffusion is assumed to be regular
and elsewhere of whether rates of growth or and continual. Industrial growth typically,
absolute increments are the relevant criteria then, will continue at high levels for back-
of the importance of a single industry by ward economies as they absorb better pro-
placing its contribution in an appropriate duction techniques.
multi-sector context. Such a prominent place for technology in
This comparative-statical framework, the schema reinforces the logical basis for
however enlightening, does not exhaust the many of the earlier Rostow conjectures con-
content of initial industrialisation. But nei- cerning the pattern of industrialisation. The
ther does it fully render the concept of take- physical presence of an industry like ma-
off. Rostow's category of lateral linkages is chinery— in contrast to dependence upon im-
an attempt to introduce further elements of ports— is indeed a good guide to successful
change into the picture. That such a grouping industrialisation. It creates a national vested
is too broad and nebulous to serve the needs interest committed to technological progress
of empirical research is not to deny his rec- and obsolescence while at the same time re-
ognition ofthe problem. In preference to that ducing uncertainty by assuring close contact
procedure, it would seem better to retain the between supplier and user of the new tech-
input-output context and to interpret the en- niques. Where maintenance needs are apt to
tire array of technical coefficients not as con- be crucial, proximity is no small comfort.
stants but as variables themselves subject to Needless to say, the portrait of an industrial
change as a function of changes in output of matrix expanding by initial augmentation of
certain industries or specific coefficients.7 sectors utilising modem technology gives
Not only do we thereby faithfully translate greater weight to the preference for manu-
the Rostow emphasis upon productivity factures over agriculture as a vehicle of
change as one of the fundamental influences
upon modern growth but we retain the appli- Whether these statements suffice or other
growth.
propositions relating to technological difffsion
7It is useful in this context to note the identity be- are substituted, the important point is that
tween a geometric index of productivity change and productivity increase is afforded an explicit
that obtained from the changing matrix of technical and central position in the hypothesis. This is
input-output coefficients. See Evsey D. Domar, "On a great advance beyond the simple capital-
the Measurement of Technological Change," Eco-
nomic Journal, Vol. LXXI, 1961, pp. 709-29. accumulation models still current — and un-
 100 DEVELOPMENT PERSPECTIVES

deriving the aggregate version of take-off it- CONCLUDING COMMENT

self. Much must be done to implement the
The appearance of The Economics of
approach, both analytically and empirically.
Historical initial conditions must be allowed Take-off marks a turning-point in a debate
that has now been carried on for almost a
to influence the propositions, thereby soften- decade. Here for the first time is a systematic
ing the claims of universality.8 confrontation of the take-off hypothesis with
The take-off phase itself must be inte-
grated into later, sustained growth. Are long a considerable segment of historical experi-
ence. The aggregate version of the concept
savings a miniature replay of the initial dis-
continuity postulated, as Rostow sometimes emerges qualified at best: Rostow's readiness
seems to suggest? How are the initital re- to accept any number of amendments is tes-
source constraints imposed by inadequate timony less to the flexibility of the notion
than to its failure to fit well the actual diverse
capital formation resolved over time in re-
sponse toindustrialisation? How do increases paths of development.
The test of the hypothesis afforded by the
in agricultural productivity become regular,
recent quantitative research upon historical
not as a pre-requisite of the industrial surge, patterns of growth constitutes the principal
but as a resultant? Without a respectable
contribution of the volume. The papers and
rate of growth of agricultural product, rapid
national income advance is almost impossi- discussion based less factually have a corre-
spondingly less impact, in no small measure
ble. Finally, much systematic historical re-
search lies ahead. We do not even know at because their relationship to the take-off is
more tenuous. Yet despite this progress the
the present what the role of productivity Conference achieved in its examination of
change was, or is, in the contemporary world
in the first stages of industrialisation! the discontinuity of modern economic
The sectoral variant of take-off thus can growth, important questions remain. It is for
this reason that I have tried to suggest a new
serve as the basis for important and relevant
research into many still unresolved problems focal point for the discussion by differentiat-
ing between the aggregate and sectoral ap-
of economic growth.9 While economic theo-
rists wrestle with the implications of embod- proaches to take-off, and to the role techno-
ied and disembodied technical progress, and logical change plays in the last. Where there
empirical workers extend the frontiers of is little reason to accept the complete expla-
measurement of productivity, it is only ap- nation of national income growth Rostow of-
propriate that historians bring to bear accu- fers, there is good cause to pursue his many-
suggestions concerning the process by which
mulated experience upon technological dif- industrialisation becomes rooted. Rostow.
fusion and its contribution to successful
industrialisation. paradoxically, is at his best read as a pros-
pectus rather than as a treatise.
The loss of grandeur in the descent from
8Recent Latin American experience where rapid
manifesto to partial hypothesis must not be
increase in the manufacturing sector has not led to exaggerated. It is a rare occasion when op-
continuing growth is an obvious case in point. For erational, albeit partial, theories pregnant
Chile, see Marto A. Ballesteros and Tom E. Davis.
with potential are put forward. The concep-
"The Growth of Output and Employment in Basic
Sectors of the Chilean Economy, 1908-1957." Eco-
tion of take-off is just such an event. Profes-
nomic Development and Cultural Change, Vol. XI. sor Clapham, some forty years ago, chided
No. 2, 1962, p. 153. his theoretical colleagues for providing him
9For one related attempt see Hollis B. Chenery. with empty boxes. Better that historians to-
Shuntaro Shishido. and Tsunehiko Watanabe. "The day recognise the important possibilities of
Pattern of Japanese Growth, 1914-1954," Econo- take-off in structuring inquiry than to con-
metrica. Vol. XXX, No. 1. 1962, pp. 114-118 espe-
cially. Although tney fail to carry the analysis of tech- tinue to expound (or to reject out of hand) a
nological change very far, their results illustrate the grand set of stages that lamentably are some-
potential returns from sectoral analysis. times vacant.
 101 HISTORICAL PERSPECTIVES

Comment
Rostow has extended his analysis in How It All Began (1975) and Why the Poor Get Richer
and the Rich Slow Down: Essays in the Marshallian Long Period (1970).
For a more detailed exposition of Rostow's general thesis, and criticisms levied against it,
see the papers presented at the International Economic Association's conference, published
in W. W. Rostow et al., Economics of Take-Off into Sustained Growth (1963).
Several other critiques deserve special mention: K. Berrill, "Historical Experience: The
Problem of Economic 'Take-Off,' " in K. Berrill (ed.), Economic Development with Special
Reference to East Asia (1964), chapter 7; Henry Rosovsky, "The Take-Off into Sustained
Controversy," Journal of Economic History (June 1965); P. Baran and E. Hobsbawm, "The
Stages of Economic Growth," Kyklos, Vol. 14, No. 2 (1961); P. T. Bauer and Charles Wilson,
"The Stages of Growth," Economica (May 1962); S. G. Checkland, "Theories of Economic
and Social Evolution: The Rostow Challenge," Scottish Journal of Political Economy (No-
vember 1960); E. E. Hagen, On the Theory of Social Change (1962), appendix 2; D. C. North,
"A Note on Professor Rostow's Take-Off' into Self-sustained Economic Growth," The Man-
chester School (January 1958); Goran Ohlin, "Reflections on the Rostow Doctrine," Eco-
nomic Development and Cultural Change (July 1961); G. L. S. Shackle, "The Stages of Eco-
nomic Growth," Political Studies (February 1962). But see Rostow's second edition of The
Stages of Economic Growth (1971) and Politics and the Stages of Growth (1971).
Of special interest is the Festschrift in honor of Rostow, Economics in the Long View
(1982), edited by C. P. Kindleberger and Guido di Telia.

II.A. 3. Economic Backwardness

in Historical Perspective*

The map of Europe in the nineteenth century 1. The more backward a country's econ-
showed a motley picture of countries varying omy, the more likely was its industrialization
with regard to the degree of their economic to start discontinuously as a sudden great
backwardness. At the same time, processes of spurt proceeding at a relatively high rate of
rapid industrialization started in several growth of manufacturing output.1
of those countries from very different levels 2. The more backward a country's econ-
of economic backwardness. Those differences omy, the more pronounced was the stress in
in points — or planes — of departure were of its industrialization on bigness of both plant
crucial significance for the nature of the sub- and enterprise.
sequent development. Depending on a given 3. The more backward a country's econ-
country's degree of economic backwardness omy, the greater was the stress upon produc-
on the eve of its industrialization, the course ers' goods as against consumers' goods.
and character of the latter tended to vary in
a number of important respects. Those vari- 'The "great spurt" is closely related to W. W. Ros-
tow's "take-off" ( The Stages of Economic Growth,
ations can be readily compressed into the Cambridge University Press, 1960, Chap. 4). Both
shorthand of six propositions. concepts stress the element of specific discontinuity in
economic development; great spurts, however, are
confined to the area of manufacturing and mining,
*From Alexander Gerschenkron, Economic Back- whereas take-offs refer to national output. Unfortu-
wardness in Historical Perspective, Cambridge, nately, inthe present state of our statistical informa-
Mass.: Harvard University Press, copyright, 1962, by tion on long-term growth of national income, there is
The President and Fellows of Harvard College, pp. hardly any way of establishing, let alone testing, the
353-9. Reprinted by permission. take-off hypotheses.
 102 DEVELOPMENT PERSPECTIVES

4. The more backward a country's econ- ment.2 Alternatively, the element of necessity
omy, the heavier was the pressure upon the may have been somewhat disguised by well-
levels of consumption of the population. meant, even though fairly meaningless, re-
5. The more backward a country's econ- marks about the choices that were open to so-
omy, the greater was the part played by spe- ciety.3 But all those schemes were dominated
cial institutional factors designed to increase by the idea of uniformity. Thus, Rostow was
supply of capital to the nascent industries at pains to assert that the process of indus-
and, in addition, to provide them with less de- trialization repeated itself from country to
centralized and better informed entrepreneu- country lumbering through his pentametric
rial guidance; the more backward the coun- rhythm. . . .
try, the more pronounced was the The point, however, is not simply that
coerciveness and comprehensiveness of those these were important occurrences which have
factors.
just claims on the historian's attention. What
6. The more backward a country, the less matters in the present connection is that ob-
likely was its agriculture to play any active serving the individual methods of financing
role by offering to the growing industries the industrial growth helps us to understand the
advantages of an expanding industrial mar- crucial problem of prerequisites for industrial
ket based in turn on the rising productivity of development.
agricultural labor. The common opinion on the subject has
. . . [T]he differences in the level of eco- been well stated by Rostow. There is said to
nomic advance among the individual Euro- be a number of certain general preconditions
pean countries or groups of countries in the or prerequisites for industrial growth, with-
last century were sufficiently large to make it out which it could not begin. Abolition of an
possible to array those countries, or group of archaic framework in agricultural organiza-
countries, along a scale of increasing degrees tion or an increase in the productivity of ag-
of backwardness and thus to render the latter riculture; creation of an influential modern
an operationally usable concept. Cutting two elite which is materially or ideally interested
notches into that scale yields three groups of in economic change; provision of what is
countries which may be roughly described as called social-overhead capital in physical
advanced, moderately backward, and very form — all these are viewed as "necessary
backward. To the extent that certain of the
preconditions," except that some reference to
variations in our six propositions can also be the multifarious forms in which the prereq-
conceived as discrete rather than continuous, uisites are fulfilled in the individual areas are
the pattern assumes the form of a series of
designed to take care of the "unique" factors
stage constructs. Understandably enough, in development. Similarly, the existence of a
this result obtains most naturally with regard value system favoring economic progress and
to factors referred to in proposition 5, where the availability of effective entrepreneurial
quantitative differences are associated with groups basking in the sun of social approval
qualitative, that is, institutional, vari- have been regarded as essential preconditions
ations. .. . of industrial growth.
Such an attempt to view the course of in- These positions are part and parcel of an
dustrialization as a schematic stagelike pro- undifferentiated approach to industrial his-
cess differs essentially from the various ef- tory. But their conceptual and empirical de-
forts in "stage making," the common feature ficiencies are very considerable, even though
of which was the assumption that all econo-
mies were supposed regularly to pass through
the same individual stages as they moved 2See, for example, Bruno Hildebrand, Die Nation-
alokonomie, der Gegenwart und Zukunft und andere
along the road of economic progress. The reg- gesammelte Schriften, 1, Jena, 1922, p. 357.
ularity may have been frankly presented as 3See Rostow, The Stages of Economic Growth, pp.
118rT.
an inescapable "law" of economic develop-
 103 HISTORICAL PERSPECTIVES

it is by no means easy to bid farewell to this industrializations proceeding under different

highly simplified way of viewing the pro- conditions. On the other hand, once the ques-
cesses ofindustrialization. It took the present tion has been asked, whole series of various
writer several years before he succeeded in substitutions become visible which could be
reformulating the concept of prerequisites so readily organized in a meaningful pattern ac-
that it could be fit into the general approach cording to the degree of economic backward-
premised upon the notion of relative ness. ... [I] t is easy to conceive of the capital
backwardness. . . . supplied to the early factories in an advanced
There should be a fine on the use of words country as stemming from previously accu-
such as "necessary" or "necessity" in histor- mulated wealth or from gradually plowed-
ical writings. As one takes a closer look at the back profits; at the same time, actions by
concept of necessity as it is appended to pre- banks and governments in less advanced
requisites ofindustrial development, it be- countries are regarded as successful attempts
comes clear that, whenever the concept is not to create in the course of industrialization
entirely destitute of meaning, it is likely to be conditions which had not been created in the
purely definitional: industrialization is de- "preindustrial" periods precisely because of
fined in terms of certain conditions which the economic backwardness of the areas
concerned. . . .
then, by an imperceptible shift of the writer's
wrist, are metamorphosed into historical . . . [T]he area of capital supply is only one
preconditions.4 instance of substitutions for missing prereq-
The recourse of tautologies and dexterous uisites. As one looks at the various patterns
manipulations has been produced by, or at of substitution in the individual countries,
any rate served to disguise, very real empiri- taking proper account of the effects of grad-
cal difficulties. After having satisfied oneself ually diminishing backwardness, one is
that in England certain factors could be rea- tempted to formulate still another general
sonably regarded as having preconditioned proposition. The more backward was a coun-
the industrialization of the country, the ten- try on the eve of its great spurt of industrial
dency was, and still is, to elevate them to the development, the more likely were the pro-
rank of ubiquitous prerequisites of all Euro- cesses ofits industrialization to present a rich
pean industrializations. Unfortunately, the and complex picture — thus providing a curi-
attempt was inconsistent with two empirical ous contrast with its own preindustrial his-
observations: (1) some of the factors that had tory that most often was found to have been
served as prerequisites in England either relatively barren. In an advanced country, on
were not present in less advanced countries or the other hand, the very richness of its eco-
at best were present to a very small extent; nomic history in the preindustrial periods
(2) the big spurt of industrial development rendered possible a relatively simple and
occurred in those countries despite the lack of straightforward course in its modern indus-
such prerequisites. trial history.
If these observations are not ignored or Thus, the concept of prerequisites must be
shrugged away, as is usually done, they quite regarded as an integral part of this writer's
naturally direct research toward a new ques- general approach to the industrial history of
tion: in what way and through the use of Europe. At the same time, it is important to
what devices did backward countries substi- keep in mind the heuristic nature of the con-
tute for the missing prerequisites? ... It ap- cept. There is no intention to suggest that
pears, on the one hand, that some of the al- backward countries necessarily engaged in
leged prerequisites were not needed in deliberate acts of "substitution" for some-
thing that had been in evidence in more ad-
vanced countries. Men in a less developed
4It is not surprising, therefore, to see Rostow at one
point (p. 49) mix conditions and preconditions of in- country may have simply groped for and
dustrial development very freely. found solutions that were consonant with the
 104 DEVELOPMENT PERSPECTIVES

existing conditions of backwardness. In fact, process of cognition and should be judged by

one could conceivably start the study of Eu- its fruits.
ropean industrializations in the east rather The main advantage of viewing European
than in the west of the Continent and view history as patterns of substitutions governed
some elements in English industrial history by the prevailing — and changing — degree of
as substitutions for the German or the Rus- backwardness lies, perhaps paradoxically, in
sian way of doing things. This would not be a its offering a set of predictabilities while at
very good way to proceed. It would make the same time placing limitations upon our
mockery of chronology and would be glar- ability to predict. To predict is not to pro-
ingly artificial. True, some artificiality also phesy. Prediction in historical research
inheres in the opposite approach. It is arbi- means addressing intelligent, that is, suffi-
trary to select England as the seat of prereq- ciently specific, questions as new materials
uisites. Yet this is the arbitrariness of the are approached.
Comment
No attempt can be made to do justice to the innumerable historical studies of development.
Only a few survey studies can be singled out: R. M. Hartwell, "The Causes of Industrial
Revolution: An Essay in Methodology," Economic History Review, Vol. 18, No. 1 (1965);
Richard A. Easterlin, "Is There a Need for Historical Research on Underdevelopment?"
American Economic Review (May 1965); Barry E. Supple, "Economic History and Economic
Underdevelopment," Canadian Journal of Economics and Political Science (November
1961); John C. H. Fei and Gustav Ranis, "Economic Development in Historical Perspective,"
American Economic Review (May 1969), J. Hughes, Industrialization and Economic History,
New York (1970), part II; J. R. Hicks, A Theory of Economic History, London (1969).
An interesting collection of "successful and promising beginnings" of development is con-
tained in Malcolm E. Falkus (ed.), Readings in the History of Economic Growth, New York
(1968); also, Barry E. Supple (ed.), The Experience of Economic Growth, New York (1963).

II.A.4. Future Development in

Historical Perspective— Note

"The historian is a prophet looking back- and present take-offs, and suggest what pol-
wards"— this dictum is apt for the economic icy implications flow from the differences.
historian concerned with development. Ros- With respect to the role of particular sec-
tow's analysis, for instance, presumes that tors of the economy, Rostow observes many
the choices now confronting the poor coun- problems and patterns familiar from the past.
tries may be revealed in the light of the stages He submits that present take-offs depend, as
of preconditions and take-off that the cur- in the past, on the allocation of resources
rently rich countries experienced in earlier
centuries, and that historical perspective may to building up and modernizing the three non-in-
contribute to the formulation of development dustrial sectors required as the matrix for indus-
trial growth: social overhead capital; agriculture;
policy. From this viewpoint, Rostow's analy- and foreign-exchange-earning sectors, rooted in
sis may be most instructive for many coun- the improved exploitation of natural resources. In
tries that have not yet passed successfully addition, they must begin to find areas of modern
through the take-off stage: it may point up processing or manufacture where the application
the similarities and differences between past of modern technique (combined with high income-
 105 HISTORICAL PERSPECTIVES

or price-elasticities of demand) are likely to permit . . . how to persuade the peasant to change his
rapid growth-rates, with a high rate of plow-back methods and shift to producing for wider markets;
of profits.1 how to build up a corps of technicians, capable of
It will be instructive to reconsider these manipulating the new techniques; how to create a
corps of entrepreneurs, oriented not towards large
conclusions after reading Chapters VI-IX profit margins at existing levels of output and tech-
below, where questions not recognized by nique, but to expand output, under a regime of
Rostow are raised regarding the allocation of regular technological change and obsolescence;
investment resources and the role of how to create a modern professional civil and mil-
industrialization. itary service, reasonably content with their sala-
Further, Rostow believes that for the pres- ries, oriented to the welfare of the nation and to
ently underdeveloped nations, the inner me- standards of efficient performance, rather than to
chanics of the take-off involve problems of graft and to ties of family, clan, or region.4
capital formation, just as in the past. If their On the basis of foregoing similarities, Ros-
take-offs are to succeed, the underdeveloped tow regards the process of development now
countries "must seek ways to tap off into the going forward in Asia, the Middle East, Af-
modern sector income above consumption rica, and Latin America as analogues to the
levels hitherto sterilized by the arrangements stages of preconditions and take-off of other
controlling traditional agriculture. They societies in earlier centuries. But there are
must seek to shift men of enterprise from also differences — by way of different kinds of
trade and money-lending to industry. And to problems now confronting poor countries,
these ends patterns of fiscal, monetary, and and in the manner in which some problems,
other policies (including education policies) although similar in kind to those of the past,
must be applied, similar to those developed are now expressed in different degrees of in-
and applied in the past."2 tensity and complexity. Especially significant
Again, this interpretation of the take-off is the fact that the poor countries now stand
should be critically reexamined after reading in a different relationship to rich countries
Chapter IV, below, where a case is made than was true when the presently rich coun-
against assigning as much importance to the tries were poor. These differences are ex-
role of capital accumulation as Rostow does. tremely important, and they deserve more at-
Rostow also notes some political and socio- tention than Rostow gives them. For insofar
cultural similarities between past and present as most of these differences aggravate the
take-offs. As in the past, political interest problems of the take-off, they warn against
groups range from defenders of the status letting the success-stories of past take-offs
quo to those prepared to force the pace of lull us into too easy an interpretation of the
modernization at whatever cost; there exists development task. Nor should we equate the
the balance between external expression of LDCs to the early stages of the presently de-
nationalism in almost every case; above all, veloped countries. The persistence of under-
"there is continuity in the role of reactive na- development in the world economy poses
tionalism, as an engine of modernization, some refractory problems that were absent in
linked effectively to or at cross-purposes with earlier cases of successful development. If we
other motives for remaking traditionalist recognize these differences, we may hesitate
society."3 to join Rostow in concluding that in the end
Historical cases of successful take-offs also
the lesson of history is that "the tricks of
indicate a contemporary catalogue of neces-
sary social change: growth are not all that difficult."5
In the first place, poor countries are at-
4Ibid. tempting to accelerate their development
'W. W. Rostow, The Stages of Economic Growth, from a lower economic level than was true for
Cambridge, 1960, p. 139.
2Ibid.
5Ibid., p. 166.
3Ibid., p. 140.
 106 DEVELOPMENT PERSPECTIVES

the presently rich countries at the time of income than did the presently developed
their rapid rates of development. As Kuznets countries, but their relative positions are also
observes, inferior compared with other countries — un-
like the position of the early comers to devel-
Output per capita is much lower in the underde-
opment that entered the industrialization
veloped countries today than it was in the pres-
ently developed countries at the date of entry — a
process from a position of superior per capita
period rather than a point of time — into modern income relative to other countries. The im-
economic growth, i.e., when growth of per-capita plications ofattempting to develop rapidly
product (with an already high rate of population from a lower level of per capita income, and
growth) began to accelerate, the shift toward non- from a relative position that entails more
agricultural sectors occurred, modern technology pressures of backwardness, should receive a
(modern by the times) was adopted, and so forth.
However difficult the comparison of per-capita
fuller treatment than Rostow's analysis
gross product at such distances of time and space, provides.
the weight of the evidence clearly suggests that, Gerschenkron's suggestive analysis, out-
with the single and significant exception of Japan lined above, can provide a more profound un-
(the records for which are still to be fully tested), derstanding ofthese implications. We should
the pre-industrialization per-capita product in the examine, as does Gerschenkron, the pro-
presently developed countries, at least $200 in cesses ofindustrial development in relation to
1958 prices (and significantly more, in the off- the degree of backwardness of the areas con-
shoots overseas), was appreciably higher than per- cerned on the eve of their great spurts of in-
capita product in underdeveloped countries in the
dustrialization. Gerschenkron's approach has
late 1950s — certainly in most of Asia and Africa,
and in a good part of Latin America. distinct advantages over Rostow's in main-
taining that it is only by comparing indus-
Yet this statistical difference in aggregate out- trialization processes in several countries at
put per capita is less important than what it rep- various levels of backwardness that we can
resents. Itimplies that even today these underde-
veloped countries still have such a low product per hope to separate what is accidental in a given
capita that they are not at the same stage as the industrial evolution from what can be attrib-
presently developed countries were at their initial uted to the historical lags in a country's de-
stage of modern economic growth. This seems to velopment, and that it is only because a de-
be the case despite access to modern technology veloping country is part of a larger area
and despite the existence of a modern sector within which comprises more advanced countries
these countries (no matter how small). These
that the historical lags are likely to be over-
underdeveloped countries are either at some ear-
lier stage within the long-term trend of the pres- come in a specifically
Another fundamental intelligible fashion.7
difference is that
ently developed countries — in terms of the West-
ern European sequence perhaps at the period of many of the poor countries have not yet ex-
city formation in the early Middle Ages; or, what perienced any significant degree of agricul-
is far more defensible, they are at some stage in a tural improvement as a basis for industriali-
sequence of long-term growth separate and dis- zation. The failure to have yet undergone an
tinct from that of the Western European cradle of agricultural revolution makes the present
the modern economic epoch and are following a problem of accelerating development far
time and phase sequence that may be quite more difficult than it was for the now devel-
different.6 oped countries when they entered upon their
Not only do poor countries now confront industrial revolutions. It is fairly conclusive
the strategic policy issues of development that productivity is lower in the agricultural
from an absolutely lower level of per capita
7Alexander Gerschenkron, Economic Backward-
6Simon Kuznets, "Notes on Stage of Economic ness in Historical Perspective, Cambridge, 1962, p.
Growth as a System Determinant," in Alexander 42. For the application of this general conception of a
Eckstein (ed.), Comparison of Economic Systems, system of gradations of backwardness to particular
Berkeley, 1971, pp. 254-5. countries, see Chaps. 1, 4, 7, 8.
 107 HISTORICAL PERSPECTIVES

sector of underdeveloped countries than it account for some obstacles to development

was in the pre-industrialization phase of the that are now more formidable than in the
presently developed countries. Although di- past. Unlike the social heritage with which
rect evidence of this is unavailable, it is in- Western countries entered the take-off stage,
directly confirmed by data suggesting that the social structure and value pattern in
the supply of agricultural land per capita is many poor countries are still inimical to de-
much lower in most underdeveloped coun- velopment. The structure of social relations
tries today than it was in presently developed tends to be hierarchical, social cleavages re-
countries during their take-off, and that there main pronounced, and mobility among
is a wider difference between per worker in- groups is limited. Instead of allowing individ-
come in agricultural and nonagricultural sec- uals to achieve status by their own efforts and
tors in the underdeveloped countries today performance, their status may be simply as-
than there was in the pre-industrial phase of cribed to them, according to their position in
presently developed countries. a system of social classification — by age, lin-
The more severe population pressures in eage, clan, or caste. A value system that re-
the underdeveloped areas constitute another mains "tradition-oriented" also tends to min-
essential difference. Rates of population in- imize the importance of economic incentives,
crease in these areas are higher than those material rewards, independence, and rational
that generally obtained during the Western calculation. When the emphasis is on an es-
cases of development in the past. Even tablished pattern of economic life, family ob-
though not all the poor countries are now ligations, and traditional religious beliefs, the
densely populated, the rate of population individual may simply adopt the attitude of
growth is, or gives indications of soon becom- accepting what happens to exist, rather than
ing, a serious problem for most of them. And, attempting to alter it— an attitude of resig-
unlike the earlier cases in which population nation rather than innovation. Within an ex-
growth was induced by, or at least paralleled, tended family system or a village community,
a higher rate of development, the present individuals may resign themselves to accept-
growth in population is simply the result of ing group loyalties and personal relationships
the introduction of public-health measures, which remain in a stable and tradition-dom-
which lower death rates. These measures act inated pattern, assigning little importance to
as an autonomous factor, quite unrelated to material accomplishments and change. Even
the rate of internal development. Moreover, though they may have latent abilities, indi-
unlike the European industrial countries that viduals may lack the motivations and stimu-
began lowering their birth rates before their lations tointroduce change; there may not be
sharpest declines in mortality, the poor coun- sufficiently large groups in the society who
tries now will not do so until long after their are "achievement-oriented," concerned with
mortality has reached a modern low level. the future, and believers in the rational mas-
Given the fact that many poor countries are tery of nature. The positive value which the
already experiencing population growth rates traditional way of life still holds for many of
that are two to three times higher than those the people in a poor country inhibits the nec-
that confronted the currently rich nations essary orientation toward the future, and
when they were in their early phases of de- change is either resisted or, if accepted, is re-
velopment, and that other poor countries may stricted tofringe areas.
face a population problem in a relatively In short, the cultural context in many poor
short time, the need to attain increases in countries may not yet be as favorable to eco-
production sufficient to outstrip potential in- nomic achievement as it was in Western
creases inpopulation is now more acute than countries. This is not, of course, to assume
it ever was in Western countries at the begin- simply that, because the West is developed,
ning of their industrialization. Western values and institutions are therefore
Sociocultural and political differences also necessary for development, and that Western
 108 DEVELOPMENT PERSPECTIVES

cultural patterns must be imported into the individual country's development where that
poor countries. Many Western values and in- country stands relative to other countries.
stitutions may be only accidentally associ- This has already been alluded to by Ger-
ated with Western development, and many schenkron for European cases of develop-
values and institutions in poor countries are ment. With more direct reference to the
not obstacles to development. But though the problems of the presently poor countries,
West need not be imitated, some institutional Professor Streeten has argued persuasively
changes and modifications in the value struc- that the fact that advanced industrial soci-
ture are necessary if the inhibiting institu- eties already exist when countries embark on
tions and values are to be removed. To allow development, makes a number of important
poor countries to enter into the development differences to the development prospects of
process with a favorable cultural framework the less developed societies.9
as did the currently developed nations, there On the one side, the coexistence of rich and
must be changes in their cultures so that new poor countries now has a number of draw-
wants, new beliefs, new motivations, and new backs for the less developed countries. A
institutions may be created. Until these cul-
tural changes are forthcoming, an accelera- suggestive list follows.10
tion of development will be more difficult to 1. The most important difference is that
achieve than it was in the past. the advanced state of medical knowledge that
If the degree of sociocultural development can be borrowed from rich countries makes it
has been less than what occurred in the past, now possible to reduce deaths cheaply and
so too has there been a difference in political rapidly, without contributing to an equiva-
development. In many poor countries, the po- lent reduction in births — thereby presenting
litical foundations for developmental efforts the LDCs with more difficult population
are not yet as firm as they were in Western problems than the now advanced countries
development. Whereas the currently devel- faced in their pre-inustrial phase.
oped countries had already enjoyed a long pe- 2. Modern technology in rich countries
riod of political independence and a stable evolved under conditions of labor scarcity
political framework before their periods of and has therefore been designed to save labor
rapid economic development, most of the cur- in relation to capital. But the transfer of
rently poor countries have only recently ac- labor-saving technology to LDCs, which is
quired a real measure of political indepen- encouraged by attitudes toward moderniza-
dence. Political instability, undifferentiated tion and by the prestige of Western technol-
and diffuse political structures, and ineffi- ogy, tends to aggravate the underutilization
cient governments are still only too prevalent. of labor in the LDCs.
In some countries, government leadership has 3. The knowledge of organizations and in-
yet to be exercised by groups that do not have stitutions that prevail in the advanced coun-
vested interests in preserving the status quo; tries may be ill-adapted to the needs of
;n others, there is still a wide gap between the LDCs. The adoption of the trade union struc-
traditional mass and a modern elite which ture, for example, may be inappropriate for
controls the central structures of government conditions of labor surplus. Or public expen-
and is the main locus of political activity.8 diture on social welfare services developed in
All the foregoing differences might be sub- advanced industrial welfare states may be
sumed under a more general observation that premature for LDCs. Or large-scale business
it matters a great deal for the course of an enterprises may be undesirable in lesser-de-
9Paul P. Streeten, "The Frontiers of Development
8For a discussion of the difference between Western Studies: Some Issues of Development Policy," Jour-
and non-Western political systems, see G. A. Almond nal of Development Studies, October 1967, pp. 2-24.
and J. S. Coleman (eds.), The Politics of the Devel- 10Streeten fully elaborates this list and some other
oping Areas, Princeton, 1 960. differences, ibid., pp. 3-7.
 109 HISTORICAL PERSPECTIVES

veloped economies. The transfer of inappro- are some dissimilarities which, on the other
priate institutions to the LDCs may impede side, make the problems less difficult. Some
the development. advantages may accrue to presently poor
4. Technical progress in advanced econo- countries from their position of being late-
mies has harmed the trade prospects of the comers to development. Most helpful now
less developed countries that depend on the may be the ability of the poor countries to
export of primary products by facilitating the draw upon the accumulated stock of knowl-
substitution of synthetics for natural prod- edge in countries that have already devel-
ucts, by reducing the input of raw materials oped. Not only may improved productive
per unit of industrial output, and by shifting techniques and equipment be derived from
demand away from products with a high pri- these countries; more generally, they may
mary import content. benefit from the transference of ideas in the
5. The land-rich or capital-rich countries realm of social techniques and social innova-
which at one time served as an outlet for tion as well as technological. As we have al-
labor surplus countries are no longer recep- ready seen, however, the value of this imita-
tive to immigration; with accelerated popu- tive ability is debatable, since it is still
lation growth and the intensified underutili- necessary to modify and adapt — not simply
zation of labor, the development pressures imitate — the technological and social inno-
are thus all the greater. At the same time, the vations within the context of the borrowing
scarce resources of capital and skilled indi- country's environment. This problem receives
viduals are drained off to the rich centers and fuller treatment in Chapter V, below. And
away from the poor peripheral countries. aside from the requirements of readaptation,
Streeten has rightly emphasized that it is there also remains the ultimate difficulty of
this coexistence of rich and poor countries, having change accepted and integrated into
rather than the international or unintentional the recipient society. We should not, there-
exploitation of colonialism or neocolonialism, fore, accept too readily the view that by
which can have detrimental effects on devel- drawing upon the lessons and experiences of
opment efforts. And he properly concludes countries that have developed earlier, the
that it is this coexistence which sets limits to latecomers are in a position to telescope the
the ready transfer of the lessons of one his- early stages of development.
torical setting to the entirely different present The existence of many advanced countries
setting of poor countries vis-a-vis much more that have already reached a high level of de-
advanced countries. velopment, which was not the case when
Beyond the coexistence of rich and poor these same countries were in their preindus-
countries, we should also recognize that the trial phase, may now, however, help to ease
attempts of many poor countries to develop the development of poor countries by provid-
simultaneously may also intensify the task ing a flow of resources from the rich to the
for any one of the LDCs. The policies poor countries. Never in the past has there
adopted by each country in its effort to de- been so much international concern with the
velop may hinder the development of another desirability of increasing trade, technical as-
country. Thus the poor countries compete sistance, private foreign investment, and the
among themselves in attempting to increase flow of public funds as objectives of develop-
their exports, in attempting to attract private ment policy. But how effective foreign eco-
capital and skilled services, in attempting to nomic assistance may actually prove to be is,
foster industrialization via import-substi- of course, another matter. Some judgment on
tuion policies, and by way of other restrictive this may be had from Chapter V, below.
nationalistic policies which have beggar-my- Finally, there is now a strong conscious de-
neighbor (usually a poor neighbor) effects. sire for development on the part of national
Although these several differences now ag- leaders in many countries. The national in-
gravate the problem of development, there terest in deliberate and rapid development,
 10

the willingness of national authorities to opment. But in the final analysis, what will
assume responsibility for directing the decide whether a poor country will succeed is
country's economic development, and the whether its government can implement effec-
knowledge of a variety of policies that a gov- tively the possible policies that might make
ernment can utilize to accelerate develop- the country's development potential realiza-
ment— all these give new dimensions to ble, and whether its people are prepared to
the role that the State may play in the de- bear the costs that accelerated development
velopment of emerging nations. Through will necessarily entail. Regardless of whether
governmental action, to a degree unknown we interpret conditions in the currently poor
in Western development, a more favorable countries as being on balance more or less fa-
environment for a take-off might be created. vorable, we must not expect these countries
Nonetheless, as will be appreciated time and to follow simply the historical patterns of
again in subsequent chapters, the mere act of presently developed countries. We must still
development planning cannot be expected give due weight to the severity of the partic-
to remove the difficult choices and deci- ular problems confronting these countries.
sions that must be made to accelerate And we must determine what policies might
development. now be most effective in removing the bar-
Depending on how much importance we riers to development. With the benefit of his-
attach to each of the various differences be- torical perspective, however, we should be
tween past and present conditions, we may better able to appraise the significance of
reach contrasting conclusions as to whether these present-day development problems and
present conditions are more or less favorable their various policy implications.
than in the past for the acceleration of devel-
 II.B. ANALYTICAL
PERSPECTIVES

II.B. 1 - Contrasting Perceptions*

Somewhat oversimplifying, we can identify It was taken for granted that organizing the
five recent changes in the perception of the march along the development path was the
development process. Although the shift in prime concern of government. . . .
perceptions has been described as "recent," The linear view begged a host of questions
some of these changes go back a considerable about the nature, causes and objectives of de-
time. Indeed, hardly any of the features of velopment. Ittended to focus on constraints
the "old" perception were generally accepted or obstacles (particularly lack of capital), the
at any stage, and qualifications, criticisms removal of which would set free the "natu-
and alternative perceptions were put forward ral" forces making for the steady move to-
almost ward ever higher incomes.
had beenas formulated.
soon as the "orthodox" perception
Applied to the area of international rela-
1 . The first change between our perception tions, this view calls on the rich countries to
of development until about 1970 and after
can be traced to differences of view between supply the "missing components" to the de-
veloping countries and thereby to help them
the First World and the Third World, and break bottlenecks or remove obstacles. These
the Right and the Left, although there are all missing components may be capital, foreign
kinds of cross-alignments. The more popular exchange, skills or management. The doc-
(though not the academic) thinking of the trine provides a rationale for international
Fifties and Sixties, codified in the Pearson capital aid, technical assistance, trade, pri-
Commission Report, was dominated by W. vate foreign investment. By breaking bottle-
W. Rostow's doctrine of the stages of growth. necks, rich countries can contribute to devel-
This perspective was, partly, superseded in opment efforts a multiple of what it costs
the early 1970s by what has sometimes been them and thus speed up the development pro-
called a "dependencia" interpretation of in- cess in underdeveloped countries. Models
ternational economic relationships. To illu- pointing to gaps between required and avail-
minate the shift in perceptions from the first able savings or foreign exchange are a ration-
of these views to the second, let us review alization offoreign assistance. The ultimate
briefly each in turn. purpose of aid is to be rid of aid.
According to Rostow's doctrine, develop- This linear or stages-of-growth view has
ment is a linear path along which all coun- come under heavy fire. It was criticized on
tries travel. The advanced countries had, at logical, moral, historical and economic
various times, passed the stage of "take-off," grounds. Logically, it should have been clear
and the developing countries are now follow- that the coexistence of more- and less-ad-
ing them. Development "was seen primarily vanced countries is bound to make a differ-
as a matter of 'economic growth,' and sec- ence (for better or worse) to the development
ondarily as a problem of securing social efforts and prospects of the less advanced,
changes necessarily associated with growth. compared with a situation where no other
country was ahead or the distance was not
very large. The larger the gap and the more
*From Paul Streeten, "Development Ideas in His- interdependent the components of the inter-
torical Perspective," in Toward a New Strategy for
Development, A Rothko Chapel Colloquium, Perga- national system, the less relevant are the les-
mon Press, New York, 1979, pp. 25-35. Reprinted by sons to be learned from the early starters.
permission. Morally and politically, the linear view ruled

11
 1 12 DEVELOPMENT PERSPECTIVES

out options of different styles of development. rich and poor societies renders the efforts of
Inexorably, we were all bound to pass the poor societies to choose their style of de-
through the Rostovian stages, in the words of velopment more difficult or impossible. Cer-
the famous limerick, like a tram, not a bus. tain groups in the developing countries — en-
Historically, the view can be criticized as ex- trepreneurs, salaried officials, employees —
cessively determinist. Economically, it is de- enjoy high incomes, wealth and status and,
ficient because it ignores the fact that the constituting the ruling class, they perpetuate
propagation of impulses from the rich to the the international system of inequality and
poor countries alters the nature of the devel- conformity. Not only Marxists but also a
opment process; that late-comers face prob- growing number of non-Marxists have come
lems essentially different from the early to attribute a large part of underdevelopment
starters, and that "late late-comers" again and of the obstacles encountered in the pro-
find themselves in a world with a range of cess of development to the existence and the
demonstration effects and other impulses, policies of the industrial countries of the
both from the advanced countries and from West, including Japan and the Soviet Union.
other late-comers, which present opportuni- According to one line of this second view,
ties and obstacles quite different from those aid is not a transitional phenomenon to be
that England or even Germany, France and ended after "take-off," but a permanent fea-
Russia, faced in their pre-industrialization ture, like an international income tax. Ac-
phase. cording to a more radical line, aid is itself
Summarized briefly, it may be thought part of the international system of exploita-
that too much weight is given to Rostow and tion, and self-reliant, independent develop-
to the linear view of the development. The ment has to get rid of it.
"stages of growth" were criticized from the The conclusion drawn from this perception
beginning by Kuznets, Gerschenkron and is that the developing countries should put up
others. The heavy concentration on physical barriers between themselves and the destruc-
capital was criticized by Cairncross, Hirsch- tive intrusions of trade, technology, transna-
man, and the human capital school of T. W. tional corporations, and educational and ide-
Schultz. There were many nonlinear theo- ological influences, and should aim at
rists, from Schumpeter to Rosenstein-Rodan "delinking" or "decoupling," at pulling down
and Nurkse. The whole debate on balanced a bamboo or poverty curtain, at insulating
versus unbalanced growth does not fit into and isolating themselves from the interna-
the linear perception. But it remains true tional system.
that, though not in academic circles, the Ros- W. W. Rostow, for the first kind of percep-
tow model had a powerful grip on the imag- tion, and Andre Gunder Frank, for the sec-
ination ofpolicy makers, planners and aid of- ond, are the popular rather than the aca-
ficials and that it was this view that gave rise demic models. Prebisch, Singer, Myrdal,
to a reaction. Hirschman, and Perroux, not to say anything
Succeeding the linear development per- of Marx and List, had long ago developed ap-
spective inpopularity in the early 1970s was proaches to development that separated
a second view, according to which the inter- "spread" or "trickle-down" effects from "po-
national system of rich-poor relationships larization," "backwash" or "dominance" ef-
produces and maintains the underdevelop- fects. And many had raised doubts as to
ment of the poor countries (the rich "under- whether everything would be fine if all coun-
develop" the poor, in Andre Gunder Frank's tries only pursued free trade policies and es-
phrase). In various ways, malignly exploita- tablished competitive markets. But probably
tive or benignly neglectful or simply as a re- because of their more careful formulations,
sult of the unintended impact of events and the impact of their thinking, important
policies in rich countries, the coexistence of
though it was, remained regional — "periph-
 1 13 ANALYTICAL PERSPECTIVES

eral," not "mainstream" — and sales of their posed the question: Are the developing coun-
books did not reach the figures of A. G. tries better off, worse off, or would it make
Frank's. little difference?
Irrespective of whether the "new" percep- The answer to the question neatly sepa-
tion istrue, what matters is that many devel- rates the adherents of different "paradigms."
oping countries see their place in the inter- The upholders of the first "paradigm" would
national system in this way and their say "worse" (pointing to Korea, Singapore,
perception is a political fact to be reckoned and Hong Kong as beneficiaries from the in-
with. This clearly does not mean that the per- ternational system and Burma and Uganda
ception should not be subjected to a critical as losers from "closing themselves off"),
analysis. those of the second "better," and Sir Arthur
The transition between the two perceptions thought it would make little difference. But I
has been overdrawn. Perceptions alternate, submit that, whichever answer one is inclined
points emphasized change and there is no to give, this is not a helpful way of presenting
rapid, large-scale conversion. At about the the problem, however useful it is in sorting
same time that the critics of the international out ideologies. The developed countries prop-
economic order became more vocal, and ad- agate alarge number of impulses to the de-
vocated "delinking," there took place a re- veloping countries. Reasonable men may dif-
birth of orthodox thinking. The work on ef- fer about the net balance of these impulses,
fective protection by Johnson, Corden and e.g., whether the exploitation by ruthless
Balassa, the Organization of Economic Co- transnational companies offsets the availabil-
operation and Development (OECD) studies ity of a stock of scientific, technological and
of industrialization and trade centering on organizational knowledge, or whether the
the work by Little, Scitovsky and Scott, re- harm done by the brain drain is greater or
search bythe Brookings Institution and the less than the benefits from foreign technical
World Bank, and the doctrines of the Chi- assistance, or whether the inflow of grants
cago School that influenced many Latin and loans at concessionary interest rates is
American policy makers reflected a recoil counterbalanced by aid-tying and capital
against inefficient protectionism and "in- flight, etc. The interesting question then is
ward-looking" planning. The conclusions not "Do the developing countries benefit or
pointed to more "outward-looking" policies. lose from their coexistence with developed
These changing analyses and perceptions al- countries?" but, "How can they pursue selec-
ternate and interact. tive policies that permit them to derive the
A reconciliation between the two percep- benefits of the positive forces, without simul-
tions (viz., that development can be speeded taneously exposing themselves to the harm of
up by the international "system" and under- the detrimental forces?" Looked at in this
development icaused
s by it) is possible along way, the question becomes one of designing
the following lines. The advanced industrial selective policies for aid, trade, foreign in-
countries emit a large number of impulses of vestment, transnational companies, technol-
two kinds: those that present opportunities ogy, foreign education, movements of people,
for faster and better development than would etc.
otherwise have been possible, and those that
present obstacles to development, those that 2. The shift from a linear theory of missing
stunt growth. Arthur Lewis invited us in components to some version of a theory of
1974 to imagine that the developed countries neocolonialism and dependence was accom-
were to sink under the sea in 1984. (He gave panied byanother change. It amounted to a
us ten years in order to allow time for ad- change in emphasis of what constitutes the
justments. He felt it necessary to add that meaning and measure of development.
this was not a recommendation.) He then Early thinking and policy making was
 1 14 DEVELOPMENT PERSPECTIVES

dominated by economic growth as the prin- ternational Labor Office (ILO) Kenya mis-
cipal performance criterion of development, sion suggested the "informal sector" as not
not so much because growth was regarded as just another name for disguised unemploy-
an appropriate objective in itself, but because ment but as a potentially productive labor
it was thought either that its fruits would force. The new emphasis on the "working
rapidly trickle down to the poor, or that cor- poor" led to a concern for redistribution of
rective government action could be relied productive assets as a path to reduced in-
upon to redistribute them, or that inequality equality. The relation between the increased
and poverty are essential for growth (which, concern for equity and conventionally mea-
through accumulation by the rich, would first sured economic growth presented a dilemma.
have to create the productive base from On the one hand, it was accepted that in poor
which to launch the attack on poverty). But societies poverty can be eradicated only
is was soon evident that none of these three through increased production. On the other
assumptions was valid. It became clear that hand, the growth experience in some coun-
growth in many countries remained concen- tries (though not in all) has shown that
trated on a narrow enclave of modern, orga- growth had reinforced inequalities in income,
nized, urban industry; that governments asset and power distribution which made it
often were unwilling or incapable of using more difficult or impossible, both economi-
taxes and services to offset growing inequal- cally and politically, for its benefits to be
ities; and that the concentration of income in spread widely. Attention, therefore, shifted
the hands of the rich was not a necessary con- to the conditions under which "redistribution
dition of development. (For example, re- with growth" is possible and desirable.
search has disclosed that small farmers with The next step was to realize that what was
access to improved agricultural technologies needed was a more direct and speedier attack
save as high a proportion of their incomes as on deprivation. Reductions in inequality do
large farmers and are often more efficient in not necessarily reduce poverty. It was not just
terms of yield per acre.) inequality as such that was offensive, but also
The expected absorption of the rapidly the fate of the destitute, whether working or
growing labor force from the subsistence sec- incapable of working, whether unemployed,
tor into the modern, industrial sector was underemployed or unemployable. The objec-
considerably slower than expected. Dualism tive narrowed down to "meeting basic human
in many countries was marked and pro- needs," which covers not only adequate earn-
longed. The golden age of growth with ing opportunities to purchase the necessities
greater equality, ushered in after a period of of life, but also access to the provision of pub-
growing inequality, or by the full absorption lic services for education, health and safe
of labor from the subsistence sector into the water. The progress was from highly aggre-
modern sector, seemed, with a few important
gated magnitudes like "national income" and
exceptions, to move into the distant future. "growth rates" to increasingly disaggregated
This awareness led to a new emphasis on objectives, like different types of employment
rural development and "employment." It was (e.g., for the young, for recent migrants) and
soon seen, however, that the problem was not reduced inequality, to meeting highly specific
"unemployment," which is a Western con- human needs of particular poverty groups.
cept that presupposes modern sector wage In focusing on these needs, it became clear
employment, labor markets, labor exchanges that measured income and its growth is only
and social security payments in the form of a part of basic needs. Adequate nutrition and
unemployment benefits. (It has been said safe water at hand, continuing employment,
that only those relatively well-off can afford secure and adequate livelihoods for the self-
to be unemployed.) The problem was rather employed, more and better schooling for
unremunerative, unproductive work of the their children, better preventive medical ser-
poor, particularly of the rural poor. The In- vices, adequate shelter, cheap transport and
 1 15 ANALYTICAL PERSPECTIVES

(but not only) a higher and growing level of Whereas, with the growing interest in differ-
measured income: some or all of these would ent societies in the Sixties, many had argued
figure on the list of urgently felt needs of poor
that there is no single, universal "science" of
people. economics, applicable from China to Peru,
In addition to these specific "economic" and Dudley Seers wrote on "the economics of
objectives, there was now a new emphasis on the special case," the wheel has now turned
"nonmaterial" needs that cannot be dis- full circle and we (with Dudley Seers in the
pensed, but, in addition to being valued in lead) now acknowledge that many of the is-
their own right, may be the conditions for sues that we had considered as belonging to
meeting "material" needs, like self-determi- the poor countries are seen to be universal, of
nation, self-reliance, political freedom and concern to the rich, too. Alienation, pollution
security, participation in making the deci- and crime can spring from underdevelopment
sions that affect workers and citizens, na- as well as from overdevelopment; "interme-
tional and cultural identity, and a sense of diate technology" is just as relevant to high-
purpose in life and work. This was accompa- income societies suffering from unemploy-
nied by attempts to evolve human and social ment in the face of resource limitations, pol-
indicators of development that would reflect lution and alienation from work; all states
the extent to which some of these needs were confront the new phenomenon of the trans-
met. national corporation; we are all affected by
the energy crisis; there are diseases of afflu-
3. A third shift in interest and emphasis ence as well as diseases of poverty; migration
was away from the specific economic prob- of workers and professionals affects rich and
lems of development and toward the world's poor countries; there is global experimenta-
common problems and shared constraints: re- tion with new lifestyles; there is a heritage
sources and, in particular, energy, the envi- common to all mankind and the unexploited
ronment and its global pollution, the sea and resources of the sea and the sea bed can be
the sea bed, and world population. The new allocated according to worldwide priorities.
emphasis was on scarcity and interdepen-
dence. 4. A fourth and closely related change is
The new emphasis on interdependence was that from a tacitly or explicitly assumed in-
regarded by some as calling for greater soli- ternational harmony of interests (which was
darity and cooperation ("spaceship earth," built into the stages-of-growth model) and
"one world"), by others, such as the advo- positive-sum games, to greater emphasis on
cates of triage and of the lifeboat philosophy, actual or potential conflict and zero-sum
as calling for partial contracting out of games. There was much talk of cooperation
human obligations, and by others again as a versus confrontation. The new perception
shift of attention from positive-sum to zero- was brought out vividly by the actions of the
sum games. The renewed emphasis after Organization of Petroleum Exporting Coun-
1972 on scarcities of resources and exhaus- tries (OPEC) and by the Intergovernmental
tion of raw materials extended not only to Council of Copper Exporting Countries
food, energy and certain metals, but also to (CIPEC) as well as similar attempts with
some previously free goods, like clean air and bauxite, phosphate and iron ore; by attempts
clean water, with the resulting concern for or threats of the developing countries to use
environmental protection. bargaining power in other fields with refusals
Interdependence does not necessarily point to participate in drug control, or the control
to solidarity; on the contrary, it may give rise of nuclear weapons, or patent conventions; by
to threats, "blackmail," the demand for ran- threats of expropriation or tougher terms for
soms, and attempts at isolation. But whatever transnational companies; and by the support
the response, the fact is that many problems by a few governments of terrorists and hi-
are now global and are shared by all people. jackers. As the old harmony has benefited the
 1 16 DEVELOPMENT PERSPECTIVES

haves, the new confrontation was intended to Conference on Trade and Development
be used to change the rules of the game in (UNCTAD) and other international forums.
favor of the have-nots. Gaps in income per head opened up more
The fact that economists point to potential widely within the Third World than between
joint gains from, say, trade liberalization or the developed and the developing countries.
commodity price stabilization, and thus to The unprecedentedly high average rate of
positive-sum games does not, of course, mean growth of income per head in the developing
that governments perceive such policies as in countries between 1950 and 1975 masked a
their national interest. Policies are shaped by wide diversity of experience. In Korea, Tai-
pressure groups and lobbies and the ubiquity wan, Hong Kong, Singapore, some OPEC
of protection testifies to the power of these in- countries and the People's Republic of
terests. On the other hand, developing coun- China, income per head rose rapidly and the
tries may on occasion have an interest in fruits of this growth were fairly evenly dis-
show-downs rather than negotiations of com- tributed (though the high growth rates for
mon interest. There are political advantages China are crucial and disputed). This group
in the publicity of such confrontations. In any of countries contains about one billion people,
case, sovereign political power is in its nature or 35 percent of the population of the devel-
a zero-sum game and the added strength of oping countries.
one country is often perceived ipso facto as a In a second group of countries, including
defeat for a rival. Pakistan, the Philippines, Thailand, Turkey
At the same time, negotiations do appeal, and many Latin American countries, espe-
if not to common interests, at least to com- cially Brazil, containing about 25 percent of
mon norms and to acceptable rules. Most na- the developing world's population, moderate
tions realize that a world in which each na- to rapid growth was combined with growing
tion and group of allied nations exercises to inequality, though it is controversial whether
the full its bargaining power to extract max- absolute poverty increased.
imum concessions from others is a world in In a third group of countries, comprising
which most nations will be worse off. Bar- some of the large countries of South Asia
gaining must, therefore, appeal not only to (India, Bangladesh) and some of the poorer
national self-interest but also to widely ac- countries of Africa (incorporating about 40
ceptable principles, rules, and norms on percent of the developing countries' popula-
which tacit agreement can either be assumed tion) slow growth was probably combined
or assumed to be more readily reached than with growing poverty in absolute terms. This
is possible with selfish demands and threats. group contains some of the largest countries
If bargaining is seen and conducted in this and it remains a disputed issue whether the
light, its power to disintegrate the world com- proportion of absolutely poor increased.
munity isgreatly weakened, even if the game The distribution of the benefits of the com-
is, in economic terms, a zero-sum one. Prop- modity boom, especially the oil price rise, and
erly used, bargaining can strengthen world the incidence of the damage done by the
cooperation through joint attempts to evolve world recession were highly uneven and fur-
a more acceptable set of rules and insti- ther sharpened differences within the Third
tutions.
World. The hoped-for benefits of the New In-
ternational Economic Order (NIEO) also are
5. A fifth change was that from treating likely to be unequally distributed.
the Third World as a homogenous group of
countries with common interests to the ac- 6. The sixth change in the perception of the
knowledgement ofthe wide variety of expe- development process, closely related to the
riences, interests and stakes in the world other five changes, is that from abounding
order — in spite of the growth of solidarity of optimism about development prospects and
the "Group of 77" at the United Nations the contribution to them by the rich (through
 117 ANALYTICAL PERSPECTIVES

aid, trade and private investment) and by rid of the need for future aid. When faced
economic analysis, which dominated the Fif- with the real problems and difficulties of de-
ties and Sixties, to the deep pessimism of the velopment, which transcended the economic
Seventies. Both optimism and pessimism variables — and, at the same time, in their
have their social origins. The optimism of the own countries, faced with what has come to
early decades, as Gunnar Myrdal has pointed be known as "stagflation," together with a
out, had its origin not only in the excitement host of new problems such as urban ghettoes,
of the discovery of new areas but also in the student unrest, a growing number of indus-
desire of the governments in the industrial trial strikes, drug addiction, racial tensions,
countries to please the new elites in the newly etc. — the pessimism of the rich countries was
independent countries and to reinforce the a convenient excuse for falling and failing
view that transfer of capital, skills and tech- commitments to development cooperation
nology from rich to poor countries will soon and for contracting out of or for reducing
lead to self-sustained growth and thereby get contributions.
Comment
Some of the earlier theories of economic growth — in particular, those of the classical econ-
omists, Marx, neoclassicists, and Schumpeter — are reviewed in G. M. Meier and R. E. Bald-
win, Economic Development: Theory, History, Policy (1957), Part 1. An outstanding inter-
pretation ofanalyses of economic development in the history of thought is provided by Lord
Robbins, The Theory of Economic Development in the History of Economic Thought (1968).
After World War II, the subject of "development economics" was formulated in its own
right. Investigating issues that went beyond the earlier growth economics of classical econo-
mists (Smith, Malthus, Ricardo), a "New Development Economics" began to be formulated
by a number of writers. Outstanding contributions were made in the 1940s and 1950s by:
Colin Clark, The Conditions of Economic Progress (1940); P. N. Rosenstein-Rodan, "Prob-
lems of Industrialization of Eastern and Southeastern Europe," Economic Journal (June-
September 1943); Kurt Mandelbaum, The Industrialization of Backward Areas (1945); Raul
Prebisch, "The Economic Development of Latin America and its Principal Problems," Eco-
nomic Bulletin for Latin America, Vol. 7 (1950); P. C. Mahalanobis, "Some Observations on
the Process of Growth in National Income," Sankhya, Vol. 12, No. 4 (1953); Ragnar Nurkse,
Problems of Capital Formation in Underdeveloped Countries (1953); Jacob Viner, Interna-
tional Trade and Economic Development (1953); Hla Myint, "An Interpretation of Economic
Backwardness," Oxford Economic Papers (June 1954); W. Arthur Lewis, The Theory of Eco-
nomic Growth (1955); P. T. Bauer and B. S. Yamey, The Economics of Under-Developed
Countries (1957); Gunnar Myrdal, Economic Theory and Underdeveloped Regions (1957);
Albert Hirschman, Strategy of Economic Development (1958); Jan Tinbergen, The Design of
Development (1958).
Comment
Several retrospective papers consider the evolution of development thought over the past
three or four decades. Most useful in tracing the changing views of development economists
are the following: P. P. Streeten, "Development Ideas in Perspective," in Toward a New Strat-
egy for Development, A Rothko Chapel Colloquium (1979); Gustav F. Papanek, "Economic
Development Theory: The Earnest Search for a Mirage," in Manning Nash (ed.), Essays on
Economic Development and Cultural Change in Honor of Bert F. Hoselitz (1977); Gustav
Ranis, "Development Theory at Three Quarter Century," in Manning Nash (ed.), Essays on
Economic Development and Cultural Change in Honor of Bert F. Hoselitz (1977); Albert O.
Hirschman, "The Rise and Decline of Development Economics," in Essays in Trespassing
(1981); Fernando Henrique Cardoso, "The Originality of a Copy: CEPAL and the Idea of
 1 18 DEVELOPMENT PERSPECTIVES

Development," CEPAL Review (1977); Dudley Seers, "The Birth, Life and Death of Devel-
opment Economics," Development and Change, Vol. 10 (1979); Hans Singer, "Thirty Years
of Changing Thought on Development Problem," in Rich and Poor Countries (1977), chapter
13; Jere R. Behrman, "Development Economics," in Sidney Weintraub (ed.), Modern Eco-
nomic Thought (1974); Ian Livingstone, "The Development of Development Economics,"
ODI Review, No. 2, 1981; Amartya Sen, "Development Which Way Now?" Economic Jour-
nal (December 1983).

II. B. 2. Structuralist Approach to

Development Policy*

Several approaches to the analysis of devel- in the 1960s casts some doubt on the signifi-
oping economies have evolved over the past cance of the structural problems that had
twenty-five years. From a methodological been identified in the previous decade. How-
standpoint, they can be grouped under three ever, in the past few years the importance of
main headings: neoclassical, neo-Marxist, structural rigidities has been reemphasized
and structuralist. The first two attempt to by several new phenomena: the limited abil-
adapt systems of thought that were initially ity of economies to absorb the growing labor
formulated for the study of industrial soci- force, the worsening of the income distribu-
eties to the less developed countries. The tion in several developing countries, and —
structuralist approach attempts to identify most recently — the disruption to world trade
specific rigidities, lags, and other character- caused by increased oil and food prices,
istics of the structure of developing econo- which will require a substantial adjustment
mies that affect economic adjustments and in productive structures. In short, develop-
the choice of development policy. ment policy again seems to be constrained by
The initial set of structural hypotheses was a number of structural factors that require a
formulated in the 1950s by writers such as more explicit analysis of the possibilities for
Paul Rosenstein-Rodan, Ragnar Nurkse, W. short-term adjustment and for longer term
Arthur Lewis, Raul Prebisch, Hans Singer, changes in the economic structure itself. . . .
and Gunnar Myrdal. They explain phenom-
ena such as balance of payments disequilib-
rium, unemployment, and worsening income METHODOLOGY
distribution on the basis of particular prop-
erties of demand and production functions The methodology of structural analysis has
and other specifications of economic behav- evolved over the past twenty-five years from
ior. A common theme in most of this work is a set of rather intuitive hypotheses to models
the failure of the equilibrating mechanisms of increasing empirical validity and analyti-
of the price system to produce steady growth cal rigor. This evolution can be summarized
or a desirable distribution of income. in three stages: formulation of hypotheses,
The success of a number of developing empirical testing, and the elaboration of
countries in accelerating their rates of growth more complete models. This sequence can be
illustrated for two of the basic elements of
structuralist systems: the concept of a dual
*From Hollis B. Chenery, "The Structuralist Ap- economy and the concept of complementarity
proach to Development Policy," American Economic
Review. Papers and Proceedings, May 1975, pp. 310- in demand, which underlies theories of bal-
315. Reprinted by permission. anced growth.
 119 ANALYTICAL PERSPECTIVES

The concept of a dual economy stems from sumed inelastic demand for exports needs to
the observation that development takes place be seriously qualified.
unevenly both within and between sectors of The third stage of theoretical refinement
an economy. Although this concept has had and policy application has proved more diffi-
many different formulations, the most influ- cult. In the first place, it has been shown that
ential isthat of Lewis. He makes three basic the structural relations posited are not suffi-
assumptions as to the structure of a develop- cient to lead to some of the policy conclusions
ing economy: (1) that technology can be di- suggested in the original formulations. As in
vided between capital-using (capitalist) and the case of the Keynesian assumptions, a
non-capital-using (subsistence); (2) that the more complete formulation of models that
labor supply is elastic at a conventional wage; can be statistically estimated has proved nec-
and (3) that saving is done largely by the re- essary inorder to reach useful policy conclu-
cipients of nonwage income (capitalists). sions. Much current work consists in devel-
These assumptions, or variants of them, have oping a second generation of models in the
been incorporated in models that explain the structuralist tradition that are designed for
acceleration of growth, the allocation of the statistical application in individual countries,
labor force, and changes in the distribution of rather than for deriving broad generali-
income. zations.
The early formulations of concepts of bal- The structuralist approach has had a sub-
anced growth by Nurkse and Rosenstein- stantial impact on both internal and external
Rodan also relied on a simple set of struc- development policies. In both instances it fo-
tural hypotheses: (1) a generalized version of cuses on identifying the consequences of var-
Engels' law, specifying that consumer de- ious kinds of structural disequilibria. In do-
mand for food, clothing, shelter, and other mestic policy, the principal phenomena
major commodity groups is mainly a function examined have been the effects of surplus
of income and little affected by relative labor on resource allocation and more re-
prices; (2) a similar assumption as to the lim- cently the interpretation of worsening income
ited price elasticity of demand for exports; distribution as resulting from a set of dis-
and (3) in Rosenstein-Rodan's formulation, equilibrium conditions. In international pol-
the importance of economies of scale in over- icy, analysis has focused on the nature of
head facilities and basic industries. The first structural disequilibrium in the balance of
two assumptions make it necessary to expand payments and its effect on trade and aid pol-
output and allocate investment in close rela- icies. The following sections examine some of
tion to the pattern of domestic demand. They the relationships between the methodology
also provide an explanation for structural dis- employed and the policy conclusions reached
equilibrium and slow growth in countries in these two areas.
that fail to do so.
Both sets of assumptions have in general
stood up well to subsequent empirical tests. INTERNAL DEVELOPMENT
The acceleration of population growth has POLICY
probably made the surplus labor assumption
more generally valid today than when it was Unlike the neoclassical assumptions, the
initially formulated for underdeveloped structuralist alternatives do not lead auto-
countries. Econometric tests by R. WeisskofT matical y to policy conclusions. To produce
and C. Lluch and A. Powell provide some such conclusions, they must also be embodied
support forthe generalized version of Engels' in an explicit general equilibrium framework.
law underlying the theory of balanced For this purpose, most analysts have used one
growth, since they show most price elastici- of two simple models: either a neoclassical
ties to be less than unity. Among the basic model with particular structural relations
structuralist assumptions listed, only the as- added or some version of a linear Leontief
 120 DEVELOPMENT PERSPECTIVES

input-output model, which excludes most terized bythe alternating pattern of expan-
forms of substitution. The neoclassical sion of production and imports described by
framework minimizes the effects of the spe- T. Scitovsky and by Chenery and L.
cific rigidities in the economic system, while Westphal.
the input-output system tends to exaggerate The elaboration of structuralist hypotheses
them. in planning models has also focused attention
Most elaborations of the dual economy and on the value of flexibility in adapting re-
surplus labor concepts have been made in a source allocation to changing circumstances.
two-sector neoclassical system, as in the work This problem does not arise in the neoclassi-
of J. C. Fei and G. Ranis, L. Lefeber, and A. cal system, which assumes perfect foresight
C. Kelley, J. G. Williamson and R. J. Chee- and a high degree of substitutability. When
tham. One policy result from this type of these assumptions are abandoned, flexibility
model is to determine the shadow price of can be provided by increased exports or cap-
labor, which can then be used in project eval- ital inflows and by planning some excess ca-
uation and in establishing the need for labor pacity in physical and human capital stocks.
subsidies. Empirical applications of these Although a formal treatment of the benefits
models have been limited by the lack of data of flexibility and the ways of achieving it has
on different aspects of dualism. It will require not been developed, this is an important prob-
some expansion of the two-sector framework lem for developing countries which cannot be
before statistical estimation of the underlying analysed in the neoclassical system.
structural relations becomes feasible. The revival of interest in income distribu-
The balanced growth hypotheses have tion has added a new dimension to structural
been widely used in empirical analysis and analysis. The traditional approach focuses on
are usually incorporated in input-output the division between wage and nonwage in-
models which include foreign trade. These come and is ill suited to developing countries,
are applied to the formulation and testing of in which modern sector wage earners are in
development plans, as illustrated in surveys the middle-income groups. Recent studies by
by Chenery (1971) and A. Manne. Apart M. S. Ahluwalia have brought out the fact
from its use in country planning, this type of that the bulk of the poorest groups in devel-
model can also be used to deduce more gen- oping countries are self-employed and largely
eral propositions through systematically rural. Their incomes depend more on the
varying the structural parameters and deter- availability of land and capital and access to
mining aset of solutions based on either sim- public facilities than on wages. Since each of
ulation or optimization. the main poverty groups — small farmers,
This form of sensitivity analysis lends con- landless laborers, urban self-employed — has
siderable support to some of the conclusions a different set of productive possibilities and
of balanced growth theorists. For example, it constraints, a new form of structural analysis
is necessary in larger countries to expand ag- based on identification of these groups is nec-
riculture at a rate that is largely determined essary for distribution-oriented policies.
by the income elasticity of domestic demand This recognition of the importance of asset
for foodstuffs because of the limited possibil- distribution does not necessarily lead to the
ities of expanding nonagricultural exports to Marxist conclusion that the redistribution of
offset a shortfall. Failure to meet this condi- existing assets is the only alternative. Mea-
tion has retarded growth in a number of sures to redistribute increments in income
countries. This conclusion does not apply and new asset formation are more likely to be
when industry is disaggregated to individual acceptable to the majority of the population
sectors such as steel or fertilizer, however, in and less disruptive of development in most
which the optimal investment allocation to countries. Development strategies based on
sectors having economies of scale is charac- this approach are elaborated by Ahluwalia
 121 ANALYTICAL PERSPECTIVES

and Chenery. The neo-Marxist policy rec- inflows in bottleneck situations, in which the
ommendations suffer from the same defects trade limitation is more restrictive than sup-
as the neoclassical in that they are implicit in plies of capital and skilled labor; (2) a re-
the initial assumptions rather than being de- statement ofthe empirical basis for assessing
rived from an analysis based on empirical es- comparative advantage over time in relation
timates ofthe underlying structural relations. to the factors limiting development; and (3)
clarification of the relation between internal
and external constraints in the evaluation of
EXTERNAL POLICY individual investment projects.
During the 1960s a number of countries
The conflict among the three analytical ap- progressed from an initial strategy of import
proaches isperhaps most acute in the area of substitution to the promotion of manufac-
external policy. The neoclassical approach tured exports after they had developed a suf-
tends to exaggerate the benefits of trade in an ficient industrial base to do so. As countries
open economy when it does not explicitly con- achieve a more diversified productive struc-
sider the effects of uncertain export prices ture and reduce their concentration on a few
and the difficulties of shifting resources to exports, the difference between the neoclass-
meet changing market conditions. Con- ical and structural prescriptions diminishes
versely, the neo-Marxist approach exagger- because some of the constraints that had pre-
ates the costs of "dependence" on external viously limited growth are no longer signifi-
trade and investment and tends to ignore the cant. An assessment of the experience of
benefits of the technological transfers that some of the industrializing countries that
accompany them. The early structuralist have made this transition — Taiwan, Israel,
views have also proved to be excessively pes- Korea, Brazil, Mexico — suggests that earlier
simistic as to the possibilities and benefits of conclusions as to the high cost of the initial
nontraditional exports of manufactures and stage of import substitution need to be reex-
services. amined in light of the subsequent ability of
The structuralist concept of development the country to shift to manufactured exports
as characterized by rigidities that limit eco- and break out of the phase of trade limited
nomic adjustments requires an analytical development. This evaluation suggests that
framework in which external policy is more the dichotomy between inward-oriented and
closely linked to domestic resource allocation outward-oriented policies has perhaps been
than does the neoclassical view, which mini- overdrawn, and that these policies can be
mizes these restrictions. Attempts to formal- more usefully viewed as sequential elements
ize these relationships started from simple of a strategy designed to bring about changes
two-gap models, which incorporate explicit in the structure of both production and trade.
limits on the rate of increase of domestic sav-
ing, investment, and exports. These models
were elaborated by including optimizing pro- CONCLUSIONS
cedures and shadow prices, the disaggrega-
tion of productive sectors, and more explicit The preceding discussion has illustrated
treatment of structural change over time, as the relations between theoretical premises
in the work of Chenery and A. MacEwan, M. and policy implications in several fields of de-
Bruno, and S. D. Tendulkar. Experiments velopment. The simplifying assumptions of
with these types of models in a number of the models currently in use tend to exagger-
countries have led to several general conclu- ate the differences between neoclassical and
sions as to development analysis and policy, structuralist prescriptions. As statistically
the most significant of which are: (1) the en- determined relations replace a priori hy-
hanced value of increased exports and capital potheses, itis predictable that these differ-
 122 DEVELOPMENT PERSPECTIVES

ences will be reduced. A similar process can greater weight in development policy to dis-
be hoped for when neo-Marxist theorists turn tributional considerations, this cannot be
their attention to verifying their hypotheses. achieved without giving equal priority to ad-
Although better knowledge of elasticities justments inexternal trade and capital flows.
of substitution in demand and production
would do quite a bit to reduce the conflicts in
policy guidance, several real differences in
basic concepts remain. Neoclassical policy References
consists essentially in removing impediments M. S. Ahluwalia, "Income Inequality: Some
to the functioning of markets so as to make Dimensions of the Problem," in H. Che-
the real world as much like the abstract nery et al., Redistribution with Growth,
model as possible. However, it will never be London, 1974.
possible to achieve perfect knowledge or in- and H. Chenery, "A Model of Dis-
stantaneous adjustment to market signals. It tribution and Growth," in H. Chenery et
is therefore necessary to incorporate these al., Redistribution with Growth, London,
1974.
"imperfections" into the model itself. Once
this has been done, it will become possible to M. Bruno, "Optimal Patterns of Trade and
take account of the existence of internal or
Development," Review of Economic Sta-
external disequilibria and to devise more re- tistics, November 1967, Vol. 49, pp. 545-
alistic policies to cope with them. In the the- 54.
oretical literature, these policies are mislead- H. Chenery, ed., Studies in Development
ingly Planning, Cambridge, Mass., 1971.
to the referred
neoclassical to as model.
"second Itbest"
wouldin be
relation
more
and A. MacEwan, "Optimal Pat-
accurate to characterize the model itself as terns of Growth and Aid: The Case of Pa-
overly simple and "first best" policies as sim- kistan," inI. Adelman and E. Thorbecke,
ply unattainable. More attention should be eds., The Theory and Practice of Economic
given to improving the realism of basically Development, Baltimore, 1966.
neoclassical models instead of discarding and L. Westphal, "Economies of
them in favor of equally oversimplified struc- Scale and Investment Over Time," in J.
turalist formulations. Margolis, ed., Public Economics, London,
In minimizing or ignoring the advantages 1969.
of market adjustments, structuralist (and J. C. Fei and G. Ranis, Development of the
Marxist) policy prescriptions usually put too Labor Surplus Economy, Homewood, 111.,
much weight on the limited administrative 1964.
apparatus of developing countries. As A. A. Hirschman, The Strategy of Economic
Hirschman has stressed, this is one of the Development, New Haven, 1958.
main limitations to development; it should be A. C. Kelley, J. G. Williamson and R. J.
allowed for by not seeking too much fine tun- Cheetham, Dualistic Economic Develop-
i-ig of development policy. Difficulties in im- ment, Chicago, 1972.
plementing a complex set of policies may L. Lefeber, "Planning in a Surplus Labor
prove much more costly than the allocative Economy," American Economic Review,
inefficiency of a simpler program that can be June 1968, Vol. 58, pp. 343-73.
more readily carried out. W. A. Lewis, "Economic Development with
Finally, it must be recognized that the task
of development has been made much more Unlimited Supplies of Labor," Manchester
School, May 1954, Vol. 22, pp. 132-91.
difficult for most countries by the recent C. Lluch and A. Powell, International Com-
changes in the world economy. Virtually parisons ofExpenditure and Saving Pat-
every country is currently suffering from terns, Int. Bank for Recon. and Dev., Res.
more or less serious disequilibria in its eco- Cent., dis. pap. no. 2, 1973.
nomic structure. Despite our desire to give A. Manne, "Multi-Sector Models for Devel-
 123 ANALYTICAL PERSPECTIVES

opment Planning: A Survey," Journal of locations ofEconomic Resources, Stan-

Development Economics, June 1974, Vol. ford, 1959.
1, pp. 43-70. H. W. Singer, "The Distribution of Gains
G. Myrdal, Economic Theory and Underde- Between Investing and Borrowing Coun-
veloped Regions, London, 1957. tries," American Economic Review, Papers
R. Nurkse, Problems of Capital Formation in and Proceedings, May 1950, Vol. 40, pp.
Underdeveloped Countries, Oxford, 1953. 473-85.
R. Prebisch, "Commercial Policy in the S. D. Tendulkar, "Interaction Between Do-
Underdeveloped Countries," American mestic and Foreign Resources in Economic
Economic Review, Papers and Proceedings,
Growth: Some Experiments for India," in
May 1959, Vol. 49, pp. 251-73. H. Chenery, ed., Studies in Development
P. Rosenstein-Rodan, "Problems of Indus- Planning, Cambridge, Mass., 1971.
trialization ofEastern and South-Eastern R. Weisskoff, "Demand Elasticities for a De-
Europe," Economic Journal, June-Sep- veloping Economy: An International Com-
tember 1943, Vol. 53, pp. 205-16. parison of Consumption Patterns," in H.
T. Scitovsky, "Growth — Balanced or Unbal- Chenery, ed., Studies in Development
anced," inM. Abramovitz et al., The Al- Planning, Cambridge, Mass., 1971.

II.B.3. Changes in Development

Theory *

It is my purpose in this section to trace some "trickle-down" is likely to work, and a re-
of the main advances in our understanding of lated body of more policy-tinged expert opin-
the development process during the past cou- ion to the effect that employment and income
ple of decades and to relate them, where ap- distribution objectives should be met via
propriate, to more recent changes in the "secondary strategies," for example, public-
objective function articulated by most devel- works programs and fiscal redistribution.
oping-country spokesmen and aid officials. Trickle-down is still sufficiently respectable
Perhaps the most important dimension of to have been the basis of the so-called Pre-
conceptual progress, in my opinion, is the bisch Report, 1970s vintage, calling for
growing awareness that the analysis of higher growth rates in Latin America in
growth, employment, and distribution must order to pull in the unemployed; and the pre-
be viewed as integrally of one cloth, with the ponderance ofwork on income distribution
focus on the existence and size of trade-offs and poverty still emphasizes the potential re-
among these objectives. The notion that em- distributive effects of tax and expenditure
ployment and distributional issues are best policies, nationalization, and public-works
treated "after the fact," that is, after all the programs.
production/allocation dust has settled, has The empirical evidence that has been ac-
died hard. There is still a substantial body of cumulating, on the other hand, indicates that
theoretical literature which claims that even if fiscal redistribution, after the fact,
were politically and administratively feasible,
it would have to assume completely unreal-
*From Gustav Ranis, "Development Theory at istic proportions to make a real difference.
Three Quarter Century," in Essays on Economic De- And, turning the problem on its head, it
velopment and Cultural Change in Honor of Bert F. would take an unreasonably large exogenous
Hoselitz (ed. by Manning Nash, University of Chi-
cago Press, 1977), pp. 256-66. Reprinted by shift in income distribution to achieve any-
permission. thing meaningful in the way of a more em-
 124 DEVELOPMENT PERSPECTIVES

ployment-intensive output mix. Moreover, creased willingness to sector the typical de-
the advice that an increase in overall growth veloping economy into meaningful compo-
targets from 5 to 8 percent would solve the nents for purposes of general equilibrium
unemployment problem is weak on two analysis. The literature on economic, social,
grounds — it is highly impractical to locate and technological dualism has a long and dis-
the additional fuel to make the old Model T tinguished history. But the revival of classical
move that much faster and, perhaps more im- economics after Arthur Lewis's pathbreaking
portantly, even if sufficient natural resources work, and its application to contemporary
and/or foreign capital could be harnessed, problems of development, has represented a
the accompanying income distribution out- major analytical advance — permitting us to
comes are not necessarily acceptable. Finally, trace the interactions over time among sec-
intrinsic administrative and organizational tors not homogeneous in structure and/or be-
difficulties, at least in the mixed-economy havior. More recently, it has been recognized
LDC context, make a solution via major re- that so-called two-sector models need to be
liance on a massive public-works program modified by extension to three or four sectors,
highly impractical. for example, along the lines of more than one
It is for these reasons that I count the grad- traditional and more than one commercial-
ually growing consensus that these new di- ized sector, with possibly two urban and two
mensions of development must be analyzed rural components emerging. While contro-
and solved as an integral part of the old as versy on this point persists, there are more
the most important single step forward. Sim- and more adherents now to the notion that
ply put, we are not in a position to "dethrone meaningful analysis in development requires
the GNP," as has been variously suggested, breaking the economy down into a few, some-
but rather we must try to place it on an an- times heterogeneous, sectors in the dualistic
alytically sturdier throne. This means ana- tradition rather than the conventional treat-
lyzing much more carefully than we have in ment of many homogeneous sectors in the
the past what the meaning of alternative input/output tradition.
growth paths — or alternative ways of achiev- A second and related advance permitting
ing a particular growth rate — might be in the more meaningful simultaneous analysis
terms of the other things we care about. of growth, employment, and equity has been
This, of course, may lead us to the conclu- in the area of recognizing the importance of
sion that a change in the nature of the growth typological differences among developing so-
path itself — that is, the way in which output cieties. In the immediate postwar period
is generated — can give us not only better em- there seemed to be two major prevailing
ployment and distribution outcomes but views — one, that every LDC is sui generis
more growth in the bargain. On the other and that only country-intensive studies were
hand, there may be trade-ofTs among these likely to advance our understanding; the
objectives; and the nature of these trade-offs, other, that a general theory of underdevel-
■hat is, the extent to which they are man- opment applicable to all countries was within
made rather than inherent in the basic struc- reach and that, in the meantime, we could
ture of development, is of great theoretical behave as if Afghanistan and Argentina had
and policy interest. It is my belief that much more in common with each other than with
of our current and prospective progress rests any so-called mature economy. In more re-
on this ability to integrate neoclassical or cent years, we may note a marked conver-
classical growth theory with a rigorous state- gence between these positions via the accep-
ment of employment and equity con- tance of the notion of halfway houses or
siderations. subfamilies of LDCs differentiated by such
Among the more important ingredients in features as size, resource endowment, and
our theoretical capacity to deal with this new other structural as well as, possibly, behav-
and broadened view of development is our in- ioral characteristics. This trend is exempli-
 125 ANALYTICAL PERSPECTIVES

fied, on the one hand, by the evolving work of that the dualism subphase may itself be typ-
Chenery and his associates, which has moved ically characterizable — to analytical advan-
away from homogeneous 50-country samples tage— by distinct subphases, including a
and toward the attempt to differentiate em- domestic-market-oriented or primary-im-
pirically among different country types; on port-substitution subphase, followed by
the other, expositors of development typolo- either an outer-oriented or export-sub-
gies, for example, of the land-surplus and stitution subphase or the prolongation of do-
labor-surplus school, have begun to open mestic-market orientation via a secondary-
these models to trade and to more empirical import-substitution subphase. The identifi-
treatment. While this work remains very cation of such subphases — based on changes
much in flux, we seem to have growing agree- in the underlying resource endowment as well
ment that it is worthwhile to differentiate as in accommodating changes in the official
countries by size — thus underlining the rela- policy package combining to fundamentally
tive importance of trade — by the extent of affect domestic as well as trade relations —
dualism or labor surplus — thus assessing the constitutes, we believe, an essential ingredi-
relevance of classical versus neoclassical con- ent in advancing our understanding of devel-
ditions inagriculture — and by the strength of opment. For example, the extent to which
their natural-resources base — thus determin- growth, employment, and income distribu-
ing the quantity of land-based fuel available tion objectives of a society are mutually rein-
for the transition effort. forcing or competitive depends very much on
Closely related to this growing acceptance the subphase in which a society finds itself —
of the usefulness of a typological approach as well as, of course, on the LDC subfamily
has been the recognition that a fuller expla- to which it belongs.
nation ofthe historical laboratory would have By way of such illustration only, for the
a substantial payoff for advances in develop- labor-surplus economy which moves from
ment theory. In the forties and fifties, primary import to export substitution, for ex-
the profession understandably was forced to ample, Taiwan, we can clearly identify turn-
concentrate on cross-sectional analysis of ing points between subphases and note sub-
the less developed world as well as on the his- stantial contrasts in the "before and after"
tory of the now-advanced societies, including performance. During primary import substi-
western Europe, Japan, and, to a lesser ex- tution, focused on infant industrial or entre-
tent, Australia, North America, and other preneurial protection and fueled by land-in-
"empty" continents. By now, however, a tensive exports and foreign capital, we would
quarter-century later, sufficient data have ac- expect trade-offs to be more pronounced;
cumulated to permit us to look at developing during export substitution, focused on pene-
societies in a historical context and to try to trating international markets by combining
isolate meaningful subphases of develop- maturing entrepreneurial capacity with "un-
ment. There is no reason to permit the unfor- limited" supplies of labor, we would expect
tunate "stages-of-growth" controversy linked such trade-offs to be substantially reduced, if
to the name of W. W. Rostow to inhibit us in not altogether eliminated. On the other hand,
this respect any longer. While no historical countries which persevere in secondary im-
inevitability connotation is intended, devel- port substitution — often combined with some
oping societies do seem to move in certain export promotion (i.e., subsidized exporta-
transitional states between the long epoch of tion)— in consumer durable, capital, and in-
open agrarianism and another long epoch of termediate goods will find such trade-offs be-
modern growth. One of the more common coming increasingly severe.
transitional states is one of dualism, whether What makes such typology and time-ori-
of the Lewis/Fei-Ranis or the Jorgensen/ ented analysis more analytically feasible
Kelley-Williamson type. today than a quarter-century ago is the num-
Moreover, there are some of us who believe ber of specific advances in our understanding
 126 DEVELOPMENT PERSPECTIVES

of alternative growth paths. One such ad- ment— in spite of a pronounced dependence
vance clearly is a revised view of technology on imported machinery in the first instance.
choice and technology change. It has not There exists, in other words, substantial po-
been all that long that the Eckaus view of es- tential flexibility both in the initial choice of
sentially fixed proportions, and consequent technology from abroad and in the domestic
technological unemployment, was dominant adaptation potential "on top of" such im-
in the LDC literature, with little flexibility in ported technology. Adaptation possibilities
either output or technology mixes. Added to may, in fact, be quantitatively more impor-
that was the notion making a virtue of capital tant than the range of shelf technology
intensity even in a labor-surplus context, choice, which is often more heavily con-
namely, the need for large profit and low strained and only partially illuminated.
wage shares to ensure high saving and low Nevertheless, the two acts are closely inter-
population growth rates. related intheory and practice; that is, econ-
Substantial differences in the industry-spe- omies which try to borrow ahead of their skill
cific technology actually found to be most levels will find it more difficult to assimilate
profitable in developing countries over time, what they borrow, quite aside from incurring
as well as the existence of substantial cross- the higher expense of the initial choice.
sectional differences within developing coun- It may be well for us to note that the adap-
tries for the same industry as one moves tive-technology argument in industry may be
across scales, have cast serious doubt on the more closely related to the situation in agri-
first proposition. The new conventional wis- culture than the profession has yet been will-
dom with respect to technology choice is ing to acknowledge. For instance, agricul-
more nearly that while some industries, es- tural economists are now coming around to
pecially continuous-process industries, are the idea that, while the contribution of inter-
clearly intrinsically not as flexible as oth- national research on new hybrids, etc., has
ers— no matter what the environment — in been substantial — and rightly ballyhooed — it
most industries substantial efficient choice has probably been overstated relative to the
does exist across countries and across scales need for adaptive national research to protect
within countries. This flexibility is most pro- new varieties against disease and ensure the
nounced inthe core processes of discrete or
continuity of such "Green Revolutions." Ag-
batch production, in machine-peripheral ac- ricultural technology is more of a public than
tivities, as well as with respect to plant-saving a private good and thus more easily appro-
possibilities. While "small is not always priated and diffused. With respect to indus-
beautiful" — since it may be inefficient — the trial technology, therefore, we might do well
conventional wisdom has now swung to the to distinguish between technology proper,
recognition that under less distorted relative where the analogies to the better-understood
price environments, growth and employment agricultural sector may hold fairly well — in
at least can be rendered much more compat- spite of the lower level of competition and
ible— via appropriate technology choices — higher level of appropriability — and the
and that, using the foreign-trade mechanism, product and taste differentiation type of
appropriate technology choices will also per- "technology," which is a horse of a different
mit a wider range of output mix variability. color and not really treated here. These an-
The overwhelming burden of the evidence alogies may be especially relevant in light
is consistent with the basic notion that, in a capital goods, cement, brick, and other rela-
developing country which is open and not too tively homogeneous product industries, and
large in size, less relative factor and commod- at the medium and small-scale end of the
ity price distortions, along with the reduction spectrum.
of institutional barriers to information, pro- The second aforementioned proposition
curement choice, etc., may be expected to has proven equally doubtful; that is, there is
produce a substantial increase in employ- no clear evidence that the admittedly higher
 127 ANALYTICAL PERSPECTIVES

saving rates out of a larger profit share are to the realization, moreover, that as impor-
sufficient to overcome the lower absolute lev- tant as international research inputs may
els of output, and profits, resulting from sub- have been, for example, in the case of the
stantial static inefficiency. The posited rela- Green Revolution, it is national adaptive re-
tionship oftechnology choice to population search responses which are going to be cru-
growth has never been established. Small cial in establishing the nature of the employ-
wonder that the explanation as to why, in ment/output generation nexus immediately
spite of large endowment differentials, tech- and, perhaps more importantly, in determin-
nology choices are not, in fact, as dissimilar ing how self-sustained technology change will
as we might expect has been shifting else- ultimately be. In industry, imported technol-
where. If countries do not, in fact, always ogy change is more nearly a privately appro-
take full advantage of the potential that ex- priable commodity in most instances, and we
ists in terms of known technology shelf are still less able to disentangle the respective
choices, this may have more to do with im- roles of public and private sector R&D and
perfect information channels or the ability of information access in the borrowing and ad-
entrepreneurs, due to the existence of monop- aptation processes. However, the overall im-
oly profits, to indulge their engineering pref- portance, for growth and labor absorption, of
erences— quite aside from the most common the size of the innovational effort apparently
explanation, that is, the effect of severely increases over time relative to the quality or
distorted relative factor and commodity bias of the innovation.
prices. In summary, there is substantial consensus
The choice of the direction of technology today that technology change, both in terms
change, a closely related issue, is still some- of its strength and in terms of its bias, rep-
thing more of a mystery, because, as in ad- resents perhaps the single most important
vanced-economy growth theory, we have no element in effecting the reduction and possi-
analytically sound innovation inducement bly the elimination of any trade-off between
mechanism on which to base our reflections. employment and growth objectives. Actual
Yet most of us do recognize that the Hayami- country experience indicates that there exists
Ruttan type of inducement mechanism as a "deviant" minority of labor-surplus devel-
loosely applied to agriculture is likely to be at oping economies which have, in fact, created
work, that is, that labor-surplus economies an environment conducive to appropriate
with expectations of a continuing relative technology choice and technology change
shortage of capital and abundance of labor during their export substitution subphase,
are more likely to seek labor-using or capital- and have apparently been able to entirely
stretching innovations — just as societies in eliminate the conflict between employment
which capital can be expected to grow secu- and growth. In Taiwan, for example, growth
larly faster than labor have shown evidence rates accelerated during the 1960s and an
of increasing the pool of labor-saving inno- unskilled labor shortage replaced labor sur-
vations. What is admittedly less clear, and a pluses bythe end of the decade.
subject of considerable theoretical and prac- While growth and employment are now in-
tical interest, is whether or not the pool of creasingly viewed as, at least potentially,
labor-using innovations, mainly in the form compatible, there is still a good deal of con-
of plant floor rearrangements, the speeding troversy surrounding the question of growth
up of machines, etc., is likely to be as easily versus income distribution. In fact, the ma-
replenished as the pool of labor-saving inno- jority view here clearly holds that, especially
vations inwestern European and U.S. expe- as growth first gets under way, "things must
rience. More likely, additional adaptations of get worse before they can get better." The re-
a labor-saving type will require the impetus sults of Kuznets, and later Oshima, using
of additional acts of shelf technology borrow- cross-sectional data, and those of Adelman
ing. In agriculture, we have gradually swung and Morris employing more sophisticated
 128 DEVELOPMENT PERSPECTIVES

techniques on a broader set of cross-sectional generation, namely, the 1960s. It is worth

socioeconomic data, point in the same direc- noting that even Adelman concedes that
tion. Most observers seem to have found this South Korea and Taiwan may constitute ex-
evidence persuasive for purposes of prognos- ceptions to the rule, even though she does not
tication for individual developing countries go into the reasons. Our own detailed exam-
over time. Kuznets. for example, sees this ination ofTaiwan leads us to the conclusion
outcome as a necessary concomitant of in- that the inverse U-shaped or Kuznets pattern
creasing levels of profit and rent-fed accu- can be substantially softened and possibly
mulation, as growth gets under way, plus the even eliminated, so that even before the com-
effect of shifts in the center of gravity from mercialization point, when labor surplus dis-
"more equal" to 'less equal" sectors (agri- appears, the trade-off between growth and
culture to nonagriculture). A. Lewis notes distribution may disappear. Our analysis is
that, while the unlimited supply of labor based on the effort to decompose the overall
phase persists and wage rates are held down, Gini into factor Ginis and then linking
the profit share must necessarily rise — tend- changes in these Ginis to such growth-rele-
ing to a worsening of the distribution of vant phenomena as reallocation under dual-
income. ism, capital intensity, and innovational bias.
These arguments have been further but- While this is not the place to dwell on our
tressed bythe actual historical experience of findings in detail, our analysis permits us to
a substantial number of contemporary LDCs. conclude that a combination of early concen-
Whether we pick the Gini coefficient or tration on agricultural productivity increase,
McNamara's favorite index relating the pro- along with rural industrialization in the dual-
portion oftotal income accruing to the lowest economy context, can produce this result.
and highest quartiles. the distribution of in- This follows basically from the fact that, even
come in Mexico, in Brazil, in the Philippines, as real wages do not rise very much, addi-
in fact, in most LDCs, has been worsening tional employment opportunities are pro-
over the past decade. Nevertheless, though vided for members of the poorest landless ag-
expert opinion is generally as pessimistic as ricultural families.
the historical evidence, we need to reassure The profession is clearly still groping to-
ourselves as to whether this record is inevi- ward a consensus on how to analyze distri-
table in the nature of things — or, once again, bution and poverty in relation to growth in a
man-made and thus just possibly avoidable. largely rural dualistic setting. But this should
Kuznets and Adelman-Morris data, after all, not deter us from recognizing that a substan-
are heavily policy distorted, with most LDCs tial conceptual shift in emphasis has already
examined remaining under essentially im- taken place in a number of important dimen-
port-substitution types of policy settings. sions. Some of these try to incorporate im-
What should give us pause is that the old portant institutional or noneconomic vari-
chestnut of an unequal distribution of income ables into our analytical framework; others
-equired to generate high saving rates is no focus heavily on structural and other typolog-
longer generally accepted. We certainly have ical differences among families of LDCs: still
examples of countries — outside of the social- others are trying to use large-scale general
ist orbit — which, like Japan and Korea, have equilibrium models and sensitivity analysis to
simultaneously experienced high saving rates help us focus our attention on the important
and a fairly equal distribution of income. behavioral relations. But what most of these
Most significant surely is the actual record of approaches have in common is a continuing
one such economy, for example, Taiwan, deemphasis on the brute forces of capital ac-
which yields not only remarkably low levels cumulation and rigid Harrod-Domar-type
of the Gini coefficient (near .3 rather than .5, output relations and an increasing emphasis
as for most LDCs) but the avoidance of any on the importance of human resources, tech-
but the slightest tendency to rise during the nology change, and other sources of flexibil-
period of fastest growth and employment ity in the system.
 129 ANALYTICAL PERSPECTIVES

II.B.4. Development Modeling:

State of the Art*

Among the developing countries, India was only, such as the "Domar — Harrod princi-
probably the first to introduce its own plan- ple," a number of more concrete and simpli-
ning system, guided by P. C. Mahalanobis, fying techniques were proposed, discussed
who was impressed very much by the Soviet and applied. One of these was planning in
accomplishments, and took over the habit of stages, where initially three levels were dis-
drawing up five-year plans. Mahalanobis was tinguished: the macro, the middle (or meso),
the first to introduce the so-called specific and the micro levels. Macro calculations used
capital goods, differentiating between capital national totals, such as national income or
goods needed to produce consumer goods and product, investments, consumption, imports
capital goods to produce capital goods. The and exports. The next stage referred to sec-
main use made by him of this concept is to tors or branches of activity in its "functional"
show that, to put it simply, initial restraint on version, or provinces in its "geographical"
consumption may later bring higher levels of version, which, in the case of a federated na-
well-being than in the absence of the initial tion, might be states. The micro-level of the
restraint. functional version would be single projects
Understandably, in developing countries, (sometimes enterprises), the subject matter
interest was more concentrated on long-term of loan negotiations. (A separate source of in-
movements of the economy than in developed formation usually resulted from such negoti-
countries where short-term movements (es- ations.) In the geographical version, either
pecially cyclical) and speculative operations cities or homogeneous regions, possibly river
evoked more interest. For any economy, how- valleys, entered the scene.
ever, short-term fluctuations cannot be fully Among the sectors, some might receive
neglected. In developing economies, they special attention; for instance, education or
probably are to a larger extent the conse- educational subsectors. Alternatively, public
quences ofcrop variations and hence almost and private, or formal and informal sectors,
random. So, short-term economic models also had to be kept apart.
became a subject for systematic research, as A technique that found widespread appli-
shown by Narasimham. cation is known as input-output analysis,
With an increasing number of former col- originating from previous concepts of produc-
onies becoming independent, at least for- tion of goods by goods (Sraffa) and the core
mally, the formulation of development poli- also of the Soviet planning system, even if not
cies as part of a government's tasks spread. infrequently in simplified ways, with the em-
This was enhanced by institutions, such as
phasis on the "material balances." A refine-
the World Bank Group and the International ment of conventional input-output analysis
Monetary Fund, which needed a basis of was obtained by a distinction between
evaluation of the loans extended to develop- tradables and nontradables and got the
ing countries. International co-operation also name of semi-input-output analysis (cf.
began to be more institutionalized, as shown Kuyvenhoven).
by the solemn proclamation, in 1961, of the Another useful tool of analysis proved to
"Development Decade," later rebaptized as be the two-gap concept. A nation, when
the First D. D. (D. D. I.). looked at as a whole, may be in disequilib-
In view of the increasing practical appli- rium in the sense that either its savings fall
cations of what erstwhile were principles short of the investmest it wants to finance or
its exports fall short of the imports it wants
*From Jan Tinbergen, "Development Modeling: to finance. The former gap is indicated as the
The State of the Art," in The Pakistan Development
Review, Vol. XIX, No. 2, Summer 1980, pp. 93-98. savings gap, the latter as the trade gap
Reprinted by permission. (where trade implies invisibles). In fact, and
 130 DEVELOPMENT PERSPECTIVES

The additional variables needed for our

"ex post" (looking at the figures of past time
units), the two are identical and in that sense models will be discussed in succession, with-
there is no point in using two words. But, es- out a claim to completeness. To begin with,
more social variables are needed and a first
timated inadvance, "ex ante," the two may
be different and one appears larger than the example is of figures on income distribution.
other. It is then sometimes called "domi- Although accurate data are hard to get,
nant" and this state of affairs may be an in- crude approximations have been estimated
dication of the measures that need to be for up to sixty developing countries. A crude
taken first. The variables that are the subject distinction into quintiles of income distribu-
of such measures are called "adjustors." tion will already be useful; for links have to
Their use, and the adjustment process, may be made with other variables in order to in-
be either automatic (part of the "free forces tegrate them into the model. An interesting
of society") or policy-induced. If more than example is given by Bornschier in a sociolog-
one adjustor exists, we may even have one or ical model, comparing up to 72 countries in a
more "degrees of freedom." Several authors cross-country study. Another very complex
have used the tool just mentioned in order to study for one country, Korea, shows what can
arrive at policy recommendations or even a be undertaken in the socio-economic field for
range of possible policies, where the policy- a country known for its wealth of data (Adel-
maker has a choice. An excellent survey of man and Robinson). A host of other social
the practice of and experience with planning data are being collected under the general
has been given by Waterston et al. name of social indicators, which claim an im-
A last remark to be made in this section is portant position in the welfare concept (Levy
on international models. These became in- and Guttman; Scheer).
creasingly necessary as a consequence of the Another extension of the number of the
task created by the institution of the United needed variables relates to the "informal sec-
Nations Development Planning Committee. tor." Depending on the type of information
This committee came into being as a conse- needed, some of it will be already part of the
quence of the concern, around 1965, in the official statistics, as handicraft and other
developing countries that the Development small-scale business. Sample surveys are the
Decade's objectives were not attained. On be- best known tool to obtain data on the infor-
half of the committee's discussions, the U.N. mal sector, and in some countries, e.g., India,
Secretariat prepared a world model, largely there is a vast wealth of them. Yet the gen-
inspired by J. L. Mosak, but unfortunately eral impression is that much remains to be
never published it. In all probability, it was done in this field. For a developed country,
the first of its kind. More recently, Leontief such as Britain, novel ideas have been for-
et al. made an impressive attempt at con- mulated on both the household and the infor-
structing various alternatives of a world mal (partly illegal) sectors by Gershuny.
model. . . . Of late, the discussion on appropriate tech-
nology has acquired much momentum. As a
ASPECTS UNDER DEBATE matter of course, most information has to
come from single enterprises, and economists
... [L]et us now discuss how the experi- will have to co-operate with engineers. Path-
ences of the past decades and the new tasks breaking work has been done by Boon, a good
as now understood have their impact on de- deal under the auspices of El Colegio de
velopment planning and the models to be Mexico, but also by Beyrard, and probably
used for it. We will discuss this impact under by many consulting engineering firms.
two main headings: (a) the additional vari- Among the United Nations organizations,
ables that must enter into our models, and UNIDO and FAO are the most involved, but
(b) the new use to be made of planning twice already a general conference on the
models. subject of science and development (Geneva,
 131 ANALYTICAL PERSPECTIVES

1963 and Vienna, 1979) was organized. The populations, practised by the Institute of Cul-
subject is so vast that each sector in each tural Affairs, located near Bombay, with af-
country required its own treatment. While filiations al over the world. Econometricians
there is much scope for the application of the are accustomed to entering non-measured
principle of self-reliance in this field also, one variables into models with the aid of "prox-
world-wide problem looms large at the hori- ies"; in a way, social indicators, discussed be-
zon of technology, namely the consequences
fore, are such proxies for "happiness" or
of the introduction of microelectronic de- "utility" in economic jargon, the very aim of
vices. Even though inertia will slow down real socio-economic policies.
developments in comparison to those theoret- Let us now turn to the (partly) new use
ically deduced from what the avant-garde that is going to be made of models.
performs, the need for case studies concern- Newer methodologies for use of economet-
ing previous generations of electronic devices ric models nowadays abound (cf. Nijkamp).
is strong. In such case studies, important One is needed to deal with a major fact of
variables to be ascertained are the capital in- life, namely that policy-makers more often
tensity ofnew processes and the elasticity of disagree than agree about their aims. Hence
demand for their services. If the latter is all sorts of compromising processes charac-
high, which is by no means impossible, the terize development policies. The problem is
threat to employment is less, of course, than more than a century old in economic science
if that elasticity is low. Some institutions in and was first tackled by the French econo-
the field of banking have announced their ex- mist Cournot in an attempt to describe the
pectation of no threat to employment, but attitude of competitors who are few (repre-
opinions widely diverge. The impact on com- senting a near-monopoly situation). Several
parative costs in developed and developing other new methods are tried out, all contain-
countries has hardly been put on the agenda ing assumptions on how politicians "play
for discussion, but, as a matter of course, it their game" (cf. the Theory of Games, due to
should be included. Von Neumann and Morgenstern).
Some technologies will experience the im- Another example of the new use made of
pact of higher energy prices; these have made models is due to Cohen. His method consists
feasible, economically speaking, village-level in maintaining a sufficient number of degrees
activities such as biogas production (Reddy, of freedom so as to alternatively use the
Prasad, and Prasad). model for maximizing welfare of different
Some of the important new variables groups. Depending on the power shifts which
which must be introduced into development may occur, another group's welfare function
models are of an environmental character. may be taken as maximum and so the effects
Among them may be the minimum surface of the power change can be studied.
under forests, i.e., the surface needed to avoid Again, another way of using models is
desertification. For an island such as Java, called interactive multiple-goal planning
Indonesia's densely populated central island, (Nijkamp and Spronk). In essence, it is a re-
this is now becoming a question of life and action to the way to which policy-makers
death; and transmigration of a part of Java's maximize their objective function. They
population to Sumatra is one of the most ur- hardly know what an objective function
gent activities to be organized at a much means, but they do know which of the differ-
larger scale than heretofore. ent situations they prefer. The planning pro-
This brings us to the category of non-mea- cess isgiven the shape of a dialogue between
sured variables which have now attracted the model-builders and policy-makers. The for-
attention of model builders. We already men- mer show a feasible situation of the economy
tioned variables concerning the informal sec- to the latter and ask them whether they want
tor, and could add the activities of self-reli- a change in any of the objective variables or
ance, by the "mobilization" of village targets. After the policy-maker has suggested
 1 32 DEVELOPMENT PERSPECTIVES

a new value for that target, the model-han- Leontief, W., et al., The Future of the World
dler tells him what other target has then to Economy, New York, Oxford University
be changed (mostly in an unfavourable Press, 1977.
sense) and whether the change is worth the Levy, S., and L. Guttman, "On the Multi-
gain on the first target. Thus the dialogue fin- variate Structure of Well-being," Social
ishes in a situation which, for the policy- Indicators Research, Vol. 2, 1975.
maker, isthe best attainable configuration. Narasimham, N. V. A., A Short-Term Plan-
ning Model for India, Amsterdam, North
Holland, 1965.
Nijkamp, P., Multidimensional Spatial Data
References and Decision Analysis, New York, Wiley,
Adelman, J., and S. Robinson, Income Dis- 1979.
tribution Policies in Developing Countries.
Nijkamp, P., and J. Spronk, "Interactive
A Case Study of Korea, London, Oxford Multiple Goal Programming," Rotterdam,
University Press, 1978 (published for the Centrum voor Bedfijfskundig Onderzoek,
World Bank). Erasmus
A). University, 1978 (Report 7803/
Boon, G. K., Technology and Sector Choice
in Economic Development, Alphen and Reddy, A. K. N., C. R. Prasad, and K.
Rijn, Sijthoff and Nordhoff, 1978. Krishna Prasad, "Bio-Gas Plants: Pros-
Bornschier, V., "Einkommensungleichheit pects, Problems and Tasks," Economic and
innerhalb von Landern in Komparative Political Weekly, Vol. IX (Special Num-
Sicht," Schweiz. Zeitschr. f Soziologie, ber), August 1974.
Vol. 4, 1978. Scheer, Lore, Quality of Life, Vienna, Ar-
Cohen, S. I., Agrarian Structures and Agrar- beitsgemeinschaft fur Lebensniveauver-
ian Reforms, Leiden/ Boston, Martinus gleich, 1977.
Nijhoff, 1978. Von Neumann, J., and O. Morgenstern, The-
Gershuny, J. L., Address to the Dutch Sci- ory and Games of Economic Behavior,
entific Council of Government Policy, The Princeton, N.J., Princeton University
Hague, 1979. Press, 1944.
Kuyvenhoven, A., Planning with the Semi- Waterston, A., et al., Development Planning,
Input-Output Method, Leiden/ Boston/ Lessons of Experience, Baltimore, Md.,
London, Martinus Nijhoff, 1978. The Johns Hopkins Press, 1965.
 II.C. RADICAL PERSPECTIVES
II.C.1. A Marxist View *

What we find, therefore, is a spotty growth protected from contamination against exter-
record within the Third World capitalist nal antagonists. For the antagonists are
countries, some of them doing well but many found within the growth process itself.
foundering and, at best, a lackluster perfor- A particular version of this general Marx-
mance inthe areas of equity and employment ist view is the notion that the capitalist ac-
opportunities. We do not find, however, de- cumulation process produces both wealth and
spite such a suggestion from some Marxist poverty. Marx originally formulated this
analyses, a uniform lack of progress among proposition, in the first volume of Capital, as
the LDCs toward industrialization, higher the general law of capitalist accumulation. In
living standards, fairer treatment of peasants that work, Marx wrote:
and the urban poor, or containment of the
The same causes [that] develop the expansive
population upsurge. In these and other direc- power of capital, develop also the labor-power at
tions, progress has been recorded here and
its disposal. The relative mass of the industrial re-
there within this group of countries, even serve army increases therefore with the potential
though the general situation is not good. energy of wealth. . . . This is the absolute general
Where, then, do these considerations leave law of capitalist accumulation. Like all other laws
us with respect to Marxian views about eco- it is modified in its working by many circum-
nomic development in the Third World? stances. .. . This law rivets the laborer to capital
more firmly than the wedges of Vulcan did Pro-
metheus to the rock. It establishes an accumula-
DIALECTICAL DEVELOPMENT: tion of misery, corresponding with accumulation of
WEALTH AND POVERTY capital. Accumulation of wealth at one pole is,
therefore, at the same time accumulation of mis-
I would say that recent experience has sup- ery, agony of toil, slavery, ignorance, brutality,
ported the Marxist contention that the devel- mental degradation, at the opposite pole, i.e., on
opment process is a dialectical one. This the side of the class that produces its own product
means that development produces not only
in the form of capital.1
equilibrium but its opposite, disequilibrium,
In this passage, Marx had in mind not only
not only continuities but discontinuities, both
material poverty (absolute and relative) at
social harmonies and social conflicts, bal-
the one pole but the poverty of alienation as
ances and imbalances, growth and stagna-
tion, and so on. In this view of economic de- well, for he added that "in proportion as cap-
ital accumulates, the lot of the laborer, be his
velopment, growth contains within it
antigrowth forces, which will lead to the in- payment high or low, must grow worse."2
evitable breaking up of the existing state of This thesis has been accepted by genera-
tions of Marxists, though it has often been
things, to the transiency of all things. So the
modified in one way or another, and it has
development process cannot be a gradual,
steady, harmonious movement toward equi- been applied in numerous ways, including ap-
librium, as seen by many advocates of neo- plications toproblems associated with under-
classical doctrine. It cannot be an uninter- developed countries. Marx and Engels paved
the way for later analyses by arguing, in a
rupted growth process in which growth is
dialectical manner, that European colonial
*From John G. Gurley, "Economic Development: expansion was an inevitable outgrowth of the
A Marxist View," in K. P. Jameson and C. K. Wilber
(eds.), Directions in Economic Development, Copy- 'Karl Marx, Capital, Vol. 1, New York, Interna-
right 1979, University of Notre Dame Press, Notre tional Publishers, 1967, p. 644-5.
Dame, Indiana, pp. 201-14. Reprinted by permission. 2Ibid., p. 645 (my emphasis).

133
 1 34 DEVELOPMENT PERSPECTIVES

development of capitalism, that it brutalized one pole and underdevelopment at the

and plundered the peoples of the colonized other"; a proposition, he emphasized, that
areas and disrupted or destroyed their liveli- "applies not only to the relations between the
hoods, but that, despite all of this, such ex- advanced capitalist countries" and the colo-
pansion was necessary to push many of the nial and semicolonial countries, but also
backward countries off dead-center so as to within both of these parts.6
implant in them the seeds of capitalist devel- This thesis was generalized and carried to
opment. Enrichment at one end was inextric- its ultimate destination by A. G. Frank, who
ably linked to impoverishment at the other. contended that capitalism had long ago en-
But the impoverishment and the brutality tered every nook and cranny of the satellite
that accompanied it were part of a historical world in such a way as to make global capi-
process in which destruction contained con- talism an integrated structure of metropoles
struction within it. The construction would and satellites that bound nations, regions,
be a capitalist one, Marx and Engels be- and urban-rural areas into dominant-depen-
lieved, inwhich the bourgeoisie would create dent relationships. A systematic transfer of
progress and wealth at one pole, but at the economic surpluses continually occurred
same time drag masses of people "through from the base of this world structure — that
blood and dirt, through misery and degrada- is, from the millions of workers and peas-
tion." This polarization would end only with ants— that benefited the metropoles and
the demise of the bourgeoisie: harmed the satellites throughout the struc-
ture, in which a lower metropolis was in turn
When a great social revolution shall have mastered
the results of the bourgeois epoch, the markets of a satellite of a higher metropolis. Frank
the world and the modern powers of production, claimed that when such exploitative ties were
and subjected them to the common control of the temporarily loosened, such as during wars or
most advanced peoples, then only will human depressions involving higher metropoles, a
progress cease to resemble that hideous pagan idol, type of development occurred in the satel-
who would not drink the nectar but from the skulls lites; when the ties were later strengthened,
of the slain.3 underdevelopment was generated. He also
That is just about as grisly a picture of capi- asserted that "the greater the wealth [that
was] available for exploitation [in the past],
talism and the wealth-poverty dichtomy it
fosters as can be found anywhere. the poorer and more underdeveloped the re-
In recent years, Paul Baran has argued gion today; and the poorer the region was as
that the British, while enriching themselves, a colony, the richer and more developed it is
"systematically destroyed all the fibres and Samir Amin has most recently analyzed
foundations of Indian society."4 He sup- this polarization thesis in greater depth. All
ported Nehru's statement that British rule today."7 or satellite formations, Amin
peripheral
and policy were responsible for the later
writes, have four main characteristics: (1)
"tragic poverty of the people."5 Baran then the predominance of agrarian capitalism; (2)
set out to show how such exploitative rela-
tionships characterize the entire structure of a local, mainly merchant bourgeoisie that is
world capitalism. In much the same way, dominated by foreign capital; (3) the growth
of a large bureaucracy, which substitutes for
Paul Sweezy asserted that "capitalist devel- an urban bourgeoisie; and (4) incomplete
opment inevitably produces development at
proletarianization,
6Ibi
which takes the form of
3These views can be found in Shlomo Avineri, ed.,
Karl Marx on Colonialism and Modernization, New massesd. of poor peasants, urban unemployed,
York, Anchor, 1969, pp. 137-9. and many marginal workers. These are the
4Paul Baran, The Political Economy of Growth,
New York, Monthly Review Press, 1957, pp. 144-
150. 7A. G. Frank, Lumpenbourgeoisie and Lumpen-
5Ibid. development, New York, Monthly Review Press,
1972, p. 19.
 135 RADICAL PERSPECTIVES

features of an incomplete and extraverted de- tonio Dos Santos, Celso Furtado, and F. H.
velopment oflocal capitalism, in which the Cardoso. But my impression is that we still
center capital reshapes the periphery into ex- fall far short of a full understanding of how
tensions ofthe center, thereby preventing au- market forces, and the values and behavior
tocentric development in the periphery and patterns that support market processes, con-
establishing unequal exchange between cen- tribute to polarization of living standards at
ter and periphery — that is, differences be- the various levels of the global capitalist
tween the returns to labor that are larger
system.
than differences between their productivi- A convincing explanation of this must
commence with the fundamental mechanism
ties.8 The dominance of foreign capital in the
periphery, Amin claims, means distorted de- of exploitation of labor by capital. The expla-
velopment toward export activities and exces- nation would then include the process that
sive development of the tertiary sector and makes the reserve army of labor a necessary
light industry. It also means extreme uneven- element of the capitalist mode of production
ness of development, the disarticulation of and an analysis of how the expansion of this
the economy, a development process that is mode dispossesses and impoverishes previ-
not cumulative, and a strong tendency to- ously self-sufficient producers. The explana-
ward periphery indebtedness to the center.9 tion would also include an analysis of how the
It is a general consensus, then, among capitalist mode necessarily builds on effi-
Marxists that capitalism produces polariza- ciency, which is often antithetical to equity;
tion, day in and day out. The basic mecha- of how it builds on specialization, which in-
nism isthat capital utilizes its economic and creases the vulnerability of people and areas
political dominance to exploit labor through- to adverse developments; and of how it trans-
out the global system by employing for this fers wealth from poor areas to wealthy ones,
purpose both plain force and market forces; via the price mechanism. It would then go on
and that, throughout the structure, big capi- to discuss the values and behavior patterns,
tal dominates small capital. engendered and supported by the capitalist
When it comes to "market forces," how- mode, which compel individuals to seek suc-
ever, Marxist theorists appear to be only in cess even at the expense of others' welfare
the early stages leading to a full formulation and which encourage and reward cheating
of the variety of mechanisms in the capital- and extortion of the poor by the rich. Finally,
accumulation process that generate biform- the explanation would show how poverty,
ity. Still, there have been some outstanding once established, feeds on itself — the poor
achievements here. Arghiri Emmanuel has lack capital, information, education, and so
analyzed how increasing inequality between on — and how wealth, once established, also
nations is rooted in an unequal exchange that feeds on itself.
tends to increase over time, so that poor na- Marx once said, in discussing the repeal of
tions become poorer and the rich ones richer.
the Corn
ers cannotLaws in England:
understand how "Ifonethenation
free-trad-
can
Samir Amin has extended the analysis into
processes of biformity within the satellite or grow rich at the expense of another, we need
periphery countries, as I previously dis- not wonder, since these same gentlemen also
cussed. Also, many notable contributions refuse to understand how within one country
have been made by economists and others one class can enrich itself at the expense of
pursuing dependency theory, including Theo- another."10 Present-day Marxists do under-
stand these things, but they are still several
8This is most fully explored in Arghiri Emmanuel, steps away from having a general theory
Unequal Exchange: A Study of the Imperialism of
Trade, New York, Monthly Review Press, 1976. about them.
9Samir Amin, Unequal Development, New York, 10Karl Marx, "On the Question of Free Trade," in
Monthly Review Press, 1976, pp. 203, 249-51, 288, Karl Marx, The Poverty of Philosophy, New York,
334-64. International Publishers, 1969, p. 223.
 136 DEVELOPMENT PERSPECTIVES

CONVENTIONAL DEVELOPMENT: would be correct to point out that young so-

LIFE-CYCLES AND MISTAKES cialist countries appear to be immune to
most, and perhaps to all, of these diseases.
Conventional development economists, Yet, in each of the examples that I just cited,
while recognizing poverty, inequalities, and there is a plausible hypothesis with more or
unemployment in the Third World countries, less supporting data that should enlist the at-
have explanations and remedies for these ills tention of Marxists, who, however, are too
that contain enough truth to be of interest to busy ascribing all ills to capitalism in general
Marxist theorists. And yet the latter have to have time to consider the orthodox
paid little attention to these orthodox arguments.
rationalizations. Conventional development economists
The principal and most general explana- have another arrow for their bow. The ills of
tion of the economic ills of the Third World poverty, inequality, and unemployment that
is that they are associated with the particular one finds in Third World countries, they say,
stage of a life-cycle that the LDCs happen to are largely owing to the gross inefficiency
be in, and so can be outgrown. Thus, as one with which these economies are run. Their
version of this explanation has it, all coun- economic managers create the maladies by
tries, provided they are not eternally damned espousing policies that are almost insane. For
to the inferno, enter a purgatory stage before example, a host of measures underprice cap-
they go onward and upward to blessedness ital and overprice labor: high minimum wage
and earthly paradise. Along this development rates, too low a rate of interest, overvalued
path, income distribution worsens before it currency (which makes imported machinery
gets better. Population growth gets out of cheap), payroll taxes on employers for their
hand before birth rates begin to fall into line employees, and exemptions from custom du-
with death rates. Countries become chronic ties for imported capital equipment. Each of
debtors before they burst into the heaven of these measures induces the substitution of
creditors. They go through the hell of hyper- capital for labor, thereby exacerbating un-
inflation before they develop their economies employment, poverty, and income inequali-
sufficiently to allow price stabilization. The ties. The same thing is accomplished when
unemployed pile up in new urban centers be- policymakers neglect agriculture, which uses
fore full employment can be enjoyed. And labor-intensive methods, and support heavy
countries might even suffer dictatorship be- industry, which utilizes capital-intensive
fore democracy wins out. Another version of methods. Other policies, it is said, also inter-
the life-cycle explanation — the straight-line fere with what would be correct relative
instead of the U-shaped model — is that prices, and the resulting distortions are
things start off bad and then get better and enough to make anyone sick.
better. . . . The World Bank has recently set forth a
These explanations are meant to suggest, similar explanation of poverty and other
of course, that, when one encounters the bourgeois ills. There has been a poor order-
growing ills of the Third World, one is merely ing, the Bank says, of national priorities. A
looking at some childhood diseases. With reordering is called for. Instead of policies to
proper care of the kiddies, all will be well foster "growth in general," policies should be
later on. In many cases, though, the data redesigned to direct investment to poverty
themselves are quite underdeveloped and just groups in the form of education, access to
barely suggest what is so confidently as- credit, health care, public facilities, and so
serted. Moreover, the results rest on cross- on. Thus, the new national priorities should
section data which are intended to apply to be "growth," as before, but growth in the
individual countries over time. The dangers lower income groups where absolute poverty
of making this leap are so well known as to resides.
require no comment from me. Also Marxists Just as in the previous set of arguments,
 137 RADICAL PERSPECTIVES

the intention here is to make one believe that harsh income inequalities, widespread lack of
the growing ills one sees in young capitalist education, the dominance of rural living, and
economies are not owing to capitalism per se the oppression of females — is closely associ-
but to the appalling way the economies are ated with high birth rates and so with high
managed. If the distortions or improper na- rates of population growth. Thus, if ruling
tional priorities were corrected, the argument classes, through their exploitation of working
goes, the economic ills would be cured. In this classes and in other ways, generate underde-
case, too, there is probably enough truth be- velopment, the consequence is likely to be
hind the argument to warrant the attention high population growth rates, which in turn
of Marxists. But even granting this, Marxists feed back to lower incomes per capita and
must go on to ask, as conventional economists more underdevelopment. The overthrow of
do not, why there is so much inefficiency and these ruling classes is a necessary, though not
ineptness. Does it serve a political purpose? a sufficient, condition for breaking this vi-
Which class does it help, and how? Or is it cious cycle.
simply due to ignorance? Conventional econ- While all of this is largely true, Marxists
omists believe much too naively that no one have not taken adequate account of the rea-
wants unemployment or poverty, and so they sons for and the consequences of the popula-
must arise from lack of knowledge or will- tion explosion that has occurred in recent
power. Marxists know that such seeming dis- decades. World population up to 1650 grew
orders are often beneficial to ruling classes, by only a small fraction of 1 percent per
and hence are no mistakes at all but out- annum. For the next 300 years the rate crept
comes of rational policies. While I believe up toward 1 percent. But in the last two or
that Marxists are mainly correct about this, three decades it has risen to around 2 percent
it is still true that the inefficiencies and ab- or more — 1 percent in the developed coun-
surdities, willful or not, are well worth study- tries and 2% percent in the less developed
ing. Marxists themselves are too prone to as- ones. Between a.d. 1 and 1750, world popu-
sume that all capitalist economies, at lation rose by 480 million. In the next 200
whatever stage of development they find years, it rose by 1,750 million, and in the last
themselves, are run with equal efficiency. 28 years by the same amount. China and
There is much for Marxists to learn in this India alone this year will add about 30 mil-
area.
lion to the world's population, a number that
exceeds the total national population of 87
CONVENTIONAL DEVELOPMENT: percent of the countries of the world.
THE POPULATION EXPLOSION The major reason for this explosion is the
decline in death rates owing to rapid techno-
Many conventional economists ascribe the logical advances in modern medicine — espe-
ills of the LDCs in part to the population ex- cially vaccination against malaria, smallpox,
plosion that is taking place throughout the yellow fever, and cholera — and to the spread
Third World. It is claimed that this is so un- of modern public health measures. The sharp
usual in history and so devastating as to ac- drop in death rates generated a very high nat-
count for much of the poverty and inequality ural rate of increase because both birth and
that one observes everywhere. death rates in the LDCs were a percentage
Marxists have assumed that the so-called point higher than they had been in preindus-
population problem is meant to divert atten- trial Europe (4% against 3%. percent). All of
tion from the real issues, which involve the this has created serious difficulties for many
transformation of the present exploitative countries in the Third World, unlike the
economic systems into socialist societies. I nineteenth-century experience of western Eu-
think that Marxists are largely correct about rope when both death and birth rates de-
this. Underdevelopment — with its low clined slowly and together.
growth rates, high levels of unemployment, Although there is evidence that a new
 138 DEVELOPMENT PERSPECTIVES

global demographic transformation is now since that brilliant analysis of Marx, his fol-
underway, involving declining birth rates to lowers have been extremely wary of all such
match the previous drop in death rates, pop- superficial evidence.
ulation pressure for many LDCs will remain But this attitude is wrong. Marx himself
intense and will retard improvements in liv- used plentiful data in his major work. These
ing standards for hundreds of millions of peo- "facial expressions" of capitalism are ex-
ple for many years to come. This is a problem tremely useful to the investigator, so long as
that concerns both socialist and capitalist he or she is aware of what is really going on
countries. Some countries in both camps have underneath it all. Since there are systematic
succeeded in recent years in getting their relations between the one world and the
population problems somewhat under con- other. Marxist theories about the underlying
trol— socialist China as well as capitalist realities can be tested against the surface
Taiwan and Singapore. Thus there would ap- data with good effect. Indeed. Marxists ought
pear to be some mechanisms within the cap- to be in a better position than conventional
italist mode of production that, to some de- economists to provide analytical depth to the
gree, control population upsurges, and these interpretation of these data.
ought to be more carefully investigated by In any event. Marxists ought to look care-
Marxist development economists than they fully at what conventional economists call the
have been in the past. "success stories" of capitalist economic de-
velopment and. indeed, at all the evidence
THE TESTING OF THEORIES that has been so laboriously compiled by the
more orthodox economists, no matter how
A serious failing of Marxist development embarrassing it may be. One would not think
economists lies in their lack of interest in it necessarv to sav that, but it is.
testing theory against the facts. Almost with-
out exception, it has been the conventional SUMMARY OF MARXIST AND
economists who have dug out the information ORTHODOX RESPONSES
on income inequalities, differential growth
rates, absolute poverty, unemployment, and A large number of non-Marxian LDCs are
the like. Even with these data available. stagnating. Many, however, are expanding,
Marxists appear uninterested in them and. but, even in these cases, their growth is often
instead, remain totally absorbed in more accompanied by growing inequities, unem-
model building. An unfriendly critic would ployment, and impoverishment. Orthodox
say that Marxists ignore these data because
economists look on these "ills" either as
they disprove their theories. A better and childhood diseases or as residues of the past
more profound explanation is that Marxists and, hence, either curable with time and pa-
are trained to be highly suspicious of surface tience or capable of being gradually mopped
phenomena and the crude empiricism often up. Marxists, on the other hand, believe that
used in analyzing these outward manifesta- capitalist development itself creates these so-
tions of deeper, underlying realities. Marx called ills, day in and day out. along with so-
prepared the way for this attitude with his ciety's wealth. Furthermore, orthodox econ-
analysis of the production and circulation
omists think that the "ills" are truly social
spheres of the accumulation process: in the maladies not desired by society, while Marx-
former, located in the underworld, so to ists are convinced that they often serve a use-
speak, resided the reality of exploitation. ful function for the ruling classes and hence
while in the latter, located on the surface of are results of rational, purposive policies.
everyday life, "equality and freedom" These are two entirely different ways of look-
reigned. Surface data, therefore, could badly ing at the development process. The first sug-
mislead one, could create fantastic illusions gests reform; the second revolution. Indeed.
about how the system really operated. Ever Marxists believe that the very processes of
 139 RADICAL PERSPECTIVES

capitalist accumulation create the revolution- In fact, for the better part of this century,
ary conditions required for overthrowing cap- capitalism has been overthrown in large parts
italism and establishing more humane social- of the world, and Marxism has taken over
ist societies. these areas.

II.C. 2. Dependency Theories of

Underdevelopment *

The general field of study of the dependency nomic growth in dependent countries would
analyses is the development of Latin Ameri- be extremely limited; the surplus they gen-
can capitalism. Its most important character- erated would be expropriated in large part by
istic isits attempt to analyse it from the point foreign capital, and otherwise squandered on
of view of the interplay between internal and luxury consumption by traditional elites.
external structures. Furthermore, not only would resources des-
There is no doubt that the "father" of this tined for investment thereby be drastically
approach is Paul Baran. His principal contri- reduced, but so would their internal multiply-
bution to the general literature on develop- ing effect, as capital goods would have to be
ment (Baran, 1957) continues the central purchased abroad. This process would nec-
line of Marxist thought regarding the contra- essarily lead to economic stagnation, and the
dictory character of the needs of imperialism only way out would be political.
and the process of industrialization and gen- Starting out with this analysis Frank at-
eral economic development of the backward tempts todevelop the thesis that the only po-
nations. Thus he affirms at the outset that litical solution is a revolution of an immedi-
ately socialist character; for within the
What is decisive is that economic development in context of the capitalist system there would
underdeveloped countries is profoundly inimical to be no alternative to underdevelopment
the dominant interests in the advanced capitalist (Frank, 1967).
countries (1957, p. 28). For the purpose of this analysis we may
distinguish three levels in Frank's "model of
To avoid such development the advanced underdevelopment." The first is that in which
nations will form alliances with pre-capital- he attempts to demonstrate that Latin Amer-
istic domestic elites (who will also be ad- ica and other areas in the periphery have
versely affected by the transformations of been incorporated into the world economy
capitalist development), intended to inhibit since the early stages of their colonial pe-
such transformations. In this way the ad- riods. The second is that in which he attempts
vanced nations would have easy access to do- to show that such incorporation into the
mestic resources and thus be able to maintain world economy has transformed the countries
traditional modes of surplus extraction. in question immediately and necessarily into
Within this context the possibilities of eco- capitalist economies. Finally, there is a third
level, in which Frank tries to prove that the
* Reprinted with permission from World Develop- integration of these supposedly capitalist
ment, Vol. 6, Gabriel Palma, "Dependency: A Formal economies into the world economy is neces-
Theory of Underdevelopment or a Methodology for
the Analysis of Concrete Situations of Underdevel- sarily achieved through an interminable me-
opment?," Copyright 1978, Pergamon Press, Ltd., pp. tropolis— satellite chain, in which the surplus
899-902. generated at each stage is successively drawn
 140 DEVELOPMENT PERSPECTIVES

off towards the centre. On account of this he made by the Latin American left — the me-
develops a subsidiary thesis: chanical determination of internal by exter-
nal structures; on the contrary, he strength-
If it is satellite status which generates underdevel- ens that mechanical determination in his
opment, then a weaker or lesser degree of metrop- attempt to construct a model to explain the
olis-satellite relations may generate less deep mechanisms through which the expropriation
structural underdevelopment and/or allow for
more possibility of local development (Frank, of the surplus takes place. Probably still un-
1967, p. 11). duly influenced by his training as an econo-
mist at the University of Chicago, he con-
But as the weakening of the satellite-me- structs a mechanico-formal model which is
tropolis network can, according to Frank, no more than a set of equations of general
only take place for reasons external to the equilibrium (static and unhistorical), in
satellite economies, of a necessarily transient which the extraction of the surplus takes
nature, it follows that there is no real possi- place through a series of satellite-metropolis
bility of sustained development within the relationships, through which the surplus gen-
system. According to this analysis, the only erated at each stage is syphoned off.
alternative becomes that of breaking com- It is not surprising that his method leads
pletely with the metropolis-satellite network Frank to displace class relations from the
through socialist revolution or continuing to centre of his analysis of economic develop-
"underdevelop" within it. ment and underdevelopment. Thus he devel-
In my opinion, the value of Frank's analy- ops a circular concept of capitalism; although
sis is his magisterial critique of the suppos- it is evident that capitalism is a system where
edly dual structure of peripheral societies. production for profit via exchange predomi-
Frank shows clearly that the different sectors nates, the opposite is not necessarily true: the
of the economies in question are and have existence of production for profits in the mar-
been since very early in their colonial history ket is not necessarily a signal of capitalist
linked closely to the world economy. More- production. For Frank, this is a sufficient con-
over, he has correctly emphasized that this dition for the existence of capitalist relations
connection has not automatically brought of production. Thus for Frank, the problem
about capitalist economic development, such of the origins of capitalism (and therefore the
as optimistic models (derived from Adam origins of the development of the few and the
Smith) would have predicted, by means of underdevelopment of the majority) comes
which the development of trade and the di- down to the origins of the expanding world
vision oflabour inevitably would bring about market and not to the emergence of a system
economic development. Nevertheless Frank's of free wage labour.
error lies in his attempt to explain this phe- Although Frank did not go very far in his
nomenon using the same economic determin- analysis of the capitalist system as a whole,
ist framework of the model he purports to its origins and development, Immanuel Wall-
transcend; in fact, he merely turns it upside- erstein tackled this tremendous challenge in
down: the development of the "core" neces- his remarkable book, The Modern World
sarily requires the underdevelopment of the System: Capitalist Agriculture and the
"periphery." Thus he criticizes both the al- Origins of the European World — Economy in
ternative proposed by the traditional Latin the Sixteenth Century (1974).
American left (the possibility of a democratic Frank has reaffirmed his ideas in a series
bourgeois revolution, because in this context of articles published jointly in 1969; a year
the only political solution is a revolution of an later he sought to enrich his analysis with the
immediately socialist character), and the pol- introduction of some elements of Latin
icies put forward by ECLA. American class structure (Frank, 1970).
Nevertheless, his critique is not directed Frank has been criticized from all sides,
towards the real weaknesses in the analysis and on almost every point in his analysis.
 14 1 RADICAL PERSPECTIVES

Prominent among his critics is Laclau of view; furthermore, they do not seem useful
(1971), who provides an excellent synthesis for demonstrating what Frank attempts to
demonstrate.
of Frank's theoretical model, and shows that
the only way in which Frank can "demon- To summarize, Frank's direct contribution
strate" that all the periphery is capitalist and to our understanding of the process of Latin
has been since the colonial period is by using American development is largely limited to
the concept of capitalism in a sense which is his critique of dualist models for Latin Amer-
erroneous from a Marxist point of view, and ica. Nevertheless, his indirect contribution is
useless for his central proposition, that of considerable. By this I mean that his work
showing that a bourgeois revolution in the pe- has inspired a significant quantity of research
riphery isimpossible. As regards this point by others (whether to support or rebut his ar-
then, Laclau concludes that Frank makes no guments), intheir respective disciplines, par-
contribution, leaving the analysis exactly ticularly inthe sociology of development.
where it started.
The central line of Frank's thought regard-
Robert Brenner (1977) takes Laclau's ing the "development of underdevelopment"
is continued, though from a critical point of
analysis of Frank (as well as Dobb's critique
of Sweezy), and demonstrates how the work view, by the Brazilian sociologist Theotonio
of Sweezy, Frank and Wallerstein — bril- dos Santos, for whom
liantly summarized and analysed by him —
are doomed to negate the model put forward the process under consideration [Latin American
first by Adam Smith in The Wealth of Na- development] rather than being one of satelliza-
tion as Frank believes, is a case of the formation
tions, Book 1, but
of a certain type of internal structures conditioned
by international relationships of dependence
because they have failed ... to discard the under-
lying individualistic — mechanist presuppositions (1969, p. 80).
of this model, they have ended up by erecting an
Dos Santos distinguishes different types of
alternative theory of capitalist development which
is, in its central aspects, the mirror image of the relations of dependency (essentially colonial,
industrial-financial and industrial-technolog-
"progressist" thesis they wish to surpass. Thus,
very much like those they criticize, they conceive ical, the latter having grown up since the Sec-
of (changing) class relations as emerging more or ond World War), and consequently distin-
less directly from the (changing) requirements for guishes different kinds of internal structures
the generation of surplus and development of pro- generated by them. Dos Santos emphasizes
duction, under the pressures and opportunities en- the differences and discontinuities between
gendered by a growing world market. Only,
the different types of dependency and be-
whereas their opponents tend to see such market- tween the internal structures which result
determined processes [the development of trade
from them, while Frank himself stresses the
and the division of labour], as setting off, auto-
matical y, adynamic of economic development, continuity and similarity of dependency re-
they see them as enforcing the rise of economic lations ina capitalist context. In other words,
backwardness. As a result, they fail to take into while Frank wishes to emphasize the similar-
account either the way in which class structures, ities between economic structures in the
once established, will in fact determine the course times of Cortez, Pizarro, Clive and Rhodes,
of economic development or underdevelopment and between those and the structures typified
over an entire epoch, or the way in which these by the activity of multinational corporations,
class structures themselves emerge: as the outcome dos Santos is more concerned with the differ-
of class struggles whose results are incomprehen- ences and discontinuities between them.
sible in terms merely of market forces (Brenner,
1977, p. 27). There is within dos Santos's analysis the
beginnings of an interesting attempt to break
Thus the way in which Frank uses the con- with the concept of a mechanical determi-
cepts "development" and "underdevelop- nation of internal by external structures
which dominated the traditional analysis of
ment" seems incorrect from a Marxist point
 142 DEVELOPMENT PERSPECTIVES

the left in Latin America, and which partic- dency which claims to be a theory of under-
ularly characterized Frank's work. One per- development should satisfy two criteria:
ceives initially in his analysis the perception
(i) it must lay down certain characteristics of de-
not only that both structures are contradic- pendent economies which are not found in non-de-
tory, but that movement is produced pre- pendent ones;
cisely through the dynamic of the contradic- (ii) these characteristics must be shown to affect
tions between the two. Nevertheless, as he adversely the course and pattern of development of
proceeds in the analysis he re-establishes, lit- the dependent countries (1975, p. 800).
tle by little, the priority of external over in-
ternal structures, separating almost meta- If crucial features of "dependence" can be
physically the two sides of the opposition — found in both dependent and "non-depen-
the internal and the external — and losing the dent" economies, the whole conceptual
schema is defective. And if it does not satisfy
notion of movement through the dynamic of
the contradictions between these structures. the second criterion, that is, if particular fea-
tures of dependency cannot be demonstrated
The analysis which begins to emerge is again
to be casually related to underdevelopment,
one typified by "antecedent causation and
we would be faced not with a "theory of
inert consequences." The culmination of this
process is his well-known formal definition of Latin American underdevelopment" but sim-
ply with a catalogue of social, political, eco-
dependency, which because of its formal na- nomic and cultural indicators, which will not
ture isboth static and unhistorical; it is found
in his 1970 article in the American Economic help us to understand the dynamic of under-
Review: development inLatin America.
Lall goes on to analyse the principal char-
Dependence is a conditioning situation in which acteristics commonly associated with depen-
the economies of one group of countries are con- dent economies and concludes that it appears
ditioned bythe development and expansion of oth- that the technique is
ers. A relationship of interdependence between
two or more economies or between such economies to pick off some salient features of modern capi-
and the world trading system becomes a dependent talism as it affects some less developed countries
relationship when some countries can expand and put them into a distinct category of depen-
through self-impulsion while others, being in a de- dence (1975, p. 806).
pendent position, can only expand as a reflection
of the dominant countries, which may have posi- He goes on to consider the possibility that
tive or negative effects on their immediate devel- the characteristics associated with the depen-
opment (1970, pp. 289-90).
dent economies could have a particular cu-
Lall (1975) offers an interesting critique of mulative effect when occurring together, but
a number of dependency studies. He argues finds no conclusive evidence. He concludes
that the characteristics to which underdevel- then that such a concept of dependency
opment in dependent countries is generally
attributed are not exclusive to these econo- applied
to less developed countries is impossible to define
mies, but are also found in so-called "non-de- and cannot be shown to be causally related to a
pendent" economies, and that therefore they continuance of underdevelopment (1975, p. 808).
are properly speaking characteristics of cap-
italist development in general and not neces- It is not surprising then that
sarily only of dependent capitalism. He fur- one sometimes gets the impression on reading the
ther argues that such analyses are not
literature that "dependence" is defined in a circu-
surprisingly unable to show causal relation- lar manner: less developed countries are poor be-
ships between these characteristics and cause they are dependent, and any characteristics
underdevelopment. that
800). they display signify dependence (1975, p.
Lall argues that any concept of depen-
 143 RADICAL PERSPECTIVES

References Lumpen Development, Dependence, Class

r» r» men r i- •»!•-• j ; and Politics in Latin America, reprinted,
Baran, P., 1957 La Econorma Pohuca del New
Yq Reyiew p ^^
r» Crecimiento,
r» ,mi Mexico,
^l F.C.E., • ■ 1969.
c , v. TLaclau,
, ^E., in-,
197 1 , „-.Feudalism
-, .. ' ,.
and capitalism
Brenner, R., 1977, The origins of capitalist T A. A „ ,r , . „ .
, , °r 0 r... in Latin America, New Left Review,
development: a critique of neo-Smithian M . ig71 iq ?e
Marxism," 7V>u> Le/f nc itev/ew,
no No. 104, TLall,„ S.,
a, 1Pk-c
1975, ttTIs ',PP
dependence, r , con-
a useful
T , A > mnn
July-August
i^ o %u 1 977, pp. 25-93.
• ,nm r j j '
cept in analysing 1 • underdevelopment?*o»
A A 1
Dos Santos, Theotonio, 1969, La dependencia „/ , , ~ , , ,. , - XT /I ir.-r
. . y las
economica . alternativ
' .. as de, F ,. World Development, Vol. 2, No. 11, 1975,
cambio en 7QQ_8in
America Latina, Mexico, November 1969. ,17 ,, 4 T m-™ tl i^ j n/
r i a ^ mr-r ^ J i- j rr j Wallerstein, I., 1974, The Modern Worldu
Frank,
, . A. G., 1967,
+ . r\' Capitalism
a and Under-
tj.'i System: Capitalist Agriculture and, the
0 ~ ' ,. . ' . . .
development in Latin America: Historical A- r.iT r tx^ /j i-
0n?T?ft
in *e f^T
the Sixteenth Century,Wo'ld- Eco"0^
New York, Ac-
Monthly of
Studies Review 1967. New York,
and Brasil,
Chile Press,
ademic Press, 1974.
Frank, A. G., 1970, Lumpenbourgeoisie:

Comment
We consider other aspects of dependency theory in connection with foreign investment and
multinational corporations in Chapter V. In Chapter VIII, the discussion of international
trade as a mechanism of international inequality is also related to dependency theory.
Within the school of dependency, different meanings are accorded to the concept of "de-
pendence," and different analyses are offered to explain underdevelopment as a result of the
interplay between internal and external structures.
Variants of dependency theory can be considered in the following works: Paul Baran, The
Political Economy of Growth (1957); Dependence and Underdevelopment in the New World
and the Old, special issue of Social and Economic Studies (March 1973); Samir Amin, "Un-
derdevelopment and Dependency," Journal of Modern African Studies, Vol. 10, No. 4 (1972);
Samir Amin, Unequal Development (1976); Fernando Henrique Cardoso, "Dependency and
Development in Latin America," New Left Review (July-August 1972); Cardoso and Enzo
Faletto, Dependency and Development in Latin America (1979); Andre Gunder Frank, Cap-
italism and Underdevelopment in Latin America (1967); Andre Gunder Frank, Latin Amer-
ica: Underdevelopment or Revolution? (1969); Andre Gunder Frank, Lumpenbourgeoisie and
Lumpendevelopment (1972); C. Furtado, Development and Underdevelopment (1976); T. dos
Santos, "The Structure of Dependence," American Economic Review (May 1970); O. Sunkel,
"National Development Policy and External Dependence in Latin America," Journal of De-
velopment Studies (October 1969); A. Emmanuel, Unequal Exchange (1972); I. Wallerstein,
"Dependence in an Interdependent World," African Studies Review, Vol. 17, No. 1 (1974);
Ronald H. Chilcote, "Theories of Dependency: The View from the Periphery," Latin Amer-
ican Perspectives (Spring 1974); Enmer L. Bacha, "An Interpretation of Unequal Exchange
from Prebisch-Singer to Emmanuel," Journal of Development Economics, Vol. 5 (1978); Her-
ald Munoz, ed., From Dependency to Development (1981).
Some critical discussions of dependency theory are Sanjaya Lall, "Is 'Dependence' a Useful
Concept in Analysing Underdevelopment?" World Development, (November/ December
1975); P. J. O'Brien, "A Critique of Latin American Theories of Dependency," in I. Oxaal
et al., eds., Beyond the Sociology of Development (1975); Benjamin J. Cohen, The Question
of Imperialism (1973); Alec Nove, "On Reading Andre Gunder Frank," Journal of Devel-
opment Studies (April-July 1974); Colin Leys, "Underdevelopment and Dependency: Criti-
cal Notes," Journal of Contemporary Asia (1977).
 144 DEVELOPMENT PERSPECTIVES

A number of country studies contradict the dependency thesis: Robert Kaufman et al., "A
Preliminary Test of the Theory of Dependency," Comparative Politics (April 1975); David
Ray, "The Dependency Model and Latin America: Three Basic Fallacies," Journal of Inter-
American Affairs and World Studies (February 1973); Patrick J. McGowan, "Economic De-
pendency and Economic Performance in Black Africa," Journal of Modern African Studies,
Vol. XIV, No. 1 (1976); Patrick J. McGowan and Dale L. Smith, "Economic Dependency in
Black Africa: An Analysis of Competing Theories," International Organization (Winter
1978).
 II.D. ECONOMICS FOR
DEVELOPMENT-NOTE

The foregoing selections give credence to economics.2 Many did argue in the 1940s and
Keynes's appraisal that "The theory of eco- 1950s that "orthodox monoeconomics" de-
nomics does not furnish a body of settled con- scended from Anglo-Saxon economics and
clusions immediately applicable to policy."1 was relevant only for the special case of ad-
In few, if any, areas of economic policy does vanced industrial capitalist economies, not
the distance between theory and policy tend for the less developed countries. But this ar-
to be as great as it is for development prob- gument has lost force as experience has dem-
lems. The various theories of development onstrated that the principles of monoecon-
discussed in this chapter are ways of thinking omics— especially those delimiting efficient
systematically about the development pro- resource allocation — cannot be overlooked in
cess— but in abstract terms and from the out- any economy.
side. They relate to the economics about de- In reconsidering development policies,
velopment. This approach, however, is far however, the development economist must
different from the economics for develop- still rethink some basic premises of both price
ment— the economics that the development theory and income theory. Price theory has to
practitioner must use in formulating and ad- be extended to consider problems of inter-
ministering an actual development program temporal allocation of resources and the im-
in a specific environment. In the administra- plications ofprices for the mobilization of re-
tion of a development program, development sources. A"correct set of prices" is needed
practitioners cannot rely simply on the ana- not only for allocational objectives during a
lyst's view of development "from the out- given period of time, but even more impor-
side." They must apply economic principles tantly as proper signals and sufficient incen-
to specific problems embedded within the tives for the fuller utilization and mobiliza-
general development process. This they can- tion of resources over time. The development
not do by merely being knowledgeable about economist needs to understand more pre-
the latest development model.
cisely just what is meant by "getting prices
To improve the quality of policymaking, right" in this dynamic context.
the literary economist is now being asked to The major concern of neoclassical econom-
provide more operational concepts and em- ics has been with the short-period analysis of
pirical constructs instead of "empty stages of resource allocation. In this analysis, it is as-
growth" or ambiguous phrases such as "big sumed that population, "state of the arts," in-
push" or "balanced growth." The mathemat- stitutions, and supply of entrepreneurship are
ical economist is being asked to be more all given. But the very essence of the devel-
aware of the limits of mathematical pro- opment process is that these parameters be-
gramming ina less developed country. The come variables. What is normally taken as
econometrician is being cautioned to recog- "given" in static analysis must actually be
nize the simplifying assumptions that are explained when the problem is one of secular
necessary to keep econometric models within change. Economics must be broadened — in-
manageable dimensions. deed, must at times become interrelated with
Moreover, most development economists
no longer look for a completely different sub- 2See Albert O. Hirschman, Essays in Trespassing,
discipline of development economics that 1981, pp. 1-14. Gunnar Myrdal remains a major ex-
would be in contrast to the use of neoclassical ception: "Need for Reforms in Underdeveloped
Countries," in S. Grossman and E. Lundberg (eds.),
The World Economic Order: Past and Prospects,
'J. M. Keynes, "General Editorial Introduction" to
Cambridge Economic Handbooks Series.
1981, pp. 501-25.

145
 146 DEVELOPMENT PERSPECTIVES

other disciplines — in order to explain the de- abra of input-output analysis, linear programming
terminants ofpopulation growth, technologi- and other mysteries, the preparation of long-term
cal progress, institutional change, and in- plans is a very beguiling operation, and one full of
crease in the supply of entrepreneurship. intellectual challenge even when the practical out-
Neoclassical economics must also be ex- come may be highly obscure. On the other hand,
the short-term management of an economy that is
tended toincorporate more analysis of distri- subject to rapid changes in circumstances offers
bution issues. Problems of absolute poverty harsher and more restricted choices with little
and inequality cannot be downgraded to the
small technical problem of efficiency in re- scope for subtlety.3
source allocation. And a welfare economics Whether it be the result of the operation of
that prides itself on avoiding interpersonal an international "demonstration effect" in
comparisons and judgments about the distri- governmental policymaking or the dominant
bution of income must appear anemic com- influence of intellectuals on planning com-
pared with the emphasis on basic needs. The missions, there has been in many developing
task is to integrate neoclassical economics countries an overly keen receptivity for the
with employment and distribution issues. most refined model, the newest technique, the
The conventional Keynesian type of in- latest element of expertise. It has. however.
come theory also leaves the development become increasingly apparent that the use of
economist perplexed about the causes of un- the latest technique is likely to be premature,
and that the attraction to the highest style
employment and what might be effective pol-
icies for fuller employment. The investment analysis may be without practical effect.
multiplier of Keynesian economics was first As the Hungarian economist Janos Kornai
identified as an employment multiplier, and has observed, the implementation of mathe-
more employment was expected to parallel matical planning is dependent on the degree
greater output. But notwithstanding the rise of maturity of nonmathematical planning. It
in their investment ratios and growth in is only clear what policy objectives are worth
GNP, many developing countries have at the striving for when there has already been es-
same time experienced rising unemployment tablished an organized institutional, non-
and underemployment. The unemployment mathematical form of planning. For. as Kor-
nai states.
problem may be related to the inappropriate
set of prices and to structural phenomena in a single model only a few hundred relationships
that have been ignored by income theory. As and constraints can be considered. But people
indicated frequently in this book, the unem- working in the central planning agencies and
ployment problem in LDCs is not of the
lower-level
hundreds of institutions
thousands ofand enterprises
further ""sense"
constraints and
Keynesian variety, and the policy implica-
tions of income theorv have to be modified for relations, and they can give expression to these in
LDCs. their own estimates. Mathematical planning will
Economists in the developing countries are develop successfully only when it develops as one
also beginning to realize that in many cases element of well-prepared and well-oriented insti-
tutional planning, connected by many threads with
they have neglected solutions to immediate
and crucial needs by overreaching — by at- real economic life in developing countries.4
tempting to use sophisticated techniques of The overly refined mathematical approach
analysis and highly formalized long-range to planning has displayed three prominent
planning models. The vogue for long-term
plans to the neglect of a more concentrated 3A. K. Cairncross. The Short Term and the Long
in Economic Planning, Tenth Anniversary Lecture.
effort to improve current policymaking has Economic Development Institute. Washington. D.C..
been decried by Sir Alec Cairncross in these January 6. 1966. p. 7.
words:
Manos Kornai. "Models and Policy: The Dialogue
For young mandarins who have returned from the between Model Building and Planner." in C. R.
Blitzer et al. (eds.). Economy-wide Models and De-
universities of the West, initiated into the abracad- velopment Planning. 1975. chapter 2.
 147 ECONOMICS FOR DEVELOPMENT— NOTE

types of bias: a bias toward macromodels in lyzing a problem within the context of a poor
plan formulation to the relative neglect of the country. In particular, this calls for special
microeconomic aspects of planning (ranging care in identifying different institutional re-
from project analysis to the installation of in- lations.5 inassessing the different quantita-
centives and enlistment of mass participa- tive importance of some variables, and in al-
tion): abias toward the quantitative aspects lowing some elements that are usually taken
of planning to the relative neglect of other de- as "given'' to become endogenous variables in
velopmental forces that are not quantifiable development analysis.
but are of crucial importance (for instance, When examining the selections of the fol-
many aspects of human resource develop- lowing chapters, we should therefore be alert
ment and sociocultural and political changes to the efforts being made to adapt our con-
for which data are limited); and a bias to- ventional way of thinking about economic
ward concentration on the formulation of a problems to the particular context of devel-
development plan without due regard for its opment problems. For some issues, the ad-
implementation. aptation ismade explicit: the key factors dis-
Although it cannot be gainsaid that prob- tinguishing the situation in a poor country
lems of underdevelopment are different in de- from that in an advanced economy are rec-
gree— and. to some extent, even in kind — ognized, and a new set of assumptions is ex-
from those encountered in developed coun- plicitly adopted; for example, the existence of
tries, nevertheless it would be overreacting to a subsistence sector, or disguised unemploy-
conclude that entirely different economic ment, or an unorganized money market may
tools and principles are needed to analyze require modification of the usual assumptions
these problems. The progress that has been of macroanalysis. In other instances, some
made in development economics has actually concepts are introduced out of consideration
been mainly within the framework of tradi- for the special conditions of poor countries —
tional economic analysis. Many tools of tra- for instance, "social dualism/' the structur-
ditional economics and many principles of alist view of inflation, or noneconomic crite-
accepted economic theory have proved di- ria for investment. Whether they are intro-
rectly applicable to the problems of poor duced explicitly or are only implicit, we
countries, and some conceptions and tech- should be aware that these considerations of
niques can become more useful with some the relevance of conventional economic anal-
ready-made modifications or extension. ysis constitute a major theme of this book.
What has become clear, however, is that This problem of adapting our accepted tools
we must frequently be prepared to depart and principles is. in a fundamental sense, an
from traditional assumptions when analyzing overriding general issue in development
problems of development. We must recognize economics.
that the premises of accepted economic
5At first sight, it might be thought that behavioral
theory may have to be altered to make the relations are also of a different kind. On closer ex-
theory more immediately relevant to coun- amination, however, the usual postulates of rational-
tries that have a different social system and ity and the principles of max;mization or minimiza-
economic structure from those to which tion appear to have quite general applicability. For
illustrative evidence, see P. T. Bauer and B. S.
Western economists are accustomed. The
Yamey. The Economics of Under-Developed Coun-
task of development economists is made dif- tries. London. 1957, pp. 91-101; W. J. Barber. "Eco-
ficult not because they must start afresh with nomic Rationality and Behavior Patterns in an
a completely new set of tools or because they Underdeveloped Area: A Case Study of African Eco-
nomic Behavior in the Rhodesias." Economic Devel-
confront wholly different problems, but be- opment and Cultural Change. April 1960. pp. 237-
cause they must acquire a sense of the differ- 51: W. O. Jones. "Economic Man in Africa." Food
ent assumptions that are appropriate to ana- Research Institute Studies. May 1960. pp. 107-34.
CHAPTER

Surplus
Labor and
Dualistic
Development

Having considered in the preceding two chapters the general context of developmental prob-
lems and some approaches to their analyses, we are now ready to turn directly to a more
detailed examination of the development process. We begin by focusing on an outstanding
characteristic of LDCs — that of "dualism." In many LDCs, a modern commercialized in-
dustrial sector has developed alongside a traditional subsistence agricultural sector, resulting
in what is termed a dual economy. Although only relatively few LDCs have yet entered into
a Rostovian process of self-sustained growth, most of them do exhibit some elements of mod-
ernization inone sector or in parts of their economies. The contrast in economic and social
organization between the advanced exchange economy and the backward indigenous economy
is one of the most striking — and puzzling — characteristics of a poor country. Professor Kuz-
nets relates the dualism of the domestic economy to the fact that the LDCs are late-devel-
oping countries.
It is a crucial characteristic of the presently underdeveloped countries that they are underdeveloped
while others are much more developed — and this situation has lasted a long enough time to affect mark-
edly the structure of the underdeveloped countries; and also to cover several phases in the changing
impact of the developed upon the underdeveloped parts of the world.
This impact, in the broadest terms, is the introduction of a modern component into the structure of
the otherwise traditional underdeveloped countries. The magnitude of this component, its specific eco-
nomic content, and the way in which it was introduced, confined, or encouraged, could and did vary
widely. In a territory with colonial status a substantial modern sector often emerged but was organized
by Western entrepreneurs of the metropolitan country. In a politically independent underdeveloped
country with an export sector oriented toward the developed countries' markets the economic organiza-
tion was and is quite different from that in a country with domestically oriented agriculture or industry.
And in some others, a few members of the native elite educated in some Western lore are participating
in administration and attempting to introduce Western elements into what may still be a purely tradi-
tional economic society.

149
 150 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

Obviously, in the countries that are still underdeveloped, this modern component, even if in existence
for a long period, has not expanded sufficiently to shift the country to developed status, and in most cases
has not even raised its per-capita product to an intermediate level. Given the long coexistence of devel-
oped and underdeveloped economies, it is no exaggeration to argue that a major result of such coexistence
in the underdeveloped economies is their dual structure. There are two distinct components, the modern
and the traditional; and in marked contrast with the past record of the presently developed countries,
the two components continue to perform without the modern one. despite its greater productivity, rapidly
outpacing the other. It is the persistence of the dual structure and the confinement or limitation of the
modern sector that are crucial, for the two have operated simultaneously but for a shorter period in the
developed countries also.1
If we are to identify the structural relationships involved in the development process, we
must understand the dual structure of the modern and traditional, and we must consider how
the acceleration of development will entail a higher rate of structural transformation in a
developing country. Instead of using an aggregative model, we may gain more insights from
the study of interactions among sectors. We therefore explore in this chapter several questions
about dualism: What conditions have given rise to a dual economy? In what sense is a dual
economy also a labor surplus economy? How can the absorption of the indigenous economy
into an expanding modern economy be accomplished?
One explanation of dualism, propounded principally by J. H. Boeke and other Dutch econ-
omists in their studies of Indonesian development, emphasized the differing social organiza-
tions and cultural contrasts that result in "social dualism." The attention to dualism now
centers less on its sociocultural aspects and more on its purely economic features, especially
in terms of the effects of a dual economy on the pattern of development. Underemployment
is commonly believed to be a dominant feature of densely populated, underdeveloped coun-
tries, and the labor force is continually increasing with population growth. It is therefore
important to consider how dualism is related to the problem of providing adequate employ-
ment opportunities for the currently underemployed workers and for the increase in the labor
force.
There has been much confusion over the phenomena of unemployment, underemployment,
and disguised unemployment in the traditional sector of a poor country. The selections in
section III. A are therefore designed to clarify the effects of a dualistic structure on employ-
ment by examining the concept of the "labor surplus economy."
Several models of development have focused on the structural formation of a dual economy
(section III.B.). Prominent among these is the Fei-Ranis model of a labor surplus economy,
which is characterized by the coexistence of a relatively large and overwhelmingly stagnant
subsistence agricultural sector in which institutional forces determine the wage rate, and a
relatively small but growing commercialized industrial sector in which competitive conditions
shape the labor market. In such a labor-surplus dualistic economy, labor is not scarce, while
capital is extremely so. Development therefore requires that "the center of gravity must con-
tinually shift towards industry through the continuous reallocation of labor from the agricul-
tural to the industrial sector: the related criterion of 'success' in the development effort is thus
a rate of industrial labor absorption sufficiently rapid to permit the economy to escape from
the ever-threatening Malthusian trap."2
Perhaps most celebrated of the labor surplus models is Sir Arthur Lewis's analysis of "De-
velopment with Unlimited Supplies of Labor." The Note on "Intersectoral Relationships in
a Dual Economy" (III.B. 1) summarizes this model, emphasizing the interaction between an
advanced "capitalist" sector and an indigenous "noncapitalist" sector in a developing econ-
'Simon Kuznets, "Notes on Stage of Economic Growth as a System Determinant." in Alexander Eckstein
(ed.), Comparison of Economic Systems (1971), pp. 256-7.
2John C. H. Fei and Gustav Ranis, Development of the Labor Surplus Economy (1964), p. 3.
 151 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

omy, and indicating how resources can be drawn into the modern exchange system through
capital accumulation in the expanding capitalist sector. The Lewis model is helpful in ex-
plaining the mechanism by which the proportion of domestic savings in the national income
increases during the course of development of a dual economy whose growth is the result of
the expansion of capitalist forms of production. As the model explains, growth in the capitalist
sector turns on the higher than average propensity to save from profit income, as well as on
the rise of the share of profits in the national income in the early stages of development. The
noncapitalist sector, in turn, serves as a reservoir from which the expanding capitalist sector
draws labor. The model therefore has significant policy implications for labor absorption and
for the employment problem — a problem to which we return frequently in subsequent
chapters.
Rural-urban migration, as examined in section III.C, is central to the problem of growing
urban unemployment. The model by Todaro emphasizes a migration function that hypothe-
sizes that the relevant urban income is the present value of expected earnings (that is, a ra-
tional calculation by an individual migrant that allows for the probability of the migrant
obtaining urban employment). In a distorted labor market, growing urban unemployment is
consistent with equilibrium in this model, and rural-urban migration is assumed to occur until
there is equality between the actual rural wage and the expected urban wage. Although the
specific character of the probability function for urban employment can be questioned, the
migration function is analytically useful in explaining why policies that are devoted only to
raising urban labor demand cannot be relied on to reduce urban unemployment.
The high rate of population growth has also been the main cause of the high density of
land occupation in the rural areas; this, in turn, has been an underlying factor in the high rate
of migration from rural to urban areas in many countries. The accelerated rural-urban labor
migration has produced extremely high rates of growth of the active population in urban areas
that have far exceeded the growth of urban employment opportunities, culminating in rising
levels of urban unemployment. Given the limited employment opportunities in the modern
sector of urban areas, an increasing share of the urban labor force has become unemployed
or has drifted into the tertiary sector, or what has come to be called the "informal sector."
As unemployment and underemployment have been transferred from the rural sector, the
absorption of labor in the informal sector has become in many respects an extension into the
urban areas of the traditional rural subsistence economy.
As discussed in selection III.C. 2, an ILO mission to Kenya has given considerable emphasis
to the role of the "informal sector." In almost every area of Kenyan activity, there appears
to be a sharp and analytically significant dualism between the protected, organized, large-
scale, foreign-influenced, formal sector and the unprotected, unorganized, small-scale, family-
based, and essentially self-reliant, informal sector. The mission argued that to infer the
growth of total employment, let alone the volume of unemployment, from the growth of for-
mal sector employment can be very misleading. It has become a general phenomenon in the
urban areas of the LDCs that during the past three decades the number in the informal sector
has increased absolutely and as a proportion of the labor force. The size of urban areas has
commonly doubled within a decade, and with such high urban growth rates, unemployment
and underemployment have also risen.
We can begin to investigate the problem of labor absorption by considering what has gone
wrong with the Lewis model of labor transfer — why urban unemployment has increased and
why the labor reservoir in the noncapitalist sector still remains so large after more than three
decades of developmental effort since Lewis propounded his model. Section III.D examines
how some important conditions of the Lewis model have failed to be fulfilled in reality. Other
selections, however, indicate ways by which the explanatory power of the model could be
improved. Specifically, it is suggested that the analysis give more attention to the interrela-
 152 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

tions between the rural and urban sectors, to the existence of an informal (nonmodern, non-
capitalist) sector alongside the organized (modern, capitalist) sector in urban areas, and to
the need for improving the quality of employment in the informal sector as well as the quan-
tity of employment throughout the economy. Capitalist and noncapitalist subsectors are
within both the rural and urban sectors, giving the economy the characteristic of "double
dualism." Unemployment and underemployment are not confined to the rural areas, but com-
parable problems of open unemployment, inadequate incomes, and low productivity have also
arisen in urban areas.
Estimates of the magnitude of the employment problem are necessarily imprecise because
of the ambiguities in the conceptual meaning of "unemployment" and "underemployment,"
lack of data, and statistical pitfalls in measurement.
Section III.E outlines various dimensions of the employment problem. This section empha-
sizes that the employment problem is not one problem but three: (1) the shortage of work
opportunities, (2) inadequate incomes from work, and (3) underutilized labor resources.
Within each of these categories, different types of unemployment and underemployment can
be identified, both visible and "disguised."
An entire set of policies is now needed that will concentrate on employment in the urban
informal sector and the rural sector, as well as on demand in the organized urban sector. The
underutilized labor reservoir in the traditional sector and open unemployment in the urban
area are problems in their own right. Employment problems, however, pervade the economy,
and pervasive policies are therefore needed, as emphasized in section III.F. Several of this
chapter's selections can be read as a plea for a set of interrelated policy measures rather than
partial measures. While some specific policies are discussed in section III.F, subsequent chap-
ters will also relate to the employment problem, especially in connection with the transfer of
technology, income distribution, education, and trade expansion. Section III.F analyzes the
central question of whether there is a conflict between output and employment. Is maximum
production compatible with maximum employment? From a comprehensive analysis of the
employment problem, we should become aware that the objectives of greater utilization of
labor, diminution of poverty, and improved income distribution are complementary, not com-
petitive objectives.
 III.A. THE LABOR SURPLUS
ECONOMY
III.A. 1 . Labor Surplus on the
Land*
The concept of surplus farm labor has at- nities may be lacking because of the rigid so-
tracted increasing attention in the underde- cial structure, and may actually be
veloped countries, especially in Asia. The decreasing because of the decline of tradi-
simplest definition of it implies that some tional handicraft industries due to the com-
labor could be withdrawn from subsistence petition ofimported manufactures.
farming without reducing the volume of farm With the growing pressure of people on the
output. In technical terms, the marginal pro- land, farms become smaller and smaller.
ductivity oflabor is believed to be zero. If this What is more, farms are divided and subdi-
is true, it has some far-reaching implications. vided into tiny strips and plots. Accordingly
But is it true? Is it even conceivable? Some it seems to me that agricultural unemploy-
economists have had serious doubts on this ment in densely populated peasant commu-
score. It must be admitted that the idea of nities may be said to take at least two basic
"disguised unemployment," as it is usually forms: (1) underemployment of peasant cul-
called, has sometimes been carelessly for- tivators due to the small size of farms; (2)
mulated and inadequately substantiated. unemployment disguised through fragmen-
The subject must be viewed in relation to tation of the individual holding. . . .
the general population problem. The crucial To the extent that the labor surplus is
fact is that world population has doubled in absorbed — and concealed — through frag-
the last hundred years. About two-thirds of mentation, itcannot be withdrawn without
the increase has taken place in the underde- bad effects on output unless the fragmenta-
veloped areas, chiefly in Asia, largely tion is reversed and the holdings are consoli-
through a fall in death-rates. This has been dated. Over a limited range the marginal pro-
part of the uneven impact of Western civili- ductivity oflabor might be zero without any
zation on the rest of the world. Now in the such reorganization. It could be zero over a
poorer countries as a rule the majority of the much wider range if the remaining factors of
population works in agriculture to start with, production were appropriately reorganized,
for basic and obvious reasons. Just as food is which would require for one thing a consoli-
the major item of consumption in low-income dation ofplots. Appropriate reorganization of
communities, so the struggle for food takes the other factors of production is clearly a
up most of their time and resources. In such necessary and a reasonable pre-requisite for
countries rapid population growth naturally purposes of policy as well as analysis.
leads, and in some has already led, to excess There are a number of empirical studies
population on the land. that tend to confirm this general picture. The
Consider for a moment the effects of pop- evidence can never be entirely satisfactory in
ulation growth in a community of peasant a matter such as this where some things, in-
cultivators. Numbers are increasing while cluding the weather, would have to be held
land, capital and techniques remain un- constant and others subjected to a reorgani-
changed. Alternative employment opportu- zation which may necessitate a revolutionary
change in rural life, bringing inevitably other
changes with it. Nevertheless the connection
*From Ragnar Nurkse, "Excess Population and
between over-population and fragmentation
Capital Construction," Malayan Economic Review, goes a long way to make the existence of sur-
October 1957, pp. 1, 3-5. Reprinted by permission. plus farm labor plausible.

153
 154 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

On a theoretical view of the matter it is

labor may well be zero or even negative.2
clear that excess population can be so great Still, it need not be as low as zero. Con-
in relation to land and capital that the mar- versely, iand
f when labor's marginal product
ginal productivity of labor is reduced to zero. is zero, the average product need not be as
There are, however, two reasons why some low as the physical subsistence level; it may
economists have found this idea difficult to be a little above that level.
accept. First, anyone trained in Western eco- In any case, excess population necessarily
nomics would have to ask: Who would em- implies that the marginal product is less than
ploy these people if their product is zero? Or the average. In this state of affairs further
else one might ask: How can these marginal population growth brings down the average
people live, what do they eat, if they really level of consumption. Why? Because the ad-
produce nothing? ditional labor contributes less than the aver-
The answer to the first question is that in age worker previously, and so it pulls down
many countries the wage-labor system, which the average product per head. If the average
Western economists are apt to take for product is as low as we know it to be, it does
granted, hardly exists. The prevailing condi- not seem far-fetched to suppose that in some
tion in subsistence farming is one of peasant cases the marginal product of labor may be
family labor. zero. The upshot of the argument does not, of
The answer to the second question — How course, depend on its being exactly zero, al-
can they live? — is that they live by sharing though this is a convenient case on which to
more or less equally in the total product of concentrate the analysis. The essential point
the farm, which includes the product of intra- is that the marginal workers live, in effect, on
marginal labor and of any land and capital a subsidy if their own contribution to output
goods the peasants may own. The product is less than their intake of food and other
from these factors goes into the same pot and necessities.
the members of the household eat out of that The relationship between total product and
same pot. These institutional arrangements total population is illustrated in the accom-
are foreign to the economies of business en- panying diagram (Figure 1). Average prod-
terprise, and so the conditions which they uct per head is reflected in the slope of the
make possible may seem paradoxical. vector from the origin to any point on the
If this sharing of food is considered a little curve. The marginal product is reflected in
further the ultimate limit to the multiplica- the slope of the curve itself at any given
tion of people on the farms becomes starkly point. The average product per head reaches
plain.1 If the average total product per person a maximum at A (where the angle AOL is
falls below the physical level of subsistence,
the outcome is the Malthusian state of star-
vation cutting down numbers or at least
checking their further increase. At the point
where this average product equals the physi-
cal subsistence level, the marginal product of
'The analysis concentates on the subsistence farm
sector. The reader should bear in mind that even the
most backward economy usually contains other sec-
tors also, including export production, commerce,
government and even some industrial activity.
2Professor J. E. Meade has shown this very clearly
in his book, The Theory of International Economic
Policy; Vol. II, Trade and Welfare, issued under the L M
auspices of the Royal Institute of International Af- Population
fairs, London, Oxford University Press, 1955, chapter
6 and appendix 1. FIGURE 1.
 155 THE LABOR SURPLUS ECONOMY

largest) and declines thereafter as population regard to population trends in the Western
increases further. The marginal product be- world. But if we consider Asia over the last
comes zero at B (where the average product hundred years, I am not sure that the objec-
measured by the angle BOM may still be tion has much force. I began by saying that
substantial). If population increases beyond the population explosion in Asia, due largely
M, the marginal product becomes negative to the fall in death-rates, reflects the uneven
and the average product continues to fall. If impact of Western civilization. The point is
we suppose that the average product at TV precisely that while population has doubled,
represents the absolute physical minimum of other things such as techniques, capital sup-
subsistence, then population cannot increase plies and cultivable land have remained too
beyond N. This supposition is of course much the same. Therein lies the whole prob-
purely arbitrary and illustrative. Actually the lem. Of course, there has been some advance
physical subsistence level of average product in these other things too, but not nearly at the
per head may lie, not in the range of negative same rate as in population. In Asia there has
marginal productivity, but conceivably at B been nothing like the advance that accom-
or somewhere between A and B where mar- panied population growth in the West. In this
ginal product is still positive. The diagram state of affairs it seems to me that the "opti-
merely illustrates the possibilities and does mum population" approach, questionable
not, of course, tell us what actually happens. though it may be in the West, has a good deal
It must be conceded that this view of the of validity in the East. The economic problem
matter is essentially that of "optimum popu- of the East has been largely a consequence of
lation" theory (in the diagram the optimum dynamic population growth in an otherwise
size of population is OL). Now this theory relatively static environment.
has sometimes been criticized as being an un- "Optimum population" theory directs at-
realistic exercise in comparative statics. It as- tention chiefly to the variation of average
sumes that nothing changes except the num- product as the size of population varies. We
ber of people — and, in response thereto, the have found it at least equally important to
volume of total product. It abstracts from, consider the marginal product. The question
and ignores, any connections that may exist might be asked: Why this obsession with the
between population size on the one hand and, margin? Why not stress the obvious fact of a
say, the state of techniques or the volume of low general level of productivity? The answer
capital on the other hand. It holds all other is that the marginal approach is useful here
things constant. Is this not bound to lead to a because of the need to take away some labor
distorted view of reality? from current production for work on capital
The criticism may be perfectly valid with construction.

III.A.2. The Concept of

"Disguised Unemployment"*

The term "disguised unemployment" is com- factors of some units of labor, nothing else of
monly used to designate a situation in which consequence or worth mentioning being
the removal from a working combination of changed, will leave the aggregate product of
the working combination undiminished, and
*From Jacob Viner, "Some Reflections on the Con- may even increase it. To say that there is
cept of 'Disguised Unemployment,' " in Contribuicoes
a Andlise do Desenvolvimento Economico, Livraria "disguised unemployment" is therefore
Agir Editora, Rio de Janeiro, 1957. Reprinted by equivalent to saying that in that working
permission. combination the marginal productivity of
 156 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

labor is zero or almost zero and may even be cific character, even though the chemical
constituency of the steel were invariant, and,
a negative quantity. The "unemployment"
may be only metaphorical, since there may moreover, it would readily find ways of
be hard work even at the margin, when "un- changing the chemical constituency of a ton
employed" must mean "unproductively em- of "steel" without reducing its suitability for
ployed." But sometimes it is intended to be its ordinary uses, and this not only in the long
realistically descriptive, as when it is used to run but in the very short run. As far as agri-
include seasonal unemployment; in such culture isconcerned, I find it impossible to
cases, I do not know what the adjective "dis- conceive of a farm of any kind on which,
guised" issupposed to mean. I will in this other factors of production being held con-
note treat "seasonal unemployment" as a dis- stant in quantity, and even in form as well, it
tinct phenomenon not obviously presenting a would not be possible, by known methods, to
serious problem and not obviously having any obtain some addition to the crop by using ad-
peculiar relationship to agriculture in under- ditional labor in more careful selection and
developed countries. As I look at the agricul- planting of the seed, more intensive weeding,
tural world with my inexpert eye, it seems to cultivation, thinning, and mulching, more
me that agricultural employment is most sea- painstaking harvesting, gleaning, and clean-
sonal, isleast continuous, in the temperate ing of the crop. Even supposing that there
zones where agriculture is most "developed" were such a farm, on which every product
and yields the highest levels of average and had technically and economically fixed ingre-
marginal product per labor-year. dients, labor would still have positive mar-
As an intermittent phenomenon, resulting ginal productivity unless there were not only
from the vagaries of weather and human fixed technical coefficients of production for
error, zero marginal productivity of labor in all the economically relevant potential prod-
agriculture is a commonplace concept. But ucts of the farm, but the proportions between
how can a priori the possibility of zero mar- the technical coefficients were uniform for all
ginal productivity of employed labor, as a of these products. For if these proportions are
chronic phenomenon, be plausibly estab- different as between different products, then
lished? One way that has been suggested is it will always be possible by appropriate
on the assumption that the (average and change in the product-mix, in the direction of
marginal) technical coefficients of production more production of those products whose
are constant, so that the addition to a work- labor technical coefficients are relatively
ing combination of more units of labor will high, to absorb productively any increment of
add nothing to the aggregate product unless labor.
additions are made also to the quantities used Unless one assumes non-economic motiva-
of all (or of some) of the other factors of pro- tion on the part of employers, there is diffi-
duction— Pareto's "fixed coefficients," or culty also in conceiving why they should hire
Frisch's "limitational factors." at any wage-rate additional units of labor be-
I am not aware that anyone has ever given yond the point at which they know the labor
a convincing illustration of a technical coef- will add less in value to the product than the
ficient which is "fixed" in a valid economic wage-cost, to say nothing of the case where
sense. The plausibility of the idea has re- the labor will add nothing to and may even
sulted,believe,
I from the confusion of chem- subtract from the product. The employer
ical ingredients of a product, or actual engi- may, of course, be ignorant as to the facts,
neering elements in a productive process, but I know of no experience to persuade me
with potential economic input-items in a pro- that the speculative economist is on such
ductive process. If iron ore, or coal, were as matters better informed than the experienced
expensive per ton as gold I am sure that the farmer in immediate touch with reality. This
steel industry would find ways of appreciably is probably what W. A. Lewis has in mind,
reducing the amounts of iron ore, or of coal, although I cannot find that he anywhere ex-
it uses to produce a ton of steel of given spe- plicitly says so, when he concedes that in ag-
 1 57 THE LABOR SURPLUS ECONOMY

riculture "disguised unemployment" occurs would be in the long run interest of the owner
only for peasant or self-employed labor, and of the crowded farm not to over-work it. Gen-
not for plantation labor.1 Since there is a uine unemployment on the farm, or employ-
good deal of plantation agriculture in under- ment of the "surplus" labor only on such
developed countries, this is an important lim- tasks as would not impair fertility, would in
itation of the applicability of the concept of the long run be more profitable than full em-
"disguised unemployment." But it raises its ployment, but the shortsightedness of the
own difficulties. In Brazil, for instance, I take owner, or the hunger of his family, might
it that agriculture is even in the same locali- nevertheless trap him into exploiting the
ties a mixture of hired labor on plantations, short-run marginal productivity of labor,
of self-employed labor on owned (or rented?) which I would expect to be always positive in
small farms, and of squatter labor. Should a situation such as here described. Lewis con-
there not be a tendency for equalization of cedes that this phenomenon of impairment of
the marginal productivity of labor in all soil-fertility through over-crowding would be
agrarian uses where labor can fairly readily present only in "over-populated areas," and
move from one type of use to another? Where lists China, India, Japan, Java, Egypt, some
there is labor-mobility, marginal productivity countries in the Middle East, Kenya, and
of labor must rise substantially above zero in some small islands as the only countries in
peasant or squatter agriculture or sink to zero this category. Latin America and Eastern
or near-zero on the plantations. This would Europe, which many writers have regarded
especially be the case ... if a member of a as subject to "disguised unemployment," are
peasant or squatter family would not lose ac- thus excluded by Lewis, and, as we have seen,
cess to the family supply of food by taking he would exclude also the plantation agricul-
employment on a nearby plantation. When I ture of any country.
was in Brazil, I heard of complaints by plan- Ragnar Nurkse has suggested that where
tation owners in districts in which there was
"disguised unemployment" prevails in agri-
also peasant and squatter agriculture of culture, itwould be desirable to transfer the
"shortage of hands" (falta de mao). I don't surplus labor off the farms to produce capital
see how this can be reconciled with the prev- goods, while keeping the consumption of food
alence ofzero marginal productivity of labor, by the population as a whole constant
whether on the plantations or for self-em- through taxation or direct controls.3 This
ployed agricultural labor. . . . would be relevant for Lewis's type of zero or
W. A. Lewis has suggested, as an expla- negative marginal productivity of agricul-
nation of "disguised unemployment" in ag- tural labor resulting from loss of fertility of
riculture that, when "there are too many per- soil through over-crowding and over-crop-
sons on too little land" the farmer cannot ping. But as I have pointed out, it would not
afford to keep cattle, so that the land gets no be a solution in the short run, for in the short
manure and land is put under the plough run transferring labor out of agriculture
which ought to be left in forest or in fallow; would, or might, reduce the total output. N.
and land is over-cropped, so that fertility is Koestner has objected against Nurkse's ar-
destroyed.2 . . . Given the situation as Lewis gument that it fails to take account of the
describes it, the long run marginal productiv- fact that the urban workingman needs more
ity of labor could be zero or negative, but it calories than the idle rural inhabitant.4 But
lThe Theory of Economic Growth, Homewood, for Lewis the "disguised unemployed" of ag-
1955, pp. 326-7. In "Economic Development with
riculture may be working as hard as anyone
Unlimited Supplies of Labour," The Manchester else, and may in fact need more calories for
School, May 1954, pp. 141-2, the presence of "dis-
guised unemployment" is claimed for hired agricul- ^Problems of Capital Formation in Underdevel-
tural labor also, although in lesser degree than for oped Countries, Oxford, 1953, pp. 36 ff.
self-employed labor. 4"Some Comments on Prof. Nurkse's Capital Ac-
2The Theory of Economic Growth, op. cit., pp. cumulation inUnderdeveloped Countries," L'Egypte
327-8. contemporaire, XLIV, April 1953, p. 9.
 158 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

farm-work than they would need for factory- duction inthe total food output of the farm
work. They are "idle" in the sense only that if any of its workers were removed.
their work is unproductive. On the other Let me now, as my last illustration, suggest
hand, Nurkse does not mention the inevitable the possibility of a special kind of unemploy-
and possibly appreciable loss of food involved ment which is not "disguised" or "hidden,"
in deterioration, spoilage, and spillage when but is open and voluntary. This is the kind of
the food is consumed in the city instead of on unemployment which would result from a
or near the farm where it is produced. rise in productivity, or in income per time-
Still another kind — or source — of zero or unit of labor, when the supply curve of labor
less marginal productivity of agricultural — was of the kind to which many years ago I
or urban — labor can be conceived of, and gave the label of a "rising-backward supply
may even be important in practice, although curve." When income per time-unit of labor
I have not encountered it in the literature, and aggregate income per laborer both rise,
and it would not be a simple matter, even if the laborer's relative valuation of marginal
it existed, to demonstrate the fact. Suppose units of leisure and of wages per unit of labor
that given the "quality" of the labor force may so change as to make a shorter working-
and the supplies of other productive re- day, week, or year attractive even at the cost
sources, the marginal productivity function of a smaller increase in the size of the pay
of labor could be represented, in the familiar envelope. (For labor paid on a piece-rate
manner, by a slowly-descending curve which basis and for self-employed labor a similar
within the range of observation remains sub- adjustment may occur through reduction in
stantially above the zero-productivity level. the intensity rather than in the duration of
Suppose, however, that the quantity of food the labor.) The English mercantilists of the
available for the farm family depends wholly eighteenth century thought that this was the
on the output of the farm, that the food is usual pattern of behavior of labor, and there-
shared by all members of the family, and that fore believed in the inexpediency of high
when it falls below a certain quantity per wages. There is no reason why such behavior
capita the energy and productive will and ca- should be peculiar to agricultural labor, but
pacity of the worker-members of the family it may be that it is more likely to be prevalent
decline. It then becomes conceivable that if for habit-ridden rural populations, as an ini-
some of the members of the family, including tial response to the availability of choice be-
working-members, were removed from the tween higher income or less — or less inten-
farm (or if the whole family were removed sive— labor.
from the farm, and the farm joined to a sim- Lewis claims that "disguised unemploy-
ilar adjoining farm) and if those removed ment" isnot confined to agriculture, but is in
could no longer draw on the food-resources of underdeveloped countries common also in
the farm, the labor remaining on the farm cities in the form of over-staffed retinues of
would acquire a sufficiently higher marginal domestic servants and of over-crowded ser-
productivity curve, so that the farm would vice occupations where self-employment is
produce more than it did before when the the rule.5 To make plausible the argument
number of workers was greater. In such a that maintenance of a large retinue of do-
case, much of what has been said about "dis- mestic servants is a symptom of "disguised
guised unemployment" and about appropri- unemployment" any more in the city of an
ate remedies for it would be relevant. Not so, underdeveloped country than it would be in
however, Nurkse's proposal of removal of London or Paris, one must assume, as Lewis
some of the workers off the farms without does, that provision of employment for per-
termination of their dependence, direct or in- sons who otherwise would be openly unem-
direct, on the farm for their food. Unless the ployed isa major motive of the employers in
per capita food consumption on the farm was
increased, there would in this case be a re- 5The Manchester School, May 1954, pp. 141-2.
 159 THE LABOR SURPLUS ECONOMY

the underdeveloped countries. Nurkse has ment or considerable seasonal unemploy-

claimed that the attractiveness of the con- ment, in the absence of convincing evidence
sumers' goods of advanced countries to the that the employment is not reasonably "pro-
population of underdeveloped countries op- ductive" when everything relevant to "real
erates as a serious barrier both to the devel- income" and to available alternatives is taken
opment oftheir own industries and to capital into account or that we know how to grow
formation.6 Since industrialization in under- spring wheat or cabbages in an American
developed countries, when not directed oth- winter. If we must do without the spring
erwise by government, tends to concentrate wheat or the cabbages if we are to escape the
on consumers' goods of advanced-country seasonal unemployment, perhaps it is sensi-
types, a shift of taste away from Lewis's type ble to reconcile ourselves to its persistence. I
of domestic service to tangible consumers' find it unhappy semantic usage also to label
goods should promote instead of retarding in- as "underemployment" and as "disguised un-
dustrialization, although I do not venture to employment" labor which would be "unnec-
guess whether it would promote or work essary" to maintain product undiminished if
against "economic welfare." "intensity
since this would of workleadperto hour" were raised,7
the conclusion that
I refrain from discussing here the appro-
priatenes ofthe application of the term "un- there was "underemployment" and "dis-
deremployment" to agriculture, even to ous guised unemployment"
American in the mostevenprosper-
urban industries when
American agriculture, merely to signify
either allegedly low-productivity employ- overtime work was common.

7Chiang Hsieh, "Underemployment in Asia," In-

^Problems of Capital Formation, op. cit., Chap. 3. ternational Labour Review, LXVI, June 1952, p. 709.

III.A.3. Choice of Techniques*

Unemployment is often not "visible." It may factors. If OL2 is the amount of labour that
be "disguised" as a result of a particular task is applied and the marginal productivity be-
being performed by more labour than is nec- comes zero at point Lu then LXL2 is the rel-
essary (given the technique and the produc- evant range. But what is the point of applying
tive resources). As Professor Nurkse puts it labour beyond L{]
"the marginal productivity of labour, over a This confusion arises because of not distin-
wide range, is zero." Thus labour can be guishing between labour and labourer. It is
taken away from these occupations without
affecting production.
The concept of "disguised unemployment" Total output
is actually less simple than it looks, and we
must say a few words on it to avoid any pos-
sible misunderstanding. We may ask if "mar-
ginal productivity of labour, over a wide
range, is zero," why is labour being applied
at all? Does it not go against rational behav-
iour? In Figure 1, curve Y gives the relation-
ship between labour and output, given the Labour
technique and the supply of other productive
*From A. K. Sen, Choice of Techniques, Basil
Blackwell, 1960, pp. 13-6. Reprinted by permission.
 160 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

not that too much labour is being spent in the marginal product of labour becomes nil with
production process, but that too many la- OLx labour-hours and labour is not applied
bourers are spending it. Disguised unemploy- beyond this point. The working population
ment thus normally takes the form of smaller being OP2, each puts in tan a hours of work.
number of working hours per head per year; Tan b represents the "normal" working hours
for example, each of three brothers shepherd- per labourer. So the job can be done by OPx
ing the sheep every third day. It is thus the labourers keeping normal hours. In this sense
marginal productivity of the labourer, so to P\?2 population is surplus. Thus while mar-
say, that is nil over a wide range and the pro- ginal productivity of labour is nil at point L\
ductivity oflabour may be just equal to zero only, that of the labourer is nil over the range
at the margin. It may also take the form of
P\Pi. This represents the volume of "dis-
lower intensity of work with people "taking it guised unemployment."
easy," e.g., the peasant having time to watch A further difficulty with the concept of dis-
the birds while working. If a number of la- guised unemployment is that any shift of la-
bourers went away, the others would be able bour from the rural area will lead to some re-
to produce about the same output working organization ofthe techniques of production
longer and harder. There is no contradiction in that area. The organization of production
between disguised unemployment and ra- may change considerably and thus the con-
tional behaviour. In a family-based peasant cept of zero marginal productivity, from the
economy unemployment will naturally put on point of view of strict theory, becomes diffi-
this disguise. A piece of land that can be cul- cult to apply. The contrast between a move-
tivated fully by two, may actually be looked ment along the productivity curve and a shift
after by four, if a family of four working men of the curve as a result of a change of the sup-
having no other employment opportunity
ply of other factors of production (e.g., "or-
happens to own it. All this is represented in ganization") isrelevant in this connection.
Figure 2. The south represents the number of But actually from the point of view of oper-
labourers, the east the number of labour- ational policy these conceptual complications
hours spent and the north the product. The need not be very important. The point is that
we are in a position to remove a considerable
part of the rural labour force away from the
Total output rural area without affecting the rural output
appreciably.
A more real difficulty will arise if it is
found that at the given supply of capital and
land, marginal productivity of labour is not in
fact nil. In this case, when some of the rural
labourers move out, some real investment
(and not merely more "organization") is nec-
Labour
essary tokeep the output constant. The "op-
portunity cost" of labour can be measured by
the amount of investment that has to be put
in to keep rural output constant as a labourer
moves out. This is a measure of the cost in
stock terms. In terms of flow, the opportunity
cost can be measured (a) as the amount of
alternative rural output sacrificed per year as
a result of drawing a labourer away from the
rural area, or (b) as the amount of alternative
output sacrificed per year as a result of put-
ting in the compensatory investment to make
FIGURE 2. good the above loss of rural output rather
 161 THE LABOR SURPLUS ECONOMY

than using that amount of investment in surprising. It is, however, likely that in this
some other field. case a relatively small amount of investment
It is in fact possible that even in some of may make good the loss of rural output and
the so-called "overpopulated" areas scarcity the opportunity cost in terms of the necessary
of capital has led to such a substitution of increase in the capital stock or in terms of the
capital by labour that the marginal product resulting loss of flow of alternative output,
of labour is not zero; thus a withdrawal of la- i.e., in sense (b), may be rather small. Thus,
bour will reduce the rural output somewhat. while labour in some of these economies may
Rural techniques of production in some of not be "free," it is likely to be cheap-
the underdeveloped countries are so primitive cheaper than the conventionally-measured
and labour-intensive that this will not be very opportunity cost, sense (a), suggests.
 III.B. DUAL SECTOR MODELS
III.B.1. Intersectoral
Relationships in a Dual
Economy— Note

When a dual economy exists, the ultimate traditional sector or the "self-employment
question for the country's future develop- sector."2 In this sector, output per head is
ment ishow the modern exchange sector is to much lower than that in the capitalist sector;
expand while the indigenous sector contracts. given the available techniques, the marginal
This requires an analysis of the interrelation- productivity of a laborer in agricultural pro-
ships between the two sectors. Sir W. Arthur duction may be zero as a limiting case. As a
Lewis has offered a perceptive analysis of this result of institutional arrangements, such as
problem.1 This note summarizes Lewis's the family farm or communal holdings of
model and assesses its relevance for contem- land, members of the farm labor force con-
porary problems of development. sume essentially the average product of the
Lewis analyzes the process of economic ex- farm's output even though the marginal
pansion in a dual economy composed of a product of some farm laborers may be well
"capitalist" sector and a "noncapitalist" sec- below the average product.
tor. The capitalist sector is defined as that A fundamental relationship between the
part of the economy which uses reproducible two sectors is that when the capitalist sector
capital, pays capitalists for the use thereof, expands, it draws labor from the reservoir in
and employs wage-labor for profit-making the noncapitalist sector. For countries that
purposes. Capitalist production need not be have experienced high rates of population
restricted to manufacturing; it may also be in growth and are densely populated, it is as-
plantations or mines that hire labor and resell sumed that the supply of unskilled labor to
its output for a profit. The capitalist sector the capitalist sector is unlimited. Labor is
may also be either private or public: again, "unlimited" in the sense that when the capi-
the distinguishing feature of the capitalist talist sector offers additional employment op-
sector is the hiring of labor and sale of its out- portunities at the existing wage rate, the
put for a profit, which can be undertaken by numbers willing to work at the existing wage
public enterprise as well as private. The sub- rate will be greater than the demand: the
sistence sector is that part of the economy
which does not use reproducible capital and 2The characterization of this traditional agricul-
does not hire labor for profit — the indigenous tural sector as the "self-employed" sector is suggested
by Kazushi Ohkawa, "Balanced Growth and the
'W. Arthur Lewis, "Economic Development with Problem of Agriculture — with Special Reference to
Unlimited Supplies of Labor," The Manchester Asian Peasant Economy," Hitotsubashi Journal of
School, May 1954, pp. 139-91; "Unlimited Labour: Economics, September 1961, pp. 13-25.
Further Notes," ibid., January 1958, pp. 1-32. "Re- Professor Reynolds has also proposed that a four-
flections on Unlimited Labor," in Luis Eugenio Di sector model would be more relevant — with the "tra-
Marco (ed.), International Economics and Develop- ditional sector" divided into the rural sector and the
ment, Essays in Honor of Raul Prebisch, New York, urban trade-service sector, and with both an industry
1972, pp. 75-96. The analysis has been extended in subsector and a government subsector in the "modern
some respects by Professors Gustav Ranis and J. C. sector." The urban trade-service sector employs peo-
H. Fei, "A Theory of Economic Development," ple with little skill and little initial capital, and there
American Economic Review, September 1961, pp. is relative freedom of entry. For a discussion of the
533-65; "Innovation, Capital Accumulation, and different production functions in these four sectors,
Economic Development," ibid., June 1963, pp. 283- see Lloyd G. Reynolds, "Economic Development with
313; Development of the Labor Surplus Economy:
Surplus Labor: Some Complications," Oxford Eco-
Theory and Policy, Homewood, 1964; "Agrarianism, nomic Papers, March 1969, pp. some
89-103. Reynolds'sto
Dualism, and Economic Development," in I. Adel- urban trade-service sector bears resemblance
man and E. Thorbecke (eds.), The Theory and De- the "informal sector" discussed in selection III.C.2,
sign ofEconomic Development, Baltimore, 1966. below.
162
 163 DUAL SECTOR MODELS

supply curve of labor is infinitely elastic at

the ruling wage. According to Lewis, one
condition for this is that the ruling wage of
the capitalist sector exceeds the earnings in
the noncapitalist sector of those who are will-
ing to transfer themselves. The other condi-
tion is that any tendency which the transfer
may set in motion for earnings per head to
rise in the noncapitalist sector must initially
be offset by the effect of increases in the labor
force (natural increase, immigration, or
greater female participation).3 A large com-
ponent ofthe unlimited supply of labor from
the noncapitalist reservoir of labor is com-
posed of those who are in disguised unem- Quantity of Labor
ployment in agriculture and in other over-
manned occupations such as domestic FIGURE 1.
service, casual odd jobs, or petty retail trad-
ing. Another source of labor is women who
transfer from the household to commercial labor. Given a fixed amount of capital at the
employment, and the labor force has also outset, the demand for labor is initially rep-
grown as a result of the population increase. resented bythe marginal productivity sched-
ule of labor, N\DX in Figure 1. If we assume
The large pool of unskilled labor enables new
industries to be created or old industries to profit maximization, capital will then be ap-
plied up to the point where the current wage
expand in the capitalist sector without en- equals the marginal productivity of labor. If
countering any shortage of unskilled labor.
The wage which the growing capitalist sec- OW'is the current wage, the amount of labor
employed in the capitalistic sector is OL; be-
tor has to pay is determined in Lewis's model yond L, workers earn whatever they can in
by what labor earns in the subsistence sector.
the subsistence sector. The total product
Peasant farmers will not leave the family
farm for wage employment unless the real NxPLO in the capitalist sector will then be di-
vided between wages in the amount OWPL
wage is at least equal to the average product
and the WPNX.
amount capitalists' surplus or profits in the
on the land.4 Capitalist wages, as a rule, will
have to be somewhat higher than subsistence
earnings in order to compensate labor for the In tracing the process of economic expan-
sion, Lewis emphasizes that the key to the
cost of transferring and to induce labor to
leave the traditional life of the subsistence process is the use of the capitalist surplus.
The driving force in the system is generated
sector. (Lewis observes that there is usually a
by the reinvestment of the capitalist surplus
gap of 30 percent or more between capitalist
wages and subsistence earnings). At the ex- in creating new capital. As the capitalist sec-
isting capitalist wages, however, the supply of tor expands, labor withdraws from the sub-
sistence sector into wage employment, the
labor is considered to be perfectly elastic.
This situation is illustrated in Figure 1 surplus then becomes even larger, there is
still more reinvestment of profits, and the
where OA represents subsistence earnings,
process continues on progressively absorbing
OW the real wage rate in the capitalistic sec- surplus labor from the subsistence sector.
tor, and WS the perfectly elastic supply of
Figure 1 illustrates this process by the out-
ward shift of the demand for labor, from
3Lewis, "Reflections on Unlimited Labor," p. 77. NXDX to N3D3 over time. When some of the
4Even though the marginal product of labor is zero initial surplus WPNX is reinvested, the
in disguised unemployment, a member of the ex-
tended family shares in the total product and receives amount of fixed capital increases, and
approximately the average product. the schedule of the marginal productivity of
 164 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

labor is then raised to the level of say, N2D2. the non-market institutional average prod-
Both the capitalist surplus and capitalist em- uct.5 When the marginal product is equal in
ployment are now larger. Further investment the capitalist and noncapitalist sectors, the
then raises the marginal productivity of labor analysis then becomes the same as in the
to, say, NiD3. And so the process continues. usual neoclassical one-sector economy.
The growth in capitalist profits is crucial in In Lewis's model, the expansion process
this process, and the share of profits in the might be cut short, however, by a rise in real
national income is of strategic importance. wages and a reduction in profits that halts
This will be determined by the share of the capital accumulation before the excess labor
capitalist sector in the national output and by supply is completely absorbed. This may be
the share of profits in the capitalist sector. As due to a rise in average product in the subsis-
the capitalist sector expands, and the wage- tence sector because the absolute number of
price ratio remains constant, the share of prof- people in this sector is being reduced without
its in national income increases. And since a fall in total output, or labor productivity
the major source of savings is profits, savings happens to increase in the subsistence sector,
and capital formation also increase as a pro- or the terms of trade turn against the capi-
portion ofthe national income. talist sector.
Barring a hitch in the process, the capital- If, for instance, the capitalist sector pro-
ist sector can expand until the absorption of duces no food, and the demand for food rises
surplus labor is complete, and the supply as the capitalist sector expands, then the
function of labor becomes less than perfectly price of food will rise in terms of capitalist
elastic. Capital accumulation has then products — that is, the terms of trade turn
caught up with the excess supply of labor; be- against the capitalist sector. In order to keep
yond this point real wages no longer remain the real income of workers constant, capital-
constant but instead rise as capital formation ists then have to pay out to labor a larger part
occurs, so that the share of profits in the na- of their product as wages, thereby reducing
tional income will not necessarily continue to their profits.
increase, and investment will no longer nec- The possibility that industrialization can
essarily grow relative to the national income. be inhibited by a deterioration in the terms of
In their two-sector model, Professors Ranis trade for the industrial sector points up the
and Fei consider disguised unemployment to extreme importance of providing an agricul-
exist in the agricultural sector when the mar- tural surplus for consumption in the expand-
ginal physical product of labor is less than its ing industrial sector. This is one of several
average product which is the institutional reasons why agricultural output must expand
wage under the extended family system in along with industrial development. This
agriculture. When labor has a marginal problem, together with other relationships
product of zero, it is termed "redundant between industry and agriculture, will be dis-
labor." In the Ranis-Fei model the horizontal cussed more fully in Chapters VI and VII
supply curve to the capitalist sector is then below.
considered to end when the redundant labor It should, of course, be recognized that if
force in the agricultural sector is taken up the country earns sufficient foreign exchange,
and a relative shortage of agricultural goods the capitalist sector could overcome the ag-
appears, so that the terms of trade turn ricultural constraint on its further expansion
against the capitalist sector that is trading by importing the necessary food and raw ma-
with the agricultural sector. This upward terials from overseas instead of being limited
trend in the labor supply curve is later accen- by domestic agricultural output. But if ex-
tuated bya rise in the agricultural real wage port earnings are insufficient, then the failure
traceable to the removal of disguised unem- of exports to keep pace with needed imports
ployment and the commercialization of agri- will constrain the rate of growth of output.
culture so that real wages become deter- 5Ranis and Fei, "A Theory of Economic Develop-
mined by competitive market forces, not by ment," pp. 539-40.
 165 DUAL SECTOR MODELS

Although the Lewis model highlights some Nobody denies that in the overpopulated countries
basic relationships in dualistic development, handicraft workers, petty traders, dock workers,
domestic servants, and casual workers have a lot
its applicability has been questioned on sev-
eral counts. Some critics believe that the of spare time on their hands, and that most of
them (except the domestic servants) would be glad
model rests on the existence of disguised un- to exchange extra work for extra income at the
employment inthe noncapitalist sector, and current rate. Neither does anybody deny that there
they contend this is unrealistic. The strict in- is much seasonal unemployment in agriculture.
terpretation ofdisguised unemployment is The dispute is confined to the situation on small
that the marginal productivity of labor, over family farms at the peak of the agricultural sea-
a wide range, is zero — that is, labor is redun- son, in some parts of Asia and the Middle East.
dant or in surplus and can be withdrawn I do not believe that the productivity of a man-
without any loss of output even if no change hour is zero in agriculture, domestic service, petty
in production techniques or use of other pro- retailing, handicrafts, or any other part of the non-
ductive resources occurs. But the existence of capitalist reservoir. Nevertheless, I have seen noth-
disguised unemployment is not necessary for ing in the now vast literature of underemployment
to alter my belief that in India or Egypt one could
the expansionary process that Lewis de- mobilize a group equal to, say, ten per cent of the
scribes; all that the model needs is the fact
unskilled noncapitalist labor force without signifi-
that supply exceeds demand at the current
cantly reducing the output of the noncapitalist sec-
wage. It is therefore not necessary to say any- tors from which they were withdrawn.8
thing about the productivity of marginal
units of labor in the reservoir, beyond noting Another type of underemployment that
that it must be less than the wage offered by characterizes some LDCs may, however, cre-
capitalists. ate more difficulties for the Lewis analysis.
Moreover, although Lewis refers to the A type of "traditional" underemployment
arises when sociocultural determinants of the
zero marginal productivity of labor as a lim- division of labor between men and women in
iting case, he means by this the marginal
product of a man, not the marginal product the traditional sector leave the men under-
employed. Insome African economies, for in-
of a manhour. "For example, in many coun- stance, itis common practice for the men to
tries the market stalls (or the handicraft in-
dustries) are crowded with people who are clear and prepare the land for cultivation
not as fully occupied as they would wish to while the women do the routine work of sow-
be. If ten percent of these people were re- ing and cultivating. The men are left in sur-
moved, the amount traded would be the plus supply in agriculture, but they then fre-
same, since those who remained would do quently become migrant laborers in the
more trade. This is the sense in which the exchange sector. As temporary immigrants
marginal product of men in that industry is from the traditional sector, they might work
in industry or mining on a seasonal basis, or
zero. It is a significant sense, and its signifi-
cance isnot diminished by pointing out that even for a year or two, and then return to
the fact that others have to do more work to their peasant farms.9 The migration of labor
keep the total product constant proves that for short periods might have only a negligible
the marginal product of manhours is effect on agricultural output, but several

positive."6 'Lewis, "Reflections on Unlimited Labor," pp.

This interpretation of zero marginal pro- 81-2.
ductivity oflabor is similar to that offered by
'Informative discussions of migrant labor are pre-
A. K. Sen, who distinguishes between the sented byGuy Hunter, The New Societies of Tropi-
amount of labor and number of laborers.7 cal Africa, Oxford, 1962, pp. 93-101, 191-203; W. J.
It is difficult to estimate empirically the Barber, The Economy of British Central Africa,
amount of surplus labor. Lewis simply states: Stanford, 1961, pp. 71-3; W. Elkan, "Migrant Labor
in Africa: An Economist's Approach," American
6Lewis, "Reflections on Unlimited Labor," p. 79. Economic Review, Papers and Proceedings, May
7A. K. Sen, Choice of Techniques, Oxford, 1968, 1959, pp. 188-97; W. Watson, Tribal Cohesion in a
pp. 3-5. [See selection III.A.3.] Money Economy, Manchester, 1959.
 66 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

studies have shown that adult manpower can- velopment in many countries still may be the
not be spared from the traditional system of absence of a capitalist class with the neces-
agriculture for more than two or three years sary ability and motivation to undertake
without reducing output.10 This special situ- long-term productive investment. We must
ation of temporary labor migration does not confront the problem of how a class of private
conform to a precise interpretation of dis- capitalists is to emerge, or else we must rely
guised unemployment. It is more enlighten- at the outset on the presence of foreign cap-
ing to analyze the labor supply as a case of italists or a class of state capitalists. The
joint supply whereby workers are being sup- analysis of the behavior in the capitalist sec-
plied jointly to the advanced sector and the tor may have to be modified, according to
traditional sector over a period of time. An which type of capitalist class exists.
important part of this problem is to deter- Further, it is assumed that whatever the
mine whether and for how long an individual capitalist sector produces, it can sell; no al-
will offer his labor for wage employment in lowance ismade for a problem of aggregate
demand. But why should this be true if the
the wage sector.11
The case of a migrant labor force, how- output is to be sold within the capitalist sec-
ever, poses special problems that cannot be tor itself, or if the product is an export good?
adquately analyzed in Lewis's model. To The remaining alternative — that the capital-
make it more relevant for this type of situa- ist sector sells to the noncapitalist sector —
tion, the Lewis model has been modified by presents a special difficulty. For then produc-
Professor Barber in his analysis of the inter- tivity must rise in the noncapitalist sector in
action between the indigenous economy and order to ensure an adequate market for the
the expanding money economy in British output of the capitalist sector. But if real
Central Africa.12 wages rise in the noncapitalist sector, the
Another difference is that even if an unlim- supply-price of labor to the capitalist sector
ited supply of unskilled labor is assumed to will then be higher, profits will be reduced,
exist, it is nonetheless generally true that in and the expansionary process may stop be-
poor countries skilled labor is in very short fore all the surplus labor is absorbed.
supply. Lewis recognizes this problem, but Despite these restrictions on its direct ap-
discounts its importance by considering it to plicability, the Lewis model retains high an-
be only a temporary bottleneck which can be alytical value for its insights into the role of
removed by providing the facilities for train- capital accumulation in the development pro-
ing more skilled labor. This will, however, at cess. What is clearly of prime significance is
best involve a time lag, and recent experience the way investment becomes a rising propor-
in developing countries indicates that the tion of national income.
problems of skill formation are not quickly Lewis wanted his model to explain rising
overcome for uneducated and untrained savings and profit ratios, and he states that
manpower. "the chief historical example on which the
A more serious limitation of the Lewis model was based was that of Great Britain
model is that it simply takes for granted the where ... net savings seems to have risen
demand side of the investment process. Can from about 5 per cent before 1780 to 7 per
we assume, as Lewis does, that a capitalist cent in the early 1800s, to 12 per cent around
class already exists? A major obstacle to de- 1870, at which level it stabilized. A similar
rise is shown for the United States [between
10Barber, Economy of British Central Africa, pp. the 1840s and 1890s] Similar changes
72-3, and other references listed in note 9.
can be found since the second world war for
"See E. J. Berg, "Backward-Sloping Supply Func-
tions in Dual Economies — The African Case," Quar- many less-developed countries such as India
terly Journal of Economics, August 1961, pp. 468-
92. Along with the expansion of the capitalist
12Barber, Economy of British Central Africa, pp. or Jamaica."13
180-8. 13Lewis, "Reflections on Unlimited Labor," p. 75.
 167 DUAL SECTOR MODELS

sector and the rise in investment, the model most of the LDCs, the persistence of surplus
also indicates that — short of the model's labor remains as acute in these countries as
turning points — labor will be continually ab- it was when Lewis first presented his model
sorbed from the reservoir of the noncapitalist in 1954. A generation later, as we review the
sector, and that disguised unemployment or Lewis model, we find that the creation of em-
underemployment or surplus labor will con- ployment isstill a major challenge for poor
tinually diminish. Despite two Development countries. We now turn to that problem —
Decades, however, and rising investment ra- and to some of its manifestations that the
tios and expansion of the capitalist sectors in Lewis model did not adequately anticipate.

Comment
Lewis responded to criticisms of his first article, "Economic Development with Unlimited
Supplies of Labor," in subsequent papers: "Unlimited Labor: Further Notes," The Manches-
ter School of Economic and Social Studies (January 1958); "Reflections on Unlimited
Labor," in Luis Eugenio Di Marco (ed.), Essays in Honor of Raul Prebisch (1972); "The
Dual Economy Revisited," The Manchester School of Economic and Social Studies (Septem-
ber 1979).
Other models of a dual economy have been offered by J. C. H. Fei and G. Ranis: "A Theory
of Economic Development," American Economic Review (September 1961); Development of
the Labor Surplus Economy: Theory and Policy (1964); "Agrarianism, Dualism and Eco-
nomic Development," in I. Adelman and E. Thorbecke (eds.), Theory and Design of Economic
Development (1966); "A Model of Growth and Employment in the Open Dualistic Economy:
The Cases of Korea and Taiwan," Journal of Development Studies (January 1975). In ad-
dition, models have been presented in these papers: D. W. Jorgenson, "The Development of
a Dual Economy," Economic Journal (June 1961); D. W. Jorgenson, "Surplus Agricultural
Labor and the Development of a Dual Economy," Oxford Economic Papers (November
1968); D. W. Jorgenson, "Testing Alternative Theories of the Development of a Dual Econ-
omy," in I. Adelman and E. Thorbecke (eds.), Theory and Design of Economic Development
(1966); A. C. Kelley, J. G. Williamson, and R. Cheetham, Dualistic Economic Development
(1972); A. K. Dixit, "Models of Dual Economies," in J. A. Mirrlees and N. H. Stern (eds.),
Models of Economic Growth (1973).

III.B.2. Technological Dualism-

Note

One of the most important effects of dualistic

'As an alternative to Boeke's sociological theory of
development is its influence on the pattern of dualism, the theory of technological dualism has been
employment. Several writers have suggested emphasized by Benjamin Higgins, Economic Devel-
that the labor employment problems of a opment, New York, 1968 (revised edition), pp. 17-
poor country are due to the existence of 20, 296-305. The theory of technological dualism in-
corporates the "factor proportions problem," as dis-
"technological dualism" — that is, to the use cussed by R. S. Eckaus, "The Factor Proportions
of different production functions in the ad- Problem in Underdeveloped Areas," American Eco-
vanced sector and the traditional sector.1 In nomic Review, September 1955. Earlier references in-
this interpretation, dualism is associated with clude Joan Robinson, The Rate of Interest and Other
Essays, London, 1952, pp. 110-11; M. Fukuoka,
"structural unemployment" or "technologi- "Full Employment and Constant Coefficients of Pro-
cal unemployment" — a situation in which duction," Quarterly Journal of Economics, February
productive employment opportunities are 1955.
 168 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

limited, not because of lack of effective de-

mand, but because of resource and techno-
logical restraints in the two sectors.
The traditional rural sector is said to have
the following characteristics: it is engaged in
peasant agriculture and handicrafts or very
small industries; the products can be pro-
duced with a wide range of techniques and
alternative combinations of labor and capital
(improved land) — that is. the sector has var-
iable technical coefficients of production: and
the factor endowment is such that labor is the
relatively abundant factor, so that techniques
of production are labor-intensive (in the
sense that relatively large amounts of labor
and relatively small amounts of capital are
used). FIGURE 1.
In contrast, the modern sector is composed
of plantations, mines, oil fields, or large-scale the actual factor endowment is to the right of
industry; there is either in fact, or entrepre- line OE — say, at point F — there must then
neurs believe there is, only a very limited de- be some unemployment of labor in this sec-
gree of technical substitutability of factors. tor. To produce an output of qy the sector will
so that production is characterized by fixed use OKx units of capital and OLx units of
technical coefficients: and the production labor; even though OL: units of labor are
processes in this sector are relatively capital- available, the excess supply of labor will have
intensive. This situation can be represented no effect on production techniques and L L:
by a production function as in Figure 1, units of labor will remain in excess supply,
where the points a, b, c. etc., denote the fixed regardless of the relative factor prices of cap-
combinations of factors — capital (AT) and ital and labor. Only if the capital stock were
labor (L) — that would be used to produce the to increase in the amount indicated by the
outputs qu qll q^ etc., irrespective of what length of the dashed line FF could the re-
the relative factor prices might be.2 The line dundant labor be absorbed in this sector.
OE joining the points a. b, c. etc.. represents Failing a sufficient accumulation of capital,
the expansion path of this sector, and its the excess labor supply will simply remain
slope is equal to a constant, relatively capital- unemployed, or must seek employment in the
intensive factor ratio. traditional sector.
Only when capital and labor are actually It is interesting to note that Marx had a
available in proportions equal to the fixed similar view of the problem of unemploy-
capital-labor ratio is it possible that both fac- ment. According to Marx, the amount of em-
tors can be fully utilized simultaneously. If ployment offered by capitalists depends on
the amount of capital in existence, and there
^Jnits of capital (K) are measured on the vertical
axis, and units of labor (L) on the horizontal axis. The is unemployment because there is insufficient
curve qx is an isoquant representing a certain level of capital to employ all the potential available
output; as drawn, the output qx can be produced only labor. If A represents the total labor availa-
with" the unique combination of factors at point a ble, and A" the amount of employment re-
(OKx of capital and OLx of labor). The curves qz. q?.
etc.. represent different levels of output, with output
quired to work the existing stock of capital at
increasing along the expansion line OE. Output can its normal capacity, then {A — X) is Marx's
be increased, however, only by increasing the use of
K and L in the constant proportions given bv the slope
"reserve army of unemployed labor."3
of OE.
3Cf. Robinson. Rate of Interest, pp. 110-11. n. 2.
 169 DUAL SECTOR MODELS

Having in mind the different production longer run, technological progress did not
functions in the two sectors, we may now ease this situation. For in the modern sector
summarize the argument that technological technological progress favored more capital-
dualism has intensified the problem of em- intensive techniques, so that it was all the
ployment indual economies. In many coun- more difficult to increase employment oppor-
tries, the advanced sector was initially devel- tunities inthis sector as investment and out-
oped by an inflow of foreign capital. As put expanded. At the same time, there was no
foreign enterprises operated under efficient incentive in the rural sector to introduce
management with modern production tech- labor-saving innovations (even if it were as-
niques, output in this sector expanded. At the sumed that the technical possibilities were
same time, however, population was grow- known and the necessary capital was
available).
ing— in some cases at a rate considerably in
excess of the rate at which capital was accu- It has also been suggested that the locus of
mulating inthe advanced sector. And since technological progress is such that a capital-
production processes in this sector were cap- intensive invention affects the choice of tech-
ital-intensive, and fixed technical coefficients nique in only those cases in which the cost of
were used, this sector did not have the capac- labor to capital is high, but not when it is low
ity to create employment opportunities at a as in the traditional sector. As Professor Lei-
rate sufficient to absorb the greater labor benstein argues,5 this is because the gradual
force. While investment and output ex- type of technological progress, which,
panded in the advanced sector, capital accu- through redesign and general improvement,
mulation was nonetheless slow relative to increases the effectiveness of given types of
population growth, and labor became a re- machines and tools, is more likely to cause a
dundant factor in this sector. Entry into the shift of points on the production function in
traditional rural sector was then the only al- this region of high rather than low capital-
ternative open to surplus labor. labor ratios, such as in the shift from q to q'
As the labor supply increased in the tra- in Figure 2. In the traditional sector, where
ditional sector, it may have been possible ini- the capital-labor ratio is low, there is less
tially to bring more land under cultivation, likelihood of recognizing opportunities for
but eventually land became relatively scarce.
Labor increasingly became the relatively
abundant factor, and since technical coeffi-
cients were variable in this sector, the pro-
duction process became ever more labor-in-
tensive in the traditional sector. Finally, all
available land became cultivated by highly
labor-intensive techniques, and the marginal
productivity of labor fell to zero or even
below: "disguised unemployment" began to
appear.4 Thus, with continuing population
growth, the limited availability of capital
caused a surplus of labor to arise in the tra-
ditional rural sector. Given the labor surplus,
there was no incentive in the traditional sec-
tor to move along the production function to-
FIGURE 2.
ward higher capital-labor ratios and thereby
achieve an increase in output per man.
Further, it is contended that, over the 5Harvey Leibenstein, "Technical Progress, the Pro-
duction Function and Dualism," Banca Nazionale del
Lavoro Quarterly Review, December 1960, pp. 13-
15.
4Higgins, Economic Development, p. 330.
 170 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

gradual inventions and improvements, the ism indicates why factor endowment and the
scale of operations may not be sufficient to differences in production functions have re-
support any new equipment, and there may sulted historically in the rise of underemploy-
be a lack of the complementary inputs ment of labor in the traditional sector, its em-
needed to adopt some type of new capital pirical relevancy can certainly be questioned.
good. Accordingly, it is maintained that there Has production in the advanced sector ac-
is little tendency for the isoquants to shift at tually been carried on with fixed coefficients?
points with low capital-labor ratios. If the ex- Even if an advanced, capital-intensive pro-
penditures line EE' in Figure 2 represents the cess was initially imported, was there subse-
existing ratio of labor cost to capital cost (in quently no adaptation to the abundant labor
this case: relatively low wage rates, as re- supply? Was technical progress actually
flected inthe slope of EE'), then the shift in labor-saving in the advanced sector? These
the capital-intensive portion of the isoquant questions call for empirical studies beyond
has no effect on the choice of technique in the the highly impressionistic statements
traditional labor-intensive sector. contained in the foregoing summary of the
Although the theory of technological dual- theory of technological dualism.
Comment
During the attention.
considerable colonial and
J. H.early postcolonial
Boeke and other periods, the concept inoftheir
Dutch economists "social dualism"
studies received
of Indonesian
development emphasized "the clashing of an imported social system with an indigenous social
system of another style." See J. H. Boeke, Economics and Economic Policy of Dual Societies
(1953). Boeke's analysis, with its conclusion that there is need for a distinctive economic and
social theory for underdeveloped countries, is criticized by Benjamin Higgins, "The 'Dualistic
Theory' of Underdeveloped Areas," Economic Development and Cultural Change (January
1956).

III.B.3. Dualism and Wage

Differentials*

We have noted that the level of wages is gen- tors, we may consider them under three
erally higher in the modern sector than in the heads.
traditional sector of the underdeveloped First, there are the wage differences which
country. How far does this imply the exis- reflect genuine economic differences in skills,
tence of a dualism in the labour market, dis- in costs of living, etc. which clearly do not
couraging the expansion of the modern sector distort the allocation of labour between the
and distorting the allocation of resources in two sectors.
the opposite direction from that caused by fi- Next, there is the less well-recognized fact
nancial dualism? Without entering into a de- that in the underdeveloped countries the
tailed enumeration of the possible causes of wage rate in the modern sector reflects the
the wage differentials between the two sec- payment to the head of the family to induce
him to move with his dependents on a per-
manent basis to the place of his work,
*From H. Myint, Economic Theory and The
Underdeveloped Countries, Oxford University Press, whereas the wage rate in the traditional sec-
New York, 1971, pp. 331-40. Reprinted by tor reflects the payment to a single worker on
permission. a casual or temporary basis. Now even in the
 17 1 DUAL SECTOR MODELS

absence of government regulations and trade bour. How are we to interpret this type of
union pressure, the larger-scale concerns in wage differential between the modern and
the modern sector may prefer to pay higher the traditional sector? In so far as the large-
wages to obtain a regular labour force, for at scale modern concerns are willing to pay a
least two reasons. First, the gains in produc- higher wage rate, voluntarily and without
tivity from a stable labour force and a low any external pressure, there can be no distor-
rate of turnover may more than pay for the tion of resource allocation originating from
higher wage bill. Second, if the concerns are the labour market. Yet the wage differential
run by foreign entrepreneurs or managers, is associated with a distinct dualism in indus-
they do not have the necessary local knowl- trial organization and may be a sign of man-
edge and skills in labour relations to cope agerial rigidity on the part of the modern sec-
with the casual type of labour. Indeed, from tor failing to make a more effective use of the
their point of view the cost of re-adapting abundant supply of casual labour. In so far
their whole system of production and orga- as this creates a distortion, it is not due to a
nization tomake use of the cheaper casual la- high wage rate discouraging the expansion of
bour would be much too high and they would the larger-scale economic units in the modern
be prepared to pay considerably higher wages sector, but due to an insufficient development
to obtain a labour force approximately simi- of the small-scale indigenous economic units
lar to the type of regular labour force they which are more likely to be able to take ad-
are used to in their own countries. Histori- vantage of the abundant supply of casual
cally, this can be illustrated by the contrast- labour.
ing labour policies adopted in the develop- Lastly, we have the factors which arbitrar-
ment of the textile industry by foreign ily widen the wage differential between the
entrepreneurs in India and by indigenous en- modern and the traditional sectors and
trepreneurs inJapan. The former recruited clearly distort the allocation of labour be-
their labour force from adult males, paying tween the two. However, these need to be
them a wage rate sufficient to maintain their disentangled from the concept of "disguised
dependents; the latter took advantage of the unemployment." A familiar argument for the
cheaper, but equally efficient, labour of protection of domestic manufacturing indus-
young farm girls available for a few years be- try based on this concept may be summarised
fore they got married.1 The present-day ex- as follows. Because of heavy population pres-
pansion ofthe modern manufacturing sector sure on existing land, the marginal product of
in the underdeveloped countries relies heavily labour in agriculture is reduced to zero. But
on foreign managers and technical experts, the income level in the traditional sector is
not to speak of the branch factories of inter- equal not to the marginal product but to the
national corporations set up to jump the tariff average product of labour on land because of
and import controls. Thus we have a pattern the prevalence of the extended family system
of wage policy and labour organization based sharing the total output among its members.
on a high differential between the ruling In recruiting labour from the traditional sec-
wage rate in the regular labour market and tor the modern sector must pay a wage rate
that in the unorganized market for casual la- equal to the income level in the traditional
sector plus an incentive margin. Thus the
'For a comparative analysis of the labour policies
adopted in the development of the textile industry in
modern sector is being penalised by having to
India and Japan see S. J. Koh, Stages of Industrial pay a wage rate high above the social oppor-
Development in Asia, University of Pennsylvania tunity cost of labour as measured by its mar-
Press, 1966, Chs. II and III; see also W. W. Lock- ginal product in agriculture, and in order to
wood, The Economic Development of Japan, Prince- correct this distortion the modern manufac-
ton, 1954, pp. 213—14; for a theoretical analysis of
turing industry should be given tariff
this point, see D. Mazumdar, "Underemployment in
Agriculture and the Industrial Wage Rate," Econ-
omica, November 1959. protection.
In order to argue that a person in the tra-
 172 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

ditional sector can enjoy an income equal to trade unions are usually able to extract out of
the average product of labour even when his the export sector tend to spread, by a series
marginal product is zero we need, first of all, of sympathetic wage rises, into the rest of the
to assume that his family owns the land and modern sector, imposing a heavy burden both
that he is being supported in a state of "dis- on the domestic manufacturing industry and
guised unemployment" out of what is, prop- on the public sector. While all these three
erly speaking, rent from the land. Once we factors can cause a serious distortion in the
introduce the landlords into the picture, then allocation of resources, none of them consti-
the income left to the family after paying tutes an argument for special protection of
economic rent must be wage income: i.e., the domestic industry. The distortions have to be
marginal product of family labour on land cured by reforms within the labour market.
must be positive. Even if the family owns the In particular the third source of distortion is
land, the marginal product of labour on the similar to the point we have discussed in con-
farm will not be zero if there are alternative
nection with "disguised unemployment" in
opportunities of using some part of the family agriculture. Here also the distortion arises
labour elsewhere. Once we introduce some from the fact that the rent income which the
form of market for agricultural labour, the government should have extracted from the
marginal product of labour on the family foreign companies in the extractive industries
farm will approximately reflect the wage rate and kept for its own use has been unjustifia-
in the neighbourhood. Finally, even in the ab- bly permitted to inflate the wage level in the
sence of landlords and a labour market, the modern sector above the marginal productiv-
marginal product of labour on the family ity of labour.
farm will not be zero unless we are prepared Finally, we may turn to the main argument
to make the highly unrealistic assumption for domestic industrialization which under-
that the marginal disutility of work on the lies much of the current writings on economic
farm is zero. But the notion of the zero mar- dualism in the underdeveloped countries.
ginal product is not needed for the purpose of This argument combines the concept of "dis-
showing the existence of a distortion in the
guised unemployment" with that of techno-
labour market. All that we need for the pur- logical dualism and attempts to support the
pose isto show that there are certain factors case for the expansion of domestic manufac-
arbitrarily or artificially widening the wage turing industry, not in terms of deviations
differential between the modern and the tra-
from the static optimum but in terms of "dy-
ditional sector beyond the extent required to namic" considerations. Thus it is argued
reflect the genuine economic differences in that, given the heavy pressure of population
the two sectors. on land, it is no longer possible for the under-
It is possible to find three such factors. The developed countries to absorb any more la-
first consists of the various government reg- bour in agriculture: the only way of absorb-
ulations on labour and minimum wage rates. ing the surplus labour is through the
These exert a differential effect in that while expansion of the manufacturing sector. But
they can be strictly enforced in the bigger the underdeveloped countries are obliged to
economic units in the modern sectors they are import modern technology in a ready-made
unenforceable for the small economic units in form from the advanced countries: this
the traditional sector. The second arises from
means that the expansion of the manufactur-
the greater ease with which the urban labour ing sector has to be based on methods of pro-
force, in the manufacturing sector can orga- duction which are not only capital-intensive
nize itself into strong trade unions restricting but also require capital and labour in fixed
free entry of labour. The third factor is im- proportions.2 In terms of the static optimum
portant incountries which have a prosperous
2Cf. R. S. Eckaus, "The Factor-Proportions Prob-
foreign-owned industry such as petroleum or lem in Underdeveloped Areas," American Economic
copper. Here the high wage rates which the Review, September 1955.
 173 DUAL SECTOR MODELS

theory, this pattern of economic development isers and improved seeds; and (b) improve-
goes against the grain of the factor endow- ments in agricultural credit and marketing.
ments in the underdeveloped countries, but it Compared with the genuinely intensive agri-
is argued that it can be justified on two culture ofNortheast Asia, the agriculture in
grounds. First, given the factor dispropor- the so-called overpopulated countries of
tionalities and technological rigidities which India and Pakistan may not unfairly be de-
characterise the dualistic economic structure scribed as an inefficient extensive type of
of the underdeveloped countries, there is farming offering great potentialities for the
really very little scope for smooth and flexible introduction of labour-intensive methods.3 In
substitutions and adjustments assumed in a terms of our analysis, the agricultural back-
neo-classical model of the economic system. wardness of these countries may be attrib-
Second, the "dynamic" gains from the ex- uted to two types of dualism: (a) unequal
pansion ofthe modern manufacturing indus- provision of government economic services to
try will tend to outweigh the static losses the modern manufacturing sector and to the
from the distortion in the allocation of agricultural sector with some degree of dual-
resources. ism within the agricultural sector itself, cre-
We shall now show that these arguments ated by unequal access to public economic fa-
do not stand up to a critical scrutiny and that cilities between the larger and the small
they seriously underestimate the scope for farmers; and (b) financial dualism, which is
the introduction of labour-intensive methods a serious obstacle to the adoption of improved
of production both in agriculture and in the methods by small farmers. Thus it is no ac-
manufacturing sector by appropriate domes- cident that
tic economic policies. Even the exponents of
the population pressure argument recognise in Japan, where the rural and small-scale indus-
trial sectors are more integrated in the organized
that the underdeveloped countries in Africa credit market (more than half of all agricultural
and Latin America are not so thickly popu- credit is provided by financial institutions), inter-
lated as some of the Asian countries. Thus est differentals are much smaller: on the average,
they tend to use countries such as India or interest rates are lower in the traditional sectors,
Pakistan as prime examples of agricultural but considerably higher in the fully organized sec-
overpopulation. But a broad survey of the tor, than in India and Pakistan.4
Asian agricultural scene is sufficient to cast Let us now turn to the argument that the
doubts on the assumption that the agricul- scope for substitution of labour for capital is
tural sector in these countries is so saturated
severely limited by the need to adopt modern
with labour that there is no further scope for
technology in the manufacturing sector and
the introduction of labour-intensive methods
that, given this technical rigidity, the only
along any known lines. As a matter of fact,
the highest population densities on land and method of absorbing surplus labour is to ex-
pand the size of the manufacturing sector on
the smallest-size peasant holdings are to be
found in countries such as Japan, Taiwan and a capital-intensive basis. This implicity iden-
tifies the manufacturing sector with the
Korea. On the other hand, these countries are
larger-scale modern style factories. But in
also outstanding illustrations of how agricul-
many underdeveloped countries small-scale
tural output can be rapidly increased by in- industries of various types employ by far the
tensive methods of farming based on small
peasant holdings provided that appropriate largest proportion of the labour and contrib-
economic policies are followed. These include
(a) the adequate provision by the government 3For two very important recent contributions to this
of agricultural inputs, notably the irrigation subject, see Shigeru Ishikawa, Economic Develop-
ment in Asian Perspective, Kinokuniya, Tokyo, 1967,
facilities which enable multiple cropping on Ch. 2 and also charts 2-4; and G. Myrdal, Asian
the same piece of land, reduce seasonal un- Drama, 1968, Vol. i, Ch. 10 and Vol. II, Part V.
employment and encourage the use of fertil- 4G. Myrdal, op cit., vol. III. appendix 8, p. 2095.
 174 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

ute a substantial proportion of the output of sector as a whole. A reduction of economic

dualism in this sense would have the effect of
the manufacturing sector.5 If we define the
manufacturing sector to include both its encouraging a greater degree of economic
modern segment of the larger-scale economic specialization between the larger and smaller
units and its traditional segment of the small- economic units and thus introduce a new pat-
scale economic units, then one thing becomes tern of complementary economic relation-
clear: even if all the large-scale modern fac- ships between the two types of manufactur-
tories operated rigidly on the basis of fixed ing industry.
technical coefficients, there would still be Complementary relationships based on
considerable scope for increasing the propor- vertical linkages between different industries
tion of labour to capital in the manufacturing are a familiar theme in development econom-
sector as a whole by substituting the output ics. But so far the analysis has been handi-
of the small-scale economic units for that of capped byconfining the concept of comple-
the larger-scale units. As we have seen, the mentarity only to the modern segment of the
possibilities in this direction are greater than manufacturing sector. Further, given the
generally allowed, for two reasons: first, the popularity of the input-output models, com-
modern factories set up for the purpose of im- plementarity tends to be interpreted in purely
port substitution in many underdeveloped technical terms: in terms of fixed technical
countries are predominantly in the field of co-efficients linking up insufficiently disag-
light consumers' goods which are also pro- gregated sectors. When, however, we
duced by the traditional sector; and second, broaden the concept of complementarity to
the lower quality of the products of the small- the possible economic linkages between the
scale industries is compensated for by their larger and smaller industries and apply the
cheaper prices. But, as we have shown, this older Adam Smithian notions of division of
overlap in the range of goods produced by the labour and specialization6 to the situation, we
larger and the smaller manufacturing units is begin to have a richer understanding of the
associated with and attributable to the glar- possible complementary relationships within
ingly unequal terms on which capital funds, the manufacturing sector which would have
foreign exchange and economic services pro- the effect of introducing a greater degree of
vided by the government are made available economic flexibility into the domestic eco-
to the two types of manufacturing industry nomic organization of the underdeveloped
by the prevailing policy of domestic indus- countries. With a reduction in dualism, the
trialization. These policies may be said to small economic units may be typically ex-
protect the larger-scale modern factories not pected to take over a larger share of the out-
only from foreign competition but also from
put in the light consumers' goods industries
the domestic competition of the small-scale while the larger economic units shift into the
economic units. Thus it would seem reason- production of inputs for the small industries.
able to expect that a reduction in the unequal As a matter of fact, the small industries in
access to scarce economic inputs between the the underdeveloped countries have shown
larger-scale and small economic units would considerable enterprise in changing over
make the latter more competitive and in- from their traditional materials to new im-
crease their share of the manufacturing out- ported materials in such inputs as dyes,
put, thereby raising the proportion of labour yarns, plastics, etc. and this is the reason why
to capital employed in the manufacturing they are so handicapped by foreign exchange
controls which restrict their access to im-
5For quantitative evidence for the importance of ports. But the division of labour between the
small-scale industries in the underdeveloped coun-
tries, see S. Kuznets, Modern Economic Growth:
Rate, Structure and Spread, 1966, table 8.1 and pp.
6See G. J. Stigler, "The Division of Labour is Lim-
417-20; also E. Staley and R. Morse, Modern Small
Industries for Developing Countries, 1965, Ch. 1. ited by the Extent of the Market," Journal of Polit-
ical Economy, June 1951.
 175 DUAL SECTOR MODELS

larger-scale and the small economic units the economic development of Japan, where
may take various other patterns: the larger the small industries have played a very im-
economic units may sub-contract any part of portant role. Significantly, the growth of the
their productive processes to the small eco- small industries was greatly facilitated by ac-
nomic units; the small economic units may cess to cheap electric power — a reduction of
set up repair shops and other servicing activ- the dualism in the supply of public economic
ities for the bigger economic units; and so on. services. W. W. Lockwood describes the de-
In general, given easier access to capital and velopment ofcomplementary economic rela-
foreign exchange, the small economic units tions between the larger-scale economic units
may be expected to increase their share of and the small economic units as "a skillful
economic activities in the manufacturing sec- utilization of Japan's limited capital re-
tor whenever their lower overhead costs and sources and technical experience to employ a
their access to cheap family and casual la- large and expanding population in productive
bour give them a comparative advantage over
the larger economic units. These various pos- 7W. W. Lockwood, The Economic Development of
sibilities are well illustrated by the history of Japan, 1954, p. 21 1. See also the rest of his Ch. 4.
pursuits."7
 III.C. RURAL-URBAN
MIGRATION
III.C. 1. A Model of Rural-Urban
Migration*

In this section I would like to set forth briefly factor in the decision to migrate. The increas-
a theoretical framework which yields some ing divergence between urban and rural in-
important insights into the causes and mech- comes has arisen both as a result of the rel-
anisms of rural-urban migration in tropical ative stagnation of agricultural earnings
Africa. No attempt will be made to describe (partly as a direct outgrowth of post-war bias
this model in any great detail since that has toward industrialisation at the expense of ag-
been done elsewhere. I believe that the model ricultural expansion) and the concomitant
can usefully serve two purposes: first, to dem- phenomenon of rapidly rising urban wage
onstrate why the continued existence of rates for unskilled workers. For example, in
rural-urban migration in the face of rising Nigeria Arthur Lewis noted that "urban
levels of urban unemployment often repre- wages" are typically at levels twice as high as
sents a rational economic decision from the average farm incomes. Between 1950 and
point of view of the private individual; and 1963 prices received by farmers through
second, to demonstrate how such a theoreti- marketing boards in southern Nigeria fell by
cal framework can be used in an analysis and 25 per cent while at the same time the mini-
evaluation of alternative public policies to al- mum wage scales of the Federal Government
leviate the growing urban unemployment increased by 200 per cent.1
problem. In Kenya average earnings of African em-
ployees in the non-agricultural sector rose
THE INDIVIDUAL DECISION TO from £97 in 1960 to £180 in 1966, a growth
MIGRATE: SOME BEHAVIOURAL rate of nearly 1 1 per cent per annum. During
ASSUMPTIONS the same period the small farm sector of
Kenya experienced a growth of estimated
The basic behavioural assumption of the family income of only 5 per cent per annum,
model is that each potential migrant decides rising from £57 in 1960 to £77 in 1966. Con-
whether or not to move to the city on the sequently, urban wages rose more than twice
basis of an implicit, "expected" income max- as fast as agricultural incomes in Kenya so
imisation objective. There are two principal that in 1966 average wages in the urban sec-
economic factors involved in this decision to tor were approximately two-and-a-half times
migrate. The first relates to the existing as high as average farm family incomes.2
urban-rural real wage differential that pre- Moreover, the urban-rural income differen-
vails for different skill and educational cate- tial in Kenya in 1968 varied considerably by
gories ofworkers. The existence of large dis- level of educational attainment. For example,
parities between wages paid to urban workers whereas farm income was approximately
and those paid to comparably skilled rural la- K£85 in 1968, individuals with zero to four
bourers has long been recognised as a crucial
'W. Arthur Lewis, Reflections on Nigeria's Eco-
nomic Growth, Paris, OECD Development Centre,
*Michael P. Todaro, "Income Expectations, Rural-
Urban Migration and Employment in Africa," Inter- 1967, p. 42.
national Labour Review, Vol. 104, No. 5, November 2Dharam P. Ghai: "Incomes policy in Kenya: need,
1971, pp. 391-5, 411-13, Copyright © International criteria and machinery," in East African Economic
Labour Organisation 1971. Review, June 1968, p. 20.

176
 177 RURAL-URBAN MIGRATION

years of primary education in urban areas farm labourer (or working his own land) for
earned on the average K£102, those with five an annual average real income of, say, 50
to eight years of primary education earned units, or migrating to the city where a worker
K£156, while migrants who had completed with his skill or educational background can
from one to six years of secondary education obtain wage employment yielding an annual
earned on the average K£290 per annum in real income of 100 units. The more com-
1968.3 monly used economic models of migration,
The second crucial element, which for the which place exclusive emphasis on the in-
most part has not been formally included in come diffential factor as the determinant of
other models of rural-urban migration, re- the decision to migrate, would indicate a
lates to the degree of probability that a mi- clear choice in this situation. The worker
grant will be successful in securing an urban should seek the higher-paying urban job. It is
job. Without introducing the probability var- important to recognise, however, that these
iable it would be extremely difficult to ex- migration models were developed largely in
plain the continued and often accelerated the context of advanced industrial economies
rate of migration in the face of sizeable and and, as such, implicitly assume the existence
growing pools of urban unemployed. Argu- of full employment or near-full employment.
ments about the irrationality of rural peas- In a full employment environment the deci-
ants who unwittingly migrate to urban areas sion to migrate can in fact be predicated
permeated by widespread unemployment are solely on securing the highest-paying job
as ill-conceived and culture-bound as earlier wherever it becomes available. Simple eco-
assertions that peasant subsistence farmers nomic theory would then indicate that such
were unresponsive to price incentives. The migration should lead to a reduction in wage
key, in my opinion, to an understanding of differentials through the interaction of the
the seemingly paradoxical phenomenon of forces of supply and demand, both in areas of
continued migration to centres of high un- out-migration and in points of in-migration.
employment lies in viewing the migration Unfortunately, such an analysis is not very
realistic in the context of the institutional and
process from an "expected" or permanent in- economic framework of most of the nations
come approach where expected income re-
lates not only to the actual wage paid to an of tropical Africa. First of all, these countries
urban worker, but also to the probability that are beset by a chronic and serious unemploy-
he will be successful in securing wage em- ment problem with the result that a typical
ployment inany given period of time. It is the migrant cannot expect to secure a high-pay-
combination and interaction of these two ing urban job immediately. In fact, it is much
variables — the urban-rural real income dif- more likely that upon entering the urban la-
ferential and the probability of securing an bour market the migrant will either become
urban job — which I believe determine the totally unemployed or will seek casual and
rate and magnitude of rural-urban migration part-time employment in the urban tradi-
in tropical Africa. tional sector. Consequently, in his decision to
Consider the following illustration. Sup- migrate the individual in effect must balance
pose the average unskilled or semi-skilled the probabilities and risks of being unem-
rural worker has a choice between being a ployed or underemployed for a considerable
3For an analysis of the relationship between edu- period of time against the positive urban-
cation and migration in Africa, see Michael P. To- rural real income differential. The fact that a
daro: "Education and rural-urban migration: theoret- typical migrant can expect to earn twice the
ical constructs and empirical evidence from Kenya," annual real income in an urban area than he
paper prepared for the Conference on Urban Unem-
ployment inAfrica, Institute for Development Stud- can in a rural environment may be of little
ies, University of Sussex, September 1971, especially consequence if his actual probability of se-
pp. 16-30. curing the higher-paying job within, say, a
 178 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

one-year period is one chance in five. In such lationship between migration flows and the
a situation we could say that his actual prob- expected income differential as expressed in
ability of being successful in securing the an "elasticity of migration response" term
higher-paying urban job is 20 per cent, so developed in the Appendix.
that his "expected" urban income for the Since the elasticity of response will itself
one-year period is in fact 20 units and not the be directly related to the probability of find-
100 units that the fully employed urban ing ajob and the size of the urban-rural real
worker receives. Thus, with a one-period time income differential, the model illustrates the
horizon and a probability of success of 20 per paradox of a completely urban solution to the
cent it would be irrational for this migrant to urban unemployment problem. Policies
seek an urban job even though the differen- which operate solely on urban labour demand
tial between urban and rural earnings capac- are not likely to be of much assistance in re-
ity is 100 per cent. On the other hand, if the ducing urban unemployment since, in accor-
probability of success were, say, 60 per cent, dance with our expected income hypothesis,
so that the expected urban income is 60 units, the growth of urban employment ceteris par-
then it would be entirely rational for our mi- ibus also increases the rate of rural-urban mi-
grant with his one-period time horizon to try gration. Ifthe increase in the growth of the
his luck in the urban area even though urban urban labour force caused by migration ex-
unemployment may be extremely high. ceeds the increase in the growth of employ-
If we now approach the situation more re- ment, the level of unemployment in absolute
alistical y byassuming a considerably longer numbers will increase and the unemployment
time horizon, especially in view of the fact rate itself might also increase. This result will
that the vast majority of migrants are be- be accentuated if, for any increase in job cre-
tween the ages of 15 and 23 years, then the ation, the urban real wage is permitted to ex-
decision to migrate should be represented on pand at a greater rate than rural real income.
the basis of a longer-term, more permanent A reduction or at least a slow growth in
income calculation. If the migrant antici- urban wages, therefore, has a dual beneficial
pates a relatively low probability of finding effect in that it tends to reduce the rate of
regular wage employment in the initial pe- rural-urban migration and increase the de-
riod but expects this probability to increase mand for labour.
over time as he is able to broaden his urban A second implication of the above model is
contacts, then it would still be rational for that traditional methods of estimating the
him to migrate even though expected urban "shadow" price of rural labour to the urban
income during the initial period or periods sector will tend to have a downward bias if
might be lower than expected rural income. the migration response parameter is not
As long as the present value of the net stream taken into account. Typically, this shadow
of expected urban income over the migrant's price has been expressed in terms of the mar-
planning horizon exceeds that of the expected ginal product of the rural worker who mi-
rural income, the decision to migrate is grates to the city to secure the additional
justified. urban job. However, if for every additional
The mathematical details of our model of urban job that is created more than one rural
rural-urban migration are set forth in the worker is induced to migrate, then the oppor-
Appendix to this article. For our present pur- tunity cost will reflect the combined loss of
poses, suffice it to say that the model at- agricultural production of all those induced
tempts to demonstrate the conditions under to migrate, not just the one who is fortunate
which the urban-rural "expected" income dif- enough to secure the urban position. It also
ferential can act to exacerbate the urban un- follows that whenever there are sizeable pools
employment situation even though urban em- of the urban unemployed, traditional esti-
ployment might expand as a direct result of mates of the shadow price of urban labour
government policy. It all depends on the re- will reflect an upward bias.
 179 RURAL-URBAN MIGRATION

APPENDIX In any one time period, the probability of

being employed in the modern sector, p{t),
A Mathematical Model of Rural-Urban will be directly related to the probability tt of
Migration having been selected in that or any previous
period from a given stock of unemployed or
Consider the following formulation of the
underemployed job seekers. If we assume
theory of rural-urban migration used in this
article. I begin by assuming that individuals that for most migrants the selection proce-
dure is random, then the probability of hav-
base their decision to migrate on consider-
ations ofincome maximisation and that their ing a job in the modern sector within x pe-
riods after migration, p(x), is:
calculations are founded on what they per-
ceive to be their expected income streams in />(D = TT(l)
urban and rural areas. It is further assumed and
that the individual who chooses to migrate is
attempting to achieve the prevailing average p(2) =tt(1) + [1 -7r(l)]7r(2)
so that
income for his level of education or skill at-
tainment inthe urban centre of his choice. p(x) =/**- 1) + [1 - p(x- l)U(x)
Nevertheless, he is assumed to be aware of or
his limited chances of immediately securing P(x) = tt(i) + x>(oIIn -«*)]
f=2 s=l
wage employment and the likelihood that he
will be unemployed or underemployed for a where
certain period of time. It follows that the mi-
grant's expected income stream is deter- w(t) equals the ratio of new job openings
mined both by the prevailing income in the relative to the number of accumulated job
modern sector and the probability of being aspirants in period t.
employed there, rather than being underem- It follows from this probability formulation
ployed in the traditional sector or totally
that for any given level of Yu(/) and Fr(f),
unemployed.
the longer the migrant has been in the city
If we let V(0) be the discounted present
the higher his probability p of having a job
value of the expected "net" urban-rural in- and the higher, therefore, is his expected in-
come stream over the migrant's time horizon; come in that period.
Yu, r(t) the average real incomes of individ- Formulating the probability variable in
uals employed in the urban and the rural
this way has two advantages: (1) it avoids the
economy; n the number of time periods in the
"all or nothing" problem of having to assume
migrant's planning horizon; and r the dis- that the migrant either earns the average in-
count rate reflecting the migrant's degree of come or earns nothing in the periods imme-
time preference, then the decision to migrate
diately following migration: consequently, it
or not will depend on whether reflects the fact that many underemployed
n migrants will be able to generate some in-
come in the urban traditional sector while
no)= t=0
J [P(t)YM searching for a regular job; and (2) it modi-
fies somewhat the assumption of random se-
- Yr{t)] ertdt - C(O) lection since the probability of a migrant hav-
ing been selected varies directly with the time
is positive or negative, where he has been in the city. This permits adjust-
ments for the fact that longer-term migrants
C( O) represents the cost of migration, and usually have more contacts and better infor-
p(t) is the probability that a migrant will mation systems so that their expected in-
have secured an urban job at the av- comes should be higher than those of newly
erage income level in period t. arrived migrants with similar skills.
 180 SURPLUS LABOR AND DUALISTIC DEVELOPMENT
dy
Suppose we now incorporate this behav- shift, the increase in the urban labour supply
iouristic theory of migration into a simple ag-
gregate dynamic equilibrium model of urban as dSdddy
labour demand and supply in the following da dd dy ga (6)
manner. We once again define the probabil-
ity w of obtaining a job in the urban sector in is—
Differentiating (3) and substituting into
any one time period as being directly related
to the rate of new employment creation and (6), we obtain-
ed dS N
inversely related to the ratio of unemployed
da da N
job seekers to the number of existing job op- TZ (7)
portunities, that is— The absolute number of urbanNdyunemployed
will increase if the increase in labour supply
exceeds the increase in the number of new
S-N (1)
jobs created,dSi.e., ifd{yN)
—
where y is the net rate of yurban N new job cre-
ation, N is the level of urban employment, da "" da
da (8)
and S is the total urban labour force.
If w is the urban real wage rate and r rep- Combining (7) and (8), we get —
resents average rural real income, then the dS N dy Ndy
"expected" urban-rural real income differ- N da (9)
ential dis—
da
d — w -w — r — r (2) Jdws^
dS/S d_ (S-N)
or, substituting (1) into (2) — (10)
dd/d w S
d = w N (3)
or, finally,
or- substituting for d-
The basic assumption 7of 7V our model once dS/S w_ 7T - r (S-N)
again is that the supply of labour to the urban (11)
sector is a function of the urban-rural ex-
pected real income differential, i.e. — Expression (11) reveals that the absolute
S = /Ad) level of unemployment will rise if the elastic-
(4) ity of urban labour supply with respect to the
If the rate of urban job creation is a func- expected urban-rural income differential,
dS/S
tion of the urban wage w and a policy param- dd/d dd/d>
eter a, e.g., a concentrated governmental ef- (what I have called elsewhere the "mi-
fort to increase employment through a gration response function") exceeds the
comprehensive programme of industrial im- urban-rural differential as a proportion of the
port substitution or, as in the case of Kenya, urban wage times the unemployment rate,
the 1964 and 1970 Tripartite Agreements to S — N
— - — . Alternatively, equation (11) shows
raise employment levels, both ofdywhich oper-
ate on labour demand, we have — that the higher the unemployment rate, the
y = fd(wa) (5) higher must be the elasticity to increase the
level of unemployment for any expected real
where it is assumed that —da- > 0. If the income differential. But note that in most de-
veloping nations the inequality (11) will be
growth in the urban labour demand is in- satisfied by a very low elasticity of supply
creased as a result of the governmental policy when realistic figures are used. For example,
 181 RURAL-URBAN MIGRATION

if the urban real wage is 60, average rural Clearly, much more needs to be known
real income is 20, the probability of getting a about the empirical value of this elasticity
job is .50 and the unemployment rate is 20 coefficient in different African nations before
per cent, then the level of unemployment will one can realistically predict what the impact
increase if the elasticity of urban labour sup- of a policy to generate more urban employ-
ply is greater than .033, i.e., substituting into ment will be on the over-all level of urban
(11) we get — unemployment.
dS/S .50 X 60 - 20
dd d
TT71 = 60
77, X -20 = .033

Comment
The key question of what determines rural-urban migration can be explored further in stud-
ies that set forth probabilistic job search models and in empirical investigations of migration
functions. Field studies and econometric analyses indicate the importance of the economic
motive in the decision to migrate. Econometric estimates of migration functions have also
demonstrated that the probability of urban employment, independent of the differences in
actual rural and urban wages, contributes significantly to the explanation of variance among
time periods and subgroups of the rural population in rates of urban migration.
See J. R. Harris and M. Todaro, "Migration, Unemployment and Development: A Two-
Sector Model, American Economic Review (March 1970); J. Bhagwati and T. N. Srinivasan,
"On Reanalyzing the Harris-Todaro Model," American Economic Review (June 1974); A.
G. Blomquist, "Urban Job Creation and Unemployment in LDCs: Todaro vs. Harris and
Todaro," Journal of Development Economics (March 1978); M. B. Levy and W. J. Wadycki,
"Education and the Decision to Migrate: An Econometric Analysis of Migration in Vene-
zuela," Econometrica (March 1974); L. Yap, "The Attraction of the Cities: A Review of the
Migration Literature," Journal of Development Economics (September 1977); Derek Byer-
lee, "Rural-Urban Migration in Africa," International Migration Review (Winter 1974); J.
Stiglitz, "Alternative Theories of Wage Determination and Unemployment in LDCs,"
Cowles Foundation Discussion Paper (1973); H. Joshi and V. Joshi, Surplus Labor and the
City: A Study of Bombay (1976); D. Mazumdar, "The Rural-Urban Wage Gap, Migration
and the Wage Gap," Oxford Economic Papers (1976); J. Connell et al., Migration from
Rural Areas (1976); Richard H. Sabot (ed.), Migration and the Labor Markets in Developing
Countries (1982), with an extensive bibliography covering both analytical and empirical stud-
ies of an econometric and descriptive character. For a review of theory, evidence, methodol-
ogy, and research priorities, see Michael Todaro, Internal Migration in Developing Countries
(1976).

EXHIBIT III.l. Urban Population, Major Areas and Regions, 1950-2000 (thousands)

1950 1960 1970 1975 1980 1990 2000

World total 724,147 1,012,084 1,354,357 1,560,860 1,806,809 2,422,293 3,208,028

More developed regions 448,929 572,730 702,876 767,302 834,401 969,226 1,092,470
Less developed regions 275,218 439,354 651,481 793,558 972,408 1,453,067 2,115,558

Source: United Nations, Patterns of Urban and Rural Population Growth, ST/ESA/Ser. A/68.
 182 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

EXHIBIT III. 2. Net Rural to Urban Migration in Selected Countries, 1950-1970

Number (thousands) As Percentage of Urban Growth

1950-1960 1960-1970 1950-1960 1960-1970

Low-income countries
36 241 56.3 81.4
Nepal
123 80.9
Uganda 287 80.2
76.8
Cambodia 123 252 53.5
Tanzania 153 314 76.9
57.9 64.7
Pakistan 56.1
2,464 3,524 48.6
Indonesia 53.5
2,476 3,486 48.0
Kenya 185 189 62.9
311 403
Burma 40.1 35.6
Sri Lanka 211 33.7
India 22.0 18.2
3,971 4,630
Middle-income countries
Ivory Coast 265 566 84.4 77.2
Ghana 594 959 68.9 66.6
Korea 3.647 92.3 63.8
Zambia 193 3,540 59.8
315 65.9
Uruguay 392 352 66.9
729 58.1
54.2
Malaysia 63.6
Tunisia 1,020
311 59.2
267 53.7
Algeria 727 952 61.5 51.4
46.3
Bolivia 83 99 48.3
Morocco 695 968 46.2
52.9
45.3 45.0
Nigeria
Venezuela 1,677 2,759 55.2 43.4
1,263 1,451
Chile 811 855 46.8 43.2
Columbia 48.6 40.8
1,432 1,840
Peru 590 708 38.0
40.4
Guatemala 218 305 52.8
39.0
Brazil 48.3
6,345 8,360 37.1
Thailand 371 552 35.8
33.4
27.5
Taiwan 673 841 30.1
El Salvador 41 29.5 24.5
Mexico 57 24.3
2,833 31.0
Argentina 742 3,803
843 28.5 20.6
Paraguay 25 15.7 13.6
34
Source: Lyn Squire, World Bank Staff Working Paper No. 336, p. 55.
 183 RURAL-URBAN MIGRATION

III.C.2. The Informal Sector*

The problem with employment is that the Nairobi, with its gleaming skyscrapers, the
statistics are incomplete, covering a major dwellings and commercial structures of the
part of wage-earning employment and some informal sector look indeed like hovels. For
self-employment in the larger and more or- observers surrounded by imported steel, glass
ganised firms but omitting a range of wage and concrete, it requires a leap of the imagi-
earners and self-employed persons, male as nation and considerable openness of mind to
well as female, in what we term "the infor- perceive the informal sector as a sector of
mal sector." thriving economic activity and a source of
The popular view of informal-sector activ- Kenya's future wealth. But throughout the
ities isthat they are primarily those of petty report we shall argue that such an imagina-
traders, street hawkers, shoeshine boys and tive leap and openness of mind is not only
other groups "underemployed" on the streets necessary to solve Kenya's employment prob-
of the big towns. The evidence suggests that lem, but is entirely called for by the evidence
the bulk of employment in the informal about the informal sector. There exists, for
sector, far from being only marginally pro- instance, considerable evidence of technical
ductive, iseconomically efficient and profit- change in the urban informal sector, as well
making, though small in scale and limited by as of regular employment at incomes above
simple technologies, little capital and lack of the average level attainable in smallholder
links with the other ("formal") sector. agriculture. The informal sector, particularly
Within the latter part of the informal sector in Nairobi but to varying degrees in all areas,
are employed a variety of carpenters, masons, has been operating under extremely debili-
tailors and other tradesmen, as well as cooks tating restrictions as a consequence of a pe-
and taxi-drivers, offering virtually the full jorative view of its nature. Thus there exists
range of basic skills needed to provide goods an imminent danger that this view could be-
and services for a large though often poor come a self-fulfilling prophecy.
section of the population. Later we explain how employment in the
Often people fail to realise the extent of informal sector has grown in spite of obsta-
economically efficient production in the in- cles and lack of outside support: the evidence
formal sector because of the low incomes re- suggests that employment has probably in-
ceived bymost workers in the sector. A com- creased a good deal faster in the informal
mon interpretation of the cause of these low than in the formal sector. It is therefore im-
incomes (in comparison to average wage lev- pos ible tojudge how the employment prob-
els in the formal sector) has been to presume lem has changed merely from the data on
that the problem lies within the informal sec- employment in the formal sector.
tor; that it is stagnant, non-dynamic, and a Our analysis lays great stress on the per-
net for the unemployed and for the thinly vasive importance of the link between formal
veiled idleness into which those who cannot and informal activities. We should therefore
find formal wage jobs must fall. It is hardly emphasise that informal activities are not
surprising that this view should be wide- confined to employment on the periphery of
spread, for academic analysts have often en- the main towns, to particular occupations or
couraged and fostered such an interpretation. even to economic activities. Rather, informal
Further, from the vantage point of central activities are the way of doing things, char-
acterised —by
*From ILO Mission, Employment, Incomes, and
Equality: A Strategy for Increasing Productive Em- 1 . ease of entry;
ployment inKenya, Geneva, 1972, pp. 5-8, 503-8.
Copyright 1972, International Labour Organisation, 2. reliance on indigenous resources;
Geneva. 3. family ownership of enterprises;
 184 SURPLUS LABOR AND DUAL1STIC DEVELOPMENT

4. small scale of operation; employment results.1 And as long as the di-

5. labour-intensive and adapted technology; vergent growth rates continue, the prospect
6. skills acquired outside the formal school of growing unemployment in the future
system; and seems inevitable. Economic growth even at
7. unregulated and competitive markets. Kenya's recent rates of 6 to 7 per cent is n )
protection, it is argued, since inappropriate
Informal-sector activities are largely ignored, technologies and rising wages increase pro-
rarely supported, often regulated and some- ductivity so rapidly that growth is robbed of
times actively discouraged by the Gov- its power to increase the number of jobs
ernment. faster than the labour force. Although this
The characteristics of formal-sector activ- explanation has sometimes been used in
ities are the obverse of these, namely — Kenya, it should by now be clear why we
have found it seriously deficient, and in some
1. difficult entry; respects seriously misleading, as the main ex-
2. frequent reliance on overseas resources; planation ofthe Kenyan situation. It ignores
3. corporate ownership; the fact that the bulk of the population works
4. large scale of operation; on the land, not in wage-earning jobs. It de-
5. capital-intensive and often imported pends crucially on the statistics of the growth
technology; of enumerated employment in the formal sec-
6. formally acquired skills, often expatriate; tor, whereas — as we stated earlier — enum-
and erated employment ignores a large and ap-
7. protected markets (through tariffs, quotas parently growing amount of employment in
and trade licenses). the "informal sector." Thirdly it focuses too
exclusively on jobs, instead of on opportuni-
Our strategy of a redistribution from ties for earning a reasonable income.
growth aims at establishing links that are at But apart from these weaknesses, the ex-
present lacking between the formal and the planation just discussed analyses the situa-
informal sectors. A transfer of incomes from
tion exclusively in terms of over-all imbal-
the top income groups to the working poor ances, giving no weight to imbalance between
would result in new types of labour-intensive the structure of skills and aspirations of the
investments in both urban and rural areas. labour force and the structure of incentives
This should not only generate demand for the and incomes from work. It pays no attention
products of the informal sector but also en- to the variations in the incidence of employ-
courage innovations in labour-intensive tech- ment problems seasonally, regionally, by age
niques in this sector. The various policies or by sex.
which we recommend in other parts of the re- There are also marked contrasts between
port are intended to reduce risk and uncer- the relative security and income levels of
tainty on the part of those employed in the those with wage-earning jobs in the bigger
nformal sector and to ensure a dynamic firms and those self-employed in the informal
growth of this large segment of the Kenyan sector. These sharp inequalities inevitably
economy. create strong ambitions to migrate to the
Unemployment is often analysed as simply towns, to strive for higher education, to
the result of the first type of basic imbalance, search for a job. As long as extreme imbal-
between a rapidly growing labour force and ances persist, so will unemployment, since
a more slowly growing number of job oppor- large differentials will always attract a mar-
tunities. Ifpopulation, it is said, grows at 3.3
ine imbalance is even greater if one compares
per cent as in Kenya, and enumerated wage- growth rates of the urban population or of the number
earning employment grows — as in Kenya of school leavers with the growth of enumerated
since 1964 — at 1.9 per cent, increasing un- employment.
 185 RURAL-URBAN MIGRATION

gin of job seekers to hover in the towns, near ble evidence to refute a view that attributes
the chances of the good jobs, in the hopes of the sources of economic and social change al-
snapping one up. This explains why the anal- most exclusively to outside forces.
ysis of inequality is fundamental to the expla- Furthermore, the traditional-modern anal-
nation of employment problems in Kenya. ysis focuses only on the positive effects of the
But unemployment is not only the result of westernisation of the Kenyan economy and
imbalance in differentials and opportunities. ignores the negative effects. In particular, it
Even with perfect equality, unemployment ignores inter-sectoral dynamics, which are
could arise. Fast rates of population growth, the key to the employment problem. The ac-
of urbanisation and school expansion inevi- cumulation ofwealth in a small part of the
tably make it more difficult to absorb the modern sector is the consequence of the con-
growing labour force and reduce the time centration ofpolitical power in that sector,
that might otherwise be available for struc- and has given rise to the development of an
tural adjustments. Here a second set of im- impoverished and economically deprived
balances arise — dynamic imbalances relat- modern sub-sector. The slums of Nairobi,
ing to the structure of economic growth in the Mombasa and to a lesser extent other urban
economy and to the constraints upon it. areas are completely modern and due to the
Rapid growth is needed, but rapid growth differences of wealth and income between dif-
can itself generate imbalances which will ferent sectors of the economy. These differ-
frustrate its continuation — most notably a ences draw migrants towards the concentra-
shortage of foreign exchange, of domestic tions, and bring about the modernisation of
savings, of skills and entrepreneurship, of de- almost the entire economy, but not the spread
mand or of the political support needed to of wealth. Because of the slow growth of
keep the system workable. For this reason high-wage employment, migration to urban
our report is not merely concerned with alle- areas by income seekers has led to the growth
viating unemployment, poverty and gross in- of a low-income periphery. This low-income
equality, but with economic growth on a pat- sector is peripheral both literally and figura-
tern which can be sustained in the future, and tively. InNairobi it sprang up, and continues
which generates wider and more productive to grow, just outside the borders of the
employment opportunities in the process. . . . wealthy urban zone, to supply goods and ser-
vices to the fortunate few inside that zone
and to its own population. Figuratively, it is
THE RELATION BETWEEN THE peripheral in that it has only fortuitous and
restricted access to the sources of wealth.
FORMAL AND INFORMAL
SECTORS
Characteristics and Dynamics of the
The process of economic transformation
and growth in Kenya has been marked by Informal Sector
growing inequalities in the distribution of We describe these two urban sectors as
wealth and income among Africans. The being the "formal" and the "informal" sec-
usual explanation is the traditional-modern tor. This designation is not intended to con-
division of the economy, in which the wester- tribute to an academic proliferation of labels;
nised modern sector is the source of dyna- we merely seek an analytical terminology to
mism and change and the traditional sector describe a duality that avoids the bias against
slowly withers away. This view does not cor- the low-incomes sector inherent in the tradi-
respond to the reality of Kenya; we reject it tional-modern dichotomy. Both sectors are
for that reason, and because it ignores the dy- modern; both are the consequence of the ur-
namism and progressive elements indigenous banisation that has taken place in Kenya over
to the Kenyan economy. We have considera- the last 50 years. We might have used the
 186 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

terms "large-scale" and "small-scale," but a livelihood in this low-income sector are
those terms are purely descriptive and tell us magnified, and the regulations ensure a high
nothing about why one sector is large-scale quality of services and commodities for the
and the other is small-scale. An explanation wealthy few at the expense of the impover-
of this is central to explaining and solving the ished many.
employment problem in Kenya. One impor- The formal-informal analysis applies
tant characteristic of the formal sector is its equally well to the agricultural sector. The
relationship to the Government. Economic parallels are obvious and striking. The di-
activities formally and officially recognised vision between favoured operators with
and fostered by the Government enjoy con- licences and those without in urban areas is
siderable advantages. First, they obtain the reproduced in agriculture between those who
direct benefits of access to credit, foreign ex- grow tea and coffee with official sanction and
change concessions, work permits for foreign those who do so illegally. Similarly, with
technicians, and a formidable list of benefits other agricultural products such as beef,
that reduce the cost of capital in relation to there are those whose wealth enables them to
that of labour. Indirectly, establishments in conform to and benefit from standards offi-
the formal sector benefit immeasurably from cially laid down, while others can make a
the restriction of competition through tariffs, livelihood only by contravening the regula-
quotas, trade licensing and product and con- tions. Inthe agricultural sector extension ser-
struction standards drawn from the rich vices take the place of the industrial estates
countries or based on their criteria. Partly be- and of loans from the Industrial and Com-
cause of its privileged access to resources, the mercial Development Corporation in the
formal sector is characterised by large enter- urban areas: farmers whose wealth and in-
prise, sophisticated technology, high wage come allow them to conform to bureaucratic
rates, high average profits and foreign criteria benefit. Perhaps the most striking
ownership. rural-urban parallel is with illegal rural
The informal sector, on the other hand, is squatters, who move unofficially on to land
often ignored and in some respects helped scheduled for resettlement and face a contin-
and in some harassed by the authorities. En- ual danger of eviction. Their similarity to
terprises and individuals within it operate urban squatters is obvious — both are irresist-
largely outside the system of government ibly drawn to real or perceived sources of
benefits and regulation, and thus have no ac- wealth, despite legal restrictions of access.
cess to the formal credit institutions and the In the Introduction to this report we con-
main sources of transfer of foreign technol- sidered the characteristics, other than rela-
ogy. Many of the economic agents in this sec- tion to the Government, that distinguish the
tor operate illegally, though often pursuing informal sector from the formal. These char-
similar economic activities to those in the for- acteristics ofthe informal sector, both agri-
mal sector — marketing foodstuffs and other cultural and non-agricultural, result in low
consumer goods, carrying out the repair and incomes for those who work in it. A natural
maintenance of machinery and consumer du- consequence of these low incomes is that
rables and running transport, for example. Il- monetary exchanges within the informal sec-
legality here is generally due not to the na- tor are different in quality from those in the
ture of the economic activity but to an official formal sector. A most important consequence
limitation of access to legitimate activity. of a low income is the primacy of risk and
Sometimes the limitations are flouted with uncertainty. The loss a small farmer or a
virtual abandon, as in the case of unlicensed small entrepreneur can bear is disproportion-
matatu taxis; sometimes the regulations are ately smaller than that which can be borne
quite effective. The consequence is always by a wealthy operator, particularly when the
twofold: the risk and uncertainty of earning former has no access to institutionalised
 187 RURAL-URBAN MIGRATION

sources of credit. As a consequence, the en- the existing nexus of restrictions and disin-
trepreneur inthe informal sector must act centives, the seeds of involutionary growth
continually to protect himself against risk. have been sown. Unlike the determinants of
Accordingly he establishes semi-permanent growth of the formal sector, the determinants
relations with suppliers and buyers, fre- of growth of the informal sector are largely
quently at the expense of his profits. For the external to it. The relevant question is not
same reason he may be hesitant to innovate, whether the informal sector is inherently evo-
particularly in agriculture, for he cannot take lutionary orinvolutionary, but what policies
the chance of failure. These characteristic should be followed to cause evolutionary
behavioural responses are not inherent in the growth. Irrespective of policy changes, the in-
informal sector; they are adaptive responses formal sector will grow in the next 1 5 years.
to low income. If policy continues as at present, the growth
As pointed out in our report, a rate of in- will be involutionary and the gap between the
crease of employment in the formal sector formal and informal sectors will widen. The
high enough to reduce the relative size of the employment problem will then be worse.
informal sector seems to us to be beyond the
bounds of possibility for the foreseeable fu-
ture. An absolute reduction is much less A Model of Inter-sectoral Flows
likely still. On the basis of any reasonable The purpose of this section is to identify
calculation, the urban informal sector in the major factors determining employment in
1985 will include a larger proportion of the the informal sector. The model used is a sim-
urban labour force than it does today. We do ple identity rather than a behavioural model.
not view this inevitable development with dis- The points may seem self-evident, but we feel
may, for we see in the informal sector not it is useful to make them, because linkages
only growth and vitality, but also the source between the small-scale, informal sector and
of a new strategy of development for Kenya. other sectors of the economy have generally
The workshops of the informal sector can been ignored.
provide a major and essential input for the From the writings of economists on urban
development of an indigenous capital goods areas in poor countries it appears that the cri-
industry, which is a key element in solving terion authors use for dividing the urban
the employment problem. The informal sec- economy is not the modernity of activities but
tor is not a problem, but a source of Kenya's their enumeration in government labour
future growth. In addition, it is in its work- force surveys.2 The relation between the two
shops that practical skills and entrepreneu- urban sectors implicit in this analysis can be
rial talents are being developed at low cost.
Many of its enterprises are inefficient tech- 2"A more meaningful distinction than that of em-
nically and economically, and will disappear ployed-underemployed-unemployed is between those
employed in establishments employing five or ten
in the process of growth; but this applies workers which are usually covered in annual labour
equally to the formal sector, where tariff and surveys . . . and the rest of the labour force. The larger
quota protection, access to capital goods establishments . . . can generally be characterised as
below world market prices and other restric- modern-sector establishments, as opposed to the
smaller-scale establishments which are better char-
tions on competition perpetuate gross oper- acterised as traditional. Employees in traditional es-
ating inefficiencies that go hand in hand with tablishments are generally underemployed under
sophisticated technology. most definitions of the term, while modern sector em-
Despite the vitality and dynamism we see ployees can be thought of as fully employed..
in the informal sector, we do not delude our- Charles R. Frank, Jr., "The problem of urban un-
selves that it will develop successfully under employment inAfrica," in R. G. Ridker and H. Lu-
bell (eds.), Employment and unemployment prob-
present conditions. Although it has the poten- lems ofthe Near East and South Asia, Delhi, Vikas,
tial for dynamic, evolutionary growth, under 1971, pp. 785-6.
 1 88 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

called the "residual" model. The residual nition given by the following input-output
model is based on the presumption that the row:
informal sector is a reservoir of unemploy- X2 — d2\Xi + (I22X2 + #23^3
ment and marginally productive activity into + a24X4 + X2i (1)
which those who cannot obtain paid jobs in
the formal sector sink, barely making ends The symbol a indicates input-use coefficients,
meet by begging, hawking or embarking on and X2f is the final demand for informal-sec-
petty crime. In short, the activities in this sec- tor output. From our observations and read-
tor are seen as providing no economic service ing in Kenya we can make up a list of the
or commodity. The demand for labour in the most important goods and services supplied
sector is presumed to be static. Popular by the informal sector to itself and the other
though this view is, we do not feel that it is three sectors.
particularly useful. It asserts a priori a char- Goods and services
acteristic ofthe informal sector (unemploy- Using sector
ment) that can be determined only empiri- supplied
cally. The mechanism of the residual model 1. Agriculture Grain-grinding, building
can be summarised briefly. The informal or
materials, transport, mar-
"traditional" urban sector has a static de-
mand for labour, or a demand for labour that keting, repair and main-
tenance.
is quite income-inelastic.3 Therefore, an in- 2. Informal Furniture for commer-
crease inthe number of persons seeking a liv- cial use, tools, transport,
ing in the informal sector (for example, after repair and maintenance.
a reduction of employment in the formal sec- 3. Private formal Marketing and distribu-
tor) merely drives down average earnings in
that sector; and if average earnings are at tion, transport, furni-
subsistence level, the influx presumably re- ture, repair and main-
tenance.
sults in unemployment. A further prediction 4. Government Construction, furniture,
of this type of model when it incorporates a transport.
rural sector is that if average earnings in the 5. Final demand Clothing, prepared food,
urban informal sector are not below those in
furniture, repair and
rural areas, migration will result. Facts men- maintenance.
tioned in the first part of the report cast
doubt on this last prediction for Kenya. In our report we have discussed in detail
This view of the informal sector ignores the ways of increasing the final demand for
linkage effects and product substitution, informal-sector products and strengthening
which we feel are at present quite significant linkages between the informal and other sec-
in Kenya, and should be strengthened in the tors. Here we restrict ourselves to a discus-
future. To identify these linkages we use a sion of the most important parameters of
four-sector model of the economy, which dis- identity (1) above. The strategy we have sug-
tinguishes between smallholder agriculture, gested for Kenya would have the effect in
the informal (non-agricultural) sector, the time of strengthening the linkages of the in-
private formal sector and the government formal sector and fostering a dynamic
sector. The outputs of these sectors are Xu growth of its final demand; this implies a
X2, X3 and XA respectively. The output of the shift in the composition of output from rela-
informal (non-agricultural) sector is by defi- tively capital-using to relatively labour-using
production processes. In short, we foresee the
3This, of course, is also an empirical question, and production of certain types of commodities
the assumption of a low income elasticity of demand
is rather arbitrary. The over-all income elasticity of
and services with more labour for a given
demand for the output of the informal sector, for ex- level of output. This will not occur, of course
ample, issensitive to the distribution of income. (indeed, the reverse will occur), unless the
 189 RURAL-URBAN MIGRATION

present development strategy is radically shift purchases of inputs toward sector 2,

altered. there are significant technological and con-
Identity (1) can be solved for informal-sec- sumer taste influences militating against suc-
tor output as follows: cess. The extent of potential linkages
between sector 2 and sectors 3 and 4 (param-
Xl = ~ (A21*l + *23*3 eters a2i and a24) is strongly related to the na-
1 - «22 ture and pace of technical change occurring
+ ^24^4 + X2F). (2) outside Kenya. This, in turn, is related to the
choice of products, particularly in sector 3.
If we assume the capital stock of the informal
sector to be constant, then a change in X2 Sector 3 products, which require sophisti-
cated technology for quality reasons (and
AX2 AX2
production is — AL2—AL,2
— , where —AL2— is, at the constantly undergo labour-saving technical
change), are unlikely to provide much scope
limit, the marginal product of labour in the for subcontracting or for intermediate inputs
informal sector. This allows us to write (2) in supplied by small producers. Thus the choice
terms of employment: of products must be determined either
AL, through the redistribution of income or
AL? = [A{a2XX{ through direct restrictions. In addition, ac-
1 - a22 AX2 tive measures must be taken to increase the
+ fl23*3 + «24*2 + *2F)]. (3) technical capacity of the informal sector to
While extremely simple, (3) is useful in an- supply inputs.
alysing the growth of employment in the in- The model used here needs to be elabo-
formal sector. rated along behavioural lines, and an attempt
It is obvious that an increase in any X will must be made to estimate the parameters.
increase employment in the informal sector if Data are lacking, though we have made very
none of the input coefficients is zero. For a rough calculations, based on arbitrary but
given increase in total output, the effect on moderate assumptions, that indicate that the
employment in the informal sector is deter- present situation in sectors 1, 2 and 3 and
mined by the values of the coefficients and shifts in government purchases within the
the composition of the increase in output in present structure of public demand imply an
intermediate demand for sector 2 output of
terms of A"!, X2i . . . X2F. The effect of the rec- almost £70 million. This very rough calcula-
om endations inthe main report is —
tion indicates that there is considerable scope
1 . to increase the size of the coefficients; and
for employment in the informal sector in
2. to shift the composition of future increases Kenya. Further, we feel that this model, with
in output toward the sectors where the
coefficients are highest. an attempt to estimate its parameters, rep-
resents amore fruitful approach to an under-
Even if steps are taken to induce the pri- standing ofthe informal sector than the an-
vate formal sector and the Government to alysis underlying the residual model.
Comment
The concept of the informal urban sector was first introduced by Keith Hart, with the
distinction between wage and self-employment as the essential difference between the formal
and informal sectors; see Keith Hart, "Informal Income Opportunities and Urban Unem-
ployment inGhana," Journal of Modern African Studies (March 1973). As noted in the
preceding selection, the ILO identifies the informal sector by a variety of other characteristics.
For analytical purposes, it may be most incisive to define the informal sector as simply that
sector in which the return to labor, whether or not it be in the form of wages, is determined
by forces of demand and supply.
For studies of the informal sector, see Dipak Mazumdar, "The Urban Informal Sector,"
 190 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

World Development, (August 1976); T. W. Merrick, "Employment and Earnings in the In-
formal Sector in Brazil," Journal of Developing Areas (1975); J. Weeks, "Policies for Ex-
panding Employment in the Informal Urban Sector of Developing Economies," International
Labour Review (January 1975); Stephen Guisinger and Mohammed Irfan, "Pakistan's In-
formal Sector," Journal of Development Studies (July 1980); Peter Lloyd, The 'Young
Towns' of Lima: Aspects of Urbanization in Peru (1980). Several studies by the International
Labour Office on urban development and employment also relate to the informal sector in
Abidjan (1967), Calcutta (1974), Sao Paulo (1976), Jakarta (1976), Bogota (1978), and
Lagos (1978). See also "Third World Migration and Urbanization: A Symposium," Eco-
nomic Development and Cultural Change (April 1982).
 III.D. GROWTH WITHOUT
EMPLOYMENT-NOTE

Although Lewis's two-sector model did not so evident. The organized sector comprises
intend it, the capitalist sector in his model plantations, estates, and mines with modern
has, in practice, become identified with in- management, advanced techniques of pro-
dustry or the urban sector, while the noncap- duction, and wage employment. Widespread,
italist sector has become identified with ag- however, is the informal subsector in which
riculture or the rural sector. It may be more production is still of the traditional subsis-
perceptive, however, to recognize that in ac- tence variety with production for household
tuality a"double dualism" has arisen within consumption.
the poor country. Not only is there rural- Many development plans have been prem-
urban dualism, but also within each of the ised with objectives of transferring resources
two sectors there are two subsectors that out of agriculture to the industrial urban sec-
might be termed the "organized" and "infor- tor and achieving a marked decline in the rel-
mal" subsectors. The organized subsector in ative size of the agricultural labor force. An
the urban sector is composed of wage earners early belief of development planning was that
in formal employment, is characterized by the process of industrialization could provide
modern management and modern techniques a substantial growth of employment oppor-
of production, and is protected by govern- tunities inthe modern urban sector. And yet,
mental policies. one of the most perplexing — and serious —
As noted in selection III.C.2, the informal problems now confronting many developing
subsector, in contrast, is composed of the self- countries is their growing level of urban un-
employed and small-scale traditional crafts employment and underemployment in the
and services, all unprotected by governmen- modern industrial sector. Perplexing — be-
tal policies. In the informal subsector we find cause levels of unemployment and underem-
the hawkers, porters, shoeshine boys, but also ployment have risen despite a rise in the rate
the small-scale craftsmen, small retail trad- of investment and an expansion in output. Se-
ers, own-account workers, and unpaid family rious— because this intensifies social resent-
workers. Employment opportunities in the in- ment and political unrest.
formal subsector are created by supply: ne- As elaborated in Chapter I, the increase in
cessity drives people to work in every con- the absolute number of unemployed and un-
ceivable way. Workers in the informal deremployed questions whether development
subsector may actually be working long is actually occurring. Unemployment, under-
hours at extremely difficult physical labor, employment, low productivity employment,
but their productivity is very low, and their and the "working poor" are all aspects of the
meager income is variable and frequently employment problem. The problem is not
shared with others. With the extensive rural- confined to one sector, but pervades the en-
urban migration, and the incapacity of the tire economy.
urban organized subsector to absorb the mi- Although we may wish for more refined
gration, the informal subsector of the city has statistics,1 a number of studies have empha-
acted as a sponge for the surplus labor. In sized the broad dimensions of the problem:
most LDCs the number in the urban infor-
'For some problems of statistical coverage and a
mal subsector has risen not only absolutely discussion of statistical methods, see Edgar O. Ed-
but also as a proportion of the total labor wards (ed.), Employment in Developing Nations,
force. In urban centers, such as Calcutta or 1974, passim; David Turnham, The Employment
Problem in Less Developed Countries (1971): Erik
Bombay, it is estimated that one-half or more
of the workforce is in the informal subsector. Thorbecke, "The Employment Problem: A Critical
Evaluation of Four ILO Comprehensive Country Re-
In the rural sector a similar subdivision is ports," International Labor Review, May 1973.

191
 192 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

industrial employment has lagged behind Lewis believed that "this brings the modern
growth in industrial output, behind growth of sector as much labor as it wants without at
the urban population, and even behind the the same time attracting much more than it
general growth rate of population. Only a can handle."3 Furthermore, the model pos-
portion of the annual increase in the urban tulates that wage rates should not rise with
labor force has been absorbed in the urban increasing productivity, but instead that cap-
organized subsector. ital formation and technical progress in the
In numerous developing countries, despite capitalist sector should raise the share of prof-
creditable rates in aggregate growth, it is not its in the national income. To the extent that
uncommon for the rate of open unemploy- the profit ratio rises, there should then be
ment (not disguised unemployment or under- capital-widening investment in the industrial
employment) inmajor urban areas to be as sector, so that the demand for labor continues
high as 15 to 20 percent. Even worse, the to rise and more industrial workers are em-
rates of urban unemployment in the age ployed at a constant real wage. Finally, after
group of 15-24 are generally about double the surplus labor is absorbed, wages begin to
the rates of unemployment among the urban rise.
labor forces as a whole.2 In actuality, however, the real income gap
The number of underemployed is consid- between the modern and rural sectors has
erably greater than those in open unemploy- been much greater than allowed for in the
ment. Underemployment can be either visi- Lewis model. The wage rate in the modern
ble or invisible underemployment. Those in sector has been higher than that needed to
visible underemployment would work more cover the cost of transfer and the higher
hours if the employment opportunity were urban costs of living. And the differential
available. Those in invisible underemploy- above rural income has widened. The wage
ment are working, but at low productivity level in the industrial sector has risen in spite
occupations that are below a cutoff point in of open unemployment and before the sur-
productivity or earnings. plus labor of the rural sector has been ab-
The labor utilization problem in a devel- sorbed. Ithas also continued to rise in many
oping economy is chronic and in large part a of the LDCs, although the average product in
problem of underemployment. This suggests agriculture may have been even stagnant in
that the Keynesian unemployment model some economies. Instead of Lewis's sug-
that focuses on short-term cyclical unem- gested 50 percent differential, the average
ployment ofan open character, caused by a real wage for workers outside of agriculture
deficiency of aggregate demand, does not fit has commonly been two to three times
the less developed country. greater than the average family income in
Furthermore, it indicates that the actual the traditional sector.
course of industrialization has deviated con- Most importantly, the inflow of labor to
siderably from Lewis's model of development the modern sector has actually been "more
with unlimited supplies of labor. As we saw than it can handle": contrary to what is to be
in selection III.B.l, the essence of the Lewis expected from the Lewis model, an exceed-
model is that wages in the modern sector are ingly high rate of unemployment and under-
based on the average product of labor in the employment has materialized in the modern
traditional rural sector, but are somewhat sector. Those formerly in disguised unem-
higher — for unskilled labor, normally about ployment inthe rural sector have, in effect,
50 percent above the income of subsistence transferred into visible unemployment and
farmers — in order to attract labor into the underemployment in the modern sector.
modern sector and compensate for the higher
cost of urban living and any nonpecuniary 3W. A. Lewis, Development Planning, London,
disadvantages. At this higher wage rate, 1966, pp. 77-8, 92. Also W. A. Lewis, "Unemploy-
ment in Developing Areas," in A. H. Whiteford (ed.),
2See Edwards, Employment in Developing Na- A Reappraisal of Economic Development, Chicago,
tions, p.13. 1967, p. 5.
 193 GROWTH WITHOUT EMPLOYMENT— NOTE

The reasons for this can be found in some sures have increased in many countries, and
of the actual deviations from the conditions labor-supported governments have shown
of the Lewis model and in some structural some sympathy to such pressures. Moreover,
distortions that have been perpetuated by in- the monopolistic structure of many product
appropriate policy measures. markets has facilitated the passing on of
The rate of urbanization has indeed been higher wages in the form of higher prices. In
high. The amenities and public services of the several countries, union pressure in crucial
urban area are attractive in themselves to sectors of the economy — for instance, in the
labor from the rural sector. But the strongest oil, copper, and bauxite industries — has been
inducement has been the widening income dif- instrumental in setting a pattern of wage in-
ference between urban wages and rural in- creases inother sectors.
come at the same time as rural employment More significantly, governmental policies
opportunities have not expanded. Fundamen- have been directly instrumental in raising
tally, itcan be submitted that the employ- urban wages. The public sector is frequently
ment problem in the urban area has been the the largest sector of wage employment and
result of a premature increase in the indus- also the only sector that is highly organized.
trial wage level combined with a premature Wages in the public sector have risen rapidly
reduction in agricultural employment.4 To a and have commonly acted as the base for a
lesser extent, but still significantly in some wider pattern of wage increases.
countries, labor has been released from the In newly independent countries, the salary
very labor-intensive indigenous handicraft scales are still basically those that were paid
industries that cannot compete with the to expatriates during the earlier colonial pe-
growth in new manufacturing activities. riod; but this scale does not now conform to
"Rationalization" of labor practices in the the utilization of the domestic supply of
tertiary sector has also tended to increase the labor, and it puts undue pressure on the wage
supply of labor to the urban industrial sector. structure. Nor can the heightened expecta-
As already noted, the urban wage level has tions from the extension of education be ful-
not been controlled by real earnings in agri- filled. Furthermore, minimum wage regula-
culture. Urban wages have commonly risen tion has been influential in raising urban
to two to four times higher than agriculture wages and in having a great impact on the
earnings. Urban wages have risen indepen- total wage structure in a developing country.
dently through the wages policies of the gov- The minimum wage in a dominant industry
ernment and trade unions.5 Trade union pres- is frequently negotiated with the government
4In Africa, the unemployment among school leav-
on a basis of "an ability to pay" criterion; but
ers is one indication of this. Not only are there school this wage tends to spread through other in-
leavers within the city, but also those who attend dustries. The increase in the minimum wage
schools in villages reject the traditional occupations will have considerable effect in raising the
on the land and migrate to the cities in search of
whole wage scale since the wages being re-
wage-paid jobs. See A. Callaway, "Education Expan- ceived bymost of the unskilled workers are
sion and the Rise of Youth Unemployment," in P. C. at or near the current minimum wage. The
Lloyd et al. (eds.), The City of Ibadan, Cambridge,
1968, pp. 197-209. generalization of a minimum wage may then
5Lewis did recognize the effects of minimum wages become highly unrealistic because it is obliv-
and union action as being among the various possibil- ious to conditions of supply and demand in
ities that could cause the process of absorbing surplus
labor from agriculture to come to a premature halt; the labor market, living standards in the tra-
ditional sector, and the effects on the wage
see W. A. Lewis, "Economic Development with Un-
limited Supplies of Labor," The Manchester School
structure as a whole. Workers who were only
of Economic and Social Studies, May 1954, pp. 172—
3. But the model is allowed to run its course without
marginally useful — but nonetheless em-
this restraint coming into effect. ployed at the lower wage — become redun-
In his Development Planning, however, Lewis does
dant when the minimum wage rises.
concentrate more on the resultant unemployment; see Minimum wage policies for unskilled labor
pp. 76-87. have the effect of making the skilled-un-
 194 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

skilled wage differential too narrow, as has rates comparable to those in the advanced in-
happened in many African and Asian coun- dustrial countries. But whereas in the indus-
tries. Market forces of supply and demand trial countries real wages have increased
are left to determine wages for skilled labor, roughly in line with average national prod-
but demand rises only slowly so that the mar- uctivities, the rise of wages in the developing
ket-determined wage for skilled labor also countries often implies an increase consider-
tends to rise slowly. If governments then in- ably faster than that in real national product
sist that unskilled wages should increase in- per capita.
dependently ofdemand and supply condi- The consequence of this has been the use
tions in the unskilled labor market, there is a of more capital-intensive production meth-
likelihood that unskilled wages will increase ods, either through the introduction of labor-
faster than skilled wages, and that relatively saving machinery in response to rising wages
low-wage labor will become overvalued. or through improvement in personnel and
In default of adequate profit taxation or production management practices that have
other tax policy, governments have also trimmed the labor requirements per unit of
found it convenient in effect to "tax" com- output.
panies— especially foreign companies — Capital-intensive methods of production
through wage increases. The government's have also been subsidized by other price dis-
policy of encouraging higher wages may ini- tortions— especially through too low a rate of
tially be directed only at foreign companies interest and too low a price for foreign ex-
in order to prevent "excess" profit repatria- change. When interest rates in the urban sec-
tion and to raise the share of income for do- tor do not reflect the true scarcity of capital,
mestic factors. But the demonstration effect a bias is imparted to capital-intensive pro-
of higher wages in the foreign enterprises also duction methods. This is often intensified by
causes a spread of higher wages to other inflation which lowers the real rate of interest
enterprises. below the nominal rate, possibly even to a
At the same time as government policies negative real rate. So too is there a bias to-
have supported urban wage increases, no par- ward more advanced production techniques
ticular attention has been given to the level when the LDCs currency is overvalued in
of agricultural wages. The result has been a terms of foreign currency, and the true cost
widening gap between the urban and agri- of importing machinery is hence underval-
cultural wage levels. Such a large differential ued. Governments have also lowered the rel-
has served to attract the disguised unem- ative price of producers' equipment by such
ployed from the rural sector to the urban sec- measures as allowing duty-free importation
tor, but it has simultaneously kept industrial of equipment, a preferential exchange rate,
labor overpriced. Moreover, the differential and making available foreign exchange for
between urban and rural wages has pro- servicing loans from overseas machinery sup-
ceeded to widen in face of the substantial and pliers. When domestic enterprises are pro-
growing urban unemployment and underem- tected bytariffs and import quotas, the pres-
ployment. With the rising expectational sure to economize on capital is also less than
wage, it has become increasingly difficult to it would be in more competitive markets.
absorb the excess supply of labor. It is most significant that the strategy of
Although the Lewis model envisages suffi- industrialization via import substitution has
cient capital-widening investment in the in- dominated the expansion of the industrial-
dustrial sector to absorb the labor inflow, the urban sector. In Chapter VIII, we examine in
actual result has been a substitution of capi- some detail the policies used to promote the
tal for labor in the modern organized sector. home replacement of imported final goods.
Contrary to the model, wage rates in many of At this point, we need only recognize that the
the LDCs have actually risen more rapidly attempt to industrialize via import substitu-
than productivity. Real wages have risen at tion has generally been accompanied by in-
 195 GROWTH WITHOUT EMPLOYMENT— NOTE

flation and an overvalued exchange rate. sector might be expected to have initially a
These policies have resulted in a distorted substantial rate of growth as imports are re-
price structure in many LDCs: too low a rate placed. But this may be a once-for-all expan-
of interest in the urban sector, too low a rate sion with little subsequent reinvestment — un-
for foreign exchange, and too high a level of less the home market continues to grow, or
urban wages. the process of import substitution can pro-
The capital-intensive bias is also supported ceed on from the final stages of production
by a number of other measures. Employers down through the production process to the
tend to seek means of reducing their labor re- replacement of intermediate goods; or the
quirements when the government uses wage import-replacement industry is able to gain a
policies as a substitute for social legislation competitive advantage in export markets.
by requiring family allowances, pensions, li- Such opportunities for the continual expan-
censing and health measures, or other fringe sion of the modern sector have not material-
benefits bordering on social insurance. Offi- ized, and the capital-widening investment of
cially required fringe benefits and wage sup- the Lewis model with an ever-expanding de-
plements may commonly amount to as much mand for surplus labor has not been
as 30 to 40 percent of the basic wage. When sustained.
the employers are foreign enterprises, they At the same time as domestic policies have
are also likely to be simply imitating the ad- had the effect of subsidizing capital-intensive
vanced techniques of production known in import-substitution industries, they have im-
the advanced country — techniques which are plicitly imposed a levy on domestic agricul-
appropriate for the factor supply of the ad- ture. This has gone against an expansion in
vanced country but not for the labor surplus labor-intensive agricultural output and a rise
of the less developed country. in rural employment.
Contrary to the Lewis model, the expan- For these various reasons the employment
sion of the modern industrial sector has also problem has become a central problem of
tended to slow down in many LDCs. Being development.
based on import substitution, the industrial
Comment
Lewis later recognized the unemployment problem and stated that
The most important ingredient in employment policy is to prevent too large a gap opening up between
wages in the modern and earnings in the traditional sectors. So long as the traditional sector is not
disturbed by a large income gap, it can hold and provide for all the people whom the modern sector is
not yet ready to employ. . . .
Other ingredients are: measures to prevent excessive capital intensity; avoidance of an overvalued
currency; adequate expenditure on developing the countryside; curbing the growth of a few large towns,
in favor of developing more numerous small urban centers; and a population policy. . . . Deliberate action
to substitute labor for machinery, in accordance with shadow pricing, even if confined to the public sector
would go a long way towards eliminating open unemployment (Lewis, Development Planning 1 966, p.
83).
 III.E. MEASUREMENT OF THE
"UNEMPLOYMENT
EQUIVALENT'-NOTE

If there is not one employment problem but (4) by the productivity criterion, he is remov-
many, there must be various approaches to able from his present employment in the
the measurement of unemployment, and the sense that his contribution to output is less
relevant approach must depend on the partic- than some normal productivity, and therefore
ular problem or policy question at hand. It his removal would not reduce output if the
has already been emphasized, however, that productivity of the remaining workers is nor-
the employment problem in LDCs cannot be malized with minor changes in technique or
interpreted as simply a Keynesian type of in- organization.
voluntary unemployment. A pervasive prob- Another study also emphasizes that there
are several dimensions of underutilization of
lem isthat of the "working poor" — those who
actually work long hours but earn only a low labor:
income below a poverty line. The disguised
In addition to the numbers of people unem-
unemployed constitute another major dimen-
sion of the employment problem. Beyond ployed, many of whom may receive minimal in-
comes through the extended family system and
measures of "open involuntary unemploy- therefore not rightly classified with the very poor,
ment" it is just as important — if not more it is also necessary to consider the dimensions of
so — to have measures of the "underem- (1) time (many of those employed would like to
work more hours per day, per week or per year),
ployed" and "disguised unemployed."
We might usefully think of a range of un- (2) intensity of work (which brings into consider-
employment, beginning at one extreme with ation matters of health and nutrition), and (3) pro-
ductivity (lack of which can often be attributed to
open unemployment in the urban area., de-
inadequate, complementary resources with which
fined as "zero hours work and zero income."
Beyond this extreme, we can apply four to work). Even these are only the most obvious di-
mensions of effective work, and factors such as
major criteria for determining whether a per- motivation, attitudes, and cultural inhibitions (as
son may be called unemployed or underem- against women, for example) must also be consid-
ployed: (1) the time criterion, (2) the income ered. Our discussions have thrown up the following
criterion, (3) the willingness criterion, and forms of underutilization of labor, which may in-
(4) the productivity criterion.1 Thus, we may dicate the diversity of the phenomenon but which
call a person unemployed or underemployed further study will probably show to be incomplete:
if either: (1) by the time criterion, he is gain- 1. Open unemployment — both voluntary (people
fully occupied during the year for a number who exclude from consideration some jobs for
of hours (or days) less than some number of which they could qualify, implying that they
normal or optimal hours (or days) defined as have some means of support other than employ-
full employment hours or days; or (2) by the ment) and involuntary.
income criterion, he earns an income per year 2. Underemployment — those working less (daily,
less than some desirable minimum; or (3) by weekly, or seasonally) than they would like to
work.
the willingness criterion, he is willing to do
more work than he is doing at present — he 3. The visibly active but underutilized — those
may either be actively searching for more who would not normally be classified as either
work or be available for more work if it is of- unemployed or underemployed by the above def-
initions, but who in fact have found alternative
fered on terms to which he is accustomed; or
means of "marking time," including
'These criteria are presented by Raj Krishna, "Un- a. Disguised underemployment. Many people
employment in India," Economic and Political seem occupied on farms or employed in gov-
Weekly, March 3, 1973, p. 475. ernment on a full-time basis even though the

196
 197 MEASUREMENT OF THE "UNEMPLOYMENT EQUIVALENT"— NOTE

services they render may actually require teria are applied. A gross measure of the em-
much less than full time. Social pressures on
private industry also may result in substan- ployment problem might be termed the "un-
tial amounts of disguised underemployment. employment equivalent" — that is, the sum of
those in open unemployment plus a measure
If available work is openly shared among
those employed, the disguise disappears and of the "equivalent" of the underemployed.
Conventional measures underestimate both
underemployment becomes explicit.
the size of the labor force and the amount of
b. Hidden unemployment. Those who are en-
gaged in "second choice" nonemployment unemployment. In addition to the "involun-
activities, perhaps notably education and tary unemployed" (those conventionally de-
household chores, primarily because job op- fined as "seeking work at the existing wage
portunities are not (i) available at the levels rate"), there are many others who are not
of education already attained, or (ii) open to seeking work because they estimate that the
women, given social mores. Thus, educa- probability of finding employment is too low,
tional institutions and households become
or they lack skill qualifications, or they suffer
"employers of last resort." Moreover, many from malnutrition or ill health. Even more,
of those enrolled for further education may
be among the less able, as indicated by their there is a part of the labor force that may be
actually employed but only for a limited time
inability to compete successfully for jobs be-
fore pursuing further education. in part employment or seasonal employment.
c. The prematurely retired. This phenomenon There is also a large portion of the labor force
is especially apparent, and apparently grow- that constitutes the "working poor." Finally
ing, in the civil service. In many countries, there are those who, while working, are un-
retirement ages are falling at the same time
that longevity is increasing, primarily as one productive: they are "underemployed," and
means of creating promotion opportunities they constitute part of the "labor reserve"
that could through effective policy measures
for some of the large numbers pressing up
be allowed to work more hours or more pro-
from below.2 ductively. Just what these policy measures
Different estimates of "unemployment" might be is the concern of much of the re-
will be possible according to which criterion mainder ofthis book. The essential point here
or various combinations of the different cri- is that a labor reserve does exist, and that be-
cause of the underutilization of labor this re-
serve exceeds the conventional measures of
2Edgar O. Edwards (ed.), Employment in Devel-
oping Nations, New York, 1974, pp. 10-11. "unemployment."
Comment
A review of several studies that attempt to measure the amount of unemployment and
underemployment can illuminate the different concepts of "unemployment" and indicate sta-
tistical difficulties in measurement. See Gunnar Myrdal, Asian Drama, (1968), part 5, ap-
pendix 6;Richard Jolly et al. (eds.), Third World Employment (1973); P. P. Streeten, "A
Critique of Development Concepts," European Journal of Sociology, Vol. 11, No. 1 (1970);
Edgar O. Edwards, Employment in Developing Nations (1974); S. Paglin, "Surplus Agricul-
tural Labor and Development: Facts and Figures," American Economic Review, (September
1965); E. Kritz and J. Ramos, "The Measurement of Urban Underemployment," Interna-
tional Labour Review (January -February 1976); Peter Gregory, "Employment, Unemploy-
ment, and Underemployment in Latin America," Statistical Bulletin of the OAS (October-
December 1980); David Turnham, The Employment Problem in Less Developed Countries,
OECD Development Center Studies (1971).
 III.F. EMPLOYMENT
STRATEGIES
III.F. 1 . Output and Employment
Trade-offs*

Neither of the objectives, maximum output At a given time, with a given stock of cap-
and maximum employment, are unambigu- ital equipment (inherited from the past), the
ous. The output objective is ambiguous be- employment of more men on that equipment
cause output at any time consists of a heter- is likely to increase output, though it could be
ogeneous collection of goods. Types of that, as a result of the reorganization of the
employment, in duration — daily, weekly and work, of less efficient production methods
seasonally — in effort and by regions, etc. also used, of people standing in one another's way
differ. In addition, both output and employ- or of a fall in efficiency for some other reason,
ment occur over time. Current levels of out- the extra workers do not add to, or even sub-
put and employment may influence future tract from, production. However, the choice
levels. Weighting therefore both intra- and facing a country is not simply a question of
inter-temporally is crucial to the definition of employing additional men with the existing
the objectives. However, we shall begin by ig- capital stock but of the type of new equip-
noring these ambiguities and assume that our ment to install, and in this decision about the
sole concern is with current levels of output nature of new investment there can be a con-
and employment, and that maximizing cur- flict between output and employment. Given
rent levels automatically leads to achieve- that the total funds available for new invest-
ment of future objectives, or put more for- ment are limited, using the funds for equip-
mally, that maximizing current levels of ment to employ people in one way will inevi-
output and employment is equivalent to max- tably mean not using the same funds for
imizing the present value of the entire some other equipment which may involve less
streams of output and employment over time. employment but might also produce more
We shall also begin by assuming that there is output. Maximizing output involves using
a single index for output and employment. scarce resources as efficiently as possible. If
capital is the scarce resource, it involves min-
CONFLICTS RESULTING FROM imizing the capital-output ratio. The type of
SCARCE COMPLEMENTARY production this requires may be, but need not
FACTORS OF PRODUCTION be, consistent with maximizing employment.
We can then rephrase the question and Suppose in the textile industry the mini-
mum capital-output ratio is associated with
ask: is maximum current production compat- fairly modern style industry. If £100,000 is
ible with maximum employment? On the available for investment in textiles, if the
face of it, the answer seems to be an obvious
capital-output ratio is 2.5, and if the capital
"yes." More men must surely be able to pro- cost per work place for this type of factory is
duce more. It is hard to picture conditions in
£1000 (assuming a firm degree of utilization
which it is impossible to find anything useful
to do for extra hands. of capacity), then investing all funds in this
modern factory will involve extra employ-
ment of 100 and extra output of £40,000. An
*From F. Stewart and P. P. Streeten, "Conflicts be- alternative way of investing the funds might
tween Output and Employment Objectives," in R. be to introduce hand-spinning. Suppose for
Robinson and P. Johnston (eds.), Prospects for Em-
ployment Opportunities in the 1970s, HMSO, 1972, this the capital-output ratio was 5.0 and the
pp. 367-73, 375-8, 381-4. Reprinted by permission. cost per work place £100; then the extra out-

198
 199 EMPLOYMENT STRATEGIES

put resulting from using the funds for hand- number of people employed with any given
spinning would be £20,000 and the extra em- machine may vary. . . .
ployment generated 1000. In this case there So long as output is responding positively
is a fairly dramatic conflict between employ- to additional workers the level of employ-
ment and output maximization. It should be ment associated with a given machine will
noted that this conflict (which is a fairly re- depend partly on the level of wages. Even
alistic one if one examines actual figures for where output is invariant with respect to em-
costs, etc. in the textile industry)1 arises be- ployment the actual employment associated
cause the more labour-intensive method in with given machinery may depend partly on
the sense of the method which uses a lower real wages since managerial effort may be
capital-labour ratio or shows lower cost per substituted for employment as real wages
work place, actually involves more capital rise. Thus the employment level associated
per unit of output than the capital-intensive with any given machine may not be indepen-
method. Some theoretical models assume this dent of the wage rate. In the examples above
can never happen. It would be true that it some wage level is implicitly assumed in as-
could not, if all techniques of production sociating each machine with a unique output
were invented and developed simultaneously, and employment level. A range to represent
since the labour-intensive methods which use output and employment at different wage
more capital would never be developed. But levels would have been a more realistic rep-
in fact methods of production are developed resentation. Ifone assumes that continuous
over an historical period with the more la- variations in output are associated with con-
bour-intensive methods generally originating tinuous variations in employment for each
from an earlier period. One reason why this machine, but that there is diminishing mar-
sort of situation develops is the existence of ginal product as employment is increased,
economies of scale; as machinery has been one is back in a neo-classical world where
adapted for larger-scale production the capi- there are variable factor proportions and any
tal costs in relation to output have tended to amount of capital (or any machine) may be
fall, so that for large-scale production the associated with any level of employment. In
later and more capital-intensive methods this neo-classical world the limit to employ-
tend to economize on capital in relation to ment is set by the real wages workers de-
output. For small-scale production the older mand. There can be no conflict between out-
machinery may remain efficient.2 Implicit in put and employment because every type of
this example is the assumption that there is a machine can be associated with any amount
specific level of employment associated with of employment. Thus if the modern factory
each technique, and thus that it is sensible to methods were employed in spinning, the
talk of a "cost per work place." In fact, the extra 900 workers could be employed in the
factory and would each add to output. At
'Bhalla (1964), suggests that the capital-output least as much employment and more output
ratio (including working capital as well as fixed cap-
ital) using factory methods in cotton spinning is about could result from choice of the factory alter-
three quarters of that using the hand Ambar Charkha native. We do not believe that this is a real-
methods. Bhalla's analysis of rice-pounding (1965), istic assumption. Though some variation in
suggests a similar conflict here; the technique which
maximizes employment (or has the lowest cost per employment is possible with any given ma-
work place), the pestle and mortar, requires nearly chine, there comes a point at which the ma-
twice the capital per unit of value added compared chine is operating at its maximum pace,
with the large sheller machine. The latter requires in- when additional workers do not increase out-
vestment per work place of about 100 times as great
as the former. put. There is thus a limited range of employ-
ment possibilities associated with each ma-
2The importance of scale in determining the effi-
cient range of production possibilities is emphasized
chine, which means that there can be a
by many empirical studies, including Boon (1964); conflict between output and employment. Put
Strassmann (1968). in another way, for any positive real wage
 200 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

there comes a point at which it is no longer REASONS FOR PREFERRING

worth while employing extra workers with a EMPLOYMENT TO OUTPUT
particular machine. This point may be
reached at lower employment level and a If a conflict between maximizing current
higher output level for one machine than for output and employment were inevitable, why
another. In this case there is a conflict be- should we wish to sacrifice output to employ-
tween output and employment, which is in- ment? Four possible answers occur to us,
dependent ofany institutional or other lower though others might think of additional rea-
limit on the level of real wages. sons. First, employment creation and the con-
Just as some economists assume that such sequential wage payments may be the only
a conflict between output and employment mechanism by which income can be redistri-
cannot arise, others assume not only that it buted to those who would otherwise remain
has arisen in the past, but also that it neces- unemployed. With an efficient fiscal system,
sarily must arise. The capital-intensive meth- taxation combined with unemployment re-
ods of production, it is claimed, will always lief, free social services and other forms of
involve lower capital costs per unit of output assistance to the unemployed could be used
(and higher costs per work place) than the la- as an engine of redistribution. In an under-
bour-intensive methods.3 developed society, loyalty to the extended
This position is as extreme as the other. family may induce the employed wage earner
There is considerable evidence that in many to share his wages with his often large ex-
industries, and in many processes, the more tended family. But if neither fiscal system
labour-intensive methods also save capital nor family provide a systematic channel of
per unit of output.4 In these cases maximiz- redistribution, job creation may have to be
ing employment and output are consistent. used for this purpose. Production will then be
Probably of more significance is the possibil- sacrificed for better distribution and, as a
ity of devoting research and development (R means to this, greater employment.
and D) efforts to the labour-intensive meth- Second, unemployment is demoralizing.
ods (in the sense of low capital cost per work To feel unwanted, not to be able to make any
place) so that they become efficient as com- contribution, lowers a man's morale and
pared with capital-intensive methods. Pres- makes him lose his self-respect. The preser-
ent possibilities reflect the fact that almost all vation of self-respect is worth sacrificing
R and D is concentrated on producing meth- some production. As Barbara Ward has said,
ods suitable for the developed world, in which "of all the evils, worklessness is the worst" —
labour is scarce; the labour-intensive meth- clearly not only and even not mainly because,
ods currently available are generally the and if, it lowers national product. It is worth
products of earlier and less sophisticated sci- sacrificing production to reduce this evil.
ence and technology. . . . Third, it might be thought that work is in-
3See, for example, Kaldor (1965), pp. 28-9: "There trinsically good, whatever its impact on mo-
is no question from every point of view of the superi- rale, self-respect and other subjective feel-
ority of the latest and more capitalistic technologies" ings. The Puritan ethic may commend job
(our italics). Similar emphasis on the overall superi-
ority of capital-intensive techniques is found in Amin creation as valuable irrespective of its contri-
(1969), pp. 269-92. bution to production. Puritanism played a
4A. K. Sen suggests that in cotton weaving the cap- valuable part in making desirable the neces-
ital-output ratio is the lowest for the most labour-in- sary but unpleasant sacrifices which pro-
tensive technique, the fly shuttle hand loom, and high- moted the industrial revolution in Britain.
est, nearly 2% times as big, for the automatic power Whether this ethic, where it has been
loom (again including working capital). Evidence for
the existence of a range of efficient techniques in a adopted in the developing world (and it is no-
number of industries below a certain critical scale of table that most Puritan-like statements tend
output is also contained in Boon (1964). to come from expatriates) should be encour-
 201 EMPLOYMENT STRATEGIES

aged and where it has not been adopted gued that it is not just a question of income
should be promoted, if it leads to a situation redistribution but of providing a chance to
which impedes rather than speeds up devel- earn, not simply receive, the higher incomes.
opment by requiring the adoption of ineffi- These are the only reasons we can think of
cient techniques to compensate for the ma- as to why the employment objective might
sochistic value placed on work, is another conflict with the output objective, and sacri-
question. Other aspects of Puritanism are fice of output to employment, properly
certainly conducive to development; to the defined, is justified. But what are the proper
extent that Puritanism is a package deal, this definitions? As argued in the first paragraph,
aspect may have to be accepted along with objectives are ambiguous. Two types of am-
the rest. biguity are relevant. First, national product
Fourth, there are obvious political disad- consists of a heterogeneous collection of
vantages and dangers in widespread unem- goods, "of shoes, and ships and sealing wax,
ployment and non-employment. This is an of cabbages" (and possibly of the services of
important reason for valuing employment kings) and it accrues to different people, in
since, in so far as anyone does, it is the poli- different regions, with varying needs. In put-
ticians who lay down "the objective function" ting all these together we must use a system
of society. Political instability may, in any of weighting the different items and different
case, eventually endanger output levels and sets of weights may lead to contradictions.
growth. . . . One set may give the impression that we are
The desire for employment which is so ap- sacrificing product for employment, another
parent in many countries cannot be entirely may not.
divorced from the desire for higher incomes. Another ambiguity arises because both
Many of those seeking urban employment production and employment occur in time
are looking for work at the going wages in the and stretch into the future. An infinite num-
organized industrial sector, where wages are ber of time profiles within any horizon that
generally considerably higher than incomes we care to consider can be drawn up. Any
obtainable elsewhere. Discussion of the need profile for either of our two objectives that
for rural employment opportunities to reduce lies all the way below another profile of the
the underutilization of labour normally takes same objective can be dismissed as ineffi-
place in the context of the need to create op- cient. But in order to choose between those
portunities for increasing incomes through that intersect at some moment of time, we
fuller labour utilization. Again the need is for must make additional choices in the light of
incomes as much as for work. It is unlikely our policy objectives. What if 5 per cent less
that the unemployed, or those scratching a employment now gives us 15 per cent more
living in the rural areas, would be prepared employment in two years' time? What if a
to suffer some loss in their incomes for the rise of 10 per cent in employment now pre-
sake of more work. What is wanted is in- vents us from employing an additional per
creased opportunities to work and earn cent of a vastly larger labour force in 10
higher incomes. Because both work and years and after? We must turn to the prob-
higher incomes are required it is difficult to lems of weighting and timing.
disentangle the two. Clearly, the desire for
redistribution of income is of prime impor- WEIGHTING: TIME
tance. To achieve this redistribution, employ-
ment opportunities may be needed but the Another serious ambiguity arises from the
sacrifice, or trade-off involved may be of the fact that sacrifices now may yield gains in the
income of the better-off for the sake of that future. We must consider two opposite sets of
of the worse-off, rather than of output for the circumstances: first, where less production
sake of employment. However, it can be ar- and more employment now lead to more pro-
 202 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

duction later than would otherwise have been ment now for the sake of producing more, we
possible; second, where less employment and can provide the men (and their children) with
more production now lead to more employ- more jobs later. This is only partly a matter
ment later than would otherwise have been of investment, i.e., producing now the ma-
possible. chines, or resources with which to buy the
In order to illustrate the first case, let us machines, that will give jobs tomorrow. A
return to the situation where men were de- greater volume of food which provides better
moralized byunemployment. We then re- nutrition for the workers and their children,
garded self-respect and high morale as ends of health measures and of certain forms of
in themselves. But we may also regard them education can also contribute to greater em-
as necessary for the continued employability ployment (and fuller labour utilization) in
of men. If men remain unemployed for long, the future. The point leads once again to in-
their skills as well as their attitudes deterio- come distribution, but this time not valued
rate and they are incapable of producing as independently as desirable, but as instrumen-
much later. This situation cannot be reme- tal to faster growth. The choice between
died by unemployment assistance, for it is maximum employment and maximum output
only on the job that ability to work and mo- reduces to one between jobs now or later, be-
tivation are maintained. Just as machines cause more output can promote more, and
sometimes have to be kept going in order to more effective, employment in the future. To
prevent attrition or rust, so workers and raise employment means sacrificing not only
teams of workers have to be kept busy to pre- output now (and, on our assumption, the rate
vent them from becoming rusty or apathetic. of growth of output) but also the rate of
Current employment, even where there is growth of employment. This means that at
nothing to show for it, can be regarded as a some future date the level of employment will
form of investment — human maintenance — be lower than it would otherwise have been.
which prevents future deterioration of pro- To go back to the example discussed earlier,
ductivity. In addition men's productive ca- suppose in each case, modern factory and
pacity, their ability to work, their initiative hand spinning wheel, 20 per cent of income
and organizational ability and their concen- generated is saved. The factory solution will
tration may not merely be maintained but involve £8000 investment available in the
may actually be increased by working. This next year, while the hand wheel alternative
form of learning by working means that the will involve £4000. The divergence will get
greater current employment opportunities, greater in subsequent years. The factory al-
the greater is future productive capacity. ternative will lead to an annual growth in in-
The second case works in the opposite di- come (and assuming the same £1000 a work
rection and is possibly the most important place technology is adopted, in employment)
way in which an apparent conflict between of 8 per cent per annum while the hand spin-
employment and output arises. Here we max- ning wheel alternative will lead to 4 per cent
imize production in the short run, even annual growth in output and employment.
though it means tolerating now more nonem- (This ignores the impact of extra consump-
ployed, because the extra production enables tion on growth.) Figure 1 illustrates the pos-
us to generate more jobs later than would sib lit es.. .
otherwise have been possible. If there is a As to the right choices, a good deal will de-
current conflict between output and employ- pend upon our time horizon and on certain
merit, it must be remembered that output is future developments. As far as employment
useful not only for itself, but can be used to is concerned, the life span of one generation
generate more employment. and perhaps its children will be relevant, but
The inter-temporal "trade-off" between few societies would be prepared to tolerate
employment now and employment tomorrow widespread unemployment over two genera-
arises because, by tolerating more unemploy- tions to improve the job prospects of their
 203 EMPLOYMENT STRATEGIES

employed and those not employed and the ac-

companying resentment may work the other
way. Poverty in the midst of affluence is
worse than plain poverty widely shared. This,
and the question of numbers, suggests that it
may be correct to give greater relative weight
to future as against current employment than
Time Time to future as against current output.
Planners must know not only their prefer-
FIGURE 1 ences between the present and the future, for
both output and employment, but also what
great-grandchildren. This is not only because opportunities there are for trade-offs. Con-
we show less concern for our great-grand- flicts between current levels of growth rates
children, but also because we may rightly of output and employment may arise either
hope that their prospects will improve for because growth rates are determined by sav-
other reasons, such as the development of ings rates (or, more generally, developmental
more appropriate technologies, improve- expenditure rates), savings rates by income
ments in motivation, administration, educa- distribution and income distribution by em-
tion, etc. Given the time horizon, we might ployment levels, or because growth rates are
say that, on the other hand, the richer society determined by the allocation of a given sav-
of tomorrow can look better after its unem- ings ratio between sectors and this allocation
ployed and to be unemployed then will be a has different effects on employment.
smaller hardship. On the other hand, with It is common to assume in this context that
present trends of growth of the labour force a capital-intensive technique leads to a
in less developed countries and likely oppor- higher savings ratio for the same income level
tunities for jobs, the total number of unem- than a labour-intensive technique. On this as-
ployed isincreasing rapidly. While the lot of sumption, lower employment now can give
a given number of unemployed will therefore faster growth of both output and employ-
be better in the future and the burden of ment. Those who make this assumption (Sen,
maintaining them lighter, the number to be 1968; Little and Mirrlees, 1968) assume:
looked after will be larger. In the more dis- 1. That a higher proportion of profits is
tant future, however, we may assume that
saved than of wages (at its most extreme
population control will have become effective
this assumption is that all profits are
or new scope for migration will have opened
up. In view of all this, it seems right to dis- saved, all wages consumed); and that con-
sumption makes no contribution to future
count future jobs and to give more weight to
more jobs now and in the near future. growth;
On balance it seems that the discount rate 2. That wage rates do not depend on
techniques;
that we should apply to employment may be
less than the one we apply to output. The 3. That the government is incapable of se-
main argument for applying a discount rate curing the savings ratio it desires by tax-
ing wage earners and generating adequate
to output is that the marginal utility of in- public savings or using inflation to reduce
come isless for a richer society. This does not real wages.
apply in the same way to employment — i.e.,
the value of extra employment generated Since the growth rate is the product of the
does not decline as the level increases — savings ratio and the output/capital ratio,
though increasing incomes per head may the effect on the growth rate of raising the
make employment in the future less impor- savings ratio by increasing the capital-inten-
tant as a means of income redistribution. On sity of technique adopted will depend on the
the other hand, the contrasts between those consequences for the output/capital ratio.
 204 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

TECHNICAL PROGRESS employment, will be less than the gains in

terms of growth of output.
Until now we have ignored technical Technical developments are likely to affect
progress and assumed that output and em- some techniques more than others. In partic-
ployment grow at the same rate if a single ular, research, development and use of tech-
technique is chosen and adhered to over time. niques in the developed countries are vir-
In practice technical possibilities available tually confined to techniques of high, and
increase over time. Generally technical prog- increasing, capital-intensity. Thus for these
ress takes a form which involves increasing techniques labour productivity, and often cap-
labour productivity, so that the rate of ital productivity as well, may rise over time,
growth of employment is less than the rate of while the more labour-intensive techniques
growth of output. This phenomenon — output may be unaffected by technical progress. The
increasing faster than employment — has labour-intensive techniques may therefore
been observed in many developing countries. become inferior over time and their use may
The precise form that technical progress involve a sacrifice of output as compared with
takes will affect the terms of any conflict be- the later techniques.
tween output and employment. However, for labour-intensive techniques
If technical progress is disembodied, af- various improvements are likely to result
fecting existing capital equipment as much as from their widespread use, including a fall in
new and therefore unrelated to the rate of in- their cost simply as a result of economies of
vestment, and if it increases the labour pro- scale in their production. Labour-intensive
ductivity associated with techniques of vary- techniques are also often easier to produce in
ing capital intensity to the same extent, then the developing countries, because they are
the technique which maximizes the rate of often of simpler design and more (in number)
growth of output will be the same as the tech- are required in a particular country so that
nique which maximizes the rate of growth of some of the economies of scale may be ex-
employment, though the latter will be lower ploited. Current relative costs and efficiency
than the former. If technical progress is em- of different techniques may therefore fail to
bodied, affecting only new investment, the reflect potential relative costs after technical
greater the rate of investment, the greater the progress through use has been realized. They
increase in labour productivity. Hence, for may also fail to take into account the differ-
any increase in growth rate, resulting from ing possibilities for local production and re-
an increase in the investment ratio, there will pair of the different techniques. This means
be a less than proportionate increase in the that current possibilities may understate the
rate of growth of employment. Similarly, if likely implications for output of labour-inten-
labour productivity is positively related to the sive techniques; the conflict between employ-
scale of production, measures which speed up ment and output may therefore be less in
the growth of output (whether they be in- reality than at first appears.
vestment or other means) will increase the
growth of employment less than proportion-
ately. Relationships of this type have been THE PRODUCT-MIX
observed for developing countries, though the
relationship appears less strong than for de- So far we have assumed the composition of
veloped countries (Oherlihy, 1970). Since the consumption goods to be determined and
growth in output remains bigger than the as- have varied only the techniques of producing
sociated growth in labour productivity, the them and the allocation of investment be-
technique which increases the growth rate of tween sectors. If different consumption goods
output will also increase the growth of em- require different proportions of labour and
ployment, but the gains in terms of growth of capital, we can raise the level of employment
 205 EMPLOYMENT STRATEGIES

without varying the techniques of producing ing, housing, shelter, etc.), there is consider-
any product by enlarging the share of labour- able scope for substitution towards more la-
intensive products at the expense of capital- bour-intensive products for the fulfilment of
intensive products. If there are opportunities each need. The possibilities of concentrating
for international trade on favourable terms, more on labour-intensive products to fulfil
this is an obvious solution. If, however, a each need may therefore extend the scope for
changing composition involves changing the using the product-mix to increase employ-
products consumed at home, the question is ment opportunities.
whether, with a proper system of weighting,
losses in consumers' welfare would arise. If
the labour-intensive products are also those
largely demanded by the poor, we have al- References
ready seen that a fall in output may be an
Amin, S. (1969), "Levels of remuneration,
optical illusion and that the weights derived factor proportions, and income differen-
from a more equal income distribution might tials with special reference to developing
show a rise. There may also be external countries," in A. Smith (ed.), Wage Pol-
diseconomies of consumption or buying as a icy Issues in Economic Development,
result of created wants or of habits. If a prod- Macmillan.
uct iswanted (a) because others buy it or (b)
Bhalla, A. S. (1964), "Investment allocation
because it was bought in the past or (c) wants and technological choice — a case of cotton
are created through advertising, and if these
spinning techniques," Economic Journal,
features are peculiar to the capital-intensive Vol. 4, pp. 611-22.
products, its elimination may lead to smaller , (1965), "Choosing techniques: hand
welfare losses (in cases (b) and (c) after a pounding vs. machine-milling of rice: an
time) than the expenditure values would in- Indian case," Oxford Economic Papers,
dicate or it may lead to welfare gains. Vol. 17, pp. 147-57.
The scope for changing the consumers' Boon, G. K. (1964), Economic Choice of
product-mix in a labour-intensive direction is Human and Physical Factors in Produc-
generally considered somewhat limited, apart tion, North Holland.
from possibilities of international trade, by Kaldor, N. (1965), in R. Robinson (ed.), In-
the need for a reasonable balance in the com- dustrialization inDeveloping Countries,
position ofdemand. We cannot expect people Cambridge University Press.
to consume all food and no clothes for ex- Little, I. M. D., and Mirrlees, J. (1968), in
ample, or to have more haircuts at the ex- Manual of Industrial Project Analaysis in
pense of bicycles. But the conclusions drawn Developing Countries, Vol. 2, p. 42.
from this, in terms of the narrow scope for
Oherlihy, C. St. J. (1970), "Wages and em-
product substitution, arise partly from a mis- ployment," Meeting of Directors of Devel-
taken definition of product. Any given need opment Training and Research Institutes,
may be fulfilled by a number of different ILO (mimeo).
products: nylon or cotton shirts fulfil the need Sen, A. K. (1968), Choice of Techniques,
for clothing, wooden houses, mud huts, rein- Blackwell.
forced concrete multi-storey buildings fulfil Strassmann, W. P. (1968), Technological
the need for shelter. While maintaining a Change and Economic Development, Cor-
reasonable balance in terms of needs (cloth- nell University Press.
 206 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

EXHIBIT III. 3. Gross Domestic Investment per New Labor Market Entrant, 1975
Increase in
Gross Domestic Percent Rate of Labor Gross Domestic
Gross Domestic Investment Labor Force Force Investment per
Investment/ Growth Labor Force
(millions of 1974-75 Entrant
GDP (percent) U.S. dollars) (thousands)
1974-75
10.5 278 2.0 234
Ethiopia 402
Bangladesh 6.1 563 1.6 1,190
1.6 1,400
Burma 10.5 376 203
Tanzania 21.3 543 2.4 152 1,850
479
Pakistan 15.6 2.5 3,570
1,718
745 199 3,590
Zaire 33.9 1.9
Sri Lanka 16.4 434 2.5 116 3,740
India 22.0 18,352 2.0 3.970
3,740
2.6
Kenya 19.6 617 132
4,619
Indonesia 19.0 2.2 4,680
Thailand 26,8 6,756 1,003
76
536 6,740
3,837 2.9 7,160
Ghana 10.6 556 2.1
Korea 26.8 3.0 7,320
5,121 562 9,110
El Salvador 20.1 370 3.3 40
Columbia 17.3 9,250
15.5 2,325 3.2 237
Senegal 316 1.7 9,810
Cameroon 21.4 467 1.4 32 9,900
10,850
410
Philippines 31.1 2.7 11,970
27.3 4,908
Egypt 2.5
125
257 13,210
Morocco 23.6 3,394 2.8 15,000
28.9 1,875 2.1
Nigeria 7,766 507 15,320
Peru 18.6 3.0 132 19,820
Turkey 22.7 2,617 1.8
8,049 302 26,650
Brazil 25.1 2.9 48
966 32,300
31,194
Syria 31.4 2.6 34,460
Mexico 28.8 1,654 3.3 550
22,764 41,390
Iran 29.8 16,266 2.6 237 68,630
Algeria 57.4 2.8 76,560
7,809 102
Argentina 21.4 10,589 1.2 118
1.3 98,740
Yugoslavia 33.0 125
10,214 81,720
United States 15.4 231,800 149,070
1.7 796
Japan 32.5 158,457 1.4 1,555 199,070
France 22.5 75,588 1.3
287 263,370

Source: Compiled by T. King and quoted in Nancy Birdsall, Population and Poverty in the Developing World, draft, February
15, 1980, p. 40.
 207 EMPLOYMENT STRATEGIES

III.F.2. Technology and

Employment*

We shall use the term local technology sector with any rate of employment. Full employ-
(or L-sector) to describe those techniques in ment of resources would occur with price
the old-style traditional sector and the new- line, PP', given limited investment resources
style informal sector; and foreign technology Oi, and labour supply OL. But distortions of
sector (or F-sector) to describe those tech- the kind mentioned result in the actual price
niques in the sector using technology devel- line of Diy. Given limited investment re-
oped recently in the advanced countries. sources, Oi, unemployment uL results. The
We may discuss three types of analysis solution to the problem is clear: "Let the en-
that have been put forward to explain the dowment speak," as one author has put it.
employment problems of underdeveloped Factor prices should be changed, or factor
countries: distorted factor prices, the factor use subsidised/taxed to bring them into line
proportions problem, and inappropriate with "undistorted" full employment prices.
technologies. A fundamental objection to this approach
is the assumption made about the nature of
technological choice — both as to the range of
techniques assumed, and to the critical deter-
"DISTORTED" FACTOR PRICES minants ofchoice. The discussion of techno-
These form the central pivot of much neo- logical choice [see selection III.B.2] demon-
classical analysis of unemployment in under- strated the unrealism, in most cases, of the
developed countries. The argument is that neo-classical assumption of a wide range of
there exists some "natural" or "undistorted" available techniques of varying labour and
set of factor prices — wage rates, interest capital intensity to produce any given prod-
rates, and prices of investment goods — which uct, while it also suggested that the relative
would reflect the opportunity cost of the var- price of capital and labour was only one
ious resources, and which would ensure full among a number of determinants of the
employment of all resources. Un- or under- choice of technique.
employment isthen attributed to various A second objection is the assumption that
market imperfections distorting the factor wages paid may be freely varied without af-
prices and thus preventing full employment fecting labour efficiency necessary for its
of resources. For example, trade unions and
government wage regulation interfere in the
labour market, pushing up wages above their
natural rate; subsidies on investment, de-
signed to increase the rate of investment,
artificially cheapen it. Similarly, overvalued
exchange rates maintain real wages above
the natural rate, and the costs of imported in-
vestment goods below the natural rate. Fig-
ure 1 illustrates the situation. There is as-
sumed to be an infinite choice of techniques
of varying labour and investment intensity,
allowing any investment rate to be associated
u D' L
Labour force P'
•Frances Stewart, Technology and Underdevelop-
ment, 2nd edition, 1978, pp. 46-50, by permission of
Macmillan, London and Basingstoke. FIGURE 1.
 208 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

working. The labour force has to be well fed, the problem of poverty as well as unemploy-
healthy, punctual and clothed. The sort of ment. However, trends over time . . . com-
wages required to get "undistorted" factor bined with the character of F-technology . . .
prices are often well below the sort of level rule out the possibility of total F-sector
necessary — and are therefore inconsistent absorption for many underdeveloped econ-
with modern technology. Here we are pri- omies.
marily concerned with the analysis of the em-
ployment situation on which this type of ap- THE FACTOR-PROPORTIONS
proach is based. It is assumed — normally PROBLEM
implicitly — that only the modern sector
counts in terms of employment, and the rest The starting position of this analysis of the
of the population are more or less equivalent employment situation is that the neo-classi-
to unemployed. Thus the factor price distor- cal assumption of a range of technological
tions referred to as the central problem are possibilities is incorrect. Assuming complete
the factor prices ruling in the modern sector. technological fixity, with a unique ratio of in-
Outside the modern sector there is no ques- vestment to employment, then changing fac-
tion of the factor endowment not speaking. tor prices will have no effect on employment.
Incomes from work are very low, interest Technological possibilities are not then rep-
rates high. In the traditional sector, activities resented bya smooth curved isoquant as in
are as labour-intensive as the technology and Figure 1, but by rigid proportions shown by
human willingness to work permits. But the a straight ray from the origin as in Figure 2.
full employment that is aimed for is the em- For any given amount of investment, say,
ployment ofeveryone in the F-sector. By ig- OK, a unique amount of employment, Ou,
noring the rest of the economy the approach will be generated, irrespective of factor
may be counterproductive. Reductions in prices. Thus total employment is determined
wage rates in the F-sector may increase by the rate of investment. Unemployment re-
employment there. But it may also reduce sults if,given the ratio of investment to em-
employment opportunities in the traditional/ ployment, the rate of investment falls below
informal sectors by a greater extent, since the that level necessary to ensure full employ-
lower wage rates will enable the F-sector to ment. In the diagram with investment OK,
undercut the L-sector. Any switch in produc- unemployment uL results. Since technology
tion from the L-sector to the F-sector tends in the modern sector is developed in advanced
to reduce total employment opportunities be- countries for their factor proportions ... it is
cause of the greater labour-intensity activity likely that it will be too investment-intensive
in the L-sector. Expansion of the F-sector for full employment in underdeveloped
may also increase the rate of open unemploy- countries.
ment through increasing the chance of ac- In so far as there is more than one good
quiring amodern sector job and hence the in- and/or more than one process, factor propor-
centive to switch jobs. This may be offset by tions may not be rigidly determined, but free
a sufficient reduction in wage-rates, reducing to vary according to the ratio in which differ-
the reward for securing such a job, But if the ent goods are produced (or processes used),
change were achieved by subsidies on the em- and the factor proportions associated with
ployment of labour, rather than changes in the alternative goods/processes. However,
wage-rates, this offsetting would not occur. the range of production possibilities may still
By concentrating entirely on the modern be too narrow to allow full employment. This
sector, the approach may thus produce coun- is shown in Figure 2b where, if the more la-
terproductive solutions. Only if the modern bour-intensive techniques are adopted, un-
sector could really be expanded to absorb the employment ML none the less results.
total labour force does it offer the possibility The rigid factor proportions theory sug-
of a solution rather than an accentuation of gests two possible solutions to the problem:
 209 EMPLOYMENT STRATEGIES

OL is total labour supply

FIGURE 2.

one, expansion of investible resources to en- sees the solution to the problem as improving
compass as many as possible of the unem- opportunities there rather than the full-scale
ployed; two, modification of the modern tech- takeover by the F-sector. While measures of
nology ina labour-intensive direction so as to underemployment are highly relevant to the
fit with the availability of resources in under- first two approaches, because they show the
developed countries. extent of the necessary expansion in the F-
The rigid factor-proportions theory is sim- sector, they are irrelevant to the alternative
ilar to the "distorted" prices in concentrating technology approach. What is relevant to this
entirely on the modern sector, and assuming approach is an examination of incomes, work
that anyone outside the sector is virtually and the technology in both sectors, and the
unemployed. As with the distorted prices the- interrelationship between the two, without
ory, the suggested solution may actually in- any attempt to define people as under-
crease open unemployment as the F-sector employed.
expands and the chance of getting a job in- The different approaches may be illus-
creases. Again, the strategy is only feasible if trated byreturning to the previous represen-
the investment rate can be increased and/or tation. InFigure 3, the shaded area shows the
the technology sufficiently adapted to encom- direction of change suggested by the ap-
pass the whole of the labour force. proaches. Figure 3a shows "distorted factor
prices" and rigid technology solutions. Both
ALTERNATIVE TECHNOLOGIES involve an expansion of employment in the F-
sector (in the distorted factor price case to be
Whereas the two previous approaches con- achieved by modifying factor prices, in the
centrate entirely on the F-sector, aiming to rigid factor proportions by modifying F-tech-
extend it to cover the whole economy, the al- nology and increasing investment), and a re-
ternative technology approach is concerned duction inaverage productivity in the F-sec-
to improve the productivity and work oppor- tor (in the distorted factor prices case
tunities ofthe L-sector, as well as modifying through inducing a changed selection of tech-
technology in the F-sector. The "distorted" niques inthe sector as a result of the modified
factor price approach and the fixed factor factor prices, in the rigid factor proportions
proportions approach both assume that un- through new technological developments in
derdevelopment prevails in the L-sector and the F-sector). In both cases, open unemploy-
can only be eliminated by expanding the F- ment remains more or less the same since the
sector. In contrast, the alternative technology effect of somewhat reduced average rewards
approach views activities in the L-sector as is offset by increased F-employment oppor-
just as legitimate as those in the F-sector, and tunities. Figure 3b shows the alternative
 210 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

lology Alternative technology

'Distorted factor prices1
Rigid technology
(a) (b)

FIGURE 3. Alternative technological strategies.

technology solution. It is two-pronged, in- ductivity inthe L-sector. The gap between
volving an increase in employment in the F- the sectors is reduced and the consequence is
sector and a reduction in its labour produc- that open unemployment may be reduced
tivity, like the other approaches, to be since there is a reduced incentive to switch
achieved by modification of technology in between the sectors, with the greater produc-
that sector; but it also involves raising pro- tivity ofthe L-sector.

III.F.3. Policy Implications— Note

Although this chapter has outlined some di- while a shortage of labor exists in another
mensions of the employment problem and sector (sugar cane cutting). Labor may de-
has suggested broad strategies of employ- cide to forgo low-income employment oppor-
ment creation, we are left with numerous tunities and queue for the few high-income
specific policy questions that need to be clar- jobs available in a segmented labor market.
ified. Itmight first be thought that the mal- Although market imperfections do exist
functioning oflabor markets is a major cause (minimum wage laws, trade unions), it is not
of the employment problem. But empirical plausible to argue that, if only these imper-
studies of labor market performance indicate fections were removed, wages would be
that labor markets actually work reasonably reduced and then unemployment and under-
well, with rational responses to incentives.1 employment would disappear. Policies other
Frequently unemployment is sector specific than the removal of labor distortions are
because of segmented labor markets. An ex- required.
cess supply of labor may appear in one sector Policies designed for increasing the effi-
(for example, in bauxite mining in Jamaica), ciency of labor markets are of limited signif-
icance compared with other major employ-
'A. Berry and R. H. Sabot, "Labor Market Perfor- ment policies that focus on raising the
mance in Developing Countries: A Survey," World demand for labor through greater output and
Development, November-December 1978, pp. 1 199—
1249; with extensive bibliography. See also Lyn higher output-elasticity of demand for labor,
Squire, "Labor Force, Employment, and Labor Mar-
while easing the supply side of the problem
kets in the Course of Development," World Bank through a slowing down in the hitherto fast
Staff Working Paper No. 336, June 1979. growth of the labor force. Not labor market
 21 1 EMPLOYMENT STRATEGIES

imbalance, but the entire set of development areas. Urban problems are in a fundamental
policies that affect the demand and supply sense rural problems: urban "pull" must be
sides of the labor problem should be of con- offset by lessening the "push" through rural
cern. Thus, all the remaining chapters will development.
concentrate on the formulation of a set of The modern sector must avoid producing
employment policies by relating to the mo- what can be produced in the rural sector —
bilization and allocation of investment re- e.g., village handicraft employment should
sources, agricultural development, industrial- not be displaced if this entails the wasteful
ization programs, and trade strategy. use of capital in the modern sector to produce
At this point, however, it may be useful to an output which could be produced equally
offer some general principles underlying pos- well by surplus labor. It is to be recalled that
sible policy options for the better utilization in Japan's case of successful development,
of labor. Their validity will then be explored both agriculture and village industry became
in greater detail in subsequent chapters. more labor-intensive. There may also be a
a. If urban unemployment is to be reduced, considerably greater scope for rural-based in-
policy measures must reduce the rural-urban dustry involving simple technology and the
drift. To this end, a reduction of urban-rural processing of agricultural materials.
real income differentials would be most help- Beyond this, however, a full-scale program
ful; but this is probably the most difficult ob- of rural development is needed to absorb and
jective to achieve. According to the Todaro retain large amounts of manpower. If the
model of labor migration, the larger the gap rural to urban migration is to be reduced, it
between urban and rural nominal wages, the is necessary to modify policies that have
higher must be the urban unemployment rate turned the terms of trade against the agri-
before migration in excess of job opportuni- cultural sector. Ceiling prices on foodstuffs,
ties ceases. As long as urban wages rise more export taxes or restrictions on primary prod-
rapidly than average rural incomes, rural- ucts, and tariff protection on industrial inputs
urban migration will continue in spite of and consumer goods have acted as disincen-
rising levels of urban unemployment. All pol- tives to agricultural producers while they
icies that would redress the imbalance be- have artificially increased the urban-rural
tween urban and rural income levels would differential.
therefore be desirable — urban wage re- Efforts should also be made to disperse to
straint, adjustment of minimum wage rates, the rural sector some of the amenities and
revision of the tax structure, a comprehensive public services now concentrated in urban
national income and wages policy. areas. Readier access to such services as pub-
A number of institutional and political lic utilities, health, education, and entertain-
considerations, however, militate against the ment in the rural areas may amount to an in-
efficacy of these policies, and it is not realistic crease inthe rural social wage, and diminish
to expect any strong downward pressure on the attractions of the city.
urban wages. An effective "wages policy" has Of greatest consequence will be the type of
proved difficult in the developed — let alone strategy pursued for developing the agricul-
newly developing — countries. tural sector. As elaborated in Chapter VII,
b. If it is difficult to institute a "wages pol- the most important factor influencing a de-
icy" that would increase urban employment, veloping country's ability to absorb a growing
it is all the more important to emphasize the labor force into productive employment is
"supplysectorside" cannot
of the absorb
problem. whether a labor-using, capital-saving type of
urban the When
inflow the
of approach to agricultural development is fol-
labor from the rural sector, special consider- lowed (as in Japan and Taiwan). For most
ation must be given to policies that will re- developing countries, the employment poten-
move the causes of the rural "supply push" tial in rural modernization can be greater
and help contain the labor force in rural than that of the modern urban sector — pro-
 212 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

vided that the countries avoid implicit taxa- the capital/output ratio. A more appropriate
tion of agriculture and "unduly labor displac- technology would in effect retain the essential
ing" measures in agriculture. quality of the tool element in physical equip-
c. If the previous strategy of industrializa- ment without the superfluous labor-saving
tion via import substitution has resulted in appendages of the advanced technology of in-
dustrial countries.
"urban bias" — that is, distortions that favor
the urban, import-substituting, modern sec- To lessen the bias toward relatively capi-
tor at the expense of the rural sector,2 then in tal-intensive techniques, it is again necessary
the future the promotion of nontraditional to stress the removal of factor price distor-
exports may allow a strategy of industriali- tions. Given that there is a positive elasticity
zation via export substitution that creates of substitution of labor against capital,3 it
more employment, among other advantages. would become less profitable to use capital-
Chapter VIII discusses various policies — no- intensive technologies if interest rates were
tably those connected with trade policy and increased, foreign exchange became more ex-
foreign investment — that are needed to make pensive interms of home currency, and the
export substitution effective. increases in urban wages were restrained.
The distortions in the price structure also e. As long as labor is induced to migrate
create divergences between domestic and in- from the rural sector and the manufacturing
ternational prices that inhibit the country's sector cannot absorb labor in sufficient quan-
exports. To the extent that the comparative tities, itwill be necessary for labor to seek
advantage of the country lies in labor-inten- employment in the tertiary sector. Labor has
sive commodities, the employment-intensity done so in many of the LDCs, and employ-
of trade can be raised by "getting prices ment in services and commerce has actually
right" and by establishing an efficient com- risen more rapidly than in other sectors.
modity composition of exports. From the standpoint of providing an em-
d. More effort is also needed to devise a ployment outlet, it is therefore advisable not
range of technological choices that are supe- to promote too rapid an increase in efficiency
rior to the country's indigenous traditional in employment practices in the service sector.
technology but are not as advanced and As remarked by Professor Galenson,4 the
labor-saving as are the modern machines and pushcarts should not be too readily replaced
equipment that have been imported from ad- by the supermarket; the bicycles by the
vanced industrial countries. As discussed in trucks; a casual but large labor force by a
Chapter V, the transfer of "appropriate" permanent and stable but smaller labor
technology has important consequences for force. The inefficient use of labor in the ter-
employment in both the urban industrial sec- tiary sector will not, of course, have the un-
tor and the rural agricultural sector. desirable cost effects that would occur if this
The choice of a more labor-intensive pro- were done in the import-replacement or ex-
duction technique may, of course, conflict port sectors. In the production of nontradable
with other investment criteria — in particular, commodities it may therefore be important to
the maximum absorption of labor may yield be unimportant about seeking the least cost
only a low return per unit of capital and not combination of factors when this would dis-
maximize the future rate of growth in output.
place labor.
The crucial consideration is the emphasis
on devising new technology that is "capital- 3For a careful empirical study that suggests that
stretching" in an efficient way — that is, the considerable substitution possibilities exist in a num-
labor-intensive equipment should raise ber of manufacturing industries, see Howard Pack,
the labor/capital ratio without also raising "The Employment-Output Trade-Off in LDCs— A
Microeconomic Approach," Oxford Economic Pa-
2Michael Lipton, "Urban Bias and Rural Plan- pers, November 1974.
ning," in Paul Streeten and Michael Lipton (eds.), 4W. Galenson, "Economic Development and the
The Crisis of Indian Planning, London, 1968, pp. 89-
95. Sectoral Expansion of Employment," International
Labour Review, January-June 1963, pp. 505-19.
 213 EMPLOYMENT STRATEGIES

Emphasizing that "Asian countries will be are to be avoided, the government will itself
forced to develop the labor-intensive sectors have to provide employment opportunities.
if jobs are to be created for the increasing On public work projects, or elsewhere in the
waves of youngsters coming into the labor public sector, labor may then be receiving a
market," Oshima has stated that wage greater than its marginal product.
the nonagricultural labor-intensive sector is very Thus, there comes about in the public sector
large, perhaps engaging two-thirds to three- a situation similar to the disguised unemploy-
fourths of the nonagricultural labor force. It is a ment that exists in agriculture (in the sense
sector that provides employment using the least of the real wage being greater than the mar-
amount of capital, in terms of capital efficiency ginal product of labor). But if it is decided
uses the less scarce type of capital and saving, re- that labor should not remain unnecessarily
quires material inputs which are domestically pro- idle, then in a labor surplus economy the
duced, utilizes labor not appropriate for modern labor will have to be used even to the point
industries, and produces goods of the traditional
where its social marginal product becomes
type, consumed by lower-income families located zero.
in various parts of the country instead of being
concentrated in the cities. It is an excellent com- g. Finally, more attention must be given to
plement to modern industrialization for underde- the "supply side" of the problem in terms of
veloped countries where modern types of inputs
population control policy and the "outputs"
and factors are scarce — whether these be capital of the country's educational system. Growth
and savings, skills, infrastructure, inputs, etc.5 in the labor force is a derivative of the pop-
An illustrative OECD study calculated ulation growth rate and the labor participa-
that on average in the developing countries tion rate. But since there is about a 1 5-year
the manufacturing sector employs 20 percent lag between a decline in the birth rate and a
of the labor force, and the unemployment decline in the labor force entry rate, any de-
celeration ofpopulation growth can only have
rate and underemployment rate together av-
erage 25 percent. The increase in labor pro- long-run effects and is not a relevant instru-
ductivity isassumed to be 2.5 percent a year. ment for short-term policy. Investment in
In order to absorb an increase in the labor human capital may, however, influence em-
ployment more readily. In this connection,
force growing at 3 percent a year (a conser- some relevant points will be noted in Chapter
vatively low estimate), industrial production
would have to increase at the exceptional rate IX on human-resource development.
From even this summary listing of policy
of 18 percent a year — a rate beyond the
achievement of any LDC. To eradicate implications, it should be apparent that em-
within a decade the existing rural and urban ployment policies make sense only within the
context of an overall development strategy.
unemployment and underemployment it
would have to increase by 30 to 35 percent a An integrated set of employment policies has
to provide for the mobilization and allocation
year. Not surprisingly, the report concludes:
of domestic and foreign resources so as to
"Thus eradication of general underemploy- achieve more labor-intensive output growth,
ment through the development of industrial
employment is a practical impossibility in the investment in human capital, a more labor-
intensive industrialization program, an agri-
medium term."6
f. If open unemployment persists and its cultural program that is itself labor-absorb-
unfavorable social and political repercussions ing and also contributory to industrialization,
access to more appropriate technology, and
the slowing down of population growth.
5Harry T. Oshima, "Labor-Force 'Explosion' and The range of policies extends from mea-
The Labor-intensive Sector in Asian Growth," Eco- sures that can be instituted immediately to
nomic Development and Cultural Change, January
1971, p. 178. other measures that can take effect only in
the long run. With respect to employment in
6Quoted in Jonathan Power, "Why Going Back to
the Land Is the Only Hope for the Third World," The
Africa, for example, Todaro has proposed the
London Times, October 18, 1974, p. 18. following short-run policies: (1) the elimina-
 214 SURPLUS LABOR AND DUALISTIC DEVELOPMENT

tion of present factor-price distortions, (2) lishment of a comprehensive incomes policy,

the establishment of a "dual" wage structure (2) the acceleration of rates of industrial and
through the use of wage subsidies in some urban output growth, and (3) the intensifi-
combination with a policy of wage restraint, cation of efforts to stimulate agricultural and
(3) the immediate creation of new types of rural development. Long-run policies would
employment opportunities through various include the establishment and maintenance
voluntary agreements, and (4) the restriction of an effective program to limit rapid popu-
of excess emigration through the use of moral lation growth, and efforts to plan for the
exhortation to return to the land, the adop- eventual development of domestic labor-in-
tion of forced controls on the movement of tensive capital goods industries.7
people, or the establishment of urban labor 7

* • the* process . . .Mlc^el P; Todar°' "ln™me Expectation s Rural-

Inter-
exchangesi to regulate
of jobly u placement tm_ and control
.* The
* Urban Migration and Emloyment in Africa,"
most important me- national Labour Review, November 1971, pp. 396,
dium-term strategies include (1) the estab- 402,407.
Comment
Some three decades after Sir Arthur Lewis formulated his dual sector model, development
economists recognize that the nature and causes of the employment problem have changed,
and so too must strategies and policies to provide additional employment opportunities and
more effective utilization of labor resources. Instead of concentrating on an expansion of mod-
ern-sector jobs, much more attention is being given to improvement in productivity and the
utilization of labor outside the modern large-scale sector. But, as emphasized in the foregoing
selections, there is no unique "employment problem" in LDCs; instead, there is a syndrome
of problems that require for their solution the coordinated pursuit of a number of policies
affecting prices, the level of investment, technological change, the foreign trade sector, the
agricultural sector, the informal sector, government expenditure, education, and training. All
the chapters in this book therefore have some bearing on the employment problem, and they
should be read with their implications for employment policy in mind.
Employment policies are considered in more detail in the following works: Paul Bairoch,
Urban Unemployment in Developing Countries (1976); Mark Blaug, Education and the Em-
ployment Problem in Developing Countries (1976); H. Bruton, "Economic Development and
Labor Use: A Review" World Development (December 1973); H. Chenery et al., Redistri-
bution with Growth (1974); ILO series of studies on employment strategies for Colombia
(1973), Ceylon (1974), Kenya (1977), Iran (1973), Philippines (1974), Sudan (1978); J.
Mouly and E. Costa, Employment Policies in Developing Countries (1974); A. K. Sen, Em-
ployment, Technology and Development (1975); Lyn Squire, Employment Policy and Devel-
oping Countries (1982); Guy Standing, Labour Force Participation and Development (1978);
Frances Stewart (ed.), Work, Income, and Inequality: Payments Systems in the Third World
(1982); Frances Stewart, "Employment Policies," in Robert Cassen and Margaret Wolfson
(eds.), Planning for Growing Populations (1978).
CHAPTER IV

Mobilizing
Domestic
Resources

Even though labor may be abundant, the output of an LDC remains limited by a shortage of
capital. It is widely recognized that the LDCs must make additional efforts to mobilize and
achieve effective use of their internal resources. The mobilization of domestic resources —
along with the mobilization of external resources (to be discussed in the next chapter) — re-
quires policies to facilitate the process of capital accumulation. Many economists emphasize
capital accumulation as the major factor governing the rate of development. Professor Ros-
tow, for example, specifies a rise in the rate of productive investment to over 10 percent of
national income as a necessary requirement for a country's take-off.
Similarly, in presenting his model of a dual economy, Sir Arthur Lewis contends that
The central problem in the theory of economic development is to understand the process by which a
community which was previously saving and investing 4 or 5 per cent of its national income or less,
converts itself into an economy where voluntary saving is running at about 1 2 to 15 per cent of national
income or more. This is the central problem because the central fact of economic development is rapid
capital accumulation (including knowledge and skills with capital).1
The discussion of technological dualism (III.B.2) also implied that development requires
primarily large amounts of capital investment, especially in the underdeveloped sector.
Certainly there has been no tendency among development economists to underestimate the
importance of capital. On the contrary, it has been stressed so much that a reaction has set
in, and there is a strong counterview that the role of capital has received excessive attention
to the neglect of other essential components of the development process. The materials in
section IV. A present opposing views on the legitimacy of emphasizing capital as the key var-
iable determining the rate of development. Since capital-output ratios have figured promi-

1 W. Arthur Lewis, "Economic Development with Unlimited Supplies of Labour," The Manchester School,
May 1954.

215
 216 MOBILIZING DOMESTIC RESOURCES

nently in the discussion of capital accumulation, the Note on capital output ratios clarifies
various interpretations of the capital-output ratio, and at the same time levies a number of
criticisms that restrict the use of a capital-output ratio in practice (selection IV.A.2).
To the extent that an increase in the rate of investment is necessary or desired, a developing
country must mobilize the required savings. The Note in section IV.B focuses on this issue by
outlining the various sources of capital formation. In examining these sources, we should as-
sess each source of capital formation from the wider standpoint of how its contribution to the
flow of resources for developmental purposes can be intensified.
Of increasing interest are the problems of financial development — the manner in which
financial institutions and financial policies may help overcome the shortage of capital and
influence a country's pattern of development. Economic theory has been generally conducted
in "real" terms, such as national product in physical units, production functions, and capital-
output ratios. And yet, for an understanding of the process of development we must consider
how the financial superstructure and the real infrastructure interact, and what are the effects
of such interaction on development.
The selections in section IV.C ask some basic questions about the contribution of financial
development to economic development. What are the alternative techniques available to each
country for mobilizing its economic surplus and channeling capital flows? What are the re-
lationships between finance and the rate and pattern of development? These general questions
should help to place in perspective the empirical studies of particular financial institutions or
the financial development of individual countries.
It is especially important to consider the need for more efficient capital markets to improve
the quantity and quality of capital formation. "Financial repression" is characteristic of in-
efficient capital markets and the consequence is the substitution of direct, microeconomic
measures that distort prices and the allocation and mobilization of resources.
Although much of the discussion in this chapter is designed to indicate various ways of
mobilizing resources for developmental expenditures without causing inflation, many coun-
tries have in practice desired a higher rate of investment than could be maintained by non-
inflationary sources of finance. Recourse to the substitute method of monetary expansion and
credit creation has become common. Increased spending of an inflationary sort is therefore
an important issue of development policy. On this issue, however, there are marked differences
of opinion regarding the causes and consequences of inflation.
We try in section IV. D to sort out these differences and reach some assessment of the effects
of inflation in LDCs. To do this, we examine in analytical terms the main forces and conse-
quences of inflation. Why are the LDCs so prone to inflation? How might inflation be
stopped? Would it do more harm than good to stop inflation in a developing country?
Against the background of this general analysis, we also give specific attention to the prob-
lem of inflation in Latin America. The selections in this section center on the issues between
the traditional "monetarist" view of inflation and the contrasting "structuralist" view. Dif-
ferent interpretations of the causes of inflation lead naturally to divergent conclusions on pol-
icies for curbing inflation. While the orthodox prescription has been to restrain demand by
the exercise of monetary and fiscal discipline, this policy is criticized by "structuralists" as
being concerned with only the "propagating" factors of inflation or the "symptoms" instead
of the underlying real structural causes. In contrast, the structural position looks to the supply
side of the problem and stresses the need for social and economic reforms to correct basic
structural imbalances.
There is no denying that underdeveloped countries are especially prone to inflationary pres-
sures, and that the policies available to an underdeveloped country for effectively controlling
inflation are more limited than those in an advanced country. What remains controversial,
however, is whether these inflationary pressures encourage or inhibit development. This issue
 217 MOBILIZING DOMESTIC RESOURCES

is discussed in section IV.D with arguments presented both in favor of and against the method
of development through inflation.
Besides the usual points advanced in favor of inflation, consideration is given to the follow-
ing special contentions: inflation permits the employment of underemployed workers; mone-
tary or credit expansion is necessary to allow the "development authorities" to bid resources
away from consumption; in the early stages of inflation, the "money illusion" may induce
factors of production to work more intensively; the period of inflation may be short, since it
will increase investment which, in turn, will expand total output, and a large portion of the
increment in output may then be saved and taxed to offset the rise in investment; and none
of the alternative methods of financing a rise in investment is any less free of hardships.
As for the case against inflation, the following arguments are especially persuasive: inflation
has harmful effects on the efficient allocation of resources, being particularly detrimental in
creating distortions in investment patterns; balance of payments pressures result from the
adverse impact on exports, the spillover into imports, and the discouragement of foreign in-
vestment; the volume of resources available for domestic investment may actually fall as vol-
untary savings decline and incentives are diminished; and the government lacks the power to
constrain the inflation and prevent the pressures from becoming progressively severe.
 IV.A. THE ROLE OF CAPITAL
IV.A.1. Capital Accumulation
and Development*

The general rate of development is always plan for development, it must be informed of
limited by shortage of productive factors. If the quantitative aspects of savings and in-
any one scarce factor associated with under- vestment, and their effects on production and
development should be singled out, it would consumption.
be capital. The final goal of development pro- These quantitative aspects are of crucial
gramming is,therefore, to find the best way importance in determining the most desirable
of breaking the vicious circle between capital rate of development. It is important, for one
shortage and under-development and to de- thing, particularly when population is grow-
sign the most efficient and optimum rate of ing rapidly, to estimate the rate of develop-
capital accumulation. ment that would be needed to bring about an
It would be an over-simplification, of improvement in per capita income or a high
course, to regard economic development as a rate of employment for the growing work
matter of capital accumulation alone. Other force. Another element which may play a
things are needed in addition, such as entre- role in estimating a minimum rate of devel-
preneurship and training of workers and pub- opment isthe necessity to give a certain min-
lic administrators. Yet these are seldom pos- imum size to some projects in order that they
sible without some increase in the stock of are at all economically sound. In some indus-
capital. Therefore capital accumulation may tries where so-called "indivisibilities" play a
very well be regarded as the core process by role, there are such minimum sizes of proj-
which all other aspects of growth are made ects. For the country as a whole, this may
possible. mean that only a "big push," as it has been
Capital increases by investment, and more called, can really help to start the process of
investment necessitates more savings or for- development. Although this may produce re-
eign assistance. Foreign assistance, if not in sults which appear ambitious in the light of
the form of grants, means some burden in the current efforts, it provides a fair indication of
future. The extent to which foreign loans can the tasks involved in the planning effort.
be serviced and repaid will ultimately depend Whatever the initial approach, there are
on what can be saved at home in the future. some useful concepts which should be borne
Domestic savings are, therefore, the more in mind in planning the rate of development.
reliable source of investment to break the These concepts may conveniently be de-
vicious circle of poverty and under-develop- scribed interms of investment. There is, first,
ment. But domestic savings can be increased the concept of a minimum rate of investment,
only by a sacrifice in consumption which has which measures the rate needed to prevent
to be compared with the future increases in per capita income from falling in the face of
consumption it promises. Investment, more- population growth. A rate of investment
over, yields different results, depending on somewhat above this minimum is the lowest
the industries in which it is made. In order, target at which any plan should aim, even
therefore, for the government of an under-de- though this may involve a heavy effort when
veloped country to design an appropriate population is growing rapidly. For some
countries this may be a rate that can be easily
*From United Nations, ECAFE, Programming attained on the basis of an effort which does
Techniques for Economic Development. Report of
the First Group of Experts on Programming Tech-
not require any fundamental policy decisions,
niques, Bangkok, 1960, pp. 8-13. Reprinted by any changes in attitudes or behaviour pat-
permission. terns, or any improvements in techniques,

219
 220 MOBILIZING DOMESTIC RESOURCES

skills and methods of business or public ad- ment is first to estimate the amount of do-
ministration. For these countries, the mini- mestic savings and capital imports that could
mum rate of investment is clearly too low, be expected with no change in economic pol-
and is useful only for reference. icies; then to calculate the rate of growth that
A second concept of use in this context is this level of savings and investment would
that of a practical maximum rate of invest- provide; and finally to compare it with the de-
ment. In theory a maximum rate of invest- sired rate of growth. Usually, the ratio of sav-
ment may refer to a level of capital accu- ing to income is fairly stable over long pe-
mulation which involves saving and investing riods of time, and these saving-income ratios
at least all income above, say, a subsistence are lower in under-developed countries
level. Clearly, such a maximum is of no prac- (under 10 per cent) than in higher advanced
tical significance. A practicable maximum countries (about 15 per cent). Any empirical
may, therefore, be determined differently in estimation of this ratio must start with the
the light of the extent to which the population observation of the rates of savings experi-
would be willing to accept austerity now, so enced by the country in the recent past. The
as to enjoy a higher standard of living in the estimates may be based on data for incomes
future. The planner must form his best judg- and the savings of households, business and
ment as to what this practical maximum government, or domestic investment minus
would be. The rate just defined above is the capital imports. It may also be possible to
one to be determined by an evaluation of peo- base the estimates on the experiences of com-
ple's potential propensity to save. parable countries, keeping in mind the differ-
A third concept is that of the highest rate ences in income levels.
of investment consistent with absorptive ca- After estimating the current rate of sav-
pacity. Absorptive capacity depends on nat- ings, the crucial question will be what
ural resources, taxes, the labour supply, the amount of net national output may be ex-
level of labour, technical and managerial pected from the investment to be made on the
skills, entrepreneurial capacity, the efficiency basis of the estimated savings. A number of
studies have been made on the amount of
of public administration, the extent of "tech-
nology-mindedness" of the population, and so capital required to increase output by one
on. Such capacity sets a limit to the amount unit per annum in each sector of the economy
of efficient investment physically possible, and for a national economy as a whole. This
and although it can itself be increased amount is called the "capital-output ratio,"
through further investment, it does effec- or "capital coefficient." . . .
tively limit the rate of development possible, Although the capital-output ratio is usu-
particularly in the short run. Maximum ab- ally calculated as the "average" capital-out-
sorptive capacity may, of course, permit of a
higher rate of investment than that allowed put ratio, what really matters is the "mar-
ginal" or "incremental" capital-output ratio:
by the ability of the population to save. In we need information on the capital required
this case, it would be the role of an ideal in- to increase the national output. If we want to
ternational policy to fill the gap and to raise increase output by 20 and estimate the capi-
investment to the highest level consistent tal-output ratio as 4, then the required addi-
with absorptive capacity. On the other hand, tion to the capital stock, to be provided by
where absorptive capacity is below the prac- new investment, is 80. Evidently the figure 4
ticable rate of savings, both national and in- in this example stands for the "incremental
ternational policies should be directed to-
wards raising such capacity. These policies capital-output ratio."
Given estimates of the current rate of sav-
would then constitute the initial phase of a ings and the capital-output ratio, the rate of
long-term plan. economic growth, in terms of national out-
Thus, one of the logical ways to start plan- put, could be projected in the following way.
ning the general rate of economic develop- If the current level of national output is
 221 THE ROLE OF CAPITAL

TABLE 1 . Rate of Economic Growth in Terms of National Output

National Saving Increase in

Saving National Output
Output Ratio Investment
Output Ratio
Capital-
(1) (2) (3) (4) (5) (6)

1,000 0.06 60 60 4 15

15 0.06 0.015
Growth rate (G) =777 (2)
1,000 (5) 4

1,000, and the saving ratio is 0.06, domestic somewhat less than shown by the preceding
savings would be 60, which may be invested calculations, if the gestation period of the in-
to generate the increased national output. vestment envisaged is large. The calculation
[See Table 1]. With a capital-output ratio of above tacitly assumes that capital created by
4, this amount of savings and investment the investment in one period can be used pro-
could generate an increase in national output ductively inthe following period. If, however,
of 15, not more. An increase in national out- the gestation period of some investment pro-
put of, say, 20 will not be possible, because ject is longer than one year, say three years,
the amount of investment required for this then capital available for productive use will
purpose is 80, which exceeds the current sav- not increase before three years. At that time,
ings of 60. Hence, the increase in output war- the level of national income will be higher,
ranted bythe savings of 60 is 60 h- 4 = 15, and hence the rise in production, as a per-
which gives the growth rate of 1.5 per cent in centage oftotal national income, is somewhat
national output. The rate of growth in na- less. This means that the extension of the ges-
tional output can thus be calculated by divid- tation period has the same effect as the de-
ing the saving ratio by the capital-output cline in the saving ratio, or the increase in the
ratio. value of the capital-output ratio. If this is the
This method of projecting the future level case, then the rate of economic growth com-
of national output can be checked by other puted in the preceding way must be adjusted
ways of forecasting, e.g., extrapolation of downward. Needless to say, a lengthening of
past figures, If the projected national income this time lag has further adverse effects,
shows a lower growth rate than actual in- owing to the additional postponement of the
come did in the past, it may be that the sav- fruits of investment.
ing ratio has been underestimated or the cap- If such projections of current trends show
ital-output ratio overestimated. If the ratios no significant rise in the people's standard of
are right, a slowing down of economic growth living, there is a definite need to increase the
must be expected. Another check would be to growth rate of national output. Suppose that
divide the projected national output by the the expected population increase is 1.5 per
numbers in the active labour force, to obtain cent a year, the saving ratio 6 per cent, and
an index of the average productivity of labour the capital-output ratio 4. This will leave the
in the future. If this index does not rise as standard of living unchanged, and represents
much as the past trend, the estimates of pa- the minimum rate of investment as defined
rameters should again be reconsidered. If [above]. If the per capita national income
they are correct, inefficiency or unemploy- must increase by, say, 2 per cent a year, the
ment must be expected in the future, unless national income must increase by 1.5 + 2.0
measures are taken to prevent them. ■■ 3.5 per cent every year. This means that,
The rate of growth of an economy will be with the same capital-output ratio, the saving
 222 MOBILIZING DOMESTIC RESOURCES

ratio must be increased from 0.06 to 0.14, re- provements inliving standard must be low-
quiring a considerable adjustment in policy ered to what was called ... the practical
measures. If such a sudden rise in the saving maximum rate of investment.
ratio is difficult to achieve, the targets for im-

IV. A. 2. Criticisms of the Capital-

Output Ratio— Note

A capital-output ratio is frequently employed ratio is used. Net investment is estimated

to estimate the amount of investment needed over the plan-period; the increase in net out-
to achieve a certain rate of growth in income. put (or income) is estimated between the
This was done explicitly in ECAFE's calcu- year before the plan-period and the last year
lations of capital requirements (IV.A.l). A of the plan. All measurements are made at
definite causal relationship between the the same price level. The use of a marginal
growth of capital and of output, however, capital-output ratio in this fashion has been
cannot be as readily assumed as the forego- inspired to a large extent by the Harrod-
ing selection would imply. And it is mislead- Domar theory of growth, which relates a
ing to suppose that the whole of any increase country's rate of growth of income to its sav-
in output is due simply to capital accu- ings-income ratio and marginal capital-out-
mulation.
put ratio.1 The Harrod-Domar analysis, how-
Many conceptual difficulties and statistical ever, relates to an advanced economy, and it
pitfalls surround the derivation and use of seeks an answer to the question of how much
capital-output ratios. Even after it is decided national income would have to grow to in-
which of the several possible definitions of duce sufficient investment to maintain this
"capital" and "output" are best to use, and rate of growth in income. For a poor country,
some solution to the problem of valuation is the relevant problem is not that of sustaining
accepted, there still remain ambiguities. It is a certain rate of growth, but rather the prior
first necessary to distinguish between the av- task of initiating or generating a higher
erage and the marginal capital-output ratio. growth rate in the first place.
The average ratio is the value of the total Moreover, it is important to be clear
stock of capital divided by total annual in- whether all other productive factors that
come; the marginal — or incremental — ratio must cooperate with capital are also assumed
for the entire economy is the value of the ad- to increase when capital increases. In an ad-
dition to capital (net investment) divided by vanced economy an adequate supply of coop-
the addition to income (net national income). erant factors is likely to exist. The institu-
The marginal ratio need not, of course, equal
the average ratio, and even though any 'Evsey Domar, "Expansion and Employment,"
change in the average ratio may be expected American Economic Review, March 1947, pp. 34-5;
to be slow, the marginal ratio can vary a "The Problem of Capital Formation," American Eco-
great deal more. nomic Review, December 1948, pp. 777-94; "Eco-
nomic Growth: An Econometric Approach," Ameri-
In framing a development plan, it is com- can Economic Review, Papers and Proceedings, May
mon practice to calculate the amount of ad- 1952, pp. 479-95; R. F. Harrod, "An Essay in Dy-
ditional capital required to produce a one namic Theory," Economic Journal, March 1939, pp.
14-33; Towards a Dynamic Economics, London,
unit increase in annual output at the margin. 1948; W. J. Baumol, Economic Dynamics, New
For this purpose, a marginal capital-output York, 1951, Chap. 4.
 223 THE ROLE OF CAPITAL

tional, political, and social prerequisites for istic view of what can be accomplished by
development also already exist. When using capital accumulation alone, or an overpes-
the marginal capital-output ratio under these simistic view of how much investment is
conditions, it is reasonable to make a mutatis needed, we should guard against a too simple
mutandis assumption that the supply of other use of capital-output ratios.
necessary factors is forthcoming. But in a For the purpose of clearly recognizing the
poor country where the cooperant factors changing circumstances that may occur
tend to be in short supply, and the other pre- when additions to the capital stock are made,
requisites for development may not yet exist, it is helpful to distinguish between the "net
it is not legitimate to consider an increase in
marginal capital-output ratio" and the "ad-
capital as a sufficient condition for an expan- justed marginal capital-output ratio."2 The
sion in output. Even though investment may net ratio interprets the marginal capital-out-
be a necessary condition, an increase in out- put ratio as net of any changes in other fac-
put may still not be produced unless other tors; itconsiders the capital-output ratio with
conditions are also fulfilled along with the in- a ceteris paribus assumption — the supplies of
crease incapital supply. Since an expansion all other factors are held constant. The ad-
in output depends on many factors of which justed ratio, however, refers to what the cap-
capital formation is only one, greater output ital-output ratio would be if it were adjusted
may require changes in other factors along to a given specific increase in the supply of
with an increase in capital. Or output may other factors; it assumes that investment is
even increase independently of investment. accompanied by changes in other output-
Even if we accept the assumption that there yielding variables. For a given increment in
is a fixed relationship between capital and output, the net marginal capital-output ratio
output as determined by technical factors, it is higher than the adjusted marginal capital-
does not follow that we can infer from this output ratio. Capital requirements will there-
relationship that only capital is needed to in- fore be underestimated if they are initially
crease output. We must also consider explic- based on an adjusted marginal capital-output
itly the effect of other variables on output — ratio, but the other output-yielding factors do
for example, the supply of trained manpower, not actually accommodate themselves to the
entrepreneurship, institutional arrange- growth of capital as expected.
ments, attitudes, etc. To ignore these other In calculating capital requirements, a de-
variables or simply to assume that accom- velopment plan usually concentrates on an
modating changes occur, and then to attrib- overall or global capital-output ratio for the
ute all of the output-increment to investment, entire economy. But this ratio depends on
is to take a too mechanical — and too easy — capital-output ratios in the various sectors of
view of the changes that are necessary for an the economy, with the overall ratio being an
increase in output. average of the sectoral ratios, weighted by
On the other hand, exclusive attention to a the increases in sectoral outputs. Since the
capital-output ratio may exaggerate the need overall ratio will be affected by the changing
for investment, insofar as output may be in- composition of output and investment among
creased bychanges in other factors without the several sectors, it is essential to analyze
requiring a sizeable amount of investment, or the capital-output relationships at the sec-
even any additional capital. If, for instance, toral level.
unutilized capacity exists, it is possible to Recognizing the problems raised above,
raise output with the fuller utilization of the W. B. Reddaway has offered a summary of
existing capital stock or without requiring what needs clarification when considering a
much more capital. Or there may be consid-
erable opportunity to raise output by apply- 2Such a distinction is suggested by Harvey Leiben-
ing better methods of production to existing stein, Economic Backwardness and Economic
plant. To avoid taking either an overoptim- Growth, New York, 1957, p. 178.
 224 MOBILIZING DOMESTIC RESOURCES

marginal capital-output ratio for a sector. He is to treat the ratio as if it were simply a tecn-
states that it would be desirable to divide the nical relationship applicable to a new plant;
increase in output for a sector between two in practice, the actual ratio is likely to differ
dates into these components.3 from r, depending on the values of the other
Output, (i) Increase due to better methods terms in the above ratio.4 Although M, D, S,
applied to old plant, involving little or no net and W may be relatively small, P will not be
capital expenditure (called P for progress), insignificant if there are large opportunities
(ii) Changes due to fuller (or lower) utiliza- for increasing output by methods which in-
tion of old plant, as a reflection of changes in volve negligible amounts of investment, and
demand (called D). (iii) Changes due to in- L will not be small if much of the period's
troduction of double-shifts, etc. (S). (iv) investment goes into projects that are not
Changes due to better weather (W). (v) completed during the period. When P is sig-
Changes of the kind for which a certain re- nificant, the observed marginal capital-out-
lationship between capital and output may put ratio will be lower than if simply r is es-
reasonably be assumed as "given" by tech- timated; and when L is significant, because
nical factors — at least if we assume a fixed new investment projects take a long time to
number of shifts, fairly full utilization, and complete and considerable construction is
no shortage of labor; the bringing into use of started in the period, then the observed ratio
new steel mills is a good example. If the cap- will be higher than simply r. These consid-
ital cost of these is x and the capital-output erations caution us against assuming that the
ratio in a new mill is r, then the increase in marginal, capital-output ratio is constant,
annual output = xjr. even at the sectoral level.
Investment. Investment in the period will At the aggregate level, the difficulties are
consist of x, plus any capital expenditure de- compounded. Even in the simplest (but most
signed to save labor without increasing out- special) of cases — namely, production coeffi-
put (M for "modernization") and plus (or cients fixed in all sectors and relatively small
minus) an adjustment for the difference be- values for all the other variables that might
tween expenditure on construction in the pe- affect output — the overall marginal capital-
riod and completion (L for "lag"). output ratio will still not be fixed, since sec-
Observed Capital-Output Ratio. If we toral output may vary with changes in de-
work from historical statistics (or from fig- mand. More generally, the overall ratio will
ures for future years included in a plan) the vary according to a number of conditions,
traditional marginal capital-output ratio for some of which may allow only a small addi-
a sector is then equal to tional income to be generated when more
x + M + L capital is accumulated, while others may
contribute to a large increment in output.
-+P+D+S+W Thus, the following conditions will tend to
r
make the capital-output ratio high: the sec-
If we consider only the first term in the nu- toral pattern of investment is biased toward
merator and the first term in the denomina- heavy users of capital, such as public utilities,
tor— ignoring changes in M, L, P. D, S, and public works, housing, industry rather than
W — we are then using the capital-output agriculture, and heavy industry rather than
ratio in an oversimplified way. Only if these light industry; there is excess capacity in the
other changes are small relative to x and xj utilization of capital; other resources are lim-
r can the marginal capital-output ratio be ited, and capital is substituted for these lim-
considered approximately equal to r. But this itational factors; capital is long-lived; the rate
of technological and organizational progress
is low; and investment is for completely new
3W. B. Reddaway, The Development of the Indian
Economy, Homewood, 1962, pp. 207-8. 4Ibid., pp. 208-9.
 225 THE ROLE OF CAPITAL

units of production rather than simply for ex- opens up new natural resources, or permits
tensions ofexisting plant. the realization of economies of scale.
In contrast, the marginal capital-output From such considerations, we must con-
ratio will be lower when the composition of clude that the marginal capital-output ratio
output is biased toward labor-intensive com- is unlikely to be constant over time. A pro-
modities, the average life of capital is shorter, jected ratio must be estimated over the
the rate of technological and organizational period for which investment requirements
progress is high, and when some capital ex- are being calculated, and it may then turn
penditure allows fuller use of previously un- out that there is a wide discrepancy between
utilized capacity, increases the productivity the actual ratio and the projected ratio.
of labor, allows capital-saving innovations,

IV.A.3. The Place of Capital in

Economic Progress*

Capital occupies a position so dominant in ties the rate of capital accumulation out of
the economic theory of production and distri- savings is equal to about 10 per cent of in-
bution that it is natural to assume that it come. Ifone were to assume that innovation
should occupy an equally important place in came to a standstill and that additional in-
the theory of economic growth. In most of the vestment could nevertheless yield an average
recent writings of economists, whether they return of 5 per cent, the consequential rate of
approach the subject historically (e.g., in an increase in the national income would nor-
attempt to explain how the industrial revo- mally be no more than % per cent per annum.
lution started) or analytically (e.g., in models We are told that the national income has in
of an expanding economy) or from the side of fact been rising in such communities at a rate
policy (e.g., in the hope of accelerating the of 2-3 per cent per annum. On this showing,
development of backward countries), it is the capital accumulation could account for, at
process of capital accumulation that occupies most, one-quarter of the recorded rate of
the front of the stage. There is an unstated economic "progress." Nor were things very
assumption that growth hinges on capital ac- different in the nineteenth century. . . .
cumulation, and that additional capital Even this way of putting things exagger-
would either promote or facilitate a more ates the role of capital in economic develop-
rapid rate of economic development even in ment. For the yield on additional capital
circumstances which no one would describe would rarely be as high as 5 per cent if there
as involving a shortage of capital. were not a discrepancy between the existing
Yet there seems no reason to suppose that stock of capital and the stock appropriate to
capital accumulation does by itself exercise the existing state of technique. If innovation
so predominant an influence on economic de- in the broadest sense of the term were at a
velopment. Inmost industrialized communi- standstill, accumulation would continue until
the rate of interest fell to a point at which
*From A. K. Cairncross, "The Place of Capital in
Economic Progress," in L. H. Dupriez (ed.), Eco- saving ceased. The sole object of accumula-
nomic Progress, Papers and Proceedings of a Round tion in those circumstances would be to take
Table held by the International Economic Associa- advantage of the progressive cheapening of
tion, Louvain, 1955, pp. 235, 236-7, 245-8; Cairn-
cross, Factors in Economic Development, George capital in order to introduce more round-
Allen and Unwin, London, 1962, pp. 111-14. Re- about methods of production, not to keep
printed bypermission. pace with current developments in technique.
 226 MOBILIZING DOMESTIC RESOURCES

Ordinary observation suggests, however, that taken place in the past; forces making for
the scope for investment in industry to take rapid increase in income may be largely nul-
advantage merely of lower rates of interest, lified unless they are reinforced by a parallel
once the long-term rate is below 5 per cent, increase in capital. It is to the third, however,
is extremely limited, although there may be that one must usually look — at least in an ad-
a good deal more scope in other directions vanced industrial country — for the main in-
where capital charges form an unusually fluences governing the rate of growth of real
high proportion of the final cost (e.g., in the income per head. Whatever may have been
erection of dwelling-houses, public buildings true in the past, it is now technical innova-
and the like). tion— the introduction of new and cheaper
The contribution of capital to economic ways of doing things — that dominates eco-
progress is not, however, confined to the usu- nomic progress. Whether technical innova-
fruct of additional capital assets, similar to tion, in the sectors of the economy in which
those already in existence. It embraces three it occurs, makes large demands on capital is,
distinct processes. First, a greater abundance however, very doubtful. Many innovations
of capital permits the introduction of more can be given effect to in the course of capital
roundabout methods of production or, to be replacement out of depreciation allowances,
more precise, of a more roundabout pattern which, in an expanding economy, may be
of consumption. This covers the freer use of fully as large as net savings. Others may ac-
capital instruments in the production of a tually reduce the stock of capital required.
given product, the use of more durable in- Existing buildings and existing machines can
struments, and a change in the pattern of often be modified so as to allow most of the
consumption in favour of goods and services advantages of the new techniques to be
with relatively high capital charges per unit gained. It is economic expansion, far more
cost. Secondly, the accumulation of capital is than technological change, that is costly in
a normal feature of economic expansion, capital. . . .
however originating. This is the process nor- Given that the national income is increas-
mally referred to as widening, as opposed to ing, whether under the influence of technical
deepening, the structure of production. It progress, population growth, or some other
may accompany industrialization, or any factor, there is good reason to expect that ad-
change in the balance between industries ditional capital will be required in some im-
that makes additional demands on capital; or portant sectors on a comparable scale. Habits
it may accompany an extension of the market of thrift — a phrase that must now be
associated with population growth, more fa- stretched to include not only the practices of
vourable terms of trade, or the discovery of corporations in adding to reserves but the
additional natural resources. Thirdly, addi- propensities of Finance Ministers — appear to
tional capital may be required to allow tech- admit of capital accumulation at a rate of
nical progress to take place. It may either fi- about !}/t per cent per annum, and this has in
nance the discovery of what was not known recent years been close to the rate of growth
before or more commonly, the adaptation of of income. Provided, therefore, that the cap-
existing knowledge so as to allow of its com- ital requirements of industry — the main sec-
mercial exploitation through some innovation tor left out of account — are also increasing at
in product, process or material. this rate, the capital-income ratio will remain
Now of these three, the first is generally of constant and the whole of the country's thrift
subordinate importance; it is unusual for cap- will be effectively mobilized. There can be no
ital accumulation, unassisted by other fac- guarantee,
tors, to bring about a rapid increase in in- ments will inhowever, that atindustry's
fact mount this rate, require-
even in
come. The second, which also abstracts from the long run. In the short run, for reasons
any change in technology, accounts for that are familiar, the whole process of capital
nearly all the capital accumulation that has accumulation may be thrown out of gear.
 227 THE ROLE OF CAPITAL

Now the significant feature of this argu- wanted only after another has been created.
ment isthat it hinges far more on the indirect Although the full consequences may be en-
than on the direct demand for capital. It as- tirely foreseeable, development does not work
sumes that technical progress operates up to its full momentum until a whole series
largely in independence of capital accumu- of changes have occurred: an extension of ca-
lation and that capital is needed, not in order pacity here, an application of the new tech-
to allow innovation to be made but in order nique there; a shift of location in one indus-
to consolidate the improvements in income try, abuilding up of new attitudes in another.
that innovation brings about. Moreover, it The introduction of the steam engine, for ex-
implies that if, at any time, the process of in- ample, brought into existence a large reser-
novation creates a bulge in the demand for voir of projects that trickled out into capital
capital, it should be possible to adapt the pat- formation all through the nineteenth century:
tern of investment so as to accommodate the the stock of capital appropriate to existing
high-yielding requirements of industry by technique was far above the existing stock
displacing part of the larger, but less remu- both because the steam engine was capable of
nerative, requirements of house-building, wide application and because many indus-
stock-building, and so forth. tries that themselves made no use of it (such
It is hardly necessary to show that this im- as bridge-building) were transformed in scale
plication may be mistaken. Public policy may or (like agriculture and many pursuits ancil-
maintain the demand for capital in the sec- lary to it) in location.
tors capable of compression or the capital Moreover, because the chain reaction
market may be so organized that industry is takes time and the innovation is, ex hypoth-
unable to draw capital from the sources that esi, a profitable one, the process is to a large
finance other forms of accumulation. But un- extent self-financing. If there is a spate of
less the bulge is a very large and consistent such innovations, interest and profits are
one it is doubtful whether innovation need suf- likely to show some response and a corre-
fer greatly. sponding shift in the ratio of savings to in-
The effect of technical progress is gener- come will ease the heavier burden of finance.
ally to widen the divergence between the ac- It may happen, however, that the situation is
tual stock of capital and the stock consistent not regulated in this way: interest rates may
with the full exploitation of current worker be sticky upwards as well as downwards. The
opportunities. Some part of the additional probable outcome will then be a series of
capital will be needed to finance the innova- spurts in investment, followed by periods of
tions in the sectors of the economy in which indigestion. . . .
they arise; some will be linked with the in- A variant of this situation is one in which
novations directly, either because associated there has been a considerable lag behind the
industries are offered a wider market or be- known opportunities for the fruitful use of
cause social capital has to be provided in an capital at existing rates of interest. A country
area where it has become insufficient; some may fail to make use of technical knowledge
will be linked indirectly, in the way already available elsewhere and suddenly become
outlined, because the increased expenditure alive to the possibilities of applying that
of consumers will give rise to a derived de- knowledge. At that stage its capital require-
mand for capital. Now it is common to find ments will increase discontinuously and the
that, particularly with a major advance in additional capital which it requires before
technique, the influence which it exerts on bumping up against the limits of technical
the scope for eventual capital accumulation is advance may be very large. It appears to be
far more profound than its immediate impact this situation that is in the minds of those
on the current flow of capital formation. who assume that the injection of additional
There is generally a chain reaction, strung capital into a country's economy will almost
out through time, one physical asset being automatically speed up its economic prog-
 228 MOBILIZING DOMESTIC RESOURCES

ress. Sometimes the argument is framed sumed to be limited by capital, whether there
more specifically in terms of a shift of em- is abundant labour or not. A high rate of cap-
ployment from agriculture to industry, with ital formation usually accompanies a rapid
a large net gain in productivity from the growth in productivity and income; but the
shift, and the large capital investment needed causal relationship between the two is com-
to accomplish it operating as a brake. plex and does not permit of any facile as-
This is a complex situation and it may exist sumption that more capital formation will of
in some underdeveloped countries. But it is itself bring about a corresponding accelera-
by no means obvious that additional capital, tion in the growth of production.
whether borrowed from abroad or accumu- In industrial countries this is only too ob-
lated through the exertions of surplus labour vious. Capital formation may assume forms,
in the countryside, would by itself suffice to such as house-building or an addition to liq-
start off a cycle of industrialization. The uid stocks, that are unlikely to add very per-
problem is often one of organization quite as ceptibly to productivity although they may
much as of capital creation: of training man- yield a sufficient return to make them worth
agements and men; of creating new attitudes while. If all capital formation were of this
towards industrial employment; of taking ad- character, or represented an enlargement of
vantage ofinnovations that need little capital the capital stock with assets broadly similar
and using the resulting gains to finance in- to those already in existence, it would be hard
vestment elsewhere. to account for the rates of growth actually re-
On the whole, there is a greater danger corded. Amoment's reflection will show that
that the importance of capital in relation to even an average return of 10 per cent to cap-
economic progress will be exaggerated than ital in a country saving 10 per cent of its in-
that it will be underrated. How many suc- come annually would raise income by no
cessful firms, looking back over their history, more than 1 per cent per annum.1 Similarly,
would single out difficulty of access to new efforts to impute the recorded expansion in
capital as the major obstacle, not to their industrial production to the additional labour
growth, but to the adoption of the most up- and capital contributing to it invariably leave
to-date technique? How many countries in a large unexplained residual.2 It is necessary,
the van of technical progress have found therefore, to take account of other influences,
themselves obliged to borrow abroad? It is such as technical progress and improvements
where there has been a lag, where technical in social and economic organization, which
progress has been too slow, that capital is may operate through investment, or indepen-
called upon to put matters right. No doubt dently ofit, so as to raise the level of produc-
where capital is plentiful, more risks can be tion. These influences, if they take effect
taken and development is speeded up, so that uniformly throughout the economy in com-
rapid development and rapid capital accu- petitive conditions, will tend to swell the na-
mulation gotogether. But the most powerful tional income without raising the average re-
influence governing development, even now, turn to capital, the extra output slipping
is not the rate of interest or the abundance of
capital; and the most powerful influence gov-
erning capital accumulation, even now, is not 'This point is developed in my "Reflections on the
technical progress. . . . Growth of Capital and Income," Scottish Journal of
Political Economy, June 1959. See also the com-
There is general agreement that, in all
ments by E. Lundberg, "The Profitability of Invest-
countries, the process of economic growth ment," Economic Journal, December 1959.
and capital accumulation are closely inter- 2See, for example, W. B. Reddaway and A. D.
con ected. Itwas in terms of this intercon- Smith, "Progress in British Manufacturing Industries
nection that the earliest theories of economic in the Period 1948-54," Economic Journal, March
development were formulated; and in the 1960, and O. Aukrust, "Investment and Economic
Growth," Productivity Measurement Review, Feb-
work of modern economists, output is still as- ruary 1959.
 229 THE ROLE OF CAPITAL

through to the consumer, the wage-earner or These suppositions are not altogether extrav-
the government. agant. Technical progress does not always in-
How far it is correct to attribute an expan- volve high net investment: indeed it may per-
sion in output to high investment, when high mit of a reduction in the stock of capital or
investment is only one of the factors at work an expansion in output without any compa-
is necessarily debatable. It would certainly be rable investment. A change in the pattern of
legitimate if capital formation was lagging investment could also, by enforcing the con-
behind, and finance could be identified as a tinued use or overloading of old types of
bottleneck in the process of expansion. It plant, make possible a far more rapid con-
might also happen that the rate of technical struction ofthose newer types which bear the
advance was itself controlled by the scale of fruits of technical progress in greatest
investment, not merely because capital for- abundance.
mation was the means by which new tech- Attempts are sometimes made to settle the
niques were adopted but also because high issue by citing the apparent constancy of the
investment created an atmosphere favourable capital-income ratio and deducing from this
to experimentation and innovation. There is the "neutrality" of technical progress. But
undoubtedly some tendency for all the symp- the capital-income ratio is affected by many
toms of rapid growth to show themselves si- things other than technical progress: the dis-
multaneously. But there is no invariable de- tribution ofconsumers' expenditure between
pendence ofgrowth on a high rate of capital capital-intensive and labour-intensive prod-
formation and it is easy to imagine circum- ucts; indivisibilities in past investment — for
stances in which efforts to increase capital example, in the transport and communication
formation may actually slow down the prog- network; changes in the pattern of trade; in-
ress of the economy. vestment in social assets such as roads,
Moreover there is some justification for schools, and hospitals to which no income is
turning the causal relationship the other way imputed; and so on. Even if these influences,
round. If income is growing fast, investment too, are neutral and if the capital-income
opportunities are likely to be expanding cor- ratio does remain constant — and neither of
respondingly fast, so that the growth in in- these assumptions seems well-founded — the
come draws capital accumulation along be- fact that capital and income grow at the
hind it. The biggest single influence on same rate tells us nothing about the causes of
capital formation is market opportunity, and growth in either. There is no reason at all
many types of capital accumulation are likely why one should rule out the suggestion that
to be embarked upon only when income is the same circumstances that favour rapid
booming. If capital formation does not re- growth of income are also favourable to a
spond, its failure to do so will certainly act as rapid growth of investment.
a drag on the expansion in output. But there This may seem a rather arid and irrelevant
is no reason why it should bring it to a halt, issue: arid, because if capital requirements
and, given a re-arrangement of the invest- must keep pace with the growth of income
ment pattern, income might grow a long way that is all we need to know for practical pur-
before the shortage of capital became acute. poses; irrelevant, because the issue relates to
In the meantime the rapid growth in income, experiences in industrial rather than prein-
particularly if it were accompanied by high dustrial countries. But when it is so com-
profits, would be likely to generate additional monly urged that countries will be able to
savings and so mitigate any symptoms of cap- take-off if only they are provided with suffi-,
ital shortage that manifested themselves. cient capital from outside, the issue seems
All this presupposes that a spurt in income neither arid nor irrelevant. For this thesis as-
could precede an acceleration of investment, sumes the very causal relationship that is in
and that capital formation is subordinate to dispute.
other elements in the process of growth.
 230 MOBILIZING DOMESTIC RESOURCES

EXHIBIT IVI. Structure of Demand

ing
Distribution of Gross Domestic Product (percent)

Gross Exports of
Goods and
Public Private Domestic Gross Domestic Nonfactor Resource --I6w
Investment — Balance
1 w
Consumption Consumption Services —-I -Iw
Sav 1 ww
-3h>
-1
-15 -
Low-income economies 1960" 1119H81* 78 w 74 w 19 w 18 w 21 H Iw -1
— — 960° -3
China and India
8h> 111 w 74198w1* 2\w
I960" 2619w81* 20 w 25 w 4 W
-5
Other low-income 10 w 11 H 81 w 84 w 12 w 14 H 101960
w
° 15 w
1981* 1960° 129h>w *
Iw
1 Kampuchea, 24 H* 1981 1960°
Dem. -6 1981*
2 Bhutan 8m> -22
3 Lao, PDR 23
13 — — —-4 — — -1
4 Chad 82 11 5 10 -14 -12
90
5 Bangladesh 6 8 86 7 17 8 2 131 1
15
6 Ethiopia 8 81 81 12 10 11 4 9 --134
7 Nepal — c — 92 — — -48 — — — -2
8 Burma c c 89 83 14 11
13 20 9 -14 -4 7
-15
c — — 12
16 —24 — 17-1 —18 —
9 Afghanistan
12 79
87 1 4
10 Mali 26 80 14 16 9 12
16 10 80
11 Malawi 10 10 21 22
18 16 88 22 -14
12 Zaire 75 59 12 33 21
16 25 55 36 9
61 -2-5
13 Uganda 9 c 11 3 3 26 1 5
79
97
14 Burundi 16 92 19 5 13 139
3 106 16 5 (•)-6
1 5 Upper Volta 10 15 96 9
94 12 -12
10 75
16 Rwanda 82 70 6 23
23 8 8 12 2 -16
17 79 -52
17 India 7 10 10 14 20 7
18 Somalia 8 — 86 —78 17 — 6 — 13 — — -37
72 - -6
19 Tanzania 9 14 22 19 8 31 3 14 5 -1
20 Viet Nam 14 -2 -12
76 -11 -15
72 -7
21 China c c 23 28 4 9 1
22 Guinea — 19 — — 78
11 — 24 14 — — (•)3 -1
93
23 Haiti c c 67
99 9 n 7 1 20 34 -120
78 -7 -
24 Sri Lanka 13 7 81 28 9 12 44 14
13 75 14
25 Benin 16 15 35 9 12 31 -7
89 -12
26 Central African 31
13 72 -16
Rep. 19 90 20 9 9 23 26
27 Sierra Leone — 11 — 91 — 13
15
— — 13 —
20 16 75 76
28 Madagascar 77 5 7 12 17
79 11
29 Niger 9 9 13 12 15 9 22
30 Pakistan 11 11 84 12 27 5 7 8 12
82 17
31 Mozambique 13 16
80 100 259
32 Sudan 8 c 12 12 (•)
15 (•)
33 Togo 8 88 11 31 4 19
17 73 68
34 Ghana 10 11 85 6 4 28 4
24 17
 23 1 THE ROLE OF CAPITAL

EXHIBIT IV. 1. Structure of Demand (Continued)

Distribution of Gross Domestic Product (percent)

Exports of -9
Gross Saving Goods and
Resource -3
Public Private Domestic Gross Domestic Nonfactor Balance -1
Consumption Consumption Investment Services -1 w
-2w
-5 -
-In -3h>4w
-6-w
22
Middle-income I960" 1981* -\w
-1
I960" 1981* -29
economies 70 w 66 w 20 96w0" 19 w 22 w 17 w -34
11 w 14 H 1 25198w1* 23 w
Oil exporters 11 w 15 w 70 w 61 K 18 w 26 w I19960w* 24
1981w* 21 K " 1 w -65
Oil importers 11 w 14* 70 w 69 h 21 w 25 w 19 w 15I96w0 w*
231981
21 H 23 h> 1960° 1981*
10 w -21 23 w -6
Lower middle-income 13 w 76 w 68 w 15 w 14 w 19 w 15 w -7
72 25 -2
35 Kenya 11 21 63 20 25 m> 16 31
40 25 -8
15
36 Senegal 22 68 83 16 17 29
17
25 71 17 -27 -1
37 Mauritania 29 62 38 4 9 15 49 -2
38 Yemen Arab 38 -89 -11-50 -5
-25
Rep. — 20 — 101 — 44 — — 6 — -5
-2
39 Yemen, PDR -13
35
40 Liberia 7 21 58 28 18 7
11 62 17
41 Indonesia 12 80 66 8 21 8 23 13
39 28
51 (•) 2
42 Lesotho 17 26 108 163 2 13 12 13
86 21
43 Bolivia 7 10 77 14 7 13 13 13
44 Honduras 11 14 24 12 18 22 32 (•)-3 -6
14 77 68 41 15 16 -6
28 48 23 -7
45 Zambia 11 25 56 36 -12 -17
46 Egypt 19
71 57
75 13 30 12 20 -1-15
17 64 16 -18
47 El Salvador 10 15 79 12 11 17
10 20 25
34
76
48 Thailand 10 12 65 16
16 28 23 31
76 14
49 Philippines 8 8 16 25 17
11 19 (•)
67 30 -6
50 Angola -2
51 Papua New -5
71 13
Guinea 28 26 63 28 1 11 16 38
52 Morocco 12 c 77
79 73 10 23 11 8 21 1 -66 -3 -4
92
53 Nicaragua 9 21 15 12 6 24 21 -4 -7
12 24
54 Nigeria 6
87 65
13 29 -72 1
23
15
15
24 25
55 Zimbabwe 11 18 23 22 22 — —
— —67 71
67 — — — —
56 Cameroon 7 21 32 1-2
22 ,
57 Cuba -10
58 Congo, People's 23 50 45
Rep. 12 98 32 38 21 62 6
59 Guatemala 8 8 81 10 17 8 11 13
84
73 17
60 Peru 9 13 25 19 14 20 17 2
15 78 27
61 Ecuador 11 64 26 11 24 16 22
61 14
62 Jamaica 7 21 73 30 16 26 12 50
10 18 67 67
63 Ivory Coast 62 15 20 34 2
13 78 27 17
64 Dominican Rep. 8 68 12 14 37 18
34 7
24 19 24
65 Mongolia
 232 MOBILIZING DOMESTIC RESOURCES

EXHIBIT IV. 1. Structure of Demand (Continued)

-5
Distribution of Gross Domestic Product (percent)
-10
-3
Gross Saving Exports of
Private Domestic Gross Domestic Goods and Resource
Public Nonfactor Balance
Consumption Consumption Investment
Services -1
-15
28 --57
73 68 21 21 24 20
16 42
12 10
66 Colombia I9660" 18981* 76 *
67 Tunisia 17
15 I960 62981*
1
31 137 23-
16 (•)
15 0"
68 Costa Rica 10 77 60 18I96 281981*
17
I960" 25 21 44
1981* 1960°
69 Korea, Dem. -1°0
Rep. 1981* 1960 *
13 1981
-5
76 70
70 Turkey 11 11 16 25 19 3 11 — 1 w
71 Syrian Arab Rep — 22 — — 41 — 9 — 18 —
— 30 — — 24 — 16 — —
72 Jordan 86
69
76 -1
73 Paraguay 8 7 74 29 19 18 541 — 1 H>
17 -2
12 w 15 w 67 w 18 w -11
Upper middle-income 65 w 22 w 25 w 21 H
24 w 23 w -7
74 Korea, Rep. of 15 12 66 11 26 1 22 3 -17 -1
84
75 Iran, Islamic 39
10 — 69 — — 21 — 42
19 — 4 — -2-
Rep. of 19
76 Iraq 18 — 48 — 17
20 — — — 13 —
-5
77 Malaysia 11 21 62 53 14 32 34
27 26 53 14 --61
21 78 0
78 Panama 11 56 16 29 11 23 31
54 40
16
79 Lebanon 10 — 85 — — 5 — — — -8
16 45
80 Algeria 15 60 42 25 31
27 34 2
12 c 81 22 20 39
23
19 5 9
81 Brazil
76 37
82 Mexico 6 15 67 62 20 25 21
18 10 13 -11
16 76
83 Portugal 11 77 19 12 8 17
70 277 -15
84 Argentina 9 15 62 22 26
27 21 23 9 -12
13 75 -13
85 Chile 9 79 12 -3 18
86 South Africa 9 13 50 14
22 22
29 12 30
14 36 5 -14 8
64 27 23
87 Yugoslavia 19 15 49 56 18 32 32 37
79 75
88 Uruguay 9 13 15 12 12
29 14 15
37 25
89 Venezuela 14 14 53 56 21 33 30 30 5
77
90 Greece 12 18 66 19 25 329 20 12
91 Hong Kong 7 8 68 18 116 14
24 82 43
111
87 20
92 Israel 18 36 59 27 30 14 5
68 33 163 212
93 Singapore 8 10 95 42 14
94 Trinidad and 57 11
Tobago 9 c 61 60 30 30 40 45 2 10
28
High-income oil —37 —
exporters
— — 22 w — 26 w
— 58 w
21 H 48 69 h 32 w
— 26 — — — — 60 —
95 Libya 26 26
— 23 — — 34 — — 68 — 33
96 Saudi Arabia 18 59 14
97 Kuwait — 15 — — — 46 — 71 — 29
39 17 72 78
98 United Arab
Emirates — 11 — — 28 — — — 44
17
 233 THE ROLE OF CAPITAL

EXHIBIT IV. 1. Structure of Demand (Continued)

Distribution of Gross Domestic Product (percent) -5 -2

-2
-2
Gross Exportsand
Goods of -2-14
Public Private Domestic Gross Domestic Nonfactor Resource -2
Consumption Consumption Investment Saving Services Balance —
1 w
-1 -3
-1
Industrial market 1960° -3
1981*
economies 15 w 17 w 631960° 1981* -2
w 60" 2219w81* 22 w 21 w 12 H 20 w
61 H 21I9 H I960"
99 Ireland 12 22 11 62 16 30
20 11 1981* 321960° 63
16 -4
70 *
100 Spain 7 12 72 18 21 18 10 171981 13 w "
I9 60 -19181*
-4
13 18 62 63 25 21 25 19
23 -1
101 Italy (•)
102 New Zealand 11 68 23 25 21 20 14
22 28
29
27
17 60
17 22 66 58 19 21
103 United Kingdom
17 17 15 -31
104 Japan 8 10 59 58 33 32 11 (•) 1
105 Austria 13 18 59 56 33
28 31
26 28 26 25 42
13 18 (•) (•)
106 Finland 58 55 30 28
26 29
25 23
15 -2 -1
107 Australia 10 17 65 60 29 27
23
25 15
34 -1 -2
108 Canada 14 20 65 55 23 25 21 18 28 (•)
109 Netherlands 13 18 62 27 18 29 21 48 58 2 3
13 20 61
66 18 18 65
110 Belgium 23 33
15
13 69 19 14
111 France 16 21 25 10
22 2
18 62 67 17
112 United States 17 18
26 19
16 19 18 5 1
28 64 64
113 Denmark 13 62 56 25 16 36
32 ()
114 Germany, Fed. 23
Rep. 14 21
57
56 26 29
28 23 19
41 48
30 2 (•)
13 59 27
30 23 8
1 1 5 Norway 19 47 25
116 Sweden 16 30 60 19 24 34
18 31
13 52 26
117 Switzerland 9 63 29 29 24 29 (•)
62
East European
nonmarket 37
economies — — — — — — — — — — — —
118 Albania ___________ _
119 Hungary c 10 74 61 28 30 26 29 — 39 — -1
120 Romania _____ 33 _ _ _ 28 — (.)
121 Bulgaria — — — — — — — — — — — —
122 Poland ___________ _
123 USSR _______ ____ _
124 Czechoslovakia — — — — — — — — — — — —
125 German Dem.
Rep. ___________ _

Source: World Bank, World Development Report, 1983, pp. 156-57.

"Figures in italics are for 1961, not 1960.
*Figures in italics are for 1980, not 1981.
cSeparate figures are not available for public consumption, which is therefore included in private consumption.
m> Indicates a weighted average.
(.) Indicates a nil amount.
 IV.B. SOURCES OF CAPITAL
FORMATION-NOTE

Whether it be financed from internal sources borders. Such agencies will not only permit small
or external, by noninflationary or inflationary amounts of savings to be handled and invested con-
means — the accumulation of capital in any veniently but will allow the owners of savings to
developing economy requires the mobiliza- retain liquidity individually but finance long-term
tion of an economic surplus. If investment is investment collectively.1
to increase, there must be a growing surplus Although the existence of a more devel-
above current consumption that can be oped capital market and financial intermedi-
tapped and directed into productive invest- aries will aid in the collection and distribu-
ment channels. The different ways of financ- tion of investible funds, they in no way lessen
ing capital formation will entail different in- the need for real saving. The rate of invest-
stitutional arrangements (for example, the ment which it is physically possible to carry
plowing back of industrial profits into invest- out is limited by saving, and a "shortage of
ment would imply a different institutional
capital" — in the sense of a shortage of real
framework from that of financing through resources available for investment pur-
taxation by the state). It should be recog- poses— cannot be solved merely ')y increas-
nized, however, that the process of capital ing the supply of finance. Indeed, it is com-
formation involves three essential steps: (1) paratively easy to introduce institutional
an increase in the volume of real savings, so arrangements to increase the supply of fi-
that resources can be released for investment nance, and a lack of finance need not persist
purposes; (2) the channeling of savings as a serious bottleneck. Once a sizeable class
through a finance and credit mechanism, so of savers and borrowers come into being, fi-
that investible funds can be collected from a nancial intermediaries are likely to appear,
wide range of different sources and claimed and lending institutions are readily created.
by investors; and (3) the act of investment it- But the creation of new financial institutions
self, bywhich resources are used for increas- is no substitute for the necessary perfor-
ing the capital stock. mance of real saving.
The first requirement — an increase in the It is therefore important to be clear on the
volume of real savings — is of fundamental various sources from which the necessary
importance if a higher rate of investment is savings can be mobilized to provide the
to be achieved without generating inflation. wherewithal for capital expenditure. From
This crucial step of mobilizing savings should internal sources, an increase of savings may
not be confused, however, with the monetary be generated voluntarily through a reduction
financing of investment. The significance of
in consumption; involuntarily through addi-
financial institutions lies in their making tional taxation, compulsory lending to the
available the means to utilize savings. As one government, or inflation; or, finally, by the
study of the role of financial institutions absorption of underemployed labor into pro-
concludes:
ductive work. From external sources, the fi-
However poor an economy may be there will be nancing of development may be met by the
a need for institutions which allow such savings as investment of foreign capital, restriction of
are currently forthcoming to be invested conve- consumption imports, or an improvement in
niently and safely, and which ensure that they are the country's terms of trade.
channelled into the most useful purposes. The An increase in voluntary saving through a
poorer a country is, in fact, the greater is the need
for agencies to collect and invest the savings of the 'Edward Nevin, Capital Funds in Underdeveloped
broad mass of persons and institutions within its Countries, London, 1961, p. 75.

234
 235 SOURCES OF CAPITAL FORMATION— NOTE

self-imposed cut in current consumption is to them is that they all suffer from a shortage of
unlikely when the average income is so low. revenue. This is partly because they have a low
At best, it can be hoped that when income "taxation potential" — which may be defined as the
rises, the marginal rate of saving may be maximum proportion of the national income that
can be diverted for public purposes by means of
greater than the average rate. Instead of re- taxation. But more important is the fact that the
lying on voluntary saving, the government taxation potential in such countries is rarely fully
will normally have to resort to "forced" sav-
ing through taxation, compulsory lending, or
credit expansion. The efficacy of credit ex- exploited.2
A country's "taxation potential" depends
pansion and its resultant inflationary conse- on a variety of conditions — the level of per
quences are discussed in section IV. D. capita real income, the degree of inequaltiy
Issues of taxation can be analyzed from in the distribution of income, the relative im-
two different perspectives — that of incentives portance ofdifferent sectors in the economy
or that of resources. If, as in supply-side (cash crops, subsistence agriculture, mining,
economics, one believes that a lack of ade- foreign trade), the political leadership, and
quate incentives inhibits investment and administrative powers of the government. It
growth, then the tax system should be im- is generally true that in the lagging LDCs the
proved through the granting of additional actual ratio of tax revenue to national income
concessions of various kinds. If, however, in- is less than in the more progressively devel-
sufficient investment and low growth are at- oping countries and is less than their tax po-
tributed to a lack of resources, then the tax tential. In many countries it might be possi-
system should be designed to increase re- ble to increase the tax effort and approach
sources available for investment through ad- closer to the tax potential — especially if more
ditional taxation. effective systems of progressive taxation were
Emphasizing the shortage of resources in to be designed that recognized the inequality
an LDC, rather than inadequate incentives, in the distribution of income, the taxable ca-
Kaldor has stated: pacity of the agriculture sector, and the ris-
ing share of industrial and commercial
The importance of public revenue from the wealth as development proceeds. The saving
point of view of accelerated economic development
that is forced by additional taxation, how-
could hardly be exaggerated. Irrespective of the
ever, will be less than the additional tax rev-
prevailing ideology or the political color of partic- enue to the extent that there is a restriction
ular governments, the economic and cultural de-
velopment ofa country requires the efficient and in private voluntary saving instead of a fall in
steadily expanding provision of a whole host of consumption by the full amount of the tax.
non- revenue-yielding services — education, health, To assess the tax performance of a devel-
communications systems, and so on, commonly oping country, we should go beyond a static
known as "infrastructure" — which require to be fi- index, such as the ratio of tax revenue to na-
nanced out of government revenue. Besides meet- tional income, and introduce more dynamic
ing these needs, taxes or other compulsory levies
provide the most appropriate instrument for in- concepts such as "tax elasticity" and "tax
effort." The "tax elasticity" indicates the
creasing savings for capital formation out of do- income-elasticity of the tax system: if the
mestic sources. By reducing the volume of spend- marginal rate of taxation exceeds the average
ing by consumers, they make it possible for the
resources of the country to be devoted to building rate, there will be an automatic increase in
up capital assets. . . . the ratio of tax revenue to national income as
Ruling out inflation as a deliberate instrument, income rises. The "tax effort" measures the,
it may be asked: What are the most appropriate
taxes that can be relied on for maximum revenue? 2Nicholas Kaldor, "Taxation for Economic Devel-
opment," Journal of Modern African Studies, Vol. 1,
The question does not admit of any general answer
No. 1, 1963, p. 7; also, Kaldor, "Will Underdeveloped
in the widely varying conditions of "underdevel- Countries
1963. Learn to Tax?" Foreign Affairs, January
oped" countries. The only feature that is common
 236 MOBILIZING DOMESTIC RESOURCES

political and administrative efforts to in- depends on the ease with which labor can be
crease effective tax rates or the coverage of attracted to investment projects, the degree
the tax system.3 to which labor can form capital directly with-
There is only narrow scope in a poor coun- out requiring additional investment expendi-
try for the practice of compulsory saving ture, the absence of an adverse effect on ag-
through the practice of compulsory purchase ricultural output, and the capacity to offset
of nonnegotiable government bonds. Of the investment with taxation.
greater practical significance may be the op- When we look to external sources of fi-
eration ofstate marketing boards which have nancing development, the capital assistance
a statutory monopoly over export crops. provided by foreign economic aid and the pri-
These boards may compel native producers to vate investment of foreign capital are of most
save by purchasing the native's produce at importance. The next chapter examines the
prices below world prices. contribution of foreign aid and private for-
Finally, another internal source of saving is eign investment. [See sections V.A and V.D.]
represented by the "investible surplus" of un- Some contribution may also come from a re-
deremployed labor. If this "investible sur- striction of consumption imports. Provided
plus" is utilized in productive activity, the that there is not simply a switch in expendi-
national output would be increased, and ture from imports to domestic consumption,
the required savings might be generated from the level of savings will then rise. Imports of
the additional output. It should also be noted capital goods can then be increased, and this
that the direct formation of capital through will represent a genuine addition to the rate
the use of underemployed labor can be ob- of capital formation: the increase in the flow
tained by what is termed the "unit multi- of investment goods imported is, in this case,
plier" method.4 If labor does have zero pro- matched by an increase in the flow of domes-
ductivity inagriculture, it can be withdrawn tic income saved. If, however, consumers in-
and put to work on investment projects (con- crease their domestic spending when they
struction, irrigation works, road building, cannot import, then resources will be di-
etc.) without a drop in agricultural output. verted from domestic capital production in
Most of the payment of the additional wages favor of the increased domestic consumer
will be directed towards foodstuffs, and ag- spending, and the increase in imports of in-
ricultural income will rise. The higher in- vestment goods will be offset by reduced do-
come may then be taxed, and the tax revenue mestic investment. An increase in saving is
can finance the investment project. If taxes therefore necessary if the restriction of con-
are levied in an amount equivalent to the ad- sumption imports is to result in an increase in
ditional wage-bill, there will be no change in
consumption but income will have risen by total net capital formation.5
A similar analysis applies to changes in the
the amount of the investment. When the in- terms of trade. When export prices rise, the
vestment projects are completed there will be
improvement in the country's commodity
an increase in output, and some of this in- terms of trade makes it possible for the coun-
crease in income may also be captured try to import larger quantities of capital
through taxation. How much scope there is goods. But again, this source of capital for-
for this method of direct investment in kind mation will not be fully exploited unless the
3For a fuller discussion, see Richard M. Bird, "As- increment in domestic money income due to
sessing Tax Performance in Developing Countries: A the increase in export proceeds is saved. If
Critical Review of the Literature," in J. Toye (ed.), the extra income merely increases consumer
Taxation and Economic Development, 1978, pp. 33-
61. spending on home produced or imported
4James S. Duesenberry, "Some Aspects of the The-
ory of Economic Development," Explorations in En- 5Cf. Ragnar Nurkse, Problems of Capital Forma-
trepreneurial History, Vol. 3, No. 2, December 1950, tion in Underdeveloped Countries, Oxford, 1953, pp.
pp. 65-7. 111-16.
 237 SOURCES OF CAPITAL FORMATION— NOTE

goods, the opportunity for new saving is lost. sumption may improve labor quality. The
The extra resources made available by the components of labor quality examined are
improvement in the terms of trade must be calories per head, investment in dwellings,
withheld from consumption and directed into higher education, health indicators, and so-
investment.6 Either a corresponding increase cial security benefits. Of these various com-
in voluntary saving or in taxation is necessary ponents, better diet is shown to have the
to give the country a command over addi- greatest impact on labor productivity and
tional imports of investment goods.
A special word should be added about con- growth
Certainof output.8
policy implications follow from the
sumption and capital formation. We have im- view that private consumption may be pro-
plied above that present consumption is at ductive. In its efforts to raise the commu-
the expense of future output; as usually nity's marginal rate of saving, the govern-
stated, it is believed that restraints on con- ment should put more emphasis on taxation
sumption are needed to divert resources from and on business saving through profits rather
the production of more consumer goods to than on individual saving through a curtail-
capital accumulation. But is this always ment of consumption. But there should at the
true? Can a case be made that — in the con- same time be an improvement in the pattern
text of a developing country — an increase in of consumption so that it might contribute as
current consumption may actually lead to an directly as possible to increasing efficiency.
expansion in future production? What is needed is a selective increase in con-
When the level of living is as low as it is in sumption. Luxury consumption, for instance,
an LDC, the distinction between consump- should be taxed, and the import-replacement
tion and investment becomes overdrawn in- of consumer goods should be limited insofar
sofar as private consumption may well have as this policy has become suboptimal. There
a positive marginal productivity. The reason should, however, be an increase in consump-
is not that consumption will augment re- tion that favors the rural population if this
sources, but that a rise in consumption may will help overcome the agricultural bottle-
improve labor quality and efficiency and neck. In this connection, we should note Pro-
hence allow better use to be made of the ex- fessor Myint's observation that "incentive
isting labor resources. The consumption of consumer goods" can be a useful means of
health-improving goods should improve the encouraging peasants to enter the money
ability to work and increase the intensity of
work. The greater consumption of foodstuffs 8W. Galenson and G. Pyatt, The Quality of Labour
and Economic Development in Certain Countries,
that aid nutrition is especially significant. For
ILO, Geneva, 1964, pp. 15-19, 87-8.
it has now been established by medical sci- Another study has shown that, in the rural areas of
entists that improper food, especially a diet Asia, an insufficiency of calories may take the form of
inadequate work effort after the peak season. See
low in protein, can in itself impair the physi-
cal and mental development of children from Harry T. Oshima, "Food Consumption, Nutrition,
and Economic Development in Asian Countries,"
birth.7 Economic Development and Cultural Change, July
In an empirical study of the impact of 1967, pp. 390-91. Cf. also, F. A. O., Nutrition and
Working Efficiency, Rome, 1962; Gunnar Myrdal,
components of "labor quality" on the growth Asian Drama, New York, 1968, pp. 1912-19. But see
of output, Professors Galenson and Pyatt
have demonstrated that an increase in con- Elliott J. Berg, "Major Issues of Wage Policy in Af-
rica," in A. M. Ross (ed.), Industrial Relations and
Economic Development London, 1966, pp. 190-96.
6Ibid., pp. 97-103. Professor Berg argues that under conditions of m\-
1Pre-School Malnutrition: Primary Deterrent to grant labor and a joint family system, higher income
Human Progress, National Academy of Sciences, does not necessarily lead to better nutrition. He also
National Research Council, Washington, D.C., 1966; contends that better nutrition is not sufficient to im-
N. S. Scrimshaw, "Infant Malnutrition and Learn- prove individual efficiency unless there are also pres-
ing," Saturday Review, March 16, 1968, pp. 64-8. ent the necessary motivation and essential cooperant
See further, IX.B, below. factors with labor.
 238 MOBILIZING DOMESTIC RESOURCES

economy: a rise in the aspiration to consume these criteria. If we take the largest view, this
these goods may lead to the sale of a food sur- brings us to the very frontier of multisectoral
plus and may encourage better methods of intertemporal models where we should at-
production that will increase the food surplus tempt to interrelate an optimal consumption
in the future.9 policy with an optimal capital policy. The
In its most general terms, the principle problem of the total amount of investment
that consumption can be productive raises that can be made in the future then becomes
the complex problem of specifying criteria a function of investment allocation and the
for intertemporal efficiency in consumption, pattern of consumption in the present period.
and then shaping policy instruments to meet We need not be overwhelmed at this point by
the complexities of such a model, but we will
9Hla Myint, The Economics of the Developing return in Chapter X to the problem of opti-
Countries, London, 1964, pp. 88. mal investment allocation.

Comment
Reviews of the theoretical and empirical literature on savings rates in LDCs are provided
by R. Mikesell and J. Zuiser, "The Nature of the Savings Function in Developing Countries,"
Journal of Economic Literature (March 1973); H. S. Houthakker, "On Some Determinants
of Saving in Developed and Underdeveloped Countries," in E. A. G. Robinson (ed.), Problems
in Economic Development (1965); G. S. Laumos and P.S. Laumos, "The Permanent Income
Hypothesis in an Underdeveloped Economy," (September 1976).
The role of the government as saver raises problems of taxation and the relative merits of
alternative fiscal policies in developing countries. These problems can be examined in the
following works: R. M. Bird and O. Oldman (eds.), Readings on Taxation in Developing
Countries, 3rd ed. (1975); R. M. Bird, Taxation of Agricultural Land in Developing Countries
(1974); John F. Due, Taxation and Economic Development in Tropical Africa (1963); John
F. Due, Indirect Taxation in Developing Countries (1970); A. R. Prest, Public Finance in
Underdeveloped Countries, 2nd ed. (1972); U. K. Hicks, Development Finance (1965); R. J.
Cheliah, Fiscal Policy in Underdeveloped Countries (1 970); Nicholas Kaldor, Indian Tax Re-
form (1956); A. M. Martin and W. A. Lewis, "Patterns of Public Revenue and Expenditure,"
The Manchester School of Economic and Social Studies (September 1956); U. Tun Wai,
"Taxation Problems and Policies of Underdeveloped Countries," IMF Staff Papers (Novem-
ber 1962); H. P. Wald, Taxation of Agricultural Land in Underdeveloped Economies (1959);
R. S. Thorn, "Evolution of Public Finances during Economic Development," The Manchester
School of Economic and Social Studies (January 1967); Donald R. Snodgrass, "The Fiscal
System as an Income Redistributor in West Malaysia," Public Finance (1974); Charles E.
McLure, "Taxation and the Urban Poor in Developing Countries," World Development
(March 1977); J. F. J. Toye (ed.), Taxation and Economic Development (1978).
Noteworthy research on tax issues and empirical studies of tax trends in developing coun-
tries have been undertaken by the Fiscal Affairs Department of the International Monetary
Fund and reported in various issues of IMF Staff Papers.
The appropriateness of different taxes for developing countries is often judged with refer-
ence to such criteria as (1) fruitfulness of revenue sources, (2) effects on resource allocation,
(3) contribution to equity, and (4) ease of administration. For an empirical study of these
criteria, see International Monetary Fund, Taxation in Sub-Saharan Africa, Occasional
Paper No. 8 (October 1981).
EXHIBIT IV.2. Forty-Seven Developing Countries: Tax Ratios 1969-71 and 1972-76

Taxes as Percent of GNP Ranking

Country
1972-76 1969-71 1969-71
(6)
Iran 32.7 21.6
Guyana 31.7 23.4 (3)
(1)
Zambia 30.8 31.3
Zaire 27.2 29.4 (2)
Venezuela 23.1 20.4 (7)
Malaysia 22.5 19.3 (11)
Trinidad and Tobago 21.9 17.7 (17)
Tunisia 20.7 21.7 (5)
Ivory Coast 20.6 19.8 (8)
Senegal 20.2
18.1 (14)
China, Republic of 19.9 17.8 (15)
Kenya 19.2 14.4 (12)
Jamaica 1910 (10)
18.9 19.4
Tanzania 13.9
Sudan 18.9 18.2 (24)
(13)
Morocco 18.6 17.8 (15)
19.6 (9)
Chile 18.4
Egypt 18.1 (12)
19.2
Brazil 18.1 22.9 (4)
Sri Lanka 17.9
17.7 (17)
Indonesia 16.3 10.0 (39)
Turkey 16.2 15.6
(20)
Singapore 15.7 13.2 (28)
Ghana 14.2 15.8
14.0 (19)
Peru 14.2
(23)
Thailand 13.9 12.4 (32)
India 13.9 13.4 (25)
Costa Rica 13.6 13.1 (30)
Korea 13.6 15.4 (21)
Argentina 13.3 13.4 (25)
Mali 12.9 13.2 (28)
Togo 12.4 11.3 (34)
Ecuador 12.0 13.4
11.8 (25)
Bolivia
8.2 (43)
Colombia 11.6 12.5 (31)
Honduras 11.5 11.3
(34)
Pakistan 11.4
11.3 8.8 (41)
Upper Volta 10.3 (38)
Lebanon 10.2 11.2 (36)
Philippines 10.1 9.1
(40)
Ethiopia 10.1 8.6
7.9 (42)
Rwanda 0.0 (44)
Burundi 9.3 11.4
Paraguay 8.8 10.9 (33)
(37)
Mexico 8.6 7.1
(46)
7.9
Guatemala (44)
8.1
Nepal 4.4 (47)
5.4
Average 16.1 15.1

Source: IMF Staff Papers, March 1979, p. 130

 IV.C. FINANCIAL DEEPENING
IV. C. 1 . Financial Structure and
Economic Development*

During economic development, as their in- cent, in more prosperous countries such as
comes per capita increase, countries usually Argentina, Brazil, Guatemala, Mexico, the
experience more rapid growth in financial Republic of Korea, Venezuela, and Yugo-
assets than in national wealth or national slavia. India, though less developed than
product.1 During the past century and a half, most of these countries, has a financial ratio
this has been true for the United States. Fi- of around 35 percent. In still more highly de-
nancial assets there were only about one-half veloped countries, the proportion of financial
the level of national wealth in the early to real wealth often lies in the range of 80 to
1880's, but they have been considerably 100 percent. France, Israel, West Germany,
larger than real wealth in recent years. Fi- and several other countries are in this range.
nancial assets have grown much faster than However, the Soviet Union has a low finan-
gross national product in the United States: cial ratio for its income per capita, about 35
the ratio increased from about unity at the percent, while Japan (150 percent), Switzer-
beginning of the last century to 4.5 now. land (over 200 percent), and the United
Japan has had a similar experience. There Kingdom (215 percent) have exceptionally
the ratio of financial assets to real wealth rose high ones.
from pehaps 10 percent in 1885 to over 150 At the present time, national stocks of fi-
percent in recent years. In the Soviet Union, nancial assets vary from 10 to more than 200
this ratio moved up from 10 percent in 1928 percent of national real wealth. As the fore-
to 35 percent a few years ago. Financial going examples suggest, differences in in-
growth in excess of real growth is apparently come per capita go a long way toward ex-
a common phenomenon around the world. plaining this variety of financial experience.
In somewhat rougher outline, this same The United States does have a higher finan-
picture is revealed by comparison among cial ratio than France, France than Mexico,
countries at any moment of time. Countries and Mexico than Afghanistan. Over time, for
that are poor in income per capita generally any one country as its income per capita in-
have very low ratios of financial to real creases, financial assets rise relative to na-
wealth. The present ratio in Afghanistan and tional real wealth.
Ethiopia, for instance, is probably little
higher than 10 or 15 percent, comparable to
the Japanese ratio in 1885 but lower than the REASONS FOR SECULARLY
ratio in the United States 1 50 years ago. The RISING FINANCIAL RATIOS
ratio is somewhat higher, from 30 to 60 per-
The relationship between growth in finan-
*From John G. Gurley and Edward S. Shaw, "Fi- cial assets and growth in real wealth and in-
nancial Development and Economic Development," come per capita may be analyzed in various
Economic Development and Cultural Change, Vol.
15, No. 3, April 1967, pp. 257-65, 267-8. Reprinted ways. For example, financial development
by permission. depends on division of labor that is feasible
'The measurements we use for relative growth in only in the context of real development.
financial assets, tangible assets, and income are based Again, financial development depends on
consistently on nominal values except when deflated
values are mentioned explicitly. Financial assets in-
conditions of demand for and supply of finan-
clude all intangible assets — claims against both non- cial assets that are sensitive to real develop-
financial spending units (primary securities) and fi- ment. We consider each of these approaches
nancial institutions (indirect securities). . . . in this section.

240
 241 FINANCIAL DEEPENING

Finance and self-sufficiency are antonyms: ultimate lenders (savers). In between lie the
Robinson Crusoes do not accumulate debt markets on which primary securities are
and financial assets. Finance is associated bought and sold. During the growth process,
with division of labor in three senses: (1) di- the division of labor between saving and in-
vision oflabor in production that involves ex- vesting becomes more intricate, and this in-
change offactor services and outputs implies stitutional evolution implies more rapid ac-
lending and borrowing. In primitive market cumulation of primary debt and financial
economies, these financial transactions are assets than of real wealth. Financial accu-
in kind. Subsequently, the diseconomies mulation falls into step with real accumula-
of finance-in-kind induce monetization. It tion when the institutional evolution ap-
seems to be the general rule that the pace of proaches itslimit. Similarly, factor inputs of
monetization exceeds the pace of real growth the financial sector increase more rapidly for
in diminishing degree. There is everywhere a a time than inputs of other sectors, but bal-
limit, in terms of real wealth and income per anced growth of inputs comes eventually.
capita, beyond which the stock of means of Primary security issues approximate 1 or 2
payment grows in step with wealth and in- percent of gross national product in the
come or even less rapidly. When money pay- poorer countries. The ratio generally lies
ments are ubiquitous, the ratio of money to within the range of 10 to 15 percent in
income or of money to wealth hovers near to wealthier countries. A similar contrast
its secular peak. emerges from data for the United States: the
In the United States, the ratio of the issues-income ratio was apparently small
money supply to gross national product was during the early years of the nineteenth cen-
only 7 percent in 1805. It doubled by 1850 tury, but it rose to more than 10 percent by
and doubled again by 1900. Thereafter, the the close of the century, and there it remains.
money-income ratio has shown no strong During the nineteenth century in the United
trend up or down: it is still around 30 percent. States, the stock of primary securities more
A similar pattern appears in cross-section than doubled relative to national wealth and
comparisons. On the basis of data for 70 income, and then changed little after 1900.
countries since World War II, one observes (3) There is division of labor in a third
that the money-income ratio varies from ap- sense that promotes growth in both quantity
proximately 10percent or less in the poorest and variety of financial assets. Specialization
countries to 20 percent in countries with develops between saving and ownership of
gross national product per capita of about primary securities. Financial intermediaries
$300. Then the rise in the ratio is retarded, solicit savings, paying a deposit rate for
though it continues to 30 percent and a little them, and assume responsibility for savings
more. allocation, charging a primary rate of inter-
(2) Finance is associated with division of est to ultimate borrowers. The spread be-
labor between saving and investment. Where tween the primary rate and the deposit rate
it is one sector in the community that releases compensates for factor costs and risks in in-
factors of production from consumer goods termediation. This spread shrinks during fi-
industries and another sector that absorbs nancial development. In combination with
such factors into accumulation of real capi- shifts in savers' tastes among financial assets,
tal, financial assets and debt accumulate in the relative rise in the deposit rate induces
both monetary and nonmonetary form. The layering of indirect debt upon primary debt
rate of accumulation depends on the mix and growth in total financial assets of savers
of techniques for transferring savings to relative to national income and wealth.
investment. There are countries which develop a com-
This division of labor leads to issues of pri- parative advantage in intermediation. They
mary securities by ultimate borrowers (inves- import primary securities and export indirect
tors) and to acquisitions of financial assets by debt, earning the spread between the primary
 242 MOBILIZING DOMESTIC RESOURCES

rate of interest and the deposit rate. Coun- controls on markets for financial assets.
tries with a comparative disadvantage in in- Briefly, both income and price phenomena
termediation prefer smaller factor inputs for associated with real development stimulate
the domestic financial sector. Exporters of in- finance.
termediation have relatively high and rising The retroactive impact of finance upon the
ratios of total financial assets to income and real world need not be explored carefully
wealth, but their portfolios of indirect finan- here. Anything that the financial sector does
cial assets follow a rising trend. to accelerate savings, improve their alloca-
Stocks of nonmonetary indirect financial tion to investment, and economize costs in
assets, including both domestic and foreign transmitting savings to investment implies an
claims, rise almost without let-up during the increase both in its own flows and stocks of
development process. They were a negligible financial assets and in flows and stocks of real
fraction of gross national product in the output.
United States at the beginning of the nine- There are degrees and kinds of differences
teenth century. They rose almost continu- in national financial systems that cannot be
ously from that time, reaching 35 percent of explained by differences in income and
gross national product by 1900 and 60 per- wealth. . . .
cent in the last few years. Cross-section data
have very much the same story to tell: time
ALTERNATIVE TECHNIQUES FOR
and savings deposits, which comprise the MOBILIZING THE ECONOMIC
bulk of nonmonetary indirect assets, are only SURPLUS
1 or 2 percent of gross national product in the
poorest countries, 10 percent in countries The principal reason for dissimilar finan-
with national products per capita of about cial structure at given national levels of in-
$300 to $400, and 40 or 50 percent in the come and wealth is that there are alternative
richer nations.
techniques for mobilizing the economic sur-
We have seen that financial assets accu- plus— for eliciting savings and allocating
mulate as income and wealth grow. Our sug- them to investment. The financial technique,
gestion isthat such accumulation reflects di- or the debt-asset system, is only one method.
vision of labor in production, saving and Each of the other techniques is a substitute
investment, and intermediation. Specializa- for it.
tion in the use of productive factors generates In its full detail, the list of alternative pro-
a rising stream of income and a rising stock cesses for putting saving to the service of se-
of both real and financial wealth. lected investment is very long. However, the
The coincidence and interaction of real list may be compressed into two major
and financial growth can be explained also in classes: processes of internal finance and pro-
terms of supply and demand conditions on cesses of external finance. In the former, the
real and financial markets. The income elas- investor draws on his own savings. In the lat-
ticity of savings is greater than unity at low ter, the investor draws on the savings of
levels of income, and the savings elasticity of others.
demand for financial assets en masse seems
Internal finance itself comprises two prin-
to remain above unity as income rises. cipal techniques: self-finance and taxation. In
Though the secular decline in the real rate of
interest that accompanies development tends self-finance, savings are put at the investors'
disposal by adjustments in relative prices on
to depress saving, trends in both explicit and commodity and factor markets and on mar-
implicit returns on financial assets and par- kets for foreign exchange. The taxation tech-
ticularly onindirect financial assets stimulate
nique employs taxes and other nonmarket al-
portfolio demands. These trends can be at- ternatives tochannel savings to the state for
tributed to technological change, to econo- either governmental or private investment.
mies of scale, and sometimes to regulatory Within external finance, the debt-asset sys-
 243 FINANCIAL DEEPENING

tern is the technique for mobilizing domestic would have said, of relative prices to the ad-
savings. In addition, savings from abroad vantage ofan investing sector, forcing invol-
may be supplied by gift or loan, on the initia- untary savings upon other sectors. These ad-
tive of either private or governmental justments inrelative prices may be imposed
sources. by government, socialist or capitalist, at a
We condense the long list of alternative stable price level. They may be induced by
processes to four: self-finance and taxation, government through inflation of the price
as internal finance processes; the debt-asset level. They may be imposed by private inves-
system and foreign aid, as external finance tors with monopoly power on some commod-
processes. This section has to do mainly with ity or factor market. The processes of self-fi-
internal finance, but there are first a few nance can operate through socialist central
comments on external finance to bring out planning, monopoly or state-directed capital-
the contrast between processes. ism, or inflation. . . .
Socialist extraction of savings depends
External Finance upon prices paid and received by state enter-
prise, to a lesser extent upon taxation, and in
The debt-asset system for mobilizing do- minor degree upon accumulation by private
mestic savings depends upon and encourages savers of financial claims, including money,
division of labor between savers and inves- against the state. Private solicitation of sav-
tors, as well as between savers and inter- ings for investment in private capital, includ-
mediaries. It belongs in the context of ing consumer durables, is unimportant,
decentralized decision-making, of market or- though expansion of consumer credit is to be
ganization, of dependence upon relative anticipated. Allocation of savings according
prices to guide economic behavior. Market to plan can be achieved by simple transfers to
rates of interest, as one class of relative state enterprise on the records of the state
prices, bear a heavy responsibility for the rate bank.
and direction of investment. Socialist centralism foregoes most of the
The issue and accumulation of government division of labor that we have stressed as one
debt is a component of the debt-asset system. basis for relatively high finance-income ra-
It draws private savings through security tios. Saving and investment are generated
markets and intermediaries to both private principally within the state sector, so that
and governmental uses. The stock of govern- market transfers of savings at explicit rates
ment debt has a role to play in portfolio di- of interest are as unnecessary as they are dis-
versification that affects deeply the perfor- tasteful insocialist doctrine. Demand for fi-
mance of the debt-asset system. nancial assets in other sectors is depressed by
constraints on personal income and wealth. It
Internal Finance is depressed, too, because the state under-
takes to supply various services for which
The processes of internal finance — self-fi- people save in capitalist societies, because
nance and taxation — are substitutes for the private bequests are minimized, and because
debt-asset system. Each of them involves reduced private risks imply reduced precau-
more centralization of decision-making, less tionary portfolios. Under these circumstances
specialization between savers and investors. of demand and supply, there is little occasion
Each of them leads to a less elaborate finan- for markets in either primary or indirect se-
cial structure. curities. Moreover^ because of pressure for
internal development, there has not been sig-
Self-finance. The processes of self-finance nificant accumulation by socialist societies of
involve adjustment in terms of trade on com- financial claims abroad, and there has been
modity, factor and foreign-exchange mar- no industry of financial intermediaries to ex-
kets— "opening the scissors," as Paul Baran port its services.
 244 MOBILIZING DOMESTIC RESOURCES

The process of socialist centralism reduces other techniques of mobilizing and allocating
the dependence of economic units on a finan- savings are underdeveloped. In the theory of
cial structure; contact is made less through structuralism, stagnation is the alternative if
financial markets and financial institutions inflation is not used to generate savings
and more through planning bureaus and among spending units with investment oppor-
other central coordinating devices; the order tunities. The "cost-push" variant of structur-
of the day is internal finance and balanced alist doctrine accepts inflation as a corrective
budgets, not external finance and the issue of for relative-price distortions that tend to re-
new securities. The theory and design of so- duce the flow of savings into the retained
cialist society are incompatible with rela- earnings of business and its self-financed cap-
tively high ratios of financial assets to income ital formation. Inflation is a second-best "in-
and wealth.
come policy" to prevent dissipation of savings
Manipulation of the terms of trade be- through consumption or underemployment.
tween agriculture and industry, whether It is inferior, the argument goes, to produc-
under socialism or capitalism, is a case of tivity ceilings on factor prices, but it is pref-
self-finance. It has been the socialist view erable to stagnation.
that a preindustrial society could not mobi- Inflation in prospect affects anticipated
lize its surplus effectively without collectivi- relative yields on various types of wealth, fi-
zation of peasant farming. Only then could nancial and real, reducing real rates of return
prices of agriculture be reduced relatively for on assets with the more inflexible nominal
the benefit of industrial investment. Or, as rates and raising real prospective returns on
one knows from American experience, the assets with more flexible nominal rates. The
scissors can be opened in the farmer's behalf inevitable result is to change rates of saving,
to increase his savings for his investment in allocation of savings, and channels by which
agricultural capital. savings flow. Self-finance gains at the ex-
Factor prices are an obvious target for pense of debt-asset finance. Foreign aid
techniques of self-finance. Peasant migration through intergovernmental loans and gifts
to the city depresses real wage rates and may gain if inflation reduces domestic saving
raises real profit rates, with the consequence of the recipient country below a target level
that industry has access to flows of new fi- of investment. The channels that lose are
nance for its own capital. Controls on rates debt-asset finance and also taxation, unless
charged by public utilities under foreign tax rates happen to be progressive to
ownership can extract involuntary savings inflation.
from abroad. Inflation transfers net worth as well as in-
Overvaluation of domestic currency on the come. Ittaxes such forms of financial wealth
foreign exchanges may transfer real income as money balances, assessing creditors for the
from an exporting sector, with taste for im- benefit of debtors, public or private. If the re-
ports of consumption goods or of foreign fi- sult isto depress the wealth position of savers
nancial assets, to an importing sector that is and the debt position of investors below pre-
accumulating plant and equipment. Depend- ferred levels, a temporary acceleration of sav-
ing upon relevant elasticities of demand and ing, investment, and debt-asset finance can
supply, overvaluation can also appropriate ensue. Needless to say, this result does not
foreign savings for domestic use. follow for gross inflation that is expected to
Inflation is a technique of self-finance. Be- continue.
cause of its impact on relative prices, relative Self-finance is centralist. It has been com-
incomes, and relative wealth positions, infla- mon in precapitalism, before the decentral-
tion appeals to "structuralists" as a tempo- izing evolution of commodity, factor, and
rarily necessary alternative in some retarded financial markets. It is inherent in anti-capi-
societies where relative prices tend to move talism, or socialism. Even in the general con-
perversely for purposes of growth and where text of capitalism, there occur changes in the
 245 FINANCIAL DEEPENING

socioeconomic structure that involve rever- real consumption. Given a social discount
sion to self-finance. Degeneration in pro- rate, the technology or combination of tech-
cesses of debt-asset finance reduces saving nologies for eliciting and allocating saving is
and investment, but it can also divert funds best which implies highest consumer welfare,
into the channels of self-finance. Self-finance counting government among consumers. The
may increase if business firms combine to optimal consumption stream has the qualities
avoid encounters, as Baran puts it, with "the of equity and stability that conform to a
social welfare function.
restraining hand of financial institutions."
Cooperative credit among, say, consumers or The contribution of each saving-invest-
farmers has been at times a way of detaching ment technology to consumer welfare has a
relatively small, homogeneous groups from positive (gross yield) component and a nega-
reliance on broader security markets. The tive (factor cost) component, and the differ-
mujin and kye of the Far East, "traditional" ence is the net yield of the technology.
credit institutions, are illustrations of coop- Additional real resources applied to a saving-
erative saving-investment processes that bor- investment technology may raise rates of sav-
der on both self-finance and the debt-asset ing and investment, the nation's capital
system. stock, and hence the future flow of consump-
tion. The gross yield of these real resources is
Taxation. The tax technique is a variant the capital value of the economy's antici-
of internal finance, a way of mobilizing the pated gross additions to its consumption
economic surplus that implies centralized de- stream. Factor cost is the capital value of the
cision-making. While it is true that there are stream of final goods that could have been
taxes which work to the benefit of debt-asset produced by the real resources if they had not
finance, there can be a structure of taxation been diverted to a saving-investment technol-
which depresses demand for financial assets, ogy. The net yield is the difference between
displaces issues of securities by investors and these two capital values.
intermediaries, and makes securities markets We suggest that there is an optimal com-
superfluous. The receipted tax bill is a sub- bination of saving-investment technologies
stitute for the financial asset in spending for each economy in each phase of its devel-
units' portfolios. opment. This optimal combination has been
attained when no gain in net yield can result
from shifting factor inputs between technol-
CRITERIA FOR CHOICE AMONG ogies. Then no transfers of resources between
ALTERNATIVE SAVING- technologies can advance the possibility fron-
INVESTMENT TECHNIQUES tier of anticipated consumption. If the opti-
mal combination varies with the aggregate of
We have defined four technologies or pro- factor inputs into all saving-investment tech-
cesses ofeliciting and allocating savings: self- nologies, one may wish to define the best
finance, taxation, debt-asset, and foreign aid. combination more narrowly, as that which
We have suggested that differences in the prevails when net yield is zero at the margin
ratio of financial assets to income and wealth, for all technologies. . . .
between countries and between times for any
one country, may reflect differences and Bias and Evolution in Choice
changes in choice between the debt-asset Between Technologies
technology and the others. In the present sec-
tion, we consider some criteria for choice Economic development is marked by iter-
among the four technologies. ative probing for the optimal combination of
The objective of public policy regarding saving-investment technologies. The search is
the saving-investment process, we assume, is guided in part by principle and prejudice, in
to maximize the capital value of anticipated part by foreign example, by trial and error,
 246 MOBILIZING DOMESTIC RESOURCES

and even by rational analysis. The combina- by undervaluation of human capital. It is in-
tions and permutations of technology are so herently unstable in the short run: variance
numerous that no two countries are likely to in the future consumption stream is high. It
follow the same probing sequence or, specifi- reeks of inequity. Needless to say, the capi-
cally, to reach the same ratio of financial as- talist's view of socialist central planning is no
sets to tangible wealth at any given level of less emphatic, using the same criteria of cap-
real wealth or income per capita. italized yield and cost.
The probing sequence is most deeply af- As development proceeds, these blacks and
fected bythe economy's choice between self- whites of choice between technologies appear
finance with central planning and technolo- to shade off to greys of varying intensity. The
gies that are compatible with decentralized probing process everywhere seems to lead to-
decision regarding saving and investment. ward a mixture of self-finance under central
This is a choice between Left and Right. Ad- planning and decentralized processes. At the
vocates of either can design imposing proof margin, neither black nor white has a long
that it is cheaper in resource inputs, more ef- life expectancy. If we dared to suggest a Law
fective ineliciting and allocating savings, less of Financial Development, it would be this:
vulnerable to instability in the consumption each economy begins its development by in-
stream, and more compatible with the ethics tensive exploitation of a saving-investment
of equity. technology that is chosen for historical, polit-
As the socialist sees decentralized finance, ical, social, or perhaps economic reasons, and
it is exorbitantly expensive in marble col- then, as this technology produces a diminish-
umns, bank presidents, and energy consumed ing net yield, experiments with alternative
in managing portfolios. It elicits too little in technologies that are marginally superior in
savings, because monopoly finance does not terms of their capitalized returns and costs.
pay savers enough, because decentralized de- Whatever the first choice may be, it is tilled
cisions pay too little attention to the con- intensively until there is obvious advantage in
sumer welfare of succeeding generations, or trying the extensive margin that involves a
because savers do not measure accurately the mix of processes for eliciting and allocating
rewards to saving. It allocates savings savings. Along the extensive margin, a so-
badly — on ostentation, on consumer dura- cialist society may tolerate market as well as
bles, on military capital. Its allocation is shadow rates of interest, and a capitalist so-
biased by indifference to externalities of pri- ciety may put up with some centralized self-
vate investment, by hostility to public goods, finance.

IV. C. 2. Organized and

Unorganized Money Markets*

The size of an organized money market in both of the following ratios, although neither
any country may be indicated by either or measurement is perfect: the ratio of deposit
money to money supply and the ratio of the
*From U Tun Wai, "Interest Rates in the Orga- banking system's claims (mostly loans, ad-
nized Money Markets of Underdeveloped Countries," vances, and bills discounted) on the private
International Monetary Fund Staff Papers, Vol. 5, sector to national income. . . .
No. 2, August 1956, pp. 249-50, 252-3, 255, 258,
The ratio of deposit money to money sup-
276-8; "Interest Rates Outside the Organized Money
ply actually measures banking development
Markets of Underdeveloped Countries," ibid., Vol. 6,
No. 1, November 1957, pp. 80-83, 107-9, 119-25. of the money market. However, to the extent
Reprinted by permission. that the development of commercial banking
 247 FINANCIAL DEEPENING

is synonymous with the development of the amounts of financial assets are available to
money market, this ratio may be used as an serve as collateral for lending at low rates. It
indicator of the growth of a money market. is usually the foreign business firms with
In most underdeveloped countries, there are longer experience and larger capital which
hardly any lending agencies of importance are able to borrow at the lower rates. Most of
other than commercial banks. There are no the indigenous firms have to pay the higher
discount houses or acceptance houses, and rates; this is especially true where foreign
savings institutions (including life insurance banks occupy an important position in the
companies) are in the early stages of devel- banking system. . . .
opment. . . In spite of the small direct dependence of
Both ratios might be expected to be low in commercial banks on the central banks for
an underdeveloped country and high in a de- funds, the latter are able to influence market
veloped one. The ratio of deposit money to rates by changes in the bank rate because of
money supply should be higher in a more de- their economic, and at times their legal, po-
veloped country because, with economic de- sition in the domestic money market, with
velopment, there is also development of the wide powers for selective credit control, open
banking system. . . . market operations, and moral pressures.
The structure of interest rates in the or- The general expectation is that the long-
ganized money markets of underdeveloped term trend of interest rates in underdevel-
countries is usually more or less the same as oped countries, at least in the organized
in the developed ones. The short-term rate of markets, should be downward. Generally
interest is generally much below the long- speaking, in these countries the banking sys-
term rate, as indicated by the spread between tems and with them the money markets are
the government treasury bill rate and the likely to develop at a faster rate than the
government bond yield; the rate at which other sectors of the economy. The long-term
bills of exchange are discounted is also lower supply of loanable funds therefore tends to
than the rate at which loans and advances are increase more rapidly than the long-term de-
granted. mand. Where, for one reason or another, the
The lowest market rates are usually the growth of banking has been restricted or the
call loan rates between commercial banks. banking system subjected by law to many re-
The next lowest are those paid by commer- strictions, including controls on interest rates
cial banks on short-term deposits, followed and of the purposes for which loans may be
by the government treasury bill rate. Then granted (as in a number of countries in Latin
come the rates at which commercial banks America), the long-term trend of interest
discount commercial paper, varying accord- rates may, however, not be downward. . . .
ing to the type of security and the date of ma- In [the above] examination ... of the in-
turity. In most countries, especially in Asia, terest rate structure and the lending practices
the government bond yield comes next, fol- of organized money markets in underdevel-
lowed by the lending rates of commercial oped countries, ... it was shown that these
banks. . . . differed much less than might have been ex-
In general, the level of interest rates in pected from those prevailing in most devel-
underdeveloped countries, even in organized oped countries. In underdeveloped countries,
money markets, is higher than in the more however, unorganized money markets also
developed countries. The more notable differ- play a very important role, and any study of
ence between the two groups of countries, credit conditions in these countries that is to
however, is that the range of interest rates is be adequate must be extended to cover the
generally much wider in underdeveloped unorganized as well as the organized mar-
countries. The volume of loans granted at rel- kets. Efforts have often been made to repair
atively low rates in an underdeveloped coun- the deficiencies of the unorganized markets
try is not very important, as only limited by government action designed to stimulate
 248 MOBILIZING DOMESTIC RESOURCES

the development of cooperative credit or to loanable funds indeed reflects the general
provide credit through agricultural banks, shortage of capital in underdeveloped
etc.; it is convenient to include these govern- countries.
ment-sponsored institutions in a study of un- The disadvantages of the high rates of in-
organized money markets in general. terest inthe unorganized money markets are
Interest rates in the unorganized money well known and include such important
markets of underdeveloped countries are gen- effects as "dead-weight" agricultural indebt-
erally very high in relation both to those in edness, alienation of land from agricultural-
the organized money markets and to what is ists to moneylenders and the agrarian unrest
needed for rapid economic development. that is thus engendered, and a general slow-
These high interest rates are caused by a dis- ing down of economic development. . . .
proportionately large demand for loanable The organized money markets in under-
funds coupled with a generally inelastic and developed countries are less fully integrated
limited supply of funds. The large demand than the money markets in developed coun-
stems from the special social and economic tries. The unorganized money markets in
factors prevalent in the rural areas of under- underdeveloped countries are even more im-
developed countries. The low level of income perfect, and indeed it is questionable whether
leaves little surplus for saving and for the ac- the existing arrangements should be referred
cumulation ofcapital for self-financing of ag- to as "markets." They are much less homo-
ricultural and handicraft production. The un- geneous than the organized markets and are
certainty ofthe weather, which affects crop generally scattered over the rural sector.
yields and incomes, causes an additional need There is very little contact between the lend-
for outside funds in bad years. A significant ers and borrowers in different localities. The
portion of the demand for loanable funds in usual textbook conditions for a perfect mar-
rural areas is for financing consumption at ket are completely nonexistent: lenders and
levels much higher than are warranted by the borrowers do not know the rates at which
low income of the peasant. . . . loans are being transacted in other parts of
The supply of loanable funds in the unor- the country; the relationship between bor-
ganized money markets is very limited and rower and lender is not only that of a debtor
inelastic because the major source is the and creditor but is also an integral part of a
moneylender, and only very small quantities much wider socioeconomic pattern of village
are supplied by indigenous bankers and or- life and rural conditions.
ganized institutions, such as cooperative In unorganized money markets, moreover,
credit societies and land mortgage banks. loans are often contracted and paid for not in
The moneylender in most cases is also a mer- money but in commodities; and the size of
chant or a landlord and therefore is willing to the average loan is very much smaller than in
lend only at rates comparable with what he the organized money markets. Both borrow-
could earn by employing his capital in alter- ers and lenders in the two markets are often
native uses which are often highly profitable. of quite different types. In the organized
The lenders in the unorganized money mar- money markets, the borrowers are mainly
kets do not have the facilities for mobilizing traders (wholesale and retail) operating in
liquid funds available to commercial banks in the large cities and, to a less extent, manu-
organized markets and therefore the supply facturers. Agriculturalists rarely account for
of funds is rather inflexible. Since the unor- a significant portion of demand except in
ganized money markets are generally not those underdeveloped countries where export
closely connected with the organized money agriculture has been developed through plan-
markets, there is little possibility of increas- tations or estates. In the unorganized money
ing the supply of loanable funds beyond the markets, the borrowers are small agricultur-
savings of the lending sector of the unorga- alists, cottage industry workers, and some
nized money markets. The limited supply of retail shopkeepers. The lenders in the orga-
 249 FINANCIAL DEEPENING

nized money markets consist almost demand for loanable funds to be large in re-
exclusively of commercial banks. In the un- lation to the available supply. Some writers
organized markets, the suppliers of credit tend to explain the high rates of interest in
consist of a few financial institutions, such as termssizeofsupply.
demand factors while others empha-
cooperatives, private and government-spon-
sored agricultural banks, indigenous bankers, The demand for funds, in relation to the
professional moneylenders, large traders, supply, is large because the average borrower
landlords, shopkeepers, relatives, and friends. in the unorganized money market has a very
Proper records of loans granted or repaid are low income and therefore has no surplus
usually not kept, and uniform accounting funds to finance his business operations. The
procedures are not adopted by the different majority of the cultivating tenants — one of
lenders. Loans are granted more on a per- the most important groups of potential bor-
sonal basis than in the organized money mar- rowers— have to borrow money not only for
kets, and most of the loans granted by the investment in land, cattle, etc., and for work-
moneylenders and by other noninstitutional ing capital to make purchases of seeds and
sources are unsecured beyond the verbal fertilizers, but also for their minimum basic
promise of the borrower to repay. necessities of food, shelter, and clothing.
The unorganized money market may be di- On the supply side, there is a general short-
vided into three major parts: (1) a part in age of capital in underdeveloped countries
which the supply is dominated by indigenous and an inadequate level of domestic savings.
bankers, cooperatives, and other institutions, Also, the small amount of domestic savings is
and the demand by rural traders and me- not channeled effectively into the unorga-
dium-sized landlords; (2) a part in which the nized money market because of the absence
demand originates mainly from small agri- of proper financial and credit institutions
culturalists with good credit ratings, who are which not only would integrate the organized
able to obtain a large portion of their funds and unorganized money markets but also
from respectable moneylenders, traders, and would facilitate the mobilization of savings in
landlords at high but reasonable rates of in- the rural areas. . . .
terest, i.e., rates that are high in relation to The difference in the levels of interest rates
those prevailing in the organized money mar- between the organized and unorganized
ket but not exorbitant by the standards of the money markets stems partly from the basic
unorganized money market; (3) a part in differences between the sources of supply of
which the demand originates from borrowers funds in the two markets. In an organized
who are not good credit risks, who do not money market, facilities for the expansion of
have suitable collateral, and who in conse- credit are open to the commercial banks,
quence are driven to shady marginal lenders which have the use of funds belonging to de-
who charge exorbitant rates of interest positors. These banks are therefore able to
Many explanations have been offered for charge relatively low rates of interest and yet
the high interest rates that generally prevail make satisfactory profits for the sharehold-
in unorganized money markets. One theory is ers. On the other hand, moneylenders in an
that interest rates are high there because unorganized money market have little influ-
they are determined by custom and have ence on the supply of funds at their disposal
always been high. This might be called the and, furthermore, their supply price tends to
theory of the customary rate of interest. . . . be influenced by the alternative uses to which
The theory of customary rates is not sat- their funds can be put.
isfactory, however, because it does not A number of institutional factors are also
explain how or why the custom of high rates responsible for high rates of interest in un-
developed. The true explanation has to be organized money markets. The size of the
found in the economic and social conditions loan is usually small and thus the fixed han-
of underdeveloped countries, which cause the dling charges are relatively high. Defaults
 250 MOBILIZING DOMESTIC RESOURCES

also tend to be larger in unorganized money proper seeds; making available adequate
markets. These higher defaults are due not so marketing facilities, including proper grad-
much to a lower standard of morality and ing, transportation, and storage of crops; and
willingness to repay debts as to the fluctua- providing an efficient agricultural extension
tions in prices and incomes derived from ag- service. . . .
ricultural products, which reduce the ability Even if it is true that the cure for high
of the agriculturalists to repay debts at in- rates of interest is to be found more on the
opportune times. . . . demand side than on the supply side, the sup-
Another general factor causing high rates ply of credit should also be increased. Supply
of interest is experience in regard to inflation. should be increased in such a way that legit-
While this is probably of importance in many imate credit needs are met at cheaper rates
Latin American countries, it hardly seems without encouraging borrowing for consump-
relevant for prewar colonial territories which, tion. This can be achieved by increasing the
by their rigid currency exchange standards, supply of institutional credit while at the
had maintained fairly stable conditions but same time taking steps to discourage borrow-
still had high interest rates in the unorga- ing from noninstitutional lenders. In this con-
nized money markets. The large development nection, icould
t be argued that legislation re-
programs in most underdeveloped countries, garding moneylenders which has had the
however, constitute possible inflationary effect of drying up noninstitutional credit
pressures and may be considered as a factor may be a blessing in disguise — although in a
in maintaining high rates of interest. manner different from that intended by
The list of causes of high interest rates legislators.
could be extended to include other social and Increasing the supply of institutional
economic factors in underdeveloped coun- credit is a difficult problem, but the efforts of
tries— even to fairly remote factors, such as governments have had a fair degree of suc-
the system of land tenure which prevents cess. One problem is that of getting the com-
land from being used as collateral. A general mercial banks to lend more to agriculture.
statement, however, is that interest rates in This problem cannot be solved merely by
the unorganized money markets of underde- opening more branches, because even at pres-
veloped countries are high because the econ- ent agriculturalists who are fairly close to the
omy is underdeveloped and the money mar- big cities are as isolated from the organized
ket unorganized. . . . money market as others living some distance
Any program to bring down interest rates away from the cities. The opening of more
in unorganized money markets must be com- bank branches is desirable, but the branches'
prehensive and should be guided by the prin- business will as likely as not be confined to
ciple that interest rates can be lowered only financing local retail and wholesale trade.
by reducing the demand for loanable funds as One way of inducing the organized finan-
well as by increasing the supply. . . . cial institutions to lend more to agriculture is
A reduction in borrowing for productive by making agriculturalists more creditwor-
purposes may not be desirable, especially as thy and generally reducing the risks of lend-
the amount of self-financing which can take ing by lessening the impact of some of the
its place is negligible. Such borrowing can be natural calamities (floods, plant and animal
reduced in the long run only through an in- diseases); improving the human factor, i.e.,
crease in savings from higher agricultural reducing carelessness and increasing honesty;
output and income. It is not sufficient that the reducing the uncertainties of the market
ability of the farmer to save be increased. through crop insurance, stabilized agricul-
The willingness to save must also be created. tural prices, etc. The lenders might also take
The problem of cheap agricultural credit is certain steps, such as spreading loans be-
inseparable from the whole problem of agri- tween different types of borrower and region
cultural development, including such mea- and supervising the use of loans for produc-
sures as increasing the use of fertilizers and tive purposes.
 251 FINANCIAL DEEPENING

IV.C.3. Demand-Following or
Supply-Leading Finance*

Typical statements indicate that the financial tization of agriculture and other traditional
system somehow accommodates — or, to the subsistence sectors. The more rapid the
extent that it malfunctions, it restricts — growth rate of real national income, the
growth of real per capita output. For greater will be the demand by enterprises for
example, external funds (the saving of others) and
therefore financial intermediation, since
It seems to be the case that where enterprise leads under most circumstances firms will be less
finance follows. The same impulses within an econ-
omy which set enterprise on foot make owners of able to finance expansion from internally
wealth venturesome, and when a strong impulse to generated depreciation allowances and re-
invest is fettered by lack of finance, devices are in- tained profits. (The proportion of external
vented to release it . . . and habits and institutions funds in the total source of enterprise funds
are developed.1 will rise.) For the same reason, with a given
Such an approach places emphasis on the de- aggregate growth rate, the greater the vari-
mand side for financial services; as the econ- ance in the growth rates among different sec-
tors or industries, the greater will be the need
omy grows it generates additional and new
demands for these services, which bring for financial intermediation to transfer sav-
about a supply response in the growth of the ings to fast-growing industries from slow-
growing industries and from individuals. The
financial system. In this view, the lack of fi- financial system can thus support and sustain
nancial institutions in underdeveloped coun- the leading sectors in the process of growth.
tries issimply an indication of the lack of de-
mand for their services. The demand-following supply response of
the growing financial system is presumed to
We may term as "demand-following" the come about more or less automatically. It is
phenomenon in which the creation of modern
financial institutions, their financial assets assumed that the supply of entrepreneurship
and liabilities, and related financial services in the financial sector is highly elastic relative
is in response to the demand for these services to the growing opportunities for profit from
by investors and savers in the real economy. provision of financial services, so that the
In this case, the evolutionary development of number and diversity of types of financial in-
the financial system is a continuing conse- stitutions expands sufficiently; and a favora-
quence of the pervasive, sweeping process of ble legal, institutional, and economic environ-
economic development. The emerging finan- ment exists. The government's attitudes,
economic goals, and economic policies, as
cial system is shaped both by changes in ob- well as the size and rate of increase of the
jective opportunities — the economic environ-
ment, the institutional framework — and by government debt, are of course important in-
fluences inany economy on the nature of the
changes in subjective responses — individual economic environment. As a consequence of
motivations, attitudes, tastes, preferences.
The nature of the demand for financial ser- real economic growth, financial markets de-
velop, widen, and become more perfect, thus
vices depends upon the growth of real output
increasing the opportunities for acquiring
and upon the commercialization and mone-
liquidity and for reducing risk, which in turn
*From Hugh T. Patrick, "Financial Development feeds back as a stimulant to real growth.2 ,
and Economic Growth in Underdeveloped Coun- The demand-following approach implies
tries," Economic Development and Cultural Change, that finance is essentially passive and permis-
Vol. 14, No. 2, January 1966, pp. 174-7. Reprinted
by permission. sive in the growth process. Late eighteenth
'Joan Robinson, "The Generalization of the Gen-
eral Theory," in The Rate of Interest and Other 2Cf. W. Arthur Lewis, The Theory of Economic
Essays. London, 1952, pp. 86-7. Growth, London, 1955, pp. 267-86.
 252 MOBILIZING DOMESTIC RESOURCES

and early nineteenth century England may sible alternatives, enabling the entrepreneur
be cited as a historical example. In fact, the to "think big." This may be the most signifi-
increased supply of financial services in re- cant effect of all, particularly in countries
sponse to demand may not be at all auto- where entrepreneurship is a major constraint
matic, flexible, or inexpensive in underdevel- on development. Moreover, as has been em-
oped countries. Examples include the phasized byRondo Cameron,4 the top man-
restrictive banking legislation in early nine- agement of financial institutions may also
teenth century France, religious barriers serve as entrepreneurs in industrial enter-
against loans and interest charges, and Ger- prises. They assist in the establishment of
schenkron's analysis of the abortive upswing firms in new industries or in the merger of
of Italian industrial development in the firms (the advantages of economies of scale
1880's "mainly, it is believed, because the may be more than offset by the establishment
modern investment bank had not yet been es- of restrictive cartels or monopolies, however),
tablished inItaly."3 In underdeveloped coun- not only by underwriting a substantial por-
tries today, similar obstacles, together with tion of the capital, but more importantly by
imperfections in the operation of the market assuming the entrepreneurial initiative.
mechanism, may dictate an inadequate de- By its very nature, a supply-leading finan-
mand-following response by the financial sys- cial system initially may not be able to oper-
tem. The lack of financial services, thus, in ate profitably by lending to the nascent mod-
one way or another restricts or inhibits effec- ern sectors.5 There are, however, several
tive growth patterns and processes. ways in which new financial institutions can
Less emphasis has been given in academic be made viable. First, they may be govern-
discussions (if not in policy actions) to what ment institutions, using government capital
may be termed the "supply-leading" phe- and perhaps receiving direct government sub-
nomenon: the creation of financial institu- sidies. This is exemplified not only by Rus-
tions and the supply of their financial assets, sian experience in the latter half of the nine-
liabilities, and related financial services in teenth century, but by many underdeveloped
advance of demand for them, especially the countries today. Second, private financial in-
demand of entrepreneurs in the modern, stitutions may receive direct or indirect gov-
growth-inducing sectors. "Supply-leading" ernment subsidies, usually the latter. Indirect
has two functions: to transfer resources from subsidies can be provided in numerous ways.
traditional (non-growth) sectors to modern Commercial banks may have the right to
sectors, and to promote and stimulate an en- issue banknotes under favorable collateral
trepreneurial response in these modern sec conditions; this technique was more impor-
tors. Financial intermediation which trans- tant in the eighteenth and nineteenth centu-
fers resources from traditional sectors,
ries (national banking in Japan in the 1870's;
whether by collecting wealth and saving from wildcat banking in the United States) than it
those sectors in exchange for its deposits and is likely to be in present underdeveloped
other financial liabilities, or by credit crea- countries, where this right is reserved for the
lion and forced saving, is akin to the Schum- central bank or treasury. Nonetheless, mod-
peterian concept of innovation financing. ern equivalents exist. They include allowing
New access to such supply-leading funds private financial institutions to create deposit
may in itself have substantial, favorable ex- money with low (theoretically, even negative)
pectational and psychological effects on en-
trepreneurs. Iopens
t new horizons as to pos-
4Rondo Cameron, "The Bank as Entrepreneur,"
Explorations in Entrepreneurial History, Series 2,
Vol. I, No. 1, (Fall 1963), pp. 50-5.
3Alexander Gerschenkron, Economic Backward- 5Except in the extreme case where inherent profit
ness in Historical Perspective — A Book of Essays, opportunities are very high, and supply-leading stim-
Cambridge, 1962, p. 363. See also Chapter 4. ulates a major entrepreneurial effort.
 253 FINANCIAL DEEPENING

reserve requirements and central bank redis- portant, and the demand-following financial
count of commercial bank loans at interest response becomes dominant. This sequential
rates effectively below those on the loans. process is also likely to occur within and
Third, new, modern financial institutions among specific industries or sectors. One in-
may initially lend a large proportion of their dustry may initially be encouraged finan-
funds to traditional (agricultural and com- cially on a supply-leading basis and as it
mercial) sectors profitably, gradually shifting develops have its financing shift to demand-
their loan portfolio to modern industries as following, while another industry remains in
these begin to emerge. This more closely re- the supply-leading phase. This would be re-
sembles the demand-following phenomenon; lated to the timing of the sequential devel-
whether such a financial institution is supply- opment of industries, particularly in cases
leading depends mainly on its attitude in where the timing is determined more by gov-
searching out and encouraging new ventures ernmental policy than by private demand
of a modern nature. forces.
It cannot be said that supply-leading fi- Japan between the 1870's and the begin-
nance is a necessary condition or precondi- ning of World War I presents an excellent
tion for inaugurating self-sustained economic example of the sequence of supply-leading
development. Rather, it presents an oppor- and demand-following finance. A modern
tunity to induce real growth by financial banking system was created in the 1870's,
means. It thus is likely to play a more signif- subsidized by the right to issue banknotes
icant role at the beginning of the growth pro- and by government deposits. These banks, in
cess than later. Gerschenkron implies that the absence of large-scale industrial demand
the more backward the economy relative to for funds, initially concentrated their funds
others in the same time period (and the on financing agriculture, domestic com-
greater the forced-draft nature of the eco- merce, and the newly important foreign
nomic development effort), the greater the trade. However, they also became the locus
emphasis which is placed on what I here term for much of the early promotional and entre-
supply-leading finance.6 At the same time, it preneurial talent which initiated the indus-
should be recognized that the supply-leading trial spurt beginning in the mid-1 880's, es-
approach to development of a country's fi- pecial y inrailroads and in cotton textiles (at
nancial system also has its dangers, and they first import-competing, and later export-
should not be underestimated. The use of re- oriented). The banks also became an early
sources, especially entrepreneurial talents important source of industrial funds, albeit
and managerial skills, and the costs of ex- via an indirect route. The modern financial
plicit or implicit subsidies in supply-leading system thus was not only created in advance
development must produce sufficient benefits of Japan's modern industrialization, but, by
in the form of stimulating real economic de- providing both funds and entrepreneurial tal-
velopment for this approach to be justified. ent on a supply-leading basis, contributed
In actual practice, there is likely to be an significantly to the initial spurt. By the mid-
interaction of supply-leading and demand- 1890's, the emphasis apparently moved from
following phenomena. Nevertheless, the fol- supply-leading to demand-following in the fi-
lowing sequence may be postulated. Before nancing of the textile and other consumer
sustained modern industrial growth gets un- goods industries. On the other hand, the fi-
derway, supply-leading may be able to in- nancing ofmost heavy manufacturing indus-
duce real innovation-type investment. As the tries continued on a supply-leading basis per-
process of real growth occurs, the supply- haps until World War I, with a considerable
leading impetus gradually becomes less im- portion of external funds provided through
the long-term loans of special banks estab-
lished at government initiative and utilizing
5Op. cit. government funds. . . .
 254 MOBILIZING DOMESTIC RESOURCES

IV. C. 4. Financial Repression

and Liberalization *

Most developing countries have formal plans systems and their unknown default potential.
for industrial development, public infrastruc- These savers, in any event, cannot easily di-
ture such as roads and utilities, the future versify their portfolios. Only a small and in-
course of education, foreign trade, agricul- adequate number of moneylenders, pawnbro-
ture, and so on. Yet planners typically do not kers, and village storekeepers can extend
set targets for, or even assess achievements limited credit directly to worthy private
in, the financial sector. Unfortunately, regu- borrowers.
lation rather than development is the usual The absence of open markets in primary
emphasis. securities means that the monetary system
However, no other sector in market-ori- has a much more important role as an inter-
ented economies is more important for pro- mediary between savers and investors. Pri-
moting indigenous entrepreneurship. Nas- vate financial savings in developing countries
cent business firms and artisans in cities or are largely currency and deposits — claims on
small farmers in the countryside do not func- central banks, commercial banks, and near
tion efficiently unless they can borrow or lend banks such as savings and loan associations,
freely on a quid pro quo basis. Only then is it financieras (development banks), postal sav-
possible to have broadly based development, ings depositories, and so on. These banking
with new technologies being widely intro- intermediaries issue liquid short-term depos-
duced, to reflect accurately the social scarcity its whose nominal value is virtually guaran-
of capital throughout the economy. . . . teed by the state, a great advantage to small
savers. This guarantee arises because society
THE BANKING SYSTEM AND must have a stable means of payment. The
LOANABLE FUNDS provision of money to provide these services
has long been recognized as one of the major
In developing countries, open markets for attributes of sovereignty. It was accepted as
primary securities are usually insignificant. a state responsibility long before the devel-
This situation does not constitute a distortion opment of banks which, as they evolved,
but merely reflects the low level of per capita tended to take it over. Consequently the su-
income and the resulting low level of individ- pervision of banks has been accepted as a
ual saving and investment. Information is in- government responsibility in most countries.
sufficient for small farmers or merchants to Institutional relationships among central
be able to issue their own notes or publicly banks, commercial banks, building societies,
traded shares. Small-scale investors cannot development of investment banks, agricul-
issue bonds, common stock, mortgages, trade tural credit banks, and so forth can vary a
bills, and so forth in an organized securities great deal from one country to another be-
market for purchase in significant quantities cause of different economic circumstances
by household savers. They and other small and regulatory practices. Many of the differ-
savers are precluded from doing so by uncer- ences are not critical for the efficient overall
tainty about the debt instruments of fledgling operation of a bank-based capital market.
entrepreneurs — by the absence of accounting Initially, therefore, consider a situation in
which the balance sheets of all organized
*From Ronald I. McKinnon, "Financial Policies," banking units in an economy are consolidated
in John Cody et al. (eds.), Policies for Industrial
Progress in Developing Countries, London, Oxford into "Monobank." Monobank collects all
University Press, 1980, pp. 93-9, 102-5, 1 10-14. Re- checking (demand) and interest-bearing time
printed bypermission. Copyright 1980 by the World deposits from, and also issues coins and cur-
Bank.
rency to, the nonbank public, private house-
 255 FINANCIAL DEEPENING

holds and firms. Monobank's balance sheet, deposits (commercial banks), or time depos-
when all interbank claims and liabilities are its (savings banks), credits can be extended
netted out, is shown in Table 1. to investing enterprises in the private or pub-
In an economy, three distinct roles can be lic sector, as represented by the left-hand side
distinguished for the banking system that are of Monobank's balance sheet. (2) Less obvi-
often complementary but sometimes in con- ously, the banking system can promote self-
flict: (1) the traditional monetary role: pro- financed saving-investment within industrial
viding astable unit of account, store of value, firms and family-owned farms. In the less de-
and means of payment in the economy; (2) veloped economy, most small enterprises and
the financial intermediation role: bringing potential new ones do not have easy access to
private savers together with public and pri- outside credits, even from the banking sys-
vate investors; (3) fiscal support for the gov- tem. But new innovative investments to buy
ernment: a source of revenue to the exche- machinery or hire workers require substan-
quer, to be allocated in parallel with tax tial discrete outlays of cash at separated
proceeds, possibly for consumption purposes points in time. To finance these outlays,
or transfer payments. small-scale entrepreneurs accumulate bank
In mobilizing resources for economic de- deposits or stash currency under their mat-
velopment, the intermediary role of the banks tresses before making their investments.
under (2) above is extremely important be- Thus the economic attractiveness of building
cause open markets for primary securities are up and holding cash balances — as repre-
not substantial and are highly illiquid. sented bythe right-hand side of Monobank's
Money in the form of currency and demand balance sheet — is itself important for self-fi-
deposits, and quasi money in the form of time nanced capital accumulation in the less de-
and savings deposits are, on the other hand, veloped economy. Here the term "finance
highly liquid. Being legal tender in an other- motive" is used to refer to the desire of inves-
wise fragmented and illiquid capital market, tor-savers tohold real cash balances.
money yields almost instant purchasing Given this important dual role of the bank-
power for goods and services outside of sub- ing system in financing new investment, how
sistence agriculture. Acceptance of interest- can achievements in the financial sector be
bearing time deposits is officially regulated to assessed empirically as industrialization pro-
ensure their safe convertibility into the ceeds and per capita income grows?
means of payment. One way is to devise a rough statistical
Thus in many poor countries the banking measure of the flow of loanable funds
system is virtually the only financial means of through Monobank, which is equivalent to
attracting voluntary private savings on a the banking system as a whole. Take M2 to
large scale. It has the potential to promote in- be a broad definition of the banking system's
vestment intwo important ways: (1) Through liabilities, as represented in Table 1 by the
the issue of currency (central bank), demand right-hand side of Monobank's balance

TABLE 1. The Monobank Balance Sheet

Assets Liabilities

Loans to private borrowers Coin and currency-

Government bonds Demand deposits
Net claims on foreigners Time deposits (interest bearing)
Net worth (capital subscription)
Total assets = Total liabilities
 256 MOBILIZING DOMESTIC RESOURCES

sheet,1 less the net worth of private banks. To substantially augment their real cash balance
what extent can data on the stock of money, holdings only if the real interest yield is ad-
as measured by the ratio M2 to gross national equate. Indeed, if initially most of the popu-
product (GNP), be an adequate measure of lation ispoor, a significantly high yield on
the /tow of loanable funds in a typical devel- longer-term deposits of the order of 8 or 9
oping country? percent — after adjusting for ongoing infla-
First, the M2/GNP ratio is indicative of tion— may well be necessary to encourage a
the absolute size of the banking system that spurt in financial saving such as that enjoyed
reinvests, in potentially new directions, funds by Taiwan.
from old loans that have matured. It also The basic problem is, of course, that banks
measures the stock of liquidity available for can pay a high yield to depositors only if they
self-financed investments. are earning an even higher yield on their loan
Second, the flow of current saving of portfolio. As a practical matter, this is rarely
households and firms shows up in part as the case, as the semi-industrialized and in-
changes in the assets and liabilities of Mon- dustrializing countries portrayed in Table 2
obank. In particular, the increase in the real have a long history of imposing both direct
stock of money is a measure of the realized and indirect ceilings on the interest rates that
net private financial saving in the prototypi- can be charged on loans from the banking
cal developing country without open markets system. Depositors then necessarily receive a
in primary securities. The flow of private sav- low yield — which can be highly negative
ing in monetary form that is eroded by infla- after allowing for price inflation. Potential
tion— part of the inflation tax collected by depositors respond by strictly limiting their
the government — is omitted from this mea- saving in financial forms, thus giving rise to
sure. However, increases in the ratio of M2/ financial repression. . . .
GNP provide an indication of real additions A repressed system may have two other as-
to the ongoing loanable funds capacity of the sociated features. First, where there are ex-
banking system. . . . cess credit demands, an unorganized finan-
cial sector — sometimes called the extra-bank
INTEREST CEILINGS AND market because it may operate illegally — is
FINANCIAL REPRESSION maintained and even stimulated. Private bor-
rowers outside the supervision of the orga-
It is easy to persuade small savers to keep nized financial system can offer higher re-
a minimum level of highly liquid "working" turns to depositors and other lenders and use
cash balances within the banking system; the resulting funds to cover the demands of
these consist mainly of coin and currency and borrowers who are prepared, or forced, to pay
perhaps demand deposits. Beyond these basic higher interest rates than prevail in the or-
working balances, small savers will increase ganized sector. Because this unorganized
their money holdings relative to their in- market is unable to provide the legal guar-
comes only if the menu of interest-bearing antees provided to depositors and others in
passbook savings, time deposits, and possibly the organized system, the supply price of
certificates of deposit of longer maturity are credit to this market tends to include a de-
made attractive. In part, this requires a phys- fault risk that depositors and others would
ical infrastructure of banking facilities and not demand in the organized market. Hence
easily accessible services — such as post of- there is a "default spread" in such markets.
fices often provide throughout the country. Paradoxically, it leads to an overall rise in the
At the bottom line, however, small savers will cost of credit resulting from legal restrictions
designed to lower this cost.
'As distinct from Mu which consists only of coin Second, lenders in the organized market
and currency and demand deposits. may be stimulated to lend indirectly to the
 257 FINANCIAL DEEPENING

TABLE 2. Bank Loanable Funds in Typical Semi-industrialized and Industrializing Devel-

oping Countries (ratio of M2 to GNP)

Mean
Country 1960 1965 1970 1975
1960-75
Latin America
Argentina 0.245 0.209 0.168 0.222
0.267
Brazil 0.148 0.156 0.205 0.168
0.164
Chile 0.123 0.130 0.183 0.099 0.134
Colombia 0.191 0.204 0.235 n.a. 0.210
Mean ratio of M2/GNP for four Latin American countries: 0.184
Asia and the Middle East
India 0.283 0.262 0.295
0.264 0.276
Philippines 0.186 0.214 0.186
0.235 0.205
Sri Lanka 0.284 0.330 0.255
0.275 0.286
Turkey 0.202 0.223
0.237 0.222 0.221
of M:
Mean ratio ;/GNP for four Asian-Middle Eastern countries:
0.247
n.a. Not available.
Source: IMF. International Financial Statistics, various issues. Mi is defined as money plus quasi money
plus deposits outside commercial banks. M2 is a stock tabulated as of June 30 for each calendar year.
whereas GNP is the flow of output for that year.

unorganized market. While admittedly "for- Malaysia is another example of a country in

mal" loans are only rarely made openly to which freely operating financial markets con-
"curb markets," many wholesalers, retailers, tributed to rapid industrial growth. . . .
and other traders receiving bank loans in ef-
fect act as moneylenders to customers with- INFLATION AND THE FINANCIAL
out direct access to bank credit. In the best
SYSTEM
of circumstances, where an unorganized
market evolves, a dual financial system de- The distortive effects of inflation are well
velops in which transaction costs are dupli- known. If monetary profits can be achieved
cated and a nepotistic credit system is en- by the stockpiling of commodities as hedges
couraged outside organized and supervised against inflation, there will be a shift from
institutions. Monopoly rents accrue to finan- productive physical investment to sterile
cial entrepreneurs who skim off the spread commodity hoarding. Any inhibition of phys-
between market and regulated interest ical investment will bear most heavily on new
charges. types of investment that are likely to appear
The financial reforms in Taiwan in the more risky (even though the potential of in-
1950s and Korea in the 1960s included at
risks). creased profits accompanies the increased
least a partial freeing of interest rates. Open-
ing the organized flow of capital to market Similarly, with prospective inflation, there
pressures served to relegate the informal is the probability that expected domestic cur-
markets to insignificant importance. While rency profits can be gained from holding for-
formalizing credit flows through the orga- eign exchange and other low-yielding foreign
nized banking sector was not the only cause assets when the inevitable exchange depre-
of the rapid industrialization of both coun- ciation is undertaken. Equally important,
tries after the monetary reforms, its alloca- firms become reluctant to borrow abroad,
tive effects did contribute to this progress. and thus to incur debts in foreign exchange.
 258 MOBILIZING DOMESTIC RESOURCES

This is another manifestation of net capital

rowings can be repaid in tomorrow's depre-
outflow from the economy. Such prospective ciated currency, the true cost of borrowing is
profits, moreover, shift the relative private re- the interest cost minus the currency depre-
turns between foreign and domestic invest- ciation over the period of the loan.
ment to the disadvantage of the latter. The market-clearing rate of interest will
Perhaps most important, prospective infla- rise if it is expected that prices will rise.
tion increases the riskiness associated with all While the inflation surcharge may be greater
investment decisions. As the expected gen- or less than the expected rate of inflation, it
eral level of prices rises, the potential range will always be positive. If institutional rates
of future prices widens. If prices are expected are limited by law or convention, the market-
to rise by something like 5 percent in the next clearing rate may be pushed above the insti-
year, there may be a general level of expec- tutional rate, with the repressive effects dis-
tation that they will settle within the 4 to 6 cussed earlier. If the credit markets are
percent range. If the general level of increase already repressed by usury restrictions, infla-
is expected to be something like 50 percent, tion will greatly aggravate the consequent
this range may become 40 to 60 percent. It is problems. Clearly a 5 percent ceiling on the
not impossible that the increased uncertainty interest rate paid to passbook savers is much
may make this range 30 to 70 percent. The more repressive when inflation rises from
effects of this increased uncertainty will not zero to 20 percent a year. A deposit rate of
be spread evenly over all investments. They interest of 25 percent may be necessary to
will bear most heavily on those investments prevent the M2/GNP ratio from falling
that take a long time to mature, because the In the short run, and within limits, infla-
price level will be more uncertain as the ma- tion may encourage a reconstitution of the
turity date approaches. A characteristic of constant price value of convenience assets
highly inflationary economies is that private (demand deposits and currency) that will
investors will undertake only projects with a stimulate their acquisition and therefore give
very short time horizon. rise to the increase in the savings ratio that
Aside from these and other general inhib- has been observed in many countries in
iting effects on industrial diversification and recent years. This is the current form of
other innovations to which attention has fre- the "inflation-forced-savings" phenomenon.
quently been directed, inflation has certain However, the limits on this process are very
direct consequences for the operation of the specific. After a point, nonfinancial assets be-
financial system. For example, the repression come effective convenience assets, and the al-
problem can be aggravated by inflation. \ locative role of the financial system largely
prospective increase in prices will alter the disappears. The banking system may con-
prospective relative values of financial assets tinue to provide minimal transaction bal-
and consumption, unless the returns on finan- ances, but not other convenience assets such
cial assets rise to compensate for their pro- as savings and time deposits.
spective decline in terms of constant prices. Inflation has led to some flexibility in in-
That is, inflation will lead to a reduction in terest rates over time, and in a few cases the
the flow of funds to the financial system un- monetary value of convenience assets and of
less the nominal deposit rate of interest is al- institutional assets have been indexed to "live
lowed to rise at least as much as the expected with inflation." In a hierarchy of policies,
increase in prices. however, indexing is clearly not the first-best
Concurrently, the anticipated constant answer to the problems of inflation. It intro-
price cost of credit will decline. If funds bor- duces an additional element of uncertainty
rowed today can be used to purchase goods and therefore, at least in subjective terms, an
that may be more expensive tomorrow, the additional cost in an already uncertain world.
price limit that borrowers will place on credit Indexation may alleviate some of the costs of
will rise. In the simplest terms, if today's bor- inflation, but it cannot eliminate them all.
 IV.D. INFLATION AND ITS
EFFECTS
IV.D.1. Effects of Inflation*

Governments are tempted to obtain resources productively without any significant drain on
by printing money or creating credit, because other currently used resources. It could not
of the political resistance which they encoun- work in factories, because there are no idle
ter ifthey raise taxes instead. The results de- factories; or produce agricultural products,
pend on how much money is created, how because there is very little idle cultivable
long inflation continues, what the money is land. But it could be used to build roads, to
used for, how people react, and how well the drain swamps, to dig irrigation canals, to re-
economy is managed. habilitate eroded lands, to build houses with
local woods or clays, and other useful ways
which require no machines and employ only
MATCHING RESOURCES locally available materials. (Ever since the
Pharaohs we have known that construction
An increase in the supply of money will not
requires very little physical capital.) Such
raise prices if it is limited to matching an in- works can increase the productive capacity of
crease inoutput or in the demand for money.
There are two cases where a limited increase the country, especially by making it possible
to cultivate more land. The obstacle is that
in the supply of money will not raise prices.
the workers, when paid for their labour,
First, the amount of goods and services to
spend most of the money on consumer goods.
be exchanged against money increases con-
tinuously. Population is increasing; produc- They produce capital goods, but demand con-
sumer goods. Hence the prices of consumer
tivity isincreasing; and the monetized sector
goods tend to rise. If they were producing
of the economy is increasing at the expense
of the subsistence sector. For these reasons consumer goods of the kind they demand
there would be no increase in prices, but the
alone the supply of money could increase by, resources required for producing more of
say, 30 to 40 per cent in a decade without
such goods (industrial equipment, cultivable
prices increasing.
land) do not exist.
Secondly, an increase in the quantity of
The effect of inflation in these circum-
money will not raise prices if there are idle stances isto redistribute consumer goods; the
resources capable of producing increased
amounts of the kinds of goods on which the previously employed get less while the previ-
ously unemployed get more. Inflation is thus
money would be spent. Idle labour is not
enough; there must also be idle factories and a substitute for taxation, used to finance in-
vestment inthe absence of voluntary savings.
idle land capable of producing more con-
sumer goods in immediate response to de-
mand. This is the situation in industrial coun-
EFFECT ON INVESTMENT
tries during a trade depression. It is not the
case in underdeveloped countries, and is not Some writers seem to deny that inflation
therefore relevant to their problems. can increase investment, but this position is
In several underdeveloped countries (es- not tenable. To analyse the problem correctly
pecially inAsia and the Middle East) un- one must distinguish between the primary
employed labour exists which could be used objective of inflation and its secondary con-
sequences. The primary objective is normally
*From W. Arthur Lewis, Development Planning,
George Allen & Unwin Ltd., London, 1966, pp. 130- achieved; if money is created for expenditure
38. Reprinted by permission. on investment projects, investment increases.

259
 260 MOBILIZING DOMESTIC RESOURCES

What these writers are denying is that, if inflation need not result in any large increase
money is created for other primary purposes, in investment if this was not its primary
investment will increase as a secondary
consequence. purpose.
Inflation will normally have some positive
Most inflations occur in wartime, to enable effect on investment, since the expectation of
the Government to get more resources for rising prices makes it profitable to acquire
making war. This purpose is achieved. The durable assets. But the pattern of investment
normal purpose of peacetime inflation is to may be grossly distorted in response to dis-
enable the Government to have more civil tortion ofprices. An inflation raises domestic
servants or soldiers than it could pay out of prices relatively to world prices. This dis-
the proceeds of taxes alone. This objective courages exports and import substitution.
also is normally achieved. Very rarely has a Thus investment is diverted from these im-
Government used inflation to create produc- portant sectors into lines which are not af-
tive capacity — to build factories or to reclaim fected by world competition. Investment of
agricultural land, for example. (The USSR the kinds which could ease the balance of
in the 1930s is one such case.) On the other payments diminishes, in favour of investment
hand, it is quite common for commercial in commercial and luxury building. To avoid
banks to create credit to finance investment this consequence requires either increasingly
by private capitalists. There is no reason to elaborate controls and subsidies to exports, or
doubt that if the purpose of creating money else frequent devaluation of the currency.
is to spend it on creating productive capacity,
this purpose will be achieved.
Money which is created for other purposes EFFECT ON PRICES
does not have the primary effect of increasing
investment. Does it have this secondary ef- Suppose the Government prints money,
fect? Itwill if the increase in prices redistrib- and uses it to finance capital formation: how
utes income in favour of people who are par- long will the inflation continue? Prices cease
ticularly prone to save and invest; otherwise to rise in one of two circumstances. Either, if
it will not. The argument about the second- the inflation changes the distribution of in-
ary effects on investment therefore turns on come in such a way that the Government is
analyzing the effect of rising prices on in- able to continue financing capital formation
come distribution.
without printing more money; or, if as a re-
The immediate effect of an inflation is to sult of the capital formation, an extra output
increase profits, since prices rise faster than of consumer goods begins to keep pace with
costs. This is not necessarily the ultimate ef- the extra flow of money.
fect. Wage earners and consumers resent the The effect on the distribution of income de-
price increase, and take steps to try to keep
pends on the public's reaction. At the start,
their money incomes rising as fast as prices. the prices of consumer goods rise faster than
They may succeed, or they may not. If they wages, in favour of profits. Take first the sim-
tail, the share of profits in the national in- plest case: suppose that the public accepts the
come increases. This is normally expected to change; i.e., suppose that the distribution of
increase investment, but whether it does so income is altered permanently in favour of
depends on the psychology of those who get profits. Part of the profits will flow to the
the profits. The capitalists who save and in- Government automatically through the in-
vest in plant and equipment tend to follow come tax. If the profit-makers also use the
conservative price policies, under the influ- rest of their profits to buy newly-issued Gov-
ence of cost accountants. A large part of in- ernment bonds, prices will cease to rise, since
flationary profits goes to get-rich-quick capi- the Government can finance continued capi-
talists, and are used rather for speculation in tal formation, merely by issuing more bonds,
commodities and in foreign exchange. Hence and without printing any more money. This
 261 INFLATION AND ITS EFFECTS

simple outcome is unlikely. The profit-mak- consequences of the inflation. Prices of goods
ers will use only part of their net profits to which enter into the cost of living of the gen-
buy Government bonds, so the Government eral population are more crucial than others;
will have to issue some more money. And in if these prices can be controlled, wage de-
any case, the public will not simply accept an mands can be moderated. The increase in
increase of prices, without trying to secure prices creates a shortage of foreign exchange;
higher incomes; so wages will start to chase if the use of foreign exchange can be con-
prices in a cumulative spiral. trolled effectively, and evasion avoided, prior-
The course of the inflation also depends on ity needs can be met. Investment is distorted;
how quickly the new capital becomes produc- export industries are discouraged; instead
tive, and how productive it is. If the money funds are put where prices are expected to
has been used to reclaim land and settle rise most, including into stocks and real es-
farmers on it, their increased output will tate. An effective licensing of investment is
check the rise of prices when it reaches the difficult, but becomes necessary if the pattern
market. Also, the public is more likely to ac- of the economy is not to be grossly distorted.
cept an unfavourable change in distribution Governments differ in the efficiency with
which is offset by an increase in real con- which they can operate such administrative
sumption per head. This is a most important controls, and populations vary widely in their
difference between the creation of money, willingness to co-operate.
say, to pay Government employees, and the
use of such money to finance the creation of
new resources which will increase commodity SPIRAL INFLATION
output.
Theoretically, an inflation can peter out No matter how efficient the Government
either because of a permanent shift to taxes may be, if the inflation is large enough (the
and savings, which flow to the original inves- rise in prices exceeding, say, 5 per cent per
tors and so enable the new investment level annum) and goes on for long enough (say,
to be held without additional new money; or more than four years) it is likely to spiral.
because there is an increased flow of con- The public tumbles to what is happening and
sumer good output which catches up with the insists on taking protective measures. Wage
increased money supply. If neither of these and salary earners arrange for their incomes
does the trick, prices will continue to rise. to rise as fast as prices. Each price increase
The reaction of the public then depends on is followed by an increase in wages; this
the length of time during which this goes on. raises costs, so prices increase again; and so
If the monetary tap is turned on for a couple on. This process can go on indefinitely, if the
of years, to finance some particular project, rate of inflation is held steady; some countries
and then turned off again, prices may have have had price increases averaging 20 per
ceased to rise before the public is sufficiently cent per annum for more than 50 years.
well organized to take protective measures. However, an "equilibrium" inflation of this
This is what happens in inflations caused by kind serves no useful purpose. An inflation is
the creation of bank credit to finance private useful only so long as the cost of living rises
investment; the tap is turned on for two or faster than the average expenditure on con-
three years, but before the public gets a sumption, for this gap is what permits an in-
chance to believe that prices will now rise crease incapital formation. Once the stage is
continuously, the monetary authorities step reached where wages and prices are rising at
in and turn it off again. the same rate, the inflation ceases to be use-
If the tap is not turned off, and prices con- ful. If the Government persists in trying to
tinue to rise, the reaction of the public will transfer resources through monetary expan-
also be influenced by the effectiveness of the sion, itwill have to expand the money supply
Government in moderating the more obvious at an accelerating rate, with prices rising
 262 MOBILIZING DOMESTIC RESOURCES

faster and faster. This is the way of hyper- This too is likely to have some inflationary ef-
inflation. This cannot go on for very long, fects. For any attempt to reduce the share of
since no population will tolerate it for long. consumption in national income, or to redis-
Spiral inflations are not only useless, but tribute consumption at the expense of large
are also difficult to stop. Wage and salary groups, is certain to be unpopular whether it
earners have grown accustomed to getting is done by increasing the supply of money or
regular pay increases to catch up with price by increasing taxes. Inflation and an increase
increases. At any point in time that a decision in indirect taxation both have the same im-
to stop the inflation is taken, some incomes mediate effect: an increase in prices; and di-
are due for a rise. If they do not get it, there rect taxes tend to be even more unpopular.
is unrest; whereas if they do get it, but em- Whatever the method of financing, the public
ployers are prohibited from passing on the in- will most probably react by demanding
crease, there will be some unemployment. higher wages or income, so any method is ca-
Investors have grown used to profits accruing pable of starting a price-wage spiral.
from a continuously augmented supply of The fact is that an increase in the ratio of
money. Once the monetary supply is stabi- capital formation to national income is al-
lized, anumber of enterprises cease to be prof- most always accompanied by some increase
itable, and employment is cut. Hence an in prices, however it may be financed. Even if
inflation cannot be stopped without a tem- the increase is financed by genuine savings, a
porary crisis in investment and employment. rise in wages in profitable industries may pro-
The crisis need not be long; six to eighteen voke sympathetic increases in prices through-
months is normal. But those who will lose op- out the economy; or expansion may come up
pose the adoption of measures which would against crucial bottlenecks. If the increase in
stop the inflation, and if they are sufficiently capital formation is financed from taxes, the
powerful, they can prevent a Government public may resist changes in consumption
from putting an end to inflation. Some Latin patterns and start a cost-push spiral. It is not
American countries have had persistent infla- true that an increase in the price level always
tion for so many decades that they are now accelerates capital formation; but it is true
probably incapable of living in any other that in most economies where the ratio of
way, unless there are major changes in the capital formation is accelerating, prices will
balance of political forces. be found to be rising. (Whether the ratio is
high or low is not the point; the emphasis is
on acceleration.)
ALTERNATIVE TO TAXES
Though taxation may also raise prices, it is
Thus the course and effects of an inflation nevertheless from the economic point of view
cannot be predicted without knowing its cir- greatly superior to inflation as a source of fi-
cumstances; what the money is being used nance for capital formation. First, its effect
for; whether a permanent shift to profits is on prices is likely to be much smaller. Sec-
possible; whether the Government can main- ondly, the discipline of controlling the money
tain effective controls; and how public opin- supply is easier to maintain if the Govern-
ion will react. One can say at once that most ment isnot committed to using inflation as a
underdeveloped countries would be unwise to means of acquiring resources. Once the tra-
launch upon an inflationary course because dition of monetary discipline is lost, Govern-
they could not control it. But one cannot rule ments take to inflation like ducks to water,
out the possibility that some of the better or- and financial control disappears. Thirdly, the
ganized societies can safely finance some incidence of taxation can be controlled more
capital formation by the creation of money. fairly and efficiently than the incidence of in-
Given the desire to accelerate capital for- flation. And fourthly, inflation is not so dan-
mation, the practical alternative to the crea- gerous to the balance of payments, or so
tion of money is the levying of higher taxes. distorting to the pattern of investment. Con-
 263 INFLATION AND ITS EFFECTS

tinuous price inflation, moving domestic costs money is to finance increased capital forma-
out of line with world prices, can be a major tion, this purpose is likely to be achieved. The
source of economic stagnation. secondary effects of inflation on investment
The idea that it is politically easier to mo- cannot be predicted with certainty. Some net
bilize resources through inflation than increase is probable; considerable distortion
through taxation is probably spurious. For a is also probable, since inflation discourages
Government which is capable of using infla- investment in export industries and import
tion intelligently is probably equally capable substituting industries, and favours sectors
of mobilizing resources through taxation; not subject to world competition, such as lux-
since if it is courageous, popular and efficient ury construction.
enough to enforce the measures which con- 4. The degree of inflation will depend on
trol inflation (and to stop it before it spirals) the willingness of the general public to accept
it is presumably equally capable of levying the redistribution of income which inflation
and collecting the necessary taxes. The situ- causes; if the public is not willing, the infla-
ation may well be that countries which could tion may spiral. The longer the inflation con-
use inflation safely do not need it; and those tinues, the more effective the public becomes
which need it cannot use it without getting in evading its consequences, and therefore
into more trouble than it is worth; but polit- the greater the chance of its spiralling.
ical analysts may find some marginal cases. 5. The degree of inflation also depends on
the purpose of the inflation. If its purpose is
to create quick-yielding productive capacity,
SUMMARY the resulting flow of consumer goods holds
prices in check; the public is also more likely
1. The supply of money can be increased to accept an unfavourable change in distri-
as fast as the supply of goods without raising bution which is offset by an increase in real
prices. An increase in the monetary sector consumption per head.
relative to the subsistence sector also makes 6. In an underdeveloped country inflation
possible an increased supply of money to is merely a substitute for taxation. From the
match increased demand. economic point of view it is an inferior
2. An increase in the supply of money will substitute.
immediately stimulate a greater output of 7. It may be easier politically to start an
consumer goods in an industrial country suf- inflation than to tax; but the measures which
fering from a trade depression. It will not control inflation, maximize its usefulness and
generally have this effect in an underdevel- minimize its disadvantages are no easier to
oped country. adopt or administer than would be an in-
3. If the primary purpose of creating crease intaxation.

IV.D.2. Inflation and

Development Policy*

The first question to be discussed is whether by development policy inevitably involves in-
the mobilization of an economy's resources flation. Itis contended here that some degree
of inflation — but a moderate degree only — is
*From Harry G. Johnson, "Is Inflation the Inevi- the logical concomitant of efficient economic
table Price of Rapid Development or a Retarding mobilization. The argument rests on two
Factor in Economic Growth?" Malayan Economic
Review, Vol. 11, No. 1, April 1966, pp. 22-8. Re- propositions. One is that so long as inflation
printed bypermission. proceeds at a rate low enough not to disturb
 264 MOBILIZING DOMESTIC RESOURCES

seriously the general confidence in the stabil- in an advanced economy is frequently defined
ity of the value of money its effects are pri- as a rate of inflation of no more than one to
marily to redistribute incomes, and to do so two per cent a year, the tolerable degree of
to an extent that does not involve serious so- stability for an underdeveloped economy
cial consequences, rather than to produce sig- might be in the range of a four to six per cent
nificant misallocations of resources, such as annual rate of price increase. (Harberger has
those which occur when people come to ex- suggested that a 10 per cent annual rate of
pect inflation and seek to protect themselves inflation represents the outside limit of infla-
from it by holding goods instead of money tion justifiable by this line of argument.)1
and by using political means to safeguard This analysis, of course, relates to purely do-
their real incomes. The other proposition is mestic considerations and ignores the bal-
that, owing to the various rigidities and im- ance-of-payments or exchange-rate implica-
mobilities characteristic of any economy, but tions of internal price trends.
paticularly of underdeveloped economies, up- The foregoing remarks relate to the ques-
ward movements of wages and prices can tion of inflation as a consequence or aspect of
help to reallocate labour and resources and to economic development policy, and to the ar-
draw labour out of traditional or subsistence gument that some degree of inflation is the
sectors into the developing sectors of the necessary price of rapid development. It has
economy. It is important to note that this been argued that a modest rate of inflation is
proposition, like the first, presumes a general a logical part of an efficient development pol-
expectation of stability in the value of money,
icy, in the sense that the "price" may pur-
as a precondition for the offer of higher chase gains in efficiency of resource alloca-
wages and prices to serve as an inducement tion and utilization that outweigh the costs.
to mobility. The argument now turns to the second prob-
The second proposition implies that some lem raised by the question, the effectiveness
inflationary pressure in the economy will as- or otherwise of inflationary financing of de-
sist the task of mobilizing resources for de- velopment programmes.
velopment; the first implies that such infla- The main theoretical arguments for infla-
tionary pressure will not introduce offsetting tionary development policies derive from two
distortions causing significant real losses to systems of economic thought, the Keynesian
the economy as a whole, but will instead theory of income and the quantity theory of
mainly involve transfers of income within the money. The Keynesian approach to the ques-
economy, the social consequences of which tion (which derives from the Keynes of the
will be small enough to be acceptable. Effi- Tract and the Treatise as much as from the
cient policy-making will therefore involve ar- Keynes of the General Theory) argues that
riving at a trade-off between the mobilizing inflation will promote growth in two ways: by
and ^distributive effects of inflation that will redistributing income from workers and
involve some positive rate of inflation. The in- peasants, who are assumed to have a low
dicated optimum rate of inflation is likely to marginal propensity to save, to capitalist en-
be significantly higher for an underdeveloped trepreneurs, who are assumed to have a high
than for an advanced economy for two rea- marginal propensity to save and invest; and
sons: first, the sophisticated financial system by raising the nominal rate of return on in-
of an advanced economy provides many more vestment relative to the rate of interest, thus
facilities for economizing on the use of money promoting investment. Neither of these ar-
in face of expected inflation; and second, the guments, however, is either theoretically
superior mobility of resources of an advanced
economy implies that the increase in total
output achievable by inflationary means is 'Arnold C. Harberger, "Some Notes on Inflation,"
in Werner Baer and Isaac Kerstenetzky (eds.), Infla-
relatively much smaller. Thus one might ex- tion and Growth in Latin America, Homewood, 111.:
Irwin, 1964.
pect that whereas "tolerable" price stability
 265 INFLATION AND ITS EFFECTS

plausible or consistent with the facts, at least holders of money have to forego each period
so far as sustained inflationary processes are in order to restore the real value of their
concerned. Both rest on the arbitrary and money holdings. The presence of this tax, in
empirically unsupported assumption that en- turn, encourages the public to attempt to
trepreneurs realize that inflation is occurring, evade the tax by reducing their holdings of
whereas the other members of the economy money, shortening payments periods, holding
do not, or at least do not realize it fully. As inventories of goods instead of cash, and so
to the first, the theoretical prediction is that forth; these efforts involve a waste of real re-
all sectors of the economy the prices of whose sources and a reduction of real income, rep-
services are upwards-flexible will come to an- resenting the "collection cost" of the infla-
ticipate inflation, so that no significant redis- tionary tax. On the other hand, the real
tributions ofincome will take place; this pre- resources collected by the inflationary tax are
diction accords with the mass of the available available for use in the development pro-
evidence. As to the second, the theoretical ex- gramme; and if they are used for investment,
pectation isthat free-market interest rates the inflationary policy may accelerate eco-
will rise sufficiently to compensate holders of nomic growth. It should be noted, however,
interest-yielding assets for the expected rate that in the transitional stages of an inflation-
of inflation; this expectation also accords ary development policy, or in the process of
with the mass of the available evidence. This acceleration of such a policy, whatever con-
argument for inflation, therefore, is valid tribution to growth there is may be out-
only under two possible sets of circum- weighed bythe increased waste of resources
stances: first, in the early stages of an infla- produced by the increase in the inflationary
tionary development programme, while the tax.
mass of the population (especially the work- In practical experience, resort to the infla-
ers and the savers) still has confidence in the tionary tax as a method of financing eco-
stability of the value of money; and second, nomic development is generally prompted by
when inflationary financing is accompanied the inability of the developing country to
by governmental policies of holding down the raise enough revenue by taxation and by bor-
wage and interest costs of business enter- rowing from the public to finance its devel-
prise. Such policies would generate distor- opment plans, either as a result of the low in-
tions in the allocation of resources, which come and taxable capacity of the economy, or
might offset any benefits to growth from the more commonly as a result of inability to
inflationary policy; in particular, the contrary command the necessary political consensus in
view has been argued that inflation will dis- support of the necessary sacrifices of current
courage the supply of saving for investment. income. Unfortunately, the same character-
The quantity theory approach, on the other istics of underdevelopment that limit the ca-
hand, adopts the more realistic assumption pacity to finance development by orthodox
that in a sustained inflationary process the fiscal methods also place rather narrow limits
behavior of all sectors of the economy will be- on the possibility of financing development
come adjusted to the expectation of inflation, by inflation. In particular, underdevelopment
and that consequently the effect of inflation implies a relatively smaller use of money
will be, not to redistribute income from work- than is common in advanced countries, and
ers or savers to capitalist entrepreneurs, but therefore a relatively smaller base on which
to redistribute it from the holders of money the inflationary tax can be levied.
balances — who are the only losers from an Before discussing this point in detail, it is
inflation that is anticipated — to the monetary appropriate to refer to a related question that
authorities who issue money the real value of figured large in the development literature of
which steadily depreciates. Inflation imposes about a decade ago, the question of the extent
an "inflationary tax" on holdings of money, to which development can be financed by
which consists in the real resources that the monetary expansion without producing infla-
 266 MOBILIZING DOMESTIC RESOURCES

tionary consequences. The answer, clearly, is which inflation can make resources available
that such financing can be safely pursued up for economic development through the infla-
to the limit set by the growth of demand for tionary tax. Ignoring the possibilities of non-
money consequent on the expected growth of inflationary financing by monetary expansion
the economy at stable prices, plus the growth due to the growth and monetization of the
of demand for money associated with the economy, the yield of the inflationary tax as
monetization of the subsistence sector (where a proportion of national income will be the
relevant), minus the portion of the growth in product of the money-to-income ratio, the
the money supply that must be created rate of inflation, and the proportion of the in-
against private debt. The magnitude of the crease in the money supply captured for fi-
resources that can be made available for fi- nancing development. Thus, with the as-
nancing development by this means, how- sumed money-to-income ratio of one-fifth
ever, depends on the magnitude of the abso- and an assumed capture rate of one-half, a 10
lute increase in the money supply permitted per cent rate of inflation would secure 1 per
by these factors or, to put it another way, on cent of national income for the development
the rate of growth of the demand for money, programme, and so on. The money-to-income
the ratio of money to income, and the portion ratio, however, is not insensitive to the rate of
of the additional money that can be used to inflation, but is on the contrary likely to de-
finance public spending. Thus, for example, crease appreciably as the rate of inflation
with a rate of growth of demand for money rises, thereby setting limits to the possibilities
of 6 per cent, half of which can be used to of development finance by these means. Fur-
finance public spending, the budget deficit fi- ther, itshould be noted that insofar as devel-
nanced by monetary expansion would be 3 opment financing depends on a growth of de-
per cent of the initial money supply. If the mand for money resulting from monetization
ratio of money supply to national income of the economy, inflation is likely to reduce
were in the neighborhood of two-fifths, as is that growth by inhibiting monetization, so
common in advanced countries, this would further reducing the net amount of resources
make 1.2 per cent of national income availa- gathered for development finance through in-
ble for development investment; if, on the flation. . .
other hand, the ratio of money supply to na- The circumstances in which inflation is re-
tional income were in the lower neighborhood sorted to in underdeveloped countries, how-
of one-fifth, as is common in underdeveloped ever, are far from the most favorable con-
countries, only one half of one per cent of na- ceivable, and their inflations are extremely
tional income would be available for devel- likely to proceed in such a way, and to be ac-
opment investment.2 The difference in the companied bysuch other economic policies,
order of magnitude of the money-to-income as to exercise a serious retarding influence on
ratio explains both why budget deficits in economic growth. Specifically, inflationary fi-
underdeveloped countries are more fre- nancing may impede growth in three major
quently associated with inflation, and why in ways, each contrary to the assumptions of the
such countries inflationary financing of de- inflationary tax model.
veloped ismore frequently resorted to, than In the first place, contrary to the assump-
in advanced countries.
tion that prices throughout the economy ad-
In the same way as it limits the scope for just freely to inflation, the government of a
non-inflationary deficit financing of develop- developing country employing inflationary
ment, the restricted use of money in under- development policies is likely to be under
developed countries limits the extent to strong political pressure to protect important
sectors of the community from the effects of
2The illustrative numbers used are derived from J.
inflation, through control of food prices,
J. Polak, "Monetary Analysis of Income Formation
and Payments Problems," International Monetary rents, urban transport fares, and so on. Such
Fund Staff Papers, November 1957, Table on p. 25. controls inevitably distort the allocation of
 267 INFLATION AND ITS EFFECTS

resources within the economy, and particu- rates and to defend them to the limit of their
larly their allocation to private investment in exchange reserves, borrowing powers, and
growth. Fixing of low prices for food inhibits ability to use exchange controls. In this kind
the development of agricultural production of setting inflation introduces a progressive
and the improvement of agricultural tech- tendency toward exchange rate overvalua-
nique; control of rents, on the other hand, tion, balance-of-payments difficulties, and re-
may unduly foster the construction of new sort to increasing protectionism, which in
housing to accommodate those who cannot turn results in the diversion of resources
find rent-controlled housing or to enable away from export industries and toward
landlords to evade rent controls. All these high-cost import-substituting industries, and
policies tend to promote urbanization, which a consequent loss of economic efficiency.
involves expenditure on social overhead and While the appearance of growth may be gen-
may increase the numbers of the urban un- erated bythe establishment of import-substi-
employed. Moreover, control of prices of tute industries, the reality may be lost in
food, and particularly of fares on state-owned misallocation of resources produced by pro-
transport facilities, may involve the state in tectionism and the inefficiency of exchange
explicit subsidies on the one hand and budget control procedures. Moreover, eventually the
deficits on the other, so that the proceeds of increasing overvaluation of the currency is
the inflationary tax are wasted in supporting likely to force a devaluation, coupled with a
the consumption of certain sections of the monetary reform involving drastic domestic
population rather than invested in develop- deflation. This experience, in addition to the
ment. Such phenomena are widely observa- immediate disturbing effects of deflation in
ble in the underdeveloped countries of the interrupting the growth of the economy, has
world. (They are also observable in advanced the long-run effect of damaging the stability
countries, but the latter can easily afford the and confidence of expectations on which the
wastes of resources involved.) process of investing in the growth of the econ-
Second, contrary to the assumptions of the omy depends.
inflationary tax model, inflation typically To summarize, this paper has argued the
does not proceed at a steady and well-antici- following propositions. First, an efficient de-
pated rate, but proceeds erratically with velopment policy should plan on some modest
large politically-determined variations in the degree of inflation as a means of more fully
rate of price increase. These variations in the mobilizing the economy's resources; in this
rate of inflation divert a great deal of the ef- limited sense, inflation is an inevitable price
fort of private business into forecasting and of rapid development. Second, while a policy
speculating on the rate of inflation, or hedg- of financing development by deliberate infla-
ing against the uncertainties involved. They tion has strong attractions theoretically and
also destroy the possibility of rational calcu- politically, the possibilities of stimulating
lation of small margins of profit and under- economic development by these means are
mine the incentives to strive constantly to re- quite limited. Third, inflationary develop-
duce costs and improve performance, which ment policies in practice are unlikely to
striving is the key to the steady increase of achieve this stimulating effect, but on the
productivity in the industrially advanced contrary likely to retard economic growth, by
nations. distorting the allocation of resources and
Finally, the inflationary tax model assumes wasting the inflation-gathered development
either a closed economy or a country on a resources on consumption, by increasing un-
floating exchange rate system. In reality, certainty and reducing incentives for inno-
countries — especially underdeveloped coun- vation and improvement, and through their
tries— are exposed to competition in and balance-of-payments effects by fostering the
from the world economy, yet they display a inefficiencies of protectionism and exchange
strong propensity to maintain fixed exchange control.
 268 MOBILIZING DOMESTIC RESOURCES

IV.D.3. Monetarism and

Structuralism in Latin America*

The controversy between "monetarists" and vocate a restrictive monetary policy in an at-
"structuralists" in Latin America has been tempt to hold down the money supply and
exaggerated beyond all bounds. While rather keep prices absolutely constant.
virulent in theoretical disquisitions, it nar- The basic controversy lies in the tendency
rows down substantially when practical pol- of economists of the structuralist persuasion
icy recommendations have to be formulated. to view monetary expansion and inflation as
One cruel fact that the structuralists must an unavoidable feature of structural change
face, when entrusted with policy responsibil- and economic growth in Latin America; and
ities outside the academic wards, is that correlatedly to claim that the attempt to
structural adjustments inevitably take a long avoid monetary expansion and inflation will
time, while the combat to inflation, if it is to only hamper structural changes and eco-
succeed, needs fairly quick, visible results; nomic growth. If other descriptions are
and those can be obtained much more expe- needed, one might say that the structuralists
ditiously (though often at the cost of painful would hold that the permissiveness of the
side effects) through the demand side — via monetary and fiscal policy of the monetary
monetary and fiscal policies — than through authorities is simply a reflection of exogenous
the supply side (unless foreign aid is available factors, particularly the decline in import ca-
in unlimited amounts). I do not, of course, pacity, while the monetarists would hold that
deny that the purely monetary or fiscal solu- this fatalistic result is not inevitable, and that
tion may not be durable or consistent with inflation cannot be blamed exclusively or
stable growth unless adjustments are also even predominantly on "exogenous" factors,
made on the supply side. The truth is, that in but rather on the "policies of inaction" pur-
the short run, all structuralists when en- sued by most of the Latin American
trusted with policy-making responsibilities governments.
become monetarists, while all monetarists Another "spurious" presentation of the
are, in the long run, structuralists. Thus, controversy is to take the view that monetar-
we might jocosely define a monetarist as a ists in Latin America are slavishly following
structuralist in a hurry and a structuralist monetary tenets of classical liberalism, by
as a monetarist without policy-making adhering strictly to the quantity theory in
responsibility. money and banking, to free trade views in
Let us try to determine at this stage what matters of international trade, to non-inter-
the controversy is not. It is not a quarrel be- vention doctrines insofar as government ac-
tween those economists and policy-makers tivities are concerned, and generally by as-
who recognize the existence of structural ri- cribing lower priority to development than to
gidities and those who do not. All recognize stabilization and by neglecting the special
that structural problems do exist. It is not a problems of structural under-employment or
controversy between those who favor eco- unemployment in underdeveloped coun-
nomic growth and development and those tries. . .
who do not. It is not a debate between those
who advocate an economic policy which per- THE MONETARIST VIEW ON
mits the economy to grow and those who ad- INFLATION AND GROWTH
*From Roberto de Oliveira Campos, "Economic
Development and Inflation with Special Reference to
Apart from a small current of thought that
Latin America," in OECD, Development Plans and approaches the question of inflation and de-
Programmes, OECD Development Center, Paris, velopment from the viewpoint of value judg-
1964, pp. 129-37. Reprinted by permission. ments, and that, by questioning the social ad-
 269 INFLATION AND ITS EFFECTS

vantage of inflation-financed development, may be offset by luxury consumption of the

tends to ascribe high social priority to stabi- entrepreneurial group, by the lower efficiency
lization, the bulk of the monetarists in Latin of government investments or by bottlenecks
America appear to recognize the social prior- in the import capacity; (iii) while inflation
ity of development, while insisting that con- encourages the adventurous entrepreneur, it
tinued and stable growth can best be attained tends to discourage risk-taking in basic en-
in an environment of monetary stability. terprises ofa long maturation period.
Their position can be thus summarized: 2. Inflation discourages the inflow of for-
1 . The only possible contributions of infla- eign capital and renders more difficult the ab-
tion to growth derive: sorption of financing from foreign govern-
a. From the forced saving mechanism, ments and institutional organizations.
through which the decline in real wages due 3. Inflation tends to impair the qualitative
to rising prices and the wage lag would in- composition of investments. It discourages
crease investible profits. Thus inflation would investments in basic industries and infras-
act as a disguised taxation on consumers and tructural services which are either price-con-
savers to the benefit of investors; trolled or require long gestation periods, or
b. From the assumption that the pressure both. It weakens or prevents the creation of
of inflationary demand would permit a fuller credit and capital markets and stimulates
utilization of manpower and resources, while speculative investments in inventories.1
conversely the attempt to combat inflation by 4. The structural thesis exaggerates the
wiping out excess demand might bring em- two basic alleged rigidities: the inelasticity of
ployment and resource-utilization down to the food supply and the inelasticity of ex-
the level permitted by bottleneck factors; ports. The reasoning of the structuralists
c. From the assumption that inflation starts from the premise of a total sectoral in-
would stimulate bold entrepreneuriship and compressibility of prices, so that any upward
reward investors at the expense of conserva- price pressure would result in a rise in the
tive savers or rentiers. general price level. Thus the pressure of pop-
Against these supposed advantages, the ulation growth and rising urban incomes
monetarists raise a formidable number of would tend to raise, through a chain-reaction
qualifications: (i) the forced savings device mechanism, first the price of agriculture
can act only temporarily for discontinuous in- goods, secondly the general price level and
flation, but has its efficacy lowered or de- thirdly wages, thus creating an inflationary
stroyed when inflation is chronic and be- spiral of a structural nature. Similarly, the
comes a part of the expectation of wage inelasticity of world demand for primary ex-
earners, who devise defense mechanisms to port would tend to lower the export capacity
prevent a decline in real wages. It is true that below import requirements for growth, ren-
in dual societies, in which because of insti- dering necessary an accelerated process of
tutional factors real wages in some urban in- import substitution. But, at least initially, im-
dustries are set above the limit of marginal port substitution tends to be inflationary be-
productivity, it is possible, because of the cause of the relative inefficiency of the new
continuous migration of under-employed industries during the learning period, this
rural labor to the city, and given the relative cost pressure being aggravated by the need
rigidity of factoral proportions of labor and for exchange devaluations in an attempt to
capital in modern technology, to depress the restore external balance. The monetarists
real urban wage for protracted periods; this, would argue that the alleged structural in-
however, only prolongs a little bit the useful-
ness of the forced savings mechanism but 'These results need not necessarily occur. For a dif-
does not guarantee its continuity; (ii) such ferent view, see Werner Baer, "Inflation and Eco-
transfer of resources as may take place from nomic Efficiency in Brazil," paper presented at the
Rio de Janeiro Conference on Inflation and Develop-
consumers to the government or to investors ment, January, 1963.
 270 MOBILIZING DOMESTIC RESOURCES

elasticity of food supplies is often not struc- tion of the monetarist approach is that it en-
tural at all. It stems frequently from the fact compasses not only the use of traditional
that the administrative control of food prices, monetary weapons — the efficacy of which is
designed to protect the urban consumer, cuts by definition very limited in the rather prim-
off the agricultural producer from price and itive financial markets characteristic of the
market stimuli, so that the inelasticity of food less developed countries — but also of fiscal
supply, rather than being an inherent struc- policies. The second is that the effectiveness
tural characteristic, may be a distortion in- of monetary weapons, stricto sensu, is consid-
duced by administrative controls. While in erably greater in the case of demand inflation
some cases the structure of land tenure would than of cost inflation. This is particularly true
prevent in any event the diffusion of market if the problem is that of a wage spiral, where
incentives, thus rendering structural reform a the reduction of the quantity of money may
precondition for increasing food production, stop the rise in prices but only at the expense
this situation occurs only rarely. As to indus- of employment. The third point is that mon-
trial import substitution, while it is not de- etary tools, as distinct from fiscal policies, are
nied that it carries a built-in inflationary inappropriate when the problem is to curb
pressure, at least if the industrial expansion consumption (which can only be achieved via
takes place at constant cost, the cost-push fiscal policies), the effectiveness of monetary
arising therefrom has been grossly exagger- restraints being greater if the objective is to
ated. Thus, assuming that imports for a typ- restrain investment.
ical underdeveloped country represent 10 per The fourth point is the asymmetrical effect
cent of national product, that real income is of monetary policies; through a combination
rising at 6 per cent per annum, that exports of quantitative and qualitative credit controls
are stationary and that the income-elasticity it is possible to orient a selective expansion of
of demand for imported products equals 2, certain economic sectors but much harder to
and finally, that the needed customs protec- enforce a selective restriction of undesirable
tion for the national substitutes is 100 per sectors.
cent — all of those factors would not entail a The net upshot of these observations is that
price increase above 1.2 per cent per annum, monetary weapons, though an indispensable
a margin vastly surpassed in the inflated ingredient in anti-inflationary programmes,
economies of Latin America. have to be used in a prudent combination
with fiscal policies. The shortcomings to their
application arise not only from asymmetrical
CRITICISM OF THE MONETARIST effects — they affect investment more than
APPROACH consumption, their efficiency is greater in re-
ducing demand than costs — but from the
The monetarist approach has curiously limited organization and responsiveness of
enough been attacked from two conflicting money and credit markets in the less devel-
vantage points. It is argued, on the one hand, oped countries. Fiscal policies, particularly
that it is too strong because controls will re- when designed to curb luxury consumption
strain investment, generating unemployment by taxation of wealthier groups so as to raise
and losses in real output, thus not only aggra- the marginal tax rate above the average rate,
vating the supply problem but also arousing have undoubtedly a fundamental role to play
political instability. Some argue, on the other in anti-inflation programmes.
hand, that it is too weak, because without the
The applicability of monetary policies de-
leverage of fiscal policies it does not really pends, of course, on the speed desired in
control excess demand nor go to the heart of checking inflation and on the degree of flexi-
structural and institutional problems. bility of the wage pattern and labor organi-
Four points may be readily granted. The sation. The latter factor is a major determi-
first is that the only meaningful interpreta- nant of the possibility of utilization of credit
 271 INFLATION AND ITS EFFECTS

controls without untoward unemployment Thus stability is achieved only at the expense
effects. of growth.

THE "STRUCTURALIST" VIEW CRITICISM OF THE

STRUCTURALIST APPROACH
The "structuralist" school stresses the
"structural vulnerability" of Latin American While the structuralists are very articulate
economies to inflation because of two alleged in their diagnosis of Latin American inflation
basic rigidities. and in their strictures on monetarism, they
The first one is the slow and unstable rate are far less explicit on practical policy rec-
of export growth, which is held to be chroni- ommendations. These do not go usually be-
cally inadequate to support the needed rate yond the expression of pious hopes for struc-
of development; the sluggish growth rate tural changes, which because of their
makes necessary a continuous and sharp ef- long-run nature are not serviceable recipes
fort of import substitution, creating a cost- for the short-run cure of inflation.
push because of the substitution effort itself; The structuralist tends therefore to advo-
the instability of the growth rate in turn cre- cate gradualism in anti-inflationary pro-
ates occasional contraction of government grammes and to postulate an increase in for-
tax revenues arising from exports, precisely eign aid and international financing as major
when government expenditures need to be in- factors in helping to buy time for the needed
creased to offset the depressant effect of the structural changes.
stagnation or recession of the export sector; One of the most cogent and articulate crit-
finally, the secular trend towards deteriora- icisms of the structuralist interpretation of
tion of the trade of primary products creates Latin American inflation has been put for-
an additional complication, further limiting ward by Arthur Lewis in his "Closing re-
the potentialities of growth of the export in- marks at the Conference on Inflation and De-
come, and reinforcing the trend toward peri- velopment inLatin America," convened in
odical exchange rate devaluation. January, 1963 in Rio de Janeiro.
The second one is the inelasticity of agri- Lewis' initial and important contribution is
cultural production, due largely to defective to stress the difference between self-liquidat-
patterns of land tenure which decrease the ing and spiral inflation, the first being rele-
responsiveness of food production to price vant and the second irrelevant for economic
stimuli.
development. Through self-liquidating infla-
The cost-push in Latin American econo- tion limited policy objectives can be attained,
mies would thus come from a fourfold direc- such as altering permanently the distribution
tion: cost of import substitution, rise in agri- of income in favor of investment. The spiral
cultural prices, deterioration of the terms of inflation, in turn, does not serve any growth
trade and exchange rate devaluation. objective but is rather a political phenome-
Fortified with those tenets, the structural- non, arising from tensions in the society.
ists proceed to inveigh against what they call The first stricture of Lewis is to question
the "orthodox" monetary policies advised by the alleged inevitable sluggishness of the ex-
the International Monetary Fund, which in port sector of primary producing countries.
their view do not go to the heart of the prob- He mentions that between 1950 and 1960,
lem— structural change — and have depres- the quantum of trade in primary products
sant effects, manifested in a decrease in the grew by 6 per cent, only slightly under the
level of investment, in the contraction of the rate of growth of manufactured exports,
private sector due to the incompressibility of which was 7 per cent. While Lewis is right in
the public sector and in the deflationary ef- excoriating the excessive importance attrib-
fects of unemployment and the wage lag. uted by the structuralist school to the "struc-
 272 MOBILIZING DOMESTIC RESOURCES

tural" sluggishness of the export sector in It is true that this does not quite solve the
Latin America, he dismisses altogether too problem, particularly in those countries
glibly the trade problem, which is real. In where the decline in export activity requires
particular, he overlooks the fact that: compensatory government policies for do-
a. Part of the statistical expansion of ex- mestic expansion, or where fiscal revenues
ports of primary products in the last decade are greatly dependent on export taxation. But
did not come from the underdeveloped coun- the brunt of the argument is taken. The two
tries but from the industrialized countries. main weaknesses of the structuralist view on
The latter expanded their export of primary trade are then:
products by 57 per cent while the former by a. that some of the sluggishness in the ex-
only 14 per cent. Thus not only world trade port growth is not really structural but re-
in manufactures expanded at a faster rate, sults plainly from the failure to exploit, be-
but the lion's share of primary exports was cause of overvalued exchange rates, export
taken by the industrialized countries them- opportunities that may still exist;
selves. Latin American exports were also b. the import quantum has been main-
greatly affected by a price decline since 1953. tained, on the average, for Latin America,
While world trade grew in volume by 50 per despite the decline in export revenues, so that
cent between 1953 and 1960, raw material there has been no inflationary decrease in the
prices declined by 7 per cent. Brazil ex- availability of import goods.
panded its volume of exports by 20 per cent The second stricture of Professor Lewis
in the above period, while unit prices fell by concerns the wage push. He tends to deny
37 per cent. any special characteristic to the Latin Amer-
b. The trade experience with which Profes- ican wage scene, as compared to other under-
sor Lewis is most familiar is that of the shel- developed regions more successful in control-
tered trade area of the British Common- ling inflation; he also writes off the wage
wealth countries. But Latin America was for spiral in Latin America as simply a political
all practical purposes, until the recent crea- problem, resulting from the low degree of
tion of LAFTA, an unsheltered trade area, sympathy of trade unions for their govern-
and the GATT report on "Trends in Inter- ments. Itis impossible to deny, however, that
there are vast differences between Latin
national Trade" presents conclusive evidence
that trade within sheltered areas has been ex- America and other underdeveloped regions
panding at a substantially faster rate than of Africa or Asia in a number of respects.
the trade of unsheltered areas. There is less surplus labor in Latin America,
Just as Professor Lewis appears to mini- the degree of labor organisation is greater,
mize unduly the importance of the decline in there is more responsiveness to Western con-
the import capacity of Latin America, the sumption habits and less passivity in claiming
"structuralists" of ECLA overestimate its social benefits. Even within Latin America,
explanatory and causal role in the process of there are substantial regional differences, the
inflation. As Professor Grunwald has pointed difficulty of preventing a wage spiral being
out, despite a decline in the export quantum directly related to the degree of labor union
over the last decade, the Latin American organisation, which seems greater in the Ar-
countries managed to maintain or even to in- gentine and Chile, much lower in Mexico,
crease their import quantum by depletion of Peru and Colombia, while Brazil holds an in-
exchange reserves and/or increase in inter- termediate position.
national indebtedness.2 The third stricture of Professor Lewis con-
cerns the alleged rigidity of agricultural pro-
2GrunwaId, "Invisible Hands in Inflation and
duction. He is right in pointing out that pres-
Growth," p. 1 1, paper submitted to the Rio de Janeiro sure on food supplies is not peculiar to the
Conference on Inflation and Development, January,
1963. [W. Baer and I. Kerstenetzky (eds.), Inflation underdeveloped countries of Latin America
and Growth in Latin America, Homewood, 1964.] and that structuralists underestimate the
 273 INFLATION AND ITS EFFECTS

possibility of corrective adjustments by in- b. the quantitative importance of the cost-

creasing the propensity to export or by reduc- push generated by import substitution, or by
ing the propensity to import. losses in import capacity through the decline
What might be said by way of conclusion in terms of trade, is greatly exaggerated;
is that the basic flaws in the structuralist ar- while those factors might account for a mod-
gument seem thus to lie in that: erate inflationary pressure, they are of little
a. no separation is made between autono- use to explain the massive and chronic infla-
mous structural rigidities and induced rigid- tion in Latin America,
ities resulting from price or exchange
controls or mismanaged government inter-
vention;

Comment
An excellent guide to the literature on the "structuralist school" is contained in Dudley
Seers, "A Theory of Inflation and Growth in Under-developed Countries Based on the Ex-
perience inLatin America," Oxford Economic Papers (June 1963). For additional readings
on inflation in Latin America, the following may be consulted: Werner Baer, "Inflation and
Economic Growth: An Interpretation of the Brazilian Case," Economic Development and
Cultural Change (October 1962); Werner Baer, "The Inflation Controversy in Latin Amer-
ica: A Survey," Latin American Research Review (Winter 1967); Werner Baer and I. Ker-
stenetzky (eds.), Inflation and Growth in Latin America (1964); David Felix, "Structural Im-
balances, Social Conflict, and Inflation," Economic Development and Cultural Change
(January 1960); David Felix, "An Alternative View of the 'Monetarist'-'Structuralist' Con-
troversy," inA. O. Hirschman (ed.), Latin American Issues (1961); G. Maynard, "Inflation
and Growth: Some Lessons to be Drawn from Latin American Experience," Oxford Eco-
nomic Papers (June 1961); Celso Furtado, "Industrialization and Inflation," International
Economic Papers, Vol. 12 (1967); Rosemary Thorp, "Inflation and the Financing of Eco-
nomic Development," in Keith Griffin (ed.), Financing Development in Latin America (1971).
The structuralist and monetarist explanations of inflation are also discussed, together with
a review of the theoretical and empirical evidence on inflation in LDCs, in C. H. Patrick and
F. I. Nixson, "The Origins of Inflation in LDCs: a Selective Review," in M. Parkin and G.
Zis (eds.), Inflation in Open Economies (1976).
General aspects of the problem of inflation in LDCs are also examined in the following: E.
M. Bernstein and I. G. Patel, "Inflation in Relation to Economic Development," IMF Staff
Papers (November 1952); Bent Hansen, Inflation Problems in Small Countries (1960); G. S.
Dorrance, "Rapid Inflation and International Payments," Finance & Development (June
1965); A. P. Thirwall, Inflation, Savings and Growth in Developing Economies (1974); P. J.
Drake, Money, Finance and Development (1980); W. R. Cline et al., World Inflation and the
Developing Countries (1981); M. J. Flanders and A. Razin (eds.), Development in an Infla-
tionary World (1981); Harvey Leibenstein, Inflation, Income Distribution and X- Efficiency
(1980); Anand G. Chandavarkar, "Monetization of Developing Economies," IMF Staff Pa-
pers (November 1977); Peter S. Heller, "Impact of Inflation on Fiscal Policy in Developing
Countries," IMF Staff Papers (December 1980).
 274 MOBILIZING DOMESTIC RESOURCES

IV.D.4. Causes and

Consequences of Inflation*
According to the structuralist view, there will corresponding increases in costs. The govern-
exist at any given time some sectors which ment budget, so it is argued, will also benefit,
are under-utilised and suffering from in- in the form of revenue from the 'inflation
adequate demand co-existing with others ex- tax', which is the loss of real purchasing
periencing excess demand and rising prices. power imposed by rising prices on holders of
Because of poor information flows, shortages money balances. So long as companies and
of entrepreneurial and capital resources, pol- the government have larger marginal propen-
icy biases and other factors making for small sities to save than the groups in society whose
elasticities of supply, large disequilibria will real incomes fall, total saving and investment
be necessary before economic forces respond will be increased. And if investment is the
to break the bottlenecks. Because of down- binding constraint on the growth of the econ-
ward stickiness, rising prices in excess de- omy, the rate of growth will accelerate.
mand sectors will not be offset by falling In opposition to this, there are others who
prices in the excess supply sectors and there argue that inflation is more likely to hamper
will thus be a continuous (but not very rapid) growth. Far from encouraging saving, rising
upward movement in the general price level. prices penalise it by eroding its real value. In-
The agricultural sector is likely to be impor- flation, itis further argued, tends to distort
tant in this context, with its small short-run the composition of investment in favour of
elasticities of supply and its tendency in quick-yielding, perhaps speculative, projects
many countries to lag behind the expansion while discouraging long-term investments in
of the remainder of the economy. A heavy research or heavy industry. Moreover, infla-
cost would be imposed in terms of reduced tion, the greater price instability that accom-
output and growth if total demand were held panies itand the expectation of counter-mea-
to such low levels that prices were stable even sures by the government will be imperfectly
in the low-elasticity sectors. At least at mod- anticipated, thus increasing uncertainty and
erate rates of inflation, a trade-off is thus pos- making forward planning more difficult. This
tulated between growth and inflation; a soci- will deter some investment altogether and re-
ety that gives priority to growth must be duce the productivity of the investments
willing to tolerate the inflation that comes which do occur. In the extreme case, hyper-
with it. inflation results in a flight out of money, a
Many economists accept the validity of breakdown in the allocative efficiency of the
this argument and of the view that govern- economy, severe bottlenecks, dysfunctional
ments should allow aggregate demand persis- changes in the distribution of income, and re-
tently to exert mild pressure on available re- tarded growth.
sources, as a stimulus to investment. Some However, it is argued, probably the most
have gone further positively to advocate infla- potent way in which inflation can harm
tionary deficit financing by the government growth is through its effect on the balance of
as a means of accelerating growth. The es- payments. If (a) inflation at home is faster
sence ofthis case is that this demand inflation than the general rate of world inflation and
will redistribute income in ways which raise (b) the exchange rate is fixed or imperfectly
saving and investment. Company profits (and flexible, then excess demand at home will
self-financed investment) will go up as firms spill over into additional imports, and export
find themselves able to raise prices without production costs will rise without compensat-
ing increases in world prices. The balance of
trade will thus be weakened and the econ-
*From Tony Killick, "Inflation in Developing
Countries: An Interpretative Survey," ODI Review I, omy may encounter a foreign-exchange
1981, pp. 4-6, 8-11, 13-15. Reprinted by permission. constraint.
 275 INFLATION AND ITS EFFECTS

These various arguments conflict in every to 25 per cent — but it is in that range that
way except one: they all treat growth as the much inflation occurs!
dependent and inflation as the independent It should, in principle, be possible to estab-
variable, with causality running from the lat- lish the nature of the relationship in question
ter to the former. It is, however, possible to through econometric testing. There are, how-
reverse this hierarchy, to argue that inflation ever, a number of complicating factors. First,
will be a diminishing function of the real there are many influences on the growth of
growth of the economy. Growth expands the an economy besides the behaviour of the
supply of goods and services which, if not ac- price level — influences which are difficult to
companied byequal increases in demand, can capture adequately in regression models.
absorb previously existing pressures of excess Second, the inverted U-shaped relationship is
demand. The growth of real incomes will in- non-linear and is thus not well suited to test-
crease the demand for money balances rela- ing by the techniques of linear regression.
tive to money supply, which will diminish any Third, it is possible for various of the influ-
excess supply of money and excess demand ences summarised earlier to be at work si-
for goods. multaneously but tending to cancel each
There is thus a wide range of viewpoints on other out: the top of the inverted-U may be
the likely connections between inflation and roughly horizontal over a range of inflation
growth, yielding quite different hypotheses values and this would result in weak corre-
and policy implications. It is, nevertheless, lations. Finally, there are difficulties in in-
possible to suggest a limited degree of con- terpretation, about the direction of causality
sensus within the profession. This would pos- between inflation and growth.
tulate an inverted-U-shaped relationship be- It is therefore unsurprising that few empir-
tween inflation and growth. At one end of the ical studies of the growth-inflation relation-
inflationary spectrum, it is suggested, there ship have arrived at strong statistical re-
would be general agreement that there is sult. . . Nevertheless, some studies have
some "safe," low range of inflation which will produced evidence consistent with the in-
be positively associated with growth because verted-U hypothesis — results which are not
it will stimulate capacity utilisation and in- individually strong but which collectively do
vestment or, more negatively, because the assume some substance. . . .
policies needed to avoid it would slow down Since the growth of an economy is likely to
the expansion of the economy. It is also in be strongly influenced by the rates of saving
this range that the growth of output will have and investment, an alternative approach to
its most noticeable effects in moderating in- the study of the influence of inflation is to
flationary pressures. At the other end of the measure its effect on these variables. Here
spectrum, there would probably be wide ac- too the evidence is not strong, with most re-
ceptance that very rapid inflation is harmful sults having unsatisfactory levels of signifi-
to growth, for reasons already outlined. If cane. . .
there is an inverted-U relationship, it implies So far the discussion has related exclu-
the existence of an optimal rate of inflation, sively to the connections between price move-
above which it becomes increasingly impor- ments and economic growth. But it is no
tant for the government to take corrective longer necessary to argue that growth is only
action. one aspect of economic development. We
However, only a limited consensus is must pay attention to the impact of inflation
claimed. Most would probably agree that sin- on the structure of the economy, on poverty
gle-figure inflation is in the "safe" range; and and on the distribution of income, although
that rates of above, say 25 per cent are likely the evidence is tentative. Take first the im-
to hamper the growth of output. The remain- pact on productive structure.
ing disagreement is about the effects of infla- It is intuitively plausible to think that ac-
tion within the notional range of 10 per cent celerating inflation will be associated with in-
 276 MOBILIZING DOMESTIC RESOURCES

creased relative movements of prices within losses will depend upon the nature of the in-
the economy. For one thing, there are liable flationary process; and upon the abilities of
to be considerable differences in the supply different groups in society to anticipate infla-
elasticities of different sectors. Indeed, tion and protect themselves against it. It is
Hirschman is among those who have advo- difficult to generalise but if we take the case
cated the use of inflation to stimulate output of demand inflation, with supply elasticities
by providing large incentives for fresh invest- in foodstuffs production lower than for most
ments in low-elasticity sectors. We are not other goods and a fixed exchange rate, then
aware of any direct evidence on the behav- the gainers are likely to include (a) the pro-
iour of relative prices during inflation in de- ducers and distributors of foodstuffs; (b)
veloping countries, but there is evidence from those who derive their incomes from profits
America that changes in relative prices are a (because final selling prices rise faster than
positive function of the general rate of costs when there is demand inflation); and (c)
inflation. distributors of imported goods (because
Changing relative prices are likely to in- buoyant demand conditions and a foreign ex-
duce changes in the composition of output if change constraint will result in a scarcity
they are of a more than transitory nature. premium on such goods). The losers from
Whether inflation persistently biases prices this type of situation will include (d) the eco-
in favour of particular sectors is not clear and nomically inactive (the unemployed, house-
Spraos has suggested, on the basis of theoret- wives, the aged); (e) the urban poor; (f) other
ical considerations, that the outcome will be members of the urban wage-labour force;
crucially influenced by whether the problem and (g) exporters (faced with rising costs and
is one of demand or cost inflation. In the case a fixed exchange rate).
of cost inflation, he suggests that relative If this is accepted as a likely outcome, then
prices will shift systematically in favour of it seems more likely than not that inflation
manufacturing; demand inflation will shift will increase inequalities, with the urban poor
relative prices in favour of primary produc- being especially vulnerable, although this
tion, of which foodstuffs is likely to be the conclusion would need to be modified if
most important component. From a different smallholder food farmers were substantial
standpoint, the structuralist school claims a beneficiaries.
persistent tendency for final food prices to
rise relative to the general level, although THE CAUSES OF INFLATION
they have supply bottlenecks and anti-agri-
cultural policy biases in mind, rather than There is much controversy, and a large ac-
general excess demand. There is, in fact, companying literature, on the causes of infla-
some evidence of a bias towards agriculture. tion in LDCs. We can make a beginning by
For example, the median inflation rate for exploring the relative importance of external
the 70 developing countries covered by [ a re- and domestic forces, for the global nature of
cent study] was 6.9 per cent p.a., while the inflation has become evident in recent years
median rate for food items alone was 8.4 per and some analysts have emphasised the infla-
cent. On alternative assumptions about the tionary effects of rising import prices. On the
weight of food in the total index, this implies basis of cross-country data, however, it is dif-
that food prices rose 35 per cent to 50 per ficult to believe that rising import prices
cent faster than non-food items. could directly explain more than a modest
Any tendency for food prices to outstrip part of LDC inflation. Remember that LDC
the general price level will also influence dis- imports are typically equivalent to about 20
tributional consequences of inflation, and the per cent of GDP, so that the direct impact of
same is true of other ways in which the price a 10 per cent increase in import prices should
level interacts with the productive structure. only be to add about 2 per cent to the general
More generally, the incidence of gains and price level. If we confine ourselves to impact
 277 INFLATION AND ITS EFFECTS

effects, even the large import price rises of widely varying inflation experiences
the post- 197 3 years could only account sta- among LDCs, nor the initiation of rapid
tistical y for a moderate increase in the infla- inflations experienced in a few of them.
tion rate. 5. Expansion of the supply of money more
Even more persuasive, however, is the fact rapidly than the growth in demand for it
that at all times LDC inflation rates have is sufficient to initiate inflation and essen-
been faster than the rate of increase in im- tial to keep inflation going. This was prob-
port prices. . . . ably the initiating force in at least some
It could, no doubt, be countered that the high-inflation countries. However, the im-
impact effects of rising import prices are pact of rising prices on the general public
magnified by the resulting attempts of wage- is limited to the extent that production
earners and others to protect their living and consumption still occur outside the
standards, so that even a modest initial im- monetised part of the economy. . . .
pact could lead to a large and continuing in-
flation. Itis doubtful, however, whether such [A summary analysis of cross-country and
powerful propagation mechanisms exist in individual-country studies is omitted.]
more than a minority of LDCs. We should The balance of the evidence . . . points
also remember the finding reported above of rather firmly to monetary expansion as the
a negative correlation between inflation and chief proximate source of inflation. If this is
openness. While the behaviour of import accepted, it has implications for the causes
prices no doubt contributed importantly to of the balance-of-payments difficulties also
the accelerated inflation of 1973-79, even for common among LDCs, for the excess de-
this period it is evident that there were also mand generated by monetary expansion will
powerful domestic forces at work. We there- increase the demand for imports and export-
fore turn to consider the relative importance ables in addition to pulling up domestic
of the various possible domestic sources of in- prices. Given government reluctance to de-
flation. Agood deal of the relevant literature preciate the currency in line with increases in
has already been surveyed by the present domestic prices and costs relative to those of
writer elsewhere and it is a useful short-cut the outside world, inflation tends to result in
to reproduce the main conclusions below, be- over-valued currencies, which reduces the in-
fore going on to consider further evidence. centives to export and to produce local im-
These were as follows: port-substitutes. Acceptance of the proposi-
tion that inflation frequently has roots in the
1. Both supply and demand factors contrib- monetary system thus lends support to the
ute to an on-going inflation. This makes it IMF view that a high proportion of countries'
difficult to establish the initiating cause. foreign-exchange difficulties have domestic
2. There is, in any case, no reason for think- origins.
ing that the initiating force will be the But some important qualifications are nec-
same for all countries, or at different times essary. First, while the available published
in a single country. evidence does lean towards monetarist expla-
3. Although import prices do have an infla- nations, there is by no means unanimity and
tionary influence, the cost-push model there are no grounds for believing that infla-
fails to provide an adequate explanation of tionary processes are uniform between coun-
the initiation of inflation in most LDCs. tries. One difficulty is that the structuralist
4. Structural considerations help to explain argument does not lend itself easily to statis-
why LDCs are generally more prone to in- tical testing, which may bias results in favour
flation than industrial countries. But the of alternative hypotheses. More fundamen-
inflationary effects of structural disequili- tally, the monetary factor can only offer a su-
bria are unlikely to be large in most cir- perficial explanation, which is why we have
cumstances and they cannot explain called it a "proximate" source of inflation.
 278 MOBILIZING DOMESTIC RESOURCES

Always assuming governments to have con- may thus be ambiguous about anti-inflation
trol over monetary aggregates (although they measures, and they are a group of key im-
may not), we are left with the question why portance. On the other side, the welfare costs
governments allow money supply to expand of inflation tend (a) to be diffused across the
so fast as to produce unwanted inflation and general consuming public but (b) to be par-
payments difficulties. ticularly concentrated on groups possessing
The answer is probably that effective mea- little political clout — the urban poor and var-
sures to halt the monetary expansion are at ious categories of economic dependents. Gov-
least as unpopular as inflation and foreign- ernments have to weigh the diffused and am-
exchange shortages themselves. Cutting back biguous unpopularity of allowing inflation
on government spending, imposing credit re- against the often more sharply focussed un-
strictions, increasing taxation are all mea- popularity of,and resistances to, counter-
sures liable to worsen unemployment in econ- measures. Inflation may cause less social dis-
omies already characterised by much harmony (and lose less popular support) than
unemployment; to reduce private consump- its alternatives, which helps to explain why
tion in countries with already low living stan- governments are so rarely willing to pursue
dards; and to reduce public-sector investment anti-inflation policies successfully for more
in a situation of capital scarcity. than temporary interludes.
It is also likely that those who gain from Even if the monetarist explanation is ac-
inflation (or who would suffer most from at- cepted, therefore, the management of infla-
tempts to control it) are among the politically tion isnot just a technical matter of regulat-
more powerful members of society, for the ing money supply and it is for reasons of this
business community can expect profit mar- kind that there is a growing literature offer-
gins to widen with demand inflation but to ing socio-political explanations of inflation.
narrow during periods of demand restraint. Economic stabilisation measures thus involve
The government itself is likely to stand highly political judgements and the sensitiv-
among the gainers. Even if government ex- ity of the issues helps explain why govern-
penditures often move ahead of tax receipts ments find it hard to pursue successful
in response to inflation the government will stabilisation or to accept the demand man-
still benefit from the lower real cost of servic- agement policies urged upon them by the
ing the domestically-owned public debt and IMF.
from the "inflation tax." Treasury officials
Comment
The structuralist view of instability, in the form of open or repressed inflation, has been
increasingly criticized by newer views on financial repression. Monetary mismanagement,
both in the form of excessive growth of the money supply and erroneous selective credit pol-
icies, holds institutional interest rates (particularly deposit rates of interest) below their mar-
ket equilibrium. This financial repression not only fails to overcome structural deficiencies but
actually exacerbates them. In contrast, it is argued that the rise in real interest rates that
results from the simultaneous introduction of a financial stabilization plan and the liberali-
zation of portfolio and interest rate regulations should promote growth. Going beyond mo-
netarists who only take into account the necessary restriction of the money supply, the pro-
ponents ofliberalization to overcome financial repression advocate financial reforms that raise
real interest rates and improve the process of financial intermediation.
The long-run negative relationship between inflation and growth in financially repressed
economies has been analyzed by the following: Ronald I. McKinnon, Money and Capital in
Economic Development (1973); E. S. Shaw, Financial Deepening in Economic Development
(1973); Maxwell J. Fry, "Saving, Investment, Growth and the Cost of Financial Repression,"
World Development (April 1980); Vicente Gablis, "Structuralism and Financial Liberaliza-
 279 INFLATION AND ITS EFFECTS

tion," Finance & Development (June 1976); Vicente Gablis "Financial Intermediation and
Economic Growth in LDCs,' Journal of Development Studies (January 1977); Warren L.
Coats and Deena R. Khatkhate (eds.), Money and Monetary Policy in LDCs (1980).
For a good survey, see Maxwell J. Fry, "Analyzing Disequilibrium Interest-Rate Systems
in Developing Countries," World Development (December 1982).
Comment
Stabilization programs to reduce the rate of domestic inflation and to correct unsustainable
balance-of-payments deficits have become a central policy issue for many developing coun-
tries and for the IMF. The IMF's criteria of loan "conditionality" have aroused considerable
controversy. See, for example, Sidney Dell and Roger Lawrence, The Balance of Payments
Adjustment Process in Developing Countries (1980); Thomas Reichmann and Richard Still-
son, "Experience with Programs of Balance of Payments Adjustment: Stand-by Arrange-
ments in the Higher Credit Tranches," IMF Staff Papers (June 1978); Joseph Gold, Condi-
tionality (1979); Manuel Guitian, Fund Conditionality: Evolution of Principles and Practices
(1981); John Williamson (ed.), IMF Conditionality (1983).
An excellent account of various stabilization programs is provided by William R. Cline and
Sidney Weintraub (eds.), Economic Stabilization in Developing Countries (1981). Some con-
clusions ofthat study are: (1) inflation has proved more difficult to deal with than balance-
of-payments problems; (2) both international and domestic factors caused stabilization prob-
lems in the 1970s, unlike the 1950s and 1960s when the absence of external shocks meant
that most stabilization difficulties were the result of domestic causes; (3) even in the 1970s,
however, the evidence from the country studies attributes greater weight to internal than to
external causes for stabilization programs; (4) another common pattern is that stabilization
prospects are affected by the government's credibility and therefore by the success or failure
of past attempts; and (5) it is clear that early action is crucial to successful stabilization with
minimum adjustment cost to real economic growth.
The recent experience with stabilization programs in Latin America in light of traditional
and monetarist models of inflation and the exchange rate is reviewed critically by Rudiger
Dornbusch, "Stabilization Policies in Development Countries: What Have We Learned?"
World Development (September 1982). Also relevant are Rosemary Thorp and Lawrence
Whitehead (eds.), Inflation and Stabilization in Latin America (1979); and Jere Behrman
and James A. Hanson (eds.), Short-Term Macroeconomic Policy in Latin America (1979).
See also Andrew D. Crockett, "Stabilization Policies in Developing Countries: Some Policy
Considerations," IMF Staff Papers (March 1981); John Odling-Smee, "Adjustment with Fi-
nancial Assistance from the Fund," Finance & Development (December 1982).
The response to external shocks is examined by Bela Balassa et al., The Balance of Pay-
ments Effects of External Shocks and of Policy Responses to These Shocks in Non-OPEC
Developing Countries (1981).
CHAPTER V

Mobilizing
Foreign
Resources

The overriding concern of this chapter is how to improve the process of transferring resources
from rich to poor countries — whether this transfer be in the form of public financial aid, loans
from private foreign banks, private foreign investment, or a nonmonetary transfer of mana-
gerial and technical knowledge. This is not only a matter of a greater amount of resources;
there is now also concern over the "appropriateness" of the transfers from the developed to
the less developed countries. In view of the redefinining of development objectives, it is now
important that the transfers to the poor countries contribute to more employment and greater
equality. It cannot simply be assumed that the stock of resources and the stock of knowledge
that now exist in the developed countries are in this sense "appropriate" for the late devel-
oping countries.
Section V.A focuses on issues that are of the most analytical and policy interest in the
transfer of foreign aid. Public financial aid — that is, concessional finance, or the "grant equiv-
alent" inthe capital inflow — has a twofold function. It supplements the LDC's low domestic
savings, and hence helps fill the resources gap or "saving gap," and also provides additional
foreign exchange and thereby helps fill the "foreign exchange gap." The "two-gap analysis"
of the role of external aid is also significant in indicating that one gap may be greater than
another ex ante: if, for example, the foreign exchange gap is greater than the saving gap,
foreign aid becomes the means of permitting the required imports, so that the full saving
potential can be realized and resources will not be left underutilized because of an import
bottleneck. The necessary identity of the savings-investment and export-import gap ex post
is brought about by a process of adjustment.
In terms of policies, we should recognize the costs of aid to the donor countries as well as
the benefits received through aid by the recipient countries. There is considerable controversy
over the contribution that aid makes to development; some economists dissent from the con-
ventional view by arguing that, as an instrument of development, aid is generally of limited

281
 282 MOBILIZING FOREIGN RESOURCES

value — let alone an indispensable instrument — because it cannot substantially affect the basic
factors that are needed to promote the material progress of the people in the aid-receiving
countries. Others, however, emphasize the need to improve the quality of the aid relation-
ship— not only from the standpoint of the more conventional "performance criteria," but now
also in the context of meeting the needs of target poverty groups, redistribution, and employ-
ment. The more socially oriented measure of "performance" associated with the redefinition
of development objectives raises some difficult questions for redirecting aid policy.
The related problem of debt servicing has become of increasing concern in view of the large
expansion in external public indebtedness by the developing countries during the 1970s —
especially to commercial banks. Although the commercial banks operated as efficient inter-
national financial intermediaries in recycling petrodollars through syndicated Eurocurrency
loans to the LDCs, the large accumulation of debt — much of it of short or medium matu-
rity— has raised fears of a liquidity crisis, if not one of solvency, for debtor nations. Analysis
of a country's creditworthiness is therefore of crucial importance, as discussed in the Note on
"External Debt and Country Risk Analysis" (section V.B).
If the real amount of resource flows through financial aid has diminished, it is important
to discover alternative means of providing aid. One alternative is related to the current de-
mands for international monetary reform. Differences of opinion now exist as to whether pro-
posals for an increase in international liquidity should also provide a link to development
finance. The Note on "The International Monetary System and Development Finance" (sec-
tion V.C) discusses this issue. Other measures of providing aid through trade policy will be
noted in Chapter VIII.
The discussion in section V.D is concerned with policy measures that a developing country
might take to obtain a more substantial contribution to its development program from private
foreign capital. To assess the potential contribution of private foreign investment, the Note
on the "Benefits and Costs of Private Foreign Investment" (V.D.I) first outlines the benefits
and costs of various forms of private foreign investment, viewed in terms of the recipient
country's development program. As an agent of private foreign investment, the multinational
enterprise merits special consideration. Against this conceptual background, we might better
appraise the nature of the bargaining process between host countries and foreign investors.
In the final section (V.E), we explore further the possibility of securing a more effective
transfer of managerial and technical knowledge to the LDCs. The technology that has been
imported from developed countries may have an excessive capital bias, be inappropriate for
the recipient country's factor endowment, and aggravate the problems of unemployment and
inequality. While the recipient country may want a technology that raises the labor-capital
ratio, it wants to avoid at the same time an increase in the capital-output ratio: efficient cap-
ital-stretching innovations are needed. This is not only a matter of new research and devel-
opment (R&D) efforts devoted to the requirements of a new technology. The problem of
technology transfer is of a much more far-reaching nature — ranging over issues that extend
from the composition of an "appropriate" product mix to various policies of bargaining over
the terms and conditions of technology importation.
 V.A. PUBLIC FINANCIAL AID
V. A. 1 . Calculation of Capital
Requirements*

The word "requirements" is used in this con- increase in the share of goods and services al-
text as indicating a need for a transfer of located toinvestment. An attempt is made to
goods and services in order to achieve certain estimate the effect upon the future growth of
targets; these may be target rates of growth income of any additional investment so as to
for the economies as a whole, or they may be be able to assess how much capital is needed
target rates of investment, although, where to achieve a given rate of growth. This is then
this is so, investment is normally thought of compared with the estimated ability of the
as being the means by which rates of growth economy to make available such a share of
of income are increased. The basic idea, how- resources. (In other words, how much saving
ever, isthat such estimates of capital inflow can an economy at a particular stage of de-
or capital requirements shall be related to a velopment carry out?) The difference be-
generally recognizable and agreed aim, and tween these two estimates will then give some
that this aim should include the transforma- indication of the required size of the capital
tion of the recipient economy in such a way inflow from abroad.
as to increase permanently its ability to con- Such estimates are made without any con-
tribute to economic welfare. They are not sideration ofthe specific form that the in-
normally regarded as subsidies to current vestments are likely to take. By contrast, the
consumption, although such capital flows second approach builds up such an overall es-
may include such subsidies within a suitable timate from below by a detailed study, on a
framework of foreign aid. project-by-project or a sector-by-sector basis,
of the need for capital. An examination is
then made of the possible sources of such
TWO APPROACHES capital locally, and a discrepancy will thus
emerge between the needs for the develop-
There are two possible approaches to mak- ment of all the projects under consideration
ing such estimates and both are based on the and the resources which can be made avail-
idea that increased income and wealth de- able. This difference is the required capital
pend fundamentally upon the application of inflow from overseas, the counterpart of
more capital, either to increase output di- which can be the estimate reached by the
rectly when used in combination with local first approach mentioned.
resources, or indirectly when the use of such These two approaches are different not
capital will lead to a more effective use of only in their nature, but in their provenance.
other resources. The first approach corresponds to the various
The first approach is to make an overall es- calculations of capital requirements which
timate ofthe uses that are made of the output have been made in the last 20 years by the
of goods and services, with particular refer- United Nations and by the U.S. Government.
ence to the share that is allocated to con- The second approach is, generally speaking,
sumption and the share that is allocated to the one followed by the World Bank in pro-
investment. It is then argued that the future ducing the estimates which have been devel-
growth of the economy depends on a suitable oped in recent years by its own thinking on
the subject.
There is a further important difference,
♦From E. K. Hawkins, "Measuring Capital Re-
quirements," Finance & Development, Vol. 5, No. 2,
however, in the manner in which this subject
1968, pp. 2-5. Reprinted by permission. has been treated by other international or-

283
 284 MOBILIZING FOREIGN RESOURCES

ganizations and the way in which it has been chase the necessary goods and services from
viewed by he World Bank Group. Estimates overseas no progress can be made.
made outside the Bank have all been con- Estimates made as a result of this ap-
cerned with the requirements aspect of capi- proach start from certain basic relationships
tal inflows. They begin by specifying a target which are generally accepted as holding true
rate of growth which is to be the aim of the for all countries. The usages of modern na-
developing country and then proceed to esti- tional accounting are designed to express the
mate the amount of capital which would be fact that the amount that can be invested in
needed to reach that target. In the estimates any country is identical with the amount that
made by the World Bank Group no explicit is saved; that is to say, only those goods and
target is used; instead, the estimates emerge services which are not consumed can be de-
from a multidimensional examination of the ployed toincrease future income through in-
economy in question and represent the vestment. At the same time, if these re-
amount of capital which the Bank feels might sources are to be supplemented from abroad,
be used in an effective way in the future, pro- such a flow of resources will appear in this
vided that certain levels of economic perfor- accounting framework as an excess of im-
mance are achieved and suitable policies are ports over exports. It will, in fact, always ap-
followed by the country concerned. . . . pear twice, first as the difference between in-
vestment and the amount that can be saved
within the economy and second as an equal
THE "TWO-GAP" APPROACH excess of imports and services over exports of
Attention is also focused, however, on what goods and services.
is thought to be a more important limitation
on development: a possible shortage of for- THE FOUR IMPORTANT
eign exchange, as a result of which countries MAGNITUDES
may be unable to acquire from abroad the
goods and services necessary for promoting Calculations of capital requirements pro-
domestic development. This approach has be- ceed on the basis of projections of the four
come known as the "two-gap" approach be- important magnitudes: savings, investments,
cause it operates in two dimensions; while exports, and imports. A target rate of growth
continuing to argue that development is a is specified, and then the amount of addi-
function of investment it also holds that such tional capital that will be required in order to
investment, which requires domestic savings, reach that target rate of growth is estimated.
is not sufficient to ensure that development This gives a figure for capital requirements
takes place. It must also be possible to obtain which can be compared with the likely avail-
from abroad the goods and services that are ability ofdomestic savings. At the same time
complementary to those available at home. it is possible to make projections of the likely
In most developing countries the structure of behavior of exports and imports. The former
the economy is so simple that it can produce will depend on the supply of goods and ser-
only a limited range of products when relying vices available, or likely to be available, for
solely on domestic sources. In these circum- export from the domestic economy, the state
stances an act of saving, by itself, even of the world markets, and the economic
though it releases resources for investment health of the developed countries which are
purposes, may not make available the correct the markets for such exports. A similar esti-
kind of resources. In physical terms a country mate can be made for import requirements.
may be unable to produce the cement, steel, All these projections (which are of course ca-
or machinery which go into the various proj- pable of such subdivision) can be made in-
ects required to raise income in the future, dependently ofone another, and it follows
although it may be able to make the neces- that there is very little likelihood that such
sary savings by cutting down on consump- independent projections will all arrive at an
tion. Unless these savings can be used to pur- answer which satisfies the basic accounting
 285 PUBLIC FINANCIAL AID

relationship already referred to above. Pro- influence on discussions of the subject. In

jected investments may well exceed projected their study, various constraints — a savings
savings; projected imports, on the other hand, constraint, the foreign exchange constraint,
may exceed the projected exports by a differ- and the limitation of absorptive capacity —
ent amount. However, national income ac- come into operation for a particular devel-
countancy demonstrates that the excess of in- oping economy at various stages of its
vestment over savings must necessarily be growth. They focus special attention on two
equal to the excess of imports over exports aspects: first, the savings behavior of the
and this must hold true at all points of time. economy concerned, or its ability to refrain
It follows that there is something inconsistent from consumption and allocate additional re-
about the projections that have been made in- sources toinvestment; and second, the ques-
dependently. The two-gap method, therefore, tion of absorptive capacity, viewed, in this in-
is essentially a way of ensuring that inconsis- stance, as a performance factor. In other
tencies, which may be implicit in other ways words, if absorptive capacity is increasing
of making projections, are brought out into over the years, it can be regarded as a mea-
the open. sure of the improved economic performance
Estimates of capital requirements made by of the country concerned, and of its ability to
this method, therefore, yield not one, but two make better use of capital from abroad. In
figures for consideration. . . . The use of this their work they spelled out the operation of
approach requires either that the projections these three constraints in terms of three dif-
include the necessary changes in the struc- ferent stages of development. An interesting
ture and the behavior of the economy which feature of this study is its extensive coverage
will make the two estimates consistent with on a country basis, since the model was ap-
each other, or that a decision has to be made plied to 50 developing countries and totals
that the larger of the two gaps will be the obtained for a spectrum of different
more significant and will be taken as the re- assumptions.
quired capital inflow from abroad. This three-phase approach can be useful
for expository purposes, but it can only be an
approximation to the real life problems of
A "THREE-PHASE" APPROACH many developing countries where the three
A variant on the approach has also been constraints often coexist and interact with
followed by . . . Hollis B. Chenery and Alan one another. It has been found, for example,
M. Strout, who, in a study1 prepared for the that the limitations of absorptive capacity
U.S. Agency for International Development may apply only to particular sectors of the
(AID), employed a three-phase approach, economy, while other parts of the economy
where an attempt was made to specify in are able to absorb more capital than can be
quantitative terms the constraint on devel- obtained from local sources of savings. This
opment arising from the limiit of absorptive coexistence of the various constraints on de-
capacity.2 This work has had a considerable velopment isin many ways a central charac-
teristic ofan underdeveloped country.
'Hollis B. Chenery and Alan M. Strout, "Foreign
Assistance and Economic Development," American One limitation of all the approaches de-
Economic Review, September 1966. tailed above is precisely that they operate at
2Absorptive capacity covers all the ways in which a highly aggregate level. Any aggregative ap-
the ability to plan and execute development projects,
to change the structure of the economy, and to real-
proach of this kind treats all units of invest-
locate resources is circumscribed by the lack of cru- ment as if each were adding to a volume of
cial factors, by institutional problems, or by unsuita- capital which is homogeneous. In practice
ble organization. Not only the structure of the this is only true of the monetary units in
economy but also the utilization of its existing capac- which capital and investment are measured.
ity will have an important bearing on a country's ab- It is convenient to measure investment and
sorptive capacity. Cf. J. H. Adler, Absorptive Capac-
ity: The Concept and Its Determinants, Brookings capital in monetary terms, but the money
Institution, Washington, 1965. values often only hide the significance of the
 286 MOBILIZING FOREIGN RESOURCES

specific nature of the investment item. Ce- will have been wasted. It is the underlying
ment once incorporated into a highway can- importance of this aspect of development
not be used for other purposes; steel once which emphasizes the project and sector ap-
built into a building must yield additional proach to the measurement of capital
wealth in that form, otherwise the investment requirements.
Comment
Some fundamental theoretical formulations of the foreign exchange constraint, two-gap
analysis, and foreign capital requirements are presented in the following studies: R. Mc-
Kinnon, "Foreign Exchange Constraints in Economic Development," Economic Journal
(June 1964); H. Chenery and I. Adelman, "Foreign Aid and Economic Development: The
Case of Greece," Review of Economics and Statistics (February 1966); H. Chenery, I. Adel-
man, and A. MacEwan, "Optimal Pattern of Growth and Aid: The Case of Pakistan," Pa-
kistan Development Review, Summer 1966; H. Chenery, I. Adelman, and A. Strout, "Foreign
Assistance and Economic Development," American Economic Review (September 1966); J.
Vanek, Estimating Foreign Resource Needs for Economic Development (1967); John Adler
(ed.), Capital Movements and Economic Development (1967); A. Sengupta, "Foreign Capital
Requirements for Economic Development," Oxford Economic Papers (March 1968); H. J.
Bruton, "The Two Gap Approach to Aid and Development: Comment," American Economic
Review (June 1969); D. Lai, "The Foreign Exchange Bottleneck Revisited," Economic De-
velopment and Cultural Change (July 1972); Vijay Joshi, "Saving and Foreign Exchange
Constraints," in Paul P. Streeten (ed.), Unfashionable Economics (1970), pp. 111-133.

V.A.2. Growth-and-Debt
Stages*
In discussing the ability of countries to ser- (b) "Conventional" loans, with interest
vice external debt, it is essential to distin- above 3 per cent but more typically between
guish among different categories of debt. A 5 and 8 per cent or higher and maturities
particular debt will or will not be serviced on over 10 years, including privately placed
schedule depending on the terms of other bonds, IBRD and regional development bank
loans.
loans contracted to finance the country's total
resource gap. It is, therefore, best to look at (c) Export or commercial credits, provided
the total blend of capital entering a country to finance export trade of industrial coun-
and consider the blend most suitable to the tries, with a variety of terms centering
country's present and prospective position. around 5-8 per cent interest (depending on
For this purpose, one may distinguish the fol- the extent of subsidy provided) and maturi-
lowing categories of official capital provided ties depending on the value of the contract
to developing countries: and the characteristics of the project for
(a) Concessionary financing, almost exclu- which the goods are used, but usually 10
sively public, including loans with interest of years or less.
3 per cent or less and grants. (d) Eastern European credits (character-
ized by low interest, about 2% per cent, and
*From Barend A. De Vries, "The Debt Bearing relatively short maturities, 10 years or less).
Capacity of Developing Countries — A Comparative
Analysis," Banca Nazionale del Lavoro Quarterly The "strongest" countries will be able to
Review, March 1971, pp. 12-18. Reprinted by finance their capital needs with predomi-
permission.
nantly hard loans — i.e. mostly with "conven-
 287 PUBLIC FINANCIAL AID

tional" loans or export credits. The weaker a graphical location, etc. Examples of these
country's ability to service loans, the softer achievements are the production of a suffi-
must be the blend, i.e. the greater its conces- cient range of import substitutes, overcoming
sionary component. stagnation in traditional export markets and
The factual analysis in the present paper is development of new export products and
based on data about grants and loans, with markets, and mobilization of sufficient do-
maturities over one year, provided by mestic resources to finance its own invest-
"donor" governments or guaranteed by recip- ments. As the gap starts to decline, the debt
contracted to finance it will continue to rise
ient governments. Assessment of countries'
debt bearing capacity must, however, also as long as new debts are needed to cover the
cover short-term credits and, more generally, resource gap. In fact, the debt continues to
the flow of finance from private creditors (in- rise even after the gap turns into a surplus
dustrial firms or banks) to private debtors and as long as the surplus is less than the in-
without government guarantee and including terest needed to pay debts outstanding. Once
short-term (mostly commercial) loans. They the surplus is large enough to cover interest
may at times bulk large in the capital flow to on outstanding debt the country will be able
individual countries. Commercial credits to start reducing its debts.
when excessive (or in arrears) have, at times, In the sequence of growth-and-debt stages
been consolidated into longer-term public described, a basic assumption is that income
debt. In addition, the assessment must con- rises in subsequent phases, i.e., the capital
sider the role played by private investment. mobilized domestically and borrowed abroad
The "servicing" of private investment does is effectively used to increase output. The
not proceed according to the same kind of more effective the use of capital the quicker
schedules as used for loan capital, but never- the country will be able to reduce its gap. Ac-
theless imposes a long-term claim on the cordingly, itwill be possible to distinguish
economy which must be taken into account in subsequent phases in which first the gap, next
analysing the country's balance of payments the debt and finally the income level in-
prospects. creases. In the later phases the gap declines
first and next the debt, while the income level
PRINCIPAL DEVELOPMENT continues to rise.G This sequence of events re-
sults in the following ap phases.
STAGES
The successive stages of development can
Phase Debt Income
be characterized by countries' dependence on
foreign capital, the level of debt contracted
and the level of income achieved. In the early IA High
Low Low
stage of development countries tend to have II A High Low
High
Low Low
a relatively small dependence on external III A High High
III B Low
High Middle
capital. They may even have a resource sur- IV A Middle
Low High
plus, being mostly exporters of primary prod- V
ucts, while the rest of the economy and soci- Low Low
ety is left undeveloped. Since the amounts to
be financed are small, external debt is also
small. As development gathers momentum, In the sequence of phases countries are de-
the external resource need — the "gap" — in- picted as evolving, over a period of years, a
creases and, after perhaps many years, stronger basic balance of payments position
reaches a peak. The gap will start to decline as measured by the relative extent of its de-
after the country has made certain critical pendence on external capital. It should not be
achievements, which will depend on its de- denied, of course, that payments positions
velopment strategy, its resource base, geo- may be subject to considerable short-term
 288 MOBILIZING FOREIGN RESOURCES

fluctuation or that, regardless of income country approaches the end of the sequence,
level, countries may follow policies which its need for external capital is reduced rela-
greatly affect their dependence on outside tive to its own resources and its ability to ser-
capital. The basic question is whether the vice debt improves.
country's policies promote growth and reduce (b) The more effective its policies are in
its longer term dependence on external assis- moving toward the next "phase" of the se-
tance. The gap may be kept small by direct quence, the greater its ability to service loans.
controls which may be harmful to growth. Clearly, debt servicing capacity will tend
Empirical evidence does not conclusively sug- to be greatest in the last phase. In this phase
gest that long-ierm development proceeds ac- countries may be able to finance their exter-
cording to the sequence presented. Various nal capital needs on hard terms even though
factors may, in fact, help countries to accel- their growth policies may not in all cases be
erate their move from one stage to another or very effective. Next in line would be coun-
may explain why countries do not neatly fit tries which have reached the higher income
into any of the phases presented: level while their debt is still relatively high.
(a) The debt burden may remain "low" be- For remaining phases the second consid-
cause the country has received loans on pre- eration iscritical. In the earlier development
dominantly soft terms. On the other hand, an phases a slow growth rate may require a very
unfavorable debt structure may increase the low average interest rate if countries' debt re-
debt burden even while the resource gap and payment capacity is not continuously and
the income level are still low. progressively to fall short of their debt
(b) A successful export orientation makes
it possible to reach middle level income with- obligations.1
out developing a large resource gap.
(c) Private investment finances a substan-
tial portion of the resource gap and enables
the country to keep down the debt burden. A
'In long-term debt-and-growth anaylsis a critical
number of countries manage to finance a sub- question is whether the debtor country can reach a
stantial portion of a relatively large resource point at which its "savings surplus" increases faster
than the interest on its debt. If, over the longer term,
gap with private capital and thus have
interest increases faster than the savings available to
at present a low public external debt
burden. . . . repay debt, the country's debt will increase continu-
ously. The interest rate at which this explosive debt
Against the background of the growth- situation occurs, the so-called critical interest rate,
and-debt sequence one can derive two basic can be derived from the country's growth parameters
(growth of product, the marginal savings rate and the
criteria debt:
service for determining a country's ability to capital output ratio). The slower the growth rate, the
lower the marginal savings rate, and the less produc-
(a) The more "developed" the country, the tive its capital investment the lower is the critical in-
greater its debt bearing capacity: as the terest rate.
EXHIBIT V.l. Total Flow of Resources from DAC Countries by Major Categories

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981

(a) Net Current Prices ($ billion)

Official Development
Assistance 6.9 7.6 9.2 9.1 11.6 13.8 14.0 15.7 20.0 22.8 27.3 25.6
Grants by private voluntary
agencies 0.9 0.9 1.0 1.4 1.2 1.3 1.4 1.5 1.7 2.0 2.4 2.0
Nonconcessional flows 8.1 8.9 10.0 11.9 9.7 29.7 30.6 34.3 48.9 50.6 45.7 60.4
Total flow of resources 15.9 17.4 20.2 22.4 22.5 44.8 46.0 51.5 70.6 75.4 75.4 88.0

(b) 1981 Prices and Exchange Rates ($ billion)

Official Development
Assistance 17.8 18.4 20.1 17.7 20.5 21.1 20.8 21.6 23.9 24.1 26.5 25.6
Grants by private voluntary
agencies 2.3 2.2 2.2 2.7 2.1 2.0 2.1 2.1 2.0 2.1 2.3 2.0
Nonconcessional flows 20.9 21.6 21.8 23.2 17.1 45.4 45.4 47.1 58.3 53.5 44.3 60.4
Total flow of resources 41.1 42.2 44.1 43.7 39.7 68.5 68.2 70.7 84.2 79.8 73.1 88.0

(c) As Percent of GNP

Official Development
Assistance 0.34 0.34 0.36 0.29 0.34 0.36 0.33 0.33 0.35 0.35 0.38 0.35
Grants by private voluntary
agencies 0.04 0.04 0.04 0.04 0.04 0.04 0.03 0.03 0.03 0.03 0.03 0.03
Nonconcessional flows 0.41 0.40 0.38 0.38 0.28 0.77 0.74 0.73 0.86 0.79 0.63 0.83
Total flow of resources 0.79 0.78 0.78 0.72 0.65 1.17 1.10 1.09 1.24 1.17 1.04 1.21

Source: OECD, Development Cooperation, 1982 Review of Members of the Development Assistance Committee (DAC), Paris,
November 1982, p. 177.

EXHIBIT V.2. Official Development Assistance from OECD Members

Amount (millions of U.S. dollars)

OECD 1960 1965 1970 1975 1978 1980 1981 1982

1977 1979
77 182 198 72
101 Italy 60 147 376 273 683 666 820
102 New Zealand — — 66 53 55 68 68 65
14
103 United Kingdom 407 472 500 904
105 458 1,114 1,465 2,157 1,852 2,194
104 Japan 244 79
1,148 1,794
105 Austria — 10 11 1,424
108 2,215 2,685
131 3,353
178 313
3,171 361
3,023
48 49 154
— 110 135
106 Finland 2 7 55 90 144
107 Australia 59 119 212 552 400 588 629 882
75 667 649
108 Canada 96 337 880 991
70 1,060 1,056 1,075
109 Netherlands 35 196 608 908 1,189 1,185
101 102 120 378 371 536
1,074 1,472
643 1,630
595 1,510
575 1,473
110 Belgium
497
1 1 1 France 823 752
971 2,093 2,705
112 United States 2,267 3,449 4,162 4,177 3,991
2,702 4,023
13 3,153 4,161 5,663 461
4,684 7,138 403 415
113 Denmark 5 59 205 4,682
258 481 5,782 8,302
223 388
114 Germany, Fed. Rep. 456 599 1,689 1,717 429
3,393 486 3,163
115 Norway 5 11 184 295 355
2,347 3,567 467
3,181 566
37 779 783
1 1 6 Sweden 7 38 117 566 988 962 919
173 987
117 Switzerland 4 12 30 104 119 213 253 251
237
Total 4,628 15,733 19,992 22,821 27,264 25,636 27,919
6,480 6,968 13,847
 290 MOBILIZING FOREIGN RESOURCES

EXHIBIT V.2. Official Development Assistance from OECD Members {Continued)

Amount ( as percentage of donor GNP)

OECD 1960 1965 1970 1975 1978 1979 1980 1981 1982
1977
.10 .08 .19
101 Italy .22 1.0 .16 .11 .14 .17 .24
.23 .33 .33 .38
102 New Zealand — — .52 .39 .34 .29 .28
.56 .45 .46 .35
103 United Kingdom .47 .41 .39 .52 .44 .29
.23
104 Japan .24 .27 .23 .21 .23 .27 .32
.23 .28
.48
105 Austria — .11 .07 .21 .22 .19 54
.27 .28 .30
106 Finland — .02 .06 .18 .16 .16 .22 .22
.59 .55
107 Australia .37 .53 .65 .42 .53 .48
.43 .41
.43 .57
.50 .48
108 Canada .19 .19 .41 .54 .52 .42
.36 .75 .82 1.08
109 Netherlands .31 .61 .86 .98 1.03
.50 1.08
.88 .46 .46 .55
110 Belgium .60 .59 .59 .59
.57
111 France 1.35 .76 .66 .62 .60 .60 .64 .73
.53 .25 .57 .74
1 12 United States .58 .27 .20 .20 .27
.32 .27 .27
113 Denmark .09 .13 .38 .58 .60 .75 .74 .73 .77
.40 .40 .33 .45
.77
114 Germany, Fed. Rep. .32 .44 .48
31
.11 .83 .90
.37 .47
115 Norway .16 .32 .66 .93 .85 .82 1.01
.05 .19 .82 .90 1.02
116 Sweden .38 .99 .79 .83 .25
.19 .97
117 Switzerland .04 .09 .15 .19 .20 .21 .24
.24
National Currencies
48
101 Italy (billions of lire) 37 92 119 175 227 585 757
319 78
102 New Zealand 1,109
(millions of dollars) — — 13 54 55 53 74 86
103 United Kingdom 66
(millions of pounds) 145 169 208 409 764
639 797
104 Japan (billions of 1,018 760 1,091 1,027
466
38 88 165 382 699 753
yen) 341 588
105 Austria (millions of
schillings) — 260 286
1,376 1,785 2,236 1,751 4,985
2,303 6,158
106 Finland (millions of
markkaa) — 6 29 226 410 583 694
177 197 351
107 Australia (millions of
dollars) 53 106 189 422 361 563 565 870
514 586
108 Canada (millions of 73
dollars) 104 353 895
109 Netherlands (millions 1,054 1,209 1,237 1,257 1,425
1,462
of dollars) 132 252 709
1,538 2,229 2,324 2,953 3,768
110 Belgium (millions of 3,241 3,933
francs) 13,902 16,880 17,400 21,350
5,050 5,100 6,000 13,298 18,852 22,708
1 1 1 France (millions of
francs) 17,589
4,063 3,713 5,393 11,139 12,207 14,674 22,700 26,230
1 1 2 United States 8,971
(millions of dollars) 2,702 4,023 3,153 4,161 4,682 4,684 7,138
113 Denmark (millions of 90 5,663
5,782 8,302
kroner) 35 443
1,178 1,549 2,140 2,425
1 1 4 Germany, Fed. Rep. 2,711 2,871 3,458
(millions of deutsche
marks) 937
1,824 2,192 4,155 3,987 4,714 6,219 6,484 7,189 7,675
 291 PUBLIC FINANCIAL AID

EXHIBIT V.2. Official Development Assistance from OECD Members (Continued)

National Currencies

OECD 1960 1965 1970 1975 1977 1978 1979 1980 1981 1982

1 1 5 Norway (millions of
kroner) 36 79 264 962 1,570 1,861 2,172 2,400 2,680 3,653

116 Sweden (millions of

kronor) 36 197 605 2,350 3,491 3,538 4,236 4,069 4,653 6,201
117 Switzerland (millions
of francs) 17 52 131 268 286 309 354 424 466 510
Summary

ODA (billions of U.S.

dollars, nominal prices) 4.63 6.48 6.97 13.85 15.73 19.99 22.82 27.26 25.63 27.92
ODA (as percentage of
GNP) .51 .49 .34 .36 .33 .35 .35 .38 .35 .39
ODA (billions of U.S.
dollars, constant 1980
prices) 16.41 20.19 18.15 21.60 21.91 24.09 24.89 27.26 25.82 28.37
GNP (trillions of U.S.
dollars, nominal prices) .90 1.30 2.00 3.90 4.70 5.70 6.50 7.20 7.30 7.24
ODA deflator .28 .32 .38 .64 .72 .83 .92 1.00 .99 .98

Net Bilateral Flow to Low-Income Countries (as percentage of donor GNP)

.06 .01 .02
101 Italy .03 .04 .02 .01 .01 .01 .01
— — — .03
102 New Zealand .14 .04 .02 .02 .13
.15 .16
103 United Kingdom .22 .23 .11 .11 .15 .11 .06
.12 .11 .08 .06
104 Japan .13 .11 .11 .02
105 Austria — .06 .05 .01 .07 .11
.02 .01 .02 .08 .08
— — — .06 .06
106 Finland .06
— .08 .10 .08
.04
107 Australia .09 .09 .07 .06
.10 .07
108 Canada .11 .22 .13 .13 .11 .31
.12
.24 .17 .30 .35
109 Netherlands .19 .08 .24 .33 .34 .26
.56 .24 .23
110 Belgium .27 .30 .31 .24 .28 .21
.12 .09
111 France .01 .09 .10 .08 .08 .10
.08 .07
.03 .03 .03 .03
112 United States .22 .26 .14
.20 .04 .26 .16
113 Denmark — .02 .10 .24 .21 .27 .25
.13 .14 .10 .12 .07 .08
114 Germany, Fed. Rep. .30 .07 .34
.02 .12 .25 .39 .07
.28
115 Norway .04 .40 .09
.41
1 16 Sweden .01 .07 .12 .37 .33 .28
.06
.10 .44
.05 .08 .06
117 Switzerland — .02 .05 .08
.18 .20 .13 .09 .09
Total .11 .09
.07 .07
Source: World Bank, World Development Report, 1983, p. 182.
EXHIBIT V.3. Non-Oil-Developing Countries: Balance-of-Payments Summaries, 1973-1983 (in billions of
U.S. dollars)

1973 1979 1980 1981 1983

1974 1975 1976 1977 1978 1982
-79.6 -52.2 -41.4
-74. -423.7 -381.8 -392.0
All non-oil developing -3-5
091. -401.1 3
-2-2
00 -238.1
-3 6. -93.7
.9
-34.6
-26.4
countries .2
5. 3 - 46.
-985.5 118.0 -158.3
5.8 -151.3 -40.8
-166.8
-27.0 -3.7 200.7 7 258.6
-5.6 175.0
139.6 -14.7
326.8
-28.1
344.1 329.6 350.5
Exports (f.o.b.) -3 3.3 117.4 -67.8
-10.3 -5.5
Imports (f.o.b.) -3.7 -86.8
-1.0 -89.0
Trade balance -41.3 -61.0
Net services and -28.9 -1.4 -107.7
-32.6
-46.3 -10.6 -32.2
private transfers -37.0
-11.3
Balance on current
account -4.1
7.5 12.5
Net official transfers 5.5 7.1 8.2 8.2 11.6 13.8 13.2 13.1
8.7 49.0 75.0 58.6
Net capital inflows 16.0 29.4 35.1 33.1 30.2 62.0 41.4 4.0
12.6 83.4
Overall balance 10.2 1.1 8.0 9.5 16.0

Source: IMF, World Economic Outlook. 1983, p. li

EXHIBIT V.4. Non-Oil-Developing Countries: Financing of Current Account Deficits and

Reserve Accretions, 1973-1983°
[973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983

In billions of U.S. dollars

Current account deficit 11.3 37.0 46.3 32.6 28.9 41.3 61.0 89.0 107.7 86.8 67.8
Increase in official 7.2
reserves 10.4 2.7 -1.7 13.0 12.5 17.4 12.6 4.5 2.1 -7.1 75.0
Total 21.7 39.7 44.6 45.6 41.4 58.7 73.6 93.5 109.8 79.7
Financed by
Non-debt-creating
flows, net 10.3 14.6 11.8 12.6 14.4 17.9 23.9 24. 28.0 25.1 24.2
Long-term capital
from official sources,
net 4.9 6.8 11.7 10.5 11.4 13.8 13.3 17.6 23.0 19.5 23.8
Long-term capital 40.3
from private sources 6.8 11.3 15.4 17.5 13.2 23.4 23.2 29.6 39.7 21.5
Other financing flows,
net -0.3 7.0 5.7 5.0 2.4 3.6 13.2 22.2 19.1 13.6 13.4

Distribution of financing In percent

flows
Non-debt-creating 32.3
flows, net 47.5 36.8 26.5 27.6 34.8 30.5 32.5 25.8 25.6 31.5
Official long-term
capital, net 22.6 17.1 26.2 23.0 27.6 23.6 18.1 18.9 21.0 24.5 31.8
Private long-term
capital, net 31.3 28.5 34.5 38.4 31.9 39.9 31.6 31.7 36.2 27.0 53.7
Other financing flows,
' net* 17.9
-1.4 17.6 12.8 11.0 5.8 6.2 18.0 23.8 17.4 17.1

Source: IMF, World Economic Outlook, 1983, p. 197.

"Excludes data for the People's Republic of China prior to 1977.
^Includes private short-term capital flows, exceptional financing, net use of Fund credit, and errors and omissions.

292
 293 PUBLIC FINANCIAL AID

V.A.3. Adverse Repercussions

of Aid*

Official Western aid has now gone to the Mayotte and Nigeria? Public and political
Third World for about thirty years, more discussion nowadays envisages the world as
than a human generation. Over this period one-third rich — the West — and two-thirds
major deficiencies, even startling anomalies,
poor or even hungry — the "Third World" or
have become apparent. These untoward re- "South." In this picture, extreme poverty is
sults might not matter much if the policy had the common feature of the Third World.
served to promote the well-being of the peo- This is altogether misleading. There is a
ples of the Third World, but it has not done continuous range of incomes in the world,
so. Only exceptionally and in the most pro- both between countries and within them,
pitious circumstances can aid promote or ac- making the line of division between rich and
celerate economic advance, and then merely poor countries quite arbitrary. One could say
to a minor extent. that the world is two-thirds rich and one-
The effects of foreign aid have been quite third poor, or one-tenth rich and nine-tenths
different. It is foreign aid that has brought poor, or choose any other two fractions which
into existence the Third World (also called add up to one. The size of the celebrated gap
the South). It thus underlies the so-called between rich countries and poor countries
North-South dialogue or confrontation. For- (i.e., the difference in their average incomes)
eign aid is the source of the North-South con- depends on the placement of the arbitrary
flict, not its solution. The paramount signifi- line of division. The picture is also misleading
cance of aid lies in this very important, in that many groups or societies in the Third
perhaps momentous, political result. World — expecially in the Far East, the Mid-
A further pervasive consequence of aid has dle East, Southeast Asia, and Latin Amer-
been to promote or exacerbate the politici- ica— are richer than large groups in the
zation of life in aid-receiving countries. This West.
major result has gravely damaged the inter- Nor is the Third World stagnant. In recent
ests of the West and the well-being and pros- decades many Third World countries have
pects of the peoples of Third World grown rapidly, as for instance have South
countries. Korea, Taiwan, Thailand, Malaysia, Singa-
The money spent by the West in no way pore, Jordan, Guatemala, Venezuela, Colom-
measures these crucial sequelae of aid. bia, Brazil, Kenya, and the Ivory Coast. In-
Whatever percentage of their national in- deed, in so far as global aggregation and
comes aid represents, the donor governments averaging of incomes and growth rates make
cannot wash their hands of the consequences any sense at all, both total and per capita in-
of their so-called caring. comes in the Third World as a whole have,
since 1950, grown no less fast than in the
West, and probably have grown faster.
THE CREATION OF
It is, of course, hardly sensible to lump to-
AND "SOUTH" NORTH" gether and average incomes of the very dif-
ferent societies of the Third World or South,
What is there in common between, say,
Thailand and Mozambique, Nepal and Ar- which comprise some two-thirds of mankind.
gentina, India and Chad, Tuvalu and Brazil, These societies live in widely different physi-
cal and social environments, display radically
different attitudes and modes of conduct, and
their governments pursue very different
*From Peter Bauer and Basil Yamey, "Foreign
Aid: What is at Stake?" The Public Interest, Sum- policies.
mer 1982, pp. 53-55, 57-61, 63, 67-69. Reprinted by But the diverse components of the Third
permission. World do indeed share one characteristic.
 294 MOBILIZING FOREIGN RESOURCES

This is not poverty, stagnation, exploitation, ment of any country, anywhere. There are, of
brotherhood, or skin colour; it is the receipt course, a number of Third World countries or
of foreign aid. The concept of the Third societies which have not progressed much in
World and the policy of official aid are insep- the postwar period. This lack of progress re-
arable. Without foreign aid there is no Third flects factors which cannot be overcome by
World. Official aid provides the only bond aid, and are indeed likely to be reinforced by
joining together its diverse and often antag- it.
onistic constituents. This has been so ever Governments or businesses in the Third
since practically all of Asia, Africa, and World that can use capital productively can
Latin America came to be lumped together borrow at home and abroad.1 This is also true
in the late 1940's, as the underdeveloped for governments borrowing to spend on the
world, and thereafter known successively as so-called infrastructure, i.e., on facilities
the less-developed world, the non-aligned which do not yield a directly appropriable re-
world, the developing world, the Third turn. Ifspending on infrastructure is produc-
World, and now the South. These expressions tive, itincreases taxable capacity so that the
never made any sense except in that they de- governments can readily service the bor-
noted a collectivity of aid recipients. . . . rowed capital. It follows that the absolute
maximum contribution of foreign aid to de-
AID AND ECONOMIC GROWTH velopment, inthe sense of the growth of the
national income, is the avoided cost of bor-
Although the case for official transfers is rowing, i.e., interest and amortization. As a
largely taken for granted, various arguments percentage of the national income for large
or rationalizations are often advanced. These Third World countries, this maximum con-
are addressed primarily to audiences not yet tribution isat best minute, and is far too
firmly committed. small to register in the national income sta-
The central argument for foreign aid has tistics. For India in recent years, the contri-
remained that without it Third World coun- bution of aid would have been at best on the
tries cannot progress at a reasonable rate, or order of one-quarter to one-half of 1 percent
cannot progress at all. But not only is foreign of recorded gross domestic product.
aid patently not required for development, it The maximum benefit from foreign aid is
is, in actual fact, much more likely to ob- thus a modest reduction in the cost of inves-
struct itthan to promote it. tible funds — a resource which is not a major
It diminishes the people of the Third independent factor in economic development.
World to suggest that, although they crave It is evident on reflection that the volume of
material progress, unlike the West they can- investible funds is not a critical determinant
not achieve it without external doles. Of of economic progress, for, if it were, countless
course, large parts of the Third World made poor individuals and societies could not have
rapid progress long before foreign aid — wit- advanced rapidly in a very short period, as
ness Southeast Asia, West Africa, and Latin they have done both in the West and in the
America. The emergence of hundreds of mil- Third World. The relative unimportance to
lions of people, both in the South and in the material progress of the volume of investible
West, from poverty to prosperity has not de- funds has been confirmed by much recent re-
pended on external gifts. Economic achieve- search, including that of Simon Kuznets, Ed-
ment has depended, as it still does depend, on
people's own faculties, motivations, and ways
of life, on their institutions and on the policies 1Where the political survival of a newly-created
of their rulers. In short, economic achieve- country is widely doubted, it may not be able to raise
capital even if it can use capital productively. These
ment depends on the conduct of people, in- apparent exceptions to the rule that development does
cluding governments. External donations not depend upon external aid are plainly irrelevant to
have never been necessary for the develop- the case for global transfers to the Third World.
 295 PUBLIC FINANCIAL AID

ward Denison, Moses Abramovitz, and Sir HOW AID CAN INHIBIT
DEVELOPMENT
Alec Cairncross.2 It is economic achievement
that produces assets and money; it is not as-
sets and money which produce economic The adverse repercussions of official aid
achievement. operate precisely on the personal, social, and
However, suppose that, contrary to the re- political factors which determine economic
sults of reflection and observation, the vol- development.
ume of investible funds was a critical deter- Most importantly, aid increases the
minant ofeconomic advance. If it were, then money, patronage, and power of the recipient
those actions of many Third World govern- governments, and thereby their grip over the
ments which restrict the inflow of foreign pri- rest of society. It thus promotes the disas-
vate capital — a course of action made easier trous politicization of life in the Third
by receipts of aid — would be evidence that World. . . . When social and economic life is
these governments gave a low priority to eco- extensively politicized, people's livelihood or
nomic development. The same applies also to even their economic and physical survival
the practice of several governments, includ- comes to depend on political and administra-
ing those of India and Nigeria, of in effect tive decisions. This result promotes conflict,
passing on to other governments part of the especially in the multiracial societies of most
aid they themselves receive. Third World countries. This sequence diverts
Any benefit from the reduction of the cost energy and attention from productive activity
of investible funds is likely to be small. It is to the political arena; and the direction of
also likely to be much more than offset by the people's activities is necessarily a crucial de-
adverse repercussions of official aid. And terminant ofeconomic performance.
these repercussions are brought about by There are further untoward implications
amounts of aid which, while small in relation and repercussions of foreign aid which are far
to the national income of recipient countries, from trivial. Aid enables governments to pur-
are nevertheless often a significant part of sue policies which patently retard growth and
their government revenues and of foreign ex- exacerbate poverty, and there is a long list of
change receipts. These are the relevant mag- such policies. These include persecution of
nitudes in assessing the principal repercus- the most productive groups, especially minor-
sion of these transfers, because aid goes to ities, and sometimes also their expulsion; re-
governments and it increases both the reve- straints on the activities of traders and even
nues and the external balances at their direct the destruction of the trading system; restric-
disposal. . . . tion on the inflow of foreign capital, enter-
prise, and skills; voluntary or compulsory
purchase of foreign enterprises (which de-
prives the country of skills very helpful to de-
2According to Kuznets, the contribution of the in- velopment, besides absorbing scarce capital);
crease inmaterial capital (both reproducible and non-
reproducible) to the increase of per-capita income forced collectivization; price policies which
over long periods in major developed countries was discourage food production; and, generally,
limited, "ranging from less than a seventh to not the imposition of economic controls which re-
much more than a fifth" (S. Kuznets, Postwar Eco- strict external contacts and domestic mobil-
nomic Growth, Cambridge: Harvard University
Press, 1964, p. 41). Much more important were im- ity, and so retard the spread of new ideas and
proved efficiency in the use of resources and the move- methods.
ment of resources from less productive to more pro- Aid also is apt to bias development policy
ductive sectors. See also P. T. Bauer, Dissent on towards unsuitable external models. Familiar
Development (London and Cambridge, Mass.: Har- examples include steel and petrochemical
vard University Press, 1971), ch. 2, and Equality, the
Third World and Economic Delusion (London and complexes and official airlines. Moreover, in
Cambridge, Mass.: Harvard University Press, 1981), some instances the adoption of external pro-
ch. 14.
totypes has gone hand in hand with attempts
 296 MOBILIZING FOREIGN RESOURCES

at more comprehensive modernization, in- poorest has very low priority.4 Indeed, to sup-
cluding attempted transformation of people's port rulers on the basis of the poverty of their
mores, values, and institutions. Such policies subjects does nothing to discourage policies
can have dangerous, even explosive, conse- that lead to impoverishment. Many Third
quences (as in Iran). World governments have persecuted and
Aid also impairs the international compet- even expelled the most productive groups,
itivenes ofeconomic activities in the recipi- such as the Chinese in Vietnam and Indone-
ent countries by helping to create or maintain sia, or the Asians in East Africa. On the cri-
overvalued exchange rates or to increase the terion of poverty, such governments then
domestic money supply.3 (These adverse ef- qualify for more aid, because incomes in their
fects could be offset if the transfers greatly countries have been reduced.
enhanced the productivity of resources, but These anomalies or paradoxes are ob-
this is extremely improbable.) Foreign aid scured when it is suggested that giving money
also makes it easier for governments to pur- to the rulers of poor countries is the same as
sue imprudent financial policies. Unless aid is giving it to poor, even destitute people. Giv-
increasingly forthcoming, these policies lead ing money to governments is certainly not the
to disruptive domestic inflation and to bal- same thing as helping the poor. On the con-
ance of payments difficulties, which in turn trary: Western aid to Third World govern-
are apt to engender a crisis atmosphere and ments, especially in Asia and Africa, has ex-
a flight of capital. This sequence encourages tensively supported disastrous economic
the imposition of specific controls, with ad- policies, which have greatly aggravated the
verse economic, social, and political results. lot of the poorest
It would be naive to suppose that adverse
policies would cease in the absence of aid. Many people think that world-wide redis-
tribution from rich to poor through foreign
What is pertinent is that aid makes it easier aid is a natural extension of domestic redis-
to pursue and continue these policies. This is tribution through progressive taxation. But
the case because, while it does little or noth- international transfers effected by foreign aid
ing to promote development, aid can relieve differ radically from internal transfers ef-
immediate shortages, especially of consumer fected byprogressive taxation (whatever the
goods and of imports. This makes it easier for case for the latter). Aid goes from govern-
governments to conceal temporarily from ment to government. Unlike progressive tax-
their populations the worst effects of the ation, aid transfers are in no way adjusted to
damaging policies. the personal circumstances of taxpayers and
recipients. Many taxpayers in donor coun-
BETWEEN AID AND THE POOR: tries are far poorer than many people in re-
THIRD WORLD RULERS cipient countries, in which aid largely bene-
fits the powerful and the relatively well-off
Foreign aid does not go to the pitiable fig- rather than the poor.
ures we see on aid posters or in aid advertise- Redistributive taxation postulates basic
ments— it goes to their rulers. The policies of similarities of conditions, and therefore of re-
these rulers who receive aid are sometimes
quirements, within its area of operation.
directly responsible for conditions such as Globally, these requirements differ widely.
those depicted. This is notably so in parts of The meaning of riches and poverty depends
Africa and Southeast Asia. But even where
this is not so the policies of the rulers, includ- ''This political reality has been recognized by oth-
ing their patterns of public spending, are de- ers. Governments "genuinely committed to improving
termined bytheir own personal and political the material standards of life of the mass of their pop-
interests, among which the position of the ulation" are "rare in the Third World." C. H. Kirk-
patrick and F. I. Nixson, "The North-South Debate:
3This effect of aid is discussed in Hans O. Schmitt's Reflections on the Brandt Commission Report,"
Three Banks Review (London, September 1981), p.
"Development Assistance: A View from Bretton
Woods," Public Policy, Fall 1973. 39.
 297 PUBLIC FINANCIAL AID

crucially on people's requirements, and thus of international taxation, as proposed for in-
on physical and social living conditions. This stance inthe recent Brandt report.
is obvious for physical conditions, notably cli- But these proposed reforms would make
mate. But it applies also to social conditions, little difference in practice unless the grant-
including customs and values. ing of aid were made deliberately discrimi-
Unlike progressive taxation within a coun- nating. Aid would have to be concentrated
try, global redistribution relies on interna- carefully on governments whose domestic
tional income comparisons. These are subject and external policies were most likely to pro-
to very large biases and errors. They greatly mote the general welfare of their peoples, and
understate the relative incomes of large notably their economic progress. Aid would
Asian and African countries compared to the have to go to governments which tried to
West. . . . achieve this end by effective administration,
the performance of the essential tasks of gov-
RESHAPING AID POLICY ernment, and the pursuit of liberal economic
policies. At present, many aid recipient gov-
Foreign aid cannot achieve its declared ob- ernments neglect even such basic tasks as
jectives and has far-reaching damaging polit- maintaining public security, while neverthe-
ical and economic results. Yet it will not be less trying to run closely controlled econo-
terminated promptly because of existing mies. Selective allocation of aid along these
commitments and because of the vested in- lines would reduce its propensity to politicize
terests behind it both in the government and life, and thereby it would reduce the extent
the market sectors, in the donor as well as in and intensity of political conflict. It would
the recipient countries. Can the worst effects also promote prosperity in the recipient coun-
of aid be mitigated by changes in its method tries, to the limited extent that external do-
of operation? nations can do so.
To begin with, it may help a little if aid Relief of need, especially humanitarian re-
were to take forms which made it possible to lief of poverty in the Third World, should be
identify its costs and possibly its benefits. left to voluntary agencies, notably to nonpol-
This rules out indirect methods of aid such as iticized charities. They are already active in
commodity agreements. Not only are their this field. They could do much more if it were
results perverse, but their overall impact can- recognized that relief of need belongs to their
not be assessed. And such schemes are not sphere. In this realm the international comity
subject to any public budgetary control. among countries calls for official aid only to
There may be some advantage in avoiding meet unforeseeable and exceptional disasters.
soft loans in favour of outright grants. The As for economic development, the West
latter avoid problems of measurement and of can best promote this by reduction of its
the confusion between gifts and loans. Aid often severe barriers to imports from poor
should also take the form of direct grants countries. External commerce is an effective
from donor government to recipient govern- stimulus to economic progress. It is commer-
ment rather than of payments to multilateral cial intercourse with the West which has
organizations to be allocated to the ultimate transformed economic life in the Far East,
recipients. Such direct grants permit a mod- Southeast Asia, part of Africa, and Latin
icum of control by the elected representatives America.
of the taxpayers in the donor countries, i.e., However, even removal of trade restric-
the real donors. The more distant the rela- tions may well do little for the economic ad-
tionship between the supplier of funds and vance of some Third World countries and
their user, the more likely that they will be groups. Where the enlargement of external
used ineffectively. Attempts to extend the opportunities would not bring about the eco-
multilateral component of aid should be re- nomic advance of particular societies, exter-
sisted; so also should attempts be resisted to nal donations to their governments would be
make aid more automatic by the introduction even less likely to do so.
 298 MOBILIZING FOREIGN RESOURCES

V.A.4. A Renewed Aid

Mandate*

1 . Does aid any longer matter — even to its re- that it has no close substitute — especially
cipients— compared to other policy instru- now, as we learn not only of the prospects . . .
ments? Development assistance will continue in the low-income countries of Africa but
to be enormously important to those very also of the continuing needs of the more pop-
poor countries — broadly, the low-income ulous, low-income nations of South Asia's
countries with per capita incomes of $400 or poverty belt.
less — which without it are not sufficiently Therefore whether aid still matters to the
creditworthy to finance the net imports they United States in good part comes down to the
need in order to develop. Most of the job of issue of whether or not these low-income
building agricultural and other capacities, countries matter to the United States. This
developing human and physical infrastruc- leads us to the next question, which concerns
ture, promoting exports, getting more of the
the purposes of aid from the donor's rather
benefits of growth to the poor, moderating
than the recipient's perspective.
population growth humanely, and achieving
broader participation in the development 2. Is aid a sufficiently coherent instrument
process must be done by these countries for advancing donor-country interests? The
themselves — and that is why ... an essential, U.S. Agency for International Development
if difficult, aspect of aid in the 1980s will re- (AID) and the other development coopera-
main the maintenance of a dialogue with host tion agencies and ministries that normally
governments concerning such self-help poli- represent their governments on the Develop-
cies. But even with their own best efforts, ment Assistance Committee [of the OECD]
these low-income countries for a long time to are all charged with acting as if the promo-
come will face real resource deficits which tion of Third World development were per se
they cannot cover with the combination of an important objective of their govern-
their unassisted export capacity, savings ca- ments— and this is indeed the bent of their
pacity, and human capital-building capacity. conscientious, professional staffs. This is of
Nor is any outside instrument but aid — not course the posture that OECD governments
commercial borrowing (except in unsustain- adopt when they participate in the World
able spurts), not foreign private investment, Bank and other multilateral, development-
and not private philanthropy on any imagin- promoting agencies — or, for example, when
able scale — an adequate or viable answer to they sign an International Development
the real resource deficits of these very poor Strategy for the Third U.N. Development
countries. . . .
Decade. Yet in formal as well as realistic po-
All members of OECD should be concen- litical terms, development promotion does
trating on better trade access and more pri- not quite ring true as an ultimate or basic
vate investment — but not instead of aid. All purpose of donor governments; for, in the
three instruments can deliver net benefits to usual view, the responsibility of the latter is
both developing and developed countries; to serve the interests of their own constitu-
when properly designed, they are not alter- ents, not to do good unto outsiders. Hence,
natives or competitors. And while aid of viewed descriptively, development promotion
course can be (and is) helpful to a consider- is better understood as an intermediate ob-
able array of middle-income developing jective toward which donor governments di-
countries, it is in the low-income countries rect aid and other policy instruments in order
to further certain of the basic goals that they
*From John P. Lewis, "Development Assistance in expect development promotion to serve.
the 1980s," in Overseas Development Council, U.S.
Foreign Policy and the Third World: Agenda 1982, The aid programs of each donor country si-
1982, pp. 102-8. Reprinted by permission. multaneously pursue a variety of purposes.
 299 PUBLIC FINANCIAL AID

The weightings given these purposes vary servation and renewable substitutes that will
across the set and for individual donors from diminish the exhaustion of nonrenewable re-
time to time. But all donor nations have sources (particularly energy), and finding
mixed purposes; indeed, commonly even a do- ways to make and keep the world's trade, fi-
nor's participation in a specific program or nancial, and monetary systems efficient and
project is variously motivated. The purposes viable.
can be grouped into three categories. These are matters of the liveliest self-inter-
First, economic aid to assist development est for the United States — and for every
often is given in part to further the donor's other country. Finding solutions clearly is a
national self-interests as traditionally con- "positive sum" game; all or nearly all hands
strued. These can include (a) strategic and will benefit. But the difference, compared
defense purposes; (b) a variety of political with matters of traditional national self-in-
and/or ideological purposes (e.g., to influ- terest, isthat in these cases the benefits can-
ence behavior in multilateral fora, to not be sought unilaterally or be sliced up into
strengthen cultural and historic ties with par- country-by-country gains. They are joint
ticular regions, to propagate social modes benefits requiring joint pursuit.
and institutions on which the donor places in- The relationship of these nontraditional
trinsic value); or (c) the donor's own eco- self-interest issues to an aid rationale is that
nomic and commercial interests, sought development promotion is an inescapable
through expanded export, increased access to part of any sensible strategy for global sys-
scarce materials, or new opportunities for tems maintenance. The positive linkage be-
private investors. tween development and a number of the sys-
Second, aid is given — again by all do- tem's sectoral needs — for example, in the
nors— partly for humanitarian, moral, or areas of food, energy, population, trade, and
ethical reasons. Whatever political theory the maintenance of orderly commercial fi-
posits against the credibility of external gen- nancial flows — is unmistakable. As for the
erosity bynation states, some of the latter's broader linkage between development and
constituents (both outside governments and global stability, . . . development, to be sure,
within the legislative and executive branches tends to unseat traditional elites and is not a
of governments) press for such action. This tranquil process. But in a Third World where
drive is clearest in response to sudden disas- change is already endemic — and in a Third
ters, but it also includes abiding concern for World that is more immune than before to
the poorest of the poor. Its force is easy to traditional instruments of power wielded
underestimate. Humanitarian reasons have from outside — those wishing to keep the level
tended to dominate aid policy in DAC's of turbulence manageable do well to bet on
front-runner countries for some time. . . . "development processes that succeed, with-
A third rationale for aid — one whose im- out excessive delay, in coupling growth with
portance isgrowing in the thinking of all do- gains in equity. . . . They are more likely than
nors— lies midway between these two ratio- non-development to yield a set of responsible,
nales. Ithas picked up a confusing variety of self-reliant, national participants in the
names — "mutal interests," "the manage-
ment of interdependence," and "world eco- global system. ..."
nomic security." Whatever the nomencla-
ture, .. . during the past ten or twelve years 3. Does aid work? Development assistance
policymakers have been waking up to what a is an awkward, difficult business. It operates
fragile and vulnerable system this planet is in unfamiliar cultures and has to deal across
and how greatly its survival depends on so- the interface between sovereign states with
lutions to a whole set of subsystem issues indigenous governments that are jealous of
such as environmental degradation, popula- their autonomy, often short on trained skills,
tion growth, balancing the world's food needs weakly managed, and sometimes venal and
and supplies, finding the combination of con- repressive. . . . Donors can try to concentrate
 300 MOBILIZING FOREIGN RESOURCES

their aid on countries making serious policy picked up by the World Bank and bilateral
and reform efforts to help themselves, but donors — aid has created a network of inter-
there is no way they can bypass the host gov- national agricultural research institutes com-
ernments, which almost always are the effec- prising the apex of a system which, as it be-
tive as well as the formal centers of power comes further articulated at both the
and decision making in their systems. The aid national and subnational levels, is injecting
input usually is a minor determinant of de- the most dynamic of all ingredients into the
velopment outcomes; it can catalyze internal transformation of traditional agriculture.
forces positively, but it can also fail to do The list of on-balance successes could be
so — or can be swamped by extraneous cir- many times longer. It constitutes no scientific
cumstances. Moreover, in times past, aid proof that aid, overall, has had positive and
sometimes has had adverse side effects (al- significant net benefits. But it fortifies a
though these are avoidable) that have damp- strong suspicion that without aid in the 1960s
ened productive incentives, delayed self-reli- and 1970s, Third World growth (which on
ance, or reinforced internal inequalities average was strong) would have been slower,
without contributing to a process that will the outcomes for the poor would have been
eventually reduce them. worse, social and political turbulence would
Thus it is obvious that the record of all de- have been greater, and less groundwork
velopment assistance by all donors up to now would have been laid for further advances. In
is mixed. It is easy enough to assemble a neg- particular, without aid, the poorest countries,
ative scorecard of mistakes, silly or counter- where growth was slower, would have lagged
productive projects, and other failures. But more.
there are some powerful counter points to be Second on the positive side is the factor of
made. experience — and the knowledge that has
First, there have been major and spreading come with it. Development assistance is now
achievements. Multifaceted aid from a num- in its fourth decade. ... To anyone who has
ber of donors was a catalyst — not the main followed aid activity over an extended period,
force, but a strong, indispensable catalyst — it is evident that the quality of aid operations
for the remarkable improvement in South has slowly continued to improve. Donor agen-
Asian food production that blossomed in the cies have learned a great deal about the de-
1970s. In the case of South Asian agricul- velopment process. They are doing better pol-
ture, "success" (by no means complete) was icy analysis. There is a keener appreciation of
achieved with a very thin spread of aid per the importance of management in both their
capita. In other cases — South Korea and own operations and those of recipient coun-
Taiwan, for example — very heavy concentra- tries and of the efficiencies that the latter
tions of economic aid (in both cases, interest- countries can gain from greater reliance on
ingly, for strategic reasons) have helped lay decentralized (including free-market) modes
the base for burgeoning economic expansion. of economic organization. There is more em-
There also has been a good deal of progress — phasis on incentives and on finding simpler,
failures as well, but mainly significant, ram- more flexible aid procedures without sacrific-
ifying successes — in helping build such key ing accountability. Donors generally are now
development-promoting institutions as agri- more sophisticated in dealing with recipients
cultural universities, technical institutes, and than they were in the past. They are rather
enterprise management training establish- better at not being counterproductively intru-
ments in many countries. Directly and indi- sive while at the same time promoting dura-
rectly aid has contributed to the downward ble self-reliance and reinforcing internal (in-
trend in birthrates that has begun to appear cluding redistributive and participatory)
in certain countries, especially in Asia. Al- reforms. A good part of the operational gains
most single-handedly — with the lead first are, moreover, attributable to the fact that
coming from American foundations and then many recipient governments are becoming
 301 PUBLIC FINANCIAL AID

more serious and businesslike development icism welled up from both these quarters. . . .
promoters. . . . However, these . . . critics. ... do not address
A third point about aid effectiveness re- the larger question of alternatives but only
lates to potential alternatives: Aid effective- score points on aid's conspicuous failings: aid
ness compared with what? In the case of the trafficks with repressive and/or corrupt re-
low-income countries particularly, there is no gimes; itimplements self-serving purposes; it
good alternative to aid for meeting needs fails to implement the right models of devel-
from which, as a matter of enlightened and opment adequately; it gives either too much
broadened self-interest, the rest of the world or too little attention to questions of equity.
cannot stand aside. The only rational course The critics are impatient; much of their quar-
is to improve aid, not scuttle it. In the end, rel seems to be with the nation-state system.
this is the answer to the spate of critics who If one posits that this system is likely to be
assault the activity. A dozen years ago those around for some time, their criticism pro-
who saw aid as a soft-headed enterprise were poses no actionable options to aid, and their
joined by critics on the New Left who, on the complaints are better read as a useful agenda
contrary, decried what they saw as neo-im- for further aid improvement than a case for
perialist efforts to shore up repressive Third aid abandonment.
World elites. It was no surprise that, in the The conclusion is that if there are good
past two or three years — a period of strained reasons for a renewal of the . . . aid mandate,
geopolitics, troubled economies, beleaguered one need have no qualms about the instru-
trade, and squeezed budgets, as well as slack ment being so flawed as to be unworthy of
leadership concerning aid — a chorus of crit- use.

Comment
Although a persistent question is whether foreign aid is successful, it is difficult to reach a
quantitative evaluation of the effectiveness of aid. To do so, the analysis at minimum would
entail three steps. First, it would have to be recognized that aid has multiple objectives, and
there would have to be as many separate assessments as there were identified purposes. It is,
however, unlikely that there is a way of adding up the collective assessments of differently
aimed activities into a single total. Second, it would be necessary to proceed, aimed activity
by aimed activity, to assess benefits — sorting out the parts and pieces of achievement attrib-
utable to public aid flows, gauging the degree to which the established objectives had been
attained, noting the efficiency with which inputs had been translated into outputs, and, if
account were to be taken also of inadvertent or by-product benefits, then identfying as well
the purposes against which these were to be measured. Finally, in order to arrive at activity-
by-activity estimates of net achievement, one would have to take account of the costs — ines-
capably inthe form of the alternative benefits foregone by doing this activity rather than some
other, but also very commonly in the form of unwanted side effects produced by the activity
undertaken.
These steps are proposed by the former chairman of the OECD's Development Advisory
Committee. (See OECD, 1980 Review, p. 54.) After reviewing the benefits and costs — even
if the complexities introduced by multiple objectives are set aside and the only focus is on
development promotion — the conclusion is that there is not "an unassailable, data-based
proof of development's macro-effectiveness that will quiet all critics and skeptics" (1980 Re-
view, p.62). There are, however, a number of robust (albeit arguable) propositions that sup-
port aid, and it is the nature of these propositions that the burden of proof should be placed
on those attacking them, not on the proponents of future aid efforts. The foregoing selection
has been argued in this fashion. (See also, Chapter III of the OECD, 1980 Review.)
 302 MOBILIZING FOREIGN RESOURCES

V.A.5. Improving the Quality of

Aid— Note

The donor nations as well as the recipients certain project is financing that particular
must now bear a responsibility for improving project: it is impossible to limit the effects of
the aid relationship. Specifically, there are aid only to the project to which aid is osten-
three major ways the donors might improve sibly tied. Aid is fungible, and the aid is ac-
the quality of foreign assistance: by untying tually financing the marginal project that
their aid, by giving more scope to program would not have been undertaken but for the
aid, and by operating more within a multilat- receipt of aid. Moreover, the efficacy of any
eral context. one project is a function of the entire invest-
Aid may be tied by both source and end- ment program; what ultimately matters is
use: the recipient country may not have the how the recipient country allocates its total
freedom to apply the aid to imports from investment expenditures. When aid is limited
sources other than the donor country, and the to specific projects, it becomes difficult to
use of aid may be restricted via specification provide more aid for education, agriculture,
of commodities or projects. Aid-tying has a small-scale industry, and administrative ser-
cost to the recipient countries; if it is tied by vices which are not visible as large projects
project as well as by source, the switching but which are extremely important for devel-
possibilities in the use of the aid are severely opment. Most significantly, it should be
limited, and the costs of tying can be quite noted that the subsequent uses of the income
significant to the recipient country. Aid-tying generated from a project are of more impor-
is essentially a protectonist device, reducing tance than the initial benefits from the pro-
the real value. The direct costs of tying aid ject. To determine the effectiveness of aid, it
can be estimated as the difference between is necessary to consider not only the initial in-
the cost of importing from the tied source and crement ofincome resulting from the receipt
the cost of importing the same commodities of aid, but also whether the increment in in-
from the cheapest source. come issubsequently used to relax one of the
In addition to the direct costs incurred, constraints on development, or is instead dis-
tying of aid may have significant indirect sipated inhigher consumption or used to sup-
costs for the recipient country by causing a port alarger population at the same low per
distortion of development priorities. The dis- capita level of income. Insofar as the most ef-
tortion in the allocation of investment re- fective use of aid depends upon the operation
sources can be especially deleterious by bias- of the whole set of development policies in the
ing the recipient's development program recipient country, program aid rather than
toward those projects that have a high com- project aid may be considered the more ap-
ponent of the special import content allowed
for under the conditions of the tied aid and propriate context from the start.1
The argument that the quality of develop-
avoiding those projects with a large amount ment assistance can be improved by untying
of "local costs" that cannot be covered by aid and shifting to nonproject aid is also an
aid. "Double-tying" — by donor procurement argument for extending the scale and range
and project restriction — can only too readily of aid efforts on a multilateral basis. In con-
artificially alter the relative priority of differ- trast tobilateral aid, multilateral aid has sev-
ent projects and bias investment toward im- eral distinct advantages: it is less influenced
port-intensive projects. by the donor's interests; the undesirable ef-
Although project-type loans have had most fects of tying are more easily avoided; it can
appeal to the donor countries, it can be ar-
gued that more development assistance
should be on a general program basis. It is 'See Hans W. Singer, "External Aid: For Plans or
Projects?" Economic Journal, September 1965, pp.
first of all illusory to believe that aid for a 539-45.
 303 PUBLIC FINANCIAL AID

more readily harmonize and improve the fi- mation of capital depends, in the last resort,
nancial terms of aid (unlike bilateral aid, on domestic action.
which allows one country to insist on hard It is appropriate therefore to emphasize
terms while another offers aid on soft terms); the necessity of self-help measures: unless re-
it facilitates coordination of aid programs cipient governments adopt policies to mobi-
among the various aid sources and with the lize fully their own resources and to imple-
development priorities of the recipient coun- ment their plans, the maximum potential
tries; and it provides the opportunity for more benefits from aid will not be realized. As the
aid consortia and consultative group arrange- record of foreign assistance in several coun-
ments to bring together the aid donors assist- tries shows, external aid may be incapable of
ing a group of developing countries. If yielding significant results unless it is accom-
regional integration can contribute to devel- panied bycomplementary domestic measures
opment, then attention should also be given such as basic reforms in land tenure systems,
to the means of distributing aid through re- additional taxation, investment in human
gional development institutions. Aid pro- capital, and more efficient government
grams can then avoid being piecemeal and administration.
fragmentary, and there can be more oppor- The success stories of foreign aid point up
tunity for the active participation of recipient the case for "performance" as the criterion
countries in the aid process. for allocating aid. The "performance crite-
What matters for securing the effective use rion" would have a country qualify for aid
of aid is not only its specific form or the terms according to its performance in accomplish-
on which it is rendered by the donor, but also ing such objectives as raising its marginal
the extent to which aid is successfully inte- rate of saving, lowering its incremental capi-
grated bythe recipient country into its devel- tal-output ratio, and reducing its balance-of-
opment plan. Clarity on the objectives of for- payments deficit.2 The purpose of this cri-
eign assistance is the necessary first step in terion isto allow aid to exert positive leverage
determining how much aid is needed by a re- in having the recipient countries meet speci-
cipient country. The essence of capital assis- fied standards in their development policies
tance isthe provision of additional resources, and to ensure that the limited amount of aid
but external assistance should add to — not is allocated to those countries where it will be
substitute for — the developing country's own most productive.
efforts. If financial assistance from abroad is Several objections may be raised, however,
to result in a higher rate of domestic invest- against the "performance criterion" for aid
ment, itmust be prevented from simply re- allocation. It does not allow a country to
placing domestic sources of financing invest- qualify for aid simply on the basis of "need";
ment or from supporting higher personal and yet, this may be an essential part of any
consumption or an increase in nondevelop- foreign assistance program. The "perfor-
mental current expenditures by the govern- mance criterion" is actually designed as a
ment.
measure of the recipient country's progress
When foreign aid is available on a general toward a termination of external aid, but it
purpose basis, the allocation of the foreign may be argued that the transfer of resources
capital is decisive in determining whether it from rich to poor countries should become a
contributes as much as possible to raising the permanent feature of the international econ-
growth-potential of the recipient country. omy. Indeed, the performance criterion side-
The efficient allocation of investment re- steps questions of equity: a country may
sources then depends upon the application of reach a state of self-sustaining growth and
investment criteria in terms of the country's thus not qualify for additional aid, but still
entire development program, and domestic
policy measures must be adopted to supple-
2H. B. Chenery and A. M. Strout, "Foreign Assis-
ment the use of foreign assistance. Regard- tance and Economic Development," American Eco-
less of the amount of aid received, the for- nomic Review, September 1966, pp. 728-9.
 304 MOBILIZING FOREIGN RESOURCES

have a per capita real income lower than an- Just as the absence of complementary do-
other country which has not yet reached self- mestic policies may limit the effectiveness of
sustaining growth and is still receiving aid aid, so too may its impact be neutralized by
even though it has a higher per capita in- changes in the other components of the total
come. Finally, the performance criterion flow of resources from rich to poor countries.
might be interpreted as simply indicating The total flow is affected by private foreign
that a country can absorb more capital, investment, export earnings, and the terms of
rather than that it needs aid: the fact that a trade, as well as by foreign aid. It is therefore
country receives low marks on performance essential to recognize the relationships be-
may be the very indication that it needs a tween capital assistance, private foreign in-
larger component of aid in its capital inflow. vestment, and international trade. The con-
It has been contended, for example, that tribution ofinternational assistance will be
many of the African countries require more greater if public loans and grants are not
aid proper. competitive with, but instead stimulate, pri-
Because the emphasis is now on the prob- vate foreign investment. Public aid for eco-
lems of absolute poverty, unemployment, and nomic overhead facilities can create oppor-
inequality, the performance criterion is less tunities for private investment, and the
controlling of aid allocations than simpler ev- private investment can, in turn, ensure fuller
idence of need. Special consideration must use of these facilities and raise their financial
now be given to the lowest income countries and economic return. Similarly, policies
that are extremely vulnerable to foreign ex- should be pursued that will bolster export
change needs. earnings so that the inflow of development
The use of income distribution criteria and capital will be able to do more than simply
employment criteria may also improve the ef- offset a weak trend of export earnings or a
ficacy of foreign assistance. An expansion of
deterioration in the recipient country's terms
lending activities in agriculture, education, of trade. Of particular concern now is the
and population sectors may support employ- need for policies that stabilize the poor coun-
ment and income distribution objectives. tries' foreign exchange earnings; this problem
More generally, the emphasis on these objec- is examined in section VIII.C, below.
tives now requires a shift in the orientation of Finally, attention should be given to ways
aid away from capital-intensive projects, in- of minimizing the burden of debt-service
vestments inurban rather than rural areas, payments. When the return flow of interest
projects in the modern rather than the tra- and amortization payments exceeds the in-
ditional and informal sectors, and away from flow of new capital assistance, the country
large rather than small projects.3 confronts a transfer problem and must gen-
Instead of imposing a rigid and quantita- erate an export surplus. If the country is to
tive interpretation of the performance crite- accomplish this without having to endure the
rion, itis more appropriate to interpret it as costs of internal and external controls, or cur-
simply reemphasizing the necessity of self- rency depreciation, its development program
help measures and the requirement that for- must give due capacity.
consideration to the country's
eign aid should avoid supporting domestic debt-servicing
policies that turn out to be counterproductive This becomes part of the problem of se-
to the recipient country's development. From lecting appropriate investment criteria. To
this viewpoint, the granting of program aid, provide for adequate servicing of the foreign
the insistence on self-help measures, and the debt, the inflow of capital should increase
concern over "performance," are all interde- productivity sufficiently to yield an increase
pendent elements in the aid relationship — a in real income greater than the interest and
relationship that is necessarily reciprocal amortization charges. If this is done, the
with respect to donor and recipient behavior. economy will have the capacity to raise the
necessary funds — either through a direct
3ILO, Time for Transition, Geneva, 1975, p. 62. commercial return or greater taxable capac-
 305 PUBLIC FINANCIAL AID

ity. Moreover, to provide a sufficient surplus cording to the criterion of productivity will
of foreign exchange to avoid a transfer prob- also be the most favorable for debt servicing,
lem, the capital should be utilized in such a since it maximizes the increase in income
way as to generate a surplus in the other from a given amount of capital, thereby con-
items of the balance of payments equal to the tributing to the growth of foreign exchange
transfer payments abroad. This does not availabilities.
mean that a project financed by foreign aid The problem of the growing burden of
must itself make a direct contribution to the debt-servcing will also be eased for the aid-
balance of payments, for the ability to create receiving country when capital assistance is
a sufficiently large export surplus depends on offered at lower interest, for longer terms and
the operation of all industries together, not with more continuity, and when the creditor
only on the use made of foreign investment. country follows a more expansionary domes-
As long as capital is distributed according tic policy and a more liberal commercial
to its most productive use and the excess policy.
spending associated with inflation is avoided, The problem of debt-servicing is discussed
the necessary export surplus can be created more fully in the next section — Note V.B.
indirectly. The allocation of foreign aid ac-

Comment
Although the previous Note suggests ways by which the effectiveness of aid might be im-
proved, the plea for more effectiveness should not be read as a substitute for a greater volume
of aid. The net flows to the poorest of the poor countries are grossly inadequate relative to the
needs posed by their lagging performance, and programs that are too small are also likely to
be ineffective.
The case for a greater volume of aid has been presented by the chairman of OECD's De-
velopment Advisory Committe, the World Bank's president in arguing for replenishment of
the International Development Association, and by the Brandt Commission.1 The latter of-
fered the following recommendations:
There must be a substantial increase in the transfer of resources to developing countries in order to
finance:
1. Projects and programmes to alleviate poverty and to expand food production, especially in the least
developed countries.
2. Exploration and development of energy and mineral resources.
3. Stabilization of the prices and earnings of commodity exports and expanded domestic processing
of commodities.
The flow of official development finance should be enlarged by:
1 . An international system of universal revenue mobilization, based on a sliding scale related to na-
tional income, in which East European and developing countries — except the poorest countries — would
participate.
2. The adoption of timetables to increase Official Development Assistance (ODA) from industrialized
countries to the level of 0.7 percent.
3. Introduction of automatic revenue transfers through international levies on some of the following:
international trade; arms production or exports; international travel; the global commons, especially sea-
bed minerals.
Lending through international financial institutions should be improved through:
1. Effective utilization of the increased borrowing capacity of the World Bank resulting from the
recent decision to double its capital to $80 billion.
2. Doubling the borrowing-to-capital ratio of the World Bank from its present gearing of 1 : 1 to 2 : 1 ,
and similar action by Regional Development Banks.

1North-South: A Programme for Survival, Report of the Independent Commission on International Devel-
opment Issues, Willy Brandt, Chairman, 1980.
306 MOBILIZING FOREIGN RESOURCES

3. Abstaining from the imposition of political conditions on the operations of multilateral financial
institutions.
4. Channeling an increasing share of development finance through regional institutions.
5. A substantial increase in programme lending.
6. The use of IMF gold reserves either for further sales, whose profits would subsidize interest on
development lending, or as collateral to borrow for on-lending to developing countries.
7. Giving borrowing countries a greater role in decision-making and management.
Resource transfers should be made more predictable by long-term commitments to provide ODA,
increasing use of automatically mobilized revenues, and the lengthening of the International Develop-
ment Association (IDA) replenishment period.
Consideration should be given to the creation of a new international financial institution — a World
Development Fund — with universal membership, and in which decision-making is more evenly shared
between lenders and borrowers, to supplement existing institutions and diversify lending policies and
practices. The World Development Fund would seek to satisfy the unmet needs in the financing struc-
ture, in particular that of programme lending. Ultimately it could serve as a channel for such resources
as may be raised on a universal and automatic basis.2

A number of reviews provide excellent critical surveys of the Brandt Report: P. D. Hen-
derson, "Survival," Development and the Report of the Brandt Commission," The World
Economy (June 1980); Graeme S. Dorrance, "North-South: The Need for Discussion," The
Banker (April-May 1980); Dudley Seers, "North-South: Muddling, Morality and Mutual-
ity," Third World Quarterly (October 1980); H. W. Singer, "The Brandt Report: A 'North-
western' Point of View," Third World Quarterly (October 1980). For a more radical view by
a structural critic of the international order, see Teresa Hayter, The Creation of World Pov-
erty: An Alternative View to the Brandt Report (1981).

2Ibid., pp. 254-5.

EXHIBIT V.5. Savings, Investment, and Balance of Payments (percent share of GNP)

Indicator 1960 1970 1975 1976 1978

1977

National Savings
Developing countries
Low income: Asia and Pacific 14.1 16.8 18.9 19.5 19.9
11.7
Africa south of 9.3
Sahara 10.8 14.1 8.4 10.0 6.6
Middle income: Africa south of
Sahara 15.5 16.8 23.9 24.4 25.0 26.3
North Africa and
Middle East 20.6 20.8 22.1 22.6 22.4
17.7
Asia and Pacific 10.3 20.6 25.8 26.2 26.3
22.4
Latin America
and Caribbean 19.4 20.0 20.6 21.8 22.8 21.8
Europe 21.1 22.5 20.2 20.1 20.9
18.0 21.5 19.7
Total 20.0 20.4 22.3 22.4
22.3 52.9 51.2 49.1
Capital surplus oil exporters 36.5 47.7
Industrialized countries 23.2 24.7 20.9 22.0 22.0 22.1

Investment, Including Stock

Developing countries 20.3
Low income: Asia and Pacific 14.5 16.7 20.6 20.9 21.3
Africa south of
Sahara 12.2 16.8 17.9 16.5 15.5
16.7
Middle income: Africa south of
Sahara 20.3 22.1 28.0 27.4 27.4 28.2
North Africa and
Middle East 24.0 23.3 32.4 32.6 34.8 32.4
Asia and Pacific 15.1 25.0 28.6 27.2 28.9
27.2
Latin America
and Caribbean 21.1 22.0 24.4 24.4 24.8 25.3
Europe 22.5 25.5 26.1 25.2 25.7
25.3 23.7
Total 20.2 23.1 25.8 25.5
25.7
30.8 19.3 26.1 29.3 31.6
-0.6 -1.3
30.9
Capital surplus oil exporters
-0.2
21.8
Industrialized countries 22.3 23.8 -2.2
20.9 22.4
-0.9 21.9 -3.2
-2.0 -0.8
Balance of Payments — Current -8.1 -2.7 -2.3
Account Surplus/Deficit -1.6 -1.3
Developing countries -4.8 -3.4
Low income: Asia and Pacific -3.6 -9.7
Africa south of -7.6 -9.4 -1.4
-0.3
Sahara — -2.4
-8.1 -0.9
-1.1 -5.5 -3.4
Middle income: Africa south of -2.3 -2.8
Sahara —-0.2 -3.1 -1.4
-4.1 -3.8
North Africa and -1.8 -3.9 -2.5 -2.6
-1.4 -4.6
Middle East -1.1 -2.9
-4.4
Asia and Pacific -1.7
Latin America -0.2
and Caribbean -0.1
Europe
Total —0.8
Capital surplus oil exporters — 2.1 20.5 19.9 16.2 14.1
Industrialized countries 0.4
0.2 0.4 0.4
Source: World Bank; and IMF, International Financial Statistics.
 308 MOBILIZING FOREIGN RESOURCES

EXHIBIT V.6. Non-Oil-Developing Countries: External Debt, 1973-1983° (in billions of U.S. dollars)
1973 1975 1976 1977 1978 1979 1980 1981 1982 1983
1974

Total outstanding debt of non-

oil developing countries 130.1 160.8 190.8 228.0 278.5 336.3 396.9 474.0 555.0 612.4 664.3
27.3 85.5 102.2 112.7 92.4
Short-term debt 18.4 22.7 33.2 42.5 49.7 58.8
111.8 138.1 163.5 194.9 235.9 286.6 338.1 388.5 452.8 499.6 571.6
Long-term debt
By type of creditor 172.4
Official creditors 51.0 60.1 70.3 152.9 193.2 218.7
82.4 98.7 117.5 133.0
Governments 37.3 43.4 50.3 57.9 67.6 79.1 87.2 98.7 108.6 120.4 135.3
International institutions 16.6 20.3 24.8 31.0 45.8 54.2 63.8 72.8 83.3
13.7 38.4
Private creditors 60.8 77.9 95.1 114.8 137.3 169.1 205.1 235.6
77.5 280.4 306.4 353.0
29.3 36.0 40.8 45.9 51.4 56.4 67.3 96.7 103.9 113.7
Unguaranteed debt
Guaranteed debt 31.5 42.0 52.4 66.6 85.9 112.7 137.8 158.1 183.7 202.2 239.3
Financial institutions 17.3 25.6 49.0 59.1 79.5 102.9 121.6 144.5 159.5 193.8
36.7 45.5
14.2 16.3 17.6 19.8 26.8 33.2 36.5 42.7
Other private creditors 34.9 39.2
By analytical group
70.5 96.5 129.0
Net oil exporters 20.4 26.0 34.1 42.4 53.3 61.2 79.4 108.1
152.5 391.5 442.6
Net oil importers 91.4 112.1 129.4 182.7 225.4 267.6 309.1 356.2
Major exporters of
73.1 145.2
manufactures 40.8 51.7 60.9 85.2 108.1 127.7 73.0 184.3 212.4
170.6
Low-income countries 25.4 29.7 33.2 38.3 46.5 53.1 59.5 67.0 80.1 90.8
25.2 30.6 35.3 41.1 51.0 64.2 96.9 112.7 127.1 139.4
Other net oil importers* 80.4
By area 26.9 75.0
Africa 14.2 17.7 21.9 35.0 42.1 49.6 60.5 67.1
55.1 100.8
Asia 30.0 34.6 39.8 46.4 57.9 67.4 76.1 115.1 131.7
88.4
Europe 14.5 17,2 20.0 23.4 28.7 38.2 49.0 57.5 69.2 73.8
10.3 13.3 63.4
20.3 24.7 28.4 32.9 35.4 39.3 43.7
Middle East 8.7 16.1
82.0 94.0 114.3 135.1
Western Hemisphere 44.4 58.2 68.6 154.7 192.6 208.9 247.4

Sources: World Bank, Debtor Reporting System; and Fund staff estimates and projections. IMF, World Economic Outlook
1983, p. 200.
"Excludes data for the People's Republic of China prior to 1977.
6Middle-income countries that, in general, export mainly primary commodities.
 309 PUBLIC FINANCIAL AID

EXHIBIT V.7. Non-Oil-Developing Countries: Long-Term and Short-Term External Debt Relative to
Exports and to GDP, 1973-1983* (in percent)
1973 1975 1976 1980 1981 1982 1983
1974 1977 1978 1979

Ratio of external debt to exports

of goods and sen ices* 104.6 122.4 125.5 119.2 112.9 124.9 143.3
All non-oil developing countries 115.4 126.4 130.2 144.4
By analytical group 179.5
154.7 124.9 169.5 179.3 176.9 144.3 154.6 192.2
Net oil exporters 162.4 128.4
Net oil importers 109.4 100.9 115.4 117.5 116.8 121.8 114.1 109.5 118.3 135.1 134.3
Major exporters of
manufactures 91.7 88.6 103.0 103.3 99.5 101.1 96.9 94.0 100.6 116.2 114.2
Low-income countries 227.9 214.5 226.1 225.1 217.8 226.3 209.8 201.4 231.1 254.1 262.9
Other net oil importersc 96.9 84.7 98.3 104.3 111.6 124.8 115.5 110.9 121.9 138.0 136.6
By region
Africa 71.5 80.9 94.2 103.1 111.4 100.8 97.4 119.9 148.6
65.4 72.5 147.4
Asia 92.9 81.0 91.6 84.4 83.3 77.7 70.2 68.2 80.9
97.1 121.6 118.2 129.6 85.4
Europe 102.4 108.0 114.9 127.4 136.4 125.6
129.7
Middle East 145.4 105.2 131.5 137.5 140.4 142.4 133.8 113.1 112.6 134.3 133.1
Western Hemisphere 176.2 163.4 195.8 204.1 194.1 211.5 192.9 178.4 207.4 245.6 242.8
Ratio of external debt to GDP6 285 27.6
22.4 21.8 23.8 25.7 27.4 27.5 31.0
All non-oil developing countries 24.7 34.7
By analytical group
26.2 25.5 38.5 34.0 43.5
Net oil exporters 27.7 32.3 39.3 37.4 36.1 44.7
Net oil importers 21.7 21.2 23.0 24.4 25.4 26.6 25.8 26.3 29.7 32.1
32.7
Major exporters of
manufactures 20.2 19.6 22.2 22.7 23.9 25.1 24.6 25.1 29.3 33.2 33.8
20.1 20.1 20.9 24.9 24.0 24.4 23.6 26.2 26.5
Low-income countries 24.4 24.7
26.2 25.2 26.2 27.7 28.6 31.5 28.8 30.6 34.1 35.8 35.6
Other net oil importers^
By region
Africa 19.4 19.6 21.6 25.8 29.4 28.9 28.8 30.6 35.2 35.1
28.4
Asia 18.9 22.3 25.2 27.1
19.7 20.4 22.4 23.4 22.2 23.2 26.7
Europe 24.5 23.1 22.8 24.6 28.6 25.1 29.0 33.1 34.5
25.7 34.7
Middle East 36.2 34.0 39.0 42.3 45.4 48.3 56.0 52.6 51.3 50.3 47.4
27.0
Western Hemisphere 23.0 22.8 25.5 26.4 28.4 30.3 28.8 38.2 38.6
31.9
Memorandum items
Ratios (including People's
Rep. of China)6
To exports:
All non-oil developing
countries — — — — 123.6 127.0 116.1 109.4 119.8 136.5
— — — — 137.7
Low-income countries 169.0 170.0 152.0 137.7 140.0 148.7 155.5
To GDP:
All non-oil developing
countries — — — — 23.9 24.6 23.9 24.1 27.3 30.5 30.5
Low-income countries — — — — 14.3 13.5 13.2 14.3
15.1 15.3
13.4

Source: IMF, World Economic Outlook 1983, p. 201.

"Excludes data for the People's Republic of China, except where noted.
*Ratio of year-end debt to exports or GDP for year indicated.
cMiddle-income countries that, in general, export mainly primary commodities.
 V.B. EXTERNAL DEBT AND
COUNTRY RISK ANALYSIS-
NOTE

The increase in lending to governments of de- of return to capital in the least productive
veloping countries by private commercial project is greater than the interest rate on the
banks has raised questions of country credit- most expensive loan, there should be no debt
worthiness. When banks engage in cross- service problem.
border lending they are obviously concerned In practice, however, it is not operationally
about the capacity of the borrowing country feasible to find the "marginal project" and to
to service the debt — that is, to pay interest calculate whether the marginal return is
and amortization in foreign currency. To ap- greater than the marginal cost of borrowing.
praise the creditworthiness of the borrower, To circumvent this difficulty, the analyst
lenders engage in country risk analysis to de- must shadow price the project and also un-
termine the borrower's ability to generate suf- dertake macroanalysis to analyze the ulti-
ficient foreign exchange to meet debt service mate effect of the loan on the total use of re-
obligations. For this analysis there is no for- sources inthe borrowing country.
mula or one definitive approach: it depends When shadow pricing the project, the an-
on the analyst's perception of the country's alyst corrects for price distortions to reflect
development process and how this relates to true social costs — raising the rate of interest,
the country's balance of payments. A debt lowering the wage rate, and increasing the
service crisis — that is, the borrower cannot or cost of foreign exchange. Externalities are
will not repay on schedule — is in essence a also calculated. Instead of focusing on simply
balance-of-payments crisis, and this, in turn, the internal financial rate of return, the
is a development crisis.1 Country risk analy- shadow pricing allows calculation of the in-
sis therefore depends on an understanding of ternal economic rate of return. If, in addition,
the development process and how the course some judgment is made on the distribution of
of development affects the balance of income, the internal social rate of return may
payments. be calculated. (This is explained later in
Even though the loan is ostensibly for a Chapter X.) This process of shadow pricing
given project, it supports in reality the mar- for the rate of return is equivalent to deter-
ginal project — the project that would not mining whether the effective rate of interest
have been undertaken but for this loan. The (nominal rate of interest corrected for infla-
loan helps to fill the savings gap and foreign tion, aid-tying, and changes in terms of
exchange gap by making more resources trade) is greater then the accounting rate of
available for capital formation and by in- interest (social return to capital).2
creasing the country's capacity to import. Be- Even though a project meets the social ef-
cause capital is fungible, the loan actually ficiency criterion based on shadow pricing,
supports the country's development program there is still the problem of whether the coun-
and balance of payments instead of being try can transfer a portion of the increased
limited to some given project. So too is the output and income into foreign exchange for
debt service on the loan a charge on the econ- debt-servicing. To determine the capacity to
omy as a whole — a charge against the total service external debt, some analysts engage
use of resources. Provided that the social rate
2For details of evaluating foreign capital inflows in
a cost-benefit framework, see Deepak Lai, "The Eval-
'Cf. Goran Ohlin, "Debts, Development and De- uation of Capital Inflows," Industry and Develop-
fault," inG. K. Helleiner (ed.), A World Divided, ment, No. 1, 1978, pp. 2-19. In Chapter X we con-
1976, pp. 207-22. sider project appraisal in greater detail.

310
 311 EXTERNAL DEBT AND COUNTRY RISK ANALYSIS— NOTE

in ratio analysis, examining such ratios as kets may have to be deepened to mobilize
debt outstanding/GNP, debt service/ex- local savings. The foreign trade regime may
ports, imports/international reserves, im- have to be liberalized and the bias against ex-
ports/exports, and so forth. If the numerator ports removed. All these policies will affect
in the fraction increases relative to the de- the debtor country's performance. These pol-
nominator, debt
a crisis might be indicated. icies are frequently suggested when the IMF
Ideally, one would like an index that would imposes conditionality on a member coun-
show the risk of a sharp fall in any kind of try's access to drawing rights in the higher
foreign exchange inflow and a rise in import credit tranches or in standby arrangements.
needs, compared with the ability to offset When the Fund imposes conditionality, this
such risks by compressing imports rapidly, may also encourage commercial banks to
obtaining compensatory financing, or using provide longer-term capital inflows.
international reserves. No one ratio captures This concentration on the policies neces-
this. Therefore, several ratios may need to be
sary to improve
in order to servicethetheeconomy's
debt alsoperformance
illuminates
used as proxies. But again, neither the history
of rescheduling cases nor the econometric other issues of debt-servicing that are fre-
analysis of default functions indicates a de-
finitive number of ratios. A suggested list quently confused.3
What is meant by the "burden" of the
might include: debt service/exports, debt ser- debt? This is usually related to the debt-ser-
vice/debt outstanding, debt outstanding/ex- vicing problem. There would, of course, be no
ports, net transfer/ imports, rate of growth of problem of debt service if capital flowed into
debt/rate of growth of exports, and imports/ the country in sufficient amount to allow the
reserves. Increases in these ratios might be developing country to meet interest and am-
warnings of a debt crisis. But in past rene- ortization payments on foreign obligations
gotiation cases there has been large variance and also to maintain its imports at a desired
and dispersion in these ratios. Other factors level. In reality, however, sooner or later —
must therefore be considered.
depending on the growth in new foreign bor-
Beyond ratios, country risk analysis should rowing, rate of interest, and amortization
monitor some "key performance indicators" rate — the debt-servicing charges may require
that will indicate how national economic a net capital outflow from the debtor country.
management is affecting the growth of the When the return flow of interest and amor-
economy and the capacity to service the debt. tization payments exceeds the inflow of new
The key performance indicators are (1) a ris- capital, the country becomes a "mature
ing ratio of savings to national income, (2) a debtor" and confronts a transfer problem in
rising ratio of taxes to income, (3) a decreas- servicing the debt. The country has to
ing incremental capital-output ratio (less un- achieve sufficient self-sustaining growth to
utilized capacity), (4) a decreasing current remove the resource gap and cover the net
account deficit, (5) high employment elastic- capital outflow. The country will have to gen-
ity— that is, the rate of growth in employ- erate an export surplus equal to its net out-
ment relative to the rate of growth in output, ward transfer of interest on current account
and (6) more equitable income distribution. and amortization on capital account in its
To realize positive changes in these perfor- balance of payments.
mance indicators the debtor country may The direct costs of debt service do not,
have to undertake a set of policies, reflecting however, constitute the burden of the debt —
a higher quality of policy analysis and imple- provided the social rate of return from the ex-
mentation. Price distortions will have to be ternal capital exceeds the interest cost. True,
removed: real interest rates raised, overval- part of the increased production from the use
ued foreign exchange rate corrected, wages
brought closer to opportunity cost. The fiscal 3This section is drawn from Gerald M. Meier, In-
deficit must be reduced. Local capital mar- ternational Economics, 1980, pp. 343-8.
 312 MOBILIZING FOREIGN RESOURCES

of the external resource inflow has to be re- ultimately a charge on the economy as a
paid abroad, and this is a reduction that whole, we can recognize that the transfer
would not be necessary if the savings had problem can still be solved without stipulat-
been provided at home. But the reality is that ing that the investment of foreign capital
the savings have been provided from over- should create its own means of payment by
seas, and the foreign savers must receive directly expanding exports or by replacing
some return. Of most importance to the bor- imports. Instead of committing the fallacy of
rowing country is that its economy has real- misplaced concreteness, we should realize
ized additional investment, and the benefits that debt capacity cannot be determined
from this should exceed the direct costs of the
without appraising the country's develop-
foreign savings that made possible the capital ment program as a whole. Analysis of the en-
formation. The direct costs of servicing the tire program is necessary for an assessment
foreign debt out of additional income should of the conditions under which the competing
not be a cause for concern. claims on total resources, on savings, and on
Of genuine concern are the indirect costs. foreign exchange can be adjusted so as to re-
These arise when the debt-servicing country lease the amount required for debt service. In
has to undertake burdensome policies of bal- the last analysis, it is not a matter of whether
ance-of-payments adjustment to acquire suf- "the project" that the foreign loan is osten-
ficient foreign exchange for debt service. Do- sibly intended to finance will be able to carry
mestic savings in the developing country will the cost of the loan — but the return on the
have to become sufficient to finance all do- marginal project or the ultimate effect on the
mestic investment and, in addition, the inter- total use of resources. If it is realized that the
est cost of accumulated debt and the repay- ability to create a sufficiently large export
ment of the principal of its loans. In order to surplus depends on the operation of all sec-
convert the surplus of savings into the foreign tors together, not simply on the use made of
exchange it needs for debt-servicing, the de- foreign capital alone, it is then apparent that
veloping country will have to generate an ex- a project financed by foreign borrowing need
port surplus through expenditure-reducing not itself make a direct contribution to the
and expenditure-switching policies so as to balance of payments. Indeed, even if foreign
expand exports or reduce the demand for im- capital were limited to financing projects that
ports. This may require some combination of earn or save foreign exchange, developments
deflation, internal and external controls over elsewhere in the economy may at the same
resource allocation, and exchange rate depre- time be affecting the supply of foreign ex-
ciation. The adverse effects of these measures change so adversely that the debt service
of balance-of-payments adjustment are the problem is aggravated even though the for-
indirect costs of foreign borrowing, and they eign capital has supposedly been directed to
constitute the burden of debt-servicing. projects that ought to be able to carry the
It might be thought that this burden can costs of the loans.
be avoided if the investment of foreign capi- Instead of such a narrow balance-of-pay-
tal creates its own means of payment by di- ments criterion for the allocation of invest-
rectly expanding exports or replacing im- ment, the basic test for the allocation of for-
ports. The lender might also believe that if it eign capital, as for any investment, is that it
lends for a project that earns foreign ex- should be invested in the form that yields the
change or saves foreign exchange there can highest social marginal product. The alloca-
be no transfer risk in debt-servicing. But this tion of capital according to its most produc-
is again to adopt the myopic and illusory view tive use will also be the most favorable for
of project financing. Once we appreciate the debt-servicing, because it maximizes the in-
relationship between total resource availabil- crease inincome from a given amount of cap-
ities and uses, the interdependence of invest- ital and thereby contributes to the growth of
ments, and the principle that debt service is foreign exchange availability. The export of
 3 13 EXTERNAL DEBT AND COUNTRY RISK ANALYSIS— NOTE

particular commodities or services through or defaults. Underlying the evaluation of bal-

which the interest is transferred abroad ance-of-payments potential is the assessment
should then be determined by the principle of of the key performance parameters. And be-
comparative costs. There are thus two prin- hind these, in turn, lie the policies that the
ciples determining debt capacity: the invest- borrowing country would have to undertake
ment criterion of social marginal productivity to raise these performance parameters and to
and the trade criterion of comparative cost. realize its balance-of-payments potential.
To allocate foreign capital to a foreign ex- In the columns of Table 1, countries are
change project is to do violence to the sepa- ranked according to their capacity for na-
ration of these two basic principles, and it is tional economic management. This involves a
to settle for the easier, but misleading, ap- judgment on the borrowing country's ability
proach ofproject appraisal instead of country and willingness (often a political matter) to
appraisal. undertake the policies that will allow the
The elements that compose the analysis of country to achieve its balance-of-payments
a country's creditworthiness can be synthe- potential. Countries in column A have the
sized in a country risk matrix (see Table 1). highest policy capability to undertake the
The vertical scale illustrates the lender's ap- measures that would allow debt-servicing.
praisal of the borrowing country's balance- All countries rated E or 5 would present the
of-payments potential — that is, the country's greatest country risk to the lender, whereas
potential to have a sufficient flow of foreign those rated A or 1 would be most creditwor-
exchange to service its debt. Countries in the thy with the highest debt capacity. Japan and
top row (numbered 1) have the highest po- Zaire are simply illustrative of the relative
tential, whereas countries at the lowest rating ratings that the lender could make within the
(numbered 5) have the least potential and matrix. Ratings would of course change over
can be expected to confront payment delays time.
Comment
Several statistical and econometric studies have attempted to refine the methodology of
country risk analysis. Using the statistical techniques of discriminant and logit methods, the
studies attempt to develop a model that will predict debt reschedulings based on historical
TABLE 1. Country Risk Matrix
— Capacity for national economic management
High
Low
A B C D E

High
1
a Japan

93
2
a.

Low
5 "Zaire"
 314 MOBILIZING FOREIGN RESOURCES

patterns. With varying degrees of predictive value, the studies attempt to establish econo-
metric default functions.
See R. Z. Aliber "A Conceptual Approach to the Analysis of External Debt of the Devel-
oping Countries, World Bank Staff Working Paper No. 421 (October 1980); R. Z. Aliber,
"Describing External Debt Situations: A Roll-Over Approach." IMF Staff Papers, Vol.
XXII, No. 1 (1975); G. Feder and R. Just, "A Study of Debt Servicing Capacity Applying
Logit Analysis," Journal of Development Economics, Vol. 4, No. 1 (1977); G. Feder, "Eco-
nomic Growth, Foreign Loans and Debt Servicing Capacity of Developing Countries," Jour-
nal ofDevelopment Studies (April 1980); Charles R. Frank, Jr., and William R. Cline, "Mea-
surement ofDebt Servicing Capacity: An Application of Discrimination Analysis," Journal
of International Economics, Vol. I, No. 3 (1971); Alice L. Mayo and Anthony G. Barrett,
"An Early Warning Model for Assessing Developing Country Risk," in S. H. Goodman (ed.),
Proceedings of a Symposium on Developing Countries' Debt (August 1977); K. Saini and P.
Bates, "Statistical Techniques for Determining Debt-Servicing Capacity for Developing
Countries," Federal Reserve Bank of New York, Research Paper No. 7818 (September 1978);
Nicholas Sargen, "Economic Indicators and Country Risk Appraisal," Federal Reserve Bank
of San Francisco Economic Review (Fall 1977).
The models all have Type I errors (failures to predict actual reschedulings) and Type II
errors (prediction of reschedulings that did not in fact occur). Moreover, the default functions
are derived from a historical sample, and their applicability to a different sample in the future
is questionable.
 V.C. THE INTERNATIONAL
MONETARY SYSTEM AND
DEVELOPMENT FINANCE-
NOTE

What emerges from the preceding selections by linking the creation of additional interna-
is that foreign aid has been declining just at tional reserves with assistance to developing
a time when there is a greater capacity to use countries. The general position of LDCs, as
aid more effectively and when the need in represented by the Lima Action Program and
many countries has become even more ur- statements of UNCTAD,1 has been in favor
gent. It is understandable that the LDCs of very limited flexibility or a return to a
have therefore sought new measures for the fixed exchange-rate system and the creation
extension of foreign aid. If aid is not forth- of additional international liquidity in the
coming in sufficient amounts in the tradi- form of a fiduciary instrument like Special
tional form of open aid explicitly granted by Drawing Rights (SDRs).
donor to recipient, then the transfer of re- In the debate over flexible exchange-rates,
sources from rich to poor countries must some special arguments relate to LDCs with
come via other policy instruments. Most respect to whether (i) rates are fixed or flex-
prominent in UNCTAD discussions are the ible between developed countries (DCs) and
attempts to have aid increased in a "dis- LDCs, and (ii) rates are fixed or flexible
guised" form through trade policy (i.e., among the DCs.
through international commodity agreements The LDCs may favor fixed rates among
and preferences on manufactures) and the DCs, but some management of a "crawl-
through reform of the international monetary ing peg" between
system. In Chapter VIII we examine the currencies. Severalthe LDCs
LDC currency
have usedandsome
DC
scope for aid through trade policy. Here we form of "crawling peg" to offset their differ-
consider the merits of introducing a stronger ential rates of inflation and to leave their real
component of development finance into the effective exchange-rate fairly constant. These
international monetary system. include Brazil (a "trotting peg"), Colombia,
New international monetary arrangements the Philippines, Korea, Malaysia, Morocco,
have been evolving since August 15, 1971 Yugoslavia, and Indonesia.
when the United States suspended converti- In favor of greater flexibility, it is claimed
bility ofthe dollar into gold, and a regime of that flexible rates depoliticize the problem of
floating exchange-rates for most major cur- devaluation. LDCs tend to maintain overval-
rencies emerged. There has been official in- ued disequilibrium rates, protected by trade
tervention inexchange markets, however, so and exchange restrictions, and are reluctant
that the floats have been "managed" within to devalue for political reasons; but this ten-
limits. Although rates were not fully flexible, dency to postpone adjustment would be over-
the old par system of the IMF was disman- come ifdepreciation of the deficit country's
tled, and a debate emerged over establish- currency came about automatically under
ment of a new exchange-rate system to im- freely fluctuating rates. Although devalua-
prove the balance-of-payments adjustment tions have not met with political favor, a
process. This debate was also closely related study of some two dozen currency devalua-
to the question of how payments imbalance
might be settled more readily with greater in-
ternational liquidity. And this, in turn, raised 'G. M. Meier, Problems of A World Monetary-
the issue of supplementing development aid Order, 2nd ed., 1982, pp. 301-14.

315
 316 MOBILIZING FOREIGN RESOURCES

tions in LDCs shows that the action has gen- LDC would be ineffective as an adjustment
erally been successful in improving the bal- mechanism to the extent that domestic infla-
ance of trade.2 tion persists, the LDCs demand for imports
Most of the LDCs, however, favor fixed is a function of its output and is unresponsive
rates between their currencies and those of to price changes, the LDCs marginal pro-
other countries with which they have most of pensity to save is a constant that cannot be
their trade and financial relations. Being gen- readily altered by relative price changes, and
erally a small open economy, the LDC be- the LDCs exports have a low price elasticity
lieves itis advantageous to maintain convert- of demand (i.e., conditions of a "structural *
ibility at a fixed price into the currency of disequilibrium"). Thus, on grounds of both ■>
some major country with which the small inefficacy and undue burden, the LDCs tend
LDC trades extensively or on which it de- to reject fluctuating exchange-rates.
pends for capital for investment. The strength of these arguments can only i
Perhaps even more persuasive than the be assessed in comparison with what are the (
positive case for pegging to a major currency, feasible alternatives to flexible rates. If rates p
is the negative case against flexible rates for are to change only infrequently, but the i
LDCs. It is argued that flexible rates have changes are very large, then the uncertainty
several special disadvantages for LDCs. The and risk aversion may be even greater than
problems of overcoming uncertainty and risk under more frequent, but smaller, rate 2
aversion may be especially severe for an LDC changes. If the alternative to adjustment
which does not have well-established forward through rate flexibility is the imposition of
exchange markets and which would have to trade restrictions, or is a policy of national in-
diversify its reserve and debt portfolios to re- come contraction, then the alternative costs
duce exchange-rate risks. The facilities for a may be greater than any burden imposed by
forward market between foreign currency rate flexibility. It is therefore not sufficient
and domestic currency are usually extremely simply to point out that there are costs in-
thin in the LDC, and it is too facile to believe volved inrate flexibility: in the absence of ac-
that forward cover can be provided the cess to unlimited international liquidity,
LDCs exporters by utilizing forward mar- there will be some costs to any remedial pol-
kets in the developed countries. Flexible rates icy for correcting a balance-of-payments def-
intensify the complexity of portfolio manage- icit. The objective is to discover the least-
ment of reserves for LDCs, and reserves held costly remedial policy.
in a depreciating currency would lose pur- Even if it is granted that the LDC wants
chasing power. The creation of uncertainty to peg its currency to a DC currency, the
for the foreign investor who wants to recon- question still remains whether fixed or flexi-
vert funds into the creditor's currency can ble rates among the DCs themselves would
also become a significant deterrent to capital be of most advantage to the LDCs. Again,
inflow. The real value of the repayment bur- this depends on the feasible alternatives for
den of servicing external debt could also be the DCs. The dominant concern of the LDC
severe if the debt is denominated in an ap- should be expansion of its exports to the DCs.
preciating currency. Trade with the DCs is by far the largest
Further, freely fluctuating rates tend to in- source of foreign exchange for the LDCs. It
hibit regional integration among LDCs un- is therefore of utmost importance that the
less the countries can agree to a joint float. DCs refrain from imposing trade barriers on
Finally, exchange-rate depreciation for an imports from the LDCs, and that the DCs
promote a high rate of domestic expansion so
2Richard N. Cooper, Currency Devaluation in De-
that imports from LDCs will be maintained.
veloping Countries, Princeton Essays in International If flexible rates among the DCs will provide
Finance, No. 86, June 1971. an effective adjustment mechanism, so that
 317 INTERNATIONAL MONETARY SYSTEM AND DEVELOPMENT FINANCE— NOTE

trade controls or deflation can be avoided, corresponding deficits are much less readily
then flexibility among DC-rates will be best tolerated, there will have to be an autono-
mous inflow of new reserve assets into the
from the viewpoint of the LDCs.3
With flexibility among DC-currencies, the system.
LDC can then peg either to a key currency In the course of discussions on interna-
(dollar, franc, mark, yen), to a bundle of cur- tional monetary reform it increasingly has
rencies, or to the SDR. Which is most desir- been argued that the SDR should become the
able for the LDC will depend on the diversi- principal asset of settlement between central
fication ofits trade and financial transactions banks, and, in time, the principal reserve
with "the rest of the world." Technically, to asset of the system. It is a common judgment
reduce its loss of control over its effective ex- that if countries must suffer deflation or im-
change-rate, the small LDC should peg to a pose restrictive measures on trade and capital
weighted average of key currencies. If the ob- movements in order to protect their interna-
jective isto keep its domestic prices in line tional reserves, then there is a shortage of in-
with the "world" price level, the weights will ternational liquidity. Even though one cannot
have to correspond to those of each major
specify the "right" amount of international
country contributing to such a price level. If liquidity,4 one can certainly maintain that
the goal is to maintain balance-of-payments countries are revealing a shortage of inter-
equilibrium by adjusting the effective ex- national liquidity when the rate of interna-
change-rate, more complicated calculations tional economic growth is retarded and there
of multiple-currency pegging will be neces- is a retreat from liberalization of trade in
sary, involving price elasticities by countries order to husband reserves. But how much li-
and forecasts of future trade patterns. Be- quidity should be created? Who is to control
cause of the complicated calculations re- the creation of the additional liquidity? How
quired to maintain some target trade- is the increase to be distributed? And what
weighted exchange-rate (and how should fi- are to be the benefits to the LDCs?
nancial flows be weighted, as compared with Even if the LDCs were not to share in the
trade flows?), the LDCs are more likely to control of the creation of additional liquidity,
opt for pegging to the SDR. The SDR is de- or were not to receive directly any of the ini-
fined in terms of a basket of five major cur- tial increase in liquidity, they would still ben-
rencies, and it is undoubtedly operationally efit indirectly from any plan that expanded
world reserves. For to the extent that such an
simplest to peg to a "standard basket" SDR.
To the extent that the mechanism of ad- expansion forestalls the developed countries
justment ismore effective through flexible from adopting beggar-my-neighbor policies,
rates and disabsorption policies, there is less or allows a relaxation of such restrictive mea-
need for the creation of additional reserve as- sures, the developing countries will benefit. If
sets. But to the extent that countries are un- as a result of access to new reserves, a devel-
willing to bear the costs of adjustment, or oped country can have a higher rate of
countries welcome continual surpluses while domestic growth without encountering a
balance-of-payments constraint, then the
demand for imports from LDCs can rise. If a
3For a comprehensive discussion, see William R.
Cline, Interests of the Developing Countries in Inter- 4It is illusory and misplaced precision to specify an
national Monetary Reform, Brookings Staff Paper, "optimum amount" of international liquidity by re-
1975. Cline concludes that the empirical evidence of lating the "need" to such variables as imports, varia-
currency changes in the early 1970s indicates that the tions in foreign trade, imports and capital flows, past
deficits, domestic money supply, or current liabilities.
LDCs' fears of flexible rates have generally been un-
warranted, and that the LDCs' gains from an im- See Fritz Machlup, "The Need for Monetary Re-
proved monetary system should greatly exceed any serves," Banca Nazionale del Lavoro, September
external costs to them from flexibility in DC-rates. 1966, pp. 175-222.
 318 MOBILIZING FOREIGN RESOURCES

developed country is enabled to remove re- that the need for reserves is to cover a tem-
strictions on the outflow of private foreign in- porary imbalance, while the need for long-
vestment, then again the LDCs may benefit. term development finance is to cover a
Or if monetary policy — previously kept chronic balance-of-payments gap. If the
"tight" out of consideration for the balance amount of liquidity to be created is deter-
of payments — is eased, the LDCs and inter- mined bydevelopment finance, then, it is ar-
national institutions specializing in develop- gued, there will be too much liquidity. More-
ment assistance may have greater access to over, it is contended that the developing
the capital markets of developed countries. countries would not conserve an addition to
And most importantly, if developed nations their reserves but would instead spend it, con-
no longer have to plead that pressures on trary to the notion that the creation of liquid-
their balance of payments prevent an in- ity should meet the demand for liquid assets c
crease in foreign aid or the untying of aid, to hold as precautionary balances, not to
then the recipient LDCs will benefit. spend as transaction balances.
Although these indirect benefits are of Against this scepticism, proponents of a di-
value, the LDCs have argued for more than rect link between monetary reform and the
this indirect assistance: they have wanted de- provision of development finance offer several
velopment finance to be linked with liquidity counter-arguments. In brief:
creation.5 A scheme of international mone- i. The need for more reserves by LDCs is
tary reform would then be of direct benefit to especially acute. It is contended that there is
the LDCs in that it would directly provide a widening trade gap, primary product ex-
additional aid in the process of creating ad- port earnings fluctuate widely, and the mech-
ditional international liquidity. anism of adjustment to a balance-of-pay-
Special Drawing Rights are a form of un- ments deficit operates more slowly in an LDC
conditional credit: they involve no fixed re- than in an advanced economy.7 There is
payment schedule and require no statement therefore a justification in initially distribut-
of need by users. Developing countries would ing an increase in international liquidity to
well prefer this to the conditionality imposed LDCs.
by the IMF on some types of drawings. Fur- ii. As for the contention that an increase in
thermore, since net users have to pay only in- reserves would allow the developing countries
terest and do not have to repay capital, pay- to escape from balance-of-payments disci-
ments on the net use of SDRs will tend to be pline, itis replied that for an LDC the so-
lower than they would be for a commercial called "discipline of the balance of pay-
bank loan of similar size. The lowest-income ments" does not in reality lead to deflationary
countries that have not had access to the Eu- (or disinflationary) monetary and fiscal poli-
rocurrency market would tend to benefit cies but instead merely to import restrictions.
An increase in reserves would therefore not
most from the "link."6
The standard argument against linking li- actually have the undesirable effect of remov-
quidity creation and development finance is, ing balance-of-payments discipline but
of course, that the need for reserves and the rather the favorable effect of supporting the
need for development finance are two distinct liberalization of trade.
issues and that therefore the LDCs are mix- iii. The argument that developing coun-
ing up two separate questions. It is argued tries are incapable of conserving an increase
in additional reserves is also denied. Contrary
5See Y. S. Park, The Link Between Special Draw- to the usual impression, the statistical evi-
ing Rights and Development Finance, Princeton Es-
says in International Finance, No. 100, September dence does not support the view that the typ-
1973.
7But cf. A. Kafka, "International Liquidity: Its
6Graham Bird, "SDR Distribution, Interest Rates Present Relevance to the Less Developed Countries,"
and Aid Flows." The World Economy, December American Economic Review, Papers and Proceed-
1981, pp. 419-27. ings, May 1968, pp. 596-607.
 319 INTERNATIONAL MONETARY SYSTEM AND DEVELOPMENT FINANCE— NOTE

ical or average LDC behavior is to increase other reserve assets. SDRs are transferred by
overseas expenditure by an amount corre- debiting the SDR account of the user and
sponding to any increase in reserves.8 crediting the SDR account of the receiver,
iv. Finally, even if the developing country with the user country acquiring convertible
did choose to spend an increase in its reserves currency from the receiving member.
on additional imports, some would argue that In the initial creation of SDRs, there was
this is quite legitimate. For if the additional no direct connection with foreign aid, but the
liquidity has not been acquired subject to LDC members of the IMF share along with
specific conditions on its use, then — as the other Fund members through additions to
owner of unconditional liquidity — the devel- their reserves in proportion to their quotas,
oping country should be able either to spend and they benefit indirectly to the extent that
it or hold it.9 From this standpoint, there is the SDRs allow an increase in aid expendi-
no reason why more severe provisions should ture and trade liberalization on the part of
be imposed on an LDC than on any other developed countries.
country that receives additional liquidity. The LDCs still advocate, however, either
When the Special Drawing Account was an organic link through an increase in SDR
established in the IMF in 1969, the emphasis allocations to the LDCs, or allocation of
was only on liquidity-creation, and a link to SDRs to development agencies that could use
development finance was not incorporated in these funds for long-term loans to the devel-
the facility for Special Drawing Rights. The oping countries, or a contribution by devel-
establishment of this facility provided, oped countries of a certain proportion of their
through deliberate international decisions SDR allocation to multilateral development
(an 85 percent majority vote of IMF mem- agencies, or the transfer to development
bers was required to create SDRs), a major agencies of reserve currencies received by the
new supplement to existing reserve assets, on IMF in exchange for a special issue of
a permanent basis, backed by the obligation SDRs.10 If the problem of liquidity-creation
of Fund members to accept the SDRs and at first dominated the establishment of the
pay a convertible currency in return. The first SDR facility, the ensuing years of discussion
allocation of SDRs amounted to $9.5 billion about international monetary reform have
over the period 1970-72. SDRs were allo- raised more critical questions regarding the
cated to each participating member country distribution of SDRs.
according to its quota in the IMF; such an Although the SDR-aid link issue poses im-
allocation did not favor the LDCs and was portant questions for both the technical op-
criticized by those who advocate a direct or eration ofthe international monetary system
"organic" link between liquidity-creation and moral judgments of international distri-
and liquidity-distribution in favor of the butional equity, the quantitative impact of
LDCs. Since the SDR was redefined in terms any flexible SDR-aid link scheme will only be
of a "standard basket" of major currencies, it marginal relative to the foreign exchange
has been increasingly used as a numeraire in that LDCs derive from trade and capital
international transactions. Whether the SDR movements.
is to become the principal reserve asset will The LDCs must support policies that pro-
depend on the future volume of new SDRs vide more aid and more trade and more cap-
that are created and their substitution for ital inflow — not policies that entail any
trade-offs among the sources of foreign
8Report of the Group of Experts, International exchange.
Monetary Issues and the Developing Countries,
UNCTAD, 1965 (66.II.D.2) paragraphs 35-6; Paul
While the SDR-aid link may have only
P. Streeten, "International Monetary Reform and the
LDCs,"
8-9. Banca Nazionale del Lavoro, June 1967, pp. 10Details of proposals for an organic link are pre-
sented in Park, Special Drawing Rights and Devel-
9Streeten, "International Monetary Reform," p. 9. opment Finance, pp. 9-1 1
 320 MOBILIZING FOREIGN RESOURCES

marginal significance for the receipt of for- The IMF should avoid inappropriate or excessive
eign exchange by the LDCs, as compared regulation of [the developing countries'] econo-
with expansion by the developed countries mies, and should not impose highly deflationary
and greater market access in the developed measures as standard adjustment policy. It should
also improve and greatly extend the scope of its
countries, nonetheless the idea of linking re- compensatory financing facility. Surplus countries
serve creation to development assistance has should accept greater responsibility for payments
symbolic value in terms of distributional adjustments, and IMF measures to encourage this
equity. If international poverty is a manifest-
should be considered.12
ation of international inequality, the chan-
neling to LDCs of a part of the share of Furthermore, the report suggests that the
SDRs formerly going to industrial countries bulk of the Fund's holdings of gold could be
would be at least a welcome symbol of a de- used as collateral for IMF borrowings from
liberate effort to transfer resources from rich financial markets that would be lent primar-
to poor countries. ily to middle-income developing countries.
Another part could be sold and the profits
It may also be argued that if "objective in- used to subsidize interest rates on loans to
dicators" are introduced imposing interna-
low-income developing countries.
tional rules that limit a country's freedom re- Although it may be unrealistic to expect
garding reserve composition, these rules
should be more lenient for LDCs. A wider that the international financial community
range for "objective indicators" might be would support the linking of development fi-
permitted for LDCs than DCs, in view of the nance to liquidity creation, or that the IMF
greater fluctuation in export earnings of would depart from its fundamental principle
LDCs than DCs, and in deference to the need of treating all member countries uniformly,
of the LDCs to acquire liquidity at less cost there is still much that the IMF can do to
than on private capital markets. play a more significant role in promoting de-
The report by the Brandt Commission ar- velopment. To allow larger standby agree-
gued that reform of the international mone- ments with developing countries, and to offset
tary system should involve improvements in the adverse repercussions of any interna-
the exchange-rate regime, the reserve sys- tional trade recession, the IMF's resources of
tem, the balance-of-payments adjustment quotas, SDRs, and compensation schemes
process, and the overall management of the should be increased. The IMF's imposition of
system that should permit greater participa- conditionality for use of the Fund's resources
tion of developing countries in the decision and the stabilization programs associated
making of the IMF. with standby agreements are also of major
The report argues that any serious reform importance to the LDCs. Actions by the IMF
must aim at greater stability of international may also influence commercial banks to con-
currencies, particularly among key curren- tinue their lending operations to LDCs.
cies. This can only be achieved if the reserve As Sir Arthur Lewis concluded about the
system and adjustment mechanism are re- importance to developing countries of an in-
formed simultaneously. The SDR should be- ternational monetary institution with wide
come the principal means of increasing responsibilities and powers:
global liquidity, and "the distribution of such [The] developed countries, with stronger econo-
unconditional liquidity should favor the de- mies, less dependence on international flows, and
veloping countries who presently bear high easy fraternal relations between central bankers,
adjustment burdens."11 There should also be could get along fairly well without an international
pressure in the world economy: ,2Ibid. institution; and having one, could man-
monetary
age quite well if its function was confined to dis-
11 Willy Brandt, North-South: A Program for Sur- cussing changes in exchange rates. To the less de-
vival, Report of the Independent Commission on in-
ternational Development Issues, 1980, chapter 13.
 321 INTERNATIONAL MONETARY SYSTEM AND DEVELOPMENT FINANCE— NOTE

veloped countries, on the other hand, an institution be to narrow the discussion, while the instinct of
which did not look beyond exchange rates to the
wider context which determines international the developing will be to widen it. . . . 13
monetary flows could be a menace rather than a
help. Given the difference in their immediate in- l3Quoted in IMF Survey, November 7, 1977,
terests, the instinct of the developed countries will
p. 349.
 V.D. PRIVATE FOREIGN
INVESTMENT
V.D. 1 . Benefits and Costs of
Private Foreign Investment-
Note

Considerable interest is now being shown in that are newly independent should not be
measures that might promote private foreign analyzed as if the undesirable features in the
investment and allow it to make a greater history of colonialism need be repeated. The
contribution to the development of the recip- central problem now is for the recipient coun-
ient countries. In an attempt to promote a try to devise policies that will succeed in both
larger flow of private capital to developing encouraging a greater inflow of private for-
nations, several capital-exporting countries eign capital and ensuring that it makes the
have adopted a range of measures that in- maximum contribution feasible toward the
clude tax incentives, investment guarantees,
achievement of the country's development
and financial assistance to private investors. objectives. The tasks of development require
International institutions are also encourag- both more effective governmental activity
ing the international flow of private capital; and more investment on the part of interna-
the International Finance Corporation, for tional private enterprise. But the private
instance, cooperates directly with private investors must be aware of the developmental
investors in financing new or expanded objectives and the priorities of the host coun-
ventures. try and understand how their investments fit
Of far greater influence, however, than the
into the country's development strategy. The
measures adopted by capital-exporting na- contribution of private foreign capital has to
tions or international organizations are the be interpreted in terms beyond private profit.
policies of the capital-recipient countries At the same time the government must rec-
themselves. Controls exercised by the host ognize that if risks are too high or the return
country over the conditions of entry of for- on investment is too low, international private
eign capital, regulations of the operation of investment will be inhibited from making any
foreign capital, and restrictions on the remit- contribution at all. Development planning
tance of profits and the repatriation of capital now requires the government to influence the
are far more decisive in determining the flow performance of private foreign investment,
of foreign capital than any policy undertaken
but in doing this, the government should ap-
by the capital-exporting country. preciate fully the potential contribution of
Policies of the host country are now being this investment, and it should devise policies
adopted in the context of development plan- that will meet the mutual interests of private
ning, and their rationale and effects have investor and host country. This calls for more
taken on new dimensions in this light. No
intensive analysis of the consequences of pri-
longer is it simply a matter of private inves- vate foreign investment and for more thought
tors dismissing investment prospects with the and ingenuity in devising new approaches
complaint that a "favorable climate" does that favor the mobilization of private foreign
not exist; nor need host countries contend
capital while ensuring its most effective
that an inflow of private capital entails noth-
"planned performance" in terms of the coun-
ing but "foreign domination." The meaning try's development program.
of a "favorable climate" calls for reinterpre- At present, the policies taken by the devel-
tation in terms of development planning, and
the effects of foreign investment in countries oping countries reveal a mixed picture of re-
strictions and incentives. On the one hand,

322
 323 PRIVATE FOREIGN INVESTMENT

is greater than the amount appropriated by

the foreign investor's freedom of action may
be restricted by a variety of governmental the investor, social returns exceed private re-
regulations that exclude private foreign in- turns. As long as foreign investment raises
vestment from certain "key" sectors of the productivity, and this increase is not wholly
economy, impose limitations on the extent of appropriated by the investor, the greater
foreign participation in ownership or man- product must be shared with others, and
there must be some direct benefits to other
agement, specify conditions for the employ-
ment of domestic and foreign labor, limit the income groups. These benefits can accrue to
amount of profits, and impose exchange con- (a) domestic labor in the form of higher real
trols on the remission of profits and the re- wages, (b) consumers by way of lower prices,
patriation ofcapital. and (c) the government through higher tax
On the other hand, a progressive liberali- revenue. Beyond this, and most importantly
zation of policy towards private foreign cap- in many cases, there are likely to be (d) in-
ital has occurred during recent years. Many direct gains through the realization of exter-
countries now recognize that an inflow of pri- nal economies.
vate capital may offer some special advan- An increase in total real wages may be one
tages over public capital, and a number of in- of the major direct benefits from an inflow of
vestment incentive measures have been foreign capital. This can be recognized in
recently adopted or are under consideration. Figure 1 where the line EG illustrates the
These incentive devices include assistance in marginal productivity of capital in the capi-
securing information on investment opportu- tal-recipient country, given the amount of
nities, the provision of supplementary fi- labor. If initially, the domestically owned
nance, establishment of economic overhead capital stock is AB, the total output is ABCE.
facilities such as in industrial estates, protec- We assume that profits per unit of capital
tive tariffs on commodities that compete with equal the marginal product of capital, and
those produced by foreign investors, exemp- that total profits on domestic capital are
tions from import duties on necessary equip- ABCD, and total real wages are CDE. Let
ment and materials, the granting of exchange there now be an inflow of foreign capital in
guarantees or privileges, tax concession the amount BE. Total output then increases
schemes for the encouragement of desired by the amount BEGC, and the profits on for-
new investments, and special legislation for eign capital are BFGH of this amount. Since
the protection of foreign investments. the profit rate on total capital has fallen, prof-
To remove the ambivalence that charac- its on domestic capital are reduced to ABHI.
terizes these policies, it is desirable to reex- But the total real wages of labor are now
amine the role of private foreign capital more GIE, with the increase in real wages amount-
systematically by appraising the prospective ing to DCGI. Although in this case, with a
benefits and costs of private foreign invest- given marginal productivity of capital sched-
ment. Such an appraisal may then provide a ule, most of labor's gain — the amount
more rational basis for determining what DCHI — is merely a redistribution from do-
type of policy is most appropriate for secur- mestic capitalists, there is still a net increase
ing the maximum contribution from private in the real incomes of domestic factors, rep-
foreign investment. resented by the rise in real wages in the
From the standpoint of national economic amount CGH.
benefit, the essence of the case for encour-
aging an inflow of capital is that the increase 'Much of the following analysis is based on Sir
in real income resulting from the act of in- Donald MacDougall, 'The Benefits and Costs of Pri-
vate Investment from Abroad: A Theoretical Ap-
vestment isgreater than the resultant in- proach," Economic Record, March 1960, pp. 13-35;
crease in the income of the investor.1 If the Paul Streeten, Economic Integration, second edition,
value added to output by the foreign capital Leyden, 1964, chapter 5.
 324 MOBILIZING FOREIGN RESOURCES

ment is cost-reducing in a particular indus-

try, consumers of the product may gain
through lower product prices. If the invest-
ment is product-improving or product-inno-
vating, consumers benefit from better quality
products or new products.
In order that labor and consumers might
enjoy part of their benefit from the higher
productivity in enterprises established by for-
eign investors, the overseas withdrawal by
the investors must be less than the increase in
output. But even if the entire increase in pro-
ductivity accrues as foreign profits, this re-
quirement may still be fulfilled when the gov-
ernment taxes foreign profits. For many
B F
countries, taxes on foreign profits or royalties
Capital Stock from concession agreements constitute a
FIGURE 1. large proportion of total government revenue.
The fiscal benefit derived from foreign in-
vestment isevident from the fact that the
share of government revenue in the national
product of countries that have received sub-
For a developing country, the inflow of for- stantial foreign investment is considerably
eign capital may be significant in not only higher than in most of the other low-income
raising the productivity of a given amount of countries.
labor but may also allow a larger labor force The most significant contribution of for-
to be employed. If, as was contended in the eign investment is likely to come from exter-
discussion of dualistic development in Chap- nal economies. Direct foreign investment
ter III, a shortage of capital in heavily pop- brings to the recipient country not only cap-
ulated poor countries limits the employment ital and foreign exchange but also manage-
of labor from the rural sector in the advanced rial ability, technical personnel, technological
sector where wages are higher, an inflow of knowledge, administrative organization, and
foreign capital may then make it possible to innovations in products and production tech-
employ more labor in the advanced sector. niques— all of which are in short supply. This
The international flow of capital can thus be ensures in the first instance that a project in-
interpreted as an alternative to labor migra- volving private foreign investment will be ad-
tion from the poor country: when outlets for equately formulated and implemented, un-
the emigration of labor are restricted, the like the situation that has frequently
substitution of domestic migration of labor confronted public economic aid when the re-
into the advanced sector becomes the most
cipient country has not had the talent or in-
feasible solution. The social benefit from the clination to undertake adequate feasibility
foreign investment in the advanced sector is studies and formulate projects that might
then greater than the profits on this invest- qualify for public capital. The preinvestment
ment, for the wages received by the newly survey, act of investment, and operation of
employed exceed their former real wage in the investment project are all ensured in pri-
the rural sector, and this excess should be vate foreign investment. One of the greatest
added as a national gain. benefits to the recipient country is the access
Domestic consumers may also benefit from to foreign knowledge that private foreign in-
direct foreign investment. When the invest- vestment may provide — knowledge that
 325 PRIVATE FOREIGN INVESTMENT

helps overcome the managerial gap and tech- tion in other industries. The foreign invest-
nological gap. The provision of this knowl- ment in the first industry can give rise to prof-
edge can be interpreted as "private technical its in industries that supply inputs to the first
assistance." The private technical assistance industry, in industries that produce comple-
and the demonstration effects that are inte- mentary products, and in industries that pro-
gral features of private foreign investment duce goods bought by the factor-owners who
may spread and have beneficial results in now have higher real incomes. Similar effects
other sectors of the economy. The rate of may also follow from investment that is prod-
technological advance in a poor country is uct-improving or product-innovating. A
highly dependent on the rate of capital in- whole series of domestic investments may
flow. New techniques accompany the inflow thus be linked to the foreign investment.
of private capital, and by the example they Against these benefits must be set the costs
set, foreign firms promote the diffusion of of foreign investment to the host country.
technological advance in the economy. In ad- These costs may arise from special conces-
dition, foreign investment may lead to the sions offered by the host country, adverse ef-
training of labor in new skills, and the knowl- fects on domestic saving, deterioration in the
edge gained by these workers can be trans- terms of trade, and problems of balance-of-
mitted to other members of the labor force, payments adjustment.
or these workers might be employed later by To encourage foreign enterprise, the gov-
local firms. ernment ofthe host country may have to pro-
Private foreign investment may also serve vide special facilities, undertake additional
as a stimulus to additional domestic invest- public services, extend financial assistance, or
ment in the recipient country. This is espe- subsidize inputs. These have a cost in absorb-
cially likely through the creation of external ing governmental resources that could be
pecuniary economies. If the foreign capital is used elsewhere. Tax concessions may also
used to develop the country's infrastructure, have to be offered, and these may have to be
it may directly facilitate more investment. extended to domestic investment since the
Even if the foreign investment is in one in- government may not be able to discriminate,
dustry, itmay still encourage domestic in- for administrative and political reasons, in
vestment byreducing costs or creating de- favor of only the foreign investor. Moreover,
mand in other industries. Profits may then when several countries compete among them-
rise and lead to expansion in these other selves inoffering inducements to foreign cap-
industries. ital, each may offer more by way of induce-
Since there are so many specific scarcities ment than is necessary: the investment may
in a poor country, it is common for invest- be of a type that would go to one country or
ment to be of a cost-reducing character by another, regardless of inducements, but the
breaking bottlenecks in production. This foreign extra
enterprise may "shop
stimulates expansion by raising profits on all secure concessions. Withoutaround" and
some form
underutilized productive capacity and by of collective agreement among capital-receiv-
now allowing the exploitation of economies of ing countries regarding the maximum
scale that had previously been restricted. concessions that will be made, the cost of
When the foreign investment in an industry "over-encouraging" certain types of foreign
makes its product cheaper, another industry investment may be considerable.
that uses this product benefits from the lower Once foreign investment has been at-
prices. This creates profits and stimulates an tracted, itshould be expected to have an in-
expansion in the second industry. come effect that will lead to a higher level of
There is also considerable scope for the ini- domestic savings. This effect may be offset,
tial foreign investment to produce external however, by a redistribution of income away
investment incentives through demand crea- from capital if the foreign investment reduces
 326 MOBILIZING FOREIGN RESOURCES

profits in domestic industries. The conse- Of greater seriousness than the foregoing
quent reduction in home savings would then costs are those associated with balance-of-
be another indirect cost of foreign invest- payments adjustments. Pressure on the bal-
ment. But it is unlikely to be of much conse- ance of payments may become acute when
quence in practice, for it would require that the foreign debt has to be serviced. If the
foreign investment be highly competitive amount of foreign exchange required to ser-
with home investment. In a poor country, it vice the debt becomes larger than the amount
is more probable that foreign capital will of foreign exchange being supplied by new
complement domestic investment and will foreign investments, the transfer mechanism
give rise to higher incomes and profits in will then have to create a surplus on current
other industries, as already noted. account equal to the debit items on account
Foreign investment might also affect the of the payment of interest, dividends, profits,
recipient country's commodity terms of trade and amortization on the foreign borrowings.2
through structural changes associated with When a net outflow of capital occurs, a real-
the pattern of development that results from location of resources becomes necessary in
the capital inflow. If the inflow of capital order to expand exports or replace imports.
leads to an increase in the country's rate of To accomplish this, the country may have to
development without any change in the terms endure internal and external controls or ex-
of trade, the country's growth of real income perience currency depreciation. The adverse
will then be the same as its growth of output. effects of these measures of balance-of-pay-
If, however, the terms of trade deteriorate, ments adjustment must then be considered as
the rise in real income will be less than that indirect costs of foreign investment, to be
in output, and the worsening terms of trade added to the direct costs of the foreign
may be considered another indirect cost of
the foreign investment. Whether the terms of The direct costs in themselves need not be
payments.
trade will turn against the capital-receiving a matter of great concern. For even though
country is problematical depending on var- part of the increased production from the use
ious possible changes at home and abroad in of foreign capital has to be paid abroad in
the supply and demand for exports, import- profits or interest — and this is a deduction
substitutes, and domestic commodities. It is that would not be necessary if the savings
unlikely, however, that private foreign invest- were provided at home — this is merely to say
ment would cause any substantial deteriora- that the country must not expect to get an in-
tion in the terms of trade. For if an unfavor- come from savings if it does not make the
able shift resulted from a rising demand for savings.3 What is fundamental is that the
imports on the side of consumption, it would country does have additional investment, and
probably be controlled through import re- the benefits from this may exceed the direct
striction. And if it resulted, on the side of costs of the foreign savings that made possi-
production, from a rising supply of exports ble the capital formation.
owing to private direct investment in the ex- The indirect costs, however, are rightly a
port sector, the inflow of foreign capital cause of concern, insofar as the capital-re-
would diminish as export prices fell, thereby ceiving country may be unable or unwilling
limiting the deterioration in the terms of to endure a loss of international reserves and
trade. Moreover, if the deterioration comes
through an export bias in production, it is still 2The length of time which elapses before this occurs
possible that the factoral and the income will depend on the growth in new foreign investment,
terms of trade might improve even though rate of interest and dividend earnings, and the amor-
the commodity terms of trade worsen, since tization rate. Cf. E. D. Domar, "The Effect of Foreign
the capital inflow may result in a sufficiently Investment on the Balance of Payments," American
Economic Review, December 1950, pp. 805-26.
large increase in productivity in the export 3Cf. J. R. Hicks, Essays in World Economics, Ox-
sector. ford, 1959, p. 191.
 327 PRIVATE FOREIGN INVESTMENT

does not want to impose measures of balance- vestment and provide wider opportunities for
of-payments adjustment in order to find suf- the foreign investor. If private foreign invest-
ficient foreign exchange for the remittance of ment is to be encouraged, it is necessary to
the external service payments. External mea- allay the investor's concern over the possibil-
sures such as import quotas, tariffs, and ex- ities of discriminatory legislation, exchange
change restrictions may suppress the demand controls, and the threat of expropriation. In-
for imports, but they do so at the expense of vestment guarantees may be utilized more ef-
productivity and efficiency. Internal mea- fectively tolessen the investor's apprehension
sures of higher taxation and credit tightness of nonbusiness risks. Either unilaterally or
involve the costs of reduced consumption and through bilateral treaties, governments can
investment. And the alternative of currency offer some assurances designed to reduce the
devaluation may cause the country to incur likelihood of expropriation or of impairments
the costs of a possible deterioration in its of investors' rights and to assure investors of
terms of trade, changes in income distribu- an adequate recourse if such impairments
tion, and necessary shifts of resources. To should occur. It has always been difficult to
avoid, or at least minimize, these indirect secure agreement by both investing and re-
costs, the role of private foreign investment cipient countries on a uniform set of substan-
must be related to the debt-servicing capacity tive rules, as in a multilateral investment
of the host country. And this depends on the charter. The World Bank has, however, suc-
country's development program as a whole, ceeded in establishing an International Cen-
since the ability to create a sufficiently large ter for Settlement of Investment Disputes,
export surplus rests on the operation of all in- which provides facilities for the settlement,
dustries together, not simply on the use made by voluntary recourse to conciliation or ar-
of foreign investment alone (see section V.B). bitration, ofinvestment disputes arising be-
In the past there has been a general ten- tween the foreign investor and the host
dency for poor countries to overestimate the government.
costs of foreign investment and to discount Finally, along with assurances against the
the benefits, especially the indirect benefits. occurrence of risk and measures for the ad-
Now, however, there is a wider appreciation justment ofinvestment disputes, considerable
that within the context of a development pro- attention is being given to the possibilities of
gram and with a careful appraisal of the pro- providing guarantees under which the inves-
spective benefits and costs of foreign invest- tors will be compensated for any loss they
ment, policies may be devised to secure a may suffer from other than normal business
greater contribution from the inflow of pri- causes. Although such a guarantee is some-
vate capital. Instead of discouraging invest- thing of a measure of last resort, it does help
ment from abroad simply because it involves to minimize the investor's risk and gives some
some costs, the developing countries are in- advance assurance of a reliable "safety mar-
creasingly recognizing that they should at- gin." Some capital-supplying nations have
tempt to devise policies that will encourage provided insurance coverage, but the insur-
the maximum feasible contribution. Al- ance of investments in developing countries
though the formulation of specific policies might be made more effective through the es-
must depend on particular conditions in each tablishment ofa multilateral investment in-
country, we may at least suggest some of the surance program as opposed to participation
principal considerations that might shape in a number of bilateral programs.
these policies. While investment guarantees may help in
In general, the attraction of private foreign
removing "disincentives" to foreign invest-
investment now depends less on fiscal action, ment, the attraction of private capital de-
upon which most countries have concen- pends even more on positive inducements in
trated, and more on other conditions and the form of greater opportunities for profit-
measures that guarantee protection of the in- making. The private investor's first concern is
 328 MOBILIZING FOREIGN RESOURCES

whether costs will be covered and a profit windfall. Moreover, the concessions are likely
earned. Some developing countries offer spe- to attract the quick speculative type of for-
cial tax concessions that provide a tax holi- eign investment which is in the country only
day or reduce the rate of tax on profits, but to take advantage of the concessions and
these measures are not effective unless the in- leaves as soon as these are withdrawn. Fi-
vestment yields a profit. The foreign investor nally, ifthe benefits of foreign investment are
is likely to be less interested in receiving an realized for the host country during the early
exemption after a profit is made than in being years of the investment, there is no case for
sure of a profit in the first instance. It is prolonging the concessions beyond the rela-
therefore most important to raise profit ex- tively short period.
pectations. Todo this, it may be necessary to Developing nations are now mainly inter-
undertake additional public expenditures, es- ested inhaving foreign enterprises contribute
pecially in developing the country's infra- to their industrialization, rather than follow-
structure and in ensuring a supply of trained ing the historical pattern of being directed to
labor. Yet rarely is a government willing to agriculture or mining. In most cases, how-
undertake expenditures expressly for the pur- ever, the size of the domestic market has re-
pose of attracting foreign investment; instead mained too small to offer much attraction. As
of incurring the present cost of additional ex- a development program proceeds, domestic
penditures, most governments prefer to assist markets may widen, and this limitation will
foreign investors through a future sacrifice in be reduced. Much can also be done to widen
revenue. Though politically more feasible, markets and establish a more substantial
tax concessions are not likely to be the most base for industry by promoting regional mar-
powerful inducements that the host country kets through arrangements for a common
can offer to encourage a flow of investment. market or a free trade association. The estab-
At the same time, their use can be over- lishment of a customs union or free trade
done. They involve a cost to the government area may have considerable potential for at-
not only in terms of a revenue foregone, but tracting investment to the development of
also in terms of equity and administrative manufacturing industry.
costs. More importantly, the LDCs may offer Even more significant than regional pref-
excessive concessions in their efforts to at- erential arrangements would be a general
tract foreign investment. If an investment is preferential system by which developed coun-
going to occur in one LDC or another (for tries granted preferences to all LDCs on their
example, in order to secure a raw material exports of manufactures and semimanufac-
supply), but each LDC competes against the tures. Such a system may offer considerable
other in offering concessions, then the LDCs attraction to private foreign investment, es-
are likely to overconcede.4 When tax conces- pecially byinducing the transfer of produc-
sions are granted to existing investments, in tion facilities required from the developed
order not to discriminate, there is simply a preference-granting countries to the less de-
veloped preference-receiving countries in
order to turn labor and raw materials into
saleable products and obtain the dual advan-
4Cf. Dudley Seers, "Big Companies and Small tages of tariff-free or preferential market ac-
Countries," Kyklos, 1963, pp. 601-3; R. H. Green cess and lower labor costs.
and Ann Seidman, Unity or Poverty?, Baltimore,
1968, pp. 99-131. In considering measures to encourage for-
Several studies indicate that tax incentive programs eign investment, a developing country does
have not been effective. See, for example, Jack Heller not want, of course, to seek foreign capital in-
and Kenneth M. Kauffman, Tax Incentives for In- discriminately. The objective is to ensure that
dustry in Less Developed Countries, Cambridge,
the investment supports activities from which
1963, pp. 60-6; M. C. Taylor, Industrial Tax Ex-
emption in Puerto Rico, Madison, 1957, pp. 143-9; the recipient nation may derive maximum
Peter Kilby, Industrialization in an Open Economy: national economic gain, as assessed through
Nigeria, 1945-66, Cambridge, 1969, pp. 132-4. benefits and costs. To achieve the most effec-
 329 PRIVATE FOREIGN INVESTMENT

tive utilization of foreign investment in terms has therefore been to focus on alternative ar-
of its entire development program, the coun- rangements for securing capital, manage-
try may have to adopt policies, such as pref- ment, technology, and marketing capabilities
erential tax treatment or other incentives, without the foreign ownership and control
that will attract private capital into activities that has commonly been associated with for-
where it will have the maximum catalytic ef- eign direct investment. It may be possible to
fect of mobilizing additional effort. From this "unwrap" the bundle of inputs that come in
standpoint, it is especially important that pol- the package of direct foreign investment and
icies affecting the allocation of foreign capi- secure inputs that are more appropriate for
tal be based on an awareness of the external the needs of the recipient country, or that
economies that can be realized from different cost less when they are not tied to equity
patterns of investment. Beyond a consider- capital.
ation of the direct increase in income result- The problem for the host country is to eval-
ing from the investment and other short-term uate the benefits and costs of alternative ar-
criteria, it is important to look to the more
rangements forimporting the investing firm's
indirect and longer-run possibilities — from capabilities, and then to secure by induce-
the widening of investment opportunities to ments and regulations the best feasible alter-
even the instigation of social and cultural native. The alternative arrangements span an
transformations. entire spectrum. At one end of the spectrum
Finally, the recipient country may be well is the traditional form of direct investment
advised to emphasize a partnership arrange- involving 100 percent foreign equity owner-
ment betwen foreign and domestic enter- ship and no time-limit on the existence of the
prise. Ajoint international business venture foreign-owned enterprise in the recipient
that involves collaboration between private country. This arrangement is appropriate for
foreign capital and local private or public only those sectors where the joint mix of cap-
capital is a promising device for protecting ital, technology, and management cannot be
international investment, integrating foreign supplied in any other way and the investment
investment within a development program Continues to provide a properly discounted
and safe-guarding against an enclave type of benefit-cost ratio greater than that of unity
investment, stimulating domestic manage- for an unlimited time period. Moving on from
ment and investment, and reducing the this arrangement, the host country may ini-
transfer burden and balance-of-payments tially allow 100 percent ownership but then
difficulties. insist that any expansion of the enterprise
The alternatives of a 1 00 percent foreign- occur through national participation. It may
owned enterprise and a joint venture are, further limit the time of foreign control by
however, only two of a number of possible ar- requiring a national majority equity holding
rangements forsecuring a mix of foreign cap- to emerge within a certain period. There may
ital, management, and technology. The be a "fadeout" provision such as Decision 24
major question is whether there are other of the Andean Common Market, which re-
means of transferring scarce managerial and quires multinational corporations to fade
technical knowledge without having to be in down from 100 percent to 49 percent owner-
joint supply with capital as in a foreign direct ship over a period of 15 or 20 years.5 There
investment. As already noted, the cost for may be other requirements and means for di-
foreign equity capital is high, even post-tax. vestment over time.6 Moving still further to a
The host country may well consider this cost dilution of the foreign equity, the host coun-
excessive for the foreign managerial and
technical knowledge which it desires but 5Andean Foreign Investment Code, Commission of
the Carthegena Agreement, International Legal Ma-
which cannot be acquired through a direct terials, Vol. 2, 1972.
foreign investment without the high payment 6Albert O. Hirschman, How to Divest in Latin
for equity capital with which it is in joint sup- America, and Why, Princeton Essays in International
ply. The new approach to foreign investment Finance, November 1969.
 330 MOBILIZING FOREIGN RESOURCES

try may insist at the very outset on a joint need is still much more for simply the adap-
venture and may also establish a limited time tation, imitation, widening, and deepening of
for any foreign equity participation. At the existing methods of investment.
other end of the spectrum, the host country Finally, it may be doubted whether there
may exclude foreign equity altogether and will be sufficient motivation when there is no
seek the managerial and technological equity interest. But managerial fees can be
knowledge through other contractual ar- graduated according to profits, and even
rangements that might allow the transfer of other objectives such as foreign exchange
technical and managerial skills without being savings realized. Other incentive devices —
tied to equity capital. Licensing agreements, short of ownership — can also be utilized.
technical services agreements, engineering In general, these contractual arrangements
and construction contracts, management are extremely flexible devices for securing
contracts, and co-production agreements can the transfer of the nonmonetary resources of
prove to be of considerable benefit to a de- management and technology; they may be
veloping country — supplying the needed for- adapted to widely diverse circumstances; and
eign knowledge at a lower cost than must be their utility in meeting a variety of objectives
paid when the knowledge comes with equity is becoming increasingly appreciated. To-
gether with the fact of development planning,
capital.7
Some doubts may, however, be expressed these new approaches to international invest-
that the LDCs can indeed realistically expect ment are causing both the foreign investor
to receive foreign technological and mana- and the recipient country to consider the ben-
gerial knowledge without equity capital. This efits and costs of foreign investment in a
may prove difficult or impossible for the lat- newer light. This approach can be particu-
est technology or research-and-development larly efficacious when activities are reserved
type of enterprise because the discoverer and under a development program for public
owner of such technology normally wishes to ownership or for majority ownership by local
retain control. But this does not apply to the nationals, but there is still a need for seeking
older, more standardized types of technology. technical information or managerial services
Assuming that the demand for such technical from abroad.
services should increase, then it can be ex- If the technical and managerial compo-
pected that a supply of such services will be nents of direct foreign investment can be se-
induced.
cured through contractual nondirect invest-
Similarly, it can be contended that a man- ment devices without having to grant its
agement contract is not an effective instru- supplier a controlling equity interest, and if
ment for the discovery of investment oppor- financial aid is provided from foreign public
tunities as distinguished from simply the sources of capital (either directly or indi-
operation of an investment. This may be true, rectly through a development bank or devel-
but in the LDCs the discovery of new invest- opment corporation), then the recipient LDC
ment opportunities is not as relevant as in may benefit from an optimal mix of public fi-
more advanced countries, and the crucial nancial aid and private technical assistance.
The potentialities for combining local pub-
7A co-production agreement allows a national firm lic or private ownership with technical assis-
in the LDC to acquire imported equipment and tech-
tance from private foreign enterprise and
nology in return for payment "in kind" by exporting
its products for a number of years to the guaranteed capital assistance from public sources de-
market overseas. serve considerable emphasis.
Comment
Cost-benefit analysis as used in project appraisal has as much applicability to the assess-
ment of private foreign investments as to domestic investment projects. Chapter X discusses
in detail the techniques of project appraisal. See also A. K. Sen, The Flow of Financial Re-
 331 PRIVATE FOREIGN INVESTMENT

sources: Methods of Evaluating the Economic Effects of Private Foreign Investment (Report
to UNCTAD, August 20, 1971).
Sen notes that the problem of negotiations on terms and conditions, including possible
concessions, is an important aspect of foreign private investment that requires special atten-
tion (pp. 20-21). He states:
Let a foreign firm F be in negotiation with the government of a developing country G. Consider a
series of contracts involving various proposals for concessions, e.g., combinations of tax benefits, repa-
triation facilities and exchange arrangements. If both F and G prefer one sort of contract to another,
then the latter can be rejected. After weeding out the non-starters, there will remain a sequence of
contracts, which can be labelled 1 to n, such that F always prefers a contract with a higher number and
G a contract with a lower number. Which particular contract will be chosen will depend on the bargain-
ing power of the two parties.
How would the government G rank the alternatives? If the alternative projects are physically identical
and differ only in the extent of the concessions given to F, then ranking them may be no great problem.
G will prefer to make fewer concessions just as F will prefer more. But what if the alternative proposals
differ from each other in production structure and not merely in terms of concessions? Or what if one
proposal involves more concessions of one type (e.g., tax benefits) and less of another (e.g., repatriation
facilities)? Then the ranking procedure must be related to the present value estimations.
The theoretically satisfactory way of solving the problem is not far to seek though its practical appli-
cability may be open to doubts. For the economy as a whole, alternative foreign investment proposals
including concessions represent alternative production possibilities. If the shadow prices are chosen ap-
popriately, no more than one alternative will have positive present value. For the evaluation of each
project the other projects are alternatives that are rejected and the maximum benefits thus foregone
must be counted as costs for the particular project under scrutiny
This procedure makes it possible to select the best alternative from a given set. By doing the exercise
again after eliminating the best project in the list, the second-best can be identified. And so on for the
whole lot, thereby obtaining the entire ranking.
In practice this procedure may be thought to be complex, though it is by no means infeasible given
the progress in computational facilities. Under a simplifying assumption a much easier procedure could
be applied. If the projects are so small that they are unlikely to alter the shadow prices for the economy
as a whole, except the shadow price of the option of selecting one of the alternatives, then the projects
could all be ranked in terms of their respective present values at given shadow prices, excluding the cost
of losing the option of doing one of the projects. Since this last cost item is not included, the "present
value" calculated this way can be positive for more than one alternative project. The rule in this case is
straightforward: choose the project with the highest "present value." (This applies only insofar as the
"present values" are positive. If a project has negative "present value," then its yield is negative even
without bringing in the cost of the option lost. Thus a negative "present value" is a ground for rejection,
though a positive "present value" is not necessarily a ground for acceptance.)
The distinction between the "present value," as calculated in this way, and the true present value of
net benefits is important. For the latter the rule is to choose all projects that yield positive present value,
while for the former the alternatives have to be ranked in terms of "present value" and the alternative
with the highest value is chosen
 332 MOBILIZING FOREIGN RESOURCES

V.D.2. Transnational
Enterprises and Linkages*

This paper reviews the literature on the re- the entry of TNC investment. The classic
lationships between transnational corpora- general discussion of the role of linkages in
tions (TNCs) in the manufacturing sector the development of LDCs is by Hirschman
and domestic enterprises as well as industrial (1958), who proposes a deliberate strategy of
structures in host LDCs. There are two broad creating imbalances to harness the forces of
sets of relationships involved, both of which entrepreneurship and growth that lie latent
are of significance for understanding the ef- in every economy. Particular investments are
fects of TNCs on host economies and to the thus supposed to create such strong external
formulation of policy. The "direct" relation- economies in sectors that supply or buy from
ships that TNCs strike up with local sup- them that new investments are undertaken in
pliers or purchasers (backward and forward order to exploit them; foreign investment is
"linkages" in the Hirschman sense) can con- assigned a vital role "to enable and to em-
stitute powerful mechanisms for stimulating bolden acountry to set out on the path of un-
(or retarding) economic, and particularly in- balanced growth . . . [and] to take the first
dustrial, growth in LDCs. The "indirect" ef- unbalancing
While it issteps obvious in growth sequences.2
that TNC investments
fects that the entry and operations of TNCs
may have on local industrial structure, con- can create strong local linkages in this sense,
duct, and performance may be equally im- and indeed the policies of many host LDCs
portant: TNCs may change the nature and to compel TNCs to maximize their purchase
evolution of concentration; they may affect of local inputs have aimed at exploiting these
the profitability and growth of indigenous linkages, it is far from clear how the actual
firms; they may alter financing, marketing, experience of TNCs in LDCs is to be evalu-
technological, or managerial practices of the ated. The normal procedure, of using the pro-
sectors that they enter; they may, by preda- portion of local to total purchases by TNCs
tory conduct, drive domestic firms out of as an indicator of backward linkages, is in-
business; and so on. adequate, since it does not take account of
externalities and does not enable us to assess
the "efficiency" of linkage creation. Though
DIRECT LINKAGES it may serve as a crude approximation to the
outer limits of the stimulus provided by
Direct linkages may be defined to consti- TNCs to local enterprises, it does not, for in-
tute those relationships between TNCs and stance, show (a) if the local enterprises would
domestic enterprises trading with them that
have led the latter to respond, positively or have been set up in the absence of TNC in-
vestments; (b)whether they gained or lost by
otherwise, to technological, pecuniary, mar- having TNCs as major customers (where this
keting, or entrepreneurial stimuli provided by is the case); (c) if the host economy could
the former.1 A "linkage" in this sense is have created the same linkages at lesser cost,
clearly different from a normal transaction in
say by replacing the TNC by a local firm; (d)
a competitive market; it refers essentially to if the linked local enterprises are desirable
externalities created for domestic industry by
from the social point of view (where the link-
ages are fostered behind heavy protective
♦"Transnational, Domestic Enterprises, and In- barriers); and (e) whether negative linkages
dustrial Structure in Host LDCs: A Survey" by San-
jaya Lall. From Oxford Economic Papers, Vol. 30 were created by stifling potential local invest-
(1978). Reprinted by permission of Oxford University
Press.
ment. The proper economic evaluation of

'See Scitovsky (1954) and Hirschman (1958). 2Hirschman (1959), pp. 205-6.
 333 PRIVATE FOREIGN INVESTMENT

linkages must be based on a case-by-case the economics of domestic purchase.3 Simi-

cost/benefit examination of actual situations larly, studies of U.S. investment in LDCs in
and plausible alternatives, necessarily a dif- general,4 Peru,5 Iran, 6 South Korea,7 Malay-
ficult and impressionistic procedure; how- sia,8 and some others,9 have discussed the use
ever, the existence and desirability of TNC of local inputs by TNCs. The general find-
linkages can only be judged by some such ings, that TNCs buy relatively few inputs
method. within the less industrialized host economies
Let us now consider how the existing lit- but may be forced or persuaded to increase
erature has treated the issue. Our review is local content by the more advanced ones, add
confined to backward linkages, since there little to our understanding of the significance
appears to be hardly any empirical work on or desirability of the externalities created in
forward linkages created by TNCs: perhaps LDCs.
not a great omission, since forward linkages There are three other studies which may
cannot be expected to be very strong. We
may consider backward linkages for the two
3For a summary see Lall and Sreeten (1977). This
main forms of TNC investment — import- study found that over half the sample firms imported
substituting and export-orientated — sepa- goods worth over 30 per cent of their total value of
rately, as the issues raised are rather sales — over 65 per cent of India is excluded — but that
different. this high degree of import dependence did not differ
significantly between TNCs and other firms.
4See Mason (1967), who attempts to quantify the
Import-Substituting TNCs local linkages of U.S. foreign investments using ag-
gregate industry data, measuring backward linkages
The vast bulk of foreign manufacturing in- simply by the "ratio of local expenditure to total
vestment in LDCs has gone into protected sales" and forward linkages by the "ratio of local
import-substituting activities, where govern- sales to total sales." See Hufbauer and Adler (1968)
for estimates of local and foreign buying propensities
ments have been able, especially in the larger of U.S. TNCs.
and more industrialized areas, to push firms
5Vaitsos (1976) gives comparative data on the im-
into buying large proportions of their inputs port propensities of local and foreign firms in Peru for
from local sources. TNCs have, conse- 1973, which show that foreign firms had higher im-
ports in eleven out of twelve broad industry groups.
quently, developed extensive and long-stand- He notes, however, that this does not necessarily
ing relationships with local enterprises in imply that local firms created more linkages (p. 40);
countries like India, Mexico, Brazil, Argen- such aggregate data do not permit a detailed exami-
tina, and so on. Despite the significance of nation ofthe technologies and products involved.
the phenomenon, however, the existing work 6Daftary and Borghey (1976) provide rather
sketchy data on local purchases by thirteen TNCs,
on TNCs has paid scant attention to exam- and conclude that few of these have set up significant
ining the economic benefits and costs of the
local linkages, except with the domestic packaging in-
linkages created.
dustry (pp. 75-6).
Most of the studies of foreign investment 7See Jo (1976), who reports that foreign firms have
in LDCs have simply noted the extent of local higher import propensities than local firms, and that
purchasing by TNCs and sometimes re- supply, cost, and quality problems limit the growth of
local purchasing.
marked on the general difficulties of local
procurement (due to technological back- 8Thoburn (1973), in the course of his study of the
tin and rubber sectors, touches upon the role of for-
wardness, small scale, high cost, poor quality, eign engineering firms in Malaysia. He finds that they
or unreliability) without attempting to ana- helped in the development of local suppliers in pro-
lyse the linkages created in any detail. Thus, cesses of low capital intensity and few scale econo-
a study of six developing countries commis- mies; "but in all cases the firms concerned were al-
ready in existence when the link with foreign firms
sioned by UNCTAD produced data on the
was formed" (p. 1 13).
import propensities of 159 firms, foreign and
9See chapters 5 and 14 of the annotated bibliog-
domestic, in Kenya, Jamaica, India, Iran, raphy on TNCs by Lall (1975) for other references
Colombia, and Malaysia, without going into on LDCs.
 334 MOBILIZING FOREIGN RESOURCES

be mentioned separately because they tackle the role of TNCs), and compares its experi-
the issue of TNC linkages more directly. ence to the highly successful one of Japan.
Reuber and associates (1973), using infor- A more pessimistic view of the virtues of
mation provided by the head offices of TNCs, local purchasing emerges from Baranson's
noted that import-substituting investments study of the Cummin's diesel-engine project
created far more local linkages than export- in India and his analysis of the automotive
orientated ones, and found, for sixty-four industry in LDCs generally.11 He comments
sample firms, that 45 per cent of inputs in at length on the problems raised by the high
1970 came from local sources. Parent com- cost, poor quality, and unreliability of local
panies were asked whether their operations suppliers in cases where the government has
had given rise to local suppliers or distribu- forced the the
pacereasons
of buying localstate
inputs,12 and
tors, and their answers indicated that some discusses for this of affairs
one-third of the investments had directly (protection, technological and skill shortages,
given rise to such local activity. Reuber made lack of experience, small scale, and the like).
no attempt to assess the costs and benefits of We must not, however, draw unfavourable
such linkages, and also qualified the esti- general conclusions about the desirability of
mates by noting that "such figures must be linkages or the capabilities of local enter-
viewed with some suspicion both because of prises from this experience: there are several
the many conceptual and practical difficulties other industries (see below) where domestic
in deriving estimates of this kind and because linkages have been economically viable, and
of the vested interest of respondents in pre- even for India the recent boom in exports of
senting the spin-off effects of their activities medium-to-high technology goods (including
in as favourable a light as possible" (p. 156). transport equipment, chemicals, and engi-
Watanabe (1972 and 1974), in his exami- neering goods) indicates that some of the
nations ofsubcontracting in LDCs, presents problems described by Baranson for the 60s
a general but useful analysis of this particu- may have been the teething difficulties of
lar (and rather strong) form of linkage. launching new and complex industrial
Though he is not concerned exclusively with
TNCs, he cites examples of foreign firms processes.
To return to TNCs, however, we find that
(like Singer in South East Asia) which have we are left with very little empirical work on
used subcontracting successfully, and con- the process and value of creating linkages.
cludes that such activity, "by stimulating en- The general impression conveyed by the lit-
trepreneurship and encouraging industrial erature isthat TNCs establish relatively few
efficiency, can help to promote the industrial- linkages in small or industrially backward
ization of the less developed countries and economies; that in larger economies they may
thus create the additional employment op- create extensive linkages, mostly because of
portunities they badly need."10 He analyses government pressure; and that a substantial
the conditions for success of such linkages part of these linkages in import-substituting
(which he terms "within-border industrial industries may be excessively costly and un-
subcontracting"), briefly notes the contribu- economical. This is all in line with a priori
tion that TNCs may make by providing as- expectation, but it is sadly inadequate in ex-
sistance with investment, technology and plaining the specific nature of the linkages
quality control, and recommends policies for that have been created, and in providing the
increasing linkages; he does not, however, ex-
amine in detail any specific instances of sub- nBaranson (1967, 1969).
contracting byforeign firms. In his 1974 12Lall (1977) notes that foreign automobile assem-
paper he examines the problems of subcon- bly plants in Malaysia have not created significant do-
tracting inIndia (though without discussing mestic linkages; despite the government's expressed
desire to increase local content, the absence of statu-
tory controls on imports has led the TNCs to continue
10Watanabe(1972), p. 425. to depend heavily on imports.
 335 PRIVATE FOREIGN INVESTMENT

' sort of evaluation of their social value that differentiation, marketing, or product inno-
was described earlier. vation are important, however, large TNCs
may predominate in production and export
Export-Oriented TNCs activity.13 In both cases, there exists a vast
potential for linkages with domestic produc-
The recent growth of manufactured ex- ers, who may manufacture components or
| ports from LDCs by foreign firms has at- whole products for foreign firms.
I tracted a great deal of attention, and the Third, there are new TNC investments in
creation of linkages, especially by subcon- "modern" industries in LDCs undertaken
tracting, has often been mentioned in this specifically for export, transferring fairly
context. In contrast to the inefficiency usually complex technologies to LDCs to service es-
| associated with linkages in import-substitut- tablished world markets. A constellation of
| ing industries, it may be expected that link- factors (labour and transport costs, the na-
, ages created by firms competing in world ture of the technology, need for short produc-
j markets will be more efficient and beneficial tion runs, managerial requirements, and, of
for the host economies (at least in a narrow
course, political stability)14 influences the de-
technical sense, without referring to distri- cision to locate such investments, good ex-
I butional, social, or political effects). It would amples of which are the Philips and General
be useful to start by distinguishing four types Electric complexes in Singapore, or some
i of export-oriented TNCs which have differ- "border industries" in Mexico; the availabil-
ent implications for the creation of domestic ity of local components is not, however, one
\ linkages. of the important factors attracting them. In
First, there are TNCs which started by most cases, such investments are tightly con-
substituting for imports and have grown in- trolled from abroad, the components and pro-
ternationally competitive enterprises with cesses may be quite advanced, and there may
substantial export interests (VW in Brazil or not be much scope for local linkages. It is
Singer in Asia may be good examples). Such possible, nevertheless, that local enterprises
activities usually involve technologies which may be able to provide some products at the
are stable and not very sophisticated, and right price and quality, and a few linkages
they are based in areas with a cheap but rel- may develop in the more advanced of the host
atively skilled labour force and an experi-
enced indigenous sector. The use of the economies.15
Fourth, there are "sourcing" investments
TNCs' marketing networks and established where only a particular (labour-intensive)
brand names are important in such export ac- process is transferred to LDCs, the more cap-
tivity. These TNCs may have established ital-intensive processes being retained in the
considerable domestic linkages in the early home countries where the requisite equip-
phases, though, of course, the extent and na- ment, skills, and R and D facilities exist. The
ture of these linkages may change as they best-known example of this is the electronics
gear themselves for world markets. industry, especially the semiconductor sector,
Second, there may be foreign firms which where the demanding specifications, the rap-
produce and export "traditional" products idly changing technology, and requirements
like footwear, textiles, processed foods, or 13See de la Torre (1974) and Helleiner (1976).
sports goods. For those industries (like tex- 14For a detailed analysis of these factors see
tiles) where technology is easily available and Sharpston(1975).
product differentiation is insignificant, the 15UNCTAD (1975) notes that this is happening in
foreign firms involved may be buying groups, South Korea, Taiwan, Hong Kong and Singapore.
retailers, or small manufacturers (sometimes However, KOnig (1975) finds that Mexico, despite its
from other LDCs, like Hong Kong firms in industrial development, is unable to provide "border
industries" with even 1 per cent of their inputs; he
Malaysia) rather than TNCs proper. For places the full blame for this on inefficiencies caused
those (like food processing) where product by protection.
 336 MOBILIZING FOREIGN RESOURCES

of cost minimization reduce the scope for do- concessions certainly have been generous;
mestic linkages to practically nothing.16 TNCs have clearly shown a marked prefer-
Of these four types of TNC investment, ence for stable regimes with little or no la-
the first two are likely to create the most link- bour problems; the incidence of "squeezing"
ages, the third rather less, and the fourth local firms needs further investigation; and
least of all. The extent of linkages created in export market instability is not a particular
particular LDCs depends upon the stage of feature of TNC exports. On the whole, the
development of indigenous industry, the benefits seem to have outweighed the costs
availability of local skills and technology, in- with LDCs, and many of them are now seek-
stitutions and government policies, changes ing to attract TNCs or foreign buying
in demand and technology in world markets
and their political attractiveness to TNCs.17 Besides the general studies of this phenom-
The main benefits of such investment are enon mentioned above, a number of country
generally supposed to be employment crea- groups.19
studies have discussed export-oriented for-
tion, export promotion (though net foreign eign investment (and subcontracting) for
exchange benefits may be very low for the Mexico,20 Hong Kong,21 Singapore,22 the Ca-
third and fourth groups that depend heavily ribbean,23 and Taiwan.24 Nearly all of
on imported components), skill and technol- them — with the exceptions of Evers (1977)
ogy transfer (particularly in the first and sec- and Fernandez (1973) — have come to fa-
ond, sometimes the third, groups), and the vourable conclusions about the net benefits of
stimulation of local linkages (see below). The such activity to host LDCs, but their discus-
main costs mentioned are the generous fiscal sion of linkages as such has remained desul-
and infrastructural incentives that LDCs tory and unsatisfactory. There are some im-
have to offer (especially for investments in pressionistic and anecdotal accounts25 of the
the fourth category), the socio-political con- potential for creating beneficial linkages
straints ofhaving to ensure a docile and low- which confirm the general analysis given
cost labour force, the danger of losing "foot- above, but none of these studies has at-
loose" TNCs when costs rise, the risk of get- tempted a systematic evaluation of the ex-
ting poor terms from monopsonistic buyers, tent, costs, and benefits of linkages from the
and the instability of demand for exports. Of viewpoint of the host economy or the local en-
these, the danger of "footloose" behaviour terprises. Of those which have touched on
does not seem to have been realized;18 fiscal linkages, the following may be mentioned:

16See UNCTAD (1975), Change (1971), Finan 19East European countries are also entering the
(1975). U.S. Tariff Commission (1970) and Lim and field, and their use of "industrial co-operation agree-
Pang (1976). Some products of the electronics indus- ments," under which Western firms provide technol-
try, mainly in consumer electronics, are amenable to ogy, and usually also equipment and intermediate in-
local manufacture in their entirety, and so fall into the puts, in return for processed goods, seems to have
third group. provided major benefits to their smaller establish-
ments without incurring the problem of direct TNC
17For general discussions of the determinants of ex-
port-orientated foreign investments and their costs investment. This arrangement, discussed by Hewett
and benefits see Helleiner (1973, 1976), Adam (1972, (1975), may serve as a model to the more advanced
LDCs.
1975), Michalet (1977), de la Torre (1974); for an
analysis of the significance of labour skills in trade see 20See Konig (1975), Baerresen (1971), Walker
Hirsch (1975) and for a recent theoretical analysis of (1969), (1974)^
tanabe Fernandez (1973), Sahagun (1976) and Wa-
TNCs and trade see Hirsch (1975, 1976); for a de-
scription ofthe role of multinational buying groups 21Reidel (1974) and Evers et al. (1977), the latter
see Hone (1974); and for an examination of the tariff concentrating on textiles.
provisions in developed countries which lead to "off- 22Lim and Pang (1976) on the electronic industry.
shore assembly" see Finger (1975). "Van Houten
24Reidel( 1975).(1973).
18Rising wage costs in Singapore have led TNCs to
upgrade the skill content of their activities rather than
leave the country. See Lim and Pang (1976) and Lall "Especially in Helleiner (1976), Watanabe (1972)
(1977). and Sharpston (1975).
 337 PRIVATE FOREIGN INVESTMENT

1. Evers et al. (1977), on the textile and higher than Hong Kong or Taiwan so that
clothing industries in Hong Kong, discuss they are constantly threatened with losing
the role of trading companies in developed their markets; they complain of little
countries that subcontract to local manu- assistance from the government; and
facturers. They find that local linkages for logy.
they are short of finance and new techno-
clothing manufacture, in terms of the pur-
chase of local cotton textiles, has weak- 3. UNCTAD (1975) reviews the electronics
ened rather than strengthened in recent industry in LDCs generally, and reaches
years with the growth of exports, for two optimistic conclusions about the effects
reasons: discrimination in developed coun- and prospects for subcontracting. It finds
tries against cotton, and the demand for that several finished electronic products
higher quality products, both leading to a can be successfully manufactured by local
greater dependence on imported textiles enterprises in South East Asia, and sub-
(often supplied by the buyers). The au- contracting has led to "a whole network of
thors comment extensively on poor work- small manufacturers that were set up as a
ing conditions, use of child labour, exces- result of the backward linkages created"
sive working hours (up to 105 hours per
week for men), and low wages that sup- Clearly, much more evidence is needed on
port the success of the industry in Hong
Kong, and draw unfavourable conclusions the experience 26).26 of different industries in dif-
(P-ferent LDCs before we can generalize about
for the distribution of benefits resulting
the impact of TNC linkages in export-based
from such export-led growth. Conditions industries. It is obvious that substantial link-
are apparently worse in small establish- ages have been created, and that in some sec-
ments, since large ones have themselves
become multinational and gone to cheaper tors, like electricals, they have been benefi-
cial to host countries; however, it is possible
areas like Malaysia and Indonesia. that in some other industries, like textiles,
2. Lim and Pang (1976), who survey the linkages have been weakening and have had
electronic industry in Singapore, note that undesirable effects on distribution and wel-
European firms buy a fair amount of their
fare. A related question which has been al-
inputs (40-50 percent) locally, while U.S. most totally neglected is whether such ex-
(under 10 percent), and Japanese (about
porting activity (perhaps excluding very
20 percent) buy much less. This is due to
the fact that U.S. firms are specialized in high-technology products) could have been
undertaken more economically by local firms
the semi-conductor sector and Japanese in countries like India, where the bulk of
firms in high-technology components, be-
"modern" manufactured exports are not in
yond the technological capabilities of do- fact accounted for by TNCs, and whether
mestic firms, while European firms man- this would have created more beneficial link-
ufacture mainly consumer electronics
ages. This whole area is of vital importance
where the scope for local purchase is
higher. However, local products tend to be to policy-making, and cries out for detailed
empirical research.
rather costly, and are purchased chiefly in
order to qualify for GSP privileges in sell-
ing to Europe (a minimum local content is
required for these exports). Local firms
face the usual problems of quality, tech-
nology, high costs, and so on, and are
sometimes assisted by the local TNCs 26 As noted above, however, this has not occurred
for "border industries" in the advanced but highly
from whom they subcontract by free tech- protected economy of Mexico, even in textiles, despite
nology transfers. Firms which subcontract the efforts of some TNCs to increase local content in
to foreign buying groups seem to face order to qualify for GSP (General System of Prefer-
greater problems; their wage costs are ences) privileges. See Konig (1975), pp. 92-4.
 338 MOBILIZING FOREIGN RESOURCES

References ployment in the less developed countries,"

Economic and Political Weekly, annual
Adam, G. (1972), "Some implications and number, February, pp. 247-62.
concomitants of worldwide sourcing," multinational firms," Economic Journal,
Acta Oeconomica, pp. 309-23.
Adam, G. (1975), "Multinational corpora- pp. 21-47.G. K. (1976), "Transnational en-
Helleiner,
tions and worldwide sourcing," in H. Rad- terprises, manufactured exports and em-
ice (ed.), International Firms and Modern Hewett, E. A. (1975), "The economics of
Imperialism (Penguin). East European technology imports from
Baerresen, D. W. (1971), The Border Indus- the West," American Economic Review,
trialization Program of Mexico (D. C.
Heath). May, pp. 377-82.
Hirsch, S. (1975), "The product cycle model
Baranson, J. (1967), Manufacturing Prob- of international trade," Oxford Bulletin of >
lems in India: the Cummins Diesel Expe- Economics and Statistics, pp. 305-17.
rience (Syracuse University Press). Hirsch, S. (1976), "An international trade
Baranson, J. (1969), Automotive Industries and investment theory of the firm," Oxford
in Developing Countries (IBRD). Economic Papers, pp. 258-70.
Chang, Y. S. (1971), The Transfer of Tech- Hirschman, A. O. (1958), The Strategy of
nology: Economics of Offshore Assembly, Economic
Press). Development (Yale University
the Case of Semiconductor Industry, Re-
port No. 1 1 (UNITAR). Hone, A. (1974), "Multinational corpora-
Daftary, F., and Borghey, M. (1976), tions and multinational buying groups,"
"Multinational enterprises and employ- World Development, February, pp. 145- i
50.
ment in Iran," World Employment Pro-
gramme, Working Paper No. 14, ILO, Hufbauer, G. C, and Adler, F. M. (1968),
mimeo. U.S. Manufacturing Investment and the
De la Torre, J. 1974), "Foreign investment Balance of Payments (U.S. Treasury
Department).
and export dependency," Economic Devel-
opment and Cultural Change, pp. 135-50. Jo, Sung-Hwen (1976), "The impact of
Evers, B., de Groot, G., and Wagenmans, W. multinational firms on employment and in-
(1977), "Hong Kong: development and comes: the case study of South Korea,"
World Employment Programme, Working
perspective of a clothing colony," trans-
lated summary of Progress Report No. 6, Paper No. 12, ILO, mimeo.
Development Research Institute, Tilburg, Konig, W. (1975). "Towards an evaluation
Netherlands, mimeo. of international subcontracting activities in
Fernandez, R. A. (1973), "The border indus- developing countries," U.N.
trialization program on the U.S. -Mexico Lall, S. (1975), Private Foreign Manufactur-
border," Review of Radical Political Eco- ing Investment and Multinational Corpo-
nomics, Spring, pp. 37-52. rations: An Annotated Bibliography
Finan, N. (1975), "The international trans- (Praeger).
fer of semiconductor technology through Lall, S. (1977), "Transfer pricing in assem-
U.S. based firms," Working Paper No. bly industries: a preliminary analysis of the
118, National Bureau of Economic Re- issues in Malaysia and Singapore," Com-
search, New York. monwealth Secretariat, London, mimeo.
Lall, S., and Streeten, P. P. (1977), Foreign
Finger, J. M. (1975), "Tariff provisions for
offshore assembly and the exports of de- Investment, Transnationals and Developing
Countries (Macmillan).
veloping countries," Economic Journal,
pp. 365-71. Lim, L., and Pang, Eng-Fong (1976), "The
Helleiner, G. K. (1973), "Manufactured ex- electronics industry in Singapore: struc-
ports from less-developed countries and ture, employment, technology and link-
 339 PRIVATE FOREIGN INVESTMENT

age," Economic Research Centre, Univer- Thoburn, J. T. (1973), "Exports and the Ma-
sity of Singapore, mimeo. laysian engineering industry: a case study
Mason, R. H. (1967), "An analysis of bene- of backward linkage," Oxford Bulletin of
fits from U.S. direct foreign investments in Economics and Statistics, pp. 91-117.
less-developed areas," Ph.D. thesis, Stan- UNCTAD (1975), International Subcon-
ford University. tracting Arrangements in Electronics be-
Michalet, C.-A. (1977), "International sub- tween Developed Market-Economy Coun-
tries and Developing Countries (U.N.).
contracting," in OECD Development
U.S. Tariff Commission (1970), Economic
Centre's "Experts" Meeting on Interna-
tional Subcontracting and Reinforcing Factors Affecting the Use of Item 807.00
LDCs' Technological Absorption Capacity, and 806.30 of the Tariff Schedules of the
OECD, mimeo. United States (Government Printing Of-
Reidel, J. (1974), The Industrialization of fice, Washington, D.C.).
Hong Kong (Institut fur Weltwirtschaft). Vaitsos, C. V. (1976), "Employment prob-
Reidel, J. (1975), "The nature and determi- lems and transnational enterprises in de-
nants of export-oriented direct foreign in- veloping countries: distortions and inequal-
vestment ina developing country: a case ity," World Employment Programme,
study of Taiwan," Weltwirtschaftliches Working Paper No. 1 1 , ILO, mimeo.
Archiv, pp. 505-28. Van Houten, J. F. (1973), "Assembly indus-
Reuber, G. L. et al. (1973), Private Foreign tries in the Caribbean," Finance and De-
Investment in Development, Oxford, Clar- velopment, June.
endon Press.
Walker, H. O. (1969), "Border industries
Sahagun, V. M. B. et al. (1976), "The im- with a Mexican accent," Columbia Jour-
pact of multinational corporations on em- nal of World Business, January-February,
ployment and incomes: the case of Mex-
ico," World Employment Programme, Watanabe, pp. 25-32.S. (1972), "International subcon-
Working Paper No. 13, ILO, mimeo. tracting, employment and skill promo-
Scitovsky, T. (1954), "Two concepts of ex- tion," International Labour Review, pp.
425-49.
ternal economies," Journal of Political
Economy, pp. 143-52. Watanabe, S. (1974), "Constraints on la-
Sharpston, M. (1975), "International sub- bour-intensive export industries in Mex-
contracting," Oxford Economic Papers, ico," International Labour Review, pp. 23-
pp. 94-135. 45.

V.D.3. Multinational Enterprises

in Developing Countries— Note

The multinational enterprise (MNE) — with that the MNE may dominate the host coun-
facilities in many countries and responsive to try or impose excessive costs.
a common management strategy — has Criticism of the MNE is but the latest at-
gained increasing prominence as an instru- tempt to dispel complacency over the rele-
ment for private foreign investment in devel- vance of the neoclassical theory of interna-
oping countries. While the capital, technol- tional trade for development problems. As we
ogy, managerial competence, and marketing shall see in Chapter VIII, critics of the neo-
capabilities of an MNE can be utilized for a classical trade theory first attempted to dis-
country's development, there is also a fear credit the theory's power to explain historical
 340 MOBILIZING FOREIGN RESOURCES

development by arguing that international gies, and marketing based upon expectations
trade had actually operated historically as a of returns and risk factors, the MNE concen-
mechanism of inequality. After the establish- trates on the total net worth of the investor's
ment of the postwar international economic interests, not that of an individual subsidiary
institutions, the argument shifted to a criti- alone.
cism of the alleged biases and deficiencies of Although these features of foreign invest-
the international institutions comprising the
ment associated with "multinationality" do
Bretton Woods system. And now the MNE not call for a different type of analysis from
has become the object of criticism, with pes- the general benefit-cost analysis already dis-
simistic warnings about future detriment to cussed, the multinational firm is character-
the developing countries if the MNE is not ized by some behavioral differences that
sufficiently regulated. broaden the reach and increase the intensity
Does the evaluation of the operation of the of both the benefits and costs of foreign in-
MNE, however, call for more than a social vestment. As Caves has indicated,1 a national
benefit-cost type of analysis, such as outlined branch of a multinational firm might behave
in selection V.D.I? Why should the MNE be differently from an equal-sized independent
analyzed differently from direct investment company for three reasons:
from one home country to a single host coun- (1) Motivation. The multinational firm
try? The essential question is what difference
maximizes profits from its activities as a
does the attribute of "multinationality" make whole, rather than telling each subsidiary to
to the analysis?
maximize independently and ignoring the
True, an MNE is likely to have the power
profit interdependences among them. The
of an oligopoly; but so too may some "sim- multinational firm also spreads its risks, and
ple" foreign enterprises or domestic indepen- could therefore behave quite differently in an
dent companies. True, an MNE may be a
vertically integrated enterprise that uses uncertain situation from an independent hav-
transfer prices for intrafirm trade; but so too ing the same risk-return preference function.
(2) Cognition and information. Its corpo-
may "simple" foreign investment. True, an rate family relations give the multinational
MNE may be involved in the costly process
unit access to more information about mar-
of import substitution; but so too may a sim- kets located in other countries— or to infor-
ple foreign enterprise or a national firm.
mation to which it can attach a higher degree
True, an MNE may be depleting too rapidly of certainty.
a wasting asset; but so too may other forms
of investment that are not multinational in (3) Opportunity set. The set of assets held
character. In assessing the contribution of a by a multinational unit can differ from a na-
foreign investment by an MNE, we have to tional firm's in various ways — perhaps most
be clear on whether what is being assessed is notably in its skill in differentiating its prod-
the investment project per se, the foreigness uct and its financial capacities.
of the investment, or some alternative insti- Although Caves's analysis is concerned
tutional arrangement for acquiring the ingre- with the differences in market behavior be-
dients ofthe direct investment package. tween a national branch of a multinational
The distinguishing feature of an MNE is firm and an equal-sized independent com-
that the range of its major decisions (finance, pany, these behavioral differences are also
investment, production, research and devel- relevant as between a multinational firm and
opment, and reaction to governmental poli- a "simple" type of foreign enterprise in the
cies) isbased on the opportunities and prob- developing country.
lems that the MNE confronts in all the
countries in which it operates. In utilizing its 'Richard E. Caves, International Trade, Interna-
tional Investment, and Imperfect Markets, Princeton
"global scanning capacity" to determine its University Special Papers in International Econom-
investment plan, worldwide sourcing strate- ics, No. 10, November 1974, pp. 21-2.
 341 PRIVATE FOREIGN INVESTMENT

These behavioral differences can be espe- fective the LDC's potential comparative ad-
cially significant in: first, promoting foreign vantage. The MNE provides the complemen-
investment; second, allowing the MNE to act tary resources of capital, technology,
as a unit of real economic integration; and management, and market outlets that may be
third, endowing the MNE with greater bar- necessary to bestow an "effective" compara-
gaining power. tive advantage to the labor-surplus factor en-
The growth of MNEs tends to promote dowment inthe host country.
more foreign investment because the MNE is This can also be evaluated as efficient in-
less of a risk-averter when it operates in a ternational production. The MNE views pro-
number of countries, produces a number of duction as a set of activities or processes, and
products, practices process specialization, the global strategy of the MNE amounts in
and enjoys greater maneuverability with re- essence to the solution of activity models of
spect to marketing opportunities and condi- production, with production processes in
tions of production than does a firm with a many countries. A competitive equilibrium
narrower range of activities. solution to the programming problem is im-
Emphasizing the interrelations between posed within the MNE when it operates effi-
output and input flows in international trade, ciently as a planning unit, and each process
Baldwin has stated: is in the solution basis. Regulatory interfer-
ence with the MNE will alter the equilibrium
When the international firm becomes economi-
cally viable in a particular industry, not only is it basis, and some processes formerly in opera-
possible to transfer knowledge, capital, and tech- tion may become inefficient to use. The like-
nical and managerial labor across borders more ef- lihood for a labor-surplus economy is that
ficiently, but these transfers tend to be economi- labor becomes unemployed for lack of coop-
cally feasible with smaller product markets than is erating factors previously supplied by the
the case when the optimum size of productive units MNE. Or if workers were formerly employed
is small. Moreover, because of the pecuniary and in a lower productivity occupation, there is a
technological externalities that exist among inter- loss in real wages.
mediate sectors in a vertical product line, it may
not be economically profitable from a private view- This interpretation of the MNE as an ef-
ficient technical and allocational unit of in-
tionpointlisttounless
add a the new international
product to a firm
country's produc-is
mechanism tegration means that while intrafirm trade
utilized.2 conforms to corporate advantage, it is also
identical with the realization of comparative
The MNE is also a unit of integration in the
world economy. The transmission of factors advantage. If the nation-state fragments the
(capital, skills, technological knowledge, world economy through restrictions on com-
modity and factor movements and thwarts
management) via the MNE, together with international economic integration, the
the MNE's economies of scale in R&D and
MNE may serve a complementary — rather
marketing, make it a unit of real interna-
than competitive — function to the nation-
tional integration. By its multinational oper-
ations and intrafirm transactions, the MNE state: the MNE may be the vehicle for evok-
ing in practice the principle of comparative
transcends the national barriers to commod-
advantage in world trade, for trade in both
ity trade and impediments to international
factor movements. As a planning unit that outputs and inputs. The internal resource al-
makes resource allocation decisions, the location inthe MNE is a substitute mecha-
nism for the market, but when it realizes
MNE becomes the mechanism for making ef-
comparative advantage in processes and ac-
tivities, the resource allocation decisions of
the MNE will be more efficient than those in
2R. E. Baldwin, "International Trade in Inputs and
unintegrated markets that are characterized
Outputs," American Economic Review, Papers and by imperfections and uncertainty. For global
Proceedings, May 1972, p. 433. technical efficiency, the world economy is the
 342 MOBILIZING FOREIGN RESOURCES

territorial unit of international production most formal terms, the problem is for the
(not the nation-state, which is a unit of inter- host government to devise an optimal welfare
national politics). tax on foreign capital which improves the
What, however, is the distribution of gains terms of foreign borrowing.4
between the MNE and host country? More The more general problems of bargaining
pointedly, how might the net benefit for the between host government and MNE have
host country be raised? This is the crucial been summarized as follows.5
question posed by the attribute of "multina-
For the transfer of a certain "package" of know-
tionality." For multinationality instills for- how, capital management and inputs there is a
eign investment by an MNE with greater
range of values which would be acceptable to both
bargaining power because of its tendency to sides but which both sides have an interest in con-
be of larger size, its capability to exercise cealing. The ability to conceal the relevant values
wider options, and its capacity to avoid some is however much greater for the MNE than for the
forms of regulation that cannot reach beyond host country.
national jurisdiction. In settling the bargain and in drawing up the
These powers are especially suspect when contract, a large number of items may be for ne-
gotiation, inaddition to income and sales tax
they coalesce in the practice of transfer pric-
concessions and tariff and nontariff protection of
ing. The host country may believe that trans-
the product. Among these are:
fer pricing allows the MNE to minimize
1. specific allowances against tax liabilities,
taxes, escape from tariff charges, or be the
such as initial or investment allowances, de-
means of remitting profits from a subsidiary
pletion allowances, tax reporting tech-
to the parent company that would otherwise niques, loss offset provisions, etc.;
not be allowed by exchange restrictions.3 2. other
royaltyfees; payments, management fees and
If the MNE is concerned with overall
profit, not profit at any particular stage, then 3. duty drawbacks on imported inputs for
the dominant motivation behind the pricing exports;
structure in a vertically integrated operation 4. content of local inputs;
5. profit and capital repatriation;
is to gain maximum advantage vis-a-vis the
different governmental rates of taxation and 6. structure of ownership and degree and tim-
ing of local participation;
regulation of international capital flows.
7. level;
local participation in management at board
Transfer prices or cost allocation techniques
then acquire an artificial quality in the ab- 8. obligations to train local labour;
sence of "arm's-length" transactions. 9. transfer pricing;
The developing country's desire to regulate 10. rules and requirements relating to
exporting;
transfer pricing is only a special instance of
the general problem of how the bargaining 1 1 . degree of competition and forms of compe-
process between the host government and rates); tition; price control and price fixing;
MNE distributes the fruits of the foreign in- 12. credit policies (e.g., subsidized interest
vestment— more technically, the extent to 13. extent of capitalization of intangibles;
which the developing country can capture
14. revalorization of assets due to currency
from the MNE a greater share of the MNE's devaluation;
quasi-rents on its supply of technological 15. subsidies, e.g., to energy, rent, transport, or
knowledge, management, and capital. In the export expenses such as insurance, freight,
promotion;
?For an analysis of the adverse effects of the trans-
fer pricing mechanism, see Constantine Vaitsos, In-
tercountry Income Distribution and Transnational 4W. M. Corden, Trade Policy and Economic Wel-
Enterprises, Oxford, 1974, chapter 6; S. Lall, "De- fare, Oxford, 1974, pp. 339-40, 345-7, 355-64.
terminants and Implications of Transfer Pricing by 5Paul Streeten, "The Multinational Enterprise and
International Firms," Bulletin of the Oxford Insti- the Theory of Development Policy," World Develop-
tute ofEconomics and Statistics, August 1973. ment, October 1973, pp. 8-9.
 343 PRIVATE FOREIGN INVESTMENT

16. place and party of jurisdiction and Ranking of Contracts in Order of Preference
arbitration;
17. time and right of termination or MNE Government
F
renegotiation.
A contract between the MNE and the host gov-
ernment will contain provisions under some of
these headings. Such possible contracts can be
A
B
C
C
(E)
(D)
l
Range of bargaining

I
ranked in an order of preference by the MNE and (D) B
by the government. If both the MNE and the gov- (E) A
ernment prefer a certain contract to another, the
latter can be eliminated. The only complication
here is that either party has an interest in con-
cealing the fact that its interest coincides with that
of the other party. For by appearing to make a and the government prefer C; F is ruled out be-
concession, when in fact no concession is made, it cause itis unacceptable to the MNE.
may be spared having to make a concession on an- At the same time, in determining the relative
other front where interests conflict.
value of the different contracts, the host govern-
But leaving this complication aside, amongst ment will find cost-benefit analysis useful. By com-
the contracts that remain when those dominated paring the present value of the stream of benefits
by others have been eliminated, the order of pref- with that of the costs the disparate components in
erence for the MNE will be the reverse of that for the bargain can, at least in principle, be made
the government. If the least attractive contract commensurable. Cost-benefit analysis and bar-
from the point of view of the MNE is outside the gaining-power analysis are not alternative
range of contracts acceptable to the government, methods of approach but are complementary.
no contract will be concluded. But if there is some Cost-benefit analysis will not tell a government
overlap, there is scope for bargaining. The precise whether a particular project is acceptable or not,
contract on which the two partners will settle will i.e., whether it falls within the acceptable bargain-
be determined by relative bargaining strength. ing range, but it will help it to rank those that are
E and D are ruled out because both the MNE acceptable.

Comment
For an evaluation of the role of multinational enterprises in LDCs, the following may be
consulted: Paul Streeten, "The Multinational Enterprise and the Theory of Development Pol-
icy," World Development (October 1973); Paul Streeten, "Costs and Benefits of Multina-
tional Enterprises in Less Developed Countries," in J. H. Dunning (ed.), The Multinational
Enterprise (1971); Stephen Hymer, "The Multinational Corporation and the Law of Uneven
Development," in J. Bhagwati (ed.), International Economics and World Order to the Year
2000, (1972); Stephen Hymer, "The Efficiency (Contradictions) of Multinational Corpora-
tions," American Economic Review (May 1970); O. Sunkel, "Big Business and 'Dependen-
cia,'" Foreign Affairs (April 1972); R. E. Caves, "International Corporations: The Industrial
Economics of Foreign Investment," Economica (February 1971); H. G. Johnson, "The Effi-
ciency and Welfare Implications of the International Corporation," in C. P. Kindleberger
(ed.), The International Corporation (1970); A. O. Hirschman, How to Divest in Latin Amer-
ica and Why (1969); G. K. Helleiner, "Manufactured Exports from Less Developed Countries
and Multinational Firms," Economic Journal (March 1973); Richard J. Barnet and Ronald
E. Muller, Global Reach (1974); C. Fred Bergsten et al., American Multinationals and Amer-
ican Interests (1978); Raymond Vernon, Sovereignty at Bay (1971); Raymond Vernon, Storm
Over the Multinationals (1977); E. Penrose, "'Ownership and Control': MNCs in LDCs," in
Gerald Helleiner (ed.), A World Divided (1976); Richard Newfarmer, Transnational Con-
glomerates and the Economics of Dependent Development (1979); Peter Evans, Dependent
Development (1979); Sanjaya Lall and Paul Streeten, Foreign Investment, Transnationals and
Developing Countries (1977); Carlos Diaz-Alejandro, "The Less Developed Countries and
 344 MOBILIZING FOREIGN RESOURCES

Transnational Enterprises," in Sven Grassman and Erik Lundberg (eds.), The World Eco-
nomic Order (1981); Theodore H. Moran, "Multinational Corporations and Dependency,"
International Organization (Winter 1978); Sanjaya Lall, The Multinational Corporation
(1980); Thomas G. Parry, The Multinational Enterprise: International Investment and Host-
Country Impacts (1980); Thomas J. Bierstaker, Distortion on Development: Perspectives on
the Multinational Corporation (1978); Isaiah Frank, Foreign Enterprise in Developing Coun-
tries (1980); For an excellent review article, see Paul Streeten, "Multinationals Revisited,"
Finance & Development (June 1979).

Comment
It is significant that the internationalization of LDC firms is growing relatively rapidly.
Local firms in parts of Africa, Asia, Latin America, and the Middle East are moving in in-
creasing numbers to establish manufacturing plants in other LDCs. As the LDC share of
world industrial production and exports grows, so too will their share of international invest-
ment and transfer of technology.
The emergence of multinationals from developing countries is attracting more attention;
for example, refer to the following: Dennis J. Encarnation, "The Political Economy of Indian
Joint Industrial Ventures Abroad," International Organization (Winter 1982); Sanjaya Lall,
"The Emergence of Third World Multinationals: Indian Joint Ventures Overseas," World
Development, (February 1982); S. Lall, Developing Countries as Exporters of Technology
(1982); Louis T. Wells, Jr., "The Internationalisation of Firms from Developing Countries,"
in Tamir Agmon and C. P. Kindleberger (eds.), Multinationals from Small Countries (1977);
David A. Heenan and Warren J. Keegan, "The Rise of Third World Multinationals," Har-
vard Business Review (January -February 1979); K. Kumar (ed.), Multinationals from De-
veloping Countries (1982).
 V.E. TRANSFER OF
TECHNOLOGY
V.E.1. Appropriate
Technology*
Technology is often identified with the hard- niques: on the one hand weak communication
ware of production — knowledge about ma- may mean that a particular country only
chines and processes. Here a much broader knows about part of the total methods known
definition is adopted, extending to all the to the world as a whole. This can be an im-
"skills, knowledge and procedures for mak- portant limitation on technological choice.
ing, using and doing useful things." Technol- On the other hand, methods may be known
ogy thus includes methods used in non-mar- but they may not be available because no one
keted activities as well as marketed ones. It is producing the machinery or other inputs
includes the nature and specification of what required. This too limits technological choice.
is produced — the product design — as well as The actual technology in use is thus cir-
how it is produced. It encompasses manage- cumscribed first by the nature of world tech-
rial and marketing techniques as well as nology, then by the availability to the country
techniques directly involved in production. of known techniques, and finally by the
Technology extends to services — administra- choice made among those available. If the
tion, education, banking and the law, for ex- technology in use is thought to be inappro-
ample— as well as to manufacturing and ag- priate, or because an inappropriate subset is
riculture. A complete description of the available to the country, or because an inap-
technology in use in a country would include propriate selection is made, or for some com-
the organisation of productive units in terms bination of the three reasons. Confusion is
of scale and ownership. Although much of caused by failing to distinguish between the
the discussion will be in terms of technologi- three.
cal development in the hardware of technol- Each technique is associated with a set of
ogy, the wider definition is of importance characteristics. These characteristics include
since there are relationships between the the nature of the product, the resource use —
hardware and the software— between, for ex- of machinery, skilled and unskilled man-
ample, mechanical process and managerial power, management, materials and energy
techniques and infrastructural services — inputs — the scale of production, the comple-
which help determine the choice made in mentary products and services involved, etc.
both spheres. Any or all of these characteristics may be im-
Technology consists of a series of tech- portant indetermining whether it is possible
niques. The technology available to a partic- and/or desirable to adopt a particular tech-
ular country is all those techniques it knows nique in a particular country and the impli-
about (or may with not too much difficulty cations ofso doing.
obtain knowledge about) and could acquire, More formally, we may think of all the
while the technology in use is that subset of known techniques as hT = {Ta, Tb, Tc, Td,
techniques it has acquired. It must be noted . . . Tn) (where "known" means known to the
that the technology available to a country world) as constituting world technology. For
cannot be identified with all known tech- a particular country, the technology available
for adoption is that subset of world technol-
*From Frances Stewart, Technology and Under- ogy known to the country in question and
development, 2nd ed., New York, Macmillan, Lon-
available. Say, cT = {Ta . . . Tn), where c
don and Basingstoke, 1977, pp. 1-3, and "Interna- denotes the country and the bar indicates
tional Technology Transfer: Issues and Policy
Options," World Bank Staff Working Paper No. 344,
that only techniques known to the country
July 1979, pp. 78, 82-88. Reprinted by permission. and available are included. Thus cT C hT.

345
 346 MOBILIZING FOREIGN RESOURCES

Each of the techniques 7a, Tb . . . etc. is a adapted (or "inappropriate") to the condi-
vector consisting of a set of characteristics, tions prevalent in poor countries. The trans-
ai, aii, aiii, bi, bii, biii. . . . Thus technology fer of such technology to poor societies tends,
can be described in matrix form, with each as a result, to cause various distortions and
column representing the characteristics of inefficiencies.
each technique, as shown in Table 1. In discussing "characteristics" of technol-
The technology in use in a particular coun- ogy, one should include all the relevant fea-
try is that subset of the technology available tures which determine its resource use, pro-
to it that has been selected and introduced, ductivity, and impact on production and
or uT = { Ta . . . Th) where uT C cT C wT. consumption patterns. These features include
The processes by which world technology the nature and design of the product, the
is narrowed down to an actual set of tech- scale and organizational system for which the
niques in use may be crudely described as technology is designed, its resource use, in-
shown in Figure 1. cluding capital and labor intensity, materials
The characteristics of technology are and fuel use, skill requirements, and the in-
largely determined by the nature of the econ- frastructural and complementary inputs it
omies for which they are designed. The most requires. The traditional economist's char-
significant determinants of the characteris- acterization oftechniques according to their
tics of new technology are the income levels, capital or labor intensity forms only one, and
resource availability and costs in the society quite often a relatively insignificant, aspect of
in and for which the technology is designed, the total characterization.
and the system of organization of production, Techniques designed for modern advanced
and the nature of the technology in use in the countries tend to produce high-income prod-
society. In each of these respects, societies of ucts, require high levels of investible re-
advanced countries differ from those of poor sources per employee, high levels of educa-
countries. Consequently, technology designed tion and skills, be of a large scale and require
to suit advanced countries tends to be ill sophisticated management techniques, be as-
socated with high levels of labor productivity,
and be linked, through inputs and outputs,
TABLE 1. Matrix of World Technology = wT
with the rest of the advanced technology sys-
Characteristics Ta Th Tr Td T. tem. If these techniques are transferred un-
modified toLDCs, the result will be a con-
Product type centration of resources, of savings and
Product nature
expenditure on human resources and infra-
Scale of production
Material inputs
structure, on a small part of the economy. In-
comes will tend to be concentrated in this
Labor input:
Skilled area, leading to markets for the high-income
Unskilled products the system produces. Resources
Managerial input available in the low-income country will tend
Investment requirements to be underutilized, including raw materials
as well as labor.

Availability Technology
cT:
available Selection
to the mechanisms
Communications country
 347 TRANSFER OF TECHNOLOGY

Many of the well-established characteris- ports the view that there is a wider range of
tics of the dual economy can be seen as fol- choice than the completely determinist view
lowing from the characteristics of advanced implies, other considerations suggest that the
country technology: the capital-intensity of potential for selecting labor-intensive tech-
productive techniques, the heavy reliance on niques may be exaggerated by calculations
imported managers, skill deficiencies, un- based on the micro-case studies. In the first
and under-employment and a relative (often place, many of the studies show that consid-
absolute) deprivation of the economy outside erations ofproduct standards/characteristics
the modern sector. Only economies which are may rule out the labor-intensive technologies.
growing very rapidly and are selective about Secondly, the labor-intensive techniques are
the choice of techniques and adept at modi- often only economic at small scale. Thirdly,
fying them are able to overcome this dualism, entrepreneurs do not have information about
by absorbing a growing proportion of their the complete "shelf of techniques in exis-
workforce into the modern sector. South tence; their access to information about dif-
Korea and Taiwan provide the obvious ex- ferent techniques depends on their channels
amples. Inother economies, dualistic tenden- of information. There tends to be a bias in
cies have been partially offset by a deliberate channels of information towards technology
attempt to protect the non-modern sector, currently in use in the supplying countries —
providing it with resources and protected i.e., the advanced countries. Appropriate
markets to prevent it being undermined by techniques, which are often older techniques
the modern sector. This is the policy of from advanced countries, or techniques re-
"walking on two legs" pursued most exten- cently developed in LDCs are less well
sively byChina (and to a less marked extent
by India). In China employment expansion in promoted.
Fourthly, many of the studies neglect the
the modern sector has been similar to that in determinants of choice or selection mecha-
many other developing economies, lagging nisms. Itis often assumed that the only rele-
well behind growth in output, while the tech- vant selection mechanism is the relative price
nology adopted has tended to be capital in- of capital and labor, and that is in the control
tensive. Overall employment policies have of government. In fact, the determinants of
succeeded because of the absorption of labor choice are far more complex. The nature and
in the agricultural and rural non-agricultural scale of the market is one critical determi-
sectors. nant. Products sold on the international mar-
There is a growing body of literature that ket or to high-income consumers may need to
questions the rather simplistic technological use the most recent technology in order to
determinist argument advanced above. It is compete. Because scale of production and the
argued that in many industries a wide choice nature of the market are of importance in de-
of efficient technologies has been established termining choice of product and technique,
by empirical research. Recently, Pack has ar- the factors determining these are significant.
gued that countries could make significant Income distribution and trading strategy
gains at the macro-level in terms of employ- help determine the nature and scale of mar-
ment, output and savings, by policies leading kets for different types of product. The dis-
to the adoption of the most appropriate tech- tribution ofinvestible resources between en-
niques inexistence. [See selection V.E.3.] It terprises ofdifferent size and type is also of
has long been established that there is consid- relevance. The very substantial wage differ-
erable potential for labor intensity in ancil- entials between enterprises of different sizes
lary processes, even if the core technology is suggest that the real wage level may be out-
more fixed. Old techniques from advanced come as well as (partial) cause of technolog-
countries offer more labor-intensive and ical choice. Moreover, a number of recent
small-scale alternatives than the most recent studies of technological choice have shown
techniques. While this sort of argument sup- that the most rational choices — both in terms
 348 MOBILIZING FOREIGN RESOURCES

of profit maximization and appropriateness tries. Technology may be more appropriate in

of techniques — are sometimes rejected in one respect, and no more so in others; or more
favor of less profitable and less appropriate in some respects, less in others. It is then a
techniques. Thus investigation of the decision matter of judgment as to whether it is to be
mechanisms involved is required if choice of preferred. More appropriate technology may
technique is to be altered in a more appro- be less efficient, either in the economic sense
priate direction. that it is less profitable (socially and/or pri-
Fifthly, the studies establishing a range of vately), or in the stricter sense (which has
choice are essentially static, depicting the sit- been defined as technical inefficiency) that it
uation at the point in time when they were uses more of all resources to produce the
made. But the so-called "shelf is moving as same output. If it is socially or technically in-
technical change proceeds, and some parts of efficient as compared with the "inappro-
the shelf are moving much faster than others. priate" alternative, then the economy would
In particular, with the current world balance suffer from some output loss in adopting it.
of Research and Development, the modern In such a situation it would only be worth-
capital-intensive part of the shelf is moving while adopting if it were believed that learn-
forward, in terms of new products and effi- ing effects would be such as to outweigh ini-
ciency of technique, as compared with the tial inefficiency, or that income distribution
labor-intensive part; and the "modern" end is and similar effects are such as to outweigh
also getting increasingly capital-intensive, the loss in output. If we assume that these ef-
larger scale, and the products more sophisti- fects may be incorporated into the measure
cated as incomes rise in the advanced coun- of social efficiency (and that we can give
tries. Thus while there may be scope for some some meaning to this concept), then an econ-
push in the direction of labor intensity and omy should only adopt more appropriate
other dimensions of appropriateness in some techniques if they can be shown to be socially
products, the aggregate effects are likely to efficient. In discussing ways in which appro-
be swamped by dynamic changes in the op- priate technology can be promoted, it is
posite direction. assumed that what is in question is the pro-
More appropriate technology may be nology. motion of socially efficient appropriate tech-
roughly defined as technology whose resource
use is more in keeping with LDC resource Broadly, one may distinguish two types of
availability, and whose products are more appropriate technology: appropriate technol-
suited to low-income consumers. This means ogy for the "modern" sector, which consists
that more appropriate technology will be in the adaptation of modern sector advanced
more labor-intensive, less skill intensive, country technologies in more labor-intensive
smaller scale, use more local materials, and directions; and appropriate technology for
produce simpler low-income products than the traditional sector, which upgrades and
most of advanced country technology cur- improves traditional technologies.
rently being transferred to low-income coun-
Comment
On the problem of "appropriate" technology, the following are noteworthy: the special issue
on "Technology," World Development (March 1974); special issue on "Science and Tech-
nology in Development," Journal of Development Studies (October 1972); G. D. Helleiner,
"The Role of Multinational Corporations in the Less Developed Countries' Trade in Tech-
nology," World Development (April 1975); Paul Streeten, "Technology Gaps Between Rich
and Poor Countries," Scottish Journal of Political Economy (November 1972); Jack Baran-
son, Industrial Technologies for Developing Countries (1969); Jack Baranson, Technology
and the Multinationals (1978); W. Berankek and G. Danis, Science and Technology and Eco-
nomic Development (1978); A. S. Bhalla (ed.), Technology and Employment in Industry
 349 TRANSFER OF TECHNOLOGY

(1975); A. S. Bhalla, Toward Global Action on Appropriate Technology (1979); C. Cooper

(ed.), Science, Technology and Development (1973); Frances Stewart, Technology and Un-
derdevelopment (1977); C. P. Timmer et al., The Choice of Technology in Developing Coun-
tries (1975); Charles Cooper, "Choice of Techniques and Technological Change as Problems
in Political Economy," International Social Science Journal (1973); E. K. Y. Chen, Multi-
national Corporations, Technology and Employment (1982).
A series of reports by the United Nations Institute for Training and Research (UNITAR)
is highly instructive; so too are the Industrial Planning and Programming Series of UNIDO,
as well as the reports of UNCTAD.

V.E.2. Capital-Stretching
Innovations*

First and foremost, it should be remembered with a given pair of capital and labor coeffi-
that, unlike in an advanced country where cients. The technology shelf is composed of
technological change is viewed as rather au- the complete set of such activities or technol-
tomatic and routinized or as capable of being ogies which have been demonstrated to be
generated through R&D expenditures ac- feasible somewhere in the advanced countries
cording to some rules of cost/benefit analysis, at some historical point in time, including the
in the contemporary developing societies present. Since there exists a number of tech-
technological change cannot either be taken nology-exporting countries — for example,
for granted or afforded through basic R&D the United States, Germany, the United
allocations. In this situation, we cannot avoid Kingdom, and Japan — with continuous tech-
K/Q
the question of what, given the existence of a
shelf of technology from abroad, is the pat-
tern bywhich the typical less developed econ-
omy, in fact, manages to innovate. This ques-
tion in turn forces us to look at least at the
following dimensions more carefully: (1) the
precise nature of that technology shelf, (2)
the availability within the LDCs of required
initial managerial and entrepreneurial capac-
ity, and (3) the changing nature of that re-
quired managerial and entrepreneurial ca-
pacity in the course of transition to modern
growth.
The technology shelf developed in the ma-
ture industrial economies abroad may be de-
scribed bya set of unit activities following a
smooth envelope curve as in Figure 1. A par-
ticular technology can be described by an L-
shaped contour producing one unit of output

*From Gustav Ranis, "Industrial Sector Labor Ab-

sorption," Economic Development and Cultural
Change, April 1973, pp. 392-7. Reprinted by
permission. FIGURE 1
 350 MOBILIZING FOREIGN RESOURCES

nological transfers among themselves as well import-substitution syndrome is that it is a

as with the LDCs, it is not unreasonable to response to a real (or imagined) shortage of
postulate the existence of a single technolog- entrepreneurship and that it permits time,
ical shelf for the lending world as a whole. through informal learning by doing or more
For example, unit technology A0 may have formal educational processes, for this entre-
been generated in Germany in 1920, Ax in preneurial capacity to develop.
the United States in 1920, and A2 in the In terms of Figure 1, this means that, al-
United States in 1950. In other words, as we though the technological shelf may look as
move to the left along the shelf we run into indicated by curve SS, the actual choices
more "modern" technology, that is, technol- available to the developing country during
ogy of more recent vintage and of higher cap- the import-substitution phase are more aptly
ital intensity. As capital per head increases described by S'S'. In other words, due to the
this means that the typical worker has inadequate state of entrepreneurial capacity
learned to cooperate with more units of cap- during the early postindependence period of
ital of increasing technical complexity. This physical controls, the efficiency of generating
capital-deepening process, in other words, is output per unit of capital in the borrowing
more complicated than the textbook version country is likely to be substantially below
of "homogeneous" labor being equipped with that in the lending country. This is likely to
more units of "homogeneous" capital. be more true the more capital intensive the
At any point in time the typical LDC is, import, that is, the further removed from the
then, theoretically free to borrow a particular cultural inheritance and experience of the
unit activity from anywhere along this shelf. borrower. Such technological imports are
What technology is chosen and what happens often accompanied by imported engineers,
as an immediate and ultimate consequence of even managers and supervisors — adding up
that choice, that is, what secondary processes to what is called a "turn-key project." The
and reactions are set off, is, of course, all part most advanced and sophisticated technology
and parcel of the innovation process taken as
can, of course, be made to "work," in the
a whole. The quality of the process, each step physical sense, even in the most backward de-
of the way, in turn depends on both the eco- veloping economy. But a shiny new plant
nomic environment, that is, the nature of the imbedded in a society many decades distant
relative price signals, and on the entrepre- is bound to be substantially less efficient. This
neurial, managerial, and skilled labor capac- is true for a thousand direct reasons, such as
ity of the borrower. the absence of even minimal skilled labor
The role of innovation must therefore be supplies, domestic subcontracting, and repair
seen as intimately related to the stage in and maintenance possibilities, as well as for
which the developing economy finds itself. In many more subtle sociological reasons which
other words, the role of technological change enter into the total milieu in which the plant
in output and employment generation must is required to operate. The more sophisti-
be viewed as sensitive to the same discernible cated and removed from the rest of the econ-
phases of growth as the economy moves in omy is the technological transplant, the
transition from open agrarianism to Kuznets' greater is the relative inefficiency, as indi-
modern economic growth. In the first post-
cated bywhen
If and the shape of the S"S"
the economy thencurve.
moves away
independence or import-substitution phase
previously mentioned, an effort is made to in- from the import-substitution phase and en-
crease the supply of domestic entrepreneur- ters into the second phase of liberalization
ship and the economy's learning capacity, and export substitution, a second important,
partly through the importation of people via if unintentional, type of innovation is likely to
aid but mainly through the system of protec- appear, namely, a reduction in the extent of
tion established by government policies. In the inefficiency of the original transplanted
fact, the most reasonable explanation for the technology. Call it X efficiency if you like,
 351 TRANSFER OF TECHNOLOGY

but the cost of the pure transplantation is choices are likely to continue to be severely
likely to be reduced, quite unintentionally, constrained — partly by a lack of illumination
that is, largely as a result of factors external of substantial portions of it, partly by such
to the profit-maximizing behavior of the pro- institutional inhibitions as prestige, aid tying,
ductive unit itself. This increase in productive and so forth — we can realistically expect rel-
efficiency over time will increase in quanti- atively less benefits from liberalization to ac-
tative significance as the import-substitution crue in the transplantation process. On the
hothouse temperature is gradually turned other hand, we can expect much more from
down and a more competitive efficiency may the assimilation type of innovational behavior
be represented by the arrows tending, over which now tends, for the same reasons, to be
time, to move S'S' back toward the original more slanted in a labor-using direction. In
SS position. the typical labor surplus type of economy —
Another more conscious and quantitatively for one likely to become one over the next
more important type of innovation begins to decade (as is probably the case in much of
gather importance during this same second Africa) — all this means a much greater pos-
phase of transition. This phenomenon may be sibility for the efficient accomodation of pure
called innovational assimilation — innovating labor services.1 Whether this will lead to a
"on top of" imported technology in the direc- sectoral output shift in favor of labor-inten-
tion of using relatively more of the abundant sive export commodities or a mix change pre-
unskilled labor supply. As the economy shifts dominantly addressed to the domestic mar-
from a natural-resource-based growth pat- ket, of course, depends on, ceteris paribus,
tern in the import-substitution phase to a income elasticities of demand, the gover-
human-resource-based system in the export- ment's fiscal prowess, and the type (e.g., size)
substitution phase, there is an increasing sen- of the economy. Moreover, no strong gener-
sitivity to the continuously changing factor alization as to the relative importance of
endowment, first in terms of the efficient uti- shifts in output mix versus changes in tech-
lization ofthe domestic unskilled-labor force, nology for given mixes is likely to be valid. It
and later in terms of the incorporation of should be clear, however, that the important
growing domestic skills and ingenuity. In issue is that the search for innovation can
other words, the appropriate type of technol- now be considered a conscious activity of the
ogy finally in place must be one in which not individual entrepreneur and — given the com-
only the initial choice from the shelf but also bination ofmore realistic relative price sig-
the adaptations and adjustments consciously nals after liberalization plus greater entre-
made thereafter, in response to changing do- preneurial capacity — that it is likely to be
mestic resource and capability constraints, mainly directed toward various forms of in-
play an important role. digenous capital stretching on top of the im-
The more liberalized the economy, in ported technology. Such capital stretching
terms of the government's performing a cat- can be represented by a reduction in the cap-
alytic role through the market by indirect ital coefficient per unit of output. The effec-
means rather than trying to impose resource tive post-assimilation set of unit technologies,
allocation by direct controls, the better the that is after domestic assimilation, may thus
chances that the millions of dispersed deci- be represented by curve S"S" in Figure 1,
sion makers can be induced, by the sheer with the strength of the indigenous labor-
force of profit maximization, to make the using innovative effort indicated by the
"right" decision. As the gap between shadow
and market prices narrows — coupled with
the expectation of continued labor surpluses 'It is important to emphasize the word "efficient"
since I am not concerned here with the, possibly also
for years to come — we would expect trans- legitimate, objective of employment creation as a sep-
plantation choices to become more flexible, arate social goal, to be weighed against output
that is, labor intensive. However, since shelf
growth.
 352 MOBILIZING FOREIGN RESOURCES

amount of the "downward" shift in the cap- In summary, once the overall policy setting
ital coefficient. has turned more favorable and permitted the
It should be noted here that a negatively economy to enter the second phase of transi-
sloped technology shelf, for example, SS, tion, it is this indigenous capital-stretching
representing pure technological transplanta- capacity which I consider to be of the great-
tion, permits, as we move to the left, higher est importance — especially for the contem-
labor productivity levels, but only at increas- porary developing economy facing the for-
ing capital cost. In a country characterized midable labor force explosion predicted for
by capital scarcity this may mean increased the seventies and eighties. It is in this specific
technical unemployment (a la Eckaus) and area also where the skepticism of planners,
hence a lower value of per capita income for engineers, and aid officials generally is most
the economy — in spite of the higher level of pronounced — especially with respect to the
labor productivity achieved. Domestic capital full range of technological choice really
stretching, however, can materially affect available when all the dust has settled. His-
that situation by enabling more workers to be torical examples from the Japanese case, as
employed per unit of the capital stock. If the well as contemporary evidence from Korea
post-assimilation unit technology set, 5"'»S", and Taiwan, permit us to demonstrate the ex-
for example, is upward sloping from left to istence and potential importance of such cap-
right, higher labor productivity levels become ital-stretching innovations for the labor sur-
consistent with lower capital/output ratios. plus developing country.

V.E.3. Factor Intensity*

Though it is not our purpose to survey the lit- observation of both DC and LDC factory
erature on choice of techniques, a summary operations and engineering specifications.
of some of the conclusions we have reached 2. Evidence is also accumulating that the
on the basis of recent literature is germane at core production process itself, whether the
this point. cooking of food or the production of woven
1 . In a large number of industrial activities cotton cloth, offers efficient possibilities
for using less expensive equipment and
a product is producible with a considera-
ble range of alternate ratios of capital to more labor per unit of output of a given
labor. Much of the potential substitution quality, i.e., the more labor-intensive pro-
of labor for capital stems from use of cesses are not relegated to the production
of inferior commodities. Adaptation of ex-
labor-intensive methods in "peripheral" isting equipment, for example, changing
production activities; labor, with little if
any capital, can be used to transport ma- the "normal" speed of operation, offers
terial within the factory, to pack cartons, still further opportunities to save capital
and to store the final product. The evi- and increase the relative use of labor. Fi-
dence for these statements is drawn from nally, extensive underutilization of indus-
trial capacity provides considerable scope
for increasing the effective labor-capital
♦Howard Pack, "Technology and Employment: ratio.
Constraints on Optimal Performance," Reprinted by
permission of Westview Press from Technology and
3. There are some industries that probably
Economic Development: A Realistic Perspective, ed- offer limited possibilities for altering the
ited by Samuel M. Rosenblatt. Copyright © 1979 by capital-labor ratio compared with that
the International Economic Studies Institute.
prevailing in advanced economies. These
 353 TRANSFER OF TECHNOLOGY

are typically activities in which most be noted that all of these studies consider
LDCs have no comparative advantage substitution among plants of efficient size, in
and where the basic problem is to forestall which no further scale economies are to be
their introduction (typically behind tariff realized. The range of choice presented is al-
walls) rather than to suggest methods of most surely to be more circumscribed than
changing production in a more labor-in- the options to be derived from a still more
tensive direction. Even these classical no- systematic search that would include ma-
technical-substitution cases offer some chinery produced in the LDCs themselves in
choice, principally in the price of equip- the set of available techniques. For example,
ment which typically varies among the the engineering data for the textile study do
sources of supply. not take account of the additional labor in-
tensity afforded by semiautomatic looms pro-
To establish some orders of magnitude a duced in Korea. These are no longer pro-
set of numerical results derived from a num- duced in the industrialized countries but are
ber of recent studies are shown in Table 1.
currently in wide use in Korean textile pro-
These figures measure the ratio of fixed cap- duction, including that destined for export.
ital per worker necessary to begin a new pro- Their inclusion would significantly widen the
duction operation using varying existing efficient range of choice in weaving.
technologies. The underlying engineering If an LDC is planning the expansion of its
analyses have been carried out by teams of industrial sector, its choices with respect to
engineers and economists. Variations in the capital intensity are certainly not narrow.
capital-labor ratio reflect several factors: (1) There need be no trade-off between output
different engineering principles used in the (in this case value added originating in man-
core process; (2) varying sources of supply of ufacturing) and employment; as long as plen-
the same equipment as some firms or coun- tiful unskilled labor can be substituted for
tries exhibit lower prices fairly consistently; scarce capital both employment and output
(3) differences in the degree of mechaniza- increase. The political authority has two di-
tion of peripheral activities such as the bag- mensions ofchoice in view of the substitution
ging of fertilizer; and (4) minor adaptations possibilities. By providing a reasonably com-
in existing plants reflecting learning accu- petitive atmosphere in product markets, par-
mulated during the production process. ticularly liberal
a international trade regime,
Thus, some of the most recent work sug- it will encourage the expansion of sectors
gests a rather substantial range of technical which are not capital-intensive. The critical
choice in modern manufacturing. It should impotance of correct sector choice can read-
TABLE 1 . Fixed Investment per Job Created (dollars)

Alternative Technolog ies

Capacity per Year 1 2 3 4 5

Sugar refining 50,000 tons

2,869 3,480 3,869
Maize milling 36,000 tons 5,987 6,204
Urea 2,917 5,465
126,302 7,081
134,800 9,730
137,631
528,000 tons 123,789
Cotton yarn 1 ton
2,942 16,434
Woven cotton cloth 1 million sq. yards 33,063
1,508 8,125 1J,870
Beer brewing 200,000 hectolitres 11,419 15,000
760
Men's leather shoes 300,000 pair
1,042 2,157
Source: For cotton yarn and woven cloth: H. Pack, "The Optimality of Used Equipment: Some Computations for the Cotton
Textile Industry," Economic Development and Cultural Change, January 1978; all other studies at the University of Strath-
clyde; summaries appear in World Development, October 1977.
 354 MOBILIZING FOREIGN RESOURCES

ily be seen in Table 1 — compare the capital- inappropriate production methods. At least
labor ratio of the least capital-intensive fer- two studies explicitly addressing the question
tilizer plant with that of the most capital-in- of whether firms, given their realtive factor
tensive shoe or cloth producing technology. If prices, have in fact chosen the cost minimiz-
policies are pursued, whether by direct gov- ing techniques or a more capital-intensive
ernment investment or by allowing distorted one, have suggested that the latter is the case.
factor and product prices to prevail, which
systematically encourage the expansion of
sectors whose lowest achievable capital cost OBSTACLES TO CORRECT CHOICE
per job created is $124,000 only a limited OF FACTOR PROPORTIONS
contribution to employment growth can be
expected from that sector. Though careful at- Assuming that private and public firms
tention todesign can reduce the cost per job often choose an inappropriate technology
from $138,000 to $124,000, the latter re- where alternatives are available, can such be-
mains considerably in excess of the capital havior be explained? The most obvious, and
available to equip new labor force entrants. until recently the most widely emphasized,
The second choice facing the government, explanation was that of incorrect factor mar-
assuming that reasonable balance in sectoral ket prices facing producers; the cost of un-
expansion is achieved, is, of course, appropri- skilled labor being above and that of capital
ate choice of technique within a given sector. below their respective opportunity costs.
Interestingly enough, the amount of direct While these are undoubtedly important ele-
evidence demonstrating inappropriate choice ments, they are unlikely to constitute the sole
of manufacturing technology is remarkably explanation.
small. Though anecdotes abound about au- Consider a firm undertaking an entirely
tomated factories employing very few work- new venture or expanding an existing one.
ers, little is known about the relative impor- The profit augmenting returns to a search for
tance of very modern equipment in a typical appropriate plant and equipment must be
LDC's annual investment. The conclusion weighed against the increase in profits to be
that inappropriate technology is widely uti- obtained by alternative uses of entrepreneu-
lized is based primarily on the discrepancy rial and technician time. Once the scarcity of
between measured growth rates of real man- both management and staff time (in the
ufacturing output and employment. This dis- larger firms) is taken into account, it is ob-
crepancy may, however, reflect increasing vious that many activities exist to which an
capital per worker or greater labor productiv- allocation of time may be profitable. Better
ity with the same capital per worker, or some inventory control; improved floor supervision;
combination. We assume that at least some searching for lower cost suppliers; identifica-
improvement in the employment-generating tion of new markets; coaxing better treat-
effects of new investment is possible; in the ment from government in the form of higher
main, this assumption is based, not on infer- protection, rebates, or tariffs on inputs; and
ences from the growth of average labor pro- exploiting the entire panoply of government
ductivity inmany LDCs, but on the level of
measures may be better uses of one's time
exports of new capital goods from DCs to than choosing "the" correct technique,
LDCs and the absence of any significant though it too could improve profits.
intra-LDC trade in capital goods, and the These possibilities suggest that, all other
lack of substantial production except in a factors being equal (the country, sector,
handful of countries. product mix, and relative factor prices), the
In addition, some of the microeconomic firms facing lower costs of acquiring the in-
work on choice of technique, particularly that formation needed for implementing a cost-
comparing local and foreign companies in the minimizing, labor-intensive technology are
same productive activity, has indicated some more likely to take that step than companies
 355 TRANSFER OF TECHNOLOGY

for whom the (explicit and implicit) infor- of origin, since this is the only technology
mation costs are very high. The adoption of a with which they are familiar, i.e., the search
readily available capital-intensive technology costs are less. However, this begs the question
may well be the profit-maximizing strategy of why these otherwise profit-maximizing en-
for the latter. It is plausible to conjecture tities (clever enough to set transfer prices to
that MNCs have lower search costs than maximize global net profits) engage in no
purely local companies involved in the same profitability calculus on the benefits and costs
activity. They can easily identify and transfer of adaptation despite the widely differing fac-
equipment among subsidiaries, especially tor prices prevailing in the home country and
items that have become too expensive to uti- the LDC.
lize in higher wage countries because of their It is often overlooked that an MNC which
labor intensity. Indeed, the parent company has been in operation for some length of time
may have established a new plant partly to is likely to have various vintages of equip-
utilize such equipment in the production of ment in its worldwide plants. As wages rise
exports. Alternately, the local manager may in the richer of the countries where it oper-
ask the parent company's purchasing office ates, the quasi-rents to be earned on older
for advice on the availability of used equip- equipment will decline. Rather than sell such
ment ifit is not available within the country. equipment in the used equipment market in
Given the well known difficulties of identify- the DCs, firms can earn higher returns by
ing reliable equipment such "costless" aid to moving it to a lower wage location. It is con-
the local manager (in both time and explicit siderably easier for an MNC than for a local
costs) clearly increases the likelihood of his LDC company to identify this equipment and
employing appropriate equipment in order to purchase it directly from the vendor in the
take advantage of the factor prices facing DC. Intracompany transactions in the face of
him. Not only can a purchasing office iden- great uncertainty about the quality of the
tify and physically evaluate the condition of equipment are likely to result in more move-
the equipment, it also probably can obtain a ment of equipment than in a market in which
better price, insofar as most studies of the one important attribute of the commodity, its
used equipment market indicate that price quality, is very difficult to determine.
setting in this market is one of bilateral mo- Thus, the possession of a considerable
nopoly rather than perfect competition. range of vintages within its international op-
Thus, if we observe the production of sim- erations may permit adaptive MNC behavior
ilar products in a country by both local firms with respect to technology. Clearly this can-
and MNCs the preceding argument suggests not be the case when the underlying produc-
that MNCs would utilize a more appropriate tion process offers virtually no substitution
technology, given the relative factor prices opportunities, for example, oil refining and
facing the two types of firms. Put slightly dif- other continuous process industries. Though
ferently, given the substitution opportunities the multinationals do utilize older vintage
in production, the MNC is more likely to ap- equipment, there is no evidence to suggest
proximate itscost-minimizing factor propor- that they engage in research specifically de-
tions than would a purely local firm. signed to modify production techniques to
This view, of course, contrasts strongly better suit LDC factor availabilities. How-
with the often polemical work that asserts, ever, because most MNC production is in
usually without evidence, that multinationals traditional branches such as food processing
use excessively capital-intensive methods. In- and hand soap, the use of traditional meth-
sofar as the logic of the argument can be sep- ods, twenty years out of date in the advanced
arated from the hyperbole of the typical dis- countries, is almost surely the cost-minimiz-
cus ion, iappears
t that the major assumption ing strategy rather than the development of
is that MNCs utilize the same production newer, labor-intensive techniques exhibiting
process in the host country as in the country improved productivity.
 356 MOBILIZING FOREIGN RESOURCES

Comment
We may raise three basic questions regarding the choice of technique by multinational
enterprises: (1) whether the technologies used by multinationals are adaptable to the low-
wage labor-abundant conditions in LDCs, (2) whether multinationals do in fact adapt the
technologies they transfer, and (3) whether multinationals adapt better or worse than local
firms. An excellent review of the literature addressed to these questions is presented by S.
Lall, "Transnational, Domestic Enterprises, and Industrial Structure in Host LDCs: A Sur-
vey," Oxford Economic Papers (July 1978), pp. 217-21, with an extensive bibliography.
Other significant references on this subject include: R. H. Mason, "Some Observations on
the Choice of Technology by Multinational Firms in Developing Countries," Review of Eco-
nomics and Statistics (1973), pp. 349-55; Samuel Morely and Gordon Smith, "Limited
Search and the Technological Choices of Multinational Firms in Brazil," Quarterly Journal
of Economics (February 1976); William H. Courtney and Danny M. Leipziger, "Multina-
tional Corporations in LDCs: The Choice of Technology," Oxford Bulletin of Economics and
Statistics (November 1975); Danny M. Leipziger, "Production Characteristics in Foreign
Enclave and Domestic Manufacturing: The Case of India," World Development (April 1976).
For additional readings on the micro-substitution possibilities in different industries in de-
veloping countries, see S. N. Acharya, "Fiscal/Financial Intervention Factor Prices and Fac-
tor Proportions: A Review of Issues," Bangladesh Development Studies (October 1975); H.
Pack, "The Substitution of Labor for Capital in Kenyan Manufacturing," Economic Journal
(March 1976); Y. Rhee and L. Westphal, "A Microeconomic Investigation of Choice of
Technology," Journal of Development Economics (September 1977); G. Ranis, "Industrial
Technology Choice and Employment: A Review of Developing Country Evidence," Inter-
sciencia (1977); L. J. White, "The Evidence on Appropriate Factor Proportions for Manu-
facturing inLess Developed Countries: A Survey," Economic Development and Cultural
Change (October 1978); Samuel M. Rosenblatt (ed.), Technology and Economic Develop-
ment: A Realistic Perspective (1979); A. S. Bhalla (ed.), Technology and Employment in
Industry, 2nd ed. (1980).
Although these studies indicate a wide potential choice in both primary and secondary pro-
duction operations and alternative commodity specifications, the empirical fact remains that
the selections actually made in the developing countries still appear to be "inappropriate" by
the standards discussed in the foregoing readings. It is therefore necessary to analyze why the
problem of inappropriate technology persists and what policy measures might be undertaken
to overcome the obstacles to the use of appropriate technology.
CHAPTER VI

Industrialization
Strategy

This chapter should be read in conjunction with the next one on agricultural strategy. For a
development program cannot afford to emphasize industrialization at the expense of agricul-
tural development. Although the initial development plans of many LDCs focused on pro-
grams of deliberate industrialization, the role of industrialization is now being reappraised.
It is not a question of concentrating resources on industry or on agriculture — as alternatives.
Rather, it is increasingly realized that the mutually supportive interactions between agricul-
ture and industry should receive prime attention.
Instead of industrialization via import substitution, which involved an output mix and
choice of techniques that conflicted with other development objectives, there is now the ad-
vocacy ofa simpler, more appropriate type of industrialization. As Streeten states,
The disenchantment with industrialization in recent writings and speeches has been based on a confusion:
it is a disenchantment with the form that economic growth had taken in some developing countries and
with the distribution of its benefits. . . . After a reorientation of goals, industrialization as the servant of
development regains its proper place in the strategy. Industry should produce the simple producer and
consumer goods required by the people, the majority of whom live in the countryside; hoes and simple
power tillers and bicycles, not air conditioners and expensive cars and equipment for luxury flats. . . .
In simple mass consumption goods, often produced in a labour-intensive, capital saving way, the de-
veloping countries have a comparative advantage and could expand their trade among themselves. But
all this depends upon countries opting for a style of development that gives priority to satisfying the
simple needs of the large number of poor people. Industries producing clothing, food, furniture, simple
household goods, electronics, buses and electric fans would thrive without the need for heavy protection
in a society that had adopted this style of industrialization and development. Much of the recent criticism
of inefficient, high-cost industrialization behind high walls of protection and quantitative restrictions
should be directed at the types of product and of technique which cater for a highly unequal income
distribution and reflect entrenched vested interests. It is in no way a criticism of industrialization for the
needs of the people.1
'Paul Streeten, "Industrialization in a Unified Development Strategy," World Development, January 1975,
p. 3.

357
 358 INDUSTRIALIZATION STRATEGY

The future of industrialization lies in a shift from industrialisation via import substitution
to a different pattern of domestic output and industrialization via export substitution. As a
corollary of this shift, the old dispute about whether to give priority to industry or agriculture
is not a real issue. The question instead is how to achieve concurrently both agricultural and
industrial development.
Section VIA. indicates that industrialization offers substantial dynamic benefits that are
important for changing the traditional structure of the less developed economy, and the ad-
vocacy ofindustrialization may be particularly compelling for primary export countries that
confront problems of a lagging export demand while having to provide employment for a
rapidly increasing labor force. Systematic support is given to the industrialization argument
in Rosenstein-Rodan's emphasis on the external economies to be realized through industrial-
ization, and his advocacy of a "big push" in the form of a high minimum amount of industrial
investment in order to jump over the economic obstacles to development. Although the "bal-
anced growth" doctrine is sometimes related to the big push theory, we should note that the
principle of balanced growth is not a necessary component of the big push theory, and bal-
anced growth need not be dependent upon a large amount of public investment or dominance
of the public sector.
Selections VI.A.4 and VI.A.5 also summarize the much-discussed "balanced growth" ver-
sus "unbalanced growth" approaches to investment. By emphasizing that investment deci-
sions are mutually reinforcing and that overall supply "creates its own demand," the balanced
growth doctrine has considerable appeal as a means of initiating development. Critics of the
doctrine, however, argue that a poor country does not have the capacity to attain balanced
investment over a wide range of industries and that, moreover, the method of balanced growth
cannot bring about as high a rate of development as can unbalanced growth. Instead of striv-
ing for balanced investment, proponents of unbalanced growth advocate the creation of stra-
tegic imbalances that will set up stimuli and pressures which are needed to induce investment
decisions. As expressed by Hirschman, for instance, "our aim must be to keep alive rather
than eliminate the disequilibria of which profits and losses are symptoms in a competitive
economy. If the economy is to be kept moving ahead, the task of development policy is to
maintain tensions, disproportions, and disequilibria."2
According to this view, the central task of a development strategy is to overcome the lack
of decision-taking actions in the economy; for this purpose, unbalanced growth is necessary
to induce investment decisions and thereby economize on the less developed economy's prin-
cipal scarce resource, namely, genuine decision making.
It has now become clear that the phrases "balanced growth" and "unbalanced growth"
initially caught on too readily and that each approach has been overdrawn. After much re-
consideration, each approach has become so highly qualified that the controversy is now es-
sentially barren. Instead of seeking to generalize either approach, we should more appropri-
ately look to the conditions under which each can claim some validity. It may be concluded
that while a newly developing country should aim at balance as an investment criterion, this
objective will be attained only by initially following in most cases a policy of unbalanced
investment. In operational terms, the crucial question has become how to determine what is
the proper sequence of investment decisions in order to create the proper amount of imbalance
in the right activities.
Section VLB. focuses on recent experience with industrialization strategies that have been
based on import substitution (VI.B.2) and on export substitution (VLB. 3). Policies that have
encouraged industrialization through import controls and tariff protection have at the same
time adversely taxed agricultural production and exports. To redress the previous bias in favor

2Albert O.Hirschman, The Strategy of Economic Development, New Haven, 1958, p. 66.
 359 INDUSTRIALIZATION STRATEGY

of import substitution, emphasis is now being given to the positive relationships between ex-
ports and industrialization. Several reasons are offered in section VLB for the superiority of
industrialization via export substitution. (Chapter VIII will offer a more detailed analysis of
export-led development.)
Finally, the relationships between agriculture and industry are outlined in section VI.C.
The view that agriculture can serve as a resource reservoir for industry is examined, but the
shift in focus now seems to be on the mutual interactions between industry and agriculture.
Without agricultural development, as formulated in the next chapter, there cannot be the
realization of a positive industrial program, as outlined in this chapter.
 VI.A. PATHS TO
INDUSTRIALIZATION

VI.A.1 . The Theory of the "Big

Push"*
An institutional framework different from taken apart from the costs of the 50% partic-
the present one is clearly necessary for the ipation inits own "Eastern European Indus-
successful carrying out of industrialisation in trial Trust" that we shall propose. It should
international depressed areas. In what fol- be counted as capital investment in the East-
lows arguments are submitted tending to ern European Industrial Trust (E.E.I.T.).
show why the whole of the industry to be cre- That is not, however, the most important
ated is to be treated and planned like one reason in favour of such a large investment
huge firm or trust. unit.
The first task of industrialisation is to pro- Complementarity of different industries
vide for training and "skilling" of labour provides the most important set of arguments
which is to transform Eastern European in favor of a large-scale planned industrialis-
peasants into full-time or part-time industrial ation. In order to illustrate the issues in-
workers. The automatism of laissez-faire volved, let us adopt the somewhat round-
never worked properly in that field. It broke about method of analysing two examples. Let
down because it is not profitable for a private us assume that 20,000 unemployed workers
entrepreneur to invest in training labour. in Eastern and South-Eastern Europe are
There are no mortgages on workers — an en- taken from the land and put into a large shoe
trepreneur who invests in training workers factory. They receive wages substantially
may lose capital if these workers contract higher than their previous meagre income in
with another firm. Although not a good in- natura. It would be impossible to put them
vestment for a private firm, it is the best in- into industry at their previous income stan-
vestment for the State. It is also a good dard, because they need more foodstuffs than
investment for the bulk of industries to be they had in their agrarian semi-unemployed
created when taken as a whole, although it existence, because these foodstuffs have to be
may represent irrecoverable costs for a transported to towns, and because the work-
smaller unit. It constitutes an important in- ers have to pay for housing accommodation.
stance of the Pigovian divergence between If these workers spent all their wages on
"private and social marginal net product" shoes, a market for the products of their en-
where the latter is greater than the former. terprise would arise representing an expan-
Training facilities (including transport and sion which does not disturb the pre-existing
housing) of one million workers per annum market, and 90% of the problem (assuming
would involve costs of certainly more than £ 10% profit) would be solved. The trouble is
100 million per annum — a sum which may that the workers will not spend all their
be too great to be borne by the State (or the wages on shoes. If, instead, one million un-
Eastern European national economy) if employed workers were taken from the land
and put, not into one industry, but into a
*From Paul N. Rosenstein-Rodan, "Problems of
Industrialization of Eastern and South-Eastern Eu-
whole series of industries which produce the
bulk of the goods on which the workers would
rope," Economic Journal, June-September 1943, pp.
204-7; "Notes on the Theory of the 'Big Push,"' in spend their wages, what was not true in the
Howard S. Ellis (ed.), Economic Development for case of one shoe factory would become true
Latin America, Macmillan and Co. Ltd., London; St. in the case of a whole system of industries: it
Martin's Press, New York, 1961, pp. 57-8, 60-62, would create its own additional market, thus
65-6. Reprinted by permission.

361
 362 INDUSTRIALIZATION STRATEGY

realising an expansion of world output with market is bound to be insufficient in this case
the minimum disturbance of the world mar- because he cannot have all the data that
kets. The industries producing the bulk of the would be available to the planning board of
wage goods can therefore be said to be com- an E.E.I.T. His subjective risk estimate is
plementary. The planned creation of such a bound to be considerably higher than the ob-
complementary system reduces the risk of jective risk. If the industrialisation of inter-
not being able to sell, and, since risk can be national depressed areas were to rely entirely
considered as cost, it reduces costs. It is in on the normal incentive of private entrepre-
this sense a special case of "external neurs, the process would not only be very
economies." much slower, the rate of investment smaller
It may be added that, while in the highly and (consequently) the national income
developed and rich countries with their more lower, but the whole economic structure of
variegated needs it is difficult to assess the the region would be different. Investment
prospective demand of the population, it is would be distributed in different proportions
not as difficult to foresee on what the for- between different industries, the final equilib-
merly unemployed workers would spend their rium would be below the optimum which a
wages in regions where a low standard of liv- large E.E.I.T could achieve. In the interna-
ing obtains. tional capital market the existing institutions
Two other types of "external economies" are mostly used to invest in, or to grant credit
will arise when a system of different indus- to, single enterprises. It might easily happen
tries iscreated. First, the strictly Marshallian that any one enterprise would not be profit-
economies external to a firm within a grow- able enough to guarantee payment of suffi-
ing industry. The same applies, however, cient interest or dividends out of its own prof-
(secondly), to economies external to one in- its. But the creation of such an enterprise,
dustry due to the growth of other industries. e.g., production of electric power, may create
It is usually tacitly assumed that the diver- new investment opportunities and profits
gence between the "private and social mar- elsewhere, e.g., in an electrical equipment in-
ginal net product" is not very considerable. dustry. Ifwe create a sufficiently large in-
This assumption may be too optimistic even vestment unit by including all the new indus-
in the case of a crystallised mature competi- tries of the region, external economies will
tive economy. It is certainly not true in the become internal profits out of which divi-
case of fundamental structural changes in dends may be paid easily.
the international depressed areas. External Professor Allyn Young's celebrated exam-
economies may there be of the same order of ple elucidates our problem. He assumed that
magnitude as profits which appear on the a Tube line was to be built in a district and
profit and loss account of the enterprise. that an accurate estimate was made of costs
The existing institutions of international and receipts. It was found that the rate of
and national investment do not take advan- profit would be below the usual rate of yield
tage of external economies. There is no incen- on investments obtainable elsewhere. The
tive within their framework for many invest- project was found not profitable and was
ments which are profitable in terms of "social abandoned. Another enterprising company
marginal net product," but do not appear bought up the land and houses along the pro-
profitable in terms of "private marginal net posed Tube line and was then able to build
product." The main driving-force of invest- the line. Although the receipts from the pas-
ment isthe profit expectation of an individual senger traffic would not pay a sufficient rate
entrepreneur which is based on experience of of profit, the capital appreciation on the
the past. Experience of the past is partly ir- houses and land more than made up the de-
relevant, however, where the whole economic ficiency. Thus the project was realised; the
structure of a region is to be changed. An in- Tube line was built. The problem is: Is it de-
dividual entrepreneur's knowledge of the sirable— i.e., does it lend to an optimum al-
 363 PATHS TO INDUSTRIALIZATION

location of resources and maximisation of na- vestment. Moreover, the allocation of invest-
tional income — that this form of capital gain ment— unlike the allocation of given stocks
(external economy) be included as an item in of consumer goods (equilibrium of consump-
the calculus of profitability, or is it not? tion), or of producers' goods (equilibrium of
Allyn Young hints that it is not desirable be- production) — necessarily occurs in an imper-
cause the capital appreciation of houses and fect market, that is, a market on which prices
land along the Tube line due to an influx of do not signal all the information required for
people from other districts has an uncompen- an optimum solution.2 Given an imperfect in-
sated counterpart in a capital depreciation of vestment market, pecuniary external econo-
houses and land in districts out of which peo- mies have the same effect in the theory of
ple moved into the Tube-line district. Agri- growth as technological external economies.
cultural land in Eastern and South-Eastern They are a cause of a possible divergence be-
Europe will, however, not depreciate when tween the private and the social marginal net
the agrarian excess of population moves out. product. Since pecuniary, unlike technologi-
In this case external economies should be in- cal, external economies are all-pervading and
cluded inthe calculus of profitability. frequent, the price mechanism does not nec-
essarily put the economy on an optimum
"There is a minimum level of resources path. Therefore, additional signalling devices
I that must be devoted to ... a development apart from market prices are required. Many
II program if it is to have any chance of success. economists, including the author, believe that
Launching a country into self-sustaining these additional signals can be provided by
growth is a little like getting an airplane off programming. Third, in addition to the risk
the ground. There is a critical ground speed phenomena and imperfections characterizing
which must be passed before the craft can be- the investment equilibrium, markets in
come airborne. . . ."' Proceeding "bit by bit" under-developed countries are even more im-
will not add up in its effects to the sum total perfect than in developed countries. The
of the single bits. A minimum quantum of in- price mechanism in such imperfect markets
l vestment is a necessary, though not sufficient, does not provide the signals which guide a
I, condition of success. This, in a nutshell, is the perfectly competitive economy towards an
i contention of the theory of the big push. optimum position. . . .
This theory seems to contradict the conclu- Indivisibilities of inputs, processes, or out-
)\ sions of the traditional static equilibrium the- puts give rise to increasing returns, that is,
I ory and to reverse its famous motto natura economies of scale, and may require a high
nonfacit saltum. It does so for three reasons. optimum size of a firm. This is not a very im-
; First, it is based on a set of more realistic as- portant obstacle to development since with
I sumptions of certain indivisibilities and "non- some exceptions (for instance in Central
appropriabilities" in the production functions America) there is usually sufficient demand,
even on the level of static equilibrium theory. even in small, poor countries, for at least one
These indivisibilities give rise to increasing optimum scale firm in many industries.
returns and to technological external econo- There may be room, however, only for one or
mies. Second, in dealing with problems of a few firms with the obvious danger of mo-
growth this theory examines the path to- nopolistic markets.
wards equilibrium, not the conditions at a As Allyn Young pointed out, increasing re-
point of equilibrium only. At a point of static turns accrue to a firm not only with the
equilibrium net investment is zero. The the- growth of its size but also with the growth of
ory of growth is very largely a theory of in-
2See P. N. Rosenstein-Rodan, "Programming in
'Massachusetts Institute of Technology, Center for Theory and in Italian Practice," in Massachusetts In-
International Studies, The Objectives of United stitute ofTechnology, Center for International Stud-
States Economic Assistance Programs, Washington, ies, Investment Criteria and Economic Growth, Cam-
D.C., 1957, p. 70. bridge, Mass., 1955.
 364 INDUSTRIALIZATION STRATEGY

the industry and with the growth of the in- doubtedly required in this lumpy field. Nor-
dustrial system as a whole. Greater speciali- mal market mechanisms will not provide an
zation and better use of resources become optimum supply.
possible when growth helps to overcome in- Social overhead capital is characterized by
divisibilities generating pecuniary external four indivisibilities. First, it is indivisible (ir-
economies. The range of increasing returns reversible) intime. It must precede other di-
seems to be very wide indeed.3 rectly productive investments. Second, its
Social overhead capital is the most impor- equipment has high minimum durability.
tant instance of indivisibility and hence of ex- Lesser durability is either technically impos-
ternal economies on the supply side. Its ser- sible or much less efficient. For this and other
vices are indirectly productive and become reasons it is very lumpy. Third, it has long
available only after long gestation periods. Its gestation periods. Fourth, an irreducible min-
most important products are investment op- imum social overhead capital industry mix is
portunities created in other industries. Social a condition for getting off the dead-end.
overhead capital comprises all those basic in- Because of these indivisibilities and be-
dustries like power, transport, or communi- cause services of social overhead capital can-
cations which must precede the more quickly not be imported, a high initial investment in
yielding, directly productive investments and social overhead capital must either precede
which constitute the framework or infra- or be known to be certainly available in order
structure and the overhead costs of the econ- to pave the way for additional more quickly
omy as a whole. Its installations are charac- yielding directly productive investments.
terized by a sizeable initial lump and low This indivisibility of social overhead capital
variable costs. Since the minimum size in constitutes one of the main obstacles to de-
these basic industries is large, excess capacity velopment ofunder-developed countries.
will be unavoidable over the initial period in Relatively few investments are made in the
under-developed countries.4 In addition, small market of an under-developed country.
there is also an irreducible minimum indus- If all investment projects were independent
try mix of different public utilities, so that an (which they are not) and if their number
under-developed country will have to invest grew, the risk of each investment project
between 30-40 per cent of its total invest- would decline by simple actuarial rules. The
ment in these channels. Since over-all vision lower marginal risk of each investment dose
is required as well as a correct appraisal of (or project) would lead to either higher or
future development, programming is un- cheaper credit facilities and these would thus
constitute internal economies. In reality,
3The capital-output ratio in the United States has
fallen over the last eighty years from around 4:1 to however, various investment decisions are not
around 3:1, while income per head, wage-rates, and independent. Investment projects have high
the relative importance of heavy industry were rising.
risks because of uncertainty as to whether
This is due to technical progress (change in produc-
tion functions), increasing returns on balance (in- their products will find a market.
creasing returns prevailing over decreasing returns), Let us restate our old example, at first for
and to the rising demand for labour-intensive services a closed economy. If a hundred workers who
characteristic of high-income economies. It is my con- were previously in disguised unemployment
viction that increasing returns played a considerable
part in it. (so that the marginal productivity of their la-
bour was equal to zero) in an underdeveloped
4We may distinguish in fact between the develop-
mental social overhead capital which provides for a country are put into a shoe factory, their
hoped for but uncertain future demand and the re- wages will constitute additional income. If
habilitation social overhead capital which caters to an the newly employed workers spend all of
unsatisfied demand of the past. The first with its ex-
their additional income on the shoes they pro-
cess capacity will necessarily have a big sectoral cap-
ital-output ratio (10-15:1); the second, through duce, the shoe factory will find a market and
breaking bottlenecks, has a certain high indirect pro- will succeed. In fact, however, they will not
ductivity and a much lower capital-output ratio. spend all of their additional income on shoes
 365 PATHS TO INDUSTRIALIZATION

There is no easy solution of creating an ad- frugality is a virtue and prodigality a vice has
ditional market in this way. The risk of not to be adapted to a situation of growing in-
finding a market reduces the incentive to in- come. Economic history does not show that
vest, and the shoe factory investment project the proportion saved from the increase in in-
will probably be abandoned. Let us vary the come was higher than the previous average
example. Instead of putting a hundred pre- rate of saving.
viously unemployed workers in one shoe fac- A zero (or very low) price elasticity of the
tory, let us put ten thousand workers in one supply of saving and a high income elasticity
hundred factories and farms which between of saving thus constitute the third
them will produce the bulk of the wage-goods indivisibility.
on which the newly employed workers will These three indivisibilities and the external
spend their wages. What was not true in the economies to which they give rise, plus the
case of one single shoe factory will become external economies of training labour, form
true for the complementary system of one the characteristic pattern of models of
hundred factories and farms. The new pro- growth of under-developed countries.
ducers will be each other's customers and will The economic factors discussed so far give
verify Says' Law by creating an additional only the necessary, but not the sufficient, con-
market. The complementarity of demand will ditions ofgrowth. A big push seems to be re-
reduce the risk of not finding a market. Re- quired tojump over the economic obstacles to
ducing such interdependent risks naturally development. There may be finally a phe-
increases the incentives to invest. . . . nomenon of indivisibility in the vigour and
A high minimum quantum of investment drive required for a successful development
requires a high volume of savings, which is policy. Isolated and small efforts may not add
difficult to achieve in low income, under-de- up to a sufficient impact on growth. An at-
veloped countries. There is a way out of this mosphere of development may only arise
vicious circle. In the first stage when income with a minimum speed or size of investment.
is increased due to an increase in investment Our knowledge of psychology is far too defi-
which mobilizes additional latent resources, cient to theorize about this phenomenon.
mechanisms must be provided which assure This does not make it a less important factor.
that in the second stage the marginal rate of It may well constitute the difference between
saving is very much higher than the average necessary and sufficient conditions for
success.
rate of saving. Adam Smith's dictum that

VI.A.2. Critique of the "Big

Push' 'Argument*

. . .[T]he theories I am considering have sev- of responsibility by the state for a greatly, in-
eral characteristics which warrant their creased rate of saving, and — extending out
being grouped together. For one thing, these from this basis according to the predilections
theories are generally strongly intervention- of the individual writer — to more or less, and
ist, at least so far as concerns the assumption generally more, control (and sometimes op-
eration) of the specific lines of investment
*From Howard S. Ellis, "Accelerated Investment and production. Secondly, these "big push"
theorists usually consider manufacture as in-
as a Force in Economic Development," Quarterly
Journal of Economics, November 1958, pp. 486, herently superior to primary production as a
491-5. Reprinted by permission. vehicle of development. These two character-
 366 INDUSTRIALIZATION STRATEGY

istics are so general that I shall terminate the ports and for marginal import substitutes,
list with these alone for greater emphasis; but where external economies are presumably
it would be tempting to point to the fre- negligible, occupies a very large part of total
quency also of an inflationary bias in writings investment. For this entire sector, the "big
of this sort, to autarkical leanings and to a push" ternal
loseseconomies.
its specific justification from ex-
fondness for general equilibrium planning as
implied by linear or nonlinear programming. We are left then with that portion of pro-
But the interventionist and other features of duction for the domestic market which does
these theories, upon which I shall want to not substitute for imports. Still, this can be a
comment later, are their overtones rather very substantial field, embracing purely local
than their substance. consumer goods production and most pub-
The substantive bases for an accelerated lic utilities — transportation, communication,
rate of investment through state intervention power, water and sewerage facilities, and the
are principally three: a demographic argu- like. Even here, however, there are limits to
ment, aline of reasoning involving the pro- potential external economies. Viner points
pensity to consume, (or to save), and thirdly, out that certain investments — presumably in
conclusions reached from the technical dis- the case of fairly inelastic demand — are cost-
continuities or "lumpiness" of investment. reducing rather than output-expanding.
Let me say clearly in advance that in no case Since external economies depend upon ex-
do I reject the reasoning completely; but that pansion ofoutput in the initial industry, they
in all cases I attach much greater weight become negligible for this category of invest-
than do the proponents of these theories to ment. Ishould like to call attention to two
the limits of possible gain, to the risks and further limitations of considerable signifi-
costs of the proposed line of action, and to the cance. In the field of purely domestic goods,
merits of alternative policies. . . . a large fraction will be personal services and
The chief basis upon which the "big push" very light industry (a good deal of food and
of investment has been justified, since its
raiment production) in which the "chunki-
original enunciation by Paul Rosenstein- ness" of fixed investment is unimportant be-
Rodan a decade and a half ago, has been the cause fixed investment is itself a small frac-
possibility of realizing extensive external tion of costs. Since external economies are
economies, and this ground is still a favorite simply internal economies in adjacent indus-
with nearly all writers of this persuasion. But tries, their significance is correspondingly
the great offset to the possibility that domes- small in these cases. It is furthermore worth
tic development programs should give rise to remembering that, in the case of public util-
further external economies has been defi- ities, potential external economies do not per-
nitely set forth by Professor Viner: foreign tain to the cost of the equipment of these in-
trade makes available to the developing dustries ifit can be more cheaply imported.
country the much more substantial econo- Taken together, all of these limitations
mies realized upon world markets, indepen- need not entirely remove the possibility of ex-
dently of home investment.1 This fact is now ternal economies. But they are neither as uni-
recognized by Professor Rosenstein. But he versal as often supposed nor, when they ac-
fails to give overt recognition to the further tually exist, as substantial. Furthermore and
fact adduced by Viner that the newly devel- finally, though their existence does increase
oping countries nowadays are chiefly primary the productivity of the economy for given
producers, and, as such, investment for ex- magnitudes of investment, they do not con-
stitute areason for a concentration of invest-
'Jacob Viner, "Stability and Progress: the Poorer ment in point of time if— as would appear
Countries' Problem," First Congress of the Interna-
tional Economic Association, Rome, September 6-1 1, probable in any but the smallest countries —
1956; mimeographed paper, pp. 27-31. [Reprinted in the "chunkiness" of individual investments
Douglas Hague (ed.), Stability and Progress in the levels out to a fairly full utilization of capac-
World Economy, New York, 1958.] ities in the aggregate for all capital facilities
 367 PATHS TO INDUSTRIALIZATION

together. This is a decidedly relevant consid- formation service, and the benefits of its
eration if"accelerated investment" is taken, information may just as well be made avail-
not as simply synonymous with more invest- able to private as to public entrepreneurs. In
ment continually, but as a "big push" fol- and of itself, linear programming does not
lowed by a lower rate. supply any rationale for accelerated invest-
Beyond its substantive theoretical basis in ment. Ifit should appear desirable to supple-
the population, savings, and external econo- ment private voluntary savings by the fiscal
mies arguments, the doctrine of accelerated arm of the state, the funds can be lent to pri-
rates of investment has overtones for policy vate firms. The theoretical underpinning of
which its proponents, I am sure, would not be accelerated investment programs pertains to
content to have ignored. One of these is the a rate of investment, and not necessarily to
predilection for manufacturing over agricul- government controlled investment. Ordinary
tural and other primary industries. In part economic motivations of the individual and
this predilection may simply reflect a senti- the firm are a powerful engine of economic
mental desire to see the country "indepen- progress. It would be regrettable if the econ-
dent" of its neighbors, particularly the richer omists of the free world created an impres-
ones; but in part it may rest on rational ar- sion to the contrary.
guments, such as the improvement in labor What, in conclusion, may be said of the
morale which is supposed to attend factory
general merits of the "big push" philosophy
production, the cultural and demographic ef- of economic development? As a starting
fects of large cities, which are supposed to be point for development some kind of impulse
favorable to economic progress, and the risks is of course, necessary; a change from stag-
of primary production from the fluctuations nation isnot likely to come by almost imper-
of world markets. On the other hand, agri- ceptible degrees. Economic historians and
cultural and primary types of production cultural anthropologists have pointed to var-
have in their favor that they utilize the rela- ious prime movers in economic change: to the
tively abundant factors of land and labor and roles of the foreign trader and foreign capital,
economize capital; that characteristically in to immigration and the transfer of tech-
the less developed countries they provide two- niques, to the process of technical innovation
thirds or more of the national income; and itself, to cultural change, and to political rev-
that, by the same token, they supply the chief olution. Among these, intensive programs of
wherewithal for industrial imports and in- state investment, as in the Japanese and Rus-
vestment ingeneral. sian cases, should certainly take their place.
It would scarcely seem necessary at the But they are by no means the only or even the
present stage of the debate concerning eco- chief channel through which development
nomic development to say that the merits of can be achieved; and the demographic advan-
investment in agriculture versus industry tages, the capital accumulation, and the ex-
have to be settled according to the peculiari- ternal economies to be expected from crash
ties of each country. By consequence, what- programs of government investment can eas-
ever merits may inhere in crash programs of ily be overrated.
investment may just as well be associated A statistical summary of recent economic
with agriculture — irrigation, drainage, trans- development throughout the world by John
portation facilities, reform of fragmented H. Adler reaches the important conclusion,
landholdings, etc., — as with building indus- among others, that "a relatively low level of
trial plants; in particular cases, indeed, more investment 'pays ofF well in the form of ad-
so.
Somewhat similar reflections would be ditional output."2 The author emphasizes this
conclusion most sharply in connection with
germane to the penchant of the "big push"
economists for planning, state direction of in- 2John H. Adler, "World Economic Growth — Re-
vestment, and extensive controls. Linear pro- trospect and Prospects," Review of Economics and
gramming, for example, is essentially an in- Statistics, Aug. 1956, p. 279; cf, also p. 283.
 368 INDUSTRIALIZATION STRATEGY

India and Pakistan; but the chief reason for income countries), and progress along the
this conclusion, the prevailing low capital- family-limitation front would seem to be
output ratio, is also characteristic of many equally critical. Taken together, or in some
other of the less developed countries of Asia cases even singly, we would seem to have
and Latin America as his statistics reveal. identified a number of factors in economic
Thus it appears that it is far from generally progress which could outweigh a burst of
true that a massive injection of capital is a state-enginered investmenet.
precondition of growth. Some food for thought concerning pro-
A general weakness of the "big push" doc- grams of intensive investment would seem to
trine isthat it frequently ignores the condi- be offered by certain points made recently by
tions for evoking the investment to which it Simon Kuznets. His statistical and historical
ascribes such potency in the general picture studies lead to the conclusion that "current
of development, as well as neglecting the con- international differences in per capita income
ditions under which investments, once made, are congealed effects of past differences in
can be fruitful. It is through the assumption the rate of growth of per capita income."
of a deus ex machina, the state, which does How far would it be necessary to go back into
all or most of the investing, that this theory the history of the more advanced countries to
is able to avoid the problems of securing not reach levels comparable to the per capita in-
only the saving, but also the willingness to comes of the currently less developed coun-
undergo risk, which is implied in investment. tries? The answer is that we should have to
And it is only through a singular narrowness go back about ten decades to reach the cur-
that the theory often implies that it tells the rent income level of Latin American and
whole story of the successful operation of the about fifteen decades for that of Africa and
economy, once the investment is made. Asia.3 Thus, even at a very early stage in the
In point of fact, the conditions for the industrialization of Western Europe, per cap-
evoking of private investment and the condi- ita incomes were probably as high as in Latin
tions for the profitable use of capital are American today and certainly higher than in
largely the same. I should place high upon Asia and Africa. The economic development
this list the existence of stable and honest of the most advanced countries, at least,
government, the absence of inflation, and the scarcely seems to be the result of crash
accessibility of the economy to the gains of
foreign trade and commerce. But other fac- programs.
tors, such as the improvement of general and 3Simon Kuznets, "Quantitative Aspects of the Eco-
nomic Growth of Nations," in Economic Develop-
technical education, the amelioration of ag- ment and Cultural Change, October 1956; see espe-
riculture (which bulks large in nearly all low- cially pp. 23-5.

VI. A. 3. Linkage Effects and

Industrialization*

[T] wo inducement mechanisms may be con- 1 . The input-provision, derived demand, or

sidered tobe at work within the directly pro- backward linkage effects, i.e., every nonpri-
ductive activities (DPA) sector: mary economic activity, will induce attempts
♦From Albert O. Hirschman, The Strategy of Eco-
to supply through domestic production the
nomic Development, New Haven, Yale University inputs needed in that activity.
Press, 1958, pp. 100-4, 109-13. Reprinted by 2. The output-utilization or forward link-
permission. age effects, i.e., every activity that does not by
 369 PATHS TO INDUSTRIALIZATION

its nature cater exclusively to final demands, puts isequal to au a2 . . . an; then the strength
will induce attempts to utilize its outputs as of the stimulus or the probability that the set-
inputs in some new activities. ting up of industry Wwill lead to the setting
Development policy must attempt to enlist up of industries producing the inputs is equal
these well-known backward and forward ef- to the ratio of the >>'s to the a's.1 Minimum
fects; but it can do so only if there is some economic size is not a technical concept, but
knowledge as to how different economic ac- is defined in economic terms relative to nor-
tivities "score" with respect to these effects. mal profits and efficient foreign suppliers. In
Ordinarily economists have been content other words, it is the size at which the do-
with general references to the advantages of mestic firm will be able both to secure normal
external economies, complementarities, cu- profits and to compete with existing foreign
mulative causation, etc. But no systematic ef- suppliers, taking into account locational ad-
fort has been made to describe how the de- vantages and disadvantages as well as, per-
velopment path ought to be modified so as to haps, some infant industry protection. In this
maximize these advantages even though the way comparative cost conditions are auto-
existence of input-output statistics supplies matically taken into account.2
us with a few tools for an analysis of this In the case of forward linkage, an interpre-
kind.
tation of the p's is less straightforward. The
First, a further note on the linkage concept concept of economic size is not helpful here,
itself. What do we imply when we speak of since the size of the market for the industries
the linkage effects emanating from industry that might be brought into being through for-
A toward industy Bl Language can be quite ward linkage does not depend on their sup-
ambiguous here, for we may have in mind the pliers. Aclue can perhaps be found in the im-
potential importance of the linkage effect in portance ofthe articles produced by industry
terms of, say, the net output of the new in- Was inputs for the output of the to-be-linked
dustries that might be called forth; or we may industry. If these inputs are a very small frac-
mean the strength of the effect, i.e., the prob- tion of the industry's eventual output, then
ability that these industries will actually their domestic availability is not likely to be
come into being. The total effect could be an important factor in calling forth that in-
measured by the sum of the products of these dustry. If,on the other hand, these articles
two elements; in other words, if the establish- are subjected to few further manufacturing
ment of industry W may lead, through link- operations, then the strength of the forward
age effects, to the establishment of n addi- stimulus is likely to be substantial, provided
tional industries with net outputs equal to xt demand is sufficient to justify domestic
(i = 1, 2, . . . n) and if the probability that production.
each one of these industries will actually be In these cases, then, importance and
set up as a result of the establishment of in- strength — the x's and the p's — of the linkage
dustry Wis pi (/ = 1,2,... n), then the total 'The ratio is to be defined as having a ceiling of 1,
linkage effect of industry W is equal to i.e., the value of the ratio is equal to unity, whenever
n
the v's are equal to or larger than the a's. Note also
that the y's are equivalent to the gross output of the
1 new industries or firms in physical terms whereas the
The probabilities can be interpreted as x's are their net outputs in value terms.
measuring the strength of the stimulus that 2Data on the economic size of plants in different in-
dustries would be the starting point for determining
is set up. For backward linkage, this strength minimum economic size in different countries. Re-
can be roughly measured as follows: suppose search in this area in relation to economic develop-
industry W requires annual inputs of yx, y2 ment is surprisingly scant, except for the pioneering
. . . yn and suppose that the minimum eco- article of K. A. Bohr, "Investment Criteria for Man-
ufacturing Industries in Underdeveloped Countries,"
nomic size (in terms of annual productive ca- Review of Economics and Statistics, Vol. 36, (May
pacity) offirms that would turn out these in- 1954), pp. 157-66.
 370 INDUSTRIALIZATION STRATEGY

effect are inversely correlated. Industries reduced to the fact that an input of the newly
where the jc's are small and the p's large are established industry is an output of the to-be-
sometimes aptly called "satellite" industries. created industry or vice versa, but the estab-
They are almost unfailingly established in lished industry would not be the principal
the wake of industry W but are of minor im- customer or supplier of the to-be-created in-
portance in comparison to that industry. dustry; infact, particularly in cases of back-
Thus defined, satellite industries can be es- ward linkage, minimum economic size of the
tablished through backward or forward link- to-be-created industry would frequently be
age. In the case of cement, for instance, the larger than that of the industry where the
manufacture of multi-wall bags for packing
purposes represents backward linkage while linkage originates.3
In spite of the importance of the nonsatel-
the establishment of a cement block industry lite type of linkage, it seems necessary to pro-
represents satellite formation through for- vide for some arbitrary cut-off point for small
ward linkage. A satellite industry usually has probabilities. It is all very well to say that the
the following characteristics: establishment of a brewery sends out a stim-
1 . It enjoys a strong location advantage from ulus in the direction of a paper industry be-
cause of the labels needed for the beer bot-
proximity to the master industry;
tles, but by itself this stimulus is not likely
2. it uses as principal input an output or by- ever to lead to the setting up of a paper mill.
product of the master industry without
Thus, if we consider in isolation the linkage
subjecting it to elaborate transformation,
effects exclusively of the beer industry on fur-
or its principal output is a — usually ther industrial or agricultural development,
minor — input of the master industry; and we should consider only those stimuli whose
3. its minimum economic size is smaller than
that of the master industry. probability exceeds a certain critical value,

While satellite industries are almost cer- sayIf one-half.4

we proceed in this way, the joint linkage
tain to be established once the master indus- effects of two industries, say beer and ce-
try is in place, the establishment of industry ment, considered as a unit, are likely to be
W also results m stimuli toward the setting larger than the sum of their individual link-
up of nonsatellite industries. In these cases, age effects, since some of the xp products
the strength of the stimulus is infinitely which are omitted in computing the individ-
weaker, but the stake is far bigger. Examples ual effects because the p's are below the crit-
of such a situation are the stimulus that the ical value will exceed this value if added to-
setting up of a multi-wall bag factory gives gether for both industries. Here we have an
toward the creation of pulp and paper indus- argument in favor of multiple development
try or, for the case of forward linkage, the that we would consider convincing were it not
stimulus given by the establishment of an that our principal argument against it is con-
iron and steel industry to all the metal-fab- 3To the extent that the minimum economic size of
rication industries. Here the establishment of an industry is larger the farther away one moves from
one industry is a contributing factor which by the finished consumer or producer goods stage. This
itself is quite unlikey to result in the creation is, of course, by no means universally true as is shown,
of the others; but when we speak of external e.g., by the large minimum size of the motor vehicles
makers.
economies and complementarities, we think
at least as much of these uncertain linkages 4It is a good rule of thumb that an industry can
properly be established in an underdeveloped but de-
as of the far more certain, but also far less veloping country as soon as existing demand is equal
significant, satellites with which any industry to one-half of the economic size of the plant as defined
of a certain size surrounds itself. The weak- above. The additional demand needed to justify the
investment can be expected to come from the growth
ness of the stimulus in the case of nonsatel- of existing demand and from the development of new
lites can be explained by the absence of the demand through forward linkage, once the plant is in
three factors that define satellites. Linkage is existence.
 371 PATHS TO INDUSTRIALIZATION

cerned with its feasibility rather than with its part is subjected to some processing indus-
desirability. tries that can be characterized as satellite
The fact that the linkage effects of two in- inasmuch as the value added by them to the
dustries viewed in combination are larger agricultural product (milling of wheat, rice,
than the sum of the linkage effects of each coffee, etc.) is small relative to the value of
industry in isolation helps to account for the the product itself. Only a comparatively
cumulative character of development. When small fraction of total agricultural output of
industry A is first set up, its satellites will underdeveloped countries receives elaborate
soon follow; but when industry B is subse- processing, which usually takes place abroad.
quently established, this may help to bring The case for inferiority of agriculture to
into existence not only its own satellites but manufacturing has most frequently been ar-
some firms which neither A nor B in isolation gued on grounds of comparative productivity.
could have called forth. And with C coming While this case has been shown not to be en-
into play some firms will follow that require tirely convincing, agriculture certainly stands
the combined stimuli not only of B and C but convicted on the count of its lack of direct
of A, B, and C. This mechanism may go far stimulus to the setting up of new activities
toward explaining the acceleration of indus- through linkage effects: the superiority of
trial growth which is so conspicuous during manufacturing in this respect is crushing.
the first stages of a country's development. This may yet be the most important reason
militating against any complete specializa-
BACKWARD LINKAGE AT WORK tion of underdeveloped countries in primary
production.
The lack of interdependence and linkage is The grudge against what has become
of course one of the most typical character- known as the "enclave" type of development
istics ofunderdeveloped economies. If we had is due to this ability of primary products from
homogeneous input-output statistics for all mines, wells, and plantations to slip out of a
countries, it would certainly be instructive to country without leaving much of a trace in
rank countries according to the proportion of the rest of the economy. Naturally hostility
intersectoral transactions to total output; it is to the profits earned by foreign companies
likely that this ranking would exhibit a close plays an important role in such attitudes; but
correlation with both income per capita and the absence of direct linkage effects of pri-
with the percentage of the population occu- mary production for export lends these views
pied in manufacturing. a plausibility that they do not have in the
Agriculture in general, and subsistence ag- case of foreign investment in manufacturing.
riculture inparticular, are of course charac- I say plausibility rather than validity, for
terized bythe scarcity of linkage effects. By while as such the primary production activi-
definition, all primary production should ex- ties leading to exports may e*xert few devel-
clude any substantial degree of backward opmental effects, they do finance imports
linkage although the introduction of modern which can become very powerful agents of
methods does bring with it considerable out- development as we shall see below.
side purchases of seeds, fertilizers, insecti- Since interdependence in the input-output
cides, and other current inputs, not to speak sense is so largely the result of industrializa-
of machines and vehicles. We may say that tion, we must now attempt to trace the var-
the more primitive the agricultural and min- ious ways in which manufacturing and the
ing activities, the more truly primary they accompanying linkage effects make their ap-
are. pearance. Inthis connection, we shall utilize
Forward linkage effects are also weak in another one of Chenery's findings, namely
agriculture and mining. A large proportion of that more than ninety percent of all input-
agricultural output is destined directly for output flows can usually be arranged in a tri-
consumption or export; another important angular pattern. Circularity — i.e., the fact
 372 INDUSTRIALIZATION STRATEGY

that coal is needed for steel-making and steel verting, assembly, and mixing plants, the
for coal mining — is undoubtedly present in pharmaceutical laboratories, the metal-fab-
the structure of a country's production, but ricating industries, and many others. This
apparently to a much smaller degree than trend has many advantages: it often provides
would be suspected upon looking at an input- an investment outlet for small amounts of
output table that has not been "triangular- capital that might not easily become availa-
ized." In other words, there is no compelling ble for ventures which require the pooling of
technological requirement for the simultane- the resources of many investors, and it makes
ous setting up of various industries, an inter- it possible to start industrial undertakings
esting complement to our case against the ex- without the heavy risk that comes in under-
istence of such a requirement on economic developed countries from having to rely on
grounds. the output of unreliable domestic producers.
In a triangular arrangement of the input- In this way underdeveloped countries often
output matrix, there is a "last" sector whose set up "last" industries first — i.e., these are
output goes entirely to final demand and "last" industries considering the input-output
which takes in inputs from a number of other flow of the advanced countries: what in these
sectors; the second-to-last sector sells its out- countries are inputs from the other sectors
put to final demand and to the last sector and are replaced in underdeveloped countries by
buys inputs from some or all other sectors ex- imports. Such industries could be termed
cept from the "last"; and so on, until we come "enclave import industries," in analogy to the
to the "first" sector whose output goes to all enclave export activities that were previously
the subsequent sectors and possibly also to mentioned. For here again we have an under-
final demand, but which does not use any in- taking that at least in its beginning is anti-
puts from other sectors. septically linkage-free; materials are im-
Industrialization can of course start only ported from abroad, some value is added to
with industries that deliver to final demand, them through mixing, assembling, packag-
since ex hypothesi no market exists as yet for ing, etc., and the finished product is rushed to
intermediate goods. This means that it will be the final consumers. The enclave nature of
possible to set up only two kinds of industries: these industries is sometimes emphasized by
1 . those that transform domestic or imported the location of the plant at a point as close as
primary products into goods needed by possible to the most convenient port of arrival
final demands; of the imported materials, and again this type
of venture has proven particularly attractive
2. those that transform imported semiman-
ufactures into goods needed by final to foreign capital — many of the branch
demands. plants owned by foreign corporations special-
ize in this kind of operation.
To the pioneer industrial countries only the But there is a considerable difference be-
first course was open, and this explains the tween the enclave export and enclave import
towering importance of a few industries (tex- activities. The former have great trouble in
tiles, iron and steel, pottery) during the early breaking out of the enclave situation. Usually
stages of the Industrial Revolution. In to- some forward linkage effects can be uti-
day's underdeveloped countries the textiles, lized— ores and cane sugar can be refined be-
food processing, and construction materials fore being shipped. But the scope for such op-
industries based on local materials are still of erations isstrictly limited. With respect to
great importance, but, to a very significant import enclave industries, the situation is
extent, industrialization is penetrating these radically different: they set up backward
countries in the second manner, through linkage effects of practically infinite range
and depth.
plants that perform the "final touches" on al-
most-finished industrial products imported In fact, much of the recent economic his-
from abroad. Examples are the many con- tory of some rapidly developing underdevel-
 373 PATHS TO INDUSTRIALIZATION

oped countries can be written in terms of in- have been gradually replaced by domestic
dustrialization working its way backward production which has been called forth by the
from the "final touches" stage to domestic existence of a large and stable market. Of
production of intermediate, and finally to considerable importance are the backward
that of basic, industrial materials. In this linkage effects that are the combined result
way, industrialization has even proven to be of the existence of several "last stage" indus-
a powerful stimulus to the development of tries. The minimum economic size of many
agriculture. By providing a reliable market, intermediate and basic industries is such that
processing industries originally based on im- in small markets a variety of user industries
ported agricultural materials such as cotton needs to be established before their combined
textiles and beer have stimulated in Colom- demand justifies a substitution of imports of
bia the domestic production of cotton and intermediate and basic goods by domestic
barley. . . .
production.
In most of these cases, imported goods

Comment
Albert O. Hirschman has broadened his linkages approach in an important paper, "A Gen-
eralized Linkage Approach to Development, with Special Reference to Staples," in Manning
Nash (ed.), Essays on Economic Development and Cultural Change in Honor of Bert F. Ho-
selitz (1977). This paper goes beyond production linkages of the input-output character to
consider consumption and fiscal linkages. The analysis is applied to export-led growth based
on staple exports. See also selections VLB. 3 and VLB. 4 for related discussions of export-led
development.

VI.A.4. The Case for Balanced

Growth*

It is no longer so certain that the less devel- the case for balanced growth which is now so
oped countries can rely on economic growth much in vogue.
being induced from the outside through an The circumstances indicated do not apply
expansion of world demand for their exports to all underdeveloped countries today: Ku-
of primary commodities. In these circum- wait and perhaps Iraq have nothing to worry
stances reliance on induced expansion about. But in so far as these cicumstances do
through international trade cannot provide a exist in reality it is clear that the poorer coun-
solution to the problem of economic devel- tries, even if they are only to keep pace with
opment. It is not surprising therefore that the richer, to say nothing about catching up
countries should be looking for other solu- with them, must expand production for their
tions. Itis important to keep these things in own domestic markets or for each others'
mind, because they form the background to markets. Now domestic markets are limited
because of mass poverty due to low produc-
♦From Ragnar Nurkse, "The Conflict Between tivity. Private investment in any single indus-
'Balanced Growth' and International Specialization," try considered by itself is discouraged by the
Lectures on Economic Development, Faculty of Eco- smallness of the existing market.
nomics (Istanbul University) and Faculty of Political
Sciences (Ankara University), Istanbul, 1958, pp. The limits set by the small size of the local
170-76. Reprinted by permission. market for manufactured goods are so
 374 INDUSTRIALIZATION STRATEGY

plainly visible to any individual businessman plication ofcapital to one industry alone may
that we are fully justified in taking for therefore be subject to sharply diminishing
granted conditions of imperfect competition, returns. As a way of escape from slowness if
and not the pure atomistic competition which not from stagnation, the balanced growth
even in advanced economies does not exist to principle envisages autonomous advance
any significant degree, outside the economics along a number of lines more or less
textbooks. simultaneously.
The solution seems to be a balanced pat- Viewed in this way, balanced growth is a
tern of investment in a number of different means of accelerated growth. Some econo-
industries, so that people working more pro- mists treat the problem of achieving balanced
ductively, with more capital and improved growth as quite separate from the problem of
techniques, become each others' customers. speeding up the rate of advance in a back-
In the absence of vigorous upward shifts in ward economy. I admit that this may be a
world demand for exports of primary prod- convenient distinction to draw on other
ucts, alow income country through a process grounds. But in my view, balanced growth is
of diversified growth can seek to bring about first and foremost a means of getting out of a
upward shifts in domestic demand schedules rut, a means of stepping up the rate of growth
by means of increased productivity and when the external forces of advance through
therefore increased real purchasing power. In trade expansion and foreign capital are slug-
this way, a pattern of mutually supporting in- gish or inoperative.
vestments indifferent lines of production can In the existing state of affairs in low in-
enlarge the size of the market and help to fill come areas the introduction of capital-using
the vacuum in the domestic economy of low techniques of production in any single indus-
income areas. This, in brief, is the notion of try is inhibited by the small size of the mar-
balanced growth. ket. Hence the weakness of private invest-
Isolated advance is not impossible. A soli- ment incentives in such areas. The balanced
tary process of investment and increased pro- growth principle points to a way out of the
ductivity inone industry alone will certainly deadlock. New enterprises set up in different
have favorable repercussions elsewhere in the industries create increased markets for each
economy. There is no denying that through other, so that in each of them the installation
the normal incentives of the price mechanism or capital equipment becomes worth while.
other industries will be induced to advance
As Marshall said, "The efficiency of special-
also. But this may be a snail's pace of prog- ized machinery ... is but one condition of its
ress. The price mechanism works but it may economic use; the other is that sufficient work
work too slowly. That is one reason for the
should be found to keep it well employed"
frequently observed fact that foreign direct {Principles, p. 264). The techniques that
investments in extractive export industries have been developed in production for mass
have created high productivity islands in low markets in advanced countries are not well
income areas and have had little impact on adapted and sometimes not adaptable at all
the level of productivity in the domestic to output on a more limited scale. It is easy
economy. to see that the relationship between the size
Within the domestic economy itself, ad- of the market and the amount of investment
vance in one direction, say in industry A, required for efficient operation is of consid-
tends to induce advance in B as well. But if erable importance for the theory of balanced
it is only a passive reaction to the stimulus
coming from A, the induced advance of B growth.
Frequently the objection is made: But why
may be slow and uncertain. And B's slowness use machinery? Why adopt capital-using
and passiveness will in turn slow down and methods in areas where labor is cheap and
discourage the initial advance of A. The ap- plentiful? Why not accordingly employ tech-
 375 PATHS TO INDUSTRIALIZATION

niques that are labor-intensive instead of cap- ducement ofgrowing markets. It is here that
ital-intensive? the element of mutual support is so useful
The answer is obvious. As an adaptation to and, for rapid growth, indispensable.
existing circumstances, including the existing It is important to note that the doctrine
factor proportions, the pursuit of labor-inten- under consideration is not itself concerned
sive production methods with a view of econ- with the question of where the capital is to be
omizing capital may be perfectly correct. But found, for all the balanced investment which
the study of economic development must con- it envisages. I have tried to make it clear in
cern itself with changing these circum- my discussion of it that the argument is pri-
stances, not accepting them as they are. marily relevant to the problem of the demand
What is wanted is progress, not simply ad- for capital; it takes an increased supply of
aptation to present conditions. And progress capital for granted. In my presentation bal-
depends largely on the use of capital, which anced growth is an exercise in economic de-
in turn depends on adequate and growing velopment with unlimited supplies of capital,
markets, which in the absence of a strongly analogous to Professor Lewis's celebrated ex-
rising world demand for the country's exports ercise in development with unlimited labor
means a diversified output expansion for do- supplies.
mestic use. In reality, of course, capital supplies are
Reference has been made to the impor- not unlimited. It may be that the case for
tance of autonomous advance in a number of state investment stems chiefly from the fact
mutually supporting lines of production. How that capital is scarce and that government ef-
is this achieved? Autonomous advance in dif- forts are necessary to mobilize all possible
ferent branches simultaneously may come domestic sources of saving. Measures to
about through the infectious influence of check the expansion of consumer demand
business psychology, through the multiplier may be necessary to make resources availa-
effects of investment anywhere which can ble for investment but may at the same time
create increased money demand elsewhere, weaken the private inducement to invest.
or through deliberate control and planning by This is a famous dilemma to which Malthus
public authorities. According to some writers first called attention in his Principles of Po-
the balanced growth argument implies that litical Economy. A case for state investment
the market mechanism is eliminated and that may clearly arise if and when the mobiliza-
investment must be effected according to a tion of capital supplies discourages private
coordinated plan. This opinion, which is investment activity and so destroys the de-
widely held, seems to be dubious. There are mand for capital. But this case is entirely sep-
many important reasons for government arate from the principle of balanced growth
planning, but this is not necessarily one of as such. It might only be added that the cap-
them. As a means of creating inducements to ital supply problem alone creates a strong
invest, balanced growth can be said to be rel- presumption against relying on the indiscrim-
evant primarily to a private enterprise sys- inate use of import restriction which may re-
tem. State investment can and often does go duce a country's real income and therefore
ahead without any market incentives. Plan- make it harder to increase the flow of saving.
ning authorities can apply capital, if they Elsewhere I have tried to explain how the
have any, wherever they may choose, though balanced growth idea is related to the classi-
if they depart too much from balance as dic- cal law of markets. Supply creates its own de-
tated by income elasticities of demand they mand, provided that supply is properly dis-
will end by creating white elephants and in- tributed among different commodities in
tolerable disproportionalities in the structure accordance with consumers' wants. An in-
of production. It is private investment that is crease in consumable output must provide a
attracted by markets and that needs the in- balanced diet. Each industry must advance
 376 INDUSTRIALIZATION STRATEGY

along an expansion path determined by the engaged on capital projects. This is the con-
income elasticity of consumer demand for its cept relevant to the problem of capital sup-
product. This simple idea must be the start- ply. Itis obvious that even a large marketable
ing point in any expansion of production of surplus of food need not involve any saving by
domestic markets in the less developed coun- the farmers. It presents a very helpful in-
tries, inso far as external demand conditions ducement, but does not in itself create the
do not favor the traditional pattern of means, for capital investment outside the ag-
"growth through trade." Yet, as often hap- ricultural sector. A fuller discussion of the in-
pens in economic discussion, critics have terrelationship between marketable and in-
tended to dismiss this idea either as a dan- vestible surpluses would take us too far from
gerous fallacy or as an obvious platitude. It is our present subject. It seemed desirable to
hardly necessary to add that the pattern of mention the distinction here merely for the
consumable output cannot be expected to re- sake of conceptual clarity. So much for the
main the same in successive stages of devel- relation between agriculture and industry.
opment. The content of a balanced diet of a Within the manufacturing field alone the
man with a thousand dollars a year will differ case for balanced investment implies a hori-
from that of a man with a hundred dollars. zontal diversification of industrial activities
The relation between agriculture and man- all pushing ahead, though naturally at vary-
ufacturing industry offers the clearest and ing rates. The objection can be made that
simplest case of balance needed for economic such diffusion of effort and resources over
growth. In a country where the peasantry is many different lines of activity must mean a
incapable of producing a surplus of food loss of dynamic momentum in the economy.
above its own subsistence needs there is little This is possible. The dispersal of investment
or no incentive for industry to establish itself: over a variety of consumer-goods industries
there is not a sufficient market for manufac- can undoubtedly be carried to excess. The
tured goods. Conversely, agricultural im- balanced growth principle can be and has
provements may be inhibited by lack of a been interpreted far too literally. Producing
market for farm products if the non-farm a little of everything is not the key to prog-
sector of the economy is backward or unde- ress. The case for balanced growth is con-
veloped. Each of the two sectors must try to cerned with establishing a pattern of mu-
move forward. If one remains passive the tually supporting investments over a range of
other is slowed down. industries wide enough to overcome the frus-
It is important in this connection to make tration ofisolated advance, in order precisely
a clear distinction between two concepts that to create a forward momentum of growth.
are frequently confused: the marketable sur- The particular factors that determine the op-
plus and investible surplus of the farm sector. timum pattern of diversification have to do
The farm sector's marketable surplus of farm with technology, physical conditions and
products determines the volume of non-farm other circumstances that vary from country
employment in manfacturing and other ac- to country. There can be no standard pre-
tivities. Itreflects simply the farm sector's scription ofuniversal applicability. We are
demand for non-agricultural commodities. concerned with a point of principle and can-
This is the concept that is relevant to the bal- not deal with the precise forms of its imple-
anced growth principle. mentation inpractice. Just as it is possible for
An investible surplus of farm products rep- manufacturing industry as a whole to lan-
resents an act of saving in the farm sector. It guish iffarmers produce too little and are too
can conceivably result from a transfer of sur- poor to buy anything from factories, so it is
plus labourers from the farms to capital con- possible for a single line of manufacturing to
struction projects: a food surplus may then fail for lack of support from other sectors in
arise through forced or voluntary saving in industry as well as agriculture; that is, for
the farm sector for maintaining the workers lack of markets.
 377 PATHS TO INDUSTRIALIZATION

VI. A. 5. Balanced versus

Unbalanced Growth *
. . . Before we enter upon a discussion of the iable. But again, on closer inspection it would
merits and faults of the doctrines of balanced seem that UG, to be most effective, does re-
growth (BG) and of unbalanced growth quire planning and preferably state planning,
(UG), it is necessary to clarify two questions because no private firm may want or be able
to which the contributors to the debate have to carry the surplus capacity and the losses,
not given clear and satisfactory answers. The and because private horizons are too narrow.
first question, most relevant in the present It is not surprising that both BG and UG
context of our discussion, concerns the role of should, to be most effective, presuppose (each
planning; the second question [concerns] the a different kind of) planning, for they are
role of supply limitations and supply both concerned with lumpy investments and
inelasticities. complementarities. Coordination is needed in
In the controversy the role of government order both to get things done that otherwise
(or for that matter private) planning has not would not be done, and in order to reap the
always been brought out clearly. In particu- rewards of complementarities. Market forces
lar, it is not always clear whether the ques- look best after adjustments that can be made
tion under consideration relates to planning in infinitesimally small steps. This is why the
or whether it relates to an attempt to explain concept "marginal" plays such an important
development that takes place without plan- part in neo-classical Western economic the-
ning, or with only an initial impulse of plan- ory. It is also one of the important differences
ning in the form of an investment project, between developed and underdeveloped
while things are thereafter left to take their countries. In the former a new profitable in-
own course with market forces responding to vestment project is normally small relative to
demand and supply. the size of existing capital equipment (how-
Nurkse thought that BG is relevant pri- ever measured), relatively to new investment,
marily to a private enterprise economy. It is and relatively to the hinterland of facilities
(he argued) private investment that needs on which it can draw. In underdeveloped
market inducements. In his doctrine, the countries indivisibilities are more prominent
choice between public and private investment and marginal adjustments rarer for at least
and betweeen direct controls and market in- four reasons.
centives ismainly a matter of administrative First, both the existing stock of equipment
expediency. But he seems to be wrong in this. and the additions to it are small compared
The indivisibilities assumed in BG imply the with those in advanced countries with com-
need for coordination, i.e., planning, al- parable populations. Since plant and equip-
though itwould, in principle, be possible to ment often have to be a minimum size for
have either private or public coordination. technical reasons, the addition of a plant or a
UG as propounded by Hirschman is con- piece of equipment makes a greater propor-
sistent with, but does not require, initial and tionate difference both to the stock of capital
continued planning. His state administrators and to total investment.
are — or should be — subject to the same kind Second, economic development is usually
of pressures as private entrepreneurs. The directed at moving people from agriculture to
role of the state is both to induce and to re- industrial enterprises. This normally implies
pair disequilibria. Thus state action becomes an increase in the number of indivisible units.
a dependent, as well as an independent, var- Third, the necessary social overhead capi-
tal and the basic structure of industry
*From Paul Streeten, "Balance versus Unbalanced
Growth," The Economic Weekly, April 20, 1963. pp.
(power, steel, transport, housing, government
669-71. Reprinted by permission. buildings) consist of large indivisible units.
 378 INDUSTRIALIZATION STRATEGY

Fourth, complementarities between enter- velopment decisions" and "developmental

prises and activities are likely to be more im-
portant in the meagre economies of under- Insofar as BG is concerned with the crea-
developed countries, so that a given tion of markets through complementary in-
tasks."
investment is more liable to require comple- vestment projects and the inducement to in-
mentary and supplementary investments. vest by providing complementary markets for
Both BG and UG give rise to external econ- final goods, it stresses a problem which is
omies. A cost incurred by A creates profit op- rarely serious in the countries of the region.
portunities for B. If steps are taken to seize Final markets can often quite easily be cre-
these opportunities at once and in one type of ated without recourse to BG, by import re-
sequence (BG), the results will be different strictions and, less easily, by export
than if they are seized later and in a different expansion.
type of sequence (UG). But there is no guar- On the other hand, although UG is correct
antee that A will be induced by market forces in pointing to the scarcity of decision-taking
to incur these costs, indeed there is a pre- in some countries, it should not be contrasted,
sumption that it will not be so induced. but it should be combined with the provision
We next turn to the role of supply limita- of more supplies. The contrast drawn by UG
tions and supply inelasticities in the contro- between scarcity of physical resources and
versy. Nurkse explicitly confined his discus- scarcity of decision-taking can be misleading.
sion to the demand side. He assumed supplies Those who stress resources say that decisions
to be available and asked what would invest- will be taken as soon as resources are avail-
ment have to be like to justify them? He able; those who stress decision-taking say
wrote: that resources will flow freely as soon as ad-
There is no suggestion here that, by taking care equate inducements to take decisions are pro-
of the demand side alone, any country could, as it vided. The former group of experts go out on
were, lift itself up by its bootstraps. We have been missions and advocate high taxation in order
considering one particular facet of our subject. to "set resources free," the latter recom-
The more fundamental difficulties that lie on the mend low taxation in order to "encourage
supply side have so far been kept off-stage for the
sake of orderly discussion.1 Both views reflect misplaced aggregation
enterprise."
and illegitimate isolation, two types of bias
Nevertheless, the position of this chapter
in his book and the emphasis laid on it have introduced by the careless use of Western
concepts and models. No general formula
led to misinterpretations. If BG stresses mar- will serve. The correct division often cuts
kets as the main limitation on growth, UG in
the Hirschman version stresses decisions. across these categories. The question is what
combination of resource policy, reform of at-
The implication of Hirschman's theory is
that supplies will be forthcoming with rela- titudes (including "incentives") and of legal,
social and cultural institutions, is necessary
tive ease if only the lack of decision-taking
can be overcome. This shift of emphasis to an in a particular situation.
attitude, usually assumed either constant or Moreover, the tendency of both BG and
automatically adjusted to precisely the re- UG to underplay supply limitations diverts
attention from the fact that planning must be
quired extent, should be welcomed. Hirsch- directed as much at restricting supplies in
man has been charged with excessive preoc- certain directions as at expanding them in
cupation with investment decisions. Much of
his book indeed focuses attention on them, others. The policy package presupposes a
but it clear that he had a wider concept in choice of allocating limited supplies, i.e., sup-
mind, as is shown by his use of the terms "de- plies growing at a limited rate, and in re-
sponse to certain stimuli, to the most impor-
1Problems of Capital Formation in Underdevel-
tant uses, combined with inducements to
oped Countries, pp. 30-31. decisions of all kinds (not only investment de-
 379 PATHS TO INDUSTRIALIZATION

cisions). These supply limitations are consid- ment projects will turn out to be wasteful and
erably less important in advanced industrial will be abandoned.
countries now and were less important in the Insofar as unbalance does create desirable
early developing phase of many now ad- attitudes, the crucial question is not whether
vanced countries, like Sweden or the regions to create unbalance, but what is the optimum
of recent settlement. These countries had al- degree of unbalance, where to unbalance and
most unlimited access to capital at low inter- how much, in order to accelerate growth;
est rates, a reserve of skilled labour and plen- which are the "growing points," where
tiful natural resources. Again, certain should the spearheads be thrust, on which
underdeveloped regions in advanced coun- slope would snowballs grow into avalanches?
tries (Southern Italy, the South of the USA) Although nobody just said "create any old
can draw on supplies but lack development unbalance," insufficient attention has been
decisions. paid to its precise composition, direction and
timing.
The models developed in the BG vs. UG
controversy seem to have drawn on this kind The second weakness of UG is that the
of experience fromto "ceilingless economies" theory concentrates on stimuli to expansion,
which is relevant South America but not and tends to neglect or minimise resistances
to the entirely different problems of South caused by UG. UG argues that the active
Asia. The two important differences between, sectors pull the other with them, BG that the
on the one hand, advanced countries now and passive sectors drag the active ones back.
in their development phase, and, on the other While the former is relevant to South Amer-
hand, the underdeveloped countries of South ica, the latter is relevant to South Asia. It
Asia are: would, of course, be better, as Nurkse would
1. that investments in advanced countries have liked it, if all sectors were active, and
can more often be treated as marginal than the wish may have been father of the thought
in underdeveloped countries, and behind these models. But the problem is how
2. that advanced countries are and were to activate them. Activation measures must
high supply-elasticity economies with re- take the form both of positive inducements
sponses and institutions already adapted to and of resistances of resistances.
economic growth. The UG model in the Hirschman version
Both doctrines have certain faults. The has the great merit, in comparison with many
trouble with advocating UG is that, for coun- other models, of including attitudes and in-
tries embarking on development, unbalance stitutions, and in particular investment in-
is inevitable, whether they want it or not, and centives, normally assumed fully adjusted to
governments and planners do not need the requirements, and of turning them from in-
admonitions of theoreticians. All investment dependent variables or constants into depen-
creates unbalances because of rigidities, in- dent variables. In particular Hirschman's
divisibilities, sluggishness of response both of discussion of forward and backward linkages
supply and of demand and because of mis- is provocative and fruitful. It brings out the
calculations. There will be, in any case, previously neglected effects of one investment
plenty of difficulties in meeting many urgent on investment at earlier and later stages of
requirements, whether of workers, techni- production. But the doctrine underplays ob-
cians, managers, machines, semimanufac- stacles and resistances (also in attitudes)
tured products, raw materials or power and called into being by imbalance. Shortages
transport facilities and in finding markets create vested interests; they give rise to mo-
permitting full utilization of equipment. nopoly gains; people may get their fingers
Market forces will be too weak or powerless burnt by malinvestments and may get fright-
to bring about the required adjustments and ened by the growth of competition. The atti-
unless coordinated planning of much more tudes and institutions evolving through de-
than investment is carried out, the invest- velopment will arouse opposition and
 380 INDUSTRIALIZATION STRATEGY

hostility. Some of these resistances may be tween BG as a method and BG as an outcome

overcome only by state compulsion, but the or objective.4 What remains of the doctrine is
governments of and
the the
"softthreat
states" are reluc- the emphasis on the complementarity of mar-
tant to use force of force. Once kets for final goods as an ultimate objective
again, the absence of this type of reaction for investment incentives. But not only is ab-
from the models is both appropriate for sence of markets not normally a serious ob-
Western countries and is opportune to the stacle todevelopment; even where it is, it is
planners in South Asia, but it introduces a by no means the main obstacle and, in any
systematic bias and neglects some of the most case, balanced growth cannot always remove
important issues. it.
Turning now to BG, we have seen that its What is sound in BG is the stress on the
main weakness is that it is concerned with the investment package, on the need for coordi-
creation of complementary domestic markets nation, on the structure of an investment
as an inducement to invest, whereas markets complex. But investment is not the only com-
in the countries of the region can usually be ponent inthis package: and there is too much
created by import restrictions, and, where stress on the complementarity of final mar-
possible, export expansion. This relates to kets. What is needed is a package of policy
measures containing
final goods and principally to consumers'
goods. As far as intermediate markets are
concerned, Nurkse came out in favour of UG 1 . complementary investments;
2. actions to reform attitudes and institu-
(vertical imbalance) in his second Istanbul
tions, including the desire to invest, but
Lecture.2 Social overhead investment pro- also the ability and willingness to work,
vides the conditions and inducements for con-
(which may involve raising consumption),
sequential direct productive investment. As
for horizontal balance, he believed that the to organise and manage and in particular
to administer politically;
case "rests on the need for a 'balanced
3. a carefully thought-out timetable showing
diet."'3 But he later drew a distinction be- the sequence of the various measures
which would be determined by technolog-
2R. Nurkse, Equilibrium and Growth in the World ical, political and sociological factors;
Economy, pp. 259-78.
3"The difficulty caused by the small size of the mar- 4. controls checking undesirable or less de-
ket relates to individual investment incentives in any sirable investments; and
single line of production taken by itself. At least in 5. policies designed to weaken or eliminate
principle, the difficulty vanishes in the case of a more obstacles and inhibitions to development,
or less synchronized application of capital to a wide
range of different industries. Here is an escape from including resistances induced by measures
the deadlock; here the result is an over-all enlarge-
1 to 4.
ment of the market. People working with more and
better tools in a number of complementary projects stems, in the last analysis, from the diversity of
become each others' customers. Most industries ca- human wants. The case for 'balanced growth' rests on
tering for mass consumption are complementary in the need for a 'balanced diet.'" Problems of Capital
the sense that they provide a market for, and thus Formation in Underdeveloped Countries, pp. 1 If.
support, each other. This basic complementarity 4Op. cit., p. 279.
 VLB. LESSONS OF RECENT
HISTORY
VI.B.1. The Industrial
Achievement*

The industrialized countries' manufacturing still relatively low. Exports of manufactures

output has grown at an unprecedented rate in cannot be regarded as an index of manufac-
the last quarter century, but the developing turing development although the character of
countries manufacturing output has grown manufactured exports might be expected to
even more rapidly. Some developing coun- change with industrial development. Small
tries have built up a manufacturing sector to "city states" such as Hong Kong and Singa-
the point at which, though they still have rel- pore, had a high level of manufactured ex-
atively low income, they may be regarded as ports from the beginning of their industriali-
industrialized; a dozen countries are clearly zation process because their size required a
semi-industrialized; most of the developing high degree of specialization to encompass
countries are well into an industrializing adequate economies of scale and it would be
stage; and for the most part only very small unreasonable for them to develop a "bal-
countries with a low population are still non- anced" industrial structure that included all
industrialized. . . . For some, very small de- sectors.
veloping countries, as for some very small Statistical difficulties add to the concep-
high income countries, industrialization re- tual problems. Countries with heavy protec-
mains alimited objective because their com- tion for manufacturing over-value their
parative advantages lies in other sectors. manufacturing output in relation to non-pro-
There have been several attempts at a tected activities and to other countries
typology of industrialization based on the with lower protection. The underpricing of
changing share of value added in manufac- agricultural production to keep down the
turing in the GDP, the proportion of value urban cost of living similarly tends to
added contributed by various manufacturing over-value manufacturing production. The dif-
branches to value added in manufacturing as ficulties, and ensuing differences in the
a whole, value added per unit of capital and measurement of the value added in service
labour employed, the share of exports in industries among countries, make the use of
manufacturing output, and of manufactured conventional ratios of value added in manu-
exports in total exports. These and similar in- facturing toGDP particularly prone to error.
dicators allcontribute to an understanding of Comparing the share of value added in man-
the growth of manufacturing and its role in ufacturing tovalue added in commodity pro-
the economy, but they have different impli- duction may be a more useful indicator, for
cations invarying circumstances. Two large, it eliminates both the difficulties of measur-
traditionally agricultural countries — China ing value added in service industries and the
and India — have each developed a highly- effects of changing weights of service indus-
complex and relatively sophisticated indus- tries in a growing economy. It corresponds
trial structure and they should be regarded as fairly closely to the share of value added in
semi-industrialized although their shares of manufacturing to GDP when the ratio of
manufacturing value added in total GDP is manufacturing to GDP is low, but the two in-
dicators diverge when the ratio of manufac-
*From Helen Hughes, "Industrialization and De- turing is relatively high because services,
velopment: AStocktaking," UNIDO, Industry and
Development, No. 1, 1978, pp. 5, 13-18. Reprinted having declined in relative terms in the early
by permission. stages of industrialization, grow quite rapidly

381
 382 INDUSTRIALIZATION STRATEGY

at the high income levels that accompany a ate commercial development. However, the
high level of industrialization. There are difficulties of distinguishing between tradi-
many difficulties in identifying export man- tional commercial services that are some-
ufacturing data, and in reconciling such data times inimical to industrial development,
with industrial production figures. The differ- and those that complement industrial devel-
ences in the national accounts methodologies opment, have thus far also precluded
of centrally planned eonomies and market measurement.
economy countries make for serious compar- A composite index of indicators of indus-
ison difficulties. Finally, standardization of trial progress would thus be difficult to cal-
data in United States dollars at official rates culate. The best single indicator of the ma-
introduces distortions. turity of industrialization seems to be the
The utilization of human and physical cap- share of value added in manufacturing in
ital is an important aspect of industrial de- commodity production. . . . The ensuing ca-
velopment. The ability to adapt and to inno- tegorization remains artificial, making arbi-
vate is critical to industrial progress and trary demarcation lines in what is essentially
therefore entrepreneurial, managerial and a continuum of experience from countries
technical capacities are perhaps the most im- barely beginning to industrialize to those
portant, and most elusive, equalities in the with mature manufacturing capacity. In this
maturing of an industrial economy. Such ca- context it is important to recognize that the
pacities are in much more ample supply in differences between developing countries at
the industrialized countries than in the non- the ends of the spectrum have become more
industrialized developing countries, but at important in many respects than the similar-
least to date they have not been meaningfully ities between them.
measured, Entrepreneurship, management
and technical expertise require the support of
SEMI-INDUSTRIALIZED
adequate labour and capital resources. The COUNTRIES
supply of labour is not homogenous. Skills
and skill potential are important. Indicators
of educational levels, the only generally (Share of manufacturing 40 to 60 per cent
of value added in commodity production)
available proxy measurements of human cap-
ital, are thus poor reflections of actual human The semi-industrialized countries are con-
capital. centrated inLatin America, the Republic of
The measurement of physical capital is no Korea is a leading example in East Asia, and
less difficult. The availability and use of elec- there are developments in this direction in
tric power is a generally used measure, but the Middle East. Hong Kong and Singapore
again it is a poor proxy of the complex of have only recently graduated from this cate-
public utility facilities required in an indus- gory and, as already suggested, China and
trialized economy and an even poorer mea- India should also be included.
sure of the capital in the sense of fixed in- In general, the semi-industrialized coun-
vestment used in manufacturing. "Capital" tries have relatively well-established private
remains conceptually illusive and the prob- or public enterprise/manufacturing sectors;
lems of measuring it are too well known to they have also clearly overcome the acute
require elaboration. early human and physical capital shortages.
The commercial infrastructure — banking, Their industrial structure is being deter-
import-export, wholesale and retail trade and mined by their resource bases, size, geo-
such auxiliary services as market research graphic location, and policies reflecting com-
and advertising — is also an important aspect parative advantages arising from these
of industrial development. Many of the ex- endowments rather than by progressive im-
ternal economies associated with industrial port substitution. This is particularly so
growth flow from interaction with appropri- where they have taken care not to lock them-
 383 LESSONS OF RECENT HISTORY

selves into an inward-oriented, import substi- significant range of intermediates, notably

tuting, excessively protected framework. The construction materials; and they have begun
export-oriented countries tend to be special- or are beginning to produce capital goods. In-
ized, whereas countries that have followed donesia and Nigeria should probably come
largely import substituting policies tend to into this category although the value added
have 'balanced" industrial development pat- in manufacturing falls short of 20 percent of
terns even though in the early days of indus- value added in commodity production.
trialization this meant inadequate scales of Again, the agricultural sectors of these very
production. large countries swamp the industrial sectors
In relative terms, the geographic concen- that have been established. Except for the
tration of manufacturing is often greater large countries, a balanced industrial struc-
than in industrialized countries, with most ture again tends to reflect inward-oriented
manufacturing plants in a handful of centres import substitution strategies and unduly
where physical infrastructure facilities are small scales of production rather than indus-
trial maturity.
available. Other industrial "poles" are now
being developed, but generally manufactur- The countries in this category have also
ing-led urban concentration is a greater overcome the initial lack of human and cap-
socio-economic problem in semi-industrial- ital resources for industrialization. A pool of
ized than in industrialized countries because indigenous enterpreneurs, managers and
the semi- industrialized countries do not have technicians has been established although it
the stock of infrastructure and levels of in- is not yet adequate for the rapid rate at which
come that make rapid improvement in the these countries wish to expand industrial pro-
urban structure economically, if not always duction. The industrializing countries have
politically, possible in high income countries. also built up at least in some areas, some of
the physical infrastructure — ports, land
INDUSTRIALIZING COUNTRIES transport, water supply and power — neces-
sary for industrial operations, but in most
(Share of manufacturing 20 to 40 per cent cases it is less adequate than in the semi-in-
of value added in commodity production) dustrialized countries. Manufacturing indus-
tries still tend to be clustered in a capital city
The largest number of developing coun- with perhaps one or two other centres, with
tries in Africa, Asia, and Latin America fall concomitant socio-economic problems. Bang-
into the category of industrializing countries. kok, Djakarta and Manila are typical of cit-
Some countries' industrial production is ies with acute urban development problems.
heavily weighted by mineral processing, and However, there are exceptions: in Colombia
in many cases high protection exaggerates and Malaysia, for example, manufacturing
the extent of industrial involvement. has, for historical reasons, developed in sev-
These countries are well beyond their first eral centres in each country, and the prob-
steps towards industrialization. Most pro- lems of urbanization are correspondingly less
duce a substantial proportion of the con- pressing.
sumer goods required by the local market; a
 384 INDUSTRIALIZATION STRATEGY

EXHIBIT VI. 1 . Structure of Production

Distribution of Gross Domestic Product

Industry
GDP (millions of dollars) Agriculture Services
(Manufac-
(percent' 1981
1960 1981 1960 1981 1960 1981 turing)0
1960 1981 1960

Low-Income
Economies 27 w
37 w 25 w 34 w —11 w —16 w 29 w
China and India 48 w 33 w 28 w 39 w 24 w
Other Low- 28 w
Income 48 48w w" 45 w
12 w 17 w 40 w
9w 10 w
1 Kampuchea, 38 w
Dem. —
2 Bhutan —
3 Lao, PDR —
4 Chad 180 — 52 — 12 — 4 — 36 —
5 Bangladesh 11,910 58 7 5 8 35
23 32
3,170 54 16
14
6 Ethiopia 900 65 50 12 6 11
410 3,870 10 55 40
7 Nepal 2,420 34
8 Burma 33 47 12 13 8
1,280 5,770
9 Afghanistan 1,190 3,230 42 10
10 Mali 270 55 11 5 6 35 47
1,120 43 20 13
11 Malawi 170 58 11 6 43
130 1,420 75 36
12 Zaire 30 32 13 3 31 44
43
5,380
13 Uganda 540 12 244 9 4 21
190 9,390 —52 —27 — — 37
14 Burundi 880 56 16 9 28
200 55 41
46 16 16 9 12 29
15 Upper Volta 1,080
16 Rwanda 120 80 7 22 1 16 13
1,260
17 India 29,550 142,010 50 20 26 14 18 30 32
71
18 Somalia 160 — 8 —15 3 — —37
1,230 37
19 Tanzania 550 52 11 5 9 32
21 33
4,350
20 Viet Nam — 57 46
21 China 42,770 264,340 47 35 33 — — 20 20
22 Guinea 370 — — 33 — 4 — 30
1,670 48 43
—
23 Haiti 270 16
1,590 37
24 Sri Lanka 4,120 32 28 20 28 15 44
1,500
160
25 Benin 850 55 44 8 13 3 7
26 Central 13
African Rep. 110 690 10 4 6 37
39 50
27 Sierra Leone — —51 31 —10 20 — 6 — 49
1,040 37 —
28 Madagascar 540 35 4 53
2,890 37 14 38
29 Niger 250 69 30 9 4 8 22 51
1,710 46
30 Pakistan 25,160 30 16 32 12 38 44
3,500
26 17
31 Mozambique —
32 Sudan — 38 —16 — 6 — 49
48
1,160 7,540
33 Togo 120 880 14 8 7 49
29
41
34 Ghana 21,260
55 24
60 10 27
12 — 7 28
1,220
 385 LESSONS OF RECENT HISTORY

EXHIBIT VI. 1. Structure of Production (Continued)

Distribut ion of G ross Domestic Product (percent)

Industry
GDP (millions of dollars) Services
Agriculture (Manufac-
1960 1960 1981 1960
1960 1981 1981 1960 1981 turing)0 1981

Middle-Income
Economies
24 w 14 w 30 w 38 w 22 w 46 w
Oil Exporters 20 w
27 w 13 w 40 w 48 u
Oil Importers 26 >v 17 w 47 w 47 w
23 w 36 w 44 w
14 w 33 w 22 w
15 h> 50 w
Lower Middle-
Income 36 w 22 w 25 w 17
13 w 43 w
35 w 15 w 25
35 Kenya 730
6,960 38 32 18 21 9 15 h> 44 47
36 Senegal 610 24 22 26 12 59 48
2,330 28
37 Mauritania 90 630 44 17
21 3 7 35 h-
39 52
38 Yemen Arab 28 24
Rep. — — 13
— 16 — 6 — 56
— 2,770 — — — —
39 Yemen, PDR 570 28 59
14
40 Liberia 220 930 36 42 8 25
41 Indonesia 84,960 50 25 27 — 12 48
8,670 43
42 Lesotho 30 320 — 24
18
— 21 —15 5 —49
34
37
43 Bolivia 460 26 31 25 13 55
7,900
44 Honduras 300 32 19 27
25 14 44
2,380
45 Zambia 680 37
11 18 63 32 4 18 46
26 50
3,430 17
46 Egypt 3,880 23,110 30 21
26 38 20 41
24 46
41 48
47 El Salvador 570 40 19 20
28 15 15
32 49 54
3,550 32 19 40
48 Thailand 2,550 36,810 13 20
26 24
23 28 20 25
49 Philippines 6,960 38,900

50 Angola — 37
5 1 Papua New 49
Guinea 230 — 13 — 164 —18 38 —
2,580
52 Morocco 14,780 23 16 52
2,040 14 26 40
53 Nicaragua 340 20 27
21 34
33 26 55
50
2,590 24
54 Nigeria 3,150 70,800 63 23 11 5 6 45
47
780 18
55 Zimbabwe 18 35 47
6,010 — — 37 —
56 Cameroon 550 37
20 278 — 53
— 6,270 27
57 Cuba 17
58 Congo, 10
130 23 9 53 5 60 50
People's Rep. 1,870 17
59 Guatemala 38
1,040 8,660 49
41
60 Peru 23,260 18 9 33
20 25
2,410 55
61 Ecuador 970 13,430 26 12 38 15
16
24 15
11 50
62 Jamaica 700 10 8 36 43
2,960 43 23 54
63 Ivory Coast 570 7 12 54 50
64 Dominican 8,670 27 14 37 15
18 23
Rep. 720
50 55
6,650
27 17
27
 386 INDUSTRIALIZATION STRATEGY

EXHIBIT VI. 1. Structure of Production (Continued)

Distribution of Gross Domestic Product (percent)

Industry Services
GDP (millions of dollars) Agriculture (Manufac-
turing)"
1960 1960 1981
1960 1981 1960 1981 1960 1981 1981

65 Mongolia — 40 42
66 Colombia 32,970 26 31 21
3,780
770 34 27
67 Tunisia 24 16
23 18
20 28 178 14 58 47
7,100
68 Costa Rica 510 26 37 14 20 49
2,630
69 Korea, Dem. 54
Rep. —
23 13
70 Turkey 53,910 41 21 32 23 38 45
8,820 26
7 1 Syrian Arab 46
Rep. 890 15,240
— 19 — 31
30 — — 50
72 Jordan — — 8 — 26 — 14 — 62
2,550
73 Paraguay 300 36 28 20 44
5,260
17 17
Upper Middle- 43
Income 20 28
18 w 10 w 33 w 39 w 23 w 24 w 49 w 51 w
74 Korea, Rep. of 14 44
3,810 65,750 37 17
75 Iran, Islamic 39
Rep. of
— 29 — 33 — 11
10 — 38 —
4,120
76 Iraq — 17 — 18 —36 — 46 —
1,580 52
77 Malaysia 24,770 36 23 139 18 31 41
2,290
78 Panama 420 23 10 21 21 10 56 69
3,490
79 Lebanon 830 — 12 — 20 — 13 — 49
68 —
16 49
80 Algeria 41,830 136 35 55 8 11 39
2,740
81 Brazil 14,540 210,660 16 35 26 27 53
16
82 Mexico 12,040 238,960 8 29 19 22 55 55
83 Portugal 21,290 25 12 36 34
37
44 29 39 44
2,340
84 Argentina 12,170 153,330 16 9 38 38 32 25 46 53
85 Chile 9 7 22 48
56 40
58
3,910 32,860 40
86 South Africa 12 7 35 35
53
43 21 23 45
6,980 74,670 45
87 Yugoslavia 24 12 36 35
30 31
9,860 63,350
88 Uruguay 19 8 28 33 21 26 53 59
1,110 9,790 45 15
89 Venezuela 67,800 6 6 22 —16 72 49
7,570
90 Greece 33,390 23 26 20 51 52
3,110
9i Hong Kong 950 27,220 4 —17 39 —31 — 78 —59
11
92 Israel 2,030 17,440 5 36 23
27 26 57
93 Singapore 700 12,910 4 1 32
18 41 12 30 57 58
94 Trinidad and 46
Tobago 470 8 2 13 46 46
6,970 52 24
High-Income Oil
Exporters — — — —
Iw 76 w 23 w
95 Libya 310 27,400 — 2 — 71
78 — 3 — 27
— — — — 4 w —
96 Saudi Arabia 115,430 1 4 20
97 Kuwait — 24,260 — 0 — 71 — 4 — 29
98 United Arab
Emirates — — 1 — 77 — 4 — 22
30,070
 387 LESSONS OF RECENT HISTORY

EXHIBIT VI. 1. Structure of Production (Continued)

Distribution of Gross Domestic Product

(percent)
Industry
GDP (millions of dollars) Agriculture Services
(Manufac-
1960 1981 1960 1981 1960 turing)"
1960
1981 1981 1960 1981

Industrial Market
Economies 6w 40 w 36 w 30 w
3w 25 w 54 w
99 Ireland 16,590 22 — 26 — — — —61 w
1,770 — —41 — —46
100 Spain 11,430 185,080 13 7 36
42 29 52
101 Italy 37,190 350,220 6 31 23
29 53
57
102 New Zealand 25,010 — 11 — 31 — — 58
3,940 20 65
103 United 43
Kingdom 71,440 496,580 3 2 33 32 42
13 45 43
44,000 4 30 54
43 53
104 Japan 1,129,500 46
105 Austria 11 4 39 23
35
34 25 49
6,270 66,240
106 Finland 48,940 7 40 36 48
5,010 57
107 Australia 16,370 171,070
17
12 5 — 23
28 —37 45 —
34 57
108 Canada 39,930 282,500 5 4 32 19 61
46
34
109 Netherlands 11,580 140,490 9 4 41 33 30 64
63
110 Belgium 11,280 96,940 7 2 34 25
24 52
111 France 60,060 568,560 10 4 39 37
35 29 25 62
112 United States 505,300 2,893,300 4 3 38 29 23 58
51 63
61
1 13 Denmark 58,260 11 4 31 34
32 21 19 58 64
5,960 46
114 Germany, Fed. 41
Rep. 72,100 708,540 6 2 53 —15 41 49
115 Norway 57,140 9 5 33 42 21 58
4,630
116 Sweden 13,950 112,420 7 3 40 31 21 53 66
54
117 Switzerland 94,260 27
8,550
East European
Nonmarket
Economies
118 Albania — 48
119 Hungary — 22,560 28 18 — — 33
120 Romania — 48,412
— 13 —39 60 — — — 34
121 Bulgaria —
27
122 Poland —
123 USSR —
124 Czechoslovakia —
125 German Dem.
Rep. —

Source: World Bank World Development Report, 1983, pp. 152-3.

"Manufacturing is a part of the industrial sector, but its share of GDP is shown separately because it typically is the most
dynamic part of the industrial sector.
^Indicates weighted average.
 388 INDUSTRIALIZATION STRATEGY

VLB. 2. Industrialization via

Import Substitution— Note

In many LDCs — especially in Latin America veloped countries can be written in terms of
and Asia — the dominant strategy of indus- industrialization working its way backward
trialization has been the production of con- from the 'final touches' stage to domestic pro-
sumer goods in substitution for imports. duction ofintermediate, and finally to that of
Given an existing demand for imported con- basic, industrial materials."2 At first, the
sumer goods, it was simple to base the post- country may import semifinished materials
war rationale for industrialization on the
and perform domestically the "final touches"
home replacement of these finished goods (in of converting or assembling the almost-fin-
most industries by importing components and ished industrial imports into final products.
engaging in the final assemblying process, in Later on, with the growth in demand for the
the hope of proceeding to "industrialize from final product, a point may be reached at
the top downwards" through the ultimate which the import demand for intermediate
production of the intermediate products and components and basic goods is sufficiently
capital goods). Besides allowing the home re- high to warrant investment in their produc-
placement ofan existing market, import sub- tion at home; the market has become suffi-
stitution also had considerable appeal by vir- ciently large to reach a domestic production
tue of the common belief that it would help
meet the developing country's balance-of- threshold.3
As with any interpretation of historical de-
payments problem. velopment, however, it is one thing to deter-
Although the widespread pursuit of import mine what has happened to make the course
substitution has in practice been based
of development in one country a "success
mainly on the objectives of industrialization story" and quite another to infer from this
and balance-of-payments support, the policy experience that the same result could now be
has been rationalized by a number of protec- induced more rapidly in another country
tionist arguments. Proponents of industrial through deliberate policy measures. The his-
protectionism have adduced several special torical evidence on the contribution of import
arguments in the context of development — substitution to industrialization applies only
arguments that should be considered more to some countries; in other countries, the re-
seriously than the usual simple assertions placement of imports was not significant.
about a "natural" inferiority of agriculture Moreover, we should recognize that the rise
or the supposed necessity of industrialization of industry through import replacement was
to achieve a rising level of income. in large part due to systematic changes in
Support for import replacement comes supply conditions, not simply to a change in
partly from an appeal to the experience of in- the composition of demand with rising in-
dustrialized countries. Historical studies of
come.4 The changes in factor supply — espe-
some countries show not only that the share cially the growth in capital stock per worker
of industiral output rises with development, and the increase in education and skills of all
but also that the growth of industries based kinds — were instrumental in causing a sys-
on import substitution accounts for a large tematic shift in comparative advantage as per
proportion of the total rise in industry.1 It is capita income rose. But for a presently
also true that "much of the recent economic underdeveloped country there is no reason to
history of some rapidly developing underde- 2A. O. Hirschman, The Strategy of Economic De-
velopment, New Haven, 1958, p. 112.
3Ibid.,p. 114.
'For evidence, see H. B. Chenery, "Patterns of In-
dustrial Growth," American Economic Review, Sep- "Chenery, "Patterns of Industrial Growth," pp.
tember 1960, pp. 639-41, 651. 624-5, 628-9, 644.
 389 LESSONS OF RECENT HISTORY

expect that a tariff on industrial imports zation may be initiated through import sub-
would cause the supplies of capital, human stitution, there still remains the problem of
skills, and natural resources to change in a sustaining the industrialization momentum
way that would favor the substitution of do- beyond the point of import replacement.
mestic production for imports. The changes Another special argument for industriali-
in supply conditions that occurred in other zation via import substitution rests on the
countries cannot now be duplicated simply by contention that a peripheral country's de-
a policy of industrial protection. mand for industrial imports increases much
Nor is industrial protection justified by ref- more rapidly than does the foreign demand
erence to the historical pattern of industrial- for its exports, so that the country must sup-
ization working its way backward from the ply all those industrial products which cannot
"final touches" stage to domestic production be imported in view of the relatively slow
of formerly imported materials. On the con- growth of its exports. If we accept the con-
trary, this pattern demonstrates that it is the tentions that there is disparity in the income
growth of imports which subsequently in- elasticities of demand for imports and ex-
duces domestic production; in offering proof ports, that the industrial imports are essential
that a market exists, the imports can fulfill and must be either imported or produced at
the important function of demand formation home, and that the country has no other
and demand reconnaissance for the country's means of increasing its capacity to import,
entrepreneurs, and the imports can act as a then there is prima facie a case for industrial
catalytic agent that will bring some of the protection to encourage import substitutes.
country's underemployed resources together What is relevant for individual primary ex-
in order to exploit the opportunities they have porting countries, however, is not the overall
revealed.5 For the objective of eventually re- income elasticity of demand for primary
placing imports with domestic production, it products, but the prospects for their individ-
would thus be self-defeating to restrict im- ual exports. It is unreasonable to believe that
ports at too early a stage and thereby forgo export prospects are equally unfavorable for
the awakening and inducing effects which foodstuffs, minerals, and raw materials, or
for all commodities in each of these broad
imports have on industrialization.6 An in-
crease in imports — not their restriction — is categories. Moreover, though the elasticity of
the effective way to prepare the ground for demand for a commodity may be low on
the eventual creation of an import-replacing world markets it may be high for the com-
industry. Only after the domestic industry modity from a particular source of supply.
has been established can the country afford Nor can the future demand of industrial
to dispense with the "creative" role played by countries for imports be inferred simply from
imports, and only then would there be a case their income elasticity of demand for im-
for protection of the domestic industry. Al- ports. Their import requirements will also de-
though in promoting the demand for import pend on their growth rates in income (a high
substitutes, restrictions on imports allow the growth rate may offset a low income elastic-
country to bypass the difficulties of having to ity of demand), on shifts of the long-term
build up internal demand simultaneously supply elasticities within the industrial coun-
with supply,7 nonetheless such a protective tries (domestic output of certain minerals
commercial policy is designed merely to re- and fuels, for example, has not kept pace
place imports; this in itself is no guarantee of with demand, so that import requirements
cumulative growth. Even though industriali- are rising relatively to income growth), and
5Hirschman, Strategy of Economic Development,
on the degree of liberalization in the import-
p. 123. ing countries' commercial policies. Without
6Ibid., p. 124. undertaking individual commodity and coun-
7Cf. Gunnar Myrdal, An International Economy,
try studies, it is therefore difficult to gauge
New York, 1956, p. 276. how applicable is the argument for indus-
 390 INDUSTRIALIZATION STRATEGY

trialization because of a weak export social objectives must first be resolved. A pol-
position. icy of industrialization through import sub-
We should also allow for the fact that a de- stitution must also be compared with a pol-
veloping country's capacity to import indus- icy of gradually inducing industrialization
trial products will depend not only on its ex- through agricultural improvement, or pro-
port earnings, but also on the inflow of moting industry through the production of
foreign capital, changes in the terms of trade, manufactured exports (as discussed in
and the capacity to replace other imports VIII. B ). There is also a tendency to exag-
(such as foodstuffs and raw materials) with gerate the amount of employment that could
domestic production. To the extent that these be provided by substituting home manufac-
other factors may raise the capacity to im- ture for imports; as country studies show, the
port industrial products, there is less need for direct employment which can be provided by
industrial protection. replacing imports with domestic manufac-
The case is also weakened if in attempting ture isgenerally limited for a poor country.
to offset the limited demand for exports, the Further, it can be questioned whether in-
policy of import substitution should in turn dustrialization through protection is the best
give rise to limitations on the supply side and remedy for underemployment. The effect of
deter exports. Such a worsening of the export surplus labor in agriculture is low productiv-
situationfinancial may occur when the country's ity, but the remedy for this is capital forma-
scarce and human resources are tion, not industrialization as such. Although
concentrated on industrialization, resources
the surplus labor constitutes an "investible
are diverted from the export sector, home
surplus," this surplus can be applied in var-
consumption limits the available export sup- ious investment outlets, and we cannot sim-
ply, or the industrialization program is ply conclude that the optimum resource use
inflationary. is in import-competing industries. We should
Another facet of the argument for replac- also recognize that other policies might be
ing industrial imports with domestic produc- more effective in stimulating labor mobility
tion is related to the objective of expanding than would protection. When occupational
employment outside of agriculture. It may be mobility is restricted by institutional and cul-
contended that industrialization is necessary tural barriers, the supply responses to the
to provide employment opportunities for the price and income stimuli of protection are
presently underemployed, to absorb man- necessarily weak, and extra-economic mea-
power that would otherwise become redun- sures are required in such forms as education
dant when agricultural productivity rises and training, land tenure reforms, and poli-
through the adoption of more advanced tech- cies that foster cultural change. Finally, we
niques, and to take up the increase in the size must distinguish between the mere availabil-
of the labor force as population grows. ity of surplus laborers and their actual trans-
The promotion of new employment oppor- ference into productive employment as effi-
tunities iscertainly a crucial component of cient and fully committed industrial workers.
development programming, and in this con- This raises all the complex problems of cre-
nection there is considerable point to the em- ating and disciplining an industrial labor
phasis on industrialization. The relevant
questions here, however, are whether invest- A more sophisticated version of the em-
ment should be directed toward import-re- force.8 ployment argument is that indsutry should be
placing industries, and whether industrial
protected by a tariff in order to offset the ef-
protectionism is the most appropriate policy fects of an excessively high wage rate for
for facilitating the expansion of nonagricul- labor in the importable manufacturing indus-
tural employment.
It is possible that the objectives of more 8Cf. W. Galenson (ed.), Labor and Economic De-
velopment, New York, 1959; W. E. Moore and A. S.
employment and a more rapid rate of devel- Feidman (eds.), Labor Commitment and Social
opment are incompatible, and this conflict in Change in Developing Areas, New York, 1960.
 391 LESSONS OF RECENT HISTORY

tries.9 It is claimed that the wage differential in several respects. Insofar as it is concerned
between the agricultural and industrial sec- with absorbing underemployed agricultural
tors overvalues labor for the industrial sector labor into import-competing industries, this
in the sense that industrial wage rates exceed aspect of the argument is subject to the same
the social opportunity costs of employing qualifications raised previously for the gen-
more labor in industry. This may be due to eral argument of expanding employment
the alleged fact that industrial wages are through industrial protection. More point-
based on agricultural earnings, which are de- edly, with regard to the alleged distortion in
termined bythe average product of labor in the labor market, it can be questioned
agriculture rather than by the marginal prod- whether the mere existence of a differential
uct of labor which is lower (compare the dis- between industrial and agricultural wages is
cussion ofLewis's model in III.B.l above); or proof of a distortion. To the extent that the
it may be due to market imperfections that wage differential might be explained entirely
make the gap between agricultural and in- by rational considerations of differences in
dustrial wages greater than can be accounted costs and preferences as between industrial
for by "net advantages" as between agricul- and agricultural work, there is no genuine
tural and industrial work. In either case, distortion.11 Considering the other possible
there is a distortion in the labor market reason for a distortion in the labor market —
which raises the private cost of labor in in- that industrial wages are related to agricul-
dustry above its social opportunity cost (the tural earnings, but these earnings exceed
marginal product of labor in agriculture). the marginal producitivity of agricultural
This results in an inefficient allocation of labor — we must recognize that this result is
labor between agriculture and industry, and based on the assumption of surplus agricul-
it also understates the profitability of trans- tural labor and the ability of the worker to
forming agriculture into manufactures.10 It is receive the average product because the sup-
therefore concluded that protection of man- ply of labor is the family which works on its
ufacturing industry may increase real income own account and not for wages. This consid-
above the free trade level by making the rel- eration isnot relevant, however, for thinly
ative price of manufactures higher and facil- populated countries or for plantation labor.
itating the redistribution of labor from agri- And, regarding the concept of surplus labor
culture toimport-competing industries. and any estimate of its extent, we should re-
This conclusion, however, can be criticized call all the reservations discussed in our ear-
lier analysis of the labor surplus economy
9For a detailed analysis of this argument, see E. E. (III.A).
Hagen, "An Economic Justification for Protection," Even if we assume, however, that the wage
Quarterly Journal of Economics, November 1958, J.
Bhagwati, "The Theory of Comparative Advantage differential does represent a genuine distor-
in the Context of Underdevelopment and Growth," tion, we must still recognize that the effects
Pakistan Development Review, Autumn 1962, pp.
of this distortion may be better offset by do-
342-5; J. Bhagwati and V. K. Ramaswami, "Domes- mestic policies rather than a tariff on indus-
tic Distortions, Tariffs and the Theory of Optimum
trial imports. The difficulty with protection
Subsidy," Journal of Political Economy, February
1963, pp. 44-50. by a tariff is that it seeks to remedy the dis-
10In technical terms, the wage differential against tortion byaffecting foreign trade whereas the
industry causes the feasible production possibility
curve to be drawn inwards within the maximum at-
distortion is in a domestic factor market.12 In
tainable production possibility curve based on a uni-
this case, a policy of subsidization of produc-
form wage. It also makes the commodity price ratio
diverge from the domestic rate of transformation, so "For a list of conditions under which wage differ-
that the optimum conditions characterized by the entials do not represent a genuine distortion, see ibid.,
equality of the foreign rate of transformation, domes-
tic rate of transformation in production, and domestic 12A tariff could make the foreign and domestic
rate of substitution in consumption are violated in the pp. 47-8.
rates of transformation equal, but it destroys the
free trade case. See Bhagwati and Ramaswami, "Do- equality between the domestic rate of substitution and
mestic Distortions," pp. 48-9. the foreign rate of transformation.
 392 INDUSTRIALIZATION STRATEGY

tion of the import-competing commodity, or cepted as a valid policy for establishing an

of taxation of agricultural production, would industry that would eventually be able to pro-
be superior to a tariff.13 A policy of subsidi- duce at lower costs and compete favorably
zation on the use of labor in the import-com- with foreign producers. Nonetheless, to jus-
peting industry, or a tax on its use in agri- tify government intervention, it is not suffi-
culture, would be an even better solution; cient to anticipate solely the realization of in-
since it directly eliminates the wage differ- ternal economies of scale. For if the future
ential, this policy yields a higher real income benefits were to accrue only to the firm, the
than would a tariff, and an even higher real investment might then still be made by a pri-
income than can be attained by tax-cum-sub- vate firm without protection insofar as the
sidy on domestic production.14 A tariff on in- firm can cover its earlier costs of growth out
dustrial imports is thus the least effective way of its later profits. Protection should instead
of offsetting a distortion in the labor market. be based on the condition that the social rate
Although we have so far been skeptical of return exceeds the private rate of return on
about the validity of protectionist arguments the investment. The social benefit is likely to
for import substitution, there remain two ar- exceed the private benefit in an infant indus-
guments that have more merit — the infant try for two special reasons that are particu-
industry case and the attraction of foreign in- larly relevant for a newly developing country:
vestment argument.15 Temporary tariff pro- the knowledge of new industrial protection
tection ofan infant industry is generally ac- techniques acquired in the protected industry
may also be shared with other producers, and
13A policy of subsidization or taxation of domestic
products could equate the domestic and foreign rates the training of the labor force may also re-
of transformation and the domestic rate of substitu- dound to the benefit of other employers.
tion. But since it does not eliminate the inefficiency of When external economies are present, social
labor-use induced by the excessive wage differential,
it achieves this equality along the production possibil-
benefits will exceed private benefits, and
ity curve that is within the maximum attainable pro- market forces would not yield the social op-
duction possibility curve. timum output. To gain the additional bene-
14A policy of tax-cum-subsidy on the use of labor fits, government aid may then be advocated.
could achieve the equality of the domestic rate of
transformation, foreign rate of transformation, and It should be realized, however, that protec-
the domestic rate of substitution, and it can do this tion causes society to bear not only the losses
along the maximum attainable production possibility that would be incurred by the industry dur-
curve. For in this case, the wage differential against ing its period of infancy, but also the cost of
the industrial sector is directly removed, and both the
inefficiency in labor allocation and the divergence of
consumption in the form of higher-priced im-
commodity prices from opportunity costs are simul- port substitutes during this period. The ulti-
taneously eliminated. mate saving in costs, therefore, ought to be
sufficient to compensate the community for
15 We omit the more general external economies ar-
gument and the terms of trade case for protection. the excess costs during the "growing up"
These arguments are analytically correct, but are not
among the most relevant for import-substitution in a period.
poor country. The external economies argument Finally, when the social rate of return ex-
merges with the balanced growth doctrine. There is, ceeds the private, the preferable policy, in a
moreover, no a priori reason why — among all possible way analogous to the other cases of domestic
alternative investment opportunities — we should ex- distortions, would be a direct subsidy on fa-
pect the net external economies to be greatest in im-
port-competing industries. And again, a policy of tax- cilities to further the "learning process" of
new production methods, or provisions by the
cum-subsidy on domestic production may be shown to
be superior to a tariff. Cf. Bhagwati and Ramaswami, government for the training of labor. These
"Domestic Distortions," pp. 45-7. subsidies are superior to a protective tariff,
Although it is possible that a nation may succeed
since they avoid the intermediate loss to con-
in improving its terms of trade by switching produc- sumption that occurs with protection.
tion from exportables to import-substitutes, this pol-
icy has little practical relevance for poor countries Protection may, however, be an effective
that cannot exercise sufficient monopoly or monop- policy for fostering an import-replacing in-
sony power in foreign trade. dustry when its successful establishment de-
 393 LESSONS OF RECENT HISTORY

pends on the acquisition of better technical The limitations and deleterious effects that
knowledge and experience. For when the have resulted in practice from actual import-
country imposes prohibitive tariffs, or other substitution policies will be discussed in
import restrictions, against foreign manufac- greater detail in Chapter VIII. We may sim-
tures, the foreign manufacturer may be in- ply note now that in many instances, the pro-
duced to escape the import controls against tectionist policies have resulted in higher
his product by establishing a branch plant or prices, a domestic product of inferior quality,
subsidiary behind the tariff wall. Although excess capacity in the import-competing in-
the protection would have little effect in at- dustries, and a restraint of agricultural out-
tracting supply-oriented industries, the in- put and on the expansion of exports. A num-
ducement may be significant for the creation ber of country studies can now document the
of "tariff factories" in market-oriented indus- contention that overinvestment has occurred
tries. Itmay be particularly effective in en- in import-replacing industry.
couraging the final stages of manufacture In some countries the promotion of import
and assembly of parts within the tariff-im- substitutes through tariff rates that escalate
posing country when there is an import duty with the degree of processing (low on im-
on finished goods while raw materials or in- ported intermediate goods and high on final
termediate goods remain untaxed. This as- goods) has actually resulted in negative value
sumes, of course, that a sufficiently high do- added. Although high protection of final
mestic demand exists for the product of the goods makes production of the import substi-
tariff factory. And in determining whether tute privately profitable in local currency, the
the attraction of additional private foreign value of inputs at world prices exceeds the
capital provides a net gain, we must again re- value of the final product at world prices; the
call the earlier discussion about the various process of import substitution is socially
costs and benefits of foreign capital. inefficient.
From the foregoing appraisal of the var- Not only has the actual process of import
ious protection arguments, we may conclude substitution been inefficient in resource use;
that they must be highly qualified, the costs it has also often intensified the foreign ex-
of protection not underestimated, and supe- change constraint. At the same time as poli-
rior alternative policies not overlooked. Be- cies have subsidized import replacement,
yond these analytical considerations, the ac- they have inhibited expansion of exports, but
tual experience of many developing countries there has not been a net saving of imports
with industrial protectionist policies also con- since the replacement of finished import com-
firms the conclusion that developing coun- modities has required heavy imports of fuels,
tries are likely to overemphasize the scope for industrial materials, and capital goods, as
replacement of industrial imports. A policy well as foodstuffs in cases where agricultural
of import-substitution industrialization be- development has also suffered.
comes increasingly difficult to follow beyond After a period of import-substitution in-
the consumer goods phase because with dustrialization, theproblems of maldistribu-
each successive import-substitution activity tion in income and unemployment have also
through the intermediate and capital goods become more serious than they were in the
phases, the capital intensity of import-substi- first place. The use of subsidies, overvalued
tution projects rises, resulting in a larger im- exchange rates, rationing of underpriced im-
port content of investment. On the demand port licenses, high levels of effective protec-
side, the projects tend to require increasingly tion, and loans at negative real interest rates
large domestic markets for the achievement have induced the production of import sub-
of a minimum efficient scale of production.16 stitutes by capital-intensive, labor-saving
methods and have resulted in industrial prof-
16David Felix, "The Dilemma of Import Substitu- its in the sheltered sector and high industrial
tion— Argentina," in Gustav F. Papanek (ed.), De-
velopment Policy — Theory and Practice, Cambridge, wages for a labor elite, aggravating inequal-
1968, pp. 60-61. ities in income distribution. As noted repeat-
 394 INDUSTRIALIZATION STRATEGY

edly, employment creation in the urban im- trialization through agricultural develop-
port-replacement industrial sector has not ment, and on the potential industrialization
kept pace with the rural-urban migration, through the export of manufactured prod-
and the unemployment problem has been ag- ucts. Having realized the limitations of an in-
gravated bythe transfer of the rural under- ward-looking strategy of industrialization,
employed into open unemployment and un- many LDCs — as represented through the
deremployment inthe urban sector. pronouncements of UNCTAD — have in fact
Country studies provide some supporting recently changed their emphasis and are now
evidence for the disenchantment with indus- seeking measures that will promote indus-
trialization via import substitution, and at trialization via the substitution of exports of
the same time they reinforce our conclusion processed primary products, semimanufac-
that an LDC must now focus on the possibil- tures, and manufactures instead of exports of
ities of inducing a gradual process of indus- primary commodities.
Comment
The import-substituion strategy of development has been critically examined by many
economists. For an analysis of theoretical issues and empirical results, see B. Balassa, The
Structure of Protection in Developing Countries (1971); M. Bruno, "The Optimal Selection of
Export Promoting and Import Substituting Projects." in United Nations, Planning the Ex-
ternal Sector (1967); M. Bruno, "Optimal Patterns of Trade and Development," Review of
Economics and Statistics, (November 1967); A. O. Hirschman, "The Political Economy of
Import Substituting Industrialization," Quarterly Journal of Economics (February 1968); K.
H. Raj and A. K. Sen, "Alternative Patterns of Growth under Conditions of Stagnant Export
Earnings," Oxford Economic Papers, (February 1961); H. Bruton, "The Import Substitution
Strategy of Economic Development," Pakistan Development Review (Summer 1970); I. M.
D. Little, T. Scitovsky, and M. FG. Scott, Industry and Trade in Some Developing Countries,
1970, chapters 2 and 3; Hollis Chenery, "Comparative Advantage and Development Policy,"
American Economic Review (March 1961); Joel Bergsman, "Commercial Policy, Allocative
Efficiency and 'X-efficiency,' " Quarterly Journal of Economics (August 1974); I. M. D. Lit-
tle, "Import Controls and Exports in Developing Countries," Finance & Development (Sep-
tember 1970); J. Bhagwati, Anatomy and Consequences of Trade Control Regimes (1978);
Anne O. Krueger, Foreign Trade Regimes and Economic Development: Liberalization At-
tempts and Consequesnces (1978).

VLB. 3. Industrialization via

Export Substitution— Note

In contrast with industrialization via import have already indicated how some countries
substitution, there is an increasing interest in have reduced the share of traditional exports
the potentialities of an industrialization strat- in total exports while diversifying into new
egy that emphasizes export substitution. exports made up largely of manufactured
"Export substitution" is the export of nontra- goods (see selections I.C.I and I.C.2).
ditional products such as processed primary This Note considers why the export-substi-
products, semimanufactures, and manufac- tution process has some distinct advantages
tured goods in substitution for the traditional over the import-substitution process. It might
exports of primary products. Other selections be though that in terms of relaxing a coun-
 395 LESSONS OF RECENT HISTORY

try's foreign exchange constraint, a unit of involves the production of labor-intensive

foreign exchange saved by import substitu- manufactures.
tion is equivalent to a unit of foreign ex- Besides private foreign investment, other
change earned by export substitution. But sources of foreign capital — such as commer-
there are other indirect effects and dynamic cial banks that syndicate Eurocurrency loans
considerations in favor of export substitution. to the LDCs — may be more willing to lend to
First, the domestic resource cost of earning a country that promotes exports. This is be-
a unit of foreign exchange tends to be less cause the debt-servicing problem is eased
than the domestic resource cost of saving a when the country has increasing export rev-
unit of foreign exchange. This means that the enue, and the ratio of debt-servicing to export
resources used in import substitution could revenue is falling.
have earned a greater amount of foreign ex- Most important, export substitution con-
change through export expansion than the tributes more than does import substitution
foreign exchange saved in import substitu- to the objectives of greater employment and
tion, which relies on high effective rates of improvement in the distribution of income.
protection. Even though the import substitu- Being labor-intensive in production tech-
tion isprofitable in local currency because of nique, and dependent on the demand of
the high protection, the economic costs in worldwide markets, the nontraditional ex-
terms of real resource used is excessive. The ports may absorb more labor than import re-
value of exports that could be produced with placement. They may also reduce the cost of
a given use of scarce factors is greater than employment in terms of the complementary
the value of imports that could be replaced. use of scarce factors of capital and imported
Moreover, to the extent that it rests on ex- inputs.
ogenous world demand, the process of indus- Export substitution also indirectly aids em-
trialization through export substitution is not ployment creation in the urban-industrial
limited to the narrow domestic market as is sector by avoiding an agricultural bottleneck
the import-substitution process. A developing that would otherwise handicap urban-indus-
economy must overcome the diseconomies of trial employment. By exporting manufac-
small size. And, as classical economists have tures and semimanufactures, the developing
emphasized, "the division of labor is limited countries are able to import agricultural
by the extent of the market." If a country can goods and thereby keep the real wage low as
export to a world market, it can enjoy econ- expressed in terms of industrial goods. If, on
omies of scale, learning effects, and the com- the contrary, there is a slow growth of agri-
petitive gains of X-efficiency.1 "X-efficiency" cultural production, and the price of agricul-
refers to forces that intensify motivation and tural goods rises relative to that of industrial
competition that result in lower cost curves goods, the real wage in terms of industrial
for the firm. goods would rise. This in turn would induce
Furthermore, a pro-trade strategy may a substitution of capital for labor, and it
also attract foreign direct investment. An in- would also reduce profit margins, thereby
flow of foreign capital to support export sub- causing savings to decline and the rate of
stitution isnot dependent on home market capital formation to decrease. Industrial em-
protection but is induced by considerations of ployment would thereby be adversely
efficiency on the side of resource costs. For-
eign investment for export substitution also affected.2
tends to have more linkages to agriculture
when it involves the processing of primary 2Countries have adjusted to agricultural stagnation
products. It also upgrades labor skills when it by expanding food imports by means of labor-inten-
sive industrial exports. Contemporary examples are
Korea and Taiwan, while historical cases include the
'Harvey Leibenstein, General X-Efficiency Theory repeal of the Corn Laws in the United Kingdom and
and Economic Development, New York, 1978. Japanese food imports from her colonies after 1900.
 396 INDUSTRIALIZATION STRATEGY

The evidence from the past two decades an absolute minimum of control. Bargaining
does show that the range of labor-intensive strength is likely to be considerable less for a coun-
manufactures exported from LDCs has in- try manufacturing components or undertaking
deed widened, and the number of LDCs en- middle-stage processing than it is even for a raw
gaged inexport substitution has increased. In material exporter; for copper or cocoa beans can,
after all, be sold on a world market or bartered
conformity with hypotheses about export- with socialist states or even used domestically.
based development, the evidence indicates
What is a country to do, however, with tuners de-
that export growth rates explain a significant signed to meet the particular specifications of
portion of the variance in income growth Philco television receivers — other than sell them to
rates, which cannot be explained by the Philco? Production for export within the multina-
growth in primary inputs; that generally the tional firm may indeed be a means for acquiring a
greatest increase in the GNP of various share of the expanding markets in products for
LDCs is better correlated with exports than which world demand is income-elastic; but it may
any other variables; that the higher-income render the host country exceptionally "dependent"
LDCs have a higher ratio of exports to GDP upon powerful foreign actors; foreign firms and/or
and a faster rate of growth; and that the governments may be in a strong position to influ-
higher rate of growth is correlated with a ence the host countries' policies — both external
and domestic — either directly or through their em-
more diversified export base. ployees or local suppliers where they dominate so
Recognizing the potential for a new indus- utterly particular sectors of their economies. The
trialization strategy, an increasing number of fundamental problem with this dependency rela-
LDCs desire to promote nontraditional tionship isthat continuation or further develop-
exports. ment in the field of these manufactured exports is
The new industrialization strategy has not, subject to the decision of foreign firms over which
however, gone unchallenged. Some analysts the host countries can have extraordinarily little
interpret the promotion of labor-intensive influence — decisions over plant location, new
product development, choice of techniques, market
processes and component manufacturing as
allocations, etc. One might therefore sensibly hes-
the replacement of a nineteenth-century itate before committing oneself overly in this
"plantation society" with the twentieth-cen-
tury creation of a "branch plant society," as direction.3
involving undue bargaining power in favor of To the extent that export substitution rests
the foreign enterprise; or as resulting in an on foreign investment by vertically integrated
unequal international distribution of the transnational firms, it has been criticized by
gains from trade and investment. Thus, Hel- Vaitsos as constituting only "shallow devel-
leiner cautions about the dependence effects: opment." Vaitsos analyzes the export-substi-
tution activities of transnational enterprises
Export-oriented labour-intensive industries sell- as follows:
ing to multinational firms, and totally uninte-
grated with the rest of the economies in which they The basic attraction offered by developing coun-
are located, would seem to combine some of the tries to such export activities by transnational
most disagreeable features of outward orientation firms is obviously due to the very low wages of un-
and foreign investment. Particularly where there skilled labor in such countries given minimum pro-
are "export processing zones," the manufactured ductivity rates. . . .
export sector constitutes an "enclave" — an "out- The development of this type of international
put of the mother country" — in as real a sense as sourcing by transnational firms has important re-
any foreign-owned mine ever did. These disagree- percussions for developing countries. Their com-
able features, moreover, are combined in a manner parative advantage in this case rests in specializing
which leaves the host country with a minimum of in unskilled labor whose wages have to stay com-
bargaining advantage. aratively low while importing a package of inputs
Not only is the export manufacturing activity (both physical and intangibles) from abroad. Since
extraordinarily "foot-loose," dependent as it is on
neither local resources nor local markets, but it is
3G. K. Helleiner, "Manufacturing for Export,
also likely to bind the host country both to sources
of inputs and to market outlets over which it has Multinational Firms July
World Development, and 1973,
Economic
p. 17. Development,"
 397 LESSONS OF RECENT HISTORY

skills and technology, capital, components and them, though, is that in the former the foreign
other goods are mobile internationally while un- investor is not very much captive once he has com-
skilled labor is not (or is preferred not to be, due mitted his activities in a country since the invest-
to the heavy social costs involved), transnational ment isvery low, the shifting of activities to other
firms will be induced to intensify such interna- nations is easy to undertake since there is no
tional sourcing, diversifying their sources of un- uniqueness in the local supply of inputs and the
skilled labor among different developing countries tapping of local resources did not imply the expen-
to assure a continuous availabilty of supply, or the sive discovery of previously unknown resources (as
products of that input. in the extractive sector). Thus, the possibility of
The "shallowness" of such a development pro- enhancing the bargaining power of the host gov-
cess isa result of the following reasons. The type ernment to share in a more equitable distribution
of labor utilized represents generally the weakest of the surplus involved is minimal or non-existent.4
and less organized part of the labor class, thus lim-
iting possibilities for increasing labor returns un- From the preceding quotations, it is clear
less a general shortage of labor takes place in the that the issues raised by a strategy of indus-
country, in which case opportunity cost consider- trialization via export substitution are con-
ations arise for the host economy. If wages in- troversial. They cannot be resolved in isola-
crease foreign investors will tend to shift to other tion. The efficacy of this industrialization
countries since their locational interests stem from
strategy is dependent upon the performance
the existence of low wages given some minimum
of foreign investment, relations between
productivity levels. The training necessary for
local labor in such activities is generally very multinational enterprises and host countries,
small, limiting spill-over effects. Of critical impor- and the trade policies of the developed im-
tance isthe absence of marketing knowhow effects porting countries. This topic should therefore
for the host country since the goods traded are be considered alongside the selections in V.D
within the captive markets of affiliates. Final prod- and V.E, above, and Chapter VIII.
uct promotion is handled abroad by the foreign
centers of decision making.
The concentration on low wage, unskilled-labor-
intensive, export promoting activities has been 4Constantine V. Vaitsos, "Employment Effects of
Foreign Direct Investments," in Edgar O. Edwards
compared to the older enclave structures in the ex- (ed.), Employment in Developing Nations, New
tractive industry. The basic difference between York, 1974, pp. 339-41.

VI.B.4. Interactions Between

Industrialization and Exports*

Sustained economic growth requires a trans- pendence on primary exports toward manu-
formation ofthe struture of production that factured goods as a source of foreign ex-
is compatible with both the evolution of do- change. There is considerable evidence that
mestic demand and the opprotunities for in- success in developing manufactured exports
ternational trade. This transforation nor- is critical to this process, and conversely that
mally involves a substantial rise in the share continued emphasis on import substitution
of industry and — except for a few specialized will ultimately lead to a slowing down of
mineral producers — a shift away from de-
growth.
Despite the amount of attention given to
*From Hollis B. Chenery, "Interactions Between
Industrialization and Exports," American Economic alternative strategies of trade and develop-
Review, Papers and Proceedings, Vol. 70, No. 2 ment since the work of I. M. D. Little, Tibor
(May 1980), pp. 281-7. Reprinted by permission. Scitovsky and M. FG. Scott, there has been
 398 INDUSTRIALIZATION STRATEGY

little attempt to examine the underlying re- lombia, Turkey, Yugoslavia, Mexico, and Is-
lationships inquantitative terms. A fuller un- rael (in ascending order of 1960 per capita
derstanding ofthe various mechanisms that income).2 Japan and Norway, which had
have been posited requires that the internal largely completed the transformation of pro-
and external aspects of industrialization be duction by1960, are added for comparative
examined together in a framework that purposes. The sample was selected primarily
brings out the several interactions among on the basis of the availability of input-out-
them. put data covering fifteen years or more.
This paper is drawn from a comparative Table 1 gives selected structural character-
study of sources of industrial growth in se- istics for the nine countries.
lected semi-industrial countries that have fol-
lowed policies ranging from the extremes of
The Sources of Growth
export promotion to import substitution. The
core of the analysis is a set of input-output An input-output model is used to provide a
accounts that permits changes in the struc- consistent framework for the analysis of
ture of demand, trade, and production to be growth and structural change. The same
analyzed in comparable terms over periods of twenty-three-sector classification is used for
fifteen to twenty years. I will compare the each country, which leads to a comparable
main effects of different types of trade and decomposition of output growth in each sec-
development strategy on industrial growth tor into the direct and indirect effects of in-
and structure. The methodology emphasizes creases indomestic demand, exports, and im-
interrelations on the demand side, which tend port substitution. For this purpose import
to be neglected in other approaches. Atten- substitution is defined for each sector by the
tion is focused on the effects of early or late reduction in the share of total supply that is
development of manufactured exports, which provided by imports.
is a major source of the differences among The model is based on the following ac-
strategies. counting balances for each sector:
Xt =«,W + A) +Et (1)
Mf - miWi ,+sDd (2)
THE TRANSFORMATION OF
PRODUCTION AND TRADE where X is total output, D is domestic final
demand, Wis intermediate demand, M is im-
The structural transformation of develop- ports, Eis exports, m, is the share of imports
ing countries is characterized by a period in in total supply, and w, is the domestic share.
which the rising share of manufacturing in
GNP approaches that of primary production Assuming that W{ = Za^ the level of out-
put can be expressed by the solution to the
and a significant portion of manufactured
corresponding Leontief model as
goods begins to be exported. Countries that
h^ve reached this stage have been altern- Xt - ZnJiUjDj + Ej) (3)
tively described as "semi-industrial" or The coefficients ry are the elements of the in-
"newly industrialized."1 Depending on the verse of a Leontief domestic matrix in which
criteria used, there were between twenty and
the coefficients (ufly) represent the amount
twenty-five such countries by 1970. supplied from domestic sources.
The present sample consists of seven of the
Equation (3) makes it possible to solve for
sixteen principal semi-industrial countries the increase in output of each sector, AXh in
identified by Bergsman: Korea, Taiwan, Co-
2The remaining nine countries identified by Bergs-
'Joel Bergsman identifies sixteen significant semi- man are: Egypt, the Philippines, Brazil, Portugal,
industrial countries, ten of which are the subject of Hong Kong, Singapore, Greece, Argentina and Spain.
the recent OECD study of "The Impact of the Newly Marginal cases include India, Uruguay. Chile, South
Industrialising Countries." Africa, Thailand and Malaysia.
 399 LESSONS OF RECENT HISTORY

TABLE 1 . Indicators of Structure and Growth

Per Capita GNP Share of GDP

Average

Population Level1970)
(U.S.S Growth Manufactured Value-Added
(millions)

Group A in Industry"*
Korea 1955 22 131 1.6rts"
Expo 13.1
1.6 Exports"
0.2
1963
27
149 Rate" 4.9 1.2 16.9
1973 33 323 8.0 31.8 24.3 29.7
Taiwan 1955 9 199 — 8.3 1.4 23.6
17.6 28.6
1963 12 252 3.0 6.2
7.4 43.8
1973 15 513 51.6 38.3
4.9
Israel 1955 2 950 — 11.5 31.6
1963 2 1429 5.2 21.4 11.2 35.5
1973 3 2374 28.3 15.1
5.2
— 36.7
Norway 1955 3 1244 40.7 10.4 35.2
1963 4 2168 7.2 39.0 12.6 33.1
1973 4 3179 3.9 43.4 19.1
30.4
Group B 18
Yugoslavia 1955 329 6.6 2.0 41.7
1963 19 510 5.6 15.6 11.8 40.8
4.8
1973 21 813 22.3 12.5 41.4
Japan 1955 89 500 — 9.1 26.5
10.7
1963 992 9.3 7.4 40.8
97 8.9
1973 108 2349 9.0 10.3 8.8 42.5

Group C
13 12.4
Colombia 1955 285 19.3
1963 17 309 1.0 11.9 0.2
23.2
23 15.5 0.5
1973 415 3.0 24.9
Turkey 1955 24 264 — 5.2 3.7
0.2 16.9
1963 30 319 5.9 0.3 19.0
3.7
1973 38 461 3.8 1.5 24.5
— 8.1
16.7 26.8
Mexico 1955 31 424
1963 40 513 2.4 10.4 3.8 27.2
9.2 1.7
1973 56 719 3.4
3.0 31.1
Source: See Kudo and Robinson.
"Shown in percent.
^Industry includes manufacturing and construction.

terms of increases in internal and external (a) The expansion of domestic demand in all
demand in all sectors (ADj and AEj) and sectors {DD):
changes in two sets of parameters (Awy and
Afly). The solution of AX( can be expressed as j
the sum of four factors:3
(b) The expansion of exports in all sectors
3This formulation is discussed in the paper by (££):
Moises Syrquin and myself. The superscripts refer to
time periods.
 400 INDUSTRIALIZATION STRATEGY

(c) Import substitution in all sectors (IS): analysis of Korea, Taiwan, and Colombia.
Thereafter the patterns diverge substantially.
J
In the four countries in Group A, the
growth of manufacturing is increasingly due
(d) Technological change (TQ: to the continued expansion of exports, which
accounts for 50 percent or more of the total
J k
increase in output. In Korea and Taiwan, ex-
port expansion led to a rapid acceleraton of
The effects of trade policy are shown by industrial growth; but in Israel, Norway, and
terms (b) and (c), export expansion and im- Yugoslavia the demand effects of export ex-
port substitution. When there is no change in pansion were largely offset by import
import proportions or in input-output coeffi- liberalization.
cients, the last two terms vanish and sectoral The countries in Group C are typical of a
growth is determined only by increases in in- large group (which includes India, Brazil,
ternal and external demands.
Chile, Uruguay, and Argentina) whose de-
velopment strategy has been based on import
substitution for several decades (see my book
The Role of Trade
with Syrquin, table 16). The decomposition
Trade and development strategies are of the sources of manufacturing growth
often characterized by a spectrum varying shows that export expansion was the smallest
from inward to outward looking or from "im- of the four factors, accounting for less than
port substituting" to "export led." The direct 10 percent of the total increase. Colombia,
effects of these policy differences on produc- the least industrialized of this group, illus-
tion are shown most clearly by changes in the trates the typical pattern of declining effects
share of manufactured exports, which are of import substitution with no offsetting rise
given in Table 1. Since the sample illustrates in export effects. After import substitution is
a wide variety of development patterns, there largely completed, manufacturing growth
is little difficulty in dividing the countries into cannot exceed that of domestic demand and
three groups on this basis. Group A: Coun- therefore tends to decline until there is a
tries with high or rapidly rising manu- change in trade policy.
factured exports: Korea, Taiwan, Israel,
Norway. Group B: Intermediate cases: Yu- TRADE-DEVELOPMENT
goslavia and Japan. Group C: Countries with
SEQUENCES
low manufactured exports: Colombia, Tur-
key, and Mexico. The previous section established a rough
In the two intermediate cases, manufac- grouping of countries based primarily on the
tured exports rose rapidly before 1960 but role of manufactured exports in the struc-
maintained a relatively high and stable share tural transformation of the economy. I will
of GDP thereafter. now examine the differences in development-
The four sources of growth of all manufac- trade sequences at a less-aggregated level to
turing for these three groups of countries are ascertain the extent to which they vary
given in Table 2. The subperiods are five to among countries and industries. Before doing
ten years, depending on the availability of so, some of the differences in trade policies
input-output data. They extend from the will be noted.
mid-1950s to the early 1970s, except for The seven developing countries all experi-
Japan where it was possible to make approx- neced some degree of balance-of-payments
imate calculations for the prewar period. In disequilibrium during the 1950s, reflected in
all countries except Norway, the data cover foreign exchange shortages and quantitative
part of the initial period of import substitu- restrictions (QRs), and exacerbated by
tion, which is particularly notable in the overvalued exchange rates. The trade and ex-
 401 LESSONS OF RECENT HISTORY

TABLE 2. Sources of Growth in Manufacturing Output -3 -11

Percent of Total
Average -2
Annual Domestic Changes in
Growth Demand Export Import Input-Output
Period Rate Substitution Coefficients
Expansion Expansion

Group A
Korea 70 12 42
1955-63 10.4
1963-70 18.9 30 0--3176 -5 0 -8
57
23.8 39 62 25 -19 2
1970-83
Taiwan 11.2 35 28
44 12
1956-61 49
1961-66 16.6 2 5
-4
1966-71 21.1 35 4 -3 4
Israel 1958-65 13.6 71 57 -7 13 -22
11.3 62 27
49 -1
1965-72 36
Norway 1953-61 5.0 65 15
17
1961-69 5.3 51 58 10

Group B 74 25
Yugoslavia 16.6 6
1962-66 72
1966-72 9.1 70 38 12
Japan 1914-35 5.5 71 33 5 -2 -1
1935-55 2.8 76 15 21
1955-60 12.6 12 15
1960-65 10.8 0
16.5 82
74 22
18 9
1965-70
Group C
Colombia 1953-66 8.3 76 7 22 11
7.4 60 5 4 15
1966-70
Turkey 1953-63 6.4 81
75 2 9 8
9.9 15
1963-68 71 165 10 10
1968-73 9.4
7.0
Mexico 72
86 3 11
1950-60
1960-70 8.6 4 14
7.2 81 8 3 8
1970-75

Source: See Table 1 .

change-rate regimes of four of these coun- tries in our sample. In Group A, Taiwan fol-
tries— Korea, Israel, Colombia and Tur- lowed a sequence similar to Korea and lib-
key— have been compared to a larger sample eralized trade even more fully by 1970. In
by Anne Krueger. While Korea and Israel Group C, Mexico followed a moderate form
show progressive liberalization and reduction of import substitution strategy with relatively
of QRs by the early 1960s, Colombia and low levels of protection.
Turkey maintained high levels of protection In summary, the trade policies of the coun-
and import substituting policies for most of tries in Group A (plus Japan) actively fa-
the period. Both the latter had intervals of vored exports since the early 1960s, while
liberalization and export expansion in the those of Group C discriminated against them
late 1960s. in varying degrees. In addition the transfor-
Yuji Kubo and Sherman Robinson have mation was affected by large inflows of for-
extended this comparison to the other coun- eign assistance to Korea, Taiwan, and Israel
 402 INDUSTRIALIZATION STRATEGY

which made possible higher growth rates and Inward- Looking Sequences. Although
more outward-looking trade and develop- there are some significant differences in tim-
ment policies. ing, the inward-looking countries of Group C
To indicate the differences in trade-devel- indicate the effects of the exhaustion of im-
opment sequences among sectors, the four- port substitution possibilities in all sectors.
teen branches of manufacturing in our The failure to develop manufactured exports
models have been aggregated into three (except on a modest scale in light industry)
groups: (a) light industry (food, textiles, has led to the decline of the rate of growth
clothing, wood products, etc.); (b) heavy in- shown in both light and heavy industry, but
dustry (chemicals, metals, petroleum, etc.), not yet in machinery.
and (c) machinery. Light industry includes Even though Colombia, Turkey, and Mex-
sectors in which both demand and factor pro- ico are relatively large countries and have
portions favor early development while ma- had fairly rapid rates of growth, the expan-
chinery typically develops at a relatively late sion of the domestic market has not offset this
stage. The three principal sources of growth. failure. A more detailed analysis shows these
expressed as percentages of each sector's in- countries lagging particularly in machinery
crease inoutput (as in Table 2), are shown in and metal products, sectors in which the
Figures 1, 2, and 3 for each of these sectors. countries in Group A have had above average

Outward-Looking Sequences. In Group growth.

In summary, the general features of the in-
A countries, each sector shows the same de- ward-looking pattern carry over to each of
cline in import substitution and rapid rise of the major sectors, as in the case of the export-
exports as a source of growth that was indi- led strategy. While the opportunities for im-
cated in Table 2. This shift is earlier and port substitution persist longer in heavy in-
more pronounced in light industry, where im- dustry and machinery and in light industry,
port substitution is only significant in the first its ultimate decline is similar.
period, and takes place last in machinery.
These differences also persist at less-aggre-
gated levels. There is a corresponding change CONCLUDING REMARKS
in the pattern of exports in each country (not
shown) with a growing component of heavy This paper illustrates an approach to the
industrial products. analysis of structural change in which inter-
Despite these differences in degree, the nal and external factors are treated together
major impression from these comparisions is in an interindustry framework. The method
a general similarity in the trade-development is adapted from techniques used in develop-
sequences of each sector in the three devel- ment planning and takes advantage of infor-
oping countries in Group A. After an initial mation collected for this purpose. It can be
period of strong import substitution, export extended to encompass production functions
eApansion became the major sources of in- and factor use by sector in order to provide a
dustrial growth in Korea and Taiwan in each more complete analysis of the source of
major sector, and also led to an acceleration growth and structural change.
of growth in each. In Israel, however, exports The effects of trade policy on industrial
predominated only in light industry. structures that are revealed by this analysis
A final phase of import liberalization in are quite striking. Although import substitu-
each sector is shown by Norway, Israel, and tion isan important feature of early stages of
Yugoslavia. In this phase export expansion is industrialization in all developing countries,
partly offset by increased imports, so that the it can be accelerated or retarded by trade
rate of sectoral growth is determined primar- policy. The later stage of expansion of man-
ily by domestic demand. ufactured exports is more susceptible to pol-
 II I I I I I I I I -a
..Jf. f
-a
-
.. 1 l

B
a-
3C
-
Q B -
© ~

!X_"Ji
-
-.
^l!^L_ 01
o
a O
T3
—
-
r-
■ —
~
n
a
p
< _ r
c
t-

=_-" —-
a z 2:

2 ^ i j a"
—
r-
T 1
_
o
—
\

r--W"1 -\:-L\i >

a-

jujojsj
,• -■
-=
I I I I
-<
TTT
t—- ^-ii
/QMV — = a -
5 &

a : - :
_ -^ _
Z _ : r

i-^Ti 5 2
A 4— J
i

c ■ U I
■IN II III
Q > L_!/__=J'^_

JU3?J?d

403
/-
 404 INDUSTRIALIZATION STRATEGY

Group A countries Norway

Taiwan Israel •is
DD
I
I
/ I
I
I
I 1.-»"
\ .• I■
-t— I-
\ I

\ I

100 200 300 200 300 500 1000 2000 1000 2000 3000
Per capita GNP (U.S. dollars 1970)
6.1 7.1
17.7 21.9 39.3 21.8 42.0 33.9 19.1 14.8
Growth rate (percent)
1Vi
l\
Group B countries Group C countries
Yugoslavia Japan Colombia Turkey Mexico
100
90 •IS
70 DD 1
80 -I
60 11 \ v DD
1
50 1 1
1 1 *"1 IS

"S"
40 £E 'IS
30 /
\/DD
1 \ 1 1 —
20 ' "-EE \ l..»-jEE
10 1 \ 1 1 DD_
-20
-10
0 h
-30 •4 --IS- ...E
-40
E 11 „ -
1 1 1
i!i ! ! f
500 700 500 1000 2000 300 500 200 300 500 500 800
Per capita GNP (U.S. dollars 1970)
18.0 11.1 24.1 13.6 22.6 12.9 13.6 4.0 11.0 14.0 10.6 14.114.0
Growth rate (percent)
DD = Domestic demand EE = Export expansion IS = Import substitution
\is
FIGURE 3. Sources of increase in machinery (percent). EE'

icy influence and is shown to have a large ef- cessful export-led strategies. This process,
fect on the subsequent course of industrial which is explored in my paper with Donald
development. Keesing, has been a major factor behind the
To complete the linkage between indus- growing share of the semi-industrial coun-
trialization and export growth, it would be tries in world markets for manufactures.
necessary to examine the changes in compar- Their success in exporting manufactured
ative advantage that result from the acceler- goods has in turn contributed to more rapid
ation of growth and learning by doing in suc- industrial growth in a cumulative process.
 405 LESSONS OF RECENT HISTORY

References Factors of Economic Growth, New York,

1979.
J. Bergsman, "Growth and Equity in Semi- Anne O. Krueger, Foreign Trade Regimes
Industrialized Countries," World Bank and Economic Development: Liberalization
Staff Working Paper No. 351, 1979.
Attempts and Consequences, New York,
H. B. Chenery and D. B. Keesing, "The 1978.
Changing Composition of Developing
Y. Kubo and S. Robinson, "Sources of In-
Country Exports," World Bank Staff dustrial Growth and Structural Change: A
Working Paper No. 314, 1979.
Hollis B. Chenery and Moises Syrquin, Pat- Comparative Analysis of Eight Coun-
terns ofDevelopment 1950-1970, London, tries," mimeo, World Bank, 1979.
1975. I. M. D. Little, Tibor Scitovsky, and M. FG.
Scott, Industry and Trade in Some Devel-
and , "A comparative Anal- oping Countries: A Comparative Study,
ysis of Industrial Growth," in R. C. O. Oxford, 1970.
Matthews, ed., Measurement, History and
 VI.C. AGRICULTURE-
INDUSTRY INTERACTIONS
VI.C. 1 . Industrialization and

Agricultural Development*

The interrelationships between growth in bility to countries of less than 1 5 million pop-
manufacturing and that in other sectors of ulation, in which imports typically are 40
the economy are critical for industrialization percent or more of the total supply of com-
as well as for overall development. Policy- modity consumption.
makers have long been conscious of the im- The position taken here is midway between
portance to manufacturing of investment in the models of international trade theorists
infrastructure. However, economic analysis and the agricultural economists. [When we
has largely neglected intersectoral links, con- deal] with food supply, agricultural produc-
centrating instead on either macroeconomic tion isassumed not to be subject to interna-
or sectoral and subsectoral issues. . . . tional trade, whereas industrial commodities
This "selection" focuses on what is gener- can be freely imported and exported. The
ally the most important of the connections analysis is conducted in terms of a dual econ-
between industrialization and other sectors — omy where agriculture provides the growing
the relationship with agriculture. urban sector with both workers and wage
Agricultural economists usually analyze goods. Attention is focused on the problem of
the role of agriculture in economic develop- producing an agricultural surplus for the
ment in the context of models of closed econ- urban market and on the connections be-
omies. Foreign trade is absent, and final de- tween economic growth and the agricultural
mand linkages and input-output relations terms of trade. [When we deal] with export
ensure a perfect complementarity in produc- agriculture, the roles are reversed: agricul-
tion between agriculture and industry. Inter- tural output is a traded commodity and in-
national trade theorists, on the other hand, dustry a producer of domestic goods. In-
caution that when trade intervenes, demand dustrialization and export agriculture are
interrelationships need not imply supply discussed in terms of a simplified analysis of
complementarities. Exports and imports may a foreign exchange-constrained economy, in
be large enough to offset the relation between which the industrial sector neither exports
domestic demand and domestic supply of nor competes with imports. The discussion
major commodities. In the stylized view of centers on the consequences of the pattern
agricultural economists, industry and agri- and rate of economic growth of the develop-
culture produce only goods for domestic ab- ing economy of a "squeeze" on agriculture
sorption which for one reason or another can- through foreign exchange controls.
not be subject to international trade. In the The purpose of these semi-open models is
simplified vision of international trade theo- not to discuss the advantages or disadvan-
rists, all agricultural and industrial goods are tages of food self-sufficiency or autarkic in-
internationally traded. Empirically, the first dustrialization, butrather to inquire into the
view aptly describes the economy of large policy issues that arise under such empiri-
countries, but the second has more applica- cally relevant scenarios.
The difficulties associated with policies re-
*From Edmar L. Bacha, "Industrialization and
lying solely on the price mechanism to pro-
Agricultural Development," in John Cody et al.
(eds.), Policies for Industrial Progress in Developing mote a "virtuous circle" of interactions be-
Countries, London, Oxford University Press, 1980, tween industry and agriculture are reviewed
pp. 259-61, 263-6, 268-72. Reprinted by permission. in the third and final section.

406
 407 AGRICULTURE-INDUSTRY INTERACTIONS

I. INDUSTRIALIZATION AND THE labor from subsistence agriculture to market-

FOOD SUPPLY oriented industry. Implicitly, industry is
defined as everything that is modern and
In this section, two-sector models are used
growing, and agriculture everything that is
to study the interrelationships between indus-
try and agriculture. While historically labor traditional and stagnant. Intersectoral prod-
market linkages have been perhaps the first uct flow are ignored because of two assump-
connection between the two, product markets tions: (1) extended family systems in the
eventually became the dominant focus of rural sector supply urban workers with the
their interaction, as described below. The required wage goods through nonmarket
channels: or (2) wage goods are imported at
need to provide food for the growing urban
labor force gives rise to so-called marketed constant prices in terms of domestic indus-
trial goods.
surplus problems. Here these are analyzed in
the context of the Soviet and the Japanese Both assumptions are untenable. The lat-
experiences. In the final part of this section, ter leaves open the question of where the
modern sector is going to obtain the foreign
the thesis that agricultural growth in mixed
market economies is causally linked to the exchange to pay for the wage goods; it does
not recognize that, culturally, wage goods
agricultural terms of trade is discussed in
may be specific to the local economy and
terms of the induced innovation hypothesis
hence not available from foreign sources; and
proposed by Hayami and Ruttan.
it ignores that governments are usually ada-
mantly opposed to letting a significant frac-
The Dual Growth Models tion of basic food consumption be imported
from abroad.
The relationships between subsistence ag- The assumption that migrant workers pro-
riculture and manufacturing industry rarely vide for their own subsistence through infor-
fit the enclave growth model in which a mod- mal rural connections is equivalent to saying
ern export sector develops by importing that the terms of trade between agriculture
labor, capital, food, and raw materials, with and industry remain constant as workers
traditional agriculture remaining as a purely move from country to town. The equivalence
subsistence activity without any connection would follow because under fixed terms of
to the growth of manufacturing. Foreign- trade it can be assumed that the urban wage
controlled plantation economies approxi- is fixed in terms of industrial goods. In fact,
mated this enclave model, as did some highly the urban wage consists of a bundle of wage
capital-intensive, foreign-dominated mining goods in which agricultural, not industrial,
operations, until developing countries began products predominate. The terms of trade
to appropriate the economic rents accruing to will remain constant as labor is withdrawn
such operations. from agriculture only if (1) the marginal
The first connection between agriculture product of labor in agriculture is zero, so that
and industry is established by the use of an total agricultural output remains constant in
unskilled labor force from agriculture in the spite of the labor transfer, and (2) the mi-
industrial sector and its ancillary service ac- grant worker consumes the same amount of
tivities. Astylized view of the labor market agricultural goods as when he was a rural
relationships between town and country in dweller. The second assumption ignores
the early stages of economic development transport and commercialization costs, which
runs as follows. Industry is a labor-using ac- are important, as well as the eventual need
tivity which can draw freely from a pool of for a higher industrial wage to induce work-
surplus agricultural labor. These labor re- ers to migrate. The zero marginal productiv-
sources are made available to industry at a ity assumption has been the object of a long
constant wage in terms of industrial goods. and heated debate in the development liter-
Economic growth is defined as the transfer of ature, with many experts suggesting that this
 408 INDUSTRIALIZATION STRATEGY

is an acceptable presumption only for clearly 1920s showed the the dilemma is particularly
overpopulated rural areas. acute in the case of a crash industrialization
In the absence of informal subsistence program. The issue was further complicated
mechanisms, the need to provide food for the in the USSR because agriculture was based
industrial labor force establishes a product on private peasant farming, whereas industry
market relationship between agriculture and was state-owned and run. Under these con-
industry. As industrial employment grows
ditions, as Preobrazhensky noted, "an ex-
and income per capita expands, the urban de- change ofthe smaller quantity of labor of the
mand for agricultural products increases. [socialist] economic system for the greater
If it can indeed be presumed that under quantity of labor of the [nonsocialist] eco-
these conditions the terms of trade will not
nomic system" was needed to secure a rapid
change, the policy implication is clear — pol- industrial advance from the low initial base.1
icymakers need worry only about industrial This famous "law of primitive socialist ac-
growth. Agriculture will respond swiftly to cumulation" stands for the whole set of gov-
the increasing urban demand by making bet- ernment market controls which serve one
ter use of its partially idle labor and land purpose: to bring about a shift of resources
resources. from the private to the socialized sector over
There is a problem with the assumption and above the share the latter could obtain as
that market-oriented agricultural growth is a result of the operation of competitive eco-
constrained only by lack of demand and that, nomic relations.
if urban demand materializes, rural output The trouble with this scheme is that the re-
will respond along a highly price-elastic sup- lationships between a modern industrial sec-
ply curve. Historically this has been the case tor and a backward peasant agricultural
in only a few countries, and they had strongly sector are not symmetrical. Food is indis-
market-oriented agricultural production. pensable for industry, while the peasants'
Several Southeast and East Asian countries, need for industrial products is secondary, if
some African countries, and such areas in not superfluous. Faced with dwindling sup-
Latin America as Argentina and southern plies of industrial goods and increasing
Brazil fit this pattern. In other countries the claims for their own products, peasants may
"benign neglect" of agriculture and even ac- simply refuse to play the game and step back
tive policies were unable to avoid the grave to a closed subsistence economy. This was the
difficulties that occasioned the voluminous situation in the USSR. The price squeeze
literature on the marketed surplus problem. that resulted from a policy of holding down
food prices was met by the peasants' massive
The Marketed Surplus withdrawal from the market, which threat-
ened to bring the Soviet economy to the brink
The important question is how to guaran- of disaster. Stalin's solution was to step up
tee a continuous supply of agricultural goods forced collectivization of the peasantry.
to the growing urban sector. The neoclassical Through state farming, he managed to break
answer is simple: raise agricultural prices. the peasants' veto power over his decisions on
But this would mean paying more for farm economic policy and so managed to indus-
products and would leave less resources for trialize peasant Russia. However, the human
industrial accumulation. The higher the vol- costs were enormous (Preobrazhensky him-
ume of industrial goods that has to be put self died in the great purge of 1937), and
aside to pay for agricultural inputs (both long-term agricultural productivity growth
wage goods and raw materials), the lower became, and still is, a bottleneck in Soviet
will be the volume of industrial goods that economic development. The lesson seems to
can be added to the capital stock of the mod-
ern sector.
'Evgeny Preobrazhensky, The New Economics
The Soviet industrialization debate of the (London; Oxford University Press, 1965), p. 91
 409 AGRICULTURE-INDUSTRY INTERACTIONS

be that developing countries not contemplat- and industry in a process of concurrent eco-
ing forced output by the rural sector will find nomic growth. But is not this question of in-
it difficult to follow Preobrazhensky's rec- tersectoral financial flows just a red herring?
om endation tofurther industrialization by Should agriculture and industry be analyzed
reducing the relative prices of agricultural as separate entities, as if they were two in-
products. dependent countries? The answer might be in
For an example of how to resolve the terms the affirmative in the USSR of the 1920s,
of trade dilemma, agricultural economists where a socialized industrial sector con-
point to Japan from the Meiji revolution to fronted an antagonistic peasant society. But
World War I. Despite its meager endowment it should not be true for planning processes in
of land, Japan's agricultural and industrial politically integrated developing countries if
development went forward concurrently. planners are concerned with the welfare of
Farm output expanded within the existing the country as a whole rather than with the
framework of small-scale agriculture with re- interests of specific social groups within it.
markably low demands on foreign exchange What this means is that, in principle, the
resources. The major factors responsible for sectoral location of an investment activity
the high rate of growth in agricultural output should not be an issue in planned investment
were the increased productivity and greater decisions. However, the planning process of
utilization of existing land and labor made many developing countries has been charac-
possible by the diffusion of new technology. terized by a considerable degree of urban
While it has been argued that this process did bias. Agriculture is often treated as inher-
not require major capital inputs, the Meiji ently low in productivity, industry as high in
period did see substantial investment in rural productivity. Empirical misconceptions, ide-
infrastructure. It is, however, true that rural- ological biases, and class interests mingle to-
to-urban capital flows occurred. A policy of gether to explain such an antirural attitude.
high land taxes was adopted, which drew a Policy measures are sometimes designed to
substantial share of the increased agricul- deliver subsidized inputs, credits, and exten-
tural productivity for investment in the in- sion activities to farmers to mitigate such
dustrial infrastructure, while avoiding the bias, but in practice these measures largely
disincentive effects of the Soviet experience tend to benefit medium- and large-scale
before collectivization. farmers. This policymaking pattern has been
It is moot whether a similar option exists criticized as one of the main reasons eco-
for contemporary developing countries. More nomic growth in developing countries since
important now are institutional and organi- World War II has failed to reach the poorer
zational reform, infrastructural investments, groups in the population.
and research and development, all highly A pro-rural strategy has been advocated in
complementary inputs in the creation of new its stead as part of a broad reconsideration of
production potentials in agriculture. The the development problem. The reasoning be-
complementarity between infrastructure in- hind this is simple enough. Most of the poor
vestments and investment in research and de- are in the countryside. Given current rates of
velopment raises a serious question about the rural population growth and likely rates of
validity of the assumption that primary em- urban labor absorption, it is argued that the
phasis on scientific progress can provide a rel- poor will remain in the rural sector for a long
atively inexpensive route to rapid growth of time. If alleviation of poverty is the main ob-
agricultural production during the early jective ofdevelopment policy, then it follows
stages of development. that the problem should be attacked at its
Where modernization of agriculture re- root, without waiting for an eventual absorp-
quires heavy initial investments, it is not pos- tion of most of the labor force in the modern
sible a priori to anticipate the direction of in- urban sector. Measures should be adopted to
tersectoral capital flows between agriculture increase the productivity and income of the
 410 INDUSTRIALIZATION STRATEGY

rural poor even if they result in less resources

for urban industrial capital accumulation,
which mostly benefits the middle- and upper-
income groups. Such a strategy is of course
in sharp contrast to the industrialization-first
doctrine with which this analysis of the inter-
actions between agriculture and industry
began.

II. INDUSTRIALIZATION AND

EXPORT AGRICULTURE
Viewed from the intersectoral perspective, FIGURE
A-+ 1 . The trade pattern between agriculture,
the main difference between the analyses of industry, and the rest of the world.
the last section and those of this section lies
in a redefinition of the role of agriculture. cruzeiro price of manufactures relative to the
Earlier, it was described as that of supplying cruzeiro price of machines. If this is done,
industry with food. Here its output is foreign the industrial sector will have more of its own
exchange. product — say, cement — left over after pay-
To facilitate the discussion, consider the ing for the machines and can add that to its
case of a developing country with a budding, capital stock. In practice, the relative price
fully protected domestic-market-oriented in- change can be done by freezing the cruzeiro-
dustrial sector, the material inputs of which dollar exchange rate at a time when the in-
are indirectly supplied by domestic agricul- ternal price of manufactures is being raised
ture through the export of primary products by domestic inflation. This certainly hap-
to the rest of the world. The relevant policy pened in Brazil from the immediate postwar
issues are illustrated by reference to the his- period to the early 1960s. In fact, the price
torical experiences of Brazil and Argentina twist was carried to such an extreme that, by
since the 1930s. the end of the 1950s, the purchasing power of
A triangular trade pattern is established a unit of domestic industrial goods in terms
between agriculture, industry, and the rest of of imported machinery was two and a half
the world, which can be visualized from the times higher than in the late 1940s. As the
Brazilian growth process from the 1930s to exchange rate lags behind industrial prices to
the early 1960s. Agriculture (A) sells coffee increase the purchasing power of domestic
to the world (W) and buys manufactured manufactures over imports, the domestic
products from industry (/), which in turn ac- price ratio of coffee to manufactures goes
quires capital goods from foreign markets. down. Domestic inflation increases the prices
(These relationships are illustrated in Figure of industrial goods, while the cruzeiro price
1.) In its role of foreign exchange supplier, of coffee, under given world coffee prices, ac-
agriculture in fact functions as the machine- companies the more sluggish growth of the
producing sector of the economy. For the cruzeiro-dollar exchange rate.
sake of the argument, take the dollar prices This method of increasing industrial capi-
of coffee and imported machines to be given tal accumulation raises several issues. First,
and assume that coffee production depends the assumption that the external terms of
only on a specific input — say, land — the trade (the dollar price of coffee divided by
growth of which is exogenously determined. the dollar price of machines) are given does
How can the growth rate of domestic indus- not fit the facts. The amount of coffee ex-
try be raised under these conditions? ported byBrazil does affect the international
One clear possibility is to increase the coffee prices.
 411 AGRICULTURE-INDUSTRY INTERACTIONS

Second, the quantity of coffee that Brazil- riculture. Trade policies similar to those ap-
ian farmers produce is a function of the do- plied in Brazil were followed from the early
mestic terms of trade (cruzeiro price of coffee 1940s to the mid-1950s. Argentina was cer-
divided by the cruzeiro price of manufac- tainly an important supplier of meat, wool,
tures). Abundant empirical evidence on the and grains in the world market, but, from a
supply response of Brazilian coffee farmers long-run perspective, foreign demand price
indicates that the quantity produced cannot elasticities for Argentinian products were
be taken as exogenously determined. Coffee high enough to justify the assumption of
production is a profit-making activity that given world prices.
shares with other sectors the available en- The degree of price discrimination in Ar-
dowments oflabor and capital. In the context gentina infavor of industry and against ag-
of the two-sector model under discussion, this riculture can be gauged by the evolution of
means that as the relative price changes, the purchasing power of domestic manufac-
there is some substitution in production be- tures over imports, as previously defined. Ac-
tween coffee and manufactures. Again, this is cording to Diaz-Alejandro, this quotient in-
true for agricultural production more creased byan average of 54 percent between
generally. 1930-39 and 1945-55. Diaz-Alehandro sub-
Third, while a surplus is generated for the mits that ofArgentina's slippagesince
behind other
industrial sector through the relative price countries recent settlement the 1930s
twist, the conclusion cannot be drawn that was causally associated with the price shifts
this surplus will be invested productively. It and related economic policies. He reasons, in
may be wasted as conspicuous consumption terms of the relationships shown in Figure 1,
or else embodied in a form — say, cement — that the industrial sector in Argentina during
with very low marginal productivity indeed. this period would be better described as a
Consider the first two issues together. home-goods sector than as a truly import-
Since Brazil has a monopoly power in the in- substituting sector. Cement, for example, is
ternational coffee market, it follows that up not exported, and domestically it is only a
to a point it is to its advantage to turn the poor substitute for imported machines. But
domestic terms of trade against the coffee this home-goods sector expands by drawing
sector in order to reduce exports and raise resources away from export agriculture,
world coffee prices. In the determination of causing a reduction in the supply of foreign
the optimal price policy from the national exchange to the economy. Additional invest-
point of view, attention has to be paid to the ment in the industrial sector can only take
fact that the long-run foreign demand curve the form of cement, to continue the example,
is more price elastic than the corresponding because foreign exchange is not available to
short-run relation. As a consequence, to keep convert domestic savings into imported ma-
competition at bay and foreign consumers chines. Because cement substitutes poorly for
loyal, the optimal price will be found at a machines, the incremental capital-output
lower level than that which would maximize ratio is very high. Diaz-Alejandro concludes
short-run earnings. Thus if a country has a that in an economy with a severe foreign ex-
monopoly power in international trade, turn- change bottleneck, not even a gross saving of
ing relative domestic prices against exports 20 percent, as was the case in Argentina, will
would be, up to a point, not a squeeze on ag- bring rapid growth. Under such conditions,
riculture but a squeeze on foreigners. The the transformation of savings into tangible
latter could be justified in terms of national machinery and equipment becomes a difficult
welfare. task.
Argentina more than Brazil provides a From this example, it would appear that
pure case where industrial development was for most developing countries the capacity to
achieved by a squeeze on domestic export ag- transform smoothly ex ante savings propen-
 412 INDUSTRIALIZATION STRATEGY

sities into nonconstruction capital goods re- address primarily the more complex indus-
quires either expansion of foreign exchange tries and key social overhead facilities for
earnings or a balanced, in-depth program of which state support is of greater strategic im-
industrialization. portance. Nevertheless, it may be argued
The contrast in the 1940s and 1950s be- that Brazil's strategy was excessively costly
tween the uncertain growth rates of Argen- and that only the country's sheer size and di-
tina, based on the development of light man- versity made it viable. Brazil eventually had
ufactures, and the vigorous expansion of to move to a less inward-oriented policy to
Brazil, based on a better integrated industrial sustain rapid industrial and overall growth,
structure, seems to confirm Diaz-Alejandro's Similar policies have not proved effective in
argument that, given the decision not to pro- the very large countries of South Asia, and
mote exports of either rural or manufactured they have been costly and ineffectual in me-
products, emphasis on light manufacturing is dium- and small-scale countries,
misplaced. A forward-looking policy should

Comment
As the foregoing selection indicates, the interactions between industry and agriculture can
be analyzed under either closed-economy or open-economy assumptions. For additional read-
ings, see Bruce F. Johnston, "The Role of Agriculture in Economic Development," American
Economic Review (December 1961); W. H. Nicholls, "An 'Agricultural Surplus' as a Factor
in Economic Devlopment," (February 1963); W. Owen, "The Double Squeeze on Agricul-
ture," American Economic Review (March 1966); M. June Flanders, "Agriculture versus In-
dustry inDevelopment Policy," Journal of Development Studies (April 1968); Bruce F. John-
ston and Peter Kilby, Agriculture and Structural Transformation (1975), chapter 7. An
excellent survey article is Bruce F. Johnston, "Agriculture and Structural Transformation in
Developing Countries: A Survey of Research," Journal of Economic Literature (June 1970).
For analysis of the role of agriculture in the context of an economy open to international
trade, see Hla Myint, "Agriculture and Economic Development in the Open Economy," in
Lloyd G. Reynolds (ed.), Agriculture in Development Theory (1975).

Comment
The early dual sector models of Lewis, Fei and Ranis, and Jorgenson (Chapter III) inter-
preted economic development as involving a major transformation of the economy from one
which is dominantly agricultural to one containing a large and growing urban-industrial sec-
tor. The inevitable relative decline in agriculture arises from (1) increased specialization in
production, which transfers many nonagricultural production jobs from the farm household
to lrban centers; (2) relatively low-income elasticites of demand for agricultural as compared
with nonagricultural commodities under conditions of rising incomes; (3) high transport costs
of particular agricultural and nonagricultural commodities, which militate against extreme
specialization in agricultural production; and (4) the inconsistency of normal input-output
relationships in a high-productivity, high-income agriculture with the existing population den-
sity of many low-income countries. Such a transformation is emphasized as a means of re-
lieving the burden of population growth and consequent fragmenting of farms and low
incomes.
As labor moves from the agricultural to the nonagricultural sector, there is the problem of
providing food supplies for the growing urban labor force. Creation of nonfarm jobs requires
a large increase in capital in the urban sector. Agriculture may be an important source of
such capital. The populous agricultural sector may also provide needed markets for industrial
 413 AGRICULTURE-INDUSTRY INTERACTIONS

output of consumer goods. The agricultural sector also depends on inputs produced in the
industrial sector.
These interactions between the agricultural and nonagricultural sectors are analyzed by
John W. Mellor, "Toward a Theory of Agricultural Development," in Herman M. South-
worth and Bruce F. Johnston (eds.), Agricultural Development and Economic Growth (1967)
and Bruce F. Johnston, "Agriculture and Structural Transformation in Developing Countries:
A Survey of Research," Journal of Economic Literature (1970), with an extensive bibliog-
raphy. See also selection VII.A.1.

VI.C.2. Complementarity of
Industry and Agriculture*

The case for rapid industrialisation in the practicable unless new employments can be
West Indies rests chiefly on over-population. found for those who would be displaced in the
The islands already carry a larger population process. Until new employments can be cre-
than agriculture can absorb, and populations ated, mechanisation and other increased use
are growing at rates of from 1.5 to 2.0 per of capital equipment will be a doubtful boon,
cent per annum. It is, therefore, urgent to and major improvements in peasant agricul-
create new opportunities for employment off ture will be impossible. . . .
the land. . . . If this impression is right, then agriculture
However carefully the peasant may work in the islands will yield a decent standard of
his holding in staple crops — and, or course, at living only if the numbers engaged in it are
present he is very backward — he cannot drastically reduced, and this will be possible
make 2 or 3 acres yield him a reasonable only if new employments can be created out-
standard of living. He cannot get from 2 or 3 side agriculture. The creation of new indus-
acres even the 6 shillings a day or more which tries isan essential part of a programme for
he hopes to earn in the town, so that it is not agricultural improvement.
at all surprising that there is a continual drift This is not generally realized. There are
to the towns. If peasant agriculture is to be still people who discuss industrialisation as if
put on its feet, the number of peasants must it were an alternative to agricultural im-
be reduced drastically, in relation to the land provement. Incountries where agriculture is
that they now occupy, so that each family not carrying surplus population, industry and
may be able to have a reasonable acreage agriculture are alternatives. New Zealand,
(more equipment to work the land will also on the one hand, or England on the other,
be needed). . . . have to weigh carefully the respective merits
In a word, if West Indian agriculture is to of industry and of agriculture, and to decide
yield a decent standard of living, the number on relative emphases. But this approach is
engaged in the present acreage must be dras- without meaning in the West Indian islands.
tically reduced — it must be something like There is no choice to be made between in-
halved, if current expectations of what is rea- dustry and agriculture. The islands need as
sonable are to be fulfilled. But this is not large an agriculture as possible, and, if they
could even get more people into agriculture,
without reducing output per head, then so
*From W. Arthur Lewis, "The Industrialisation of much the better. But, even when they are em-
the British West Indies," Caribbean Economic Re-
view, May 1950, pp. 1, 6-7, 16-17, 38. Reprinted by ploying inagriculture the maximum number
permission. that agriculture will absorb at a reasonable
 414 INDUSTRIALIZATION STRATEGY

standard of living, there will still be a large Now, if the islands were not short of good,
surplus of labour, and even the greatest ex- cultivable land, the conclusion would be to
pansion of industry which is conceivable put the surplus population on to the land to
within the next twenty years will not create a grow this food. Since they are short of land,
labour shortage in agriculture. It is not the they will have to depend for their food sup-
case that agriculture cannot continue to de- plies increasingly on importing food, and on
velop ifindustry is developed. Exactly the op- putting people instead into producing manu-
posite istrue: agriculture cannot be put on to factures which will be exported to pay for im-
a basis where it will yield a reasonable stan- ported food. . . .
dard of living unless new jobs are created off The islands cannot be industrialised to
the land. . . . anything like the extent that is necessary
Two propositions follow from this. The without a considerable inflow of foreign cap-
first is that, if the West Indian market is it- ital and capitalists, and a period of wooing
self to support a large manufacturing indus- and fawning upon such people. Foreign cap-
try, then the standard of living must first be ital is needed because industrialisation is a
greatly increased. The way to increase the frightfully expensive business quite beyond
standard of living, as we have already seen, the resources of the islands. Light industries,
is to make it possible for each worker to have based on imported materials, will require,
at his disposal the produce of a large acreage; say, £600 of capital or more, for each person
i.e., greatly to reduce the number of persons employed, and heavy industries, using local
in agriculture per 1 ,000 acres, without reduc- materials, may easily demand £2,000 per
ing the total output of agriculture. But this, head. To provide employment for 100,000
as we have also seen, means putting more persons calls for an investment in manufac-
persons into manufacturing industry. Thus turing alone (i.e., excluding all the subsidiary
the full complementarity of industry and ag- activities that will bring employment up to
riculture stands revealed. If agriculture is to 400,000) of something like £130,000,000.
give a higher standard of living, then industry There is no prospect of the West Indies pro-
must be developed. But equally, if industry is viding such a sum out of its own savings; it
to be developed, then agriculture must give a must be borrowed abroad. Then, some people
higher standard of living, in order to provide may say, let it be borrowed by the govern-
a demand for manufactures. The agricultural ment. Let the government set up the facto-
and the industrial revolutions thus reinforce ries, and hire foreign technicians to run them.
each other, and neither can go very far unless It is very doubtful if the governments could
the other is occurring at the same time. raise any such sum on their own credit. But
Those who speak as if the choice in the West even if they could, and even if they were able
Indies lay between agricultural development to hire technicians, the problem would still
and industrial development have failed com- not be solved. For successful industrialisation
pletely to understand the problem. on a big scale is possible only if the islands
The second proposition is that manufactur- can export manufactures. And, since it is dif-
ing industries cannot provide employment for ficult and expensive to break into a foreign
an extra 1 20,000 in the next ten years unless market by building up new distribution out-
the islands start to export manufactures to lets, this is most likely to succeed if the is-
outside destinations. Neither their own grow- lands concentrate on inviting manufacturers
ing demands, nor the replacement of imports, who are already well established in foreign
can provide a large enough market. The do- markets. If an industry is to supply only the
mestic market for manufactures is too small local West Indian market, it is quite feasible
to support more than a fraction of what is for the government to start and run it. But
those industries which are to try to sell in
needed.
the local For, at the ispresent
demand standard
for food, rather of*living,
than for Latin America, or in the United States, or in
manufactures. England, are better left to foreign capitalists.
415 AGRICULTURE-INDUSTRY INTERACTIONS

EXHIBIT VI. 2. Growth of Production, 1960-1981

Average Annual Growth Rate (percent)
turing
Industry
GDP Agriculture Manufac- Services

1960-70 1970-81 1960-70 1970-81 1960-70 1970-81 1960-70 1970-81 1960-70 1970-81

Low-income
economies 4.6 W 4.5 w 2.3 m 6.6 m 3.6 m 5.4 m 2.9 m 4.2 m 4.6 m
China and India 4.5 w 4.8 w 1.8 m 2.4 m 8.3 m 6.4 m — — 5.2 m 4.8 m
4.6 m
Other low-income 4.7 w 2.7 m m* 2.3 m 6.6 m 3.2 m 5.9 m 2.8 m 4.2 m
3.6 w
2.2
1 Kampuchea,
Dem. 3.1
— — — — — — — — —
2 Bhutan — — — — — — — — — —
3 Lao, PDR — — — — — — — — — —
4 Chad 0.5 — — — — — — — —
4.2 —
8.0
5 Bangladesh 3.7 4.2 9.0 11.2 5.3
2.4 6.6 7.8 4.2
2.7 -0.4
6 Ethiopia 4.4 2.2 2.2 7.4 1.8 8.0 2.8-2.3
7 Nepal 2.5 — —
0.9 — — — —-9.
4.6 — —-0.7
2.1
4.8 3
8 Burma 2.6 4.1 2.8 -0.8
5.6 1.5 4.7
9 Afghanistan 2.0 3.9 — 4.7
3.2 — -9.8 —
3.4 2.8 — 5.3
4.0 — — — —
10 Mali 3.3 -4.6
0.2 — -0.8 3.2
2.4
11 Malawi 4.9 -1.6
5.6 — — — — — — — —
5.9

12 Zaire 3.4
— 1.5 — — —
13 Uganda 5.6 — — — —
14 Burundi 4.4 — 2.2 — 8.5 — 5.9 — 3.5
3.2 —
15 Upper Volta 3.0 — — 2.9 — 5.8
3.6 -0.6 1.4 3.4
16 Rwanda 2.7 5.3 — — — — — — —
4.6 —
17 India 3.4 3.6 1.9 4.4 4.7
4.0 5.0 5.2
18 Somalia 1.0 —
1.9 5.4 — -0.1 — 4.2 —
3.9 —
19 Tanzania 6.0 — —
3.4 2.2 — 2.9
5.1 — — — —
20 Viet Nam 3.8 — — —
5.5 — — 5.4
-0.6
21 China 5.2 5.5 1.6 28 11.2 8.3 — —
22 Guinea 3.5 3.0 — — — — — —-4.3
7.6 —
5.7 —
4.4
23 Haiti 0.2 1.1 0.2 7.1 1.1
4.6 3.5
4.6 3.4
4.3
24 Sir Lanka 3.0 3.0 6.6 4.2 2.1 5.0
25 Benin 2.6 3.3 — — — -—3.6 —
6.3 — — —
26 Central
African -3.0 4.0
Rep. 1.9 1.6 0.8 2.3 1.8 4.5
27 Sierra Leone 4.3 — —
5.4 —
5.4 — (■)
28 Madagascar
1.9
0.3 — 2.4
0.3 — 0.3 — —
-1 —
7.0 0.4
29 Niger
2.9
3.3 13.9 11.4 — — .0
3.7 (•)
2.9 3.1 4.9 -2.2 -14.4
0.4
30 Pakistan 4.8 2.6 10.0 5.5 9.4 6.1
6.9
6.7
31 Mozambique — — — — — — — — — —
32 Sudan 1.3 -04.1
.2 — 2.3 — — 1.5 — 6.0
33 Togo 8.5 — 1.5 — 3.2
6.2 — — 3.2
34 Ghana 2.1
3.2
— 0.0 — — —
Middle-income
5.5 m 0.4
economies 7.4 m 6.1
6.0 w 5.6 w 3.4 m 3.0 m 6.8 m 6.7 m 5.9 m 7.2 mm
Oil exporters 3.3 m 7.4 m 7.6 m 7.4 m 8.7 m 4.8 m
6.3 w 6.2 w 3.4 m 5.7 m
Oil importers 3.5 m 7.0 m 5.9 m 6.5 m 5.6 m 5.7 m
5.8 w 5.4 w 2.9 m
416 INDUSTRIALIZATION STRATEGY

EXHIBIT VI. 2. Growth of Production, 1960-1981 (Continued)

Average Annual Growth Rate (percent)
turing
Industry
GDP Manufac- Services
Agriculture

1960-70 1970-81 1960-70 1970-81 1960-70 1970-81 1970-81 1960-70 1970-81

1960-70
-4.0
Lower middle- 7.4 m 5.8 m
3.2 m 6.8 m 7.1 m 5.3 m 6.0 m
income 5.6 h
35 Kenya 5.8 — 4.2 — 8.5
4.1
— 9.5 — 6.0
5.0 w-
5.9
-0.7
36 Senegal 2.5 2.0 3.0 mb
2.9
2.6 4.4 2.0
4.6 7.4 0.9
37 Mauritania 1.4 14.1 6.2
9.2 1.7
6.7 1.7 3.1
38 Yemen Arab — — — 13.9 — 12.1 — 5.2
8.7 3.6
39 Yemen, PDR — — — — — — — — — —
11.7

40 Liberia 1.3 — 4.0 — — 5.6 —

4.8
5.1
41 Indonesia 3.9 7.8 5.2 -0.4
11.2 3.3 13.9 9.5
1.4
2.7
— 4.3
3.8 — — —
42 Lesotho 5.2 12.9 9.6 9.2
8.4
43 Bolivia 5.2 3.0 2.9 6.2 4.9 5.3 4.6
5.2
4.4
44 Honduras 5.3 1.9 5.4
4.5 4.7 4.8
5.4
3.8 5.7 5.4 3.7
45 Zambia 5.0 — 1.8 — 7.6
— 0.3 — 1.2
0.4
46 Egypt 4.3 8.1 2.9 5.4 4.8 11.8
2.9 8.7 7.5
47 El Salvador 5.9 3.0 2.3 3.3 8.8 6.5
4.7 3.3-0.8
3.1
7.2 4.5 8.5
48 Thailand 8.4 5.6 11.9 9.9 11.4 2.4
10.3 9.1
49 Philippines 5.1 6.2 4.3 4.9 6.0 5.2 5.3
8.4 6.9
50 Angola — — — — — — —
6.7 — — —
5 1 Papua New -—
Guinea 6.7 1.9 -0—.4
0.4 — — —
4.2 — — —
4.2
52 Morocco 4.4 5.2 4.7
7.8
0 4.4 6.4
7.3 0.8 5.8
2.1 11.4 5.4
2.8 5.8
53 Nicaragua 2.7 10.4
54 Nigeria 3.1 4.5 6.0 9.1 2.3 7.4
14.7 12.4
-1.8
55 Zimbabwe 4.3 1.8 — — — — — — — —
65 Cameroon 6.3 — — — 5.3 — 6.8
57 Cuba
3.7
— — — 3.9
— — —
9.4 — — — —
58 Congo, -2.6
7.3
-3.
People's
Rep. 2.3 5.1 1.8
4.3
7.4
7.8 13.6 6 7.4 1.1 2.2
4.3 2.1
59 Guatemala 5.6 5.5 8.2
-1.2 5.9 5.5 5.4
60 Peru 4.9 3.0 0.3 5.0 2.9 5.3
3.7 — — —
61 Ecuador — 8.6 — 2.9 3.4
12.5 5.7 4.6 3.5
10.8
62 Jamaica 4.4 1.5 0.5 4.8 8.7
4.2 5.8
63 Ivory Coast 8.0 6.2 11.5 9.3 11.6 ()
4.7 5.7 5.8 9.7
64 Dominican 7.6
Rep. 4.5 6.3 2.1 6.0 5.0 6.1 5.0
3.2
— — — — — — — — — 6.7
—
65 Mongolia
66 Colombia 5.1 3.5 4.7 6.0 4.7 7.8 7.5
6.8
7.3
5.7 2.0 5.7 4.5
5.7
67 Tunisia 4.7 4.1 8.2 9.3
9.4 7.4 11.7
5.7
7.1 5.2
68 Costa Rica 6.5 5.2 10.6
.57 2.2
69 Korea, Dem.
Rep. — — — — — — — — —
5.7 —
70 Turkey 6.0 5.4 2.5 3.2 9.6 6.9
6.1 10.9 5.5
71 Syrian Arab 6.2
Rep. 4.6 10.0 — 8.2 — 9.8 — 8.2 — 10.8
72 Jordan — — — — — — — — — —
4.2 — 7.0 — 11.0 — —
73 Paraguay 8.8 8.1
9.1
 417 AGRICULTURE-INDUSTRY INTERACTIONS

EXHIBIT VI. 2. Growth of Production, 1960-1981 (Continued)

Average Annual Growth Rate

turing
(percent)
Industry
GDP Manufac- Services
Agriculture

1960-70 1970-81 1960-70 1970-81 1960-70 1970-81 1960-70 1970-81 1960-70 1970-81

Upper middle- 7.8 m 7.1 m 6.5 m

income 4.0 m 2.6 m 8.8 m 4.5 m 6.3 m
6.4 w 5.6 w
17.6 8.2
74 Korea, Rep. of 8.6 9.1 4.4 17.2 14.4 15.6
3.0 -1.1
75 Iran, Islamic 8.9
Rep. Of 11.3 — 4.4 — 13.4 — 12.0 — 10.0 —
76 Iraq 6.1 — — — 5.9 — 8.3 —
7.8 5.7
77 Malaysia -5.4 — 5.2 —
4.7 9.3 — 11.1 —
7.6
6.5 4.6
78 Panama 7.8 2.0 4.1 10.5 8.5
5.7 10.4
2.7 5.3
79 Lebanon 4.9 6.3 — 4.5 —
7.6 5.0
7.8 — 4.8 —
-0.8 11.6 11.6
80 Algeria 4.3 6.9 0.1 3.9
81 Brazil 8.4 — 5.2 — — 7.1 —
7.3 8.3
6.4
5.4 6.5 4.5
82 Mexico 7.6 9.4 9.1
7.4 8.7
4.5 6.6
10.1
4.4 1.3 3.4 6.0
83 Portugal 6.2 8.8 4.4 8.9
4.6
5.9
84 Argentina 4.3 1.9 1.8 2.5 5.8 5.6 2.2
85 Chile 2.1 3.0 4.4 1.4 5.5 0
0.7 3.8
4.4 3.1
— — — —
0.7 — — — —
86 South Africa 6.3 4.3 2.9
5.8 3.7 3.3 6.8 7.1 6.9
87 Yugoslavia 5.7 2.6 6.2
88 Uruguay 1.2 3.1 1.9 -1.2
3.0 1.1 5.7
1.5 1.0 2.9
5.6
3.5 5.3 7.3
4.6
89 Venezuela 6.0 4.5 5.8 4.5 7.1 5.9
90 Greece 4.4 3.4 9.4 2.7 6.4
10.2 5.5
6.9 3.5
91 Hong Kong 10.0 9.9 — 1.7
-1.8
— — — 10.1 — —
5.2
92 Israel 8.1 4.0 — — — — — — — —
93 Singapore 8.8 8.5 5.0 12.5 9.0 13.0 7.7 8.5
1.7 9.7
94 Trinidad and
Tobago 4.0 5.5 — — 4.0 — 1.3 —
-2.2 6.9
High-income oil
— 5.3 w — 7.1m — 3.1m — 9.2 m — 12.2 m
exporters
95 Libya 24.4 2.3 — 10.5 — — — 17.1
— 14.7 —
96 Saudi Arabia — 10.6 — — 3.1
10.2 6.5 12.2
7.1
5.3 — — —
97 Kuwait 5.7 2.3 — 9.2 9.6
98 United Arab
Emirates — — — — — — — — — —

Industrial market
economies 1.4 m 1.6 m 5.7 m 2.9 m 5.9 m -0.m5
3.1 4.6 m
5.1 w 3.0 w 4.3 3.6 m
99 Ireland 4.2 4.0 0.9 — — — — —
100 Spain
7.1 — 2.1 —
6.1 — 6.0 — 4.5
3.2 3.9
101 Italy 5.5 2.9 2.6 1.3 8.0 5.1
102 New Zealand — — —
6.6 —
2.9 — —
3.7 — —
3.2
3.6 2.0
103 United
2.9 1.6 3.1 3.3 2.8
Kingdom 1.7 2.2 0.4 4.2
4.5 0.2 13.0 5.6 13.6 6.5
104 Japan 10.4 10.2
4.6 2.1
105 Austria 3.5 1.2 1.9 3.2 5.2
.05 0.0 5.4 6.1 3.4 4.0
5.0
4.4
106 Finland 4.3 3.9
3.1 — 5.2 —
3.3 — —
107 Australia 5.6 3.3 5.9 5.5 3.6
2.0 3.7 4.3
108 Canada 2.5 1.8 6.3 2.9 6.8 5.5
5.6 3.8
3.2
 418 INDUSTRIALIZATION STRATEGY

EXHIBIT VI. 2. Growth of Production, 1960-1981 (Continued)

Average Annual Growth Rate (percent)

turing
Industry Manufac
GDP Services
Agriculture

1960-70 -0.5 1970-81 1970-81

1970-81 1960-70 1970-81 1960-70 1970-81 1960-70 1960-70

109 Netherlands 5.2 2.8 6.8 2.0 6.6 2.6

3.0 4.6
5.1
2.7 3.9
110 Belgium 4.7 3.0 5.5 6.2 3.5
3.7
0.7 7.1 3.1
111 France 5.5 3.3 1.6 0.5 2.3 7.8 3.2 5.0 4.2
112 United States 4.3 .05 1.6 4.6 2.7 5.3 2.9 4.4
4.6
2.9 3.3
1 13 Denmark 4.5 2.1 0.1 2.6 5.2 1.1 3.1 2.5
-1.1 5.2
114 Germany, Fed.
Rep. 4.4 2.6 1.5 1.3 4.8 — 5.4 2.1 4.2 4.5
2.5
1 1 5 Norway 4.3 4.5 2.2 5.5 5.0 4.8 1.3 5.0
1.8 0.7
116 Sweden 4.4 0.8 6.2 0.8 5.9 3.9
117 Switzerland 4.3 — — — — — —
0.7 — —
2.7
0.7
East European
nonmarket
economies — — — — — — — — — —
118 Albania — — — — — — — — — —
119 Hungary 5.3 5.0 2.9 6.3 5.8 6.5 5.9 5.8 5.0
3.2
120 Romania 8.6 9.1 1.7 4.9 12.8 9.2 — — — —
121 Bulgaria — — — — — — — — — —
122 Poland — — — — — — — — — —
123 USSR — — — — — — — — — —
124 Czechoslovakia — — — — — — — — — —
124 German Dem.
Rep. — — — — — — — — — —

Source: World Bank, World Development Report, 1983, pp. 150-51

"Indicates weighted average.
Indicates median value.

VI.C.3. Overcoming the

Weakness of Agriculture*

Agriculture has been the weakest link in the erywhere bright until one turns to agricul-
development chain. Industry in LDCs has ture, where the dominant fact is that, in
grown at around 7 percent per annum, the LDCs as a whole, food production has failed
number of children in school has multiplied to keep pace with the demand for food,
by four, the domestic savings ratio has risen thereby causing or aggravating a whole series
by three percentage points — the picture is ev- of other problems.
The basic reasons for this failure are well
*From Sir W. Arthur Lewis, "Development Strat- known, so I will list but not dwell on them.
egy in a Limping World Economy," The Elmhurst The first has been fast population growth.
Lecture, The International Conference of Agricul-
tural Economists, Banff, Canada, September 3-12, Population has grown at around 2.5 percent
1979 (processed), pp. 2-9. Reprinted by permission. per annum, and per capita demand has
 419 AGRICULTURE-INDUSTRY INTERACTIONS

pushed the growth of total demand well be- while industrial prices rose all the time. This
yond three percent, while output has grown was anomalous, since prosperity usually im-
at significantly less than 3 percent, turning proves agriculture's terms of trade. The basic
what used to be an export surplus into an im- factor was the enormous increase in agricul-
port surplus of food. tural productivity in the United States, re-
Secondly, the technological revolution in sulting in the build up of stocks of cereals;
tropical food production has only just begun, since agricultural commodities compete with
research in the colonial days having been each other either on the demand side or on
confined almost but not exclusively to com- the supply side, this depressed all other agri-
mercial crops exportable to the world mar- cultural prices. Add to this that in several
ket. We have made spectacular progress with LDCs governments wanted to keep farm rev-
maize, wheat for subtropical conditions, and enues low, whether by imposing taxes on ex-
rice for areas of controlled irrigation, but portable crops, or by placing price ceilings on
have still far to go with other rice, with sor- food for the domestic market. This is at first
ghums, and millets, and with livestock sight a curious phenomenon. One would ex-
management. pect that farm populations, being more than
Third, even where there is new technology half the nation (in most cases), would carry
to impart, the agricultural extension services enough political clout to be able to defend
and the network for supplying modern inputs themselves against such measures — and
to the farmer — especially seeds, fertilisers would on the contrary be manipulating the
and pesticides, are gravely deficient, and in terms of trade in their favour, but this is not
many areas virtually non-existent. automatic. European farmers were doing this
Fourth, investment in rural infrastructure at the end of the nineteenth century, but the
is inadequate. Road systems have improved contemporaneous efforts of American farm-
immensely, and the penetration of the coun- ers— though they were still in the majority —
tryside bybuses and trucks is altering the were a failure.
patterns of rural life. But not enough has Let me now turn from the causes of the low
been invested in irrigation, or in storage level of agricultural output in the LDCs to
facilities. some of its effects. Agricultural failure is not
Fifth, everyone speaks in favour of land re- the sole cause of the problems I shall men-
form, but very few governments have done it tion, but makes in each case a significant
in any of its various forms, whether distrib- contribution.
uting land to the landless, or converting from Take first the probability that inequality of
rental to ownership tenures, or fixing rental the income distribution has increased along
ceilings. The case for some sort of land re- with recent growth. This is not a novel phe-
form remains unquestionable from the stand- nomenon. Increased inequality is inherent in
point ofjustice; the case from the standpoint the classical system of economics because
of its effects on production is now stated with population growth keeps labour income down
greater sophistication, recognising the extent while profits and rents increase. Given the
to which higher output is tied to improved long and strident debate between economic
technology, extension and investment. Indeed historians as to what happened to the Euro-
several writers now speak not of land reform pean living standards in the first half of the
but of "the land reform package," to distin- century, no modern economist should have
guish what they see as good land reform from assumed that economic growth would auto-
bad land reform. matically raise the incomes of those at the
And finally to complete our list of factors lower end of the scale. Rapid population
that have inhibited agricultural output we growth has also played its negative role in our
must add poor terms of trade. The prices of day, restraining the wage level and farm in-
agricultural commodities in world trade fell come per head. Since the majority of the la-
throughout the 1950s and most of the 1960s, bour force in LDCs consists of farm people,
 420 INDUSTRIALIZATION STRATEGY

who also have the lowest incomes, the stan- make the countryside economically viable,
dard of living of the great bulk of the popu- with a larger cultivated area, with rising pro-
lation can be raised only by raising farm in- ductivity on the farms, more rural industry,
come. Decisions of the effects of growth on and better social amenities.
income distribution or income distribution on
Note "the larger cultivated area." Devel-
growth lead nowhere unless farm income is at opment economists have been mesmerized by
the centre of the alleged relationship. European experience into assuming that the
The worst effects of population growth development process always involves a de-
combined with technological standstill are to cline in the number of persons in agriculture.
be seen in the arid zones of the tropical This is true of relative decline, but it extends
world, where some 500 million people live, es- to an absolute decline only in the later stages
pecially along the fringes of the African and of development. For example, around 1850 in
Asian deserts. There we have the largest con- Western Europe the agricultural population
centration ofhuman poverty; the numbers was only 50 percent of the whole, and the
continue to grow rapidly; and we have not yet rate of natural increase about one and a
had the technological breakthrough in dry quarter percent. So the agricultural popula-
farming that might promise higher produc- tion would decline absolutely if the non-ag-
tivity. To raise the living standards of these ricultural population grew at over 2.5 percent
hundreds of millions is the greatest challenge a year. Whereas with 70 percent in agricul-
to those who work for development. ture and a rate of natural increase of 2.5 per-
Consider next the huge flow of migrants cent, an absolute decline of the agricultural
from the countryside into the towns. Central labour force requires non-agricultural em-
to this of course is the growth of population. ployment to expand at 8.3 percent per
Relatively under-populated countries can annum, which it cannot do.
cope with population growth by opening up An increase in the absolute numbers en-
new land, as has been happening over much gaged in agriculture is therefore an essential
of Africa, but in less favoured countries pop- item in coping with the current flood of pop-
ulation growth means smaller farms, more ulation. The fact that the green revolution in
landless labourers and lower output per head. cereals is labour-intensive helps, especially if
Unless a green revolution is set in motion, the the natural propensity of the more enterpris-
natural reaction of farmers caught in this sit- ing farmers to invest in labour saving ma-
uation isto put pressure on the young to mi- chinery can be restrained. But there is no es-
grate to the cities, which they will do if the caping the need to bring more land under
cities show signs of expanding employment. cultivation, by opening up roads, irrigation,
This is not a complete solution. The towns terracing, drainage, and other investment in
cannot provide employment for the whole of infrastructure. Some governments are ac-
the natural increase in the countryside, not to tively engaged in colonisation schemes of this
speak of women now also leaving the family sort, which, if highly planned to meet modern
ta^ks and seeking wage employment; so un- standards, are costly and troublesome. The
employment mounts. The government is also subject is neglected in our textbooks. It needs
trapped. The towns exert great pressure for more research and experimentation, leading
expansion of the public services — of water, to action.
bus transport, schools, hospitals, and so on — A third consequence of the weakness of ag-
eating up more funds than exist, and leaving riculture isthat it is one of the reasons why
nothing to spend in the countryside. So the so many LDCs have had balance of payments
differential in amenities between town and troubles, have incurred large external debts,
country widens all the more, and the stream or have found themselves defaulting on their
of migrants is increased. Unemployment in obligations. It is not just that a larger output
the towns cannot be ended by spending more would earn more foreign exchange, or save on
in the towns. The basic solution is rather to
food imports. Indirectly it would reduce ur-
 421 AGRICULTURE-INDUSTRY INTERACTIONS

banization, the high cost of which is the mestic production were also low, for reasons
prime cause of their needing so much capital which we have already considered. So import
and having to borrow so much. Also, in coun- substitution of manufactures, which was the
tries suffering from the two-gap disease, it starting point of industrialization, was lim-
would facilitate the translation of domestic ited by the narrowness of the domestic mar-
saving into foreign exchange. ket. LDCs soon discovered that if industry is
A fourth and final consequence of the to grow at 7 percent per annum, in the face
weakness of agriculture has been to inhibit of a peasantry with only a small marketable
the growth of manufacturing industry be- surplus, industry must look to foreign mar-
cause of the farmers' low purchasing power. kets. By the year 1970 this lesson had been
The physical output of LDC commercial ex- learnt, and nearly every LDC had begun ex-
port crops grew rapidly, aided on the supply porting some manufactures to developed
side by the expansion of internal transport, countries. Unfortunately this range was very
and on the demand side by the unusually narrow, dominated by textiles and clothing;
rapid growth of the developed countries. But broadening only as the protests and restric-
the prices at which these commodities sold tions of MDCs forced the more advanced
were poor; exports are a small part of agri- LDCs into light metals, electronics and other
cultural output, so their prices are linked on fields. The LDC effort was clearly successful,
the supply side to the price of food, which as since LDC exports of manufactures were
we saw earlier, was depressed by American growing at ten percent a year, despite the
surpluses. The individual LDC can do well barriers erected by the MDCs. Whether
out of exporting agricultural raw materials or world trade will revive, and if so whether
tropical beverages; but for the group of LDCs LDC exports of manufactures will again
as a whole the elasticity of supply of these grow at ten percent are crucial questions for
commodities is so high, at prices yielding LDC development strategy. . . . But no mat-
roughly the same incomes as domestic food ter how they may be answered, it will be to
production, that the factoral terms of trade the advantage of LDCs to raise their agricul-
stay much the same despite increases in de- tural productivity, since this would simulta-
mand or improvements in technology. The neously raise the living standards of their
road to riches does not run in these farmers, create a domestic market for their
directions. manufactures, and improve their terms of
At the same time farm incomes from do- trade.

Comment
The preceding two selections indicate a shift in the thinking of Lewis with respect to the
role of agriculture vis-a-vis industry. The first of the two selections by Lewis (VI. C. 2), written
in 1950, focused on the labor surplus in agriculture and emphasized putting more persons into
manufacturing industry. So too did Lewis's 1954 dual-sector model, examined in Chapter III.
The last selection (VI. C. 3) is written against the background of a fast population growth, the
poverty of subsistence farmers, the migration from the countryside, and unemployment and
underemployment. It emphasizes the necessity to absorb labor in agriculture.
Agriculture is also now emphasized in order to increase the productivity of food producers
and thereby raise the export prices of commercial tropical crops, to increase the demand for
domestically produced manufactures, and to reduce the balance-bf-payments deficit by in-
creasing export revenue from primary products and by saving foreign exchange through im-
port substitution for food imports. As Lewis states, "the most important item on the agenda
of development is to transform the food sector, create agricultural surpluses to feed the urban
population, and thereby create the domestic basis for industry and modern services." For
elaboration of his latest position, see W. Arthur Lewis, The Evolution of the International
 422 INDUSTRIALIZATION STRATEGY

Economic Order (1978); Growth and Fluctuations 1870-1913 (1978); and "Development
Strategy in a Limping World Economy," The Elmhurst Lecture, Banff, Canada, September
3, 1979.
The next chapter examines policies that might accomplish Lewis's objectives for agricul-
tural development.
CHAPTER VII

Agricultural Strategy

We have repeatedly referred to the removal of the agricultural bottleneck as a strategic policy
issue: the attainment of a proper balance between the establishment of industries and the
expansion of agriculture is a persistently troublesome problem for developing nations. In the
earlier days of development planning, deliberate and rapid industrialization was often advo-
cated. Experience, however, has shown the limitations of overemphasizing industrialization,
and it is increasingly recognized that agricultural progress must have a vital role in the de-
velopment process. The earlier confrontation of industrial development versus agriculture has
been shown to be a false issue, and the concern now is rather with the interrelationships be-
tween industry and agriculture and the contribution that each can make to the other.
Agricultural progress is essential to provide food for a growing nonagricultural labor force,
raw materials for industrial production, savings and tax revenue to support development of
the rest of the economy, to earn more foreign exchange (or save foreign exchange when pri-
mary products are imported), and to provide a growing market for domestic manufactures.
As the theme of section VILA emphasizes, the intersectoral relations between agriculture and
industry will determine the course of structural transformation in a developing economy: if
in the longer run, there is to be a diminishing share of agriculture in output, there first must
be in the short run successful policies of agricultural development to facilitate this
transformation.
Another theme in this chapter is the special effort that must be made to help the rural poor
take advantage of the potential now provided by the green revolution. Section VI I. B examines
some of tire issues raised by the new seed-fertilizer technology, especially policies that might
increase yields among small-scale, low-income producers of traditional crops, and policies that
might provide more employment for small farmers and landless labor. With the present con-
cern over absolute poverty and unemployment, rural development takes on a new importance
in its own right — and not simply for its instrumental value for industrial development. Agri-

423
 424 AGRICULTURAL STRATEGY

cultural development is now crucial for amelioration of absolute poverty and for labor
absorption.
Even with allowance for the imprecision of the data, the general enormity of the problem
can be sensed from the World Bank's estimates that, of the 800 million people in LDCs who
are in absolute poverty, some 80 percent, or 640 million, are in the rural areas. Given the
magnitude of the numbers in rural poverty and the incapacity to reduce significantly these
numbers by any feasible amount of rural-urban migration, the rural proportion of those in
poverty is likely to remain high in many LDCs.
Only about a third of the population in developing countries, however, live in countries that
have had a satisfactory performance in agricultural production.
If past trends in food production and agriculture simply continue, the outcome by the year
2000 will be not only unsatisfactory but alarming. As a study of the Food and Agriculture
Organization concludes, trend production performances in the majority of developing coun-
tries would perpetuate the weakness of the agricultural basis of economic development and
would not provide the food supplies needed as a prerequisite to relieve many millions of people
of their grossly inadequate nutritional conditions. Between 1980 and 2000, it is projected that
the food and agricultural experience in a large part of the developing world will be charac-
terized bysemi-stagnation and a tendency to deficits in output with the majority of the coun-
tries showing little improvement in per capita production.1
For most of the developing countries, rural development on a massive scale will be needed
for decades. But the progress of rural development will be crucially dependent on the outcome
of the green revolution, measures of land reform, land settlement at the extensive margin,
new forms of rural institutions, and various special programs designed to increase the pro-
ductivity and incomes of the rural poor. The selections in section VII. C therefore focus on
various components of a rural-based development strategy.
While the green revolution provides a potential for increasing farm output by technical
innovations that increase the productivity of labor and land, it is necessary — from the stand-
point of labor absorption — to realize the advantages of a labor-using and yield-increasing
strategy of agricultural development. In this connection, the transformation of Japanese ag-
riculture isinstructive. The "Japanese model" has three important characteristics: first, ag-
ricultural output has been increased within the unchanged organizational framework of the
existing small-scale farming system. This was possible because of increases in the productivity
of the existing on-farm resources of land and labor, and was associated with small demands
on the scarce resources of capital and foreign exchange. Second, the bulk of the nation's farm-
ers have been involved in increases in agricultural productivity associated with the use of
improved varieties of seeds, fertilizers, and other current inputs; and technological progress
of this type has continually been the source of greater agricultural productivity. Third, agri-
cultural and industrial development moved forward together in a process of "concurrent"
growth.2
Emphasis is therefore placed in selection VII. C. 1 on the relevance of a "unimodal strategy"
such as occurred historically in the Japanese pattern of a small-scale farming system and the
participation of the bulk of the nation's farmers in the increase of agricultural productivity.
Such a strategy, however, must also be linked to changes in the land tenure system, techno-
logical development in agriculture, and persistent concern with the effects on employment
generation.

'Food and Agriculture Organization of the United Nations, Agriculture: Toward 2000, 1981, chapter 2.
2Kazushi Ohkawa and Bruce F. Johnston, "The Transferability of the Japanese Pattern of Modernizing
Traditional Agriculture," in Erik Thorbecke (ed.), The Role of Agriculture in Economic Development, 1969,
pp. 277-8.
 425 AGRICULTURAL STRATEGY

Higher per capita output must be realized. But to do this, the selections in section VII. C
emphasize a number of requirements. It is necessary to use more "modern" inputs per unit
of land, while also providing for an increasing total use of labor. There must also be improved
incentives for farm producers, especially through pricing policies that do not overvalue inputs
to farmers while undervaluing their output. Better distribution of ownership or access to pro-
ductive resources is also required, especially for small landholders, and the distribution of
income must be improved.
 VILA. IMPORTANCE OF
AGRICULTURE

VILA. 1. Agriculture's
Contribution to Development-
Note

Ever since at least the time of Ricardo, the duction items from other sectors at home or
abroad; (2) selling some of its product, not only to
"theology" of development has emphasized
that agricultural progress contributes to the pay for the purchases listed under (1) but also to
support of greater productivity throughout purchase consumer goods from other sectors or
the economy. In his Principles of Political from abroad, or to dispose of the product in any
Economy and Taxation, Ricardo viewed the way other than consumption within the sector. In
all these ways, agriculture makes it feasible for
problem of diminishing returns in agriculture
as crucial. He believed that a limitation on other sectors to emerge and grow and for interna-
tional flows to develop; just as these other sectors
the growth of agricultural output set the and the international flows make it feasible for the
upper limit to the growth of the nonagricul- agricultural sector to operate more efficiently as a
tural sector and to capital formation for eco- producing unit and use its product more effectively
nomic expansion. as a consuming unit.1
It is now customary to summarize in four
The "factor contribution" occurs "when
ways how greater agricultural productivity there is a transfer or loan of resources from
and output contribute to an economy's devel- the given sector to others. Thus if agriculture
opment: (1) by supplying foodstuffs and raw itself grows, it makes a product contribution;
materials to other expanding sectors in the if it trades with others, it renders a market
economy; (2) providing an "investible sur- contribution; if it transfers resources to other
plus" of savings and taxes to support invest- sectors, these resources being productive fac-
ment in another expanding sector; (3) selling
tors, itmakes a factor contribution."2
for cash a "marketable surplus" that will In this traditional interpretation, the de-
raise the demand of the rural population for
products of other expanding sectors; and (4) velopment process is viewed as one of struc-
tural transformation from an economy in
relaxing the foreign exchange constraint by which agricultural employment and output
earning foreign exchange through exports or dominate to a decline in the share of the
by saving foreign exchange through import labor force in agriculture and a decrease in
substitution.
the share of agriculture in GNP. But this
Kuznets summarizes these contributions as
structural transformation is itself dependent
the "market contribution" and the "factor
on agricultural progress. Industrial develop-
contribution." ment will be cut short by lack of agricultural
A given sector makes a contribution to an economy progress — unless the economy is in the ex-
when it provides opportunities for other sectors to ceptional situation of being able to export
manufactures for imports of foodstuffs and
emerge, or for the economy as a whole to partici-
pate in international trade and other international raw materials (compare, Hong Kong). In
economic flows. We designate this contribution the Lewis' dual-sector model, we saw that if food
market type because the given sector provides such supplies to the modern sector do not keep up
opportunities by offering part of its product on do- with the modern sector's demand for labor,
mestic or foreign markets in exchange for goods the modern sector will have to consume a
produced by the other sectors, at home or
abroad. . . .
'Simon Kuznets, Economic Growth and Structure,
Thus agriculture makes a market contribution New York, 1965, pp. 244-5.
to economic growth by (1) purchasing some pro- 2Ibid., p. 250.

427
 428 AGRICULTURAL STRATEGY

larger share of its output in feeding its labor The increase in farm output in Japan between
force, and this will leave a smaller surplus for the 1880's and 191 1-20, which seems to have been
capital accumulation. More generally, it is of about the same magnitude as the growth of de-
mand during that period, corresponded to an an-
widely believed that "both in concept and in nual rate of increase in demand of approximately
practice it is possible for the agricultural sec- 2 percent. With current rates of population growth
tor to make large net transfers of resources to
and a modest rise in per capita incomes, the annual
other sectors. If these transfers are used pro- rate of increase of demand for food in a developing
ductively, the rate of economic growth can be economy can easily exceed 3 percent, a formidable
accelerated."3 challenge for the agriculture of an underdeveloped
Agriculture's contribution of foodstuffs — country. Moreover, as a result of the expansion of
the "wage good" in classical terminology — is population in cities and in mining and industrial
clear. If the labor force for manufacturing or centers dependent upon purchased food, the
another expanding sector is drawn from ag- growth of demand for marketed supplies is a good
deal more rapid than the overall rate of increase.
riculture, the new workers must "take their Thus there are additional problems in developing
lunch" with them when they leave the rural transportation links and marketing facilities in
sector. A growing urban labor force must be
order to satisfy the requirements of the nonagri-
supported by an expanding supply of food- cultural population.
stuffs. A growing population must also be If food supplies fail to expand in pace with the
supported with increased food supplies. The growth of demand the result is likely to be a sub-
annual rate of increase in demand for food is stantial rise in food prices leading to political dis-
given by D = p + rjg, where p and g are the content and pressure on wage rates with conse-
rate of growth of population and per capita quent adverse effects on industrial profits,
income, and r\ is the income elasticity of de- investment, and economic growth. There is scant
evidence concerning the price elasticity of demand
mand for agricultural products.4 As indi- for food in underdeveloped countries. At least in
cated by Johnston and Mellor,5 not only are the case of an increase in prices as a result of de-
there high rates of population growth in the
LDCs, but the income elasticity of demand mand outstripping supply, there is a strong pre-
for food in these countries is considerably sumption that the price elasticity for "all food" is
extremely low, probably lower than in economi-
higher than in high-income countries — prob- cally advanced countries. Cheap starchy staple
ably on the order of .6 or higher in the low- foods — cereals and root crops — provide something
income countries versus .2 or .3 in Western like 60 to 85 percent of the total calorie intake in
Europe, the United States, and Canada. A low-income countries, so there is relatively limited
given rate of increase in per capita income scope for offsetting a rise in food prices by shifting
therefore has a considerably stronger impact from expensive to less costly foods; and the pres-
on the demand for agricultural products in sure to resist a reduction in calorie intake is strong.
The inflationary impact of a given percentage
the lower-income countries than in the eco-
increase in food prices is much more severe in an
nomically advanced countries.
Johnston and Mellor observe that: underdeveloped country than in a high-income
economy. This is a simple consequence of the dom-
inant position of food as a wage good in lower-in-
come countries where 50 to 60 percent of total con-
sumption expenditure is devoted to food
consumption compared with 20 to 30 percent in
3John W. Mellor, "Accelerated Growth in Agricul-
tural Production and the Intersectoral Transfer of Re- developed economies.
sources," Economic Development and Cultural Owing to the severe economic and political re-
Change, October 1973, p. 5. percus ions ofa substantial rise in food prices, do-
4K. Ohkawa, "Economic Growth and Agriculture," mestic shortages are likely to be offset by ex-
Annals Hitotsubashi Academy, October 1956, pp. panded food imports, provided that foreign
45-60.
exchange or credits are available.6
5Bruce F. Johnston and John W. Mellor, "The Role
of Agriculture in Economic Development," American 'Ibid., p. 573.
Economic Review, September 1961, pp. 571-81.
 429 IMPORTANCE OF AGRICULTURE

Through the transfer of capital and labor ing price controls on agricultural products,
to nonfarm activities, agriculture may also taxation, or the use of multiple exchange-
provide an investible surplus. The transfer of rates that discriminate against agriculture. If
labor has been repeatedly discussed in the the improvement in the terms of trade in the
context of the Lewis model and needs no fur- nonagricultural sectors raises nonagricul-
ther attention here. But it should be noted tural incomes, and the beneficiaries save at a
that agriculture can be a source of capital higher marginal rate than the decreased ag-
formation in ways other than the simple lend- ricultural incomes, aggregate saving rates
ing of voluntary savings. There may be a will increase, and agriculture will have made
compulsory transfer from agriculture for the a net contribution to total saving in an indi-
benefit of other sectors, ordinarily through rect manner. The next selection gives some
taxation in which the burden on agriculture indication of what has actually occurred in
is greater than the governmental services recent years by way of income transfers from
provided to agriculture. Kuznets remarks: agriculture.
A "marketable surplus" from agriculture
The measurement of such forced contributions is needed not only to provide the wage good
of agriculture to economic growth is not easy; the to industry, but also to widen the home mar-
incidence of some indirect taxes is difficult to as- ket for the industrial products. The demand
certain and the allocation of government expendi- for industrial products depends on growth of
tures interms of benefits to agriculture and to eco- farm cash income, unless the country can ex-
nomic growth elsewhere is far from simple. But port its growing industrial output. Barring
this factor contribution by agriculture was clearly unlimited export possibilities, and with 70 to
quite large in the early phases of economic growth 90 percent of the home market in the rural
in some countries. Thus in Japan in the last two sector, the nature of rural demand will affect
decades of the nineteenth century the land tax was
over 80 percent of central government taxation, the growth of non-farm employment and out-
and the direct tax ratio to income produced was put. Increased agricultural productivity, a
between 12 and 22 percent in agriculture, com- growing marketable surplus, and rising real
pared with from 2 to 3 percent in the nonagricul- income are necessary to raise the rural sec-
tural sectors. Forced extraction of surplus from ag- tor's demand for industrial output.
riculture by taxation, confiscation, and other Finally, agriculture may be a major source
measures also probably financed a considerable of foreign exchange. It is clear that agricul-
part of industrialization in the Soviet Union. In- tural exports dominate in a country's early
deed, one of the crucial problems of modern eco- phase of development. But also important in
nomic growth is how to extract from the product relaxing the foreign exchange constraint is
of agriculture a surplus for the financing of capital the possibility in several developing countries
formation for industrial growth without at the
same time blighting the growth of agriculture, to save foreign exchange by replacing im-
under conditions where no easy quid pro quo for ports of foodstuffs with home production. Ex-
such surplus is available within the country. It is port promotion and import substitution are
only the open economy, with access to the markets activities not only for the industrial sector but
of the more highly developed countries, both for also for agriculture.
goods and for capital loans, that can minimize this Considering these various contributions of
painful task of initial capital accumulation.7 agriculture, development economists have in-
sisted that if there is to be in the longer run
Another way of transferring resources a structural transformation in output and
from the agricultural to the nonagricultural labor force, there must first be in the short
sectors is by the government turning the run "successful policies of agricultural devel-
terms of trade against agriculture by impos- opment" to facilitate this transformation.
But what specifically do "successful policies
of agricultural development" entail? And is
7Kuznets, Economic Growth, pp. 250-51. the purpose of agricultural development sim-
 430 AGRICULTURAL STRATEGY

ply to underwrite the expansion of nonagri- ment services. The tariff, supported as providing
cultural sectors — even at the expense of an better employment opportunities for the agricul-
tural population, may itself be a major instrument
"agricultural squeeze"? Now, in view of the whereby agricultural real incomes are depressed.
emphasis on absolute poverty and the em-
Government may also operate to depress agricul-
ployment problem, is it not necessary to con- tural real incomes by imposing its taxes, mainly or
centrate on agricultural development for the largely, directly or indirectly, on agriculture, and
sake of employment and a diminution in in- directing its expenditures mainly to the benefit of
equality? Even though the longer-term objec- the urban population. Even though the rural pop-
tive is structural transformation — the ab- ulation may have lower per capita incomes than
sorption of a larger fraction of the rural the urban, it may nevertheless be the only econom-
population in new income-earning opportu- ically healthy part of the population, the only part
nities— there remains the complex problem which gives good value to the community in
of the timing of this transformation and the ecxhange for what it gets from the community.
Where the situation is one — as it often is— of
intertemporal sequence of policies to accom-
plish it. The lessons of recent history have urban exploitation of the rural population, to pro-
pose as a remedy the further subsidization of
shown that an "urban bias" can discriminate urban industry as a means of drawing rural work-
against agriculture;8 and that, as illustrated ers to the city is equivalent to proposing to remedy
in the next selection, the net outflow of re- the exploitation of worker bees by the drones by
sources from agriculture may be excessive. transforming the worker bees also to drones. It is
Not only may there be an inefficient use of obvious that it can work at all only as long as there
the resources transferred to the nonagricul- still remain worker bees in the fields to be
tural sectors, but the transfer may itself be at exploited.
the expense of more employment and higher The refutation of bad argument does not nec-
income in the agricultural sector. Should not essarily refute the conclusions reached by such ar-
gument. Itis not my position that the path to eco-
the "growth-promoting interactions between nomic progress is not, for many countries and even
agriculture and industrial development" for most countries, by way of industrialization and
mean more than that agricultural develop- urbanization. I have in fact conceded that as any
ment should have simply instrumental value country or any region becomes more prosperous it
for industrial development? will normally tend to increase the ratio of its pop-
In the early years of development plan- ulation which is non-agricultural. My position is a
ning, Viner wisely anticipated the answer to different one, and I will now state it frankly and
this question: positively for the first time. The real problem in
Let us now suppose that real incomes are lower poor countries is not agriculture as such, or the ab-
sence of manufactures as such, but poverty and
in agriculture than in industry, and that by tariff
backwardness, poor agriculture, or poor agricul-
protection or subsidies industry can be made to ex- ture and poor manufacturing. The remedy is to re-
pand and to draw workers from the country into
move the basic causes of the poverty and back-
the cities. Is this sound economic policy? wardness. This is as true in principle, and probably
The correct answer depends on why per capita
real incomes are lower in agriculture than in man- nearly as true in practice, for industrialized coun-
tries as for predominantly agricultural countries.
ufacturing. There may be urban exploitation of ag-
Misallocation of resources as between agricul-
riculture, through monopolistic pricing by employ- ture and manufactures is probably rarely a major
ers, or through labour monopolies in the factories
cause of poverty and backwardness, except where
which by forcing wages up force up also the prices
which the agricultural population has to pay for government, through tariffs, discriminatory taxa-
urban products and services — including govern- tion and expenditure policies, and failure to pro-
vide on a regionally non-discriminatory pattern fa-
cilities for education, health promotion, and
8Michael Lipton, "Strategy for Agriculture: Urban technical training, is itself responsible for this mis-
Bias and Rural Planning," in Paul Streeten and Mi- allocation. Where there is such government-in-
chael Lipton (eds), The Crisis of Indian Planning, duced misallocation, it is today more likely to
London, 1968, chapter 4. consist of the diversion of agrarian-produced
 431 IMPORTANCE OF AGRICULTURE

resources to the support of parasitic cities than of carefully husbanded, and the sore disappointment
overinvestment of resources in primary industries of the wishes of the great masses of population
and in workers in such industries.
crying to be relieved of their crushing poverty.9
Economic improvement may call for greater in-
dustrialization, but this should be a natural The emphasis on agricultural development
growth, appropriately facilitated by government now is not only for its instrumental value in
but not maintained under hot-house conditions. In sustaining expansion elsewhere in the nona-
many countries, the most promising field for rapid gricultural sectors, but for its own absorption
economic development lies in agriculture, and the of labor and its own increase of real income
measures needed are primarily such as will pro- among the rural poverty target groups of the
mote health, general education, technical training, small farmers and the landless laborers. The
better transportation facilities, and cheap rural widely expressed view now is that the root of
credit for productive use. There are no inherent
the employment problems lies in the fact that
advantages of manufacturing over agriculture, or,
for that matter, of agriculture over manufacturing. modern economic activity is not being dif-
It is only arbitrarily in fact that the line separating fused to the countryside. An agricultural
the two can be drawn. The choice between expan- strategy that would improve the rural-urban
sion of agriculture and expansion of manufactures balance now requires the extension of plan-
can for the most part best be left to the free deci- ning, infrastructure, appropriate technology,
sions of capitalists, entrepreneurs, and workers. To and complementary resources to the rural
the extent that there is need for government deci- sector.10 If in earlier decades of development,
sion, itshould be made on rational grounds, in the agricultural development had instrumental
light of considerations of costs and of comparative
value, in future decades it must have an in-
returns from alternative allocations of scarce na- trinsic value of its own.
tional resources, human and material. If direction
is accepted from maxims and arbitrary dogmas
and prejudices, from unsubstantiated and incred- 'Jacob Viner, International Trade and Economic
ible natural laws of the inherent inferiority of one Development, London, 1953, pp. 51-3.
type of industry over another, then it is highly ,0See, for instance, Edgar O. Edwards (ed.), Em-
probable that the result will be the squandering of ployment inDeveloping Nations, New York, 1974,
resources so scanty in supply that they need to be
pp. 30-31.

VII.A.2. Strategies for

Transferring Agricultural

Surplus*

The agricultural sector is of particular inter- bulk of national income, labor and capital re-
est and concern with respect to capital accu- sources. In the long run, the nonagricultural
mulation inthe early stages of economic de- sectors must grow at a substantially more
velopment. The agricultural sector is initially rapid rate than the agricultural sector, grad-
dominant in the economy, containing the ually providing a transformation of the econ-

*From T. H. Lee, "Strategies for Transferring Ag-

ricultural Surplus under Different Agricultural Situ- University Press, Ithaca, 1971, pp. 8-12, 131-4, 139-
ations in Taiwan," Conference on Agriculture and 42. Reprinted from Teng-hui Lee: Intersectoral Cap-
Economic Development, Japan Economic Research ital Flows in the Economic Development of Taiwan,
Center, September 6-10, 1971, pp. 1-7, 12-15, 21- 1895-1960. Copyright © 1971 by Cornell University.
5; also in Lee, Intersectoral Capital Flows in the Eco- Used by permission of the publisher, Cornell Univer-
nomic Development of Taiwan, 1895-1960, Cornell sity Press.
 432 AGRICULTURAL STRATEGY

omy from one dominated by agriculture to tions and across national boundaries. There-
one dominated by other sectors. It is logical fore, provided that increased production is
to presume that this process of economic actually obtained by the presently developing
transformation will proceed more rapidly if a countries, a new set of agricultural problems
net transfer of income and savings can be can be anticipated, for which new solutions
made from the agricultural sector to other must be discovered. How to maintain the cur-
sectors of the economy. rent increase in agricultural production with
The recent literature on the role of agri- more inputs and investment in agriculture,
culture in economic development has been and how to effectively siphon off such gains
enhanced by a number of essays on intersec- in agricultural production for the nation's in-
toral capital flows in relation to the transfor- dustrial development, are most important
mation of traditional agriculture in less de- and urgent tasks in the Southeast Asian
veloped countries. Two different viewpoints countries.
can be categorized from these papers. One is The experience of Taiwan in economic de-
that agriculture does not require a large velopment isan example of a country with a
amount of capital for its transformation. Ag- traditional agricultural pattern successfully
riculture, therefore, is a great contributor of advancing and transforming its economy as a
capital to industrialization. The other opinion whole. . . .
is that the investment requirements for agri-
cultural transformation are so large that cap-
ital may have net inflow from nonagriculture FRAMEWORK FOR
to agriculture. These views are based primar- MEASUREMENT
ily on different emphases of how to modern-
ize the traditional agriculture. The interrelationships between agricul-
Obviously, there are many differences be- tural production and nonagricultural produc-
tween countries in social-institutional ar- tion are dealt with here by dividing the areas
rangements with which to mobilize the re- participating in the economic transactions of
sources from the agricultural sector to the the national economy of Taiwan into six sec-
nonagricultural sector. According to the ini- tors: agricultural production, agricultural
tial level of agricultural productivity and re- household, nonagricultural production, non-
source endowment, capital requirement for agricultural household, government, and for-
transforming traditional agriculture will be eign trade.
different between countries. In the agricultural production sector, ser-
Infusions of the new technology into tra- vices of primary production factors such as
ditional agriculture are gaining dramatic re- land and labor flow from the agricultural
sults in Southeast Asia. Several countries in
household sector (A,).1 The sector also pro-
this area have begun shifts in production as duces agricultural output (Ya). The agricul-
strategic moves toward agricultural develop- tural production sector consumes production
ment. There is a spreading optimistic belief goods such as chemical fertilizer, feed, and
that the transformation is already sufficient other material manufactured in the nonagri-
to lift away the spectre of famine and to post- cultural production sector (Ra„). Agricultural
pone, at least, the materialization of the Mal- products used in agricultural production are
thusian trap. These gains have been most no- provided from the gross agricultural output
ticeable inbasic food grains (rice, corn and within the sector. The net agricultural output
wheat) and have been realized mainly from is partially consumed by the agricultural
new high-yield varieties with adequate sup-
plies of fertilizer, pesticides, water and mod- 'The letters in brackets are the symbols used for the
ern implements. It is expected that the new given item in the algebraic expression of these vari-
technology will quickly spread within the na- ables and relationships presented in equations below.
 433 IMPORTANCE OF AGRICULTURE
= Da
= Dn
household sector {Caa). The remaining Inflows . + M,
Outflows
amount of net output is sold to the nonagri- + Ga = Ca + Cna + Rna + E a (1)
= cn„ + C„ + Cf + K + / + En
cultural production sector as raw materials (2)
Ran + M + Gn
Da + Rna
Dn
Ca + <?„ + Sa
(Ra), to the nonagricultural household sector
Ca
C| ++ 5,
C„ + s„ = Ga+ Gn (4)
(3)
for consumption (C£), and directly to foreign (5)
trade as exports (Ea). The total quantity of
agricultural products sold amounts to the
sum of Ra + Ca + Ea. In nonagricultural Adding the five equations and cancelling out
production, services of production factors similar terms on both sides of the resulting
flow from the nonagricultural household sec- equality, we have:
tor (Dn). The sector produces two products, sa + sn + sg
consumer goods and capital goods. Consumer
-/+(£.+ En) - (Mc + M() (6)
goods flow from the nonagricultural produc-
tion sector to the nonagricultural household
or
sector (C"), to the agricultural household sec-
/„ = (Sa - Ia) + Sn + Sg+ F (7)
tor (C"),(C£), t0 tne government sector (Cf),
and to exports (En). Capital goods are dis- where / = Ia + /„, and F = (Ea + En) -
tributed to the agricultural production sector (Mc + M{). The terms Sa, Sn, and Sg in the
as intermediate goods (Ran), as investment above equations denote the savings of agri-
goods (Ia), and for investment in its own sec- cultural household, nonagricultural house-
tor (/„). No capital goods export is assumed hold, and government. Equation (6) is the fi-
in this case. The government sector collects nancing equation indicating the relationship
taxes (Ga) from the agricultural household between savings and investment for the na-
sector and (Gn) from the nonagricultural tional economy as a whole. Equation (7) in-
household sector, and allocates the revenue dicates the sectoral interdependence. The in-
for consumption of industrial goods (Cf), and vestment in the nonagricultural sector
for government savings (Sg). In the foreign depends upon the amount of net capital flow
trade sector, the government exports agricul- from agriculture, size of savings r: in its own
tural products (fsfl), and industrial consumer and government sectors, and the export sur-
goods (En), in exchange for consumer goods plus. Adding equations (1) and (3) for the
(Mc), and capital goods (Af,-). The balance of agricultural sector, we have:
international trade represents an additional
variable (F). Sn — Ca + Ra + Ec- c°n
Income generation is represented by flows (8)
in the opposite direction from the commodity As government taxing on agriculture is not
flows between sectors. In addition to the com- greatly in the form of commodities, the term
modity transactions between sectors, income (Ga) in equation (8) may be better included
also flows from the agricultural household in the term (Sa) from equation (8) and the
sector to the nonagricultural sector in the
term (Sa - Ia) in equation (7); then we can
form of payment of land rent, wages, and in- draw the following cases, indicating the bal-
terest. The agricultural household sector also ance of commodity flows between agriculture
receives income from the nonagricultural and nonagriculture. R°n = h
household sector. a,
These commodity and income flows can be Cna + Rna + Ec
summarized in the following accounting < (9)
equations.2 or
2J. C. Fei and G. Ranis, Development of the Labor
Surplus Economy, Theory and Policy, Homewood, Cn + Ra + F'a ~ C°r Ia= B (90
1964, p. 57.
 434 AGRICULTURAL STRATEGY

The terms on the left side of equation (90 in- changes in price ratios or sectoral terms of
dicate the commodity transactions between trade on sectoral capital flows are not re-
two sectors, and the term B is the balance flected inequations (90 and (10). The term
showing the physical aspect of capital outflow B in the equations, therefore, should be ad-
from agriculture. The term B is also the bal- justed for changes in the price ratios. The
ance of capital accounting between two equation (90 in real terms thus can be ex-
sectors.
Generally speaking, it is more common pressed:
and more useful to set up both capital and (Cna + Rna + Ea) I Pa - (C°
current operating (income) accounts in order + Ran + Ia) I Pn = Bf (11)
to investigate the sectoral commodity and fi-
nancial transactions. Captial account shows where Pa and Pn are price indices for agri-
the changes in assets and liabilities. An in- cultural products and nonagricultural prod-
crease in assets or decrease in liabilities in- ucts bought by the agricultural sector. When
dicates the outflow of capital. A decrease in capital flows out from the agricultural sector,
assets or an increase in liabilities indicates
the term B' can be expressed:
the inflow of capital. The term B can, there-
fore, be expressed as follows: B' = B/Pa + (Ci + K
B = R + K (10) + h) I Pn (Pn/Pa 1) (12)
The term R on the right side of the above
The first term on the right side of the equa-
equation is the balance of current financial tion isthe financial amount of capital outflow
transactions between sectors, including the
from agriculture in real terms, and the sec-
net payments of land rent, wages, and inter- ond term is the amount of capital outflow
est, and government taxing and subsidies. caused by the change in the sectoral terms of
The term K is the balance of the capital ac- trade between agriculture and nonagricul-
count between sectors, including the net ture. We call the former the visible net real
changes in outstanding short-term and long- capital outflow and the latter the invisible net
term loans and investment.
real capital outflow. . . .
The above exposition of the accounting
system of sectoral interdependence between
agriculture and nonagriculture is based on
commodity and income flows and the sectoral STRATEGIC MEASURES FOR
capital accounting. The important fact is that AGRICULTURAL DEVELOPMENT
both of the above sectoral accounts of income
AND CAPITAL TRANSFER
and capital are related to the expense ac-
counts ofincome, consumption, and savings- With a systematic examination of Tai-
investment in the agricultural and the non- wan's experience for the period 1895-1960,
agricultural sector. This means that the we set out to provide a statistical framework
above sectoral accounts can be derived statis- for empirical analysis of intersectoral capital
tically from the social income accounts in- flow between the agricultural sector and the
cluding income, consumption, and savings-in- nonagricultural sector. A statistical scale for
vestment ina sector. When we construct the measuring the intersectoral capital flows was
social income account for the agricultural developed by the social accounting system
sector, the sectoral accounts can be system- based on the definition of capital, and an ef-
atically derived from it. fort was made to identify the determinants of
Equations (90 and (10) are generally val- the net capital outflow from the agricultural
ued at current prices of commodities and ser- sector. Conclusions may be summarized as
vices in the transactions. The effects of follows:
 435 IMPORTANCE OF AGRICULTURE

(a) The direction of the intersectoral net nonagricultural sector with it. This important
capital flow was identified as an outflow from aspect of agricultural development particu-
the agricultural sector in Taiwan throughout larly with reference to the role of government
the entire period. The amount of net capital and technological progress in agriculture is
outflow showed a slightly increasing trend in discussed below.
terms of real price up to 1940, but has tended (c) The empirical tests showed that Tai-
to decline since 1950. Invisible net real capi- wan's experience departed appreciably from
tal outflow caused by terms of trade against the conventional hypotheses regarding net
agriculture was less important in the prewar capital outflow from agriculture. (1) Taiwan
period but increased in relative importance to has maintained a continuous outflow of net
more than 50 percent of the total net real capital from the agricultural sector under the
capital outflow in the postwar period. Finan- high growth rate of agricultural population
cially, current transfers of rent payment and and labor force. This fact disproves the
taxes occupied the most important role in the broadly held viewpoint that decelerating the
financial accommodation of net agricultural rate of population growth is a necessary con-
surplus in the prewar period, and direct cap- dition for accelerating the growth of agricul-
ital transfer of farmers' savings became in- tural surplus. (2) The agricultural wage rate
creasingly important in the postwar period. or per capita consumption of farmers im-
(b) The size of the intersectoral capital proved through time, despite the increase of
flow is dependent in part on the changes in population in agriculture. However, the pro-
the terms of trade, but it is also significantly portion oftotal agricultural income going to
dependent on the physical and financial mea- agricultural labor has tended to decline rel-
sures bywhich development can be achieved. atively incontrast to the nonagricultural sec-
Certain measures and conditions signifi- tor where the proportion of total income to
cantly influenced the intersectoral capital labor increased. In the context of net capital
outflow in Taiwan: (1) Under the Japanese outflow from the agricultural sector, the rel-
administration a new system of government ative decline of the share of labor income in
taxes and levies was imposed while the inher- agriculture is thus a more important concept
ited system of agricultural squeeze was not than that of constant institutional wage rate
abolished. Since institution of land-reform in agriculture. (3) Heavy investment in irri-
program in the postwar period, taxation and gation isnecessary in order to transform tra-
levies by means of both direct and hidden ditional agriculture in the paddy farming
methods have been strenthened. (2) Despite areas. Intensive innovation in the use of cap-
the high-gross squeeze, agricultural produc- ital has been witnessed in the period of trans-
tivity of land and labor in the sector was not formation oftraditional agriculture. This de-
affected. After the shift from traditional ag- parts from the conventional viewpoint of a
riculture in the period 1926-1930, the in- complementary relation between capital and
crease in agricultural productivity was ac- labor in agricultural innovation. (4) As for
celerated. Neither the initial resource endow- the amount of net capital outflow, it is clearly
ment nor the level of agricultural productiv- shown by the statistical comparison in the
ity in Taiwan in 1895 was any more favora- text that the concept of "net agricultural sav-
ble than in countries presently developing. ings" is not appropriate. (5) Financial ad-
However, the successful transformation of justment ofthe net agricultural surplus is one
traditional agriculture in Taiwan could be of the important factors determining the
accomplished while maintaining a continuous magnitude of the net capital outflow from the
net outflow of capital from the agricultural agricultural sector. The problem of intersec-
sector. A heavy investment in irrigation was toral capital flow can be discussed from the
initiated in the transformation period, yet it viewpoint of financial adjustment and the
did not bring a net inflow of capital from the commodity transferring process as well as
 436 AGRICULTURAL STRATEGY

from the viewpoint of increase in agricultural wan's experience, since a big push in irriga-
productivity. tion and land improvement had not been un-
(d) Finally, agricultural development is dertaken inTaiwan before a surplus in the
primarily concerned with the feasiblity of in- government budget and technological prog-
creasing net agricultural surplus or net capi- ress were realized. Two important factors
tal outflow from the agricultural sector. In need to be noted in this context: (a) deter-
less-developed countries like Taiwan, mobi- mined government action, and (b) technolog-
lization of internal capital must depend on ical relation between the fixed capital input
agricultural development. The development and biological technology. The former is re-
of agriculture and the application of eco- lated to the basic problem of capital alloca-
nomic squeeze on agriculture are closely re- tion in the whole national economy. Since
lated to government strategies for agricul- agriculture is generally considered the main-
tural development. stay of the economy, better utilization of
In relation to the intersectoral capital out- slack in agriculture is preferred to additional
flow from agriculture, four important govern- input of scarce capital funds. The latter is
ment measures toward agricultural develop- concerned with the availability of new vari-
ment can be derived: (1) allocation of capital eties of seeds, with the farmers' skill in ap-
to agriculture, (b) technological progress, (c) plication ofchemical fertilizer, and with the
agricultural taxation, and (d) organizational method of cultivation in relation to the heavy
improvements. . . . irrigation investment. The requirement for
heavy irrigation investment seems to be large
in the period of transition from extensive to
IMPLICATIONS OF TAIWAN'S intensive farming in paddy farming areas.
EXPERIENCE
With the high pressure of population, there is
In considering the implications of the a general tendency for labor intensive culti-
above discussion, it is important to gereralize vation in agriculture. To absorb more labor
the relationship between determinants of the input in farming, expansion of productive ca-
intersectoral capital flow in order to provide pacity interms of land is naturally necessary.
a measurement of agricultural development. However, the intensity of farming is greatly
The resource endowment and the level of ag- dependent on the demand for crops and live-
ricultural productivity determine the size of stock as well as the quantitative and quali-
agricultural investment that can be under- tative relationships between inputs. Land-
taken to achieve a given rate of agricultural owners, as receivers of large shares of land
growth. Basically, land productivity and per rent from the additional increase of output,
capita land area or man-land ratio are the de- will play some role in encouraging such in-
terminants ofthe level of agricultural pro- tensive farming. In Taiwan, promotion of
ductivity interms of labor. Consequently, new varieties of seeds, chemical fertilizers,
given the increase in population, limited land and irrigation investment represented such
resource and heavy food requirements in a an effort on the part of landlords.
low income country, a big increase is re- Agricultural transformation which simul-
quired in irrigation and land improvements. taneously maintains a net capital outflow
For this reason, Ishikawa has concluded that calls for a variety of strategies. The more im-
the agricultural sector may require a net in- portant of these are: (a) the basic agricul-
flow of capital from the nonagricultural sec- tural investment should be accompanied by
tor for the transformation of agriculture in technological improvement; (b) an appropri-
Asia.3 This obviously does not apply to Tai- ate investment scheme with large labor input
and less input of capital goods should be se-
3Shigeru Ishikawa, Economic Development in lected, and (c) a capital transfer mechanism
Asian Perspective, Tokyo, 1967, pp. 346-347. should be established. According to the dif-
 437 IMPORTANCE OF AGRICULTURE

ferent conditions or stages of agricultural de- will change in relative importance, as the ex-
velopment, the above strategic components perience ofTaiwan has shown. . . .

VII.A.3. Urban Bias*

The most important class conflict in the poor early development. . . . This huge welfare gap
countries of the world today is not between is demonstrably inefficient, as well as ineq-
labour and capital. Nor is it between foreign uitable. ... It persists mainly because less
and national interests. It is between the rural than 20 percent of investment for develop-
classes and the urban classes. The rural sec- ment has gone to the agricultural sector . . .
tor contains most of the poverty, and most of although over 65 percent of the people of
the low-cost sources of potential advance; but less-developed countries (LDCs), and over 80
the urban sector contains most of the articu- percent of the really poor who live on $1 a
lateness, organisation and power. So the week each or less, depend for a living on ag-
urban classes have been able to "win" most riculture. The proportion of skilled people
of the rounds of the struggle with the coun- who support development — doctors, bankers,
tryside; but in so doing they have made the engineers — going to rural areas has been
development process needlessly slow and un- lower still; and the rural-urban imbalances
fair. Scarce land, which might grow millets have in general been even greater than those
and beansprouts for hungry villagers, instead between agriculture and industry. Moreover,
produces a trickle of costly calories from in most LDCs, governments have taken nu-
meat and milk, which few except the urban merous measures with the unhappy side-ef-
rich (who have ample protein anyway) can af- fect of accentuating rural-urban disparities:
ford. Scarce investment, instead of going into their own allocation of public expenditure
water-pumps to grow rice, is wasted on urban and taxation; measures raising the price of
motorways. Scarce human skills design and industrial production relative to farm pro-
administer, not clean village wells and agri- duction, thus encouraging private rural sav-
cultural extension services, but world boxing ing to flow into industrial investment because
championships in showpiece stadia. Resource the value of industrial output has been arti-
allocations, within the city and the village as ficially boosted; and educational facilities en-
well as between them, reflect urban priorities couraging bright villagers to train in cities for
rather than equity or efficiency. The damage urban jobs.
has been increased by misguided ideological Such processes have been extremely inef-
imports, liberal and Marxian, and by the ficient. For instance, the impact on output of
town's success in buying off part of the rural $1 of carefully selected investment is in most
elite, thus transferring most of the costs of countries two to three times as high in agri-
the process to the rural poor. culture as elsewhere yet public policy and
The disparity between urban and rural private market power have combined to push
welfare is much greater in poor countries now domestic savings and foreign aid into nom
than it was in rich countries during their agricultural uses. The process has also been
inequitable. Agriculture starts with about
one-third the income per head of the rest of
*From Michael Lipton, Why Poor People Stay
Poor, London, Temple Smith, 1977, pp. 13, 16-7, 19- the economy, so that the people who depend
21, 23-4. Reprinted by permission. on it should in equity receive special attention
 438 AGRICULTURAL STRATEGY

not special mulcting. Finally, the misalloca- come, therefore, a policy that concentrates on
tion between sectors has created a needless raising income in the urban sector will
and acute conflict between efficiency and eq- worsen inequalities in two ways: by transfer-
uity. In agriculture the poor farmer with lit- ring not only from poor to rich, but also from
tle land is usually efficient in his use of both more equal to less equal. Concentration on
land and capital, whereas power, construc- urban enrichment is triply inequitable: be-
tion and industry often do best in big, capital- cause countryfolk start poorer; because such
intensive units; and rural income and power, concentration allots rural resources largely to
while far from equal, are less unequal then in the rural rich (who sell food to the cities);
the cities. So concentration on urban devel- and because the great inequality of power
opment and neglect of agriculture have within the towns renders urban resources es-
pushed resources away from activities where pecially likely to go to the resident elites.
they can help growth and benefit the poor, However, urban bias does not rest on a
and towards activities where they do either of conspiracy, but on convergent interests. In-
these, if at all, at the expense of the other. dustrialists, urban workers, even big farmers
Urban bias also increases inefficiency and all benefit if agriculture gets squeezed, pro-
inequity within the sectors. Poor farmers vided its few resources are steered, heavily
have little land and much underused family subsidised, to the big farmer, to produce
labour. Hence they tend to complement any cheap food and raw materials for the cities.
extra developmental resources received — Nobody conspires; all the powerful are satis-
pumpsets, fertilisers, virgin land — with much fied; the labour-intensive small farmer stays
more extra labour than do large farmers. efficient, poor and powerless, and had better
Poor farmers thus tend to get most output shut up. Meanwhile, the economist, often in
from such extra resources (as well as needing the blinkers of industrial determinism, con-
the extra income most). But rich farmers gratulates all concerned on resolutely ex-
(because they sell their extra output to the tracting an agricultural surplus to finance
cities instead of eating it themselves, and be- industrialisation. Conspiracy? Who needs
cause they are likely to use much of their conspiracy?
extra income to support urban investment) Thirdly, how far does the urban bias thesis
are naturally favoured by urban-biased poli- go towards an agricultural or rural empha-
cies; itis they, not the efficient small farmers, sis? It was noted that there is a rather low
who get the cheap loans and the fertiliser limit to the shifts than can swiftly be made in
subsidies. The patterns of allocation and dis- allocations of key resources like doctors or
tribution within the cities are damaged too. savings between huge, structured areas of
Farm inputs are produced inefficiently, in- economic life like agriculture and industry.
stead of imported, and the farmer has to pay, In the longer run, if the arguments of this
even if the price is nominally "subsi- book are right, how high do they push the al-
dised". .. . The processing of farm outputs, locations that should go to agriculture in poor
notably grain milling, is shifted into big countries: from the typical 20 percent of var-
urban units and the profits are no longer rein- ious sorts of scarce resource (for the poorest
vested in agriculture. And equalisation be- two-thirds of the people, who are also those
tween classes inside the cities becomes more
normally using scarce resources more effi-
risky, because the investment-starved farm ciently, as will be shown) up to 50 percent, or
sector might prove unable to deliver the food 70 percent, or (absurdly) 100 percent?
that a better-off urban mass would seek to Clearly the answer will differ according to
buy. the resource being reallocated, the length of
Moreover, income in poor countries is usu- time for the reallocation, and the national sit-
ally more equally distributed within the rural uation under review. The optimal extra pro-
sector than within the urban sector. Since in- portion ofdoctors for rural India, of invest-
come creates the power to distribute extra in- ment for rural Peru, and of increase in farm
 439 IMPORTANCE OF AGRICULTURE

prices for rural Nigeria will naturally differ. The second stage in policy for rural devel-
However, it remains true that pressures exist opment usually arises out of the failures of
to set all these levels far below their optima. Stage I. In Stage II, policy-makers argue
To acquire the right to advise against letting that agriculture cannot be safely neglected if
children go naked in winter, do I need to pre- it is adequately to provide workers, materials,
scribe the ideal designs of babies' bonnets? markets and saving to industry. Hence a lot
Linked to the question "Is there a limit to of resources need to be put into those parts of
the share of resources agriculture ought to agriculture (mainly big farms, though this is
get?" is a more fundamental question. Does seldom stated openly) that supply industry
the need for a high share of rural resources with raw materials, and industrial workers
last for ever? Does not development imply a with food. That is the stage that many poor
move out of agriculture and away from vil- countries have reached in their official pro-
lages? Since all developed countries have a nouncements, and some in their actual deci-
very high proportion of resources outside ag- sions. Stage II is still permeated by urban
riculture, can it make sense for underdevel- bias, because the farm sector is allocated re-
oped countries to push more resources into sources not mainly to raise economic welfare,
agriculture? And — a related question — as a but because, and insofar as, it uses the re-
poor country develops, does it not approach sources tofeed urban-industrial growth. De-
the British or U.S. style of farming, where it velopment ofthe rural sector is advocated,
is workers rather than machines or land that but not for the people who live and work
are scarce, so that the concentration of farm there.
resources upon big labour-saving farms be- In Stage III, the argument shifts. It is re-
gins to make more sense? alised that, so long as resources are concen-
The best way to look at this question is to trated on big farmers to provide urban in-
posit four stages in the analysis of policy in a puts, those resources will neither relieve need
developing country towards agriculture. nor — because big farmers use little labour
Stage I is to advocate leaving farming alone, per acre — be used very productively. So the
allowing it few resources, taxing it heavily if sequence is taken one step further back. It is
possible, and getting its outputs cheaply to fi- recognised, not only (as in Stage II) that ef-
nance industrial development, which has top ficient industrialisation is unlikely without
priority. This belief often rests on such com- major growth in rural inputs, but also (and
fortable assumptions as that agricultural this is the distinctive contribution of Stage
growth is ensured by rapid technical change; III) that such growth cannot be achieved ef-
does not require or cannot absorb investment; ficiently or equitably — or maybe at all — on
and can be directed to the poor while the rich the basis of immediately "extracting sur-
farmers alone are squeezed to provide the plus." Stage III therefore involves accepting
surpluses. Such a squeeze on agriculture was the need for a transformation of the mass
overtly Stalin's policy, and in effect (though rural sector, through major resource inputs,
much more humanely) the policy of the Sec- prior to substantial industrialisation, except
ond Indian Plan (1956-61) as articulated by insofar as such industrialisation is a more ef-
Mahalanobis, its chief architect. The bridge ficient way than (say) imports of providing
between the two was the economic analysis of the mass rural sector with farm requirements
Preobrazhensky and Feldman. The underly- or processing facilities. For development to
ing argument, that it is better to make ma- "march on two legs," the best foot must be
chines than to make consumer goods, espe- put forward first.
cially if one can make machines to make It is at Stage III that I stop. I do not be-
machines, ignores both the possible case for lieve that poor countries should "stay agri-
international specialisation, and the decided cultural" inorder to develop, let alone in-
inefficiency of using scarce resources to do stead of developing. The argument that
the right thing at the wrong time. neither the carrying capacity of the land, nor
 440 AGRICULTURAL STRATEGY

the market for farm products, is such as to oped agriculture, to conclude that, in order to
permit the masses in poor countries to reach begin the process of learning, a general at-
high levels of living without a major shift to tack on numerous branches of industrial ac-
non-farm activities seems conclusive. The ex- tivity should be initiated. A far better strat-
istence of a Stage IV must be recognised, egy is to concentrate first upon high-yielding
however. Stage IV is the belief that industri- mass rural development, supported (partly
alism degrades; that one should keep rural for learning's sake) by such selective ancil-
for ever. This is attractive to some people in lary industry as rural development makes vi-
poor countries because it marks a total rejec- able. Rapid industrialisation on a broad
tion of imitativeness. Neither Western nor front, doomed to self-strangulation for want
Soviet industrialism, but a "national path," is of the wage goods and savings capacity that
advocated. Other people, notably in rich only a developed agricultural sector can pro-
countries, argue that environmental factors vide, is likely to discredit industrialisation
preclude an industrialised world where all itself.
consume at U.S. levels; that there would be The arguments for rapid general industri-
too little of one or more key minerals, or that alisation, prior to or alongside agricultural
the use of so much energy would disastrously development, assume against most of the ev-
damage the world's air, water, climate or idence that such a sequence is likely to suc-
other aspects of the ecosystem. . . . ceed. But no national self-esteem, no learn-
The learning process, needed for modern ing-by-doing, nojam tomorrow, can come
industrialisation, is sometimes long; but it is from a mass of false starts. If you wish
fallacious for a nation, comprising above all for industrialisation, prepare to develop
a promising but overwhelmingly underdevel- agriculture.

EXHIBIT VII. 1. Growth Rates of Agricultural and Food Output by Major World Regions (excluding
China), 1960-1980

Agricultural Output Food Output -0.3

-1.1
Total Per capita Total Per capita
Region and Country
Group
1960-70 1970-80 1960-70 1970-80 1960-70 1970-80 1960-70 1970-80
-1.4 2.8
Developing countries 2.8 2.7 0.3 0.3 2.9 0.4
2.5 2.6
Low-income 2.1 0.2 2.2 0.4
0.4 0.2
Middle-income 2.9 3.1 3.3 0.9
0.4 2.6
3.2 0.7
Africa 2.7 1.3 0.7 1.6
0.2 0.1
Middle East 2.5 2.7 0.0 0.0 2.6 2.9 0.1 0.2
Latin America 2.9 3.0 0.1 0.6 3.6 3.3 0.1 0.6
Southeast Asia 2.9 3.8 0.3 2.8 0.3
2.5 1.4 3.8 1.4
South Asia 2.2 0.1 0.0 2.6 0.0
3.1 3.5 1.8 1.9 2.2
3.5 0.1
1.8 1.9
Southern Europe
Industrial market 3.2
1.1 2.0 1.3
economies 2.1 2.0 1.2 2.3 1.1
Nonmarket industrial
economies 3.2 1.7 2.2 0.9 3.2 2.2 0.9
Total world 2.6 2.2 0.4 1.7
2.3 0.8 0.5
0.7 2.7

Note: Production data are weighted by world export unit prices. Decade growth rates are based on midpoints of five-year
averages, except that 1970 is the average for 1969-71.
Source: World Development Report, 1982.
 VII.B. FOODGRAIN
REVOLUTION
VII.B. 1 . The Green Revolution:
Seven Generalizations*

By the mid-1970s, a substantial body of em- The rate of adoption of the new wheat and
pirical evidence had emerged that permits rice varieties has declined since the early
some clarification on the initial impact of the 1970s. In the case of the new wheat varieties
adoption of the new varieties on production the largest yield increments have been
and on the functional and personal distribu- achieved in relatively arid areas where farm-
tion of income. The conclusions that emerge ers have had access to effective tubewell or
from these studies can be summarized in a gravity irrigation systems. In the case of rice,
series of seven generalizations. Broad gener- the largest yield increments have been
alizations ofthe type presented below are achieved on irrigated rice during the dry sea-
never able to capture the rich detail of the son in areas such as Central Luzon (Philip-
particular location specific investigations on pines) or Western Uttar Pradesh (India).
which they are based. The net effect of this The agro-climatic regions where the wheat
review of the literature does, however, add up varieties developed at CYMMIT and the rice
to a quite different perspective on the impact varieties developed at IRRI were best
of the green revolution than the views that adapted have achieved relatively rapid and
dominate much of the earlier social sciences high level of adoption. Where diffusion of the
literature. CYMMIT and IRRI based varieties has de-
The new wheat and rice varieties were pended on the adaptation or development of
adopted at exceptionally rapid rates in those varieties suited to other environmental con-
areas where they were technically and eco- ditions, or on the modifications of environ-
nomically superior to local varieties. In the mental conditions, the rate of diffusion has
Indian Punjab, the proportion of total wheat been slower and the yield impact has been
area planted to the new high yielding vari- lower. Diffusion to other areas will depend, to
eties of wheat rose from 3.6 percent in the a very substantial degree, on the development
1966-67 the year of initial introduction, to of varieties of wheat and rice suited to other
65.6 percent in 1969-70. In three important ecological niches and on investment in irri-
wheat producing districts in the Pakistan gation and drainage in those areas where well
Punjab, 73 percent of wheat acreage was adapted varieties are available. It will also
sown with Mexican wheats during the 1969— depend on the successful development of high
70 rabi (winter) season. yielding varieties of other food grains, coarse
In the Philippines, 95 percent of the farm- grains and grain legumes.
ers in the barrios and almost 60 percent of Neither farm size nor tenure has been a se-
the farmers in the entire municipality where rious constraint to the adoption of new high-
the new rice varieties were initially intro- yielding grain varieties. This is not to say that
duced had adopted the new varieties in 1969, differential rates of adoption by farm size
four years after initial introduction. These and tenure have not been observed. What the
rates compare favorably to the diffusion rates available data do seem to imply is that within
of new crop varieties in developed countries. a relatively few years after introduction, lags
in adoption rates due to size or tenure have
*From Vernon W. Ruttan, "The Green Revolution:
typically disappeared.
Seven Generalizations," International Development Neither farm size nor tenure has been an
Review, December 1977, pp. 16-22. Reprinted by
permission. important source of differential growth in

441
 442 AGRICULTURAL STRATEGY

productivity. Sidhu's evidence from the In- input per hectare might be expected to de-
dian Punjab indicates that the new wheat cline substantially under fully mechanized
technology has been approximately neutral techniques combined with adoption of the
with respect to scale — it has not been high yielding variety technology.
strongly biased in either a labor-saving or a At this stage it seems more accurate to
capital-saving direction, and small and large view the growth of tractor mechanization, in
farmers have achieved approximately equal areas such as the Punjab, as an economic re-
gains in efficiency. sponse to the rising demand for labor associ-
Azam interprets the data that are availa- ated with the intensification of crop produc-
ble from a number of studies in the Pakistan tion rather than as an autonomous source of
Punjab to indicate "that while the smaller technological change leading to labor dis-
farmers do face relatively more severe con- placement. This process was underway prior
straints ofirrigation water and credit, the dif- to the introduction of the new wheat varieties
ferences inthe severity of these constraints is and has been reinforced by the more rapid
not serious enough to have caused any signif- growth in demand for labor since their
icant differences in the yields obtained by the introduction.
small farmers as compared with the large Landowners have gained relative to tenants
farmers." Similar results have been reported and laborers from the adoption of the higher
from the Philippines by Mangahas and from yielding grain varieties. Data assembled by
Indonesia by Soejono. Mellor indicate that although the percentage
The introduction of the new high yielding increase in labor earnings from increased em-
wheat and rice technology has resulted in an ployment and wages is often fairly large, the
increase in the demand for labor. Sidhu's re- percentage of the increased output allocated
sults indicated a very substantial shift to the to labor is relatively small. . . .
right in the labor demand function on wheat Much of the discussion about the income
producing farms, as a result of the introduc- distribution effects of the green revolution
tion of the new wheat varieties in the Indian has failed to distinguish between its absolute
Punjab. . . . The net effect of the increase in and relative effects on income distribution.
demand for labor has been a significant rise Many authors who refer to the worsening po-
in real wages in the Punjab at a time when sition of the smaller owners and tenants or
real wages were constant or declining in most landless laborers are apparently referring to
states in India. the widening absolute gap in the income dis-
tribution rather than to an actual decline in
the income of those who occupy the lower
end of the income distribution in rural
HIGH-YIELDING VARIETY communities. . . .
TECHNOLOGY
The effect of introduction of the new high
An extensive review of the literature by yielding varieties has been to contribute to a
Bartsch indicates that the introduction of widening of wage and income differentials
high yielding varieties into traditional wheat among regions. As mentioned in the first gen-
and rice production systems has typically re- eralization, the varieties have been developed
sulted insubstantial increases in annual labor to respond most favorably to those elements
use per hectare. The increase in labor use has in the environment which are subject to
been due to greater labor utilization per unit man's control. They are more responsive than
of cropped area, and in some cases, to higher the varieties they replaced to higher levels of
cropping intensity. fertilization, to more effective irrigation and
Even mechanized farms typically were uti- drainage; and more effective control of path-
lizing increased labor inputs per hectare al- ogens, insects, weeds. Reductions in sensitiv-
though simulation results conducted by ity to certain natural variations such as day
Bartsch and by Singh indicate that labor length and temperature make them more
 443 FOODGRAIN REVOLUTION

adaptable to intensive systems of crop come strata, relative to the higher income
production. strata, the more favorable will be the distri-
The contribution of the new varieties to butional benefits.
productivity growth has, therefore, been This is illustrated quite dramatically by
greatest in those regions where there has the impact of the new rice technology on con-
been substantial investment in physical and sumer welfare in Colombia since the mid-
institutional infrastructure development. 1960s. Between 1966 and 1974 the percent-
This pattern is reinforced by the location spe- age of the area planted to modern varieties
cific character of agricultural technology. . . . rose from 10 percent to 99 percent. Yields on
The contribution of the new seed-fertilizer irrigated land rose from 3.1 to 5.4 metric tons
technology to the widening of regional in- per hectare, and total rice production in-
come disparities has apparently been greater creased from 600 thousand to 1,570 thousand
than its impact on disparities in income metric tons. Most of the increased production
within communities and regions. The associ- was absorbed in the local market. The bene-
ated changes in the regional distribution of fits were transmitted to consumers through
political resources can be expected to become both lower prices and increased per capita
a more important source of institutional consumption.
stress and institutional change than the stress The benefits were strongly biased in favor
at the community level which has received so of low income consumers. The lowest income
much attention in recent literature. quartile of Colombian households, which re-
The effect of the introduction of the new ceived only 4 percent of household income,
high yielding varieties has been to dampen captured 28 percent of the consumer benefits
the rate of increase in food grain prices at the resulting from the shift to the right in the
consumer level. During the 1974-75 crop supply curve for rice.
year the new higher yielding or modern va- In most countries, increases in food grain
rieties ofwheat were planted on over 20 mil- production have generally not been adequate
lion hectares and the new high yielding vari- to prevent substantial increases in food grain
eties of rice were planted on close to 25 prices, when measured in current dollars, in
million hectares in Asia and the Near East. the face of the general inflationary pressures
In Asia over 60 percent of the wheat area that have dominated world commodity mar-
and over one quarter of the rice area are kets between the late 1960s and the early
planted to the modern varieties developed 1970s. It is clear, however, that in the ab-
since the mid-1960s. Evenson has estimated sence of the contribution of the new high
that in crop year 1974-75, the supply of rice yielding varieties, food grain prices would be
in all developing countries was approximately even higher in many countries of Asia, Africa
12 percent higher than it would have been if and Latin America. Part of the new income
the same total resources had been devoted to streams generated by the new varieties have
production of rice using only the traditional been transferred from producers to con-
rice varieties available prior to the mid- sumers either through the market or through
1960s. . . . administered distribution schemes.
The impact of a shift to the right in the Thus while there may be some ambiguity
supply of food grain is particularly significant regarding the distribution of the gains by size
for both the urban and rural poor. The dis- of farm or by economic or social class within
tribution ofgrains among consumers depends the agricultural sector, there can be little
primarily on the relative amount of a partic- question that the distributional effect on the
ular commodity consumed by each income consumption side has been positive. And
stratum and on the price elasticity of demand among those who have gained on the con-
in each stratum. The larger the quantity con- sumption side have been the landless and
sumed and the higher the absolute value of near landless workers in rural areas.
the price elasticity of demand in the lower in- Issues of the distribution of gains between
 444 AGRICULTURAL STRATEGY

producers and consumers in the developing weak and the pattern of inequity has been
countries which have benefited from the new reinforced. The differential impact of the
high yielding varieties have been given less technology on income growth has apparently
attention than in developed countries. This been greater among regions than among eco-
may be in part because the rapid growth of nomic factors and social classes within
demand stemming primarily from population regions.
growth has tended to equal or exceed the rate It is still premature to attempt a definitive
of growth in supply even when the growth in evaluation of the impact of the green revolu-
supply has been augmented by rapid techni- tion technology on institutional change. Few
cal change. attempts have been made to separate the ef-
Perhaps the more important factor is that fects of technical change from the other dy-
with relatively few exceptions the peasant namic changes that have also impinged on
producers of food crops in Asia are not effec- the rural areas which are experiencing rapid
tively organized to reflect their economic in- productivity gains.
terests at the policy level. Resistance to low It does seem clear, however, that the con-
price levels tends to take the form of attempts tribution ofthe new seed-fertilizer technol-
to adjust the crop mix to relative price shifts ogy to food grain production has weakened
rather than to directly influence price policy. the potential for revolutionary change in po-
litical and economic institutions in rural
CONCLUSION areas in many countries in Asia and in other
parts of the developing world. The green rev-
The picture that emerges from this review olution has not turned red. In spite of wid-
of the evidence on the initial impact of the ening income differentials, the gains from
green revolution can be summarized as fol- productivity growth, in those areas where the
lows— a technology that is essentially neutral new seed-fertilizer technology has been effec-
with respect to scale has been introduced into tive, have been sufficiently diffused to rein-
environments in which the economic, social force interests of most classes in an evolution-
and political institutions have varied widely ary, rather than a revolutionary, pattern of
with respect to their neutrality. This view rural development.
has been eloquently expressed by Wolf The most serious criticism that can be lev-
Ladejinsky: elled at the green revolution is that it has not
When all is said and done, it is not the fault of the yet become sufficiently pervasive. The most
disturbing evidence of failure in agricultural
new technology that the credit service doesn't development in Asia over the last decade is
serve those for whom it was originally intended;
that the extension service is not living up to expec- the evidence of declining real wages in many
tations; that the panchayats are essentially politi- areas where the impact of the new seed-fer-
cal rather than development bodies; that security tilizer technology on productivity growth has
of tenure is a luxury of the few; that rentals are been marginal.
exorbitant; that ceilings on land are merely no- Furthermore there is evidence that even in
tional; that for the greater part tenurial legislation the more favored areas the productivity gains
is deliberately miscarried, or wage scales are
hardly sufficient to keep soul and body together.
are coming more slowly and at greater cost
than during the last decade. A combination
Where the technology has been introduced of continued rapid growth in the rural labor
in areas characterized by a reasonable degree force pressing against inadequate rate of
of equity in the distribution of resources, the growth in productivity is resulting in increas-
effect has been favorable both in terms of ing immiserization of the landless and near
productivity and equity. When the technol- landless in many areas in Asia. It is possible
ogy has been introduced in areas character- that this increasing immiserization may, in
ized by great inequality in the distribution of the next decade, induce the revolutionary
resources, the productivity impact has been changes in rural institutions which were an-
 445 FOODGRAIN REVOLUTION

ticipated in the 1970s by the radical critics of Kosi Area: A Field Trip," Economic and
the green revolution. Political Weekly, Vol. 4, September 27,
1959, pp. 1-14.
References Mangahas, Mahar. "Economic Aspects of
Agrarian Reform under New Society,"
Azam, K. M. "The Future of the Green Rev- Philippine Review of Business and Eco-
olution in West Pakistan: A Choice of nomics, Vol. 11, December 1974, pp. 175-
87.
Strategy," International Journal of Agrar-
ian Affairs, Vol. 5, March 1973. Mellor, John W. The New Economics of
Bartsch, William G. Employment Effects of Growth: A Strategy for India, Ithaca,
Alternative Technologies and Techniques 1976.
in Asian Crop Production: A Survey of Ev- Sidhu, Surjit S. "Economics of Technical
idence. Geneva, International Labor Of- Change in Wheat Production in the Indian
fice, Provisional Draft, 1973.
Punjab," American Journal of Agricul-
Evenson, Robert E. "Comparative Evidence tural Economics, Vol. 56, May 1974, p.
on Returns to Investment in National and 221.
International Research Institutions," in . "Relative Efficiency in Wheat Pro-
Thomas M. Arndt, Dana G. Dalrymple duction inthe Indian Punjab," The Amer-
and Vernon W. Ruttan (eds.), Resource ican Economic Review, Vol. 64, September
Allocation and Productivity in National
and International Agriculture Research, 1974, pp. 742-51.
Singh, Srinath. Modernization of Agricul-
Minneapolis, 1977, pp. 237-65. ture, New Delhi, 1976.
Ladejinsky, Wolf. "The Green Revolution in Soejono, Irian. "Growth and Distributional
the Punjab: A Field Trip," Economic and Changes in Paddy Farm Income in Central
Political Weekly, Vol. 4, June 28, 1969, Jave," Prisma — Indonesian Journal of So-
pp. A73-A83. cial and Economic Affairs, May 1976, pp.
. "Green Revolution in Bihar, The 26-32.

Comment
Studies of the green revolution include: T. T. Poleman and D. K. Freebairn (eds.), Food,
Population, and Employment: The Impact of the Green Revolution (1973); Clive Bell, "The
Acquisition of Agricultural Technology," Journal of Development Studies (October 1972);
B. F. Johnston and J. Cownie, "The Seed-Fertilizer Revolution and Labor Force Absorption,"
(September 1969); B. F. Johnston and Peter Kilby, Agriculture and Structural Transforma-
tion (1975).
J. W. Mellor, The New Economics of Growth (1976); C. Wharton, "The Green Revolution:
Cornucopia or Pandora's Box?" Foreign Affairs (April 1969); W. Ladejinsky, Agrarian Re-
form as Unfinished Business (1978); Michael Lipton, "The Technology, the System, and the
Poor: The Case of the New Cereal Varieties," in K. Post (ed.), Developing Societies: The
Next 25 Years (1979).
Radical political economists have argued that the green revolution's technology tends to be
monopolized by large commercial farmers who have better access to new information and
better financial capacity. A large profit resulting from the exclusive adoption of modern va-
rieties oftechnology by large farmers stimulates them to enlarge their operational holdings
by consolidating the farms of small nonadopters through purchase or tenant eviction. As a
result, polarization of rural communities into large commercial farmers and landless prole-
tariat ispromoted. See Harry M. Cleaver, "The Contradictions of the Green Revolution,"
American Economic Review (May 1972); AH M. S. Fatami, "The Green Revolution: An Ap-
praisal," Monthly Review (June 1972); Keith Griffin, The Political Economy of Agrarian
Change (1974); Richard Grabowski, "The Implications of an Induced Innovation Model,"
 446 AGRICULTURAL STRATEGY

Economic Development and Cultural Change (July 1979); Richard Grabowski, "Reply", Eco-
nomic Development and Cultural Change (October 1981).

Comment
The green revolution is often compared to the "Japanese model" of increases in agricultural
productivity associated with the use of improved varieties, fertilizers, implements, and other
complementary inputs within the framework of Japan's small-scale farming system. For a
comparative study of Japan's experience and what has been brought about by the green rev-
olution, see Kazushi Ohkawa, Differential Structure and Agriculture — Essays on Dualistic
Growth (1972).
Some other important studies of the Japanese case are: B. F. Johnston, "Agriculture and
Economic Development: The Relevance of the Japanese Experience," Food Research Insti-
tute Studies, Vol. 5 (1966); W. W. Lockwood (ed.), The State and Economic Enterprise in
Japan (1965), Chapters 2 and 6; Saburo Yamada, "Changes in Output and in Conventional
and Nonconventional Inputs in Japanese Agriculture Since 1880," Food Research Institute
Studies, Vol. 7 (1967); Lawrence Klein and Kazushi Ohkawa (eds.), Economic Growth: The
Japanese Experience Since the Meiji Era (1968), part I; K. Ohkawa and B. F. Johnston, "The
Transferability of the Japanese Pattern of Modernizing Traditional Agriculture," in E. Thor-
becke (ed.), The Role of Agriculture in Economic Development (1969); M. Akino and Y.
Hayami, "Agricultural Growth in Japan, 1880-1965," Quarterly Journal of Economics
(August 1974).

VII. B. 2. Generations of
Problems*
In the regions of Asia where the production system. Large, uncovered piles of rice accu-
revolution has occurred, the impact on mar- mulated at railheads, and prices to farmers
keted surplus has been nothing short of phe- fell substantially. Millers were working
nomenal. Even with a moderately high on- equipment at capacity and were running into
farm demand from increased output, market- severe inventory and working capital con-
ings have risen much more than proportion- straints. (As usual, they were blamed for the
ately to production. While the response of decline in price.) It nearly required a French-
public and private sectors in a few regions style, pitchfork rebellion to obtain more rail
has been good, the pace of change, the preoc- cars, to change government policy to permit
cupation with production, and the ability of trucks to deliver rice to the port, etc. In the
policy makers to handle only a few issues si- meantime, however,
multaneously have meant that few policy ac- least relative to whatfarmers
would were
have "hurt," at
been the
tions were taken before crises erupted. Trans- case with a better transport system and
portation bottlenecks often have been a faster-moving government policy machinery.
problem, as an example from West Pakistan Similar stories on milling, grading, stor-
will illustrate. In Sind (the lower half of the age, and transport can be told for other coun-
hour-glass-shaped Indus Basin), rail market- tries as well. The problem of limited, old-
ings of rice in 1 969 completely swamped the style mills, unable to handle increased sup-
*From Walter P. Falcon, "The Green Revolution: plies or to produce "export-quality" rice, is
well documented in reports and government
Second-Generation Problems," American Journal of documents that I have seen for at least five
Agricultural Economics, December 1970, pp. 701-9.
Reprinted by permission. Asian countries. These physical problems of
 447 FOODGRAIN REVOLUTION

marketing have been exacerbated by social enough to emerge as surplus areas the prob-
factors in several countries of Southeast lems of breaking into international grain
Asia, where specific ethnic or racial groups markets have rarely appeared so difficult.
have traditionally controlled most of the The International Wheat Agreement appears
commerce. Regardless of the efficiency of the to be seriously undermined, and there has
marketing system, rural problems have been a considerable softening in rice prices,
tended to be blamed on these groups. Justly particularly in the lower-quality grades.
or unjustly, middlemen are an important fac- Several elements of the international di-
tor in social and political unrest in these mension deserve mention. What happens to
countries. This unrest, in turn, has posed the "world prices" for wheat and rice is obviously
problem of either taking over milling in the dependent on what happens to the green rev-
public sector or developing a set of incentives olution inthe developing countries as well as
and guidelines for the private trade that will to the agricultural policies of the developed
protect the public interest as seen by the pol- nations. As indicated previously, there are
icy makers. Efficiency and ideology are often reasons to believe that portions of Indonesia,
in conflict on this point, and the net result in India, and East Pakistan are likely to be net
many regions has been that the developments importers for some time. On the other hand,
in marketing skills have lagged. the quantity traded internationally is so small
There have also been varietal-quality ques- relative to production — less than four per-
tions that have posed difficulties both domes- cent in the case of rice — that increases in
tically and internationally. . . . However im- production in key countries such as India are
portant these difficulties may have been in likely to have important international price
the short run, they are clearly transitional in repercussions. Perhaps even more important
nature. New varieties are already being de- than what happens in the developing coun-
veloped and introduced that will overcome tries iswhat happens to agricultural policy in
many of the most severe quality problems. the advanced countries. Unable to adapt to
In addition to the readily identifiable mill- rapid technological advances and structural
ing, transport, and grading questions, there change themselves, these countries have in-
are also formidable second-generation prob- stituted support systems that use commodity
lems concerned with pricing and markets. exports to solve sectoral income distribution
There are economic and political dimensions problems. In short, less-developed countries
to these questions, and both aspects must be breaking into export markets will be faced
incorporated into meaningful answers. with three kinds of problems: (a) a tenacity
A number of the food-deficit countries among developed countries in fighting for
have historically had a structure of relative shares of the commerical market and a will-
prices that bore little relationship to world ingness to cut prices to retain them; (b) an
prices. Although in allocative efficiency increasing amount of foodgrains being sup-
terms such a structure has always had draw- plied by developed countries at concessional
backs, the problem takes on even more seri- terms to countries that might "normally" be
ous proportions when countries and regions trading partners of developing countries; and
close their food import gap and become po- (c) an inability, or at least difficulty, of the
tential exporters. Adjusting domestic support
less-developed
markets in terms worldof to compete
specific in "buyers'
grades, quality,"
prices, which at the official exchange rate are
often double or more the world price, is no deliverability, etc. This does not mean that
easier politically in these countries than in the developing countries cannot sell in inter-
the United States. national markets. What it does mean is that
In addition to internal pricing difficulties planners in these countries must be hard-
an even larger problem looms ahead on the headed about the quantities and especially
international side. For those regions "lucky" the prices at which wheat and rice can be ex-
 448 AGRICULTURAL STRATEGY

ported and about the concomitant internal and nonagricultural commodities. The real
price adjustment (or export subsidy) that will tragedy would be for those countries to retain
be required at these levels. outmoded pricing policies which lead to great
The foregoing marketing and demand inefficiencies in resource use, stock accumu-
problems, any one of which could be the sub- lations, and/or highly subsidized agricultural
ject of a major paper, suggest several exports — exports that were uneconomically
conclusions: grown in real terms in the first place. Unfor-
First, the production gains in certain re- tunately, experience in dealing with such
gions have shown how rapidly second-gener- problems in developed countries does not in-
ation marketing problems can arise. It is to spire confidence, nor do recent policies in a
be hoped that in the future policy makers will number of Asian countries.
heed earlier the warnings given by marketing Third, the advanced countries must con-
specialists and will react before crisis situa- sider more seriously the distorting effects of
tions develop. Unless these milling and trans- their dumping programs. The talk of a world
port problems are solved, farm prices will de- market price for wheat or for rice is largely a
cline steeply and the quality problems of fiction, and concessional pricing arrange-
exporting will be all the more difficult. What ments will be a sharp deterrent to the gener-
is particularly needed in several Asian coun- ation of third-country foodgrain exports.
tries is a marketing strategy that resolves Fourth, since there is little reason to have
the basic public/private/foreign investment confidence
question on marketing facilities. Also needed to deal with intheir
the sectoral
developedincome
countries' ability
distribution
is an explicit recognition of the interaction of problems without resort to concessional ef-
price support policies and techniques with the forts, the developing countries should look in-
behavior and efficiency of marketing firms. creasingly todomestic markets for absorbing
Second, planners must pay increasing at- additional supplies. On this point there is
tention to the adjustment and pricing prob- some room for optimism. What has been se-
lems attendant on the new varieties. The nar- riously underestimated, I believe, is the in-
row focus on foodgrains and relative neglect vestment and employment uses to which
of other crops must be reevaluated in a mul- wheat and rice, the wage goods, can be put.
ticrop setting. In particular, the cropping The basic elements in this argument can be
patterns of many of the irrigated areas of stated as follows: With significant increases
Asia which can best use the new varieties are in production, foodgrain prices in a closed
quite sensitive to profitability changes. What economy would fall. However, given the fact
constitutes an appropriate incentive price for that much of the increase came from cost-
foodgrains in these areas with the new vari- free technological change, prices could fall
eties has changed substantially; unfortu- somewhat and still provide adequate incen-
nately, the rhetoric of the later 1950's and tives to farmers. In addition, with substantial
early 1960's regarding the need for ever supplies of grain, the government can have a
higher agricultural prices has not changed. much more expansionary fiscal and monetary
Vested interests in agriculture are already a policy. (Indeed, in India and Indonesia the
fact of life in these countries, and economists lack of adequate food supplies and a fear of
concerned with agriculture must keep in rising prices have been constraints on the size
mind the overall needs of development, not of the development budget.) The more ex-
just the needs of the agricultural sector. pansionary monetary and fiscal policy — par-
Since most agricultural goods are tradable, ticularly ifit is directed toward labor-inten-
what is especially needed in the less-devel- sive public projects — can shift the demand
oped countries is an assessment of the do- curve for grains, helping to counteract some
mestic costs of earning or saving foreign ex- of the decline in prices. Given the fact that
change from producing various agricultural the price of the wage good is a major devel-
 449 FOODGRAIN REVOLUTION

opment constraint in much of Asia, especially techniques, which are often of a labor-dis-
as seen by finance ministers, the increases in placing nature. The resulting dilemma can be
production from the green revolution can baldly stated: The Asian countries need ag-
thus continue after initial import substitution ricultural growth if ever they are to break the
has been exhausted. These increases can be chains of poverty; but they need equity as
converted into investible resources through well, for obvious humanitarian reasons and in
fiscal and monetary expansion, and the coun- order not to find themselves in a continuous
try (perhaps even the agricultural sector, if cycle of violence and repression. The chal-
the investments are rural) could be much bet- lenge of these forces is far greater in magni-
ter off. This should be one element of devel- tude than the problems ever faced by the
opment strategy for countries moving into United States and most other presently de-
foodgrain surpluses; moreover, it seems es- veloped nations. Moreover, the latter are not
pecially important for countries who find in a position to help. Although they are per-
themselves with seriously distorted internal haps capable of exporting the growth tech-
prices. This approach should provide time nology, they have few institutional forms to
both to solve the institutional problems of en- export that can come to grips with the income
tering international trade and to make tran- distribution and employment questions that
sitional changes in relative and absolute price now plague Asia.
levels without having to rely on stock accu- India, Pakistan, and Indonesia — three
mulation or"excessive" subsidies to agricul- enormously large and regionally heteroge-
ture. Such a strategy also has much in com- neous countries — present stark examples of
mon with a sensible P.L. 480 policy which the problems outlined above.
can effect shifts in the demand curve through Perhaps even more important than the di-
investment policies, thereby helping to coun- rect effects, and often neglected in discus-
teract much of the decline in prices that sions on employment, are the side effects of
would have resulted from increased supplies. increased food supplies and lower food prices
The first-generation production problems on public and private savings and investment
and the second-generation marketing and de- generally. As noted earlier, the food-price
mand difficulties created by the green revo- constraint is an important one and has a per-
lution are a formidable list. Nevertheless, vasiveness that extends far beyond the agri-
they are largely short-run issues on which cultural sector. Here, too, the green revolu-
economists have worked for years. By con- tion helps, provided that its potential for
trast, the third-generation problems, having increasing savings is realized and is trans-
to do with equity, welfare, employment, and ferred into real investment.
social institutions generally are questions Far more disturbing, however, are two
that have received inadequate attention even other effects of the green revolution on em-
in the developed countries. . . . ployment, welfare, and stability. Both of
These third-generation factors arise from these derive basically from the unequal re-
four principal sources: (a) population growth gional growth that seems to be a concomitant
rates in excess of 2.5 percent annually in of the new technology. The process is as fol-
areas already extraordinarily densely popu- lows: The regions with irrigation, such as the
lated; (b) very low average income levels, Punjab, have the ability to respond rapidly to
coupled simultaneously with great regional the new technology. A combination of the re-
and personal disparities in income, wealth, sulting production plus an agricultural price
and political power; (c) limited opportunities policy that reflects concerns for nongrowing
for nonfarm employment, even if the manu- districts as well as vested agricultural inter-
facturing and service sectors grow very rap- ests will mean that incomes in the irrigated
idly; and (d) the possibility for technological regions will grow at phenomenal rates. That
leap-frogging with agricultural inputs and is all to the good; the difficulty is that welfare,
 450 AGRICULTURAL STRATEGY

between regions as among people, is more a years. This labor-displacement process was
relative concept than an absolute idea. In this not "easy" in the United States; in Asia the
interregional sense, therefore, the green rev- situation is distressing even to
olution ishardly a stabilizing influence. contemplate. . . .
Within a given region, the mechanism pro- Several recommendations and reconsider-
ducing greater income inequality is much the ations are suggested in the light of these
same, and the form is even more virulent. Al- third-generation questions. First, as long as
though intheory the new seeds and fertilizer the new varieties remain limited to a few re-
are neutral to scale, in practice they are not. gions and as long as farm incomes are pri-
Under rationed conditions, and unfortunately marily dependent on acreage rather than
these often prevail for inputs in Asia, it is the people, it is naive to believe that the new
larger farmers who obtain the fertilizer and technology for agriculture is likely to be a
receive the irrigation water. Moreover, with stabilizing influence. Growth generally is
the prices and technology now prevailing, ag- destabilizing, and this form of unequal agri-
ricultural incomes of large farmers have risen cultural development is particularly so. Even
dramatically. This too is not "bad," but the if the first borrowings of technology are neu-
side effects may be. Land prices are rising tral to scale (which in practice probably they
rapidly, as farmers seek to expand size and are not), subsequent borrowings are likely to
find new outlets for their increased incomes. be labor-displacing unless strong policy mea-
Even more important is the drive that these sures are introduced. The magnitude of this
windfall gains are providing for certain types phenomenon will vary by commodity and re-
of mechanization. Although this is a broad gion, but the direction seems fairly clear.
question, deserving also of a separate study, Second, some way must be found to close the
several points deserve mention. First, there gap between social and private benefits from
are powerful forces that are pressing for certain forms of agricultural technology. It is
mechanization of all kinds. Large farmers,
not sufficient to appeal to the "Japanese
foreign and domestic industrialists, politi- method" of cultivation, to urge labor-inten-
cians, and even aid agencies have vested sive techniques for agriculture and industry,
interests in promoting various implements, or to proclaim the virtues of small-scale in-
including tractors. Some forms of mechani- dustry. Such pronouncements must be trans-
zation may be labor-displacing, others not. formed into instruments of direction and con-
However, large farms in wheat areas are an trol: high taxes on tractors; a possible
example of where tractors and combines will lowering of wheat and rice prices as a stim-
be introduced, barring strong government ac- ulant to the rest of the economy; much higher
tion to the contrary. The net result will be to interest rates on capital and higher de facto
make tenants into laborers and to increase rates for foreign exchange; progressive land
the number of people displaced from agricul- taxes; and perhaps even ceilings on farm size
ture. Just as in the interregional illustration, so as to make uneconomical, from a private
the intraregional effects of the green revolu- point of view, certain forms of technology.
tion are likely to increase the inequality of in- And in any Asian country, no one should dis-
comes within agriculture. There will indeed count the size and power of the forces that
be agricultural growth in these areas, but are likely to be against most of these policies.
probably increasing tension among classes as Third, neither growth nor equity problems
well Perhaps the growth in service and sup- in Asia can be solved by the green revolution
ply industries in small towns can absorb this or even by the agricultural sector alone. The
additional displacement. But the adjustment
problems, with which the United States had employment problem in particular is total-
economy in character, whose solution re-
trouble in coping under much more favorable
quires increased savings, more foreign ex-
demographic circumstances and over a cen- change, higher investment rates, altered
tury, must be dealt with in Pakistan in 20 factor- and product-pricing structures — in
 451 FOODGRAIN REVOLUTION

short, economic development. While agricul- credit, marketing, and pump facilties at the
tural policies should not aggrevate the situa- village level. On the other hand, most of the
tion, meangingful answers to these issues cooperatives of South and Southeast Asia
must look to other sectors as well. have been run as heavy-handed government
Fourth, given the tearing effect that un- agencies with little local support except
equal regional growth has on the national among the rural elite who have benefited
fabric, there is need to stress again the im- from them. Similarly, loan programs espe-
portance of developing new technology for cially designed for small farmers have gen-
the monsoon/dryland areas. erally had little success because of prohibi-
Finally, while there is need to keep social tively high transaction costs for issuing and
and private benefits from diverging among monitoring small loans. Perhaps most prom-
the large farms, the oppostie side of the coin ising as an aid to the smaller operator is the
is to assist small farmers. . . . Given the re- provision of adequate supplies of irrigation
sources available and the political interests water. The employment effects from this type
that are involved, a broad-based welfare sys- of infrastructure are substantial, and reliable
tem does not seem to be the answer. Nor do water supplies may provide the flexibility for
special loan or credit arrangements to small diversifying and intensifying output. On the
farmers which are used for unproductive in- whole, however, the outlook is far from
vestments. There is reason to be even more bright for the smaller farmer.
skeptical, as amply demonstrated in the
United States, about price support or input CONCLUDING COMMENTS
subsidies as an instrument. It is the large
farmer who has the marketed surplus and The foregoing assessment of the green rev-
who uses most of the inputs. (Nearly one- olution ishardly one of wild enthusiasm. The
third of the farmers in Indonesia, Pakistan, purpose has not been to argue that it should
and India, for example, are net purchasers of not have happened or to deny its great pro-
grains.) It is spurious to argue for higher duction successes in certain regions. Rather,
the intent has been to indicate how limited a
farm prices or increased subsidies to "help
the small farmer," for it would be hard to de- solution the revolution is, given the broader
sign a more inefficient system for reaching development problems of South and South-
them. (Some rough calculations for India east Asia.
and Pakistan indicate that of $10 transferred Four central themes stand out in the anal-
via a price-support system, only about $1 ysis. First, impressive as the gains to date
goes to "small" farmers.) The small-farmer have been, the term "revolution" can be ap-
argument, which is always offered by the rep- plied correctly only to about 10 to 15 percent
resentatives oflarger farmers whenever pric- of Asia. One of the greatest second-genera-
ing is an issue, should be viewed very tion obstacles is that set of individuals who
skeptically. believe, explicitly or implicitly, that the first-
Except for the obvious and important point generation solutions have been found. Many
of assuming a ready supply of inputs such as additional answers are needed, and any com-
fertilizer, the literature of agricultural devel- placency on varietal research would be most
opment has little to offer in the way of posi- unfortunate both in terms of growth and re-
tive suggestions for dealing with the agricul- gional equity. Clearly also, a real revolution
tural production alternatives for millions of will require greatly expanded investments in
small Asian farmers. Providing credit in kind irrigation and substantial improvements in
(as under the BIMAS program in parts of In- systems for pest control.
donesia) has worked in some circumstances, Second, the sudden increases in agricul-
as have a few cooperative arrangements. The tural output have already or will soon neces-
program at Comilla in East Pakistan, for ex- sitate basic pricing decisions on the parts of
ample, has shown the merit of cooperative governments. It would be a great pity if the
 452 AGRICULTURAL STRATEGY

nations of Asia, in the face of remarkable between regions. Lest this view be regarded
productivity changes, maintained pricing as too bleak, it should also be emphasized
structures that did not keep in mind the that without the green revolution, the devel-
needs of the entire economy. As a result of opment situation in these countries would
the increased production from the green rev- now be even more dire.
olution, there is considerable potential for ex- Finally, although it is important to recog-
panding the development effort with invest- nize and understand what has happened in
ment programs that are wage-good intensive. the past, the great challenge of the future will
As regards exports, the developed countries be to forge institutions that can deal simul-
could play a major facilitating role; however, taneously with the demographic explosion,
their probable increased use of dumping pro- rapid economic growth, and equality of in-
grams will provide the most formidable kinds come distribution. Certain obvious mistakes
of competition for those developing nations in policies can be avoided, such as the subsi-
who generate export surpluses. Hence, the in- dization oftractors. However, there is little in
ternal market opportunities and the external the way of a broad, institutional blueprint in
market difficulties indicate the probable need the history of the developed countries or in
for downward adjustments in relative grain the general writings of agricultural econo-
prices in several Asian nations. mists that is now of much help on this issue.
Third, the limited technological revolution The Asian challenge of the future will be to
in agriculture has permitted an easing of one encourage growth elements in the econ-
critical development constraint. It has not, omy— such as the green revolution — while at
however, provided a panacea for the employ- the same time fostering equity so as to pre-
ment and equity problems, and indeed has vent an ascending spiral of violence and
probably been destabilizing in the sense that repression.
it has widened income disparities within and
 VII. C. Designing an Agricultural
Strategy
VII.C.1. The Case for a
Unimodal Strategy*

The historical experience in a number of

pecially relevant to the design of strategies
countries, and the recent technical break-
for agriculture that are efficient in this broad
throughs of the Green Revolution, justify sense:
major emphasis on increases in factor pro-
ductivity. Itis, however, the experience of (1) Contributing to the overall rate of eco-
nomic growth and the process of structural
Japan and Taiwan that is especially useful in transformation,
demonstrating that an appropriate sequence
of innovations based on modern scientific (2) Achieving a satisfactory rate of in-
knowledge and experimental methods makes crease infarm output at minimum cost by en-
possible an expansion path for the agricul- couraging sequences of innovations which ex-
ploit the possibilities for technical change
tural sector that is characterized by large in-
creases infactor productivity throughout the most appropriate to a country's factor
endowments,
agricultural sector. Such a strategy enables a
widening fraction of the working population (3) Achieving a broadly based improve-
ment in the welfare of the rural population,
in agriculture to be associated with increas- and
ingly productive technologies, based mainly
on expanded use of purchased inputs that are (4) Facilitating the process of social mod-
divisible and neutral to scale. It is because ernization (including the lowering of birth-
rates, the extension and improvement of rural
the new inputs of seed and fertilizer, that are
the essence of the Green Revolution, are education, and the strengthening of entrepre-
complementary to the large amounts of labor neurial capacities) by encouraging wide-
spread attitudinal and behavioral changes
and land already committed to agriculture
among farm households.
that these increases in factor productivity can
have such a large impact on total farm out- I believe that it is useful to assess the "total
put. At the same time, by involving an in- efficiency" of alternative agricultural strate-
gies in terms of their relative success in
creasingly large fraction of the rural popula- achieving those four objectives.
tion in the process of technical change, such
a strategy means that the fruits of economic The concept of total efficiency is, quite ob-
viously, difficult to define operationally. That
progress are widely shared. is inherent in the nature of the problem. But
The thrust of this argument is that it is
the problem must be confronted because only
possible and desirable to devise and imple-
ment agricultural strategies which are effi- when a country's agricultural strategy is ef-
ficient irthis broad sense is the trade-off be-
cient in terms of a number of objectives, in- tween the goal of increased output and other
cluding but not confined to the objective of
objectives likely to be minimized. Indeed, it
achieving desired increases in farm output at
is my contention that with an agricultural
low cost. The following objectives, which are
strategy that is designed with those multiple
examined later in some detail, seem to be es-
objectives in view, the trade-off is likely to be
*From Bruce F. Johnston, "Criteria for the Design small or nonexistent.
of Agricultural Development Strategies," Food Re- A country's overall strategy for agriculture
search Institute Studies in Agricultural Economics.
Trade, and Development, Vol. 11, No. 1, 1972, pp. is a composite of substrategies relating to re-
35-7, 42-54. Copyright 1972 by the Board of Trust- search, education, water resources develop-
ees of Leland Stanford Junior University.
ment, promotion of farmers' organizations,

453
 454 AGRICULTURAL STRATEGY

marketing and price policy, credit and the cial demand is small in relation to a farm
distribution of inputs, agricultural taxation, labor force that still represents some 60 to 80
land tenure, policies affecting the nature and percent of the working population.
pace of mechanization, and other elements. The late developing countries face a wide
The total efficiency of the strategy depends choice of farm equipment embodying large
on the complementarities among those var- investments in research and development ac-
ious activities and the quality of implemen- tivity in the economically advanced coun-
tation as well as decisions with respect to the tries. The performance characteristics of
allocation of funds and personnel and policies these machines are impressive, and represen-
for individual substrategies. . . . tatives of the major manufacturing firms in
the economically advanced countries are ex-
THE CHOICE BETWEEN perienced and skillful in demonstrating their
UNIMODAL AND BIMODAL equipment. And they now have added incen-
AGRICULTURAL STRATEGIES tive to promote sales in the developing coun-
tries to more fully utilize their plant capacity
The most fundamental issue of agricul- which is large relative to domestic demand
tural strategy faced by the late developing (mainly a replacement demand since the pe-
countries is to choose between a bimodal riod of rapid expansion of tractors and trac-
strategy whereby resources are concentrated tor-drawn equipment in the developed coun-
within a subsector of large, capital-intensive tries has ended). The availability of credit
units or a unimodal strategy which seeks to under bilateral and international aid pro-
encourage a more progressive and wider dif- grams temporarily eliminates the foreign ex-
fusion oftechnical innovations adapted to the change constraint to acquiring such equip-
factor proportions of the sector as a whole. ment; and when such loans are readily
The essential distinction between the two ap- available it may even appear to be an attrac-
proaches isthat the unimodal strategy em- tive means of increasing the availability of
phasizes sequences of innovations that are resources — in the short run. Within devel-
highly divisible and largely scale-neutral. oping countries there is often considerable
These are innovations that can be used effi- enthusiasm for the latest in modern technol-
ciently bysmall-scale farmers and adopted ogies. But little attention is given to research
progressively. A unimodal approach does not and development activity and support ser-
mean that all farmers or all agricultural re- vices to promote the manufacture and wide
gions would adopt innovations and expand use of simple, inexpensive equipment of good
output at uniform rates. Rather it means that design, low import content, and suited to the
the type of innovations emphasized are ap- factor proportions prevailing in countries
propriate toa progressive pattern of adoption where labor is relatively abundant and capi-
in the twofold sense that there will be pro- tal scarce. . . .
gressive diffusion of innovations within par- Under a bimodal strategy frontier firms
ticular areas and extension of the benefits of with their high capital to labor ratio would
technical change to new areas as changes in account for the bulk of commercial produc-
environmental conditions, notably irrigation tion and would have the cash income re-
facilities, or improved market opportunities quired to make extensive use of purchased in-
or changes in the nature of the innovations puts. Inasmuch as the schedule of aggregate
available enable farmers in new areas to par- commercial demand for agricultural prod-
ticipate inthe process of modernization. Al- ucts is inelastic and its rightward shift over
though abimodal strategy entails a much time is essentially a function of the rate of
more rapid adoption of a wider range of mod- structural transformation, to concentrate re-
ern technologies, this is necessarily confined sources within a subsector of agriculture in-
to a small fraction of farm units because of evitably implies a reduction in the ability of
the structure of economies in which commer- farm households outside that subsector to
 455 DESIGNING AN AGRICULTURAL STRATEGY

adopt new purchased inputs and technolo- innovations, the best firms in the agrarian
gies. In addition, the high foreign exchange sector display essentially the same factor in-
content of many of the capital inputs em- tensities as average firms. Interfarm differ-
ployed inthe frontier sector implies a reduc- ences in performance will be large, especially
tion in the amount of foreign exchange avail- during transitional periods as farmers are
able for imported inputs for other farm firms learning how to use new inputs efficiently,
(or for other sectors). It is, of course, because but this will reflect mainly differences in out-
of these purchasing power and foreign ex- put per unit of input rather than major dif-
change constraints that it is impossible for ferences infactor proportions. Inasmuch as
the agricultural sector as a whole to pursue a the expansion path for the agricultural sector
crash modernization strategy. It might be ar- associated with a unimodal strategy implies a
gued that a proper farm credit program could level of capital intensity and foreign ex-
eliminate the purchasing power constraint, change requirements that are compatible
but the availability of credit (assuming that with a late developing country's economic
repayment takes place) merely alters the structure, more firms within the agricultural
shape of the time horizon over which the con- sector are able to expand their use of fertil-
straint operates. And capital and government izer and the other divisible inputs that dom-
revenue are such scarce resources in a devel- inate purchases under this strategy. Thus, the
oping country that government subsidy pro- diffusion of innovations and associated inputs
grams are not feasible means of escaping will be more broadly based, and the diver-
from this constraint. In brief, bimodal and gence in factor intensities between frontier
unimodal strategies are to a considerable ex- firms and average firms will be moderate.
tent mutually exclusive. Although the foregoing has emphasized
Under the bimodal approach the diver- the contrast in the pattern of technical
gence between the factor intensities and the change, it is apparent that the two strategies
technical efficiency of "best" and average will have significantly different impacts on
firms is likely to become progressively greater many dimensions of economic and social
as agricultural transformation takes place. change. Most obvious are the differences in
Moreover, both the initial and subsequent di- the nature of demand for farm inputs, but the
vergences between the technologies used in structure of rural demand for consumer
the two sectors are likely to be accentuated goods will also be very different under a uni-
because the factor prices, including the price modal as compared to a bimodal strategy.
of imported capital equipment, faced by the A major difference in income distribution
modern sector in contemporary developing is to be expected because of the likelihood
countries typically diverge from social oppor- that under a bimodal strategy the difficult
tunity cost. This divergence is obvious when problem of absorbing a rapidly growing labor
subsidized credit is made available on a ra- force into productive employment would be
tioned basis to large farmers and when equip- exacerbated whereas under a unimodal strat-
ment can be imported with a zero or low tar- egy there is a good prospect that the rate of
iff at an official exchange rate that is increase in demand for labor would be more
overvalued. In addition, the large-scale farm- rapid than the growth of the labor force. Un-
ers depend on hired labor rather than unpaid deremployment and unemployment would
family labor. The wages paid hired labor may thus be reduced as a result of wider partici-
be determined by minimum wage legislation, pation ofthe rural population in improved in-
and even without a statutory minimum the come-earning opportunities. This improve-
price of hired labor is characteristically ment in income opportunities available to
higher than the opportunity cost of labor to members of the rural work force would result
small farm units. . . . in part from increased earnings as hired labor
Under the unimodal strategy with its em- since rising demand for labor would tend to
phasis on highly divisible and scale-neutral raise wage rates and the number of days of
 456 AGRICULTURAL STRATEGY

work available during the year for landless economy. Outward labor migration and in-
laborers and for very small farmers whose in- creased farm purchasing power are synchro-
comes derive to a considerable extent from nized with the growing importance of com-
work on farms that are above average size. modity flows between agriculture and other
Most important, however, would be the in- sectors: a flow of food and raw materials out
creased incomes earned by farm households of agriculture and a return flow of farm in-
cultivating their own or rented land. The ex- puts and consumer goods from the manufac-
tent to which tenants would be able to share turing sector. Tertiary activities of govern-
in the increased productivity resulting from ment, transport, marketing and other service
yield-increasing innovations will be deter- industries expand to meet the needs of indi-
mined by forces related to land reform as an vidual sectors and to facilitate the linkages
aspect of broadly based improvement in the between them.
welfare of the rural population. Basically, Agricultural exports have special signifi-
however, it will depend upon the rate of cance here for two reasons. First, in countries
growth of the rural population of working age that have experienced little structural trans-
seeking a livelihood in farming or in nonfarm formation there are usually few alternative
activities relative to the rate of expansion of means of meeting the growing demands for
income-earning opportunities. The latter will foreign exchange that characterize a devel-
be influenced strongly by the demand on the oping economy. Secondly, expanded produc-
part of landowners for labor "hired" indi- tion for export makes it possible to enlarge
rectly through tenancy arrangements, or farm cash incomes when the domestic market
hired directly as laborers on owner-operated for purchased food is still very small, and at
farms. the same time it provides a stimulus and the
means to establish some of the physical infra-
THE MULTIPLE OBJECTIVES OF structure and institutions that are necessary
AN AGRICULTURAL STRATEGY for the creation of a national, market-ori-
ented economy.
In the paragraphs that follow I comment The structure of rural demand for farm in-
briefly on some of the reasons why the design puts associated with alternative agricultural
of an efficient strategy for agriculture should strategies exerts an important influence on
be guided by explicit consideration of four the growth of local manufacturing as well as
major objectives of an agricultural strategy on the pattern of productivity advance within
and the interrelationships among them. . . . agriculture. I emphasize the composition of
Contributions to overall economic growth this demand because the capacity of the ag-
and structural transformation. It is conven- ricultural sector to purchase inputs from
tional when considering agriculture's role in other sectors is powerfully constrained by the
economic development to catalog a number proportion of the population living outside
of specific "contributions." Several of these agriculture. Pathological growth of popula-
contributions imply a net transfer of factors tion in urban areas only loosely related to the
of production out of the agricultural sector as growth of ofT-farm employment opportunities
the process of structural transformation is a conspicuous and distressing feature of
takes place. Typically the farm sector pro- many of the contemporary less developed
vides foreign exchange, public and private in- countries, but basically this growth of urban
vestment resources, and labor to the more population depends on the transformation of
rapidly expanding sectors of the economy as a country's occupational structure that is a
well as increased supplies of food and raw concomitant of economic growth.
materials to support a growing urban popu- The nature of the linkages between agri-
lation and manufacturing sector. culture and the local manufacturing sector
These contributions are, of course, synon- and the seriousness of foreign exchange and
ymous with the increased sectoral interde- investment constraints on development will
pendence that characterizes a developing be influenced significantly by the structure of
 457 DESIGNING AN AGRICULTURAL STRATEGY

rural demand for both inputs and consumer in their use of fixed and working capital and
goods. Because of their differential effects on associated differences in their investments in
the sequence of innovations and on rural in- human resources that affect the level and ef-
come distribution, a bimodal and a unimodal ficiency of agricultural research and other
strategy will differ greatly in their aggregate supporting services as well as the knowledge,
capital and foreign exchange requirements. skills, and innovativeness of the farm
The more capital-intensive bimodal strat-
egy emphasizes rapid adoption of mechanical population.
The importance of distinguishing between
innovations such as tractors along with chem- inputs and innovations that are mainly in-
ical fertilizers and other inputs essential for strumental inincreasing output per acre and
increasing crop yields. Even if that type of those that make it possible for each farm
machinery is manufactured locally, the for- worker to cultivate a larger area has already
eign exchange requirements for capital been noted. Biological and chemical innova-
equipment and for components are high, and tions increase agricultural productivity
the production processes require a high level mainly through increasing yields per acre. In
of technical sophistication, large plants, and general the effect on yield of farm mechani-
capital-intensive technologies. zation per se is slight, although certain me-
The unimodal strategy with its emphasis chanical innovations, notably tubewells and
on mechanical innovations of lower technical low-lift pumps may be highly complementary
sophistication and foreign exchange content, to yield-increasing innovations. Indeed, for
such as improved bullock implements and some high-yielding varieties, especially rice,
low-lift pumps, appears to offer greater an ample and reliable supply of water is a
promise for the development of local manu- necessary precondition for realizing the ge-
facturing which is less demanding in its tech- netic potential of the new varieties. This dis-
nical requirements and which is character- tinction between yield-increasing and labor-
ized by lower capital-labor ratios and lower saving innovations is significant because the
foreign exchange content. On the basis of ex- relative emphasis given to these two types of
perience inJapan and Taiwan as well as an innovations largely determines whether de-
analysis of the nature of the supply response velopment ofagriculture will follow a uni-
to the two patterns of demand, it seems clear modal or bimodal pattern.
that a unimodal strategy will have a much The thrust of a unimodal strategy is to en-
more favorable impact on the growth of out- courage general diffusion of yield-increasing
put and especially on the growth of employ- innovations and such mechanical innovations
ment in local manufacturing and supporting as are complementary with the new seed-fer-
service industries. The reasons cannot be pur- tilizer technology. The bimodal strategy em-
sued here except to note the wider diffusion phasizes simultaneous adoption of innova-
of opportunities to develop entrepreneurial tions that increase substantially the amount
and technical skills through "learning by of land which individual cultivators can effi-
doing" that leads to increasing competence in ciently work in addition to the yield-increas-
manufacturing. Progress in metalworking ing innovations emphasized in the unimodal
and in the domestic manufacture of capital
goods are especially significant because they approach.
For reasons discussed above, it is not pos-
are necessary to the creation of an industrial sible for developing countries to pursue the
sector adapted to the factor proportions of a unimodal and bimodal options simulta-
late developing economy. neously. In placing emphasis on reinforcing
Increasing farm productivity and output. success within a subsector of large and capi-
The differences in farm productivity between tal-intensive farms, a bimodal strategy may
modern and traditional agriculture are, of have an advantage in maximizing the rate of
course, to be attributed mainly to their use of increase in the short run because it bypasses
widely different technologies. Those differ- the problems and costs associated with in-
ences in turn are based on large differences volving alarge fraction of the farm popula-
 458 AGRICULTURAL STRATEGY

tion in the modernization process. In a longer same time building infrastructure important
view, however, a unimodal strategy appears to agriculture and other sectors. But because
to be more efficient, especially in minimizing of the organizational problems and particu-
requirements for the scarce resources of for- larly the severe fiscal constraints that char-
eign exchange and loanable funds. acterize a developing country, it seems
Policies and programs to ensure that the doubtful whether this approach can have a
seed-fertilizer revolution is exploited as very substantial effect on underemployment
widely and as fully as possible are clearly of and unemployment in rural areas. . . .
central importance. This emphasizes the im- Other programs also merit attention be-
portance ofadaptive research and of training cause they offer the promise of substantial
and extension programs to promote further benefits relative to their cost, and some of
diffusion of new varieties and to narrow the them can also make a substantial contribu-
gap between yields at the farm level and the tion to the expansion of output by improving
potential yields obtainable. Investments in in- the health and productivity of the rural pop-
frastructure and in land and water develop- ulation. Public health programs such as ma-
ment required to provide environmental con- laria control are notable examples. The suc-
ditions favorable to the introduction of more cess of such programs is, of course, a major
productive technologies are also priority factor underlying the population explosion
needs. . . . and the urgent need for policies and pro-
The distribution of land ownership and, grams that will have both direct and indirect
more particularly, the size distribution of op- effects in encouraging the spread of family
erational units are highly important factors planning. Nutritional programs also deserve
influencing the choice of technique and the attention. The effects on well-being of in-
factor proportions that characterize the ex- creased farm productivity and incomes can
pansion path of the agricultural sector. Both be enhanced considerably if diet changes are
are influenced by policies and practices af- informed by practical programs of nutrition
fecting land tenure which are discussed in the education. . . .
following section. Although it is foolhardy to attempt to treat
Achieving broadly based improvement in the complex and controversial subject of land
the welfare of the rural population. In a tenure in a few paragraphs, the positive and
longer term view substantial improvement in negative effects on rural welfare of land re-
the welfare of the rural population depends form programs cannot be ignored. In Asia
upon the process of structural change which, the land tenure situation is dominated by the
inter alia, makes possible a reduction in the fact that the area of arable land is small rel-
absolute size of the rural population, a large ative to the large and growing farm popula-
increase in commercial demand for farm tion entirely or mainly dependent on agricul-
products, and large increases in the capital- ture for their livelihood. One implication of
labor ratio in agriculture. There are, how- this, which is distressing but beyond dispute,
ever, some more direct relationships between is that for the agricultural sector as a whole
strategies for agriculture and the improve- in these countries the average farm size will
ment of rural welfare that need to be become even smaller — or at least that the
considered. number of agricultural workers per acre of
Rural works programs are probably the arable land will continue to increase for sev-
most frequently discussed measure aimed di- eral decades until a structural transformation
rectly atimproving the welfare of the poorest turning point is reached.
segments of the farm population. There is It is sometimes argued that because of the
much to be said for such programs as a connection between size of holding and
means of providing supplemental employ- choice of technique, redistributive land re-
ment and income to the most disadvantaged form isa necessary condition for a unimodal
members of the rural population and at the strategy. Indeed it is even claimed that the
 459 DESIGNING AN AGRICULTURAL STRATEGY

success of unimodal strategies in Japan and amount of land rented to individual tenants,
Taiwan is attributable to their postwar land the cropping pattern and other farm prac-
reforms, notwithstanding the fact that in tices, and sharing of expenses of inputs.
both countries the basic pattern of progres- These arrangements will tend to maximize
sive modernization of small-scale, labor-in- the landowner's rental income subject to the
tensive, but technically progressive farm constraint that a tenant and members of his
units was established long before World War household must obtain residual income that
II.
represents a "wage" approximately equal to
I am persuaded that an effectively imple- his best alternative earnings or they will not
mented land reform program that brings enter into the agreement. To the extent that
about a more equal distribution of landed the proposition is valid, it means that im-
wealth will not only contribute to the goal of provement inthe welfare of tenants must de-
equity but will also tend to facilitate low-cost pend primarily on improving the income-
expansion of farm output based primarily on earning opportunities available, including the
yield-increasing innovations. Although such possibility of enlarging their own holdings by
a program would appear to be desirable, redistributive land reform as well as the in-
there is reason to believe that for a good crease in demand for labor within and out-
many Asian countries it is not a likely out- side agriculture.
come. It therefore seems important to em- The advantages of organizing agricultural
phasize that historical evidence and logic production primarily on the basis of small-
both contradict the view that in the absence scale units appropriate to the unfavorable
of land reform the pattern of agricultural de- man-land ratios that characterize the agri-
velopment will inevitably accentuate the cultural sector in late developing countries
problems of rural underemployment and un- are enhanced by the new technical possibili-
employment and the inequality of income ties resulting from the seed-fertilizer revolu-
distribution. tion. Although those advantages are to a con-
The critical factor determining the choice siderable extent a function of the size of
of technique and factor proportions in agri- operational units, there are some specific ad-
culture isthe size distribution of operational vantages ofowner cultivation related to pro-
(management) units rather than ownership ductivity considerations as well as the more
units. Past experience, for example in prewar obvious effects on income distribution. Al-
Japan and Taiwan, demonstrates that a though in principle, investments in land im-
highly skewed pattern of land ownership is provement that are profitable will be made by
not incompatible with a unimodal size distri- the landowner, by the tenant, or under some
bution ofoperational units. To a considerable joint agreement, the division of responsibility
extent the widespread condemnation of ten- in decision-making is likely to delay or pre-
ancy, particularly of share tenancy, seems to vent investments even though they would be
stem from a tendency to confuse what is to the advantage of both parties. Owner cul-
really a symptom with the root cause of the tivation also avoids the difficulties that arise
miserable existence that is the plight of so when landlords, responding to higher yields,
many tenant households in underdeveloped raise the percentage share of output that they
countries. The fact that tenants are prepared demand as rent. But if redistributive land re-
to accept rental arrangements that leave form isnot a realistic possibility, widespread
them such a meager residual income is fun- renting of land seems clearly preferable to
damental y a consequence of the extreme the further concentration of land in large op-
lack of alternative income-earning opportu- erational units and the bimodal pattern
nities. The proposition, briefly stated, is that which is thereby accentuated. . . .
bargaining between landowners and tenants Facilitating the processes of social mod-
will tend to result in equilibrium arrange- ernization byencouraging widespread at tit u-
ments with respect to the rental share, the dinal and behavioral changes. The spread of
 460 AGRICULTURAL STRATEGY

economic and technical change among the countries where average incomes are consid-
rural population, buttressed by a network of erably lower seems to provide some support
institutions and communication links, un- for this generalization.
doubtedly has significant effects on the pro- Thirdly, and most important, the reduction
cess of social modernization that go beyond in birthrates in the countryside, resulting
their effects on economic growth. It seems from spontaneous changes in attitudes and
likely that the broad impact of a unimodal behavior as well as behavioral changes in-
strategy would have favorable effects in three duced by government population programs,
areas important to this process of social are likely to be more widespread and have a
change. First, the wide diffusion of familiar- greater effect on the national birthrate under
ity with the calculation of costs and returns a unimodal than a bimodal strategy. For rea-
and of opportunities to acquire managerial sons examined earlier, the bulk of the popu-
experience would appear to provide a favor- lation in the late developing countries is
able environment for the training and re- going to be in the agricultural sector for sev-
cruitment ofentrepreneurs. The same would eral decades or more. Under those circum-
apply, of course, to the wider diffusion of stances rapid reduction in a country's birth-
learning experiences in manufacturing which rate to bring it into tolerable balance with a
is associated with a unimodal strategy. sharply reduced death rate cannot be
Secondly, a broadly based approach to ag- achieved unless family planning spreads in
ricultural development seems likely to gen- the countryside as well as in towns and cities.
erate strong support for rural education as It seems probable that reasonably rapid
well as the institutions more directly related changes in this domain of behavior are more
to promoting increased agricultural produc- likely to take place if the dynamic processes
tivity. It is sometimes argued that large- of economic and technical change affect a
scale, highly commercialized farm enter- large fraction of a rural population involved
prises are easier to tax than millions of small to an increasing extent in formal and infor-
units. Because of the power structure main- mal education and communication networks
tained or created by a bimodal strategy, how- (including mass media). It also seems likely
ever, the greater administrative convenience that the wider spread of improved income
may in practice mean very little. The fact and educational opportunities will affect mo-
that public education, and especially rural tivations inways favorable to the practice of
education, in most of South America seems family planning. . . .
to lag behind progress in other developing

Comment
Considering the efficiency of farmer decision making, several studies have examined the
allocational behavior of peasant producers from the viewpoint of efficiency across farm size
groups, risk, pricing policy, credit, and marketing. See T. W. Schultz, Transforming Tradi-
tional Agriculture (1964); Amartya K. Sen, "Peasants and Dualism with or without Surplus
Labor," Journal of Political Economy (October 1966); D. W. Hopper, "Allocational Effi-
ciency in Traditional Indian Agriculture," Journal of Farm Economics (August 1965); M.
Lipton, "The Theory of the Optimizing Peasant," Journal of Development Studies (August
1968); M. Schluter and T. Mount, "Some Management Objectives of the Peasant Farmer,"
Journal of Development Studies (August 1977).
 461 DESIGNING AN AGRICULTURAL STRATEGY

VII.C.2. A Three-Pronged
Strategy for Rural
Development*
A general conclusion suggested by a policy eludes education as is now generally recog-
analysis perspective is that we should focus nized. Because of their importance and past
on the more serious failings of rural devel- neglect, we stress particularly the need for in-
opment programs and on the most feasible terventions directed at the interrelated prob-
and desirable opportunities for mitigating lems of malnutrition and chronic ill-health
those failures. Our analysis of the constraints among infants and small children. In addi-
and opportunities relevant to today's low-in- tion to the high human and economic costs
come countries leads us to conclude that associated with impaired child development
three areas of historical failure merit priority and the excessive mortality rates among
attention: (1) the failure to design and imple- those vulnerable groups, the risk that small
ment effective strategies for fostering broad- children will not survive to maturity is also a
based agricultural development; (2) the fail- major obstacle to the spread of family plan-
ure to reduce excessive mortality and mor- ning. Considerable attention has been given
bidity among infants and small children and in the past few years to the decline in fertility
the interrelated failure to slow the rapid in the developing world. A number of middle-
growth of population and labor force; and (3) income countries and two low-income coun-
the failure to create effective problem-solv- tries— China and Sri Lanka — have indeed
ing organizations at the local level to im- achieved remarkably rapid declines in fertil-
prove the performance of development ity. Itis hardly coincidental that virtually all
bureaucracies. of the countries that have realized significant
This diagnosis is the basis for our advocacy declines in fertility since 1960 have also
of a three-pronged strategy for rural devel- sharply reduced infant and child mortaility
rates.
opment. The first "prong" relates to the need
for a broadly based, employment-oriented Our third prong concerns organization
pattern of agricultural development. Expan- programs designed to strengthen the institu-
sion of opportunities for productive employ- tional infrastructure and managerial skills
ment, both inside and outside agriculture, is needed for rural development. The neglect of
of central importance. Fuller and more effi- these problems is in part a consequence of the
cient utilization of the relatively abundant re- tendency for economists to focus on what to
source of human labor can facilitate the ex- do while neglecting the "details" of how to do
pansion of output while at the same time it. More than a decade and a half ago Hsieh
generating the incomes that enable the poor and Lee asserted that "the main secret of
to raise their levels of consumption. This Taiwan's development" was "her ability to
analysis and empirical evidence emphasize meet the organizational requirements."2 This
the importance of tightening the labor sup- lesson has, however been largely ignored in
ply/demand situation if increases in returns spite of the accumulating evidence that or-
to labor are to be achieved. ganizational requirements are not easily ful-
The second prong involves strengthening a filled and require serious and sustained atten-
very limited set of social services. This in- tion. Success in implementing the first and

'Bruce F. Johnston and William C. Clark, Rede-

*From Bruce F. Johnston and William C. Clark, signing Rural Development: A Strategic Perspective,
"Rural Development Programs: A Critical Review of Baltimore, 1982, pp. 145-47.
Past Experience," paper prepared for the XVIII In- 2S. C. Hsieh, and T. H. Lee, Agricultural Devel-
ternational Conference of Agricultural Economists, opment and Its Contributions to Economic Growth in
Jakarta, Indonesia, August 24-September 2, 1982 Taiwan, Joint Commission on Rural Reconstruction,
(processed), pp. 6-9. Reprinted by permission. Taipei, 1966.
 462 AGRICULTURAL STRATEGY

second "prongs" of an effective strategy for sumption programs can be expanded only
rural development is unlikely unless the or- with the growth of organizational capacity
ganizational requirements are met. for their effective implementation. This focus
Analysis of the interrelationships among on the need to develop organizational com-
activities in these three areas emphasizes that petence directs attention at another major
program interventions can have significant theme of our strategic perspective — the im-
complementary as well as competitive im- portance oftime and timing.
pacts on each other. We have suggested that Improving rural well-being is an inherently
the components of a rural development strat- time-consuming process. Historical experi-
egy can be designed (and redesigned) in ways ence teaches that, at best, broadly based and
that maximize their capacity for mutual sup- sustainable progress will be measured in dec-
port and minimize the risk that action in one ades, not years. In part, the pace of develop-
program area will compromise efforts in ment is limited by fundamental structural-
other areas. This is obviously a difficult chal- demographic constraints: even under the
lenge, especially since development special- most optimistic of assumptions, today's late-
ists are far from being in agreement concern- developing countries will remain predomi-
ing the need for priority attention to nantly rural societies into the twenty-first
simultaneous action in these three program century. Perhaps even more significant for
areas. policy design, however, is that development is
The greatest danger of the three-pronged essentially a learning process: the growth of
perspective described here is that it will be organizational competence and personal
interpreted as a recommendation to adopt knowledge are as fundamental a requirement
any and all programs which may be advo- as the growth of capital in the conventional
cated in the production, consumption, or or- sense. Because most of the "answers" to de-
ganization areas. No interpretation could be velopment questions aren't known by anyone,
further from our actual intent. On the con- because even those answers which are known
trary, widespread and sustained progress to- by some must be laboriously passed on to oth-
wards improving rural well-being requires ers, because all answers must be adapted to
that only a very few of the highest priority, specific places and times, the learning process
most mutually reinforcing programs should of development necessarily entails a great
be undertaken at all. Strategic choice means deal of trial and error.
choosing not to do a vast number of tactically This much-needed persistence has too
attractive things. Our emphasis on the neces- often been lacking in rural development
sity of an appropriate balance between pro- strategies. The desire to bypass the errors, to
duction- and consumption-oriented activities teach answers instead of facilitating learning,
highlights this key requirement of strategic is strong — and disastrous. Impatience for im-
design. mediate results has often encouraged relief-
The need to conserve scarce resources, to and-welfare approaches to improving rural
undertake only those tasks most central to well-being. Such short-term activities, how-
the improvement of rural well-being, is ever, almost invariably lead to the neglect of
equally acute when we consider the organi- more fundamental long-term tasks: building
zational components of a development strat- organizational and technical capabilities and
egy. A strategic perspective means using all strengthening the capacities of local people to
available forms of social organization — in- meet their own needs. Failure to pursue rural
cluding political, market, and traditional development strategies that are focused on
structures — for the tasks which they can be programs to strengthen indigenous capacities
made to perform reasonably well, while pre- to accelerate the growth of output, expand
serving scarce administrative talent for fo- employment, improve health and nutrition,
cused intervention when and where it can do and slow population growth will mean that
the most good. Effective production and con- the number of people suffering the depriva-
 463 DESIGNING AN AGRICULTURAL STRATEGY

tions of rural poverty will continue to given to the task of evolving equipment and
increase. tillage innovations for that purpose that meet
Those of us who have emphasized the rel- the needs of smallholders with very limited
evance of the experience of Japan and Tai- cash income and purchasing power.
wan have been guilty of teaching answers in- The development "failures" that we have
stead of stressing the need for a learning noted are failures about which something can
process leading to a continually improving se- be done. It must not be forgotten that only a
quence ofprogram designs. In particular, we decade or two ago many of today's middle-
have been much too slow to take account of income countries faced constraints and con-
a major limitation of technology transfer
based on their experience. Because of the rel- comeditions comparable
countries. to those ofof those
A number today'scountries
low-in-
atively homogeneous character of agriculture have made substantial progress in improving
in those two countries and the dominant im- the well-being of their people because they
portance of irrigated production, there was did pursue strategies for rural development
enormous scope for expanding farm produc- emphasizing vigorous and sustained produc-
tivity and output by relying mainly on im- tion, consumption, and organization pro-
proved seed-fertilizer combinations for rice grams similar to the three "prongs" of the
and a few other major crops. In countries strategy that we have advocated. China and
with different and more heterogenous condi- Taiwan are only two of the specific examples
tions, including major reliance on rainfed in which these critical choices played a piv-
production, it is necessary to confront the otal role.
more difficult task of identifying a number of We believe that totheemulate
efforts of
strategies adapted to a variety of agrocli- income countries suchtoday's low-
successes
matic conditions and crops. Experience at will be facilitated to the extent that the de-
ICRISAT, for example, suggests that in velopment debate within these countries and
areas where rainfall is variable and only mar- the international development community
ginally adequate, there is a need for improved can be focused constructively on fundamen-
methods of soil and water management in tal constraints and opportunities, on the
order to realize the yield potential of im- longer term implications of strategic policy
proved varieties and fertilizer. Little sus- decisions, and on systematic efforts to learn
tained and systematic attention has been from both successes and failures.

Comment
General theories of development, such as those of the dual sector variety, are concerned
with trends of real rural wages over the long period. But such general models are silent on
the behavioral and institutional features of the rural sector. More attention needs to be given
to the labor relations and the mechanisms that determine wages, employment, and earnings
in rural areas. For a critical review of the existing literature on labor and other factor markets
in rural areas, see Hans P. Binswanger and Mark R. Rosenzweig, Contractual Arrangements,
Employment and Wages in Rural Markets (1981). This monograph looks at the various
models and theories of labor markets and tenancy with attention to the issues of absent mar-
kets, market failure, collusive power, and the interdependence of markets. A central conclu-
sion isthat an understanding of institutional arrangements or imperfections in any one market
(e.g., the labor market) requires attention to the imperfections in or constraints on other mar-
kets (e.g., the land or credit markets)'.
Other important literature on rural labor markets may be cited: Pranab K. Bardhan, "In-
terlocking Factor Markets and Agrarian Development," Oxford Economic Papers, Vol. 32,
(1980); P. K. Bardhan and T. N. Srinivasan, "Cropsharing Tenancy in Agriculture: A The-
oretical and Empirical Analysis," American Economic Review (March 1971); Clive Bell, "Al-
ternative Theories of Sharecropping," Journal of Development Studies (July 1977); S. N. S.
 464 AGRICULTURAL STRATEGY

Cheung, The Theory of Share Tenancy (1969); Lloyd G. Reynolds (ed.), Agriculture in De-
velopment Theory (1975); Joseph E. Stiglitz, "Alternative Theories of Wage Determination
and Unemployment in LDCs: The Labor Turnover Model," Quarterly Journal of Economics
(May 1974); A. Braverman and T. N. Srinivasan, "Credit and Sharecropping in Agrarian
Societies," Journal of Development Economics (December 1981); A. Braverman and J. E.
Stiglitz, "Sharecropping and the Interlinking of Agrarian Markets," American Economic Re-
view (September 1982).

VII.C.3. The New Rural-Based

Development Strategy*

In low income countries agriculture occupies they can only be met if there is a large in-
half to two-thirds of the labor force, produces crease indomestic food production. The need
on the order of half the gross national prod- cannot be met by marginal redistribution of
uct, provides well over half the consumption income from the more well off to the poor be-
goods (much more, of course, for the poor) cause of sharp differences in food consump-
and is geographically pervasive. A rural- tion patterns. In India, for example, the lower
based strategy emphasizes mobilization of 20 percent of the income distribution spend
labor for productive purposes, recognizes the 60 percent of its increments to income on
importance of capital and attempts to reduce food grains alone and 85-90 percent on food
the capital per worker by turning to a labor and agricultural commodities in total. In
intensive structure of production. Because sharp contrast, the top 10 percent of the in-
the laboring classes spend the bulk of even in- come distribution spend only 5 percent of in-
crements to their income on food, mobiliza- crements toincome on grain and 35 percent
tion of labor requires increased agricultural on food and agricultural commodities in
production. If employment increases without total. One dollar of income removed from the
a commensurate increase in food supplies, in- rich will only reduce demand for grain by 50.
flationary pressure will mount and demand Given to the poor the same dollar increases
policies which reduce employment will be in- the demand for grain by 600. Thus a fiscal
stituted. Thus accelerated growth in agricul- policy balance of a dollar from the rich and
tural production is essential to the mobiliza- a dollar to the poor brings about a twelvefold
tion of labor. imbalance for food. With variation in the
Of course, the relation of agriculture to composition of the food basket these relation-
poverty abatement is as close as its relation ships hold broadly throughout the Third
to growth strategy. Increased food consump- World.
tion is essential to improved welfare for the The converse of these growth and equity
poor. The food needs of the massive numbers relations is equally correct. If the rate of
of the Third World poor are so great that growth of food production is to be substan-
tially increased on a sustained basis it must
almost inevitably be accompanied by policies
*From John W. Mellor, "The New Rural-Based which succeed in providing higher rates of
Development Strategy," presented at the Joint UNI- consumption growth by the poor. They are
TAR/EDI Seminar on Economic Development and
Its International Setting, Washington, D.C., March the people with the latent demand for in-
27, 1979, as a precis of the arguments documented in creased quantities of food. In the short run,
John W. Mellor, The New Economics of Growth — A exports may be increased, often however only
Strategy for India and the Developing World, a
Twentieth Century Fund Study, Cornell University under subsidy, or storage stocks may be en-
Press, Ithaca, New York, 1976. Reprinted by larged. Inthe longer run failure to match in-
permission. creased food production with increased
 465 DESIGNING AN AGRICULTURAL STRATEGY

domestic demand will cause depressed agri- Second, modern agricultural technology
cultural prices and hence loss of the produc- requires an immense supply of purchased in-
tion incentives essential to maintenance of puts such as fertilizers, pesticides and in-
high agricultural growth rates. creased control of water as well as ready
Thus we describe a two-way street. If em- communications with sources of new technol-
ployment isto increase, if labor is to contrib- ogy. The consequent commercialization of
ute to growth, if the poor are to prosper, there agriculture calls for vast investment in roads
must be increased food production. If there is and electrification while providing the basis
to be increased food production there must be for growth stimulating linkages with other
commensurate increase in the effective de- sectors of the economy. Thus while agricul-
mand of the poor which may best be accom- tural development is essential to improve wel-
plished by employing them in productive fare for the poor, it will only occur when vast
processes. investments are made and large numbers of
people trained, many to highly advanced lev-
THE KEY ROLE OF els. Primitive agriculture is no more the sal-
TECHNOLOGICAL CHANGE IN vation of the poor than capital intensive
industry.
AGRICULTURE
Unfortunately, achieving accelerated CONSUMER DEMAND
growth in agriculture is not effortless. An- STIMULATED GROWTH
nual growth rates of 2.5 percent, more or less
keeping up with population growth, have The new agricultural technology creates a
generally been achieved by gradual increases prosperous peasantry and hence demand for
in cultivated area and greater intensity of production inputs such as pumps, improved
labor use inevitably accompanied by declin- tools and fertilizers as well as for consumer
ing productivity and income. A dynamic role goods. The peasantry already with adequate
in development requires agricultural growth calories wish more of vegetables and livestock
of 4 percent or even higher. That can rarely products, travel, housing and furnishings,
be achieved except by application of modern textiles, and electronics. These are largely
science to development of high yielding crop goods and services which are efficiently pro-
varieties and practices. New technology not duced by labor intensive processes. Thus the
only increases yields per acre, thereby cir- economic ring is closed. More food is pro-
cumventing the troublesome problem of rap- duced, the producers exchange that food for
idly approaching limits to expanding the cul- goods and services from other sectors, em-
tivatable area, but also increases labor ployment ismade in the other sectors for low
productivity and overall efficiency in resource income people who spend the bulk of those
use. The requisite new agricultural technol- increments to income on food. Because de-
ogy has two characteristics of importance to mand patterns are skewed towards non-agri-
development planning. cultural goods and services, those other sec-
First, because of the variability of agricul- tors, when stimulated by agriculture will
tural conditions it is only infrequently that accelerate to 10-15 percent growth rates
new technology can be transferred from one compared to the 4 percent considered re-
environment to another. There is a conse- markable in agriculture. Thus the overall
quent need for a vast system of agricultural growth rates accelerate as the faster growing
experiment stations covering a wide range of sectors gradually become relatively more im-
conditions. The development of trained per- portant. Industrial development is funda-
sonnel and the institutions to effectively uti- mental in the old and new strategies in
lize that personnel in creating effective new growth. The difference lies in how industrial
technology is one of the most important and growth is achieved and its composition.
difficult to meet features of agricultural A rural led pattern of growth has the fur-
development. ther advantage of being geographically dif-
 466 AGRICULTURAL STRATEGY

fused. While it requires large investments in production processes must be held down as
roads and electrification, it thereby offers op- much as possible. A necessary component of
portunity for manufacturing to be more this effort is to import those essential goods
widely diffused, jobs coming to the rural poor such as fertilizers, aluminum and even steel
rather than the rural poor migrating to the which are efficiently produced only by highly
megalopolis. capital intensive processes. These goods are
The rural based, high employment strat- best paid for by export of labor intensive
egy depends on a delicate balance of suffi- manufactures which serve to further increase
cient demand for industrial goods and ser- employment, broaden distribution of income
vices to make it profitable to invest in their and raise the demand for food. Further, when
production (retained profits and induced sav- the strategy is working at its best, growth in
ings being the engine of capital formation in demand for food may temporarily outpace
this model) and sufficiently labor intensive domestic production while inclement weather
production systems to provide increased em- may reduce food production. In general,
ployment to the low income people who will trade is a much cheaper way of meeting these
spend their income on food and thereby pro- problems than halting the strategy or build-
vide adequate demand and incentive prices ing large food stocks.
for food production. Thus public policy must If the industrial countries do not provide a
be sensitive to the needs for the measures to favorable environment for trade, then the
assist growth in industry on one hand and ef- growth strategy for developing countries will
fective demand for food on the other. have to be more autarkic — more capital to
A more basic difficulty arises if land own- heavy industry and less to agriculture and
ership ishighly concentrated in the hands of consumer goods industries — and of course
a small, wealthy, land owning class. Their slower growth, less employment, and less
consumption patterns may be oriented to- amenities and food for the poor.
wards imported goods or very capital inten-
sive goods. In that case, the strategy will fail
from lack of domestic demand for agricul- THE ACCEPTABILITY OF THE
tural production and inadequate employment NEW STRATEGY
growth. Participation in development will be
limited to a small elite. Why would a country fail to opt for the ag-
ricultural based, high employment strategy
which probably gives faster growth than al-
THE ROLE OF INTERNATIONAL ternative strategies and certainly benefits the
TRADE poor more? The strategy of course has little
appeal to narrowly based elitist governments
The new rural-based strategy is highly de- for it is, after all, a strategy which delivers to
pendent on rapidly growing international the bulk of the farmers and to labor. It is a
trade. It is not a subsistence oriented ap- strategy of broadly based governments. The
proach. Trade is essential as a means of con- dependence of the strategy on international
serving scarce capital, providing an outlet for trade may seem a dubious policy given the
abundant labor and may even be needed to uncertainties of trade policy in the developed
meet short run food needs. nations and the almost complete absence of
The capital demands of the new strategy Third World influence on world trade rules
are immense, not only for modernizing agri- and regulations. Finally, the strategy gives
culture but also to combine with the greatly little emphasis, at least initially, to the heavy
increased labor supply that is mobilized by industry on which military defense depends.
the strategy — all in the context of acceler- Many governments may find this too uncer-
ated growth. Thus while every effort is made tain a world for faith in an agrarian
to raise savings rates, the capital intensity of
approach.
 467 DESIGNING AN AGRICULTURAL STRATEGY

One may ask further, why might the de- fers the essential conditions for such political
veloped nations choose to encourage the new systems. Major groups in developed countries
strategy and thus provide trade and aid guar- claim humanitarian concern for the eradica-
antees for the strategy? The developed na- tion of poverty. It is the rural-based strategy
tions have expressed much concern with the that promises the food and employment es-
long term implications of population growth, sential to poverty abatement. Finally, the
It is the broad participation in the rural- trade requirements of the strategy are in the
based strategy that brings down population long term economic interest of the developed
growth rates. The Western developed nations nations even though they pose short term
espouse broadly representative governments, problems of economic and political adjust-
Again it is the broad participation of farmers ment.
and labor in the rural-based strategy that of-

Comment
Several major studies of strategies for agricultural development can be cited. Especially
informative are: Bruce F. Johnston and Peter Kilby, Agriculture and Structural Transfor-
mation (1975); J. W. Mellor, Economics of Agricultural Development, ed. (1974); T. W.
Schultz, Transforming Traditional Agriculture, (1964); T. W. Schultz, Economic Growth and
Agriculture (1968); C. K. Eicher and L. W. Witt (eds.), Agriculture in Economic Develop-
ment (1964); H. M. Southworth and B. F. Johnston (eds.), Agricultural Development and
Economic Growth (1967); Clifton R. Wharton (ed.), Subsistence Agriculture and Economic
Development (1969); Y. Hayami and V. W. Ruttan, Agricultural Development: An Interna-
tonal Perspective (1971); Arthur T. Mosher, Getting Agriculture Moving (1966); Uma J.
Lele, The Design of Rural Development: Lessons from Africa (1975); Sterling Wortman and
Ralph W. Cummings, Jr., To Feed the World (1978); Bruce F. Johnston and William C.
Clark, Redesigning Rural Development: A Strategic Perspective (1982); C. Peter Timmer,
Walter P. Falcon, and Scott R. Pearson, Food Policy Analysis (1983).
International policies to stabilize commodity prices are analyzed in section VIII. C.

Comment
Special attention should be given to the problems of rural development in sub-Saharan
Africa. Annual rates of increase of major staple food crops in sub-Saharan Africa in 1960 to
1975 were only about 2 percent, compared with almost 3 percent in Asia and over 3.5 percent
in Latin America. Because of higher population growth, annual rates of increase in production
required to meet consumption needs by 1990 are also estimated to be higher in sub-Saharan
Africa, about 4.5 percent, compared with not quite 4 percent for Asia and less than 3 percent
for Latin America. (International Food Policy Research Institute, Food Needs of Developing
Countries, Research Report No. 3, December 1977).
For excellent analyses of agriculture in sub-Saharan Africa, see World Bank, Accelerated
Development in Sub-Saharan Africa: An Agenda for Action (1981); Uma Lele, "Rural Africa:
Modernization, Equity, and Long-Term Development," Science (February 1981); Uma Lele,
The Design of Rural Development: Lessons from Africa (1975); K. Anthony et al., Agricul-
tural Change in Tropical Africa (1979); Shankar N. Acharya, "Perspectives and Problems of
Development in sub-Saharan Africa," World Development (Feburary 1981); Carl K. Eicher
and Doyle C. Baker, Research on Agricultural Development in Sub-Saharan Africa: A Crit-
ical Survey, MSU International Development Paper No. 1 (1982), with an extensive bibli-
ography and literature reviews.
 468 AGRICULTURAL STRATEGY

VII.C.4. Land Reform*

If we compare the different forms of land usually extensively cultivated, or used as cat-
tenure, three distinct patterns emerge, and tle ranges. We may call them latifundia,
we can say that from the standpoint of eco- since this is the term used in the countries
nomic analysis there are really three distinct where they prevail; they are the direct de-
problems of reform. We can leave out some scendants ofthe slave-tilled ranches of the
land systems altogether as irrelevant to our Roman Empire.
subject — the peasant systems, in which land 3. Plantation estates. These are also both
ownership is more or less equally distributed, large properties and large enterprises. They
and communal tenure systems, in which the are usually owned by a company with foreign
land is communally owned (mainly prevalent capital and foreign management, though es-
in Africa). These may need other types of re- tates of a plantation type may also be found
form— reforms of the agrarian structure — in private ownership. The methods of culti-
but they do not need redistribution of own- vation are usually intensive.
ership. We can concentrate on the land sys- Many countries have agrarian structures
tems in which the large estate is the predom- which include estates of two or even three of
inant form of tenure. these types. The land system of Egypt in cer-
We must, however, distinguish sharply be- tain features resembles the Asian form of
tween the different types of large estate. One ownership, while in other features it is a plan-
of the great difficulties in the study of this tation system.
subject is that we have no accepted vocabu- These forms of ownership and enterprise
lary. Much confusion arises from lack of pre- have very little in common with the types of
cise terminology. "Large estate" itself is an large-scale farming found in advanced coun-
ambiguous term, referring to at least three tries, i.e., in countries with an industrialised
different forms of tenure and three different economy and commercialised agriculture.
types of economic organisation. The three The Asian system is found principally in sub-
types are: sistence economies, while latifundia and
1. The type of ownership characteristic of plantations produce mainly for export.
Asian countries, in which the land holding is From the standpoint of economic analysis,
only a property and not a large farm or large the most obvious feature of all these types of
producing unit. The property is leased in ownership is the existence of an institutional
small units to tenant cultivators, either on the monopoly. In Asian countries, where demo-
basis of money rent or on a basis of share- graphic pressure is high, the level of rents is
cropping rents. determined not by the fertility of the land,
2. The large estate, characteristic of South but by the fertility of human beings. Land is
European countries and of Latin America, a scarce factor of production, and would
which is both a large property and a large en- command a high price in terms of its pro-
terprise. This type of estate is managed by duce, whatever the system of land tenure.
salaried officials and worked by labourers The existence of institutional monopoly al-
and people of indeterminate status, squatters lows the landowner to raise rents to a still
or share-croppers. Estates of this kind are higher level. In latifundian systems and in
plantation systems, the estate owner is a mo-
*From Doreen Warriner, Land Reform and Eco- nopoly buyer of labour, controlling the use of
nomic Development, National Bank of Egypt Fiftieth
Anniversary Commemoration Lectures, Cairo, 1955, land rather than its price, and he uses his mo-
Lecture II. (Corrected version reprinted in Carl nopoly power to keep wages low.
Eicher and Lawrence Witt (eds.), Agriculture in Eco- The main economic argument for land re-
nomic Development, New York, McGraw-Hill, 1964, form is the need for securing a more equal
pp. 280-90.) Reprinted by permission. distribution of income by eliminating these
 469 DESIGNING AN AGRICULTURAL STRATEGY

monopoly elements. In the first case the aim and Pakistan speak of "uneconomic farms,"
is to reduce the price for the use of land, i.e., they mean farms which fall below a subsis-
a reduction in rents, and in the second case, tence minimum, not below a technical opti-
the aim is to subdivide big holdings and se- mum. Nor does the argument about efficient
cure a fuller use of land, an increased de- large estates apply generally to latifundian
mand for labour, and higher wages for the systems. The haciendas in Mexico and many
farm worker. of the latifundia in Southern Italy were not
But, it may be objected, will not this redis- efficient large estates on any standard. They
tribution ofownership reduce productivity by wasted both land and labour.
dividing up efficient large estates? If we wish So generally speaking, the argument about
to use this argument, we must consider in "efficient large estates" does not seem to
what sense these estates are to be regarded as apply to the first type or the second type of
efficient. The theory of the firm is always dif- estates which we have distinguished. It does
ficult to apply in agriculture, and as far as the seem to apply to plantation estates which use
under-developed countries are concerned, it intensive methods of cultivation and modern
seems to have very limited application. methods. Every plantation system is a special
The argument that the division of large ag- case. Where there is reason to believe that
ricultural enterprises will cause a decline in sub-division of the estate would lead to a de-
productivity is true on two assumptions: (1) cline in production, then the monopoly effect
that there is competition between the factors on labour must be tackled by a policy for
of production and (2) that there are econo- raising wages, and taxing profits to secure
mies of large scale production. These as- reinvestment in other types of farming pro-
sumptions are generally valid in industri- ducing for local needs. Or the estate may be
alised countries. In England, for example, a divided with safeguards for maintaining effi-
large farm has generally become large be- ciency, as under the Egyptian Land
cause itis a more efficient producer, i.e., it Reform. . . .
produces at lower costs; it can compete more Several other arguments used against land
effectively for the factors of production and reform are false because they are based on
combine them more efficiently, using more projections of conditions in advanced coun-
capital and using it more fully; it can also use tries, and do not take these basic differences
more efficient management and more speci- into account. One argument frequently en-
alised labour. In such conditions there is a countered ininternational discussions is that
presumption that the size of farms is more or because tenancy works very well in England
less adjusted to an optimum scale of output there can be no reason for Asian countries to
for certain types of farming. This optimum abolish tenancy by redistribution of owner-
scale of output is difficult to define pre- ship: what they need is legislation to improve
cisely, and in practice means the minimum the security of tenure for tenants. This ar-
area needed to utilise power-driven gument overlooks the monopoly influence in
machinery. . . . Asian countries. It is true that tenancy works
When we try to apply this argument about well in England, because the conditions of
the scale of production to the under-devel- tenancy are regulated by law, and also be-
oped countries, we shall find that over a very cause land is only one of the many forms of
wide range of conditions it has no validity at holding wealth. If a landowner attempts to
all. In Asian land systems, large estates are take too high a rent, the tenant will prefer to
not large producing units. Land reform in invest his capital in other ways. But in Asian
such systems simply means the transfer of conditions tenancy laws will never suffice
ownership from the landowner to the culti- to counteract the effects of monopoly
vator of the existing small holding. The size ownership.
of the farm is not affected, for there are no Another argument of this kind is that there
large farms. When the Governments of India is no need for expropriation by compulsion.
 470 AGRICULTURAL STRATEGY

This argument runs as follows: "If govern- dian systems the redistribution of ownership
ments wish to encourage ownership, they can will not have adverse effects on production
do this by giving tenants special credit facil- through the division of efficient large units,
ities enabling them to buy their holdings. In though in plantation systems sub-division
Switzerland (or Denmark or Sweden) the may have bad effects, and other ways of
land system has evolved itself by gradual ad- equalizing incomes may have to be used.
justment to modern conditions, and Asian These arguments, however, tell us nothing
countries should therefore adjust their sys- about the positive effects of reform on devel-
tems gradually, without drastic legislation to opment. They are negative arguments which
show it will not do harm. If we are to consider
expropriate owners of land." The logical fal-
lacy in this argument is obvious. In advanced the effects on economic development, this is
countries, an improvement in the economic not enough, and it is the investment aspect
position of agriculture will enable the tenant that must be considered.
to buy his land, and special credit facilities The general economic argument for land
can encourage the acquisition of ownership. reform as distinct from the social argument
In the United States, the proportion of own- for more equality is that these systems of
ership to tenancy rises when agriculture is ownership give rise to large incomes which
prosperous, and special legislation aids farm are not reinvested in production. They give
purchase. In European countries, particu- rise also to social attitudes inimical to invest-
larly in Scandinavia, governments have ment. Landowners spend conspicuously; buy
helped tenants to become owners by giving more land; or invest in urban house property;
them easy credit terms. But in Asian coun- or lend at extortionate rates of interest to cul-
tries, the market price of land is too high in tivators for non-productive purposes. This ar-
terms of what it produces to allow the tenant gument applies with great force to Asian ten-
to purchase his land. If agriculture becomes ancy systems and to latifundian systems. It
more prosperous, either as a result of higher does not apply generally to plantation sys-
prices or better harvests, the share-cropping tems. These may have bad social conse-
tenant will not be able to buy his holding, be- quences, but whatever their defects may
cause the landlord benefits equally from the be, failure to invest productively is not one
increased income, and the tenant's position in of them, or not generally one. (There are
relation to the landlord has not improved. exceptions where plantations keep land
There is no price which the tenant can afford out of cultivation, and these systems cause
to pay which the landlord will be willing to trouble.)
accept. If the tenant is to acquire ownership, In general, the land systems of Asia and
the price of land must be fixed at a level Latin America are strong deterrents to in-
which he can pay, and this will inevitably be vestment and aggravate the shortage of cap-
much lower than the market value of the ital by draining capital from agriculture.
land. All land reforms involve expropriation They undervalue the future. The land own-
to some extent for this reason.
ers' preference for land as a form of holding
In economic terms, there can be no ground wealth can be explained simply as a result of
for paying compensation at all, since the ex- the secure and high return on capital which
isting prices of land are monopoly prices. The results from institutional monopoly. . . .
price that is actually fixed in reform legisla- The crucial question is whether land re-
tion is determined by political bargaining form— the change to small ownership — will
power. give better results in the future. Can it pro-
We can conclude therefore that the exis- mote more investment? . . .
tence of institutional monopoly creates a All we can say as to the investment effect
strong argument for land reform on the is that results depend mainly on what can be
ground of equalizing incomes. We can con- done to give inducements to invest, through
clude that in Asian systems and in latifun- special credit facilities and special forms of
 471 DESIGNING AN AGRICULTURAL STRATEGY

village organisation. We cannot say that re- effects of reform, when there is actually sub-
form will cause more investment: but we can division ofthe land?
certainly say that it is a condition, for with- Here too we can only say that results will
out more income in the hands of the cultiva- depend on how far the new owners can inten-
tor, no investment programme for agriculture sify farming, either by the use of more labour
is likely to have much effect. on the land, or by the use of more labour and
Can we say anything about the production more capital. . . .

Comment
The case for a small farm or reformed agricultural system claims several advantages —
more employment, more equitable distribution of income, and a wider home market for the
manufacturing sector. But the case for small holdings must also address itself to other re-
quirements. Itis also essential to consider the effects of land reform on agricultural produc-
tion, both for exports and for increasing food production. Capital formation is also necessary
in both agricultural and industrial sectors. And efficiency must be achieved, as well as em-
ployment and equity.
The economic gains to be had from a radical modification of land tenure patterns appear
to be greatest during periods when rapid technical changes are opening up new production
possibilities which are inhibited by existing tenure relationships. See Yujiro Hayami and Ver-
non W. Ruttan, Agricultural Development: An International Perspective (1971). In view of
the technical revolution in grain production and the explosive rate of growth in the agricul-
tural labor force, the authors conclude that the payoff to tenure reforms, involving a shift
from share tenure, plantation, and collective tenure systems to smallholder owner-operator
systems, may be greatly increased in the future.
On the need for and reform in connection with the development of agriculture, see D. War-
riner, Land Reform and Development in the Middle East (1957); D. Warriner, Land Reform
in Principle and Practice, (1969); D. Felix, "Agrarian Reform and Industrial Growth," In-
ternational Development Review (October 1960); P. M. Raup, "The Contribution of Land
Reforms to Economic Development," Economic Development and Cultural Change (October
1963); Anthony Y. C. Koo, The Role of Land Reform in Economic Development — A Case
Study of Taiwan (1968); Dale Adams, "The Economics of Land Reform: Comment," Food
Research Institute Studies, Vol. 12, No. 2 (1973); Peter Dorner, Land Reform and Economic
Development (1972); R. Albert Berry, "Land Reform and Agricultural Income Distribution,"
Pakistan Development Review (Spring 1971); Folke Dovering, "Land Reform: A Key to
Change in Agriculture," in Nurual Islam (ed.) Agricultural Policy in Developing Countries
(1972); David Lehmann, Peasants, Landlords and Governments (1974).

VII.C.5. Price Incentives*

Most developing countries for the last quar- lower the prices of food and other agricul-
ter century have had policies designed to tural goods and to increase the prices of man-
ufactured goods. Trade and foreign-ex-
*From Gilbert T. Brown, "Agricultural Pricing change practices have been major
Policies in Developing Countries," Theodore W.
Schultz (ed.), Distortions of Agricultural Incentives, instruments of these policies, along with tax,
Bloomington, Indiana University Press, 1978, pp. 84- direct price, and other market-control mea-
89. Reprinted by permission. sures. The conventional wisdom supporting
 472 AGRICULTURAL STRATEGY

this twisting of the terms of trade against ag- Gale Johnson, greater agricultural productiv-
riculture has had four main pillars, based on ity per hectare in industrialized than in de-
the assumptions: veloping countries is a phenomenon that has
occurred largely since the 1930s, and within
1. that aggregate agricultural production is
the period of differential price policies. In the
not very responsive to price changes;
2. that the chief beneficiaries of higher period 1934-38, grain yields averaged 1.15
tons per hectare in industrial countries, and
prices would be the larger size farmers;
1.14 tons (i.e., the same level) in developing
3. that higher food and other agriculture-re- countries. By 1975, however, grain yields in
lated prices such as clothing would most industrial countries were more than double
adversely affect low-income consumers; those in developing countries, 3.0 tons versus
and
1.4 tons per hectare. This is not to argue, of
4. that manufacturing provides a more rapid
means of growth, and that achieving that course, that the more rapid growth in agri-
cultural yields and production in the indus-
growth depends upon large transfers of in- trialized countries is due solely to differences
come (profits) and foreign exchange from
in price policies. A plausible hypothesis is
agriculture to manufacturing.
that both the more favorable farm prices and
Thus, policies that depress agricultural prices the more rapid growth of farm yields in in-
and increase manufacturing prices will result dustrial countries reflect efforts to support
both in more rapid economic growth and in a farm incomes and increase agricultural pro-
more equal distribution of income. ductivity, while developing countries have
Lagging agricultural production and over- generally been much more concerned about
all economic growth in many developing increasing industrial incentives and produc-
countries that have followed such policies, tion, and urban incomes.
however, have led to increasing concern The focus of this paper is not on the factors
about measures that reduce farm incomes that explain the differences in growth rates
and incentives. Argentina, Egypt, Kenya, the between developed and developing countries,
Ivory Coast, Pakistan, Peru, the Philippines, however, but on the effects of recent pricing
Thailand, and Uruguay are among the coun- policies on today's developing countries. My
tries which have acted to significantly in- hypothesis is that agricultural production, in-
crease agricultural relative to nonagricul- come distribution, and economic growth
tural prices in the last several years. Greater would all benefit from reduction or elimina-
emphasis in the development literature on
tion of distortions that reduce agriculture's
employment-led and rural-development domestic terms of trade. The relation of ag-
strategies and on strategies to meet basic riculture's domestic terms of trade to income
needs and to benefit persons in the lowest 40 distribution and economic growth particu-
percent of the income distribution, has also larly are problems whose complexities outrun
provided new intellectual support for price our capacity to measure by formal models,
policies more favorable to agriculture, in- but they are also problems that are too im-
cluding lessened subsidies for capital goods, portant to neglect.
less over-valued exchange rates and protec- Several general problems of terminology or
tion for industry, and increased production of analysis should be briefly mentioned. First,
foodstuffs and other wage goods. references to "free-market" prices are to
In -contrast to those of the developing coun- those that would exist in absence of specific
tries, price policies in the industrialized coun- controls and policies that now distort the re-
tries during this period have been distorted in lation of agricultural to nonagricultural
favor of rather than against the farmer, and prices, including those relationships that
the typical problem has been overproduction magnify differences between domestic and
rather than underproduction of agricultural world-market prices. This is not to imply that
commodities. Moreover, as pointed out by D.
world-market prices are "free," or that policy
 473 DESIGNING AN AGRICULTURAL STRATEGY

should aim to equate domestic and world- correlation analysis of the acreage or output
market prices or price relatives at all times. of a particular crop in relation either to the
Neither is it to imply that shadow free-mar- relative prices of that crop versus others, or
ket prices can be known with exactitude, or to several presumably independent variables,
to ignore that we are dealing with neither a such as output price and use of fertilizer,
first nor even a second-best world. Rather, labor, and water. The price elasticity of ag-
the comparison to probable prices "in the ab- gregate agricultural production is less than
sence of controls" reflects only a judgment for individual crops, of course, since substi-
about approximate price relationships. Sec- tution between crops may account for an im-
ond, references to changes in prices are to portant part of the response of a single crop
changes from existing distorted price rela- to changes in its relative prices. This multi-
tions, not from the traditional equilibrium- ple-correlation methodology can be faulted
price starting point of economic theory. Usu- on several grounds, including the lack of in-
ally the discussion concerns what would hap- dependence between prices on one hand and
pen if the distortion between existing prices the level of use of fertilizer and other inputs.
and free-market or "equilibrium" (using that The use of acreage rather than output in
term loosely) prices were reduced. Thirdly, most such studies underestimates the price
prices are an incomplete measure of incen- elasticity of production by not taking account
tives. Ifagriculture's relative productivity is of changes in yields, which have been ac-
increasing, it is possible for agricultural in- counting for an increasingly large part (now
comes and incentives to be growing at the more than half) of the annual growth in
same time that agricultural prices are declin- world food-grain production. Also, it is
ing. Fourthly, a fundamental distinction doubtful that yearly fluctuations in prices are
must be made between measures such as an adequate proxy for changes in the income
price controls, taxes, and subsidies that arti- expectations that determine rates of invest-
ficially lower food and agricultural prices and ment in agriculture and its supporting infra-
measures such as on-farm investment, tech- structure, labor inputs, and the adoption of
nological advances, and rural infrastructure new techniques. These decisions appear to
development (e.g., roads, electricity, and depend more upon whether expected profita-
water) that lower prices by lowering the real
bility isabove or below a "threshold" level of
costs of production. Reductions in real re- acceptability, and may be little affected by
source costs of production (i.e., increases in yearly price fluctuations unless these cause
input yields) may benefit urban dwellers, longer-run expectations of profitability to
farmers, and rural laborers alike. change. Moreover, incentives are a function
of net income and not just prices, and there-
EFFICIENT USE OF RESOURCES fore a new technology may importantly
change the incentive effect of a given set of
Higher prices may stimulate agricultural prices as happened during the Green Revo-
production by (1) causing producers to move lution inwheat and rice
closer to their production-possibility frontier Price incentives may cause farmers to use
by better use of resources, (2) encouraging improved seeds, along with more fertilizer,
use of more labor and other variable resource pesticides, and other purchased inputs, to
inputs to reach higher production-function adopt improved cultural practices, and to
and output levels, or (3) inducing investment apply more family or hired labor. All of these
and the discovery and adoption of new agri- ways of increasing the efficiency of resource
cultural technologies that result in new, use may occur at once, for example, if a small
lower-cost production functions. farmer decides to reduce or give up his off-
The usual empirical basis of arguments farm employment — or a son does not migrate
that agricultural production is not very re- to the city — in order to adopt the more labor-
sponsive toprice changes rests upon multiple- intensive techniques required to reap the ben-
 474 AGRICULTURAL STRATEGY

efits of new varieties. Generalities about most through price-induced shifts in the produc-
small farmers as "subsistence" farmers unaf- tion function, rather than through greater ef-
fected by agricultural prices are at best ficiency ofresource use with existing produc-
misleading. tion functions. These long-run effects depend
Timmer and Falcon have found a close upon the extent to which the incentive struc-
rank correlation between unhulled rice ture has an important effect upon technolog-
(paddy) prices and rice yields among Asian ical change (both on research and on adop-
countries. For 1970, they found that rice tion of new techniques by farmers), on public
yields varied from 5.64 and 4.53 metric tons and private investment related to agriculture,
per hectare in Japan and Korea to 2.1 to 1.7 and on institutional change affecting agricul-
tons per hectare in Indonesia, Thailand, the tural output (land reform, extension services,
Philippines, and Burma. At the same time, marketing, and distribution facilities).
the ratio of the price of a kilogram of rice to Clear links have been demonstrated be-
the price of a kilogram of fertilizer nutrients tween pricing policy and both the discovery
showed a similar wide variation, from 1.43 and nature of new technological discoveries,
and 0.96 in Japan and Korea to 0.4 to 0.1 in and the adoption of these techniques by
Indonesia, Thailand, the Philippines, and farmers. The relationship between relative
Burma. This pattern of inter-country rela- prices of land and labor in Japan and the
tionships between prices and yields cannot be United States, and the very different direc-
explained as short-term elasticity response, tions of technological change in these two
but may well be indicative of the long-run re- countries in the last century, has been well
sponsiveness of production to incentives. documented. The late-1973 jump in oil prices
While the price data used in this study were has also presented ample evidence of the link
only for one year, the difference in national between prices and research. Much of the ag-
prices are very substantial, and reflect long- ricultural research initiated since then is "en-
standing differences in price policies among
ergy saving," with heavy emphasis on reduc-
these countries. It is noteworthy that the ing fertilizer requirements, for example,
three countries with the highest yields — through nitrogen fixation, seed treatment,
Japan, South Korea, and Taiwan — have and placing a small amount of fertilizer close
some of the poorest soils among the nine to the seed or seedling roots.
countries studied. The high prices to farmers The link between price incentives and pri-
in these countries appear necessary to cover vate investment in chemical inputs, labor,
the costs of achieving these yields under the land leveling, irrigation, and other measures
given climate, soil, and other conditions, even to increase farm output is almost axiomatic.
though the decisions to pay such high prices Studies of diffusion of new techniques and of
may have been primarily political. It is inter- new investment in developing countries indi-
esting to speculate to what levels rice yields cate that where profitability of adopting a
and production would fall in Japan if farmers new technique or investing, say, in a tube
there received the same price as Thai farm- well is very high, that the new techniques will
ers, or conversely the levels to which rice be rapidly adopted and the new investment
yields and production would climb in Thai- quickly made. At the margin at least there
land within a few years if Thai farmers faced are wealthy individuals who make choices be-
the same rice and fertilizer prices as Japa- tween agricultural and nonfarm investments.
nese farmers.
A large proportion of upper and even middle-
class individuals in South Asia reside in cities
SHIFTS IN THE PRODUCTION and have a profession, a business, or a gov-
FUNCTION ernment job, but own substantial farmland
operated by a hired manager or tenant
The most important long-run effects of (sometimes a relative). For these individuals,
price incentives on production may be as for the small farmer deciding how to di-
 475 DESIGNING AN AGRICULTURAL STRATEGY

vide his time between his own farm and other tries) an important part of that saving is
employment, the choice between farm and likely to be invested in farming,
nonfarm investment (and employment) is a Public investment and institutional
recurring one. Higher returns to farming cer- changes may also be importantly affected by
tainly attract more labor and more invest- prices, but research on these topics is in its
ment (ceteris paribus) into farming. More- infancy. Evidence has been presented that
over, the higher returns generate more public investment in irrigation in the Philip-
income and saving, and, therefore, the ability pines has been closely correlated with the im-
to finance more investment in agriculture, port price of rice. There has certainly been a
Nearly all studies show high marginal saving tendency for some officials to think of non-
rates among even poor farmers. Thus, an im- agricultural development as somehow more
portant part of the additional income accru- important than progress in agriculture,
ing to farmers through higher prices is likely Higher prices for agricultural products will
to result in greater savings and investment. In make at least the nonshadow-priced value of
countries where farm yields are low and re- agricultural projects more attractive. The
turns to additional investment in agriculture more attractive financial return may induce
are high (e.g., where fertilizer use is well highly productive institutional development,
below optimal levels because farmers lack fi- such as more effective agricultural research
nancial resources, as is true in much of South and extension systems, and input distribution
Asia and other food-deficit low-income coun- and output-marketing systems.

Comment
The effects of government intervention in agricultural pricing have become a major concern
to development economists. This concern is directed to an analysis of the interrelationships
among agricultural prices and resource allocation, incentives, income distribution, and
employment.
Nobel Laureate Theodore Schultz, among others, has argued strongly that the economic
potential for agricultural development has been largely unexploited because of government
intervention to suppress economic incentives by pricing agricultural products below compet-
itive market equilibrium relative to inputs. These interventions are based on: (1) underval-
uation of agriculture's contribution to growth derived from the misunderstanding of agricul-
ture as an inherently backward sector incapable of innovation; (2) the false assumption that
the market mechanism is an instrument for middlemen to exploit small farmers and poor
urban consumers; and (3) the false assumption that there is a trade-off between agricultural
efficiency and equity in income distribution, both within agriculture and between rural and
urban sectors. See Theordore W. Schultz (ed.), Distortions of Agricultural Incentives (1978).
A number of studies also present evidence that agricultural pricing policies have had an
adverse effect on (1) the gap between rural and urban income, (2) the incentive to produce
food and export crops, (3) the ability of governments to establish and maintain food reserves,
and (4) employment opportunities in farming, processing, and rural industries. These studies
are surveyed by Carl K. Eicher and Doyle C. Baker, Research on Agricultural Development
in Sub-Saharan Africa: A Critical Survey, M.S.U. International Development Paper No. 1
(1982).
There is also more attention being given to relative food prices as important determinants
of change in the relative and absolute real income of low-income people. For a study of im-
portant trade-offs and conflicts among various direct short-run influences and indirect long-
run effects of price policy on the real incomes of the poor, see John W. Mellor, "Food Price
Policy and Income Distribution in Low-Income Countries," Economic Development and Cul-
tural Change (October 1978).
476 AGRICULTURAL STRATEGY

Comment
A number of empirical studies offer evidence on the positive supply elasticity of agricultural
production in response to price incentives. An excellent summary of empirical estimates of
supply elasticities is presented by Hossein Askari and John T. Cummings, Agricultural Sup-
ply Response: A Survey of the Econometric Evidence (1976). See also Raj Krishna, "Agri-
cultural Price Policy and Economic Development," in Herman M. Southworth and Bruce F.
Johnston (eds.), Agricultural Development and Economic Growth (1967); T. W. Schultz
(ed.), Distortions of Agricultural Incentives (1978); W. Falcon, "Farmer Response to Price
in a Subsistence Economy," American Economic Review, Papers and Proceedings, (May
1964); K. Bardhan, "Price and Output Response of Marketed Surplus of Foodgrains," Amer-
ican Journal of Agricultural Economics (February 1970).

VII.C.6. Food Prices*

Two very different views of the role of food traditional peasant economies. In the pres-
prices in the economic development process ence of new biological and chemical technol-
have dominated thinking in academic and de- ogies that offer significantly higher yields for
cisionmakers' circles since the Second World basic food crops when used properly, price in-
War. General development economists, fol- centives become the major factor in deter-
lowing the implications of the classic eco- mining yields. As empirical evidence has
nomic growth model developed by Arthur been gathered over the past decade demon-
Lewis and others, argued that food prices strating adramatic long-run response to
should be kept low to keep real wages low and price, this neoclassical view has increasingly
thus promote rapid industrialization. A var- been pushed on Third World leaders who are
iant of this argument emerged primarily in
urged to set their prices carefully.1
the Latin American context, where the struc- This article attempts to reconcile these two
tural analysts argued that food prices are ir- views of the role of food prices. It does this
relevant to the long-run development process by examining the role of prices in the pro-
since both producers and consumers are in- duction sector, where the evidence for impact
sensitive to changes in prices. Consequently, in both the short run and long run is quite
political leaders can feel free to manipulate persuasive. In other words, as far as it goes
food prices for whatever short-run political ef- the neoclassical view has increasingly been
fect is desirable. Usually this manipulation
takes the form of keeping urban food prices
'The empirical evidence is reviewed in C. P. Tim-
low to satisfy workers, politically active stu- mer, "Fertiliser and Food Policies in LDCs," Food
dents, and the urban middle class. Policy, Vol. 1, No. 2, 1976, pp. 143-154. More de-
The second, or neoclassical view, holds tailed discussions are in C. P. Timmer and W. P. Fal-
that food prices are a critical factor in farm- con, "The Impact of Price on Rice Trade in Asia," in
G. Tolley and P. Zadrozny (eds.), Trade, Agriculture
ers' decisions about which crops to grow and and Development, 1975;andandtheWillis
how intensively to grow them, even in fairly national Farm Prices SocialPeterson,
Cost of "Inter-
Cheap
Food Policies," American Journal of Agricultural
Economics, Vol. 61, No. 1, February 1979, pp. 12-
*C. Peter Timmer, "Food Prices and Food Policy 21. The production-oriented neoclassical policy ad-
Analysis in LDCs," Food Policy, Vol. 5, No. 3, pp. vice is most clearly argued in T. W. Schultz (ed.),
188-93, 197-99, Butterworth Scientific Ltd, Journals Distortions of Agricultural Incentives, Indiana Uni-
Division, Guildford, August 1980. versity Press, Bloomington, 1978.
 477 DESIGNING AN AGRICULTURAL STRATEGY

pushed on Third World leaders important profitability which, for given technologies
roles that neoclassical analysis has largely ig- and input prices, will depend directly on the
nored on an empirical basis:
relative farm gate prices. Krishna's review of
The differential impact on levels of food con- estimated price elasticities for long-run
sumption (and hence on nutritional status) acreage response reports values greater than
by poor and rich households. Serious con- one for cotton and jute in pre-war India and
straints on data and on modelling methodol- Pakistan, and values exceeding 0.5 for im-
portant food crops for much of the developing
ogy have prevented significant empirical in-
vestigation ofthis important issue. world.2 Clearly, the composition of agricul-
tural output is largely determined by the rel-
The mechanisms by which short-run and ative prices farmers receive for their
long-run food prices determine the distribu- produce. . . .
tion of household income levels. The type of technology used by farmers
When these two added dimensions of the im- and within the food system more generally
also depends on food prices. The issue here is
pact of food prices are included in the neo-
the nature of the production process itself
classical analysis, the policy advice to "get and particularly whether it should be capital
prices right" becomes much more compli- or labour intensive. With low food prices in
cated than assuring farmers prices deter-
mined byworld markets and realistic foreign poor countries, few peasants are able to af-
ford modern agricultural equipment; the
exchange rates. Indeed, the income distribu-
tion and differential consumption effects of level of food prices determines whether cap-
food price changes legitimize much of the ital-intensive agricultural techniques are fea-
sible without government subsidy. The choice
short-run political concern over urban food
prices, even if the advice based on the clas- between capital- and labour-intensive farm-
sical growth model or structuralist analysis is ing depends on the relative factor prices, i.e.,
faulty on empirical grounds. The article con- how expensive capital is relative to farm la-
cludes, then, with a discussion of the role of bour in the face of their differential contri-
food prices in the policy process and the type bution to output. Barker has shown that four-
of analysis modern political economies need wheeled tractors make relatively small pro-
to cope with the complex trade-offs in an in- ductivity contributions for much of Asian ag-
riculture and need subsidies to compete with
terrelated multi-staple food system.
labour in most settings.3 Similarly, the appro-
priate choice of rice milling technology in
THE PRODUCTION SECTOR Java is neither hand-pounding nor large rice
Direct Roles mills. Small rice mills are optimal under a
fairly wide range of factor and output price
Traditional economic theory and more conditions likely to prevail in Indonesia.4
than half a century of empirical investigation
confirm three important direct roles for 2Raj Krishna, "Agricultural Price Policy," in H.
prices in the production sector: in the choice M. Southworth and B. F. Johnston (ed.), Agricul-
tural Development and Economic Growth, Cornell
of crops to be grown; in the choice of tech- University Press, 1967.
nologies used to grow the crops; and, in the
3Randolph Barker, "Barriers to Efficient Capital
choice of input levels needed to produce out- Investment in Agriculture," in T. W. Schultz, (ed.).
put levels, i.e., to aggregate agricultural Distortions of Agricultural Incentives, Cornell Uni-.
output. versity Press, Ithaca, N.Y., 1967.
The evidence is overwhelming that farmers *The evidence is presented in C. Peter Timmer,
are quite sensitive to changes in relative out- "Choice of Techniques in Rice Milling on Java," Bul-
letin ofIndonesian Economic Studies, Vol. 9, No. 2,
put prices between alternative feasible crops.
1973, and in C. P. Timmer et al. The Choice of Tech-
Where permitted (and frequently even where nology in Developing Countries: Some Cautionary
they are not), farmers change the crops they Tales, Harvard Studies in International Affairs, No.
plant in close correspondence to their relative 32, 1975.
 478 AGRICULTURAL STRATEGY

Again, prices are the crucial variables in the tural research directions and its social pay-
socially desirable choice once the technical off: and, on the directions and effectiveness of
productivities of the alternatives are under- institutional change.
stood. Itmust be emphasized, however, that The factors determining the rate of diffu-
improving technical productivities of inputs sion of new agricultural technology have
is at least as important for raising the profit- been studied quite intensively for several dec-
ability of food production as manipulating ades. The profitability of the new technology
the relative prices of inputs and outputs. . . . relative to existing varieties or techniques is
Evidence from the Stanford Project on the the most powerful factor explaining the speed
Political Economy of Rice in Asia indicates a and magnitude of its impact on national
yield elasticity for rice production (with con- agricultural productivity. Given the risks of
stant rice and other cropped areas) of 0.14 experimenting with new techniques and
when varieties and cultivation techniques are varieties in traditional agricultural settings,
kept constant, and 0.33 when these can also this price impact works more slowly than
change in response to price incentives. The farmers' adjustments in fertilizer use when
strikingly positive relationship between the prices change.
relative price of rice to fertilizer and per hec- Vernon Ruttan has been the leading pro-
tare rice yields for ten important Asian rice ponent ofa long-run role for prices in the ag-
producers and the USA is shown in Figure 1 . ricultural sector that is ultimately more pow-
erful than any discussed so far. The induced
Indirect Roles innovation hypothesis grows out of the rec-
ognition of severe limits to agricultural
The long-run impact of prices on agricul- growth from factors of production in abun-
tural production depends partly on letting the dant supply such as labour in many Asian
direct effects discussed above accumulate countries. The nature of the agricultural pro-
and gradually take hold. But three areas of duction process is such that strongly dimin-
indirect impact also contribute to a more sig- ishing returns are likely when any single fac-
nificant long-run price response: on the rate tor is used very intensively. Consequently,
of adoption of new technology; on agricul- few growth opportunities exist using tradi-

1.4 ~* / .Tropical J J,aPanJ., frontier

J, /Overall
/ /frontier •/
l.l
/ r r . South
| 1.0 . J Korea
*- 1.7 ton/ha-
/
10.7 °-9
2a 0.8
Sri Lanka /
i /
o
/
1 °-6
"g, 0.5 Malaysia/ /•
Philippines /Taiwan • ! pr China (above quota; high,
/• / stable yield areas)
£ 0.4 Indonesia PR China (average)
/*/ '
0.2
£ 0.0- 3
1 ■/ > Thailand
- r i ^ ^TV i i 1_ J
3 4
Rice yields (ton/ha)

FIGURE 1 . Relationship between rice yields and the rice price to fertilizer price ratio in selected
countries, 1970.
 479 DESIGNING AN AGRICULTURAL STRATEGY

tional agricultural techniques for countries and social background. Most important, con-
with unbalanced factor endowments.5 sumption response to food price changes de-
In response to food prices rising relative to
level. pends critically on the household's income
the prices of the abundant factors of produc-
tion, societies find it profitable to invest in re- Understanding the extent to which con-
search that supplements the availability of sumers at different income levels change
scarce factors of production, such as land in their food intake as food prices change is ex-
Asia, while being complementary to the tremely important, for only with such infor-
abundant factors. Thus, Japan invested heav- mation isit possible to trace the impact of
ily in biological and chemical innovations to food price changes on nutritional status of at-
raise yields per hectare in the face of severe risk populations. The impetus to do this
land constraints and rising agricultural land comes from economic planners concerned
prices. The fertilizer-responsive seeds devel- about guaranteeing basic needs and from nu-
oped by Japanese scientists thus extended the tritionists concerned about the nutritional
productivity potential of Japan's scarcest ag- status of populations at risk. This broader
ricultural resource while complementing its food policy analysis is just now receiving
abundant labour supply serious methodological and empirical
treatment.
THE CONSUMPTION SECTOR Policy impact is traced through one of
three mechanisms: price effects, exogenous
Ample economic analysis and business ev- income effects, and endogenous income ef-
idence confirm that consumers substitute be- fects. Pinstrup-Anderson, et al. have investi-
tween various commodities as their relative gated the impact of a shift in food supply
prices change. This is increasingly true as the curves (a possible food supply policy) on the
commodity becomes more narrowly defined calorie and protein intake of various income
and, hence, as the scope for substitution in- strata of urban households in Cali, Colombia.
creases. Consumers may find it quite difficult This is the simplest of all possible policy ef-
to substitute between food and non-food fects to trace out because there is a single
items, although they shift quite flexibly be- chain of causation from food supply policy
tween chicken and turkey when the relative through price effects to nutritional (i.e., con-
prices shift significantly. But farmers do not sumption) impact on the urban households.
grow food or non-food and the scope for con- Even so, the methodology requires a full own-
sumption changes of particular commodities and cross-price elasticity matrix by income
is extremely important for agricultural policy strata to translate food price changes into in-
purposes. Traditionally this was the major come-strata-specific consumption changes
reason for analysing consumption patterns, to and a full set of market equations to translate
determine how much and which crops farm- neutral supply shifts into price changes.
ers should be encouraged to grow. Extending the analysis to broader policies
The overall market changes in consump- (food marketing or economic development
tion induced by price changes are composed policies) or to other vulnerable groups (the
of changes from thousands and millions of in- landless rural poor or subsistence-oriented
dividual households, and each household small farmers) adds entirely new dimensions
reacts to price changes according to its own to the complexity of the impact. Exogeneous
circumstances. These circumstances depend income effects via changed employment pat-
on a host of individual factors including terns or opportunities and endogenous in-
tastes, household composition, knowledge, come effects for farmers due to output or
price changes must be added to the price ef-
fects. Income-strata-specific income elastici-
5The most important presentation of this perspec- ties will be needed to translate the income
tive is in T. W. Schultz, Transforming Traditional
Agriculture, Yale University Press, 1964. changes into consumption changes. Much
 480 AGRICULTURAL STRATEGY

more difficult will be the corresponding func- much attention in the past thirty years as a
tional relationships that translate the policy focus of nutrition policy concern, most ex-
changes into income-strata-specific income perts now view the primary nutritional prob-
changes. For the exogenous income effects it lem in the Third World as inadequate energy
will be necessary to specify any changes in intake from traditional foods. Protein tends
employment and/or wages (by income class) to be supplied in adequate amounts from
caused by the policies under analysis. The en- such foods when calorific needs are met, with
dogenous income effects are especially diffi- perhaps the important exceptions of weaning
cult to specify because they depend on both and toddler children and in areas with high
output and price components. Both of these, reliance on tropical roots as the main calorific
but especially the output component, are source. In most environments, substitutions
likely to vary systematically by income which maintain or improve calorific intake,
strata. even if cassava is substituted for rice, will not
More important, the translation of policy result in serious deterioration of nutritional
change into resulting effects on prices and in- status if high protein foods such as legumes
comes must necessarily be done in the spe- and fish continue to be consumed with
cific political, social, and economic context of starchy roots in their traditional pattern.
the change itself. For instance, any price Certainly, little nutritional harm arises from
changes due to a shift in supply will depend substituting equal calories of sorghum or mil-
on whether the country is an importer or ex- let for wheat or rice.
porter, the state of the marketing sector, and The second question is analagous to asking
existing institutional mechanisms of price whether aggregate agricultural output is re-
formation. Similarly, the income effects of a sponsive to the terms of trade between agri-
marketing change will depend on the extent cultural and non-agricultural goods. Just as
of open or disguised unemployment, choice of that question must be answered positively on
technique in processing and distribution, and the basis of long-run evidence drawn from
mechanisms of wage formation. It is difficult cross-section data, so too must the consump-
to generalize about these in the absence of a tion equivalent. On the basis of data from 16
significant number of reliable agricultural developing countries for an average of 10
sector models able to trace through price and years each, Weisskoff estimated the overall
income effects by income class. No such price elasticity of demand for food at —0.87,
models exist at present although several a figure that is predictably much larger in ab-
models are able to trace these effects without solute terms than the —0.16 that Houthak-
disaggregating income. . . . ker derived with similar methodology for Eu-
Substitution of one basic foodstuff for an- ropean countries. At the level of national
other as relative prices change promotes effi- aggregates, consumers do substitute food for
ciency in an economy, encouraging con- non-food and this substitution is more re-
sumers to buy the relatively more abundant, sponsive inpoor countries than in rich.
and hence cheaper commodity. Does such The evidence is similar for cross-section
substitution have any welfare costs? In par- analysis of consumers within countries, with
ticular, are some basic foods nutritionally an intriguing twist. Indonesian data. ... in-
better than their substitutes, and does any dicate significant price elasticities for calories
significant substitution of food in general from rice, maize, and cassava (providing on
take place when overall food prices rise and average three-quarters of Indonesian calor-
fall in real terms? ific intake), but the lowest income group has
The first question is more difficult to an- a smaller absolute price response than the
swer than would first appear since the nutri- higher income groups. There is a real phys-
tional value of foods depends on what is rel- iological constraint below which poor con-
atively least adequately supplied in the diet. sumers cannot go, and nearly all the budget
Although protein deficiencies have received must be devoted to calorie purchases.
 48 1 DESIGNING AN AGRICULTURAL STRATEGY

The final role of food prices in determining larger for the middle and upper classes. The
consumption patterns relates to the forma- direct impact on producers tends to have a
tion of long-run tastes. Although economists similarly skewed effect. Farmers producing
usually assume that tastes are given, the significant market surpluses who will benefit
duality between equilibrium quantities and from higher food prices are typically in the
equilibrium prices that derives from formal upper income strata of rural societies and the
planning models provides the framework for direct effect of higher food prices will tend to
a new interpretation of mechanisms of taste skew further the rural income distribution.
formation that rely on natural climatic ad- This effect will be further exacerbated if sig-
vantage and physical availability of particu- nificant numbers of rural landless workers
lar crops as the basis for a society's basic food must purchase most of their food from the
tastes. The dual of such factors is the (im- market.
plicit) relative price of, say, rice to wheat, or It is only with the impact on food produc-
potatoes to corn, and if these prices were con- tion that the income distribution consequence
sciously manipulated for long periods of time of higher food prices tends to reverse. Given
then tastes would similarly be changed. Ja- the high marginal propensity of the poor to
pan's new taste for wheat products may be a buy food from additional income, economic
case in point. development strategies aimed at raising the
income of the poor often run afoul of inade-
INCOME DETERMINATION quate food supplies. If higher food prices can
promote technical change in agriculture and
The impact of food prices on the level and significantly larger food supplies in the fu-
distribution of incomes is the least under- ture, a growth strategy aimed at reaching the
stood of the three major areas discussed here. poor may become economically and politi-
Much of what is known quantitatively about cally feasible.
these relationships draws from a major re- The impact of food prices on employment
search project at Cornell University directed is the most complicated area of the four. Em-
by John Mellor, and heavily oriented toward ployment isaffected indirectly through na-
India.6 The project looked at the four basic tional food consumption patterns because of
topics of most relevance to the impact of food varying employment intensities of different
prices: the direct impact on the real income foods (vegetables require more labour than
of consumers: the direct impact on producers; wheat). Employment is also affected via price
the impact on the level of food production; impact on the total level of agricultural pro-
and, the impact on employment. duction and hence on demand for farm la-
The obvious point about food prices and bour in general; through cropping patterns
the real income of consumers is the differen- that may include crops grown for export or
tial budget shares of poor versus rich con- industrial use (groundnuts in India require
sumers devoted to food purchases. Higher nearly twice as much labour per acre as cot-
food prices have a disporportionately large ton); and through the choice of inputs used to
impact on the poor although the absolute im- produce food. Obviously, mechanized equip-
pact— the total real loss in income — may be ment may affect employment, although very
careful attention to the mechanical require-
^he most succinct statement of the project results ments of double or triple cropping is impor-
tant. However, more subtle employment
appears in John Mellor, "Food Price Policy and In-
come Distribution in Low-Income Countries," Eco- changes are also caused by input choices.
nomic Development and Cultural Change, October Fertilizer, for example, may substitute indi-
1978. More of the earlier conceptual analysis for the rectly for labour if similar aggregate output
project is contained in John Mellor, The New Eco-
nomics ofGrowth: A Strategy for India and the De- could be achieved by more extensive cultiva-
veloping World, Cornell University Press, Ithaca, tion with higher labour input per unit of
N.Y., 1976. output.
 482 AGRICULTURAL STRATEGY

The major impact of food prices on em- mies is obviously much more complicated
ployment, discussed in the economic devel- than in single food grain economies. The con-
opment literature, is through the level of real sumption picture becomes more complicated
wages. The classic Lewis model of develop- because of the need to know multiple own-
ment with unlimited supplies of labour relies price effects by income class, and cross-price
almost entirely on keeping food prices low as effects also become important. The produc-
a mechanism for rapidly expanding both tion side is made more complicated by the
wage employment and the capitalist surplus substitute possibilities if they are produced
for reinvestment in the modern industrial domestically. Planning intensification pro-
sector. The strategy appears to have worked grammes for rice, for example, when maize,
well for Japan, Taiwan, South Korea, and wheat, or barley are alternatives, requires a
perhaps now for China. ... complex balancing of output price incentives,
The best that can be said now is that the input subsidies, credit programmes, and de-
answers are likely to be country-specific and velopment of suitable seed and production
will require careful integration of detailed technologies. Attempts to raise rice prices to
micro level studies with a sensitive macro increase production, while keeping maize
model capable of capturing income-distribu- prices low to protect poor consumers, may
tion effects, commodity substitution in pro- simply be frustrated by the production sub-
duction and consumption, and price deter- stitution options and the level of alternative
mination mechanisms in both open and technologies, unless dual price systems with
closed economies. extensive subsidies can be implemented.
The complications extend to the import
THE POLICY PROCESS and domestic marketing arenas. Planning
food grain imports, especially if much of the
Many food deficit countries urgently need grain will be available under food aid terms,
higher real food prices as an incentive to mil- is far more complicated if several grains are
lions of small farmers to raise the agricul- being imported (or some exported) and
tural productivity through adoption of mod- changes at the margin in their rate of substi-
ern technology. But those same higher, tution are being attempted. On the domestic
incentive food grain prices will have a dispro- side, the marketing structure for the pre-
portionate impact on food consumption of the ferred grain, typically rice or wheat, is usu-
poor. Many of these people are already suf- ally much more fully developed than for sec-
fering from inadequate protein-calorie in- ondary grains and root crops. The latter are
take, and further reduction in their food con- usually viewed by government planners as
sumption may mean serious malnutrition or inferior foods produced primarily for sub-
death. sistence which deserve little government at-
This dilemma has been resolved histori- tention to production, marketing, or con-
cally in two ways. First, food grain imports sumption issues.
can be used to fill the gap between inade- A number of important countries with
quate domestic production and consumption large populations now seem to be facing the
levels generated by low food prices. Second, prospect of inadequate internal or external
in some countries prices of the preferred food resources to increase availability of favoured
grain have been raised as an incentive to do- food grains fast enough to meet market de-
mestic farmers while secondary grains and mand at constant prices, not to mention the
root crops have been kept cheap, or subsi- latent nutritional demand that would be
dized, to protect the poor. The substitute op- forthcoming at significantly lower prices. In
portunities increase policy flexibility to deal the absence of massive food aid transfers,
with this fundamental dilemma of modern these countries will have to seek food grain
political economies. substitutes for the poorest parts of their pop-
Policy analysis in multi staple-food econo- ulations until long-term investment in agri-
 483 DESIGNING AN AGRICULTURAL STRATEGY

cultural infrastructure, made profitable by protein-calorie intake without subsidizing the

higher incentive prices for the preferred food consumption of the entire population and
grain, begins to transform the domestic pro- hence bankrupting the nation. Such a strat-
duction outlook.
egy calls for high political commitment to in-
A differential price policy by commodity, creasing the access of the poor to adequate
even if it includes direct subsidies on those food. It may also be the only financially fea-
foodstuffs such as cassava and maize that are sible way of coping with protein-calorie mal-
consumed primarily by the poor as a means nutrition over the next several decades.
of implementing the policy, offers the poten- The analytical costs of such a multi-com-
tial to target the nutritional impact without modity price policy are quite high. But these
many of the associated enforcement costs high analytical costs must be compared with
and leakages of target-oriented programmes the costs of subsidizing food consumption for
using more preferred foods. Such a strategy much of the population or attempting to en-
relies on self enforcement. The desire, or ne- force target-oriented food distribution. The
ces ity, ofthe poor to eat staple foodstuffs no combination of high analytical and political
longer attractive to the more wealthy pre- costs is not attractive, but neither are the
sents an opportunity to deal with inadequate alternatives.
CHAPTER VIII

Trade Strategy

This chapter concentrates on the problems connected with the transmission of development
through trade. It is especially concerned with whether there is a conflict between market-
determined comparative advantage and the acceleration of development — whether in pursu-
ing the gains from trade, a country might limit its attainment of the gains from growth.
Materials in the first section present opposing views on this question. Some economists
argue that the accrual of the gains from trade is biased in favor of the advanced industrial
countries, that foreign trade has inhibited industrial development in the poorer nations, and
that — contrary to what would be expected from classical trade doctrine — free trade has in
reality accentuated international inequalities. In contrast, others maintain the traditional po-
sition that foreign trade can contribute substantially to the development of primary exporting
countries and that the gains from international specialization merge with the gains from
growth. A more eclectic approach attempts to identify the various conditions that favor — or
inhibit — a process of export-led development.
Phenomena associated with a changing international division of labor may also exert strong
influence on the rate and structural pattern of a country's development. The concern over the
role of the LDCs in a new international economic order is connected with problems of facil-
itating anew international division of labor: To what extent can primary-producing countries
exercise "resource bargaining power"? What is the scope for import substitution? What are
the potentials for export of manufactured products?
The trade policy of a developing country is closely related to the problems raised by the
foreign exchange constraint on the country's development. A number of theoretically sophis-
ticated arguments support a protectionist trade policy for a developing country. In actual
practice, however, protectionist policies have rarely been adopted out of a reasoned consid-
eration ofhow protectionism might improve the terms of trade, raise the savings ratio, take
account of external economies, or overcome distortions in the labor market. On the contrary,

485
 486 TRADE STRATEGY

it has been the persistent shortage of foreign exchange that has dominated considerations of
trade policy, and protectionist policies have been much more in the nature of ad hoc responses
to recurrent balance-of-payment crises.
These policies have, however, been disappointing in most countries. The policies have not
succeeded in reducing the foreign exchange constraint; indeed, in some cases, it can be
claimed that import-substitution policies have actually intensified the shortage. Nor have the
policies of import replacement succeeded in achieving any widespread degree of industriali-
zation beyond the immediate replacement of the final imported consumer goods; nor has there
been the expected progression from import replacement to production for export markets (as
in the earlier Japanese case); nor has industrial protection been an effective means of amelio-
rating the labor absorption problem. Nonetheless, import-substituting policies have become
self-justifying in LDCs as their trade gap has widened, and the countries have resorted to yet
another round of import restrictions to meet the balance-of-payments problem. These disap-
pointing results of import substitution in practice are reviewed in section VIII. B.
There is now a notable shift of emphasis away from import-replacing policies to the out-
ward-looking policies of export promotion, particularly of semimanufactured and manufac-
tured exports. This requires consideration of a variety of policies — not only the granting of
tariff preferences for exports from the LDCs but, even more important, a reduction of the
effective rates of protection as well as a removal of nontariff barriers. If their nontraditional
exports are to be promoted, the LDCs will also have to pursue policies to take advantage of
the new opportunities provided by the underlying real forces of a changing international di-
vision oflabor.
There has now been considerable experience of inward-oriented and outward-oriented de-
velopment strategies. Section VI I LB reviews the lessons of this experience, emphasizing pol-
icies that liberalize the foreign trade regime and promote exports. Some of the selections also
look to the future, asking about the future growth in world trade and whether the success
stories of East Asian countries can be generalized to a larger group of countries.
In demanding the reform of international trading arrangements, the LDCs in UNCTAD
are seeking not merely more trade, but more trade at higher export prices. In essence, the
trade proposals of UNCTAD attempt to internationalize protection or invert protection in the
sense of having it practiced by developed importing countries in favor of the less developed
exporting countries. Both international commodity agreements for primary products and pref-
erences for manufactured exports have the objective of improving the LDCs' terms of trade
by raising export prices, thereby effecting a transfer of real resources from consumers in de-
veloped countries to producers in LDCs. If balance-of-payments and budgetary considerations
make it unlikely that more resources can be transferred through taxpayers in the developed
countries in the form of "open aid," then the transfer may have to come more covertly through
implicit taxation of consumers in the developed importing countries. The selections in section
VIII.C discuss how the character of trade in primary products might be changed through the
use of "producer power" and international commodity agreements.
Proposals for regional integration, as a means of lessening the dependence on primary ex-
ports and accelerating development, have also gained increasing favor. With slower growth
in the more developed countries and the slow down of trade between the LDCs and MDCs,
there is also now more consideration of how LDC-LDC trade might be promoted. The Note
in section VIII. D appraises the contributions that regional integration, in the form of a cus-
toms union or free trade area, might make to the development of its member nations.
The issues raised in this chapter tend to cut across the objectives of international efficiency
in resource allocation, international stabilization of primary export revenue, and international
redistribution of income. In the more orthodox view of trade theory, these objectives are kept
quite separate, and different policy instruments are advocated as being "first-best" policy for
 487 TRADE STRATEGY

each objective. In international policy discussions, however, the problems become only too
easily intermixed. And just as there is a political controversy over whether the international
monetary system should be linked to development finance, so too there is political disagree-
ment over "aid through trade policy," over the sacrifice of efficiency in international resource
allocation for the sake of an international transfer of resources, and over the burden that must
be shared as a result of adjustments to a changing international division of labor.
 VIII.A. INTERNATIONAL TRADE
AND INTERNATIONAL
INEQUALITY

VIII.A. 1. Trade as an "Engine of

Growth"-Note

An overriding issue in the relations between duction powers, and to augment its annual produce
trade and development is the ultimate ques- to the utmost, and thereby to increase the real rev-
tion of whether there is a conflict between the enue and wealth of the society.2
gains from trade and the gains from growth.1 This idea of "vent for surplus" assumes
Can foreign trade have a propulsive role in that resources are not fully employed prior to
the development of a country? Or, on the trade, and that exports are increased without
contrary, are the dictates of comparative ad- a decrease in domestic production, with the
vantage incompatible with the requirements result that trade raises the level of economic
of accelerated development? activity.
The orthodox interpretation as expounded
by classical and neoclassical economists is More generally, classical economists con-
sidered comparative advantage as determin-
that foreign trade can be a propelling force in
ing the pattern of trade. Not the use of sur-
development. Adam Smith's model of foreign plus resources but resource reallocation
trade postulates the existence of idle land and allowed trade to benefit a country by pro-
labor before a country is opened to world
markets. The excess resources are used to moting a more efficient international alloca-
tion of resources. Without any increase in re-
produce a surplus of goods for export, and sources or technological change, every
trade thereby "vents" a surplus productive trading country is able to enjoy a higher real
capacity that would otherwise be unused. In
income by specializing in production accord-
Smith's words, ing to its comparative advantage and trading.
Between whatever places foreign trade is carried The exports have instrumental significance as
on, they all of them derive two distinct benefits the intermediate goods used for the "indirect
from it. It carries out that surplus part of the pro- production" of imports: exports allow the
duce of their land and labour for which there is no
country to "buy" imports on more favorable
demand among them, and brings back in return
for it something else for which there is a demand.
It gives a value to their superfluities, by exchang- 2Adam Smith, An Inquiry into the Nature and
ing them for something else, which may satisfy a Causes of the Wealth of Nations, Edwin Cannan
part of their wants, and increase their enjoyments. (ed.), 1937, p. 415. For more detailed discussion of
By means of it, the narrowness of the home market Smith's theory, See Hla Myint, "The Classical The-
does not hinder the division of labour in any par ory of International Trade and the Underdeveloped
ticular branch of art or manufacture from being Countries," Economic Journal, 1958, Vol. 68, pp.
carried to the highest perfection. By opening a 317-31. Myint indicates that Smith's concept of sur-
more extensive market for whatever part of the plus productive capacity is not merely a matter of sur-
plus land by itself but surplus land combined with sur-
produce of their labour may exceed the home con- plus labor; and the surplus labor is then linked with
sumption, itencourages them to improve its pro- his concept of "unproductive labor" (p. 323). This in-
terpretation allows Smith's vent-for-surplus model of
'This Note is an abbreviated version of Gerald M. trade and growth to be consistent with W. Arthur
Meier, "External Trade and Internal Development," Lewis's model of development with unlimited supplies
in Peter Duignan and L. H. Gann (eds.), Colonialism
of labor. See R. E. Caves, "'Vent for Surplus' Models
in Africa 1870-1960, Vol. 4, New York, 1975. See of Trade and Growth," in R. E. Baldwin et al., Trade,
also Gerald M. Meier, The International Economics Growth, and the Balance of Payments: Essays in
of Development, New York, 1968, chapter 8. Honor of Gottfried Haberler, 1965, pp. 95-1 15.

489
 490 TRADE STRATEGY

terms than if produced directly at home. The fied or entirely undeveloped, and they may fail to
gain from trade is on the import side; and it put forth the whole of their productive energies for
is significant that the gains are also mutual, want of any sufficient object of desire. The opening
of a foreign trade, by making them acquainted
realized by all the trading countries. By spe-
cializing incommodities for which its costs with new objects, or tempting them by the easier
acquisition of things which they had not previously
are comparatively lowest, a trading nation
thought attainable, sometimes works a sort of in-
would, in Ricardo's words, increase "the sum dustrial revolution in a country whose resources
of commodities and mass of enjoyments"; in were previously undeveloped for want of energy
modern jargon, trade optimizes production. and ambition in the people: inducing those who
Although specialization according to com- were satisfied with scanty comforts and little work,
parative advantage yields the direct benefits to work harder for the gratification of their new
of international exchange, there are in addi- tastes, and even to save, and accumulate capital,
tion dynamic aspects of trade that are rele- for still more complete satisfaction of those tastes
vant for the growth-transmitting effects of at a future time.3
trade above and beyond the static gains. Further, Mill stated that trade benefits the
Classical and neoclassical economists did not less developed countries through
make the dynamic aspects of trade central to
their thought; but to the extent that they did the introduction of foreign arts, which raises the
consider the effects of trade on development, returns derivable from additional capital to a rate
corresponding to the low strength of accumulation;
they saw no conflict between a country's con- and the importation of foreign capital which ren-
formity with its comparative advantage and
ders the increase of production no longer exclu-
the acceleration of its development. Indeed, sively dependent on the thrift or providence of the
John Stuart Mill stated that trade, according inhabitants themselves, while it places before them
to comparative advantage, results in a "more a stimulating example, and by instilling new ideas
efficient employment of the productive forces and breaking the chain of habit, if not by improv-
ing the actual condition of the population, tends to
of the world," and that this might be consid- create in them new wants, increased ambition, and
ered the "direct economical advantage of for-
eign trade. But there are, besides, indirect ef- greater thought for the future.4
fects, which must be counted as benefits of a
The indirect benefits of trade on develop-
high order." A most important "indirect" dy- ment are therefore of three kinds: (1) those
namic benefit, according to Mill, is that widen the extent of the market, induce
the tendency of every extension of the market to innovations, and increase productivity; (2)
improve the processes of production. A country those that increase savings and capital accu-
which produces for a larger market than its own, mulation; and (3) those that have an educa-
can introduce a more extended division of labour, tive effect in instilling new wants and tastes
can make greater use of machinery, and is more and in transferring technology, skills, and en-
likely to make inventions and improvements in the
processes of production. trepreneurship. This emphasis is on the sup-
ply side of the development process— the op-
Widening the extent of the market, inducing portunity that trade gives a poor country to
innovations and increasing productivity remove domestic shortages, to overcome the
through foreign trade allow a country to diseconomies of the small size of its domestic
overcome the diseconomies of being a small market, and to accelerate the "learning rate"
country. of its economy.
Another important consideration, accord- For these several reasons, the traditional
ing to Mill, "principally applicable to an
early age of industrial advancement," is that 3John Stuart Mill, Principles of Political Econ-
omy, 2vols., London, 1848, Vol. II, book III, sec. 5,
a people may be in a quiescent, indolent, unculti- chapter 17.
vated state, with all their tastes either fully satis- 4Ibid., Vol. 1, book 1, sec. 1, chapter 13.
 491 INTERNATIONAL TRADE AND INTERNATIONAL INEQUALITY

conclusion has been that the gains from trade responses within the economy that increase
do not result merely in a once-over change in the productivity of the exporting economy.
resource allocation, but are also continually The staple theory has some relation also to
merging with the gains from development: Rostow's leading-sector analysis insofar as
international trade transforms existing pro- the staple-export sector may be the leading
duction functions and increases the produc- sector of the economy, growing more rapidly
tivity of the economy over time. If trade in- and propelling the rest of the economy along
creases the capacity for development, then with its growth. In Rostow's analysis, how-
the larger the volume of trade, the greater ever, a primary-producing sector can be a
should be the potential for development. leading sector only if it also involves process-
More recently, various export-based ing of the primary product.
models of growth have been formulated to A more general analysis of the effects of
present a macro-dynamic view of how an trade on the rate of growth has been consid-
economy's growth can be determined by ex- ered by Corden.5 Instead of the "demand-
pansion inits exports. One version of the ex- motored" model of the staple theory, Corden
port-based model is that of the staple theory analyzes a "supply-motored" model that em-
of growth. phasizes growth in factor supplies and pro-
The term "staple" designates a raw mate- ductivity. After a country is opened to world
rial or resource-intensive commodity occu- trade, five different effects may be distin-
pying adominant position in the country's guished. First is the "impact effect" corre-
exports. It has a structural similarity to the sponding to the static gain from trade: cur-
vent-for-surplus view insofar as "surplus" re- rent real income is raised. Then there may be
sources initially exist and are subsequently the "capital-accumulation effect": an in-
exported. It also has some affinity with Lew- crease in capital accumulation results when
is's model of development with an unlimited parts of the static gain are invested. This
supply of labor when the surplus to be vented amounts to a transfer of real income from the
through trade is one of labor and not natural present to the future instead of an increase in
resources.
present consumption. Third may be the "sub-
The staple theory postulates that with the stitution effect." This may result from a pos-
discovery of a primary product in which the sible fall in the relative price of investment
country has a comparative advantage, or goods to consumption goods if investment
with an increase in the demand for its com- goods are import-intensive. This would lead
parative advantage commodity, there is an to an increase in the ratio of investment to
expansion of a resource-based export com- consumption and an increase in the rate of
modity; this in turn induces higher rates of growth. The fourth possibility is an "income-
growth of aggregate and per capita income. distribution effect": there will be a shift in in-
Previously idle or undiscovered resources are come toward the factors that are used inten-
brought into use, creating a return to these sively in the production of exports. If the
resources and being consistent with venting a savings propensities differ between sectors or
surplus through trade. The export of a pri- factors, this will have an effect on the overall
mary product also has effects on the rest of savings propensity and hence on capital ac-
the economy through diminishing underem- cumulation. Finally, there is the "factor-
ployment or unemployment, inducing a weight effect." This considers the relative
higher rate of domestic saving and invest- productivity of capital and labor and recog-
ment, attracting an inflow of factor inputs
into the expanding export sector, and estab- 5W. Max Corden, "The Effects of Trade on the
lishing linkages with other sectors of the Rate of Growth," in Trade, Balance of Payments and
economy. Although the rise in exports is in- Growth, Jagdish N. Bhagwati et al. (eds.), Amster-
duced by greater demand, there are supply dam, 1971, pp. 117-43.
 492 TRADE STRATEGY

nizes that if the rate of growth of output is a real income over time as a result of opening
weighted average of capital and labor growth a country to foreign trade,
rates (with a consant returns-to-scale aggre- The positive view of trade and develop-
gate production function), then if exports ment thus emphasizes the direct gain that
rise, and exports use the faster-growing fac- comes from international specialization plus
tor of production, the rate of growth of ex- the additional support to a country's devel-
ports will rise more rapidly. These effects are opment through a number of spread effects
all cumulative, and intensify the increase in within the domestic economy.

Comment
Critics of the view that trade will transmit development often deny the relevance of the
conclusions of neoclassical trade theory. Some claim that the theory of comparative advantage
is static and misses the essence of change in the development process. But the conclusions
derived from the theory of comparative advantage need not be limited to a "cross section"
view and a given once-for-all set of conditions. The comparative cost doctrine still has validity
among countries undergoing differential rates of development. For a simple exposition of how
to incorporate different types of factor growth, technological progress, and changes in the
structure of demand into the neoclassical theory of comparative advantage, see G. M. Meier,
International Trade and Development (1963), chapter 2. For some illustrations of the chang-
ing international division of labor, see "A 'Stages' Approach to Comparative Advantage," in
Irma Adelman (ed.), Economic Growth and Resources, Vol. 5, National and International
Issues (1979).
Another criticism is directed against the "factor-price equalization theorem" derived from
neoclassical theory. This theorem states that, under certain conditions, free trade is a perfect
substitute for complete international mobility of factors and is sufficient to equalize, in the
trading countries, not only the prices of products but also the prices of factors. Against this
proposition, however, critics argue that, in reality, the international distribution of income has
become more unequal. Some, like Myrdal, would contrapose against the factor-price equali-
zation theorem a theory of "cumulative causation" or a "cumulative process away from equi-
librium infactor proportions and factor prices, engendered by international trade." See Gun-
nar Myrdal, An International Economy (1956) and Rich Lands and Poor (1957), chapter 11.
This criticism is, however, highly overdrawn. The critics attribute much more importance
to the theorem than did its expositors who recognized its highly restrictive assumptions and
hence never maintained that equalization will actually occur. The very special set of assump-
tions are: identical production functions in all countries; constant returns to scale exist in the
production of each commodity; and factor and product markets are perfectly competitive.
Insofar as these restrictive conditions have been violated in reality, it should not be surprising
that factor returns have not been equalized between rich and poor countries. See P. A. Sam-
uelson, "International Factor-Price Equalization Once Again," Economic Journal (June
1949).
Although the factor price equalization theorem cannot be claimed to be a valid empirical
generalization, the fundamental contention of neoclassical trade theory does hold — the real
income of each country will be higher with trade than without trade.
 493 INTERNATIONAL TRADE AND INTERNATIONAL INEQUALITY

VIII.A.2. Patterns of Trade and

Development*

[It is] instructive to take a look at past ex- the study of international trade."3 In the sec-
perience and see how economic growth in ond half of the twentieth century this may
certain areas was induced through interna- seem to us a curious statement. It can be un-
tional trade in the nineteenth century. The derstood only in the light of certain historical
areas involved in this process of growth conditions, and it embodies the particular ex-
through trade were chiefly the so-called re- perience ofBritain's economic relations with
gions of recent settlement in the temperate the new countries overseas. Economic growth
latitudes outside Europe. These areas, in in these areas was due not to international
which the United States may be included, re- specialization alone but more particularly to
ceived alarge inflow of labour as well as cap- the fact that the character of trade was such
ital from Europe, but a basic inducement that the rapid growth which was taking place
that caused them to develop was the tremen- in the centre was transmitted to the outlying
dous expansion of Western Europe's, and es- new countries through a vigorous increase in
pecially Great Britain's, demand for the the demand for primary products.
foodstuffs and raw materials which they were Trade in the nineteenth century was not
well suited to produce. Growth at the periph- simply a device for the optimum allocation of
ery was induced, through trade, by growth in a given stock of resources. It was that too, but
the rising industrial centre. it was more than that. It was above all "an
Alfred Marshall referred to "the splendid enginetantofobservation
growth." Thiswhich
profoundly
markets which the old world has offered to is one we owe impor-
to Sir
the products of the new."1 He forgot to men- Dennis Robertson.4
tion the crucial point that these were growing It helps us to see things in perspective, but
markets, but this he probably assumed as a in doing so it serves also to limit the signifi-
matter of course. The penultimate chapter of cance of classical trade theory to its proper
his Principles is entitled "General Influences sphere. The conventional tendency has been
of Economic Progress" and begins as follows: to credit international specialization as such
"The field of employment which any place of- with the spectacular growth of the new coun-
fers for labour and capital depends, firstly, on tries in the nineteenth century. In the light of
its natural resources; secondly, on . . . knowl- Robertson's remark it may perhaps be sug-
edge and . . . organization; and thirdly, on . . . gested that classical specialization theory,
markets in which it has a superfluity. The im- which in the nature of the case is a static
portance ofthis last condition is often under- analysis, has derived more prestige from
rated; but it stands out prominently when we nineteenth-century experience than it has de-
look at the history of new countries."2 served. The dynamic nature of trade as a
It was under the impression of this expe- transmitter of growth was overlooked during
rience that Marshall made the following pro- an era in which progress was taken for
nouncement: "The causes which determine granted, like the air we breathe.
the economic progress of nations belong to There is no doubt that international trade
was peculiarly important in the conditions of
*From Ragnar Nurkse, "Trade Theory and Devel-
the nineteenth century. In real volume it in-
creased tenfold between 1850 and 1913,
opment Policy," in H. S. Ellis (ed.), Economic Devel-
opment for Latin America, Macmillan and Co. Ltd., twice as fast as world production. Imperial-
London; St. Martin's Press, New York, 1961, pp. ism had very little to do with the expansion
236-45. Reprinted by permission.
3Ibid., p. 270.
'Alfred Marshall, Principles of Economics, 8th ed.,
London, 1920, pp. 668-9. 4D. H. Robertson, Essays in Monetary Theory,
2Ibid., p. 668. London, 1940, p. 214.
 494 TRADE STRATEGY

of trade. As was shown by J. A. Hobson him- to be discouraged. Indeed, I should like to as-
self,5 the tropical colonies took a minor share sume that all opportunities in this direction
in the growth of British trade. Continental are fully exploited. The trouble is that in the
Europe and the new countries outside as well mid-twentieth century, with a few notable
as within the British Empire took the major exceptions, conditions for this type of growth
share. The regions of recent settlement were do not, by and large, appear to be as prom-
high-income countries from the start, effec- ising as they were a hundred years ago.
tive markets as well as efficient producers. Since 1913 the quantum of world trade has
Their development was part of the growth of increased less than world production. To be
international trade itself. sure, in the last five or six years we find the
So much for the new countries. Elsewhere, volume of trade in the non-communist world
in the truly backward areas, economic increasing at just about the same pace as pro-
growth induced through international trade duction. But when we look at it more closely
in some cases carried with it certain features we find that it is chiefly among the advanced
that were, and still are, regarded as undesir- industrial countries that international trade
able. It sometimes led to a lopsided pattern has been expanding in the recent past. These
of growth in which production of primary countries, including above all the United
products for export was carried on with the States, are themselves efficient primary pro-
aid of substantial investment of foreign cap- ducers, especially for food. Their demand for
ital, while the domestic economy remained exotic raw materials like crude rubber, silk,
far less developed, if not altogether primitive. nitrates, jute, vegetable oils, hides, and skins
This picture applies especially to tropical has been, and will probably continue to be,
areas. It is the familiar picture of the dual affected by the growth of the chemical indus-
economy resulting from trade and from for- try in the twentieth century. . . . Professor D.
eign business investment induced by trade. D. Humphrey in his voluminous study,
Areas of outpost investment producing for American Imports,1 attaches great impor-
foreign markets often showed a lack of social tance to the technological factor. He esti-
as well as economic integration internally. mates that, in its effect on total United States
Moreover, their export activities were subject imports, the displacement of imported raw
to the familiar hazards of cyclical instability. materials by synthetic products has more
Nevertheless, even unsteady growth than offset the 75 per cent reduction in the
through foreign trade is surely better than no American tariff which has taken place in the
growth at all. Mr. Bauer has given impressive last twenty years partly through duty reduc-
examples of progress resulting from peasant tions and partly through the effect of price
production for export in some parts of West inflation on the burden of specific duties.
Africa during the early half of the twentieth While tariff changes have mainly affected
century.6 Elsewhere foreign capital working imports of manufactured goods from other
for export has usually led to an additional de- industrial countries, technological displace-
mand for local labour, increased wage in- ment has particularly affected United States
comes, expenditures on local materials, new imports from the less developed countries.
sources of taxation, and, in the case of min- Only for minerals are conditions generally
eral concessions, lucrative profit-sharing ar- favorable, although even here it should be
rangements. Al these benefits have helped to noted that, first, the demand for metals is af-
promote expansion in the domestic economy. fected by the increasing efficiency of scrap
The traditional pattern of development collection and recovery in the industrial
through production for expanding export countries. Second, mineral deposits are gifts
markets is not to be despised and ought not of nature, and if a country does not happen
5J. A. Hobson, Imperialism, 3rd ed., London, 1938,
to have any, it can do nothing in response to
ch. 2.
6P. T. Bauer, West African Trade, Cambridge, 7D. D. Humphrey, American Imports, New York,
1955.
1955.
 495 INTERNATIONAL TRADE AND INTERNATIONAL INEQUALITY

the rise in world demands. Some countries It is therefore not surprising that, according
that have deposits fail to exploit them. to the report of the Contracting Parties to the
Nevertheless, the point remains that while General Agreement on Tariffs and Trade, we
Guatemala, for example, can at least try to find the following distribution of interna-
grow chicle, she cannot try to grow nickel. tional trade in the non-communist world in
Third, the export of minerals involves in an 1955. The exports of twenty advanced indus-
obvious sense an element of living on capital. trial countries (United States, Canada,
The growth of synthetic materials is un- Japan, and Western Europe) to each other
doubtedly one explanation of the findings constitute as much as 40 per cent of total ex-
which Professor Kindleberger reaches in his ports. Exports from these twenty countries to
book on The Terms of Trade: A European all less developed countries outside the com-
Case Study. This study lends some support to munist orbit amount to 25 per cent of the
the view that the poorer countries' terms of total. Exports from the less developed to the
trade have shown a tendency to deteriorate. advanced countries represent another 25 per
Kindleberger has calculated industrial Eu- cent. Only 10 per cent of the total are exports
rope's terms of trade separately for various of the less developed countries to each other,
parts of the world, including in particular two even though the more than hundred countries
groups of countries overseas, the areas of re- in this group contain two-thirds of the total
cent settlement, not including the United
population of the non-communist world.10
States, and the poorer countries (the rest of Why is it that so little of the coffee, tea, rub-
the world in his grouping). Difficulties due to ber, and tin produced in these countries goes
quality changes and transport costs apply to to other countries in the same group? Obvi-
both groups. Both the new countries and the ously the main explanation is the low pur-
poor countries are exporters of primary prod- chasing power of people in these countries,
ucts and importers of manufactured goods. which in turn is a reflection of their low
From 1913 to 1952, according to these esti- productivity.
mates, Europe's terms of trade with the areas The fact that the economically advanced
of recent settlement showed a 20 per cent im- countries are each others' best customers is
provement, while in trade with the poorer now more than ever a central feature of
countries Europe's terms seem to have im- world trade. It is chiefly within this small cir-
cle of countries that international trade is
proved byas much as 55 per cent.8
Other recent studies have provided evi- now expanding. With the leading exception
dence that world demand for the poorer of petroleum and a few other minerals, it can
countries' export products has tended to rise hardly be said that primary producing coun-
much less than in proportion to the produc- tries are enjoying a dynamic expansion in
tion and incomes of the advanced countries.9 world demand for their exports. . . .
Professor T. W. Schultz in his paper on
8C. P. Kindleberger, The Terms of Trade: A Eu- "Economic Prospect of Primary Products"
ropean Case Study, New York, 1956, p. 234. shows that the demand for all raw materials,
9For the post-war period this conclusion is docu-
whether imported or domestically produced,
mented in United Nations, World Economic Survey, has lagged far behind the increase in output
1956, and also in the annual report of the Contracting in the United States. What we are consider-
Parties to the General Agreement on Tariffs and
Trade, International Trade, 1955, Geneva, 1956. ing therefore is merely the international as-
For a longer period, Professor Cairncross has made pect of a fairly general tendency. In a coun-
a careful statistical study of world exports of manu- try amply supplied with capital and technical
factured goods since 1900 showing that the manufac- know-how, it seems a perfectly natural ten-
tured goods which the industrial countries export to
each other have constituted a steadily increasing pro- 10 International Trade, 1955. The figures given in
portion oftheir total exports of manufactured articles; this report exclude trade within the communist orbit.
A. K. Carincross, "World Trade in Manufactures For the sake of comparability I have adjusted them so
since 1900," Economia Internazionale, November as to exclude trade between communist and non-com-
1955. munist countries as well.
 496 TRADE STRATEGY

dency for investment in research and devel- The slight increase which has occurred in
opment todisplace crude materials with syn- the last few years in the United States import
thetic products made from a few basic ratio has been due to increased imports of fin-
elements of mostly local origin. These trends ished and semi-finished manufactures. This
are not confined to the United States. They has meant increased trade with other indus-
are affecting the trade of other advanced trial countries, Canada, Western Europe,
areas as well.11 Japan. Imports of crude materials, largely
If this is the situation of the mid-twentieth from under-developed areas, have not re-
century, the mental habits which economists gained their pre-war position in relation to
have inherited from the mid-nineteenth may United States gross national product. All this
no longer be altogether adequate. It will be does not mean that the absolute volume of
recalled that Professor Hicks' analysis of the United States imports has failed to expand.
long-run dollar problem was based on what It increased by 44 per cent from 1929 to
he described as "a change in economic at- 195515 But notice two things. This increase is
mosphere between the nineteenth and twen- much less than proportional to the growth of
tieth centuries."12 His analysis in regard to United States output. Moreover, it is much
the dollar problem was open to criticism, yet less than the rate of growth of British imports
I believe that in emphasizing the varying in- in the nineteenth century, which during any
cidence of productivity changes on interna- comparable period showed a two to threefold
tional trade he made an important point, a increase in volume.
point that had been noted some years earlier It is useful to keep in mind these elemen-
by Professor Haberler.13 While Britain's tary facts about American imports because
ratio of imports to national income showed a the United States is now the dominant econ-
rising tendency during most of the nineteenth omy not only in world production but also in
century, the United States import ratio has world trade. Some economists are more in-
been practically halved in the last five dec- clined to stress the future prospect of expan-
ades.14 This has happened in spite of the fact sion in United States imports, but that is a
that in short period comparisons the United debatable matter. It is never quite safe, and
States typically shows a rather high income for present purposes really unnecessary, to
elasticity of imports. There seems to have engage in predictions. The facts for the re-
been a long-run downward shift in the United cent past are sufficient to indicate a change
States import function, resulting from in the economic atmosphere of international
changes in economic structure. It is not cer- trade between the nineteenth and twentieth
tain that tariff policy provides the major part centuries.
of the explanation. It seems very likely that It will be remembered that in Hicks' anal-
the incidence of technological advance has ysis of the dollar shortage, the balance of
had a good deal to do with it. payments problem resolves itself into a terms
of trade problem. This seems a plausible sim-
nA. K. Cairncross and J. Faaland, "Long-Term
Trends in Europe's Trade," Economic Journal,
March 1952, pp. 26-7.
15The quantum of crude material imports, as al-
12J. R. Hicks, "An Inaugural Lecture," Oxford ready stated, increased by only 23 per cent. The other
Economic Papers, June 1953, p. 130.
commodity groups showed the following percentage
13G. Haberler, "Dollar Shortage?" in S. E. Harris increases from 1929 to 1955: crude foodstuffs, 33 per
(ed.), Foreign Economic Policy for the United States,
cent; manufactured foodstuffs, 55 per cent; semi-
Cambridge, Massachusetts, 1948, pp. 438-9. manufactures, 76per cent; finished manufactures, 52
14United States exports as a percentage of gross na- per cent. Is it not possible, however, that the relatively
tional product fell from 5.7 per cent in the period small rise in imports of crude commodities may be
1896-1914 to 2.97 per cent in 1955. See W. Lederer, due, not to a low rate of growth of United States de-
"Major Developments Affecting the United States mand, but rather to a deficiency on the supply side?
Balance of International Payments," Review of Eco- The answer is in Professor Schultz's paper, where the
nomics and Statistics, May 1956, p. 184. strategic role of demand is clearly demonstrated.
 497 INTERNATIONAL TRADE AND INTERNATIONAL INEQUALITY

plification. Any country in foreign exchange fully, if at all, in changes in the terms of
difficulties can normally restore its balance of trade.
payments by accepting a worsening in its In considering the international mecha-
terms of trade. In Hicks' model external bal- nism of development it is necessary at any
ance is maintained by changes in terms of rate to admit the possibility of variation in
trade. the conditions of growth transmission
But can we not go a step further? There through trade. Just as the limited extent to
has been a tendency, in Britain and else- which the United States economy transmits
where, to exaggerate both the actual extent its own growth rate to primary producing
and the economic significance of changes in countries is fully understandable in the light
the terms of trade. We are sometimes apt to of its own abundant natural resources com-
think of these changes as if the resources of bined with its ample capital supplies and
each country were for ever committed to the technical know-how, so the nineteenth cen-
existing export industries. This view may be tury experience was conditioned by the fact
all right for the short run, but in the longer that the industrial revolution happened to
run labour and capital within each country originate on a small island with a limited
can usually move to other occupations, and range of resources, at a time when the chem-
do in fact move. If the relationship of export ical industry was yet unborn.
prices to import prices undergoes a marked As a result, the rate of growth in the im-
increase or decline, it is entirely natural that port demand of the dominant economy of the
factors of production should tend to move twentieth century seems different from that
from export industries to import-competing of the nineteenth. If this is so, it is not certain
industries or vice versa. This may involve that the less developed countries can rely on
simply changes in the allocation of increases economic growth being induced from the out-
in factor supplies rather than movements of side through an expansion of world demand
existing factors. In any event, the point is for their exports of crude materials.16 In
that a change in the terms of trade tends to these circumstances reliance on induced ex-
induce shifts in production and in the distri- pansion through international trade may not
bution ofresources, which will tend to reverse be able to provide the main solution to the
or counteract the change in the terms of problem of development. It is not surprising,
trade. What remains is growth and change in therefore, that countries should be looking
the volume of productive activity induced for other types of solution. It will be useful to
through international trade. On this view, keep these things in mind, because they form
changes in the terms as well as in the balance the background to the case for balanced
of trade are a transient and relatively insig- growth which is now so greatly in vogue.
nificant element in the mechanism by which
processes of economic growth (or decline)
may be transmitted from one country to
others. 16To ask the less developed countries to increase
their export quantities of primary products in the face
This does not imply that shifts in external of a price-inelastic and not an upward-shifting de-
demand do not matter. Fortunate indeed is mand schedule would be to ask, in effect, for an in-
come transfer from poor to rich countries though a
the country with an expanding export market
change in the terms of trade in favour of the latter If
for the commodity in whose production it has one of several countries exporting the same primary
a comparative advantage; for it can then commodity were to cut its export costs and prices, its
draw increasing supplies in limitless variety export proceeds could indeed increase, but only at the
from the outside world. The suggestion is expense of a fall in the other countries' export pro-
ceeds. The balance of payments adjustment process
merely that, because of the possibility of in- alone (whether through exchange rate variations or
ternal factor shifts in response to varying domestic price changes) would lead the latter to cut
price relationships, long-term trends in exter- their export prices too, and all will be worse off at the
nal demand conditions need not be reflected end than they were at the start.
 498 TRADE STRATEGY

VIII. A. 3. Trade as a Mechanism

of International Inequality*

Our inherited economic theory would . . . of stating the law of comparative costs in
lead us to expect that international inequali- terms of only a single factor, labour — which,
ties should not be so large as they are and not however, could have different "qualities" or
be growing. In any case this theory does not degrees of "effectiveness" — turned the em-
furnish us with an explanation in causal phasis inother directions.
terms of these inequalities and their tendency After Eli F. Heckscher's paper on the
to increase. equalising influence of trade on factor prices
"The fact that many under-developed and Bertil Ohlin's restatement of the classi-
countries do not derive the advantages from cal theory of international trade in terms of
modern transportation and commerce that a general equilibrium theory of the Lausanne
theory seems to demand is one of the most school type,2 trade appeared more clearly as
pertinent facts in the present international a substitute, or an alternative, to factor
situation and cannot be easily dismissed" — I movements in permitting an adjustment of
am quoting from a recent paper by a Swedish industrial activity to adapt itself to the local-
economist, Mr. Folke Hilgert.1. . . isation of natural and population resources
Hilgert points out that huge movements of with the result that the relative scarcity of la-
labour and capital from Europe have trans- bour and capital became less different. Upon
formed the plains in the temperate belts into this foundation there has in recent years been
"white man's land" with high, rapid and sus- a lively discussion between the econometri-
tained economic development and rising lev- cians elaborating, under specific, abstract
els of living. "Yet the gradual filling of the and static, conditions, the relative effective-
'empty spaces' has not reduced the pressure ness of this tendency to equalisation of factor
of population in, for instance, Asia's over- prices as a result of international trade.3
populated regions where labour is most The inadequacy of such theories for ex-
abundant." plaining reality cannot be accounted for by
Let us remember, however, that according pointing to the relative breakdown of the
to the classical doctrine movements of labour multilateral trading system as it functioned
and capital between countries would not be prior to the First World War, a change which
necessary for bringing about a development is related as both effect and cause to the in-
towards equilisation of factor prices and, con- crease of national trade and payments re-
sequently, earnings and incomes; in fact, the strictions. For, as Hilgert observes, a similar
theory of international trade was largely de- confrontation of the facts of international in-
veloped on the abstract assumption of inter- equality with the theory of international
national immobility of all factors of produc- 2Eli F. Heckscher, "The Effect of Foreign Trade on
tion. That trade itself initiated a tendency the Distribution of Income," Readings in the Theory
towards a gradual equalisation of factor of International Trade, selected by a committee of
the American Economic Association, Allen & Unwin,
prices was implicit already in the expositions
London, 1950 (translation from the Swedish original
by the classical authors, though their method 1919); Bertil Ohlin, Interregional and International
Trade, Harvard University Press, Cambridge, Mass.,
♦From Gunnar Myrdal, Development and Under- 1933.
development, National Bank of Egypt Fiftieth Anni- 3The recent discussion of the problem of factor
versary Commemoration Lectures, Cairo, 1956, pp. price equalisation as a result of international trade
9-10, 47-51. Reprinted by permission. was initiated by Professor Paul A. Samuelson in two
'"Uses and Limitations of International Trade in articles in the Economic Journal, 1948 and 1949; for
Overcoming Inequalities in World Distribution of fuller reference see Svend Laursen, "Production
Population and Resources," World Population Con- Functions and the Theory of International Trade,"
ference, Rome, 1954. American Economic Review, 1955, pp. 540 ff.
 499 INTERNATIONAL TRADE AND INTERNATIONAL INEQUALITY

trade for the period before 1914 reveals the mostly also less progressive economically. In
same discord. And I would add that it is not a rather close correlation to their poverty
self-evident but, indeed, very much up to they are ridden by internal economic inequal-
doubt whether today a freer trade would nec- ities, which also tend to weaken the effective-
essarily lead to less of international inequal- ness of their democratic systems of govern-
ity or whether in general trade between ment in the cases where they are not under
developed and (densely populated) under- one form or another of oligarchic or forth-
developed countries has ever had that ef- right dictatorial rule.
fect. . . The relations between relative lack of na-
Contrary to what the equilibrium theory of tional economic integration and relative eco-
international trade would seem to suggest, nomic backwardness run, according to my
the play of the market forces does not work hypothesis of circular cumulative causation,
towards equality in the renumerations to fac- both ways. With a low level of economic de-
tors of production and, consequently, in in- velopment follow low levels of social mobility,
comes. If left to take its own course, eco- communications, popular education and na-
nomic development is a process of circular tional sharing in beliefs and valuations,
and cumulative causation which tends to which imply greater impediments to the
award its favours to those who are already spread effects of expansionary momentum; at
well endowed and even to thwart the efforts the same time the poorer states have for
of those who happen to live in regions that much the same reasons and because of the
are lagging behind. The backsetting effects of very fact of existing internal inequalities
economic expansion in other regions domi- often been less democratic and, in any case,
nate the more powerfully, the poorer a coun- they have, because they are poorer, been up
try is. against narrower financial and, at bottom,
Within the national boundaries of the psychological limitations on policies seeking
richer countries an integration process has to equalise opportunities. Inequality of op-
taken place: on a higher level of economic de- portunities has, on the other hand, contrib-
velopment expansionary momentum tends to uted to preserving a low "quality" of their
spread more effectively to other localities and factors of production and a low "effective-
regions than those where starts happen to ness" in their production efforts, to use the
have been made and successfully sustained; classical terms, and this has hampered their
and inequality has there also been mitigated economic development.
through interferences in the play of the mar- On the international as on the national
ket forces by organised society. In a few level trade does not by itself necessarily work
highly advanced countries — comprising only for equality. A widening of markets strength-
about one-sixth of the population in the non- ens often on the first hand the progressive
Soviet world — this national integration pro- countries whose manufacturing industries
cess is now being carried forward towards a have the lead and are already fortified in sur-
very high level of equality of opportunity to roundings ofexternal economies, while the
all, wherever, and in whatever circumstances under-developed countries are in continuous
they happen to be born. These countries are danger of seeing even what they have of in-
approaching a national harmony of interest dustry and, in particular, their small scale in-
which, because of the role played by state dustry and handicrafts outcompeted by
policies, has to be characterized as a "created cheap imports from the industrial countries,
harmony"; and this has increasingly sus- if they do not protect them.
tained also their further economic devel- It is easy to observe how in most underde-
opment. veloped countries the trading contacts with
Outside this small group of highly devel- the outside world have actually impoverished
oped and progressive countries, all other them culturally. Skills in many crafts inher-
countries are in various degrees poorer and ited from centuries back have been lost. A
 500 TRADE STRATEGY

city like Baghdad, with whose name such glo- self does not lead to such a development; it
rious associations are connected, today does rather tends to have backsetting effects and
not harbour any of the old crafts, except to strengthen the forces maintaining stagna-
some silver smithies, and they have adopted tion or regression. Economic development
patterns from abroad requiring less crafts- has to be brought about by policy interfer-
manship; similarly it is only with the greatest ences which, however, are not under our pur-
difficulties that one can buy a book of Arabic view at this stage of the argument when we
literature, while cheap magazines in English are analysing only the effects of the play of
or Arabic are in abundance. the market forces.
If international trade did not stimulate Neither can the capital movements be re-
manufacturing industry in the under-devel- lied upon to counteract international inequal-
oped countries but instead robbed them of ities between the countries which are here in
what they had of old-established crafts, it did question. Under the circumstances described,
promote the production of primary products, capital will, on the whole, shun the under-de-
and such production, employing mostly un- veloped countries, particularly as the ad-
skilled labour, came to constitute the basis vanced countries themselves are rapidly de-
for the bulk of their exports. In these lines, veloping further and can offer their owners of
however, they often meet inelastic demands capital both good profits and security.
in the export market, often also a demand There has, in fact, never been much of a
trend which is not rising very rapidly, and ex- capital movement to the countries which
cessive price fluctuations. When, further- today we call under-developed, even in ear-
more, population is rapidly rising while the lier times — except tiny streams to the eco-
larger part of it lives at, or near, the subsis- nomic enclaves, mainly devoted to export
tence level — which means that there is no production of primary products which, how-
scarcity of common labour — any technologi- ever, usually were so profitable to their own-
cal improvement in their export production ers that they rapidly became self-supporting
tends to confer the advantages from the so far as investment capital was concerned
cheapening of production to the importing and, in addition, the considerably larger but
countries. Because of inelastic demands the still relatively small investments in railways
result will often not even be a very great en- and other public utilities which had their se-
largement ofthe markets and of production curity in the political controls held by colo-
and employment. In any case the wages and nial governments. The bulk of European
the export returns per unit of product will overseas capital exports went to the settle-
tend to remain low as the supply of unskilled ments in the free spaces in the temperate
labour is almost unlimited. zones which were becoming populated by
The advice — and assistance — which the emigration from Europe. After the collapse
poor countries receive from the rich is even of the international capital market in the
nowadays often directed towards increasing early 'thirties, which has not been remedied,
their production of primary goods for export. and later the breakdown of the colonial sys-
The advice is certainly given in good faith tem, which had given security to the foreign
and it may even be rational from the short investor, it would be almost against nature if
term point of view of each under-developed capital in large quantities were voluntarily to
country seen in isolation. Under a broader seek its way to under-developed countries in
perspective and from a long term point of order to play a role in their economic
view, what would be rational is above all to development.
increase productivity, incomes and living True, capital in these countries is scarce.
standards in the larger agricultural subsis- But the need for it does not represent an ef-
tence sectors, so as to raise the supply price fective demand in the capital market.
of labour, and in manufacturing industry. Rather, if there were no exchange controls
This would engender economic development and if, at the same time, there were no ele-
and raise incomes per capita. But trade by it- ments in their national development policies
 501 INTERNATIONAL TRADE AND INTERNATIONAL INEQUALITY

securing high profits for capital — i.e. if the setting effects in the under-developed world,
forces in the capital market were given un- and their mode of operation would be very
hampered play— capitalists in under-devel- much the same as it is in the circular cumu-
oped countries would be exporting their cap- lation of causes in the development process
ital. Even with such controls and policies in within a single country. . . . Internationally,
existence, there is actually a steady capital these effects will, however, dominate the out-
flight going on from under-developed coun- come much more, as the countervailing
tries, which in a realistic analysis should be spread effects of expansionary momentum
counted against what there is of capital in- are so very much weaker. Differences in leg-
flow to these countries. islation, administration and mores generally,
Labour migration, finally, can safely be in language, in basic valuations and beliefs,
counted out as a factor of importance for in- in levels of living, production capacities and
ternational economic adjustment as between facilities, etcetera make the national bound-
under-developed and developed countries. aries effective barriers to the spread to a de-
The population pressure in most under-devel- gree which no demarcation lines within one
oped countries implies, of course, that they country approach.
do not need immigration and the consequent Even more important as impediments to
low wages that immigrants are not tempted the spread effects of expansionary momen-
to come. Emigration from these countries tum from abroad than the boundaries and
would instead be the natural thing. For var- everything they stand for is, however, the
ious reasons emigration could, however, not very fact of great poverty and weak spread
be much of a real aid to economic develop- effects within the under-developed countries
ment, even if it were possible. themselves. Where, for instance, interna-
And the whole world is since the First
tional trade and shipping actually do trans-
World War gradually settling down to a sit- form the immediate surroundings of a port to
uation where immigrants are not welcomed a centre of economic expansion, which hap-
almost anywhere from wherever they come; pens almost everywhere in the world, the ex-
people have pretty well to stay in the country pansionary momentum usually does not
where they are born, except for touristing by spread out to other regions of the country,
those who can afford it. And so far as the which tend to remain backward if the forces
larger part of the under-developed world is in the markets are left free to take their
concerned, where people are "coloured" ac- course. Basically, the weak spread effects as
cording to the definition in the advanced between countries are thus for the larger
countries, emigration is usually stopped al- part only a reflection of the weak spread ef-
together bythe colour bar as defined by the fects within the under-developed countries
legislation, or in the administration, of the themselves.
countries which are white-dominated and at Under these circumstances the forces in
the same time better off economically. the markets will in a cumulative way tend to
If left unregulated, international trade and cause ever greater international inequalities
capital movements would thus often be the between countries as to their level of eco-
media through which the economic progress nomic development and average national in-
in the advanced countries would have back- come per capita.
Comment
Some assessments of colonialism and neocolonialism are also relevant in considering the
argument that international trade has operated as a mechanism of international inequality.
Many writers have traced forces of dependency through history, and the legacy of imperialism
is often referred to in explaining underdevelopment.
Studies that emphasize features of imperialism or "exploitation" are: R. Rhodes (ed.), Im-
perialism and Underdevelopment (1970); H. Magdoff, The Age of Imperialism (1969); K.
Boulding and T. Mukerjee (eds.), Economic Imperialism (1972); P. A. Baran, The Political
 502 TRADE STRATEGY

Economy of Growth (1962), chapters 6-8; K. T. Fann and D. C. Hodges (eds.), Readings in
U.S. Imperialism (1971); Walter Rodney, How Europe Underdeveloped Africa (1972); D. S.
Landes, "Some Thoughts on the Nature of Economic Imperialism," Journal of Economic
History (December 1961); A. Emmanuel, Unequal Exchange (1972); S. Hymer and S. Res-
nick, "International Trade and Uneven Development," in J. Bhagwati (ed.), Trade, Balance
of Payments, and Growth (1971); B. J. Cohen, The Question of Imperialism (1973); R. J.
Owen and R. B. Sutcliffe (eds.), Studies in the Theory of Imperialism (1972); D. K. Field-
house (ed.), The Theory of Capitalist Imperialism (1967); Johan Galtung, "A Structural
Theory of Imperialism," Journal of Peace Research (1971); B. Warren, Imperialism: Pioneer
of Capitalism (1980). An extensive bibliographic essay is presented by Gavin Williams, "Im-
perialism and Development: A Critique," World Development (July-August 1978).
The literature on dependency, as* listed in Chapter II above, is also relevant.
Comment
During the 1950s and 1960s, the United Nation's Economic Commission for Latin America
(ECLA) was influential in formulating ideas on development that contradicted orthodox, neo-
classical thought. In doing so, ECLA was in many respects a precursor of the dependencia
school and also the North-South dialogue.
The main ECLA document is entitled The Economic Development of Latin America and
Its Principal Problems (1950). ECLA publications offer a critique of the neoclassical theory
of international trade and emphasize the tendency for international trade to reproduce the
inequality among nations through relations between the center and periphery. Special em-
phasis isgiven to the consequences of differential technical progress and the worsening of the
periphery's terms of trade. In development programming, ECLA's early objectives were in-
dustrialization ofthe periphery, "healthy protectionism," programming of import substitu-
tion, foreign currency allocation policies, and the avoidance of a reduction in real wages.
See Fernando Henrique Cardoso, "The Originality of the Copy: The Economic Commis-
sion for Latin America and the Idea of Development," in Toward a New Strategy for Devel-
opment, ARothko Chapel Colloquium (1979), chapter 2.

Comment
The terms of trade issue remains very much alive — despite the many critical questions
raised about the empirical evidence, its welfare significance, and policy implications. The
Singer-Prebisch thesis on the alleged tendency for the long-term movement of terms of trade
to be against developing countries is still prominent in discussions of the distribution of gains
from trade, and its implications for "distributive justice" carry over to demands for a New
International Economic Order. For an account of the emergence of this thesis, see Joseph
Love, "Raul Prebisch and the Origins of the Doctrine of Unequal Exchange," Latin American
Research Review (November 1980).
The empirical evidence has recently been reviewed more closely by John Spraos, "The Sta-
tistical Debate on the Net Barter Terms of Trade between Primary Commodities and Man-
ufactures," Economic Journal (March 1980); Michael Michaely, "The Terms of Trade be-
tween Poor and Rich Nations," Institute for International Economic Studies, University of
Stockholm, Seminar Paper No. 162 (November 1980); Irving B. Kravis and Robert E. Lipsey,
"Prices and Terms of Trade for Developed-Country Exports of Manufactured Goods," Na-
tional Bureau of Economic Research Working Paper No. 774 (September 1981).
The theory of "unequal exchange," as presented recently in A. Emmanuel's Unequal Ex-
change: AStudy of the Imperialism of Trade (1972), stems from Marxist theory and contends
that there is a transfer of reinvestible surplus (surplus value) from the low-wage developing
country to the high-wage industrial country via the terms of trade. Trade is believed to be
unequal to the "South" because their terms of trade are lower than they would be under a
 503 INTERNATIONAL TRADE AND INTERNATIONAL INEQUALITY

Pareto efficient trade arrangement, which would allow perfect international labor mobility.
This argument relates to the absolute level of the terms of trade rather than to secular dete-
rioration or cyclical fluctuations. For a discussion of Emmanuel's analysis, see Edmar L.
Bacha, "An Interpretation of Unequal Exchange from Prebisch-Singer to Emmanuel," Jour-
nal ofDevelopment Economics (December 1978); David Evans, "Unequal Exchange and Eco-
nomic Policies," Economic and Political Weekly (February 1976). Another Marxist analysis
of unequal exchange is offered by Samir Amin in three works: Accumulation on a World Scale
(1974), Unequal Development (1976), and Imperialism and Unequal Development (1977).
The influence of a developing country's agricultural productivity on its terms of trade
should be emphasized. Sir Arthur Lewis has argued that so long as the bulk of tropical peoples
are food farmers with relatively low productivity, their tropical export products are available
to the rest of the world on an essentially low-wage basis. Since farmers can grow cash crops
for export or foods, the prices of these two must tend to move together. The newer trade in
light manufactures from the tropics is essentially of the same kind; it is an additional oppor-
tunity to sell low-wage labor. The low factoral terms of trade derive from the low productivity
of the bulk of tropical producers — the food farmers. If the terms of trade are to improve, it
is necessary to raise the productivity of food farmers who constitute the source of wage labor
into the export sector. See selection VI. C. 3 and W. A. Lewis, The Evolution of the Interna-
tional Economic Order (1978).
Lewis also concludes that

It is easy to exaggerate the potential contribution of foreign trade to development. To take an extreme
example, India's exports are only 5 per cent of her national income. If as a result of a new international
economic order India were paid five times as much for exports (without an increase in prices of imports)
her national income would be raised only by 20 per cent in the first instance, say from $100 to $120 per
head. The basic cause of India's poverty is not her terms of trade but the fact that an Indian farmer
produces only one-eleventh as much food as an American farmer. Other countries depend on foreign
trade to a greater extent than India, and would benefit more directly from better terms of trade. One
must also take into account the indirect effects. But the main points remain: the poverty of the tropical
countries is due mainly to their low productivity and only secondarily to their terms of trade. Their
productivity is not low because their terms of trade are poor; their terms of trade are poor because their
agricultural productivity is low {Growth and Fluctuations 1870-1913, 1978, p. 244).

VIII.A.4. Conditions of Export-Led

Development — Note

The foregoing materials raise a central ques- countries have not acted as a key propulsive
tion: Under what conditions can a process of sector, propelling the rest of the economy for-
export-induced development follow upon an ward. Although the classical belief that de-
expansion of the export sector? How can ex- velopment can be transmitted through trade
port expansion also act as an engine of do- has been confirmed by the experience of some
mestic development? Notwithstanding the countries that are now among the richest in
possibility of lagging exports in more recent the world, trade has not had a similar stim-
decades, most of the underdeveloped coun- ulating effect for countries that have re-
tries have experienced long periods of export mained underdeveloped. Why has not the
growth. In most cases, after a country was growth in exports in these countries car-
exposed to the world economy, its exports ried over to other sectors and led to more
grew markedly in volume and in variety. Yet, widespread development in the domestic
despite their secular rise, exports in many economy?
 504 TRADE STRATEGY

As noted above, some critics of the classi- change economy.2 This positive effect on the
cal position contend that the very forces of side of factor supply may have more than off-
international trade have been responsible for set any negative effect on saving.
inhibiting development. They argue that the More serious is the argument that inter-
development of the export sector by foreign national market forces have transferred in-
capital has created a "dual economy" in come from the poor to rich nations through a
which production has been export-biased, deterioration in the terms of trade of the less
and the resultant pattern of resource utiliza- developed countries. The significance of this
tion has deterred development. This argu- argument is also overdrawn, and it can be
ment, however, tends to contrast the pattern questioned on both theoretical and empirical
of resource utliization that actually occurred grounds. The alleged trend is not based on
with some other ideal pattern. More relevant the measurement of prices within the poor
is a comparison between the actual pattern countries, but rather on inferences from the
and the allocation that would have occurred
United Kingdom's commodity terms of trade
in the absence of the capital inflow. There is or the terms of trade between primary prod-
little foundation to the assertion that if there
ucts and manufactured products.3 This does
had been no foreign investment, a poor coun- not provide a sufficiently strong statistical
try would have generated more domestic in- foundation for any adequate generalization
vestment; or that, in the absence of foreign about the terms of trade of poor countries.4
entrepreneurs, the supply of domestic entre- The import-price index conceals the hetero-
preneurs would have been larger. Contrary to geneous price movements within and among
what is often implied by the critics of foreign the broad categories of foodstuffs, raw ma-
investment, the real choice was not between terials, and minerals; no allowance is made
employing the resources in the export sector for changes in the quality of exports and im-
or in domestic production, but rather be- ports; there is inadequate consideration of
tween giving employment to the surplus re- new commodities; and the recorded terms of
sources inexport production or leaving them
2For an instructive analysis of the general process
idle.1 It is difficult to substantiate the argu- by which the money economy has developed through
ment that foreign investment was competi- expansion of export production induced by the growth
tive with home investment, or that the utili- of new wants for imported consumers' goods, see Hla
zation of resources in the export sector was Myint, The Economics of Developing Countries,
at the expense of home production. London, 1964, Chapters 1-5.
Another contention is that trade has 3United Nations, Department of Economic Affairs,
Relative Prices of Exports and Imports of Under-De-
impeded development by the "demonstration veloped Countries, New York, 1949, pp. 7, 13-24; W.
effect": the international demonstration of A. Lewis, "World Production, Prices, and Trade,
higher consumption standards in more devel- 1870-1960," Manchester School, May 1952, p. 118.
oped countries has allegedly raised the pro- 4For detailed criticisms, see R. E. Baldwin, "Secu-
pensity to consume in the less developed lar Movements in the Terms of Trade," American
countries and reduced attainable saving Economic Review, Papers and Proceedings, May
rates. By stimulating the desire to consume, 1955, pp. 267ff.; P. T. Ellsworth, "The Terms of
Trade Between Primary Producing and Industrial
however, the international demonstration ef- Countries," Inter-American Economic Affairs, Sum-
fect may also have operated on incentives and mer 1956, pp. 47-65; T. Morgan, "The Long-Run
been instrumental in increasing the supply of Terms of Trade Between Agriculture and Manufac-
effort and productive services — especially as turing," Economic Development and Cultural
between the subsistence sector and the ex- Change, October 1959, pp. 6-17; Gottfried Haberler,
"Terms of Trade and Economic Development," in H.
S. Ellis (ed.), Economic Development for Latin
'Cf. Hla Myint, "The Gains from International America, New York, 1961, pp. 275-97; Jagdish
Trade and the Backward Countries," Review of Eco- Bhagwati, "A Skeptical Note on the Adverse Secular
nomic Studies, Vol. 22, No. 58, 1954-55; "The 'Clas- Trend in the Terms of Trade of Underdeveloped
sical Theory' of International Trade and the Under- Countries," Pakistan Economic Journal, December
developed Countries," Economic Journal, June 1958, 1960; G. M. Meier, International Economics of De-
pp. 317-37. velopment, New York, 1968, Chapter 3.
 505 INTERNATIONAL TRADE AND INTERNATIONAL INEQUALITY

trade are not corrected for the substantial de- trade operated as a mechanism of interna-
cline in transportation costs. The introduc- tional inequality, we must look to other fac-
tion of new products and qualitative improve- tors for an understanding of why trade has
ments have been greater in manufactured not had a more stimulating effect in under-
than in primary products, and a large pro- developed countries. If the export sector is to
portion ofthe fall in British prices of primary be a propelling force in development, it is es-
products can be attributed to the great de- sential that the export sector not remain an
cline in inward freight rates. The simple use enclave, separate from the rest of the econ-
of the "inverse" of the United Kingdom's omy; instead, an integrated process should be
terms of trade to indicate the terms of trade established, diffusing stimuli from the export
of primary producing countries involves sector and creating responses elsewhere in
therefore a systematic bias which makes the economy. A more convincing explanation
changes appear more unfavorable to the pri- of why export-led development has occurred
mary exporting countries than they actually in some countries, but not in others, would
were. therefore distinguish the differential effects
Even if it were true that the less developed of the integrative process by focusing on the
countries experienced a secular deterioration varying strength of the stimuli in different
in their commodity terms of trade, the ques- countries from their exports and on the dif-
tion would still remain whether this consti- ferent response mechanisms within the ex-
tuted a significant obstacle to their develop-
ment. The answer depends on what caused porting countries.5
Different export commodities will provide
the deterioration and whether the country's different stimuli, according to the technolog-
factoral terms of trade and income terms also ical characteristics of their production. The
deteriorated. If the deterioration in the com- naturetion hasof antheinfluence
export good's
modity terms is due to increased productivity on theproduction func-
extent of other
in the export sector, the single-factoral terms secondary changes elsewhere in the economy,
of trade (commodity terms corrected for beyond the primary increase in export out-
changes in productivity in producing exports) put.6 With the use of different input coeffi-
can improve at the same time. As long as pro-
ductivity inits export industries is increasing 5The following paragraphs draw upon G. M. Meier,
"External Trade and Internal Development," in Peter
more rapidly than export prices are falling, Duignan and L. H. Gann (eds.), Colonialism in Af-
the country's real income can rise despite the rica 1870-1960, Vol. 4, The Economics of Colonial-
deterioration in the commodity terms of ism, New York, 1975, Chapter 11.
trade: when its factoral terms improve, the 6Although more empirical research is needed, some
country benefits from the ability to obtain a illustrative cases are suggested by D. C. North, "Lo-
greater quantity of imports per unit of factors cation Theory and Regional Economic Growth,"
Journal of Political Economy, June 1955, pp. 249-
embodied in its exports. Also possible is an 51; Dudley Seers, "An Approach to the Short-Period
improvement in the country's income terms Analysis of Primary-Producing Economies," Oxford
Economic Papers, February 1959, pp. 6-9; R. E.
of trade (commodity terms multiplied by Caves and R. H. Holton, The Canadian Economy,
quantity of exports) at the same time as its Cambridge, 1959, pp. 41-7; J. V. Levin, The Export
commodity terms deteriorate. The country's Economies, Cambridge, 1960; R. E. Baldwin, "Ex-
capacity to import is then greater, and this port Technology and Development from a Subsistence
will ease development efforts. When due Level," Economic Journal, March 1963, pp. 80-92;
weight is given to the increase in productivity M. H. Watkins, "A Staple Theory of Economic
Growth," Canadian Journal of Economics and Pol it1
in export production and the rise in export ical Science, May 1963, pp. 141-58; V. D. Wickizer,
volume, it would appear that the single-fac- "The Plantation System in the Development of Trop-
toral terms and income terms of trade ac- ical Economics," Journal of Farm Economics, Feb-
tually improved for many poor countries, not- ruary 1958, pp. 63-77; Dudley Seers, "The Mecha-
withstanding any possible deterioration in nism of an Open Petroleum Economy," Social and
Economic Studies, June 1964, pp. 233-42; M. J.
their commodity terms of trade. Herskovitz and M. Horwitz, Economic Transition in
Having rejected the view that international Africa, Evanston, 1964, pp. 312-18.
 506 TRADE STRATEGY

cients to produce different types of export lenge and instills abilities usable in other sec-
commodities, there will be different rates of tors, but is not so high as to require the im-
learning and different linkage effects. The porting of a transient class of skilled
degree to which the various exports are pro- managerial labor.
cessed ishighly significant in the determina- Although the processing of a primary
tion of external economies associated with product provides forward linkages in the
the learning process; the processing of pri- sense that the output of one sector becomes
mary-product exports by modern methods is an input for another sector, it is also impor-
likely to benefit other activities through the tant to have backward linkages. When some
spread of technical knowledge, training of exports grow, they provide a strong stimulus
labor, demonstration of new production tech- for expansion in the input-supplying indus-
niques that might be adapted elsewhere in tries elsewhere in the economy. These back-
the economy, and the acquisition of organi- ward linkages may be in agriculture or in
zational and supervisory skills. other industries supplying inputs to the ex-
In contrast, growth of the export sector panding export sector, or in social overhead
will have a negligible carry-over if its tech- capital. The importance of linkages has been
niques ofproduction are the same as those al- stressed by Hirschman.8
ready in use in other sectors, or if its expan- The notion is emphasized also by Perroux,
sion occurs by a simple widening of who refers to a developing enterprise as a
production without any change in production "motor unit" when it increases its demands
functions. If the introduction or expansion of on its suppliers for raw materials or commu-
export crops involves simple methods of pro- nicates new techniques to another enterprise.
duction that do not differ markedly from the The "induction effect" that the motor unit
traditional techniques already used in subsis- exerts upon another unit may be considered
tence agriculture, the stimulus to develop- in two components that frequently occur in
ment will clearly be less than if the growth in combination: (1) a dimension effect that is
exports entailed the introduction of new skills the augmentation of demand by one enter-
and more productive recombinations of fac- prise to another by increasing its supply; and
tors of production. More favorable linkages (2) an innovation effect that introduces an in-
may stem from exports that require skilled novation which for a given quantity of factors
labor than from those using unskilled labor. of production yields the same quantity of pro-
The influence of skill requirements may op- duction at a lower price and/or a better qual-
erate in various ways: greater incentives for ity. When a motor unit is interlinked with its
capital formation may be provided through surrounding environment, Perroux refers to a
education; on-the-job training in the export growth pole or a development pole.9 The em-
sector may be disseminated at little real cost phasis on generating new skills, innovations
through the movement of workers into other in the export sector or other sectors linked to
sectors or occupations; skilled workers may exports, and technical change are important
be a source of entrepreneurship; skilled work- in determining the learning rate of the
ers may save more of their wage incomes economy.
than unskilled workers.7 The level of entre- Beyond this, the nature of the production
preneurial skill induced by the development function of the export commodity will also
of an export is also highly significant. The
level will be expanded if the development of 8Albert O. Hirschman, The Strategy of Economic
the export commodity offers significant chal- Development, New Haven, 1958, Chap. 9.
'Francois Perroux, "Multinational Investment and
the Analysis of Development and Integration Poles,"
7Richard E. Caves, "Export-led growth and the in Multinational Investment in the Economic Devel-
new economic history," in Trade, Balance of Pay- opment and Integration of Latin America, Bogota,
ments and Growth, J. N. Bhagwati et al. (eds.), Am- Inter-American Development Bank, April 1968, pp.
99-103.
sterdam, 1971, pp. 403-42.
 507 INTERNATIONAL TRADE AND INTERNATIONAL INEQUALITY

determine the distribution of income, and, in to import according to the degree of ampli-
turn, the pattern of local demand and impact tude of fluctuation in foreign exchange re-
on local employment. The use of different ceipts. To the extent that different exports
factor combinations affects the distribution vary in their degree of fluctuation, and in rev-
of income in the sense that the relative shares enue earned and retained at home, their re-
of profits, wages, interest, and rent will vary percussions on the domestic economy will
according to the labor intensity or capital in- also differ. Depending on the various char-
tensity ofthe export production and the na- acteristics ofthe country's export, we may
ture of its organization — whether it is min- thus infer how the strength of the integrative
ing, plantation agriculture, or peasant process, in terms of the stimulus from ex-
farming. If the internal distribution of the ex- ports, will differ among countries.
port income favors groups with a higher pro- In summary, we would normally expect
pensity to consume domestic goods than to the stimulating forces of the integrative pro-
import, the resultant distribution of income cess to be stronger under the following con-
will be more effective in raising the demand ditions: the higher the growth-rate of the ex-
for home-produced products; and to the ex- port sector, the greater the direct impact of
tent that these home-produced products are the export sector on employment and per-
labor-intensive, there will be more of an im- sonal income, the more the expansion of ex-
pact on employment. In contrast, if income is ports has a "learning effect" in terms of in-
distributed to those who have a higher pro- creasing productivity and instilling new
pensity toimport, the leakage through con- skills, the more the export sector is supplied
sumption ofimported goods will be greater. through domestic inputs instead of imports,
If income increments go to those who are the more the distribution of export income
likely to save large portions, the export sector favors those with a marginal propensity to
may also make a greater contribution to the consume domestic goods instead of imports,
financing of growth in other sectors. the more productive is the investment result-
If the export commodity is subject to sub- ing from any saving of export income, the
stantial economies of scale in its production, more extensive are the externalities and link-
this will tend to imply large capital require- ages connected with the export sector, and
ments for the establishment of enterprises, the more stable are the export receipts that
and hence extra-regional or foreign borrow- are retained at home. Some exports fulfill
ing. This may then lead to an outward flow these conditions more readily than others,
of profits instead of providing profit income and countries specializing on these exports
for local reinvestment. But this is only part of will enjoy greater opportunities for
the impact of the foreign investment. For a development.
full appraisal, it would be necessary to con- Even with a strong stimulus from exports,
sider all the benefits and costs of the foreign however, the transmission of growth from the
investment. And these too will vary accord- export-base to the rest of the economy will
ing to the nature of the export sector in which still be contingent upon other conditions in
the foreign investment occurs. the economy. The weak penetrative power of
Finally, the repercussions from exports will exports in underdeveloped countries can be
also differ according to the degree of fluctua- explained not only by a possibly weak stim-
tion in export proceeds. Disruptions in the ulus from a particular type of export, but also
flow of foreign exchange receipts make by the host of domestic impediments that
the development process discontinuous; the limit the transmission of the gains from ex-
greater the degree of instability, the more dif- ports to other sectors even when the stimulus
ficult itis to maintain steady employment, may be strong.
because there will be disturbing effects on After analyzing the character of a coun-
real income, government revenue, capital for- try's export base for an indication of the
mation, resource allocation, and the capacity strength of the stimulus to development pro-
 508 TRADE STRATEGY

vided by its export commodities, we must go It follows that if a more extensive carry-
on to examine the strength of the response or over from exports is to be achieved, it is neces-
diffusion mechanism within the domestic sary to remove the domestic impediments
economy for evidence of how receptive the that cut short the stimulus from exports.
domestic economy is to the stimulus from ex- Many of the policy recommendations in this
ports. The strength of the integrative process, book refer to the need for reducing the frag-
in terms of the response mechanism to the ex- mentation and compartmentalization of the
port stimulus, will depend on the extent of economy by overcoming the narrow and iso-
market imperfections in the domestic econ- lated markets, ignorance of technological
omy and also on noneconomic barriers in the possibilities, limited infrastructure, and slow
general environment. The integrative forces rate of human-resource development. To ac-
are stronger under the following conditions: complish this, alternative forms of economic
the more developed the infrastructure of the and social organization are required, and pol-
economy, the more market institutions are icy measures must aim at diminishing the
developed, the more extensive the develop- prevalence of semimonopolistic and monopo-
ment of human resources, the less are the listic practices, removing restraints on land
price distortions that affect resource alloca- tenure and land use, widening and expanding
tions, and the greater is the capacity to bear financial markets, promoting market facili-
risks. Our view of the carry-over should stress ties and increasing investment in economic
not only the mechanical linkages but also a and social infrastructure, and in promoting
more evolutionary (and hence biological human-resource development.
rather than mechanical) analogy that recog- It also follows that the stimulus from the
nizes societal responses. What matters is not export-base should be as strong as possible.
simply the creation of modern enterprise or While domestic limitations and impediments
modern sectors but modernization as a pro- may have accounted for the weak carry-over
cess. This involves not simply physical pro- of exports in the past when export markets
duction or mechanical linkages but a change were expanding, it may be contended (as
in socioeconomic traits throughout the soci- Nurkse does, VIII. A.2) that exports no
ety, and an intangible atmosphere that re- longer enjoy a strongly rising world demand
lates to changes in values, in character, in at- and do not now provide a sufficient stimulus
titudes, in the learning of new behavior for development in the first instance. If ex-
patterns, and in institutions. ports are confined to a slow rate of growth,
In sum, the effects of a strong integrative then there can be little scope for development
process will be the following: (1) an acceler- through trade even if the domestic obstacles
ation inthe learning rate of the economy; (2) are removed. To counter this "export pessi-
an enrichment of the economic and social in- mism," itis all the more necessary for under-
frastructure (transportation, public services, developed countries to raise productivity in
health, education); (3) an expansion of the agriculture in order to ensure that their pri-
supply of entrepreneurship (and a manage- mary exports are competitive on world mar-
rial and administrative class); and (4) a mo- kets, and to prevent home consumption from
bilization ofa larger surplus above consump- causing a limitation of their export supplies.
tion in the form of taxation and saving. These Further, it is important that the less devel-
effects constitute the country's development oped countries pursue policies that will en-
foundations. Once these foundations are laid, sure that they specialize as much as they can
the country's economy can be more readily in exports with the highest growth prospects.
transformed through diversification in pri- To do this, a country must have the capacity
mary production and the service industries, to reallocate resources — to shift, for instance,
new commodity exports, and industrializa- from exporting a foodstuff which may have
tion via import substitution and export only a slowly growing demand, to the export
substitution. of an industrial raw material or a mineral for
 509 INTERNATIONAL TRADE AND INTERNATIONAL INEQUALITY

which the demand may be rising more rap- traditional primary exports, but also for en-
idly. Of special significance is the country's couraging new manufactured exports. The
potential for taking advantage of new export export market for primary products might
opportunities in manufactured goods. The also be expanded if industrialized nations
exportation of manufactured commodities avoided artificial supports for the substitution
may play a strategic role in transmitting de- of primary products by synthetic materials,
velopment to some poor countries that have a And along with the liberalization of trade, it
favorable factor endowment and can gain a is vitally important that the LDCs should be
comparative advantage by utilizing labor-in- able to look forward to the maintenance of
tensive methods. In selection VIII. B.4, we high rates of growth in the advanced coun-
shall consider more fully the issue of export tries to which they export,
stimulation, giving particular attention to the Finally, advanced countries can contribute
opportunity for exporting manufactures. Ef- by supporting policies to stablize export pro-
forts at regional integration may also pro- ceeds. High and stable levels of employment
mote trade in manufactures among the de- in industrial nations will help reduce the
veloping nations themselves, as will be short-term fluctuations in export proceeds;
discussed in section VIII. D below. but in addition, national and international
Export prospects may also be improved by measures might be advocated to achieve
a more liberal importation policy on the part greater short-run stability in international
of advanced countries. A removal of trade re- commodity markets,
strictions is beneficial not only for the more

Comment
The following readings offer various interpretations of the potentials and limitations of de-
velopment through trade: R. E. Baldwin, "Patterns of Development in Newly Settled Re-
gions," Manchester School (May 1956); R. E. Baldwin, "Export Technology and Develop-
ment from a Subsistence Level," Economic Journal (March 1963); K. Berrill, "International
Trade and the Rate of Economic Growth," Economic History Review (April 1960); J. Bhag-
wati, "Immiserizing Growth: A Geometrical Note," Review of Economic Studies (June
1958); J. Bhagwati, "Growth, Terms of Trade, and Comparative Advantage," Economia In-
ternazionale (August 1959); J. Bhagwati, Foreign Trade Regimes and Economic Development
(1978); H. B. Chenery, "Comparative Advantage and Development Policy," American Eco-
nomic Review (March 1961); G. Haberler, International Trade and Economic Development
(1959); J. R. Hicks, Essays in World Economics (1959), chapter 8; H. G. Johnson, Interna-
tional Trade and Economic Growth (1958), chapter 3; D. Keesing, "Outward-Looking Poli-
cies and Economic Development," Economic Journal (June 1967); Hla Myint, "The Gains
from International Trade and the Backward Countries," Review of Economic Studies, Vol.
22, No. 58 (1954-55); Hla Myint, "The 'Classical Theory' of International Trade and the
Underdeveloped Countries," Economic Journal (June 1958); Hla Myint, Economics of the
Developing Countries, London (1964), chapters 2-5; Hla Myint, "Inward and Outward-Look-
ing Countries of Southeast Asia," Malayan Economic Review (April 1967); R. Nurske,
"Some International Aspects of the Problem of Economic Development," American Economic
Review (May 1952); R. Nurkse, Patterns of Trade and Development (1956); H. W. Singer,
"The Distribution of Gains Between Investing and Borrowing Countries," American Eco-
nomic Review (May 1950); W. A. Lewis, Aspects of Tropical Trade 1883-1965 (1969); H.
W. Singer, The Evolution of the International Economic Order (1978); I. Kravis, "Trade as
a Handmaiden of Growth," Economic Journal (December 1970); Anne O. Krueger, Foreign
Trade Regimes and Economic Development: Liberalization Attempts and Consequences
(1978).
 VIII.B. INWARD- VERSUS
OUTWARD-ORIENTED
DEVELOPMENT STRATEGIES
VIII.B. 1 . Trade Policy as an
Input to Development*

My topic is the question: what difference in lines where domestic opportunity cost ex-
does the set of commercial policies chosen by ceeds the international price ratio. An allo-
a developing country make to its rate of eco- cation of resources satisfying this criterion
nomic growth? Three points are salient. would be optimal in the absence of any dy-
First, in its present state, trade theory pro- namic considerations.
vides little guidance as to the role of trade Theory does not, however, indicate how
policy and trade strategy in promoting many activities are likely to be undertaken.
growth. Second, the empirical evidence over- Nor does it suggest the relative importance of
whelmingly indicates that there are impor- exporting and import-competing activities in
tant links between them. Third, a number of an optimum allocation, or how that alloca-
hypotheses as to the reasons for these links tion would change with growth. Worse yet,
have been put forward, but there is not as yet there is nothing in theory to indicate why a
sufficient evidence to enable us to estimate deviation from the optimum should affect the
their relative importance. rate of economic growth. Most growth
Turning first to theory, there are many models suggest that there are once-and-for-
static propositions but few useful theorems all losses arising from nonoptimal policies
about the effects of alternative trade policies with lower levels of income resulting from
on growth. Clearly there are gains to be them but no change in growth rates.
achieved through trade in the development Turning from theory to practice, develop-
process. Even the trade and growth models ing countries' trade policies have fallen into
along Corden- Johnson lines are based upon two distinct categories. One group of devel-
differential rates of change in capital-labor oping countries has adopted trade policies
ratios in two-country, two-commodity worlds which diverge from the optimality criterion,
under assumptions of free trade. They pro- often by a large amount, by protecting their
vide little indication of the quantitative im- domestic industries. These "import substitu-
portance oftrade as a contributor to growth, tion" policies have been employed to stimu-
and still less insight into the probable orders late domestic production on the theory that
of magnitude of the losses in attainable nonagricultural sectors must grow at a rate
growth rates that may be incurred with de- above the rate of growth of domestic demand,
partures from free trade. and can do so only insofar as additional pro-
To be sure, once the assumption that there duction substitutes for imports. The other
are only two goods is abandoned, theory sug- category, "export promotion," has consisted
gests that activity in production of tradables of encouragement to exports, usually beyond
should be undertaken to the point where the the extent that would conform to the IMRT
international marginal rate of transformation = DMRT criterion. Countries adopting an
(IMRT) equals the domestic marginal rate of export-oriented trade strategy have generally
transformation (DMRT), with no production experienced rapid growth of traditional ex-
*From Anne O. Krueger, "Trade Policy as an
ports, but even more rapid growth of nontra-
ditional exports.
Input to Development," American Economic Review,
Papers and Proceedings, May 1980, pp. 288-92. Re- Experience has been that growth perfor-
printed bypermission. mance has been more satisfactory under ex-

510
 511 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

port promotion strategies (meant as a general associated with export promotion contrasted
bias toward exports and not as a package of with import substitution. There are three
specific measures to encourage selective ex- major hypotheses, and each undoubtedly
ports of particular items themselves induced contains some explanatory power. . . .
by a bias toward import substitution) than The first hypothesis is that technological-
under import-substitution strategies. While it economic factors imply an overwhelming su-
is impossible to specify a particular model of periority for development through export
growth process that will simultaneously sat- promotion. These factors include such phe-
isfy all observers, the relationship between nomena as minimum efficient size of plant,
export performance and growth is sufficiently increasing returns to scale, indivisibilities in
strong that it seems to bear up under many the production process, and the necessity for
different specifications of the relationship. It competition. According to this hypothesis,
has been tested over many countries for: (1) failure to take advantage of the opportunities
rates of growth of real GNP and of exports to exploit these phenomena through trade
(see Michael Michaely); (2) for real GNP significantly impairs the attainable rate of
net of exports and exports (see Bela Balassa); growth. A second hypothesis is that differ-
and (3) for rates of growth of GNP as a func- ences in growth rates have resulted, not from
tion of rate of capital formation, aid receipts, the choice of trade strategy per se, but rather
and export growth (see Constantine Michal- from excesses in the ways in which import
opoulos and Keith Jay). Time-series and substitution policies were administered. The
cross-section data have been pooled, so that third hypothesis is that policies adopted in
deviations of countries' growth rates from pursuit of an export promotion strategy are
their trends have been estimated as a func- generally far closer to an optimum, both in
tion of the growth of export earnings (see my the DMRT = IMRT sense and with respect
book, p. 27 Iff). In all of these specifications, to the domestic market, than are those
rate of growth of exports has turned out to be adopted under import substitution. Under
a highly significant variable. While the "suc- this interpretation, the role of trade policy is
cess stories" of Korea, Taiwan, and Brazil to constrain policymakers in such a way that
are well known, there are enough other ob- they do not impede the growth rate as much
servations, both for different time periods in as they otherwise would.
the same country (as for example Turkey and Both the first and second hypotheses are
the Philippines) and of countries (including consistent with the notion that the nonagri-
on the positive side Ivory Coast, Colombia, cultural sector of most developing countries
and Malaysia and on the negative side India, is, in some sense, an "infant industry," and
Argentina, and Egypt), so that there is little requires some stimulus for growth. The third,
doubt about the link between export perfor- by contrast, essentially takes the negative
mance and growth rates. view, that markets would function well and
Moreover, it seems clear the export perfor- provide satisfactory growth if only policy-
mance isa function in large part of govern- makers would abstain from counter-produc-
mental policies. While an export promotion tive intervention.
strategy will not always be successful in gen- The first hypothesis really amounts to an
erating more export growth (especially if pol- assertion that the gains from trade, especially
icies affecting the domestic market are inap- for developing countries, are so sizable that
propriate), certainly policies adopted to the losses associated with import substitution
encourage import substitution, especially significantly reduce the rate of return on fac-
when they include overvalued exchange rates tor accumulation. On the negative side, do-
and quantitative restrictions upon imports, mestic markets are extremely small in most
retard the growth of exports. . . . developing countries, and attempts to replace
The central question, then, is why such a imports result in the construction of plants of
difference in growth performance should be less-than-efficient minimum size, while si-
 512 TRADE STRATEGY

multaneously generating an oligopolistic or one in capital-labor ratios between activities

monopolistic market structure. As import at the prevailing wage-rental ratio, while the
substitution proceeds, new activities are in- rate of capital accumulation is the binding
creasingly capital intensive and inefficiencies constraint on expansion of employment in the
from below-minimum-efficient size increase. urban sector, an allocation of additional cap-
On the positive side, so the argument runs, ital to the labor-intensive activity for export
export promotion permits entrepreneurs to will permit an upward shift in the demand for
base their plans on whatever size plant seems labor three times as great as that which
appropriate: size of domestic market is no would occur if import substitution dictates
longer a virtually binding constraint, as it is the start of the more capital-intensive activ-
when the activity is profitable only because of ity. To be sure, the expectation is that the
very high rates of effective protection. More- more rapid rate of growth of demand for in-
over, monopoly positions arise less frequently dustrial labor would drive up the urban wage
under export promotion, as exporters face once the demand for labor was rising more
competition from abroad as well as from rapidly than the labor force, but this is pre-
other domestic producers. cisely adesired outcome of policy. . . .
Export promotion may also be more effi- The second hypothesis focuses upon the
cient in permitting rapid expansion of profit- costs of import substitution policies as in fact
able activities; by contrast, under import sub- carried out, and suggests that alternative
stitution, most activities are constrained to means of achieving import substitution might
expanding at approximately the same rate: have avoided them. According to this view,
inefficient firms and sectors expand approxi- the failure of import substitution resulted
mately as rapidly as efficient ones. In this from the excesses of the particular ways in
view, potential export lines consist of a num- which domestic industries were encouraged:
ber of industrial products (girls' sneakers, extreme currency overvaluation combined
wigs, tennis rackets, engine parts, plywood, with quantitative restrictions provided the
and so on) and it is as much a matter of the equivalent of prohibitive tariff protection;
right entrepreneur, and the right specialized techniques of allocating import licenses were
product, as choosing the "right industry" employed which prevented competition
that is necessary for rapid growth. To be among domestic firms and rewarded entre-
sure, factor proportions and comparative ad- preneurs for license-getting abilities rather
vantage may result in greater profitability of than their cost-minimizing performance; and
relatively labor-using industries, but the excessive and detailed quanitative controls
basic notion is that there are thousands of in- were employed over many aspects of eco-
dustrial products, and that, among relatively nomic activity. One of the costs was the fail-
labor-intensive activities, the ones which will ure of export earnings to grow as much as
develop into exports will be those in which
there are firms with good management and they would have under "better" import-sub-
an ability to utilize factors of production stitution policies; that in turn led to "stop-go"
patterns with their attendant costs. Simulta-
efficiently.
neously, the emerging "foreign exchange bot-
A final aspect of the technology-related tleneck" had both direct and indirect impacts
view of the advantages of export promotion upon the structure and growth of the econ-
has to do with factor proportions. Given the
omy. In particular, efforts at "import substi-
vast disparity in capital-labor ratios of the in- tution" stopped being geared toward devel-
dustrial sectors of the developed and devel- opment of economic new industries, and
oping countries, the opportunity for trade
became focused upon "foreign exchange sav-
represents a means for shifting the demand
ing," often in highly irrational and indiscrim-
for labor outward more rapidly than the im- inate ways, which further distorted the
port-substitution strategy permits. If there
are differences of, say, two-to-one and six-to- system.
The third view denies the need for any bias
 513 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

toward exports and implicitly or explicitly as- both of entrepreneurs and of government of-
serts that growth would be optimal in the ab- ficials, and simultaneously provides feedback
sence of intervention. A bias toward exports to them as to the success or failure of policies
is therefore better than one toward import- in terms of their objectives.
substitutes only because policies are less dis- Undoubtedly, all three approaches to the
tortive. In this view, an export-oriented strat- differential in economic performance contain
egy imposes constraints on policymakers, elements of truth. There are export opportu-
both in what they can attempt to do, and in nities that are passed up under import sub-
making them aware of the costs of mistakes. stitution where indivisibilities or increasing
Policymakers receive feedback in a relatively returns within a range would permit sizable
short time period as to the costs of their pol- gains in output. There are also high-cost im-
icies. Also, it is infeasible to rely upon quan- port substitution activities which, if never un-
titative controls: the international price, at dertaken, would have freed resources for con-
least, cannot be administered and to that ex- siderably more productive use, even within
tent, more generalized forms of incentive, in- an import-substitution strategy. Likewise,
cluding a relatively realistic exchange rate, the international market has served to con-
must be employed. Indeed, it is argued that strain policymakers and induce them to
incentives cannot be as biased toward export abandon uneconomic policies sooner than
promotion as they can be toward import sub- they otherwise would have done. Knowledge
stitution. This is precisely because to do so is not yet far enough advanced to determine
would require either export subsidization the relative importance of the alternatives. It
(whose costs would be immediately evident will not be until we have far more informa-
through the drain on the budget) or such a tion than is currently available about the
degree of currency undervaluation that a cur- order of magnitude of indivisibilities and
rent account surplus would absorb much of minimum size plant contrasted with size of
the country's savings potential. markets in LDCs, and also about the deter-
According to this third line of argument,
minants of politicians' and bureaucrats' be-
constraints upon policymakers go well be- havior. Moreover, it is certain that the pri-
yond the inability to impose too great a bias mary sources of growth are internal, and that
toward exports. For example, it is virtually there is no magic formula, or single policy
impossible to administer any highly protec- change, that can by itself account mean-
tive system for intermediate and capital ingfully for differences in economic
goods imports if exporters are expected to performance.
compete in international markets: they must Nonetheless, experience has clearly dem-
be permitted ready access to imported raw onstrated the importance of access to inter-
materials, intermediate goods, and capital national markets in providing a means of per-
equipment. To impose any comprehensive mitting more rapid growth than would
system of licensing or controls would entail otherwise be feasible. Given the enormous dif-
delays and other costs, inconsistent with the ficulties and costs of achieving the institu-
export strategy. Thus, the commitment to an tional and other changes that economic
export-oriented development strategy implies growth requires, it is probable that trade pol-
a fairly liberal and efficient trade regime, and icy changes have a higher rate of return to
thus prevents paperwork, delays, bureau- LDCs than most other feasible policy
cratic regulation, and other costs that can changes. It is, of course, to be hoped that pro-
arise under import substitution. This in turn tectionist pressures in the developed coun-
limits the restrictions that can be imposed on tries do not result in fewer opportunities for
capital account. More generally, under an the LDCs. If such protectionist measures are
export promotion strategy, there is an inter- taken, they will lower the rate of return to
national market in the background: it func- outward-oriented trade strategies. They will
tions as a constraint upon economic behavior, however, for the foreseeable future, still leave
 514 TRADE STRATEGY

that rate distinctly above the returns from a I. Kravis, "Trade as a Handmaiden of
policy of persisting with inward-oriented Growth: Similarities between the Nine-
growth. teenth and Twentieth Centuries," Eco-
nomic Journal, December 1970, Vol. 80,
References Anne O. Krueger, Foreign Trade Regimes
pp. 850-72.
B. Balassa, "Exports and Economic Growth: and Economic Development: Liberalization
Further Evidence," Journal of Develop- Attempts and Consequences, Cambridge,
ment Economics, June 1978, Vol. 5, pp. Mass., 1978.
181-89. et al., Alternative Trade Strategies
W. M. Corden, "The Effects of Trade on the and Employment, Chicago, 1981.
Rate of Growth," in Jagdish Bhagwati et M. Michaely, "Exports and Growth: An Em-
al. (eds.), Trade, Balance of Payments, and pirical Investigation," Journal of Develop-
Growth, Amsterdam, 1971. ment Economics, March 1977, Vol. 4, pp.
49-53.
H. G. Johnson, "The Theory of Trade and
Growth: A Diagrammatic Analysis," in C. Michalopoulos and K. Jay, "Growth of
Jagdish Bhagwati et al. (eds.), Trade, Bal- Exports and Income in the Developing
ance of Payments, and Growth, Amster- World: A Neoclassical View," disc, paper
dam, 1971. No. 28, A.I.D., November 1975.

VIII. B. 2. Phases in Exchange

Control Regimes*

[We may delineate] exchange control re- pattern of its use of control and price
gimes into certain basic types or, what we instruments.
term, Phases that have a broadly differen-
tiated impact on economic efficiency and
A. PHASE I
performance.
This notion of Phases, shortly to be defined This Phase is characterized by the system-
but essentially based on the restrictionist con- atic and significant imposition of Quantita-
tent and associated blend of control and price tive Restrictions (QR). It might start in re-
instruments, reflected initial familiarity with sponse to an unsustainable payments deficit
the evolving exchange control regimes in spe- resulting from intense or sustained prior in-
cific countries and the strongly suggested hy- flationary pressure (due, for example, to the
pothesis that these regimes went through a initiation of a large-scale development plan
sequencing. These Phases in the exchange and consequent extraordinary increases in
control regimes were designed essentially as government expenditures) or from a sharp
a classificatory and descriptive device to cap- drop in world prices for some major exports
ture meaningfully the evolution of the ex- (as in 1953). These reasons for instituting
change control system in terms of its restric- controls are of interest in ascertaining the
tionist content and the dimensions and
logic of the evolution of exchange control re-
gimes but are not critical to the definition of
*From J. Bhagwati's Foreign Trade Regimes and Phase I.
Economic Development: Anatomy and Consequences
of Exchange Control Regimes, copyright © 1978, Throughout Phase I (which, of course, can
National Bureau of Economic Research. Reprinted be of varying duration) controls are generally
with permission from Ballinger Publishing Company, maintained and often intensified. This con-
pp. 56-9. tinuation ofincreased severity might result
 515 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

from any of several interrelated factors: (1) is a concern with evasions, both legal and il-
the continuation of initial controls might be legal, of the system. Thus, export licensing
necessary to contain an ex ante unsustainable may be adopted to ensure that exporters do
payments deficit; (2) the initial set of controls not succeed in their attempts at capital flight;
might result in evasions of the system controls over tourist and other invisible trans-
through illegal transactions that negate any actions are aimed at preventing the emer-
possible impact on the balance of payments; gence of a black market, and so on.
(3) once controls are instituted, policy-mak- In Phase II, however, the rules of the re-
ers may perceive the new instruments as gime have become more complex and differ-
handy for a variety of purposes in addition to entiated. The growth of the bureaucracy to
controlling the payments situation and begin administer the allocations and the inevitable
employing them for such purposes; and (4) tendency to differentiate increasingly among
policy-makers may perceive their actions as alternative end uses and claimants as the
freed from a balance of payments constraint control regime is perceived to be a continuing
and adopt policies that in fact require restric- state of affairs interact to produce a complex
tiveness afresh to offset further adverse ef-
system.
fects on the balance of payments. 2. The other major feature of Phase II is
an increasing resort to price measures to sup-
B. PHASE II plement the QR regime. This generally oc-
curs with respect to both exports and imports.
This phase is characterized by continued The continuation or intensification of foreign
reliance upon quantitative restrictions and, exchange shortage leads to recognition that
indeed, generally increased restrictiveness of additional export earnings would be desira-
the entire control system. However, Phase II ble. Rebate schemes, import replenishment
is distinguished by two additional and related schemes, special credits for exporters, and a
aspects of the QR system, both relatively un- variety of other devices may be instituted
important during Phase I: (1) for a variety of that offset part or all of the discrimination
reasons, indicated below, the detailed work- against exports implicit in an overvalued ex-
ings of the control system become increas- change rate. Like the quantitative restric-
ingly complex; and (2) price measures are tions discussed above, however, export incen-
adopted to buttress the functioning of the tives tend to be adopted in a piecemeal and
control system. Both of these characteristics fragmented fashion. As for imports, price
of Phase II stem from dissatisfaction with the measures are also adopted to absorb part of
results of an undifferentiated system and are the excess demand for imports. Tariffs may
often the result of many small decisions be increased or surcharges added to the cost
rather than an overall policy design. of importing. Guarantee deposits are gener-
1. Although neither term is quite appro- ally required on various categories of im-
priate, Phase I decision-making and policies ports. These and other measures tend to re-
may be characterized as "crude" and "unso- duce (but hardly eliminate) the windfall gain
phisticated." Quantitative restrictions are or premium accruing to the recipients of im-
applied with relatively few rules, and treat- port licenses.
ment of competing claims to import licenses The following aspects of the price situation
tends to be relatively undifferentiated in the in Phase II are then evident: (1) the exchange
sense that rules for allocation tend to be of an
parity is "overlaid" by tariffs and subsidies;
"across the board" nature, such as allocating levied in lieu of formal parity change; (2) the
to everyone a certain percentage of this per- effective exchange rate for exports, on the av-
son's imports over a specified number of ear- erage, israrely as high as the effective ex-
lier years or allotting licenses pro rata with change rate for imports, implying a bias
applications. The major motivation for against the (relative) incentive to export; (3)
changes in the control system during Phase I the domestic currency is overvalued at the
 516 TRADE STRATEGY

current parity plus trade tariffs and subsi- D. PHASE IV

dies, implying a premium (on the average) on Phase IV is entered when Phase III has re-
imports; and (4) this import premium varies sulted incontinued liberalization. This means
across different activities, implying differen- that the average import premium has fallen,
tial incentives to produce and invest in these
different activities. the bias against exports has substantially di-
minished (the effective exchange rate for ex-
ports having come closer to the effective ex-
C. PHASE III change rate for imports), and the degree of
Phase III is entered against the backdrop dispersion in the incentives to expand differ-
ent activities has diminished through greater
just defined, and it can take various forms. It
uniformity in these incentives via the foreign
may consist of a mere "tidying-up" operation trade sector.
directed at replacing the diverse import pre- Needless to say, this outcome can only be
miums by reasonably uniform tariffs such anticipated from a Phase III episode when
that the differential incentive effects caused
the latter takes the ambitious form of a con-
by diverse premiums on different imports are
certed effort of the kind presupposed in a de-
greatly reduced or virtually eliminated. Al- valuation cum liberalization package.
ternatively., thetidying-up operation may re-
place the existing tariffs and export subsidies E. PHASE V
with a formal parity change, the result being
that the (average) effective exchange rates on The transition from Phase V occurs when
exports and imports do not change much but the exchange regime is virtually liberalized in
the dispersion of tariffs is replaced by the uni- the sense that there is full convertibility on
form devaluation — an action again of ration- current account and quantitative restrictions
alization inthis instance. On the other hand, are not employed as a means of regulating
Phase III may be substantially ambitious and the balance of payments. Thus, an economy
take the form of what might be described as in Phase V is not an exchange control regime
a devaluation cum liberalization package. in the usual sense of the term, and Phase V
Such a package may have a gross devaluation represents a total alternative to the QR re-
large enough to leave a net devaluation de- gimes of Phases I and II. The pegged ex-
spite the removal of the trade tariffs and sub- change rate is at its equilibrium level, a flex-
sidies, with accompanying grants by donors ible exchange rate policy is in operation, or
of additional credits to facilitate early expan- monetary and fiscal policies are employed as
sion of imports, debt rescheduling, and simi- the instruments to achieve payments balance
lar measures aimed at quick and easy liber- in contrast to a reliance on the exchange con-
alization ofimports. trol mechanism.

VIII. B. 3. Inward-Oriented
Strategies*

Industrial development generally begins in the primary sector that also provides investi-
response to domestic demand generated in ble funds for manufacturing industries. De-
mand for industrial products and investible
♦From Bela Balassa, The Process of Industrial De- savings represent possible uses of the surplus
velopment and Alternative Development Strategies,
Princeton University, International Finance Section, generated in agriculture, or in mining. The
Essays in International Finance No. 141, December surplus is generated as primary output comes
1980, pp. 4-11. Reprinted by permission. to exceed subsistence needs and, more often
 517 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

than not, it is associated with export cipient manufacturing industries producing

expansion. for domestic markets. There were differ-
At the same time, the effects of primary ences, however, as regards the rate and the
exports on industrial development depend to form of protection. While the industrial
a considerable extent on input-output rela- countries of today relied on relatively low tar-
tionships and on the disposition of incomes iffs, a number of present-day developing
generated in the export sector. Infrastructure countries applied high tariffs or quantitative
in the form of ports, railways, and roads are restrictions that limited, or even excluded,
often important inputs for primary exports, competition from imports.
and their availability may contribute to the At the same time, high protection discrim-
development of industrial activities. inates against exports through the explicit or
The disposition of incomes generated in implicit taxation of export activities. Explicit
the export sector is affected by ownership taxation may take the form of export taxes
conditions. In the case of foreign ownership, while implicit taxation occurs as a result of
a substantial part of the surplus may be re- the effects of protection on the exchange rate.
patriated, albeit taxing the earnings of for- The higher the rate of protection, the lower
eign capital does add to domestic incomes. will be the exchange rate necessary to ensure
There are leakages in the form of investing equilibrium in the balance of payments, and
and spending abroad, as well as consuming the lower the amount of domestic currency
imported luxuries, in the case of domestic exporters receive per unit of foreign exchange
ownership in a system of plantation-type ag- earned.
riculture and large-scale mining, too. And, as The adverse effects of high protection are
Douglas North noted, plantation owners have exemplified in the case of Ghana, where im-
little incentive to finance human investment port prohibitions encouraged inefficient,
in the form of general education. high-cost production in manufacturing in-
By contrast, in cases when family-size dustries: taxes on the main export crop,
farms predominate, demand is generated for cocoa, discouraged its production: and other
the necessities and the conveniences of life as crops were adversely affected by the unfavor-
well as for education. Such demand contrib- able exchange rate. Ghana's neighbor, the
utes to the development of domestic industry Ivory Coast, in turn, followed a policy en-
that enjoys "natural" protection from im- couraging the development of both primary
ports in the form of transportation costs. It and manufacturing activities. As a result, it
further contributes to the accumulation of increased its share in cocoa exports, devel-
human capital that finds use in manufactur- oped new primary exports, and expanded
ing industries. manufacturing industries.
The process of industrial development may Differences in the policies applied may
be accelerated if natural protection is com- largely explain that, between 1960 and 1978,
plemented bytariff or quota protection. This per capita incomes fell from $430 to $390 in
last point, in turn, leads me to the discussion Ghana, in terms of 1978 prices, as compared
of the next step in the industrialization pro- to an increase from $540 to $840 in the Ivory
cess: the first, or "easy" stage of import Coast. This has occurred notwithstanding the
substitution. fact that the two countries have similar nat-
ural resource endowments and, at the time of
independence, Ghana had the advantage of a
THE FIRST STAGE OF IMPORT higher educational level and an indigenous
SUBSTITUTION civil service corps.
Indeed, there is no need for high protection
With the exception of Britain at the time at the first stage of import substitution, en-
of the Industrial Revolution, and, more re- tailing the replacement of the imports of non-
cently, Hong Kong, all present-day industrial durable consumer goods, such as clothing,
and developing countries protected their in- shoes, and household goods, and of their in-
 518 TRADE STRATEGY

puts, such as textile fabrics, leather and ket conditions for primary exports and lack of
wood, by domestic production, since these competitiveness in manufactured exports
commodities suit the conditions existing in would not permit developing countries to at-
developing countries that are at the begin- tain high rates of economic growth by relying
ning of the industrialization process. The on export production. Rather, Prebisch sug-
commodities in question are intensive in un- gested that these countries should expand
skilled labor; the efficient scale of output is their manufacturing industries oriented to-
relatively low and costs do not rise substan- wards domestic markets. This purpose was to
tially at lower output levels; production does be served by industrial protection that was
not involve the use of sophisticated technol- said to bring additional benefits through im-
ogy; and a network of suppliers of parts, com- provements inthe terms of trade.
ponents, and accessories is not required for ef- Similar ideas were expressed by Gunnar
ficient operations. Myrdal. Myrdal influenced the policies fol-
The relative advantages of developing lowed by India, which were also affected by
countries in these commodities explain the the example of the Soviet Union that chose
frequent references made to the "easy" stage an autarkical pattern of industrial develop-
of import substitution. At the same time, to ment. And, the European socialist countries
the extent that the domestic production of faithfully imitated the Soviet example; they
these commodities generates external econo- attempted to reproduce the Soviet pattern in
mies in the form of labor training, the devel- the framework of much smaller domestic
opment of entrepreneurship, and the spread markets and also lacking the natural resource
of technology, there is an argument for mod- base of the Soviet Union.
erate infant industry protection or promotion. Second-stage import substitution involves
the replacement of the imports of intermedi-
THE CHOICE OF SECOND-STAGE ate goods and producer and consumer dura-
IMPORT SUBSTITUTION bles by domestic production. These commod-
ities have rather different characteristics
In the course of first-stage import substi- from those replaced at the first stage. Inter-
tution, domestic production will rise more mediate goods, such as petrochemicals and
rapidly than domestic consumption, since it steel, tend to be highly capital-intensive.
not only provides for increases in consump- They are also subject to important economies
tion but also replaces imports. The rate of of scale, with efficient plant size being large
growth of output will however decline to that compared to the domestic needs of most de-
of consumption, once the process of import veloping countries and costs rising rapidly at
substitution has been completed. lower output levels. Moreover, the margin of
Maintaining high industrial growth rates, processing is relatively small and organiza-
then, necessitates turning to the exportation tional and technical inefficiencies may con-
of manufactured goods or moving to second- tribute tohigh costs.
stage import substitution. This choice, in Producer durables, such as machinery, and
fact, represents alternative industrial devel- consumer durables, such as automobiles and
opment strategies that may be applied after refrigerators, are also subject to economies of
the completion of the first stage of import scale. But, in these industries, economies of
substitution. scale relate not so much to plant size as to
Second-stage import substitution was un- horizontal and to vertical specialization, en-
dertaken in the postwar period in several tailing reductions in product variety and the
Latin American countries, some South Asian manufacture of parts, components, and ac-
countries, in particular India, as well as in cessories on an efficient scale in separate
the European socialist countries. In Latin
America, it responded to the ideas of Raul Reducing product variety will permit
plants.
Prebisch, in whose view adverse foreign mar- longer production runs that lower production
 519 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

costs through improvements in manufactur- This will occur as goods produced at earlier
ing efficiency along the "learning curve," sav- stages have come to saturate domestic mar-
ings in expenses incurred in moving from one kets. High protection, in turn, discriminates
operation to another, and the use of special- against manufactured and primary exports
purpose machinery. Horizontal specialization and against primary activities in general.
is however limited by the smallness of do-
mestic markets in the developing countries.
Similar conclusions apply to vertical special- CHARACTERISTICS OF INWARD-
ization that leads to cost reductions through ORIENTED DEVELOPMENT
the subdivision of the production process STRATEGIES
among plants of efficient size.
Given the relative scarcity of physical and In the postwar period, several capitalist
human capital in developing countries that countries in Latin America and in South
completed the first stage of import substitu- Asia and the socialist countries of Central
tion, they are at a disadvantage in the man- and Eastern Europe adopted inward-oriented
ufacture ofhighly physical capital-intensive industrial development strategies, entailing
intermediate goods and skill-intensive pro- second-stage import substitution. Capitalist
ducer and consumer durables. In limiting the countries generally utilized a mixture of tar-
scope for the exploitation of economies of iffs and import controls to protect their in-
scale, the relatively small size of their na- dustries whereas socialist countries relied on
tional markets also contributes to high do- import prohibitions and on industry level
mestic costs in these countries. At the same planning.
time, net foreign exchange savings tend to be Notwithstanding differences in the mea-
small because of the need for importing ma- sures applied, the principal characteristics of
terials and machinery. the industrial development strategies applied
The domestic resource cost ratio relates in the two groups of countries show consid-
the domestic resource cost of production, in erable similarities. To begin with, while the
terms of the labor, capital, and natural re- infant industry argument calls for temporary
sources utilized, to net foreign exchange sav- protection until industries become interna-
ings (in the case of import substitution) or tionally competitive, in both groups of coun-
net foreign exchange earnings (in the case of tries protection was regarded as permanent.
exports). In the absence of serious distortions Also, in all the countries concerned, there
in factor markets, the domestic resource cost was a tendency towards what a Latin Amer-
(DRC) ratio will be low for exported com- ican economist aptly described as "import
modities. Itis also relatively low for con- substitution at any cost."
sumer nondurables and their inputs, in the Furthermore, in all the countries con-
production of which developing countries cerned, there were considerable variations in
have a comparative advantage. However, for rates of explicit and implicit protection
the reasons noted beforehand, DRC ratios among industrial activities. This was the
tend to be high in the manufacture of inter- case, first of all, as continued import substi-
mediate goods and producer and consumer tution involved undertaking activities with
durables. increasingly high domestic costs per unit of
Correspondingly, the establishment of foreign exchange saved. In capitalist coun-
these industries to serve narrow domestic tries, the generally uncritical acceptance of
markets is predicated on high protection. demands for protection contributed to this re-
Also, rates of protection may need to be sult, when, in the absence of price compari-
raised as countries "travel up the staircase," sons, the protective effects of quantitative re-
represented by DRC ratios, in embarking on strictions could not even be established. In
the production of commodities that less and socialist countries, the stated objective was to
less conform to their comparative advantage. limit imports to commodities that could not
 520 TRADE STRATEGY

be produced domestically, or were not avail- on short-term objectives on the part of man-
able in sufficient quantities, and no attempt agers discouraged technological change.
was made to examine the implicit protection The managers' emphasis on short-term ob-
the pursuit of this objective entailed. jectives insocialist countries had to do with
In both groups of countries, the neglect of uncertainty as to the planners' future inten-
intraindustry relationships further increased tions. In capitalist countries, fluctuations in
the dispersion of protection rates on value real exchange rates (nominal exchange rates,
added in processing, or effective protection, adjusted for changes in inflation rates at
with adverse effects on economic efficiency. home and abroad) created uncertainty for
In Argentina, high tariffs imposed on caustic business decisions. These fluctuations, result-
soda at the request of a would-be producer ing from intermittent devaluations in the face
made the theretofore thriving soap exports of rapid domestic inflation, aggravated the
unprofitable. In Hungary, the high cost of do- existing bias against exports as the domestic
mestic steel, based largely on imported iron currency equivalent of export earnings varied
ore and coking coals, raised costs for steel- with the devaluations, the timing of which
using industries and large investments in the was uncertain.
steel industry delayed the substitution of alu- In countries engaging in second-stage im-
minum for steel, although Hungary had con- port substitution, distortions were further ap-
siderable bauxite reserves. parent in the valuation of time. In capitalist
Countries applying inward-oriented indus- countries, negative real interest rates ad-
trial development strategies were further versely affected domestic savings, encour-
characterized by the prevalence of sellers' aged self-investment, including inventory ac-
markets. In capitalist countries, the size of cumulation, at low returns, and provided
national markets limited the possibilities for inducements for the transfer of funds abroad.
domestic competition in industries estab- Negative interest rates also necessitated
lished at the second stage of import substi- credit rationing that generally favored im-
tution while import competition was practi- port-substituting investments, whether it was
cally excluded by high protection. In socialist done by the banks or by the government. In
countries, the system of central planning ap- the first case, the lower risk of investments in
plied did not permit competition among do- production for domestic as compared to ex-
mestic firms or from imports and buyers had port markets gave rise to such a result; in the
no choice among domestic producers or ac- second case, it reflected government priori-
cess to imported commodities. ties. Finally, in socialist countries, ideological
considerations led to the exclusion of interest
The existence of sellers' markets provides
little inducement for catering to the users' rates as a charge for capital and as an ele-
needs. In the case of industrial users, it led to ment in the evaluation of investment projects.
backward integration as producers undertook There was also a tendency to underprice
the manufacture of parts, components, and public utilities in countries following an in-
accessories in order to minimize supply diffi- ward-oriented strategy, either because of low
culties. This outcome, observed in capitalist interest charges in these capital-intensive ac-
as well as in socialist countries, led to higher tivities or as a result of a conscious decision.
costs, since economies of scale were foregone. The underpricing of utilities benefited, in
Also, in sellers' markets, firms had little in- particular, energy-intensive industries and
centive toimprove productivity. In capitalist promoted the use of capital.
countries, monopolies and oligopolies as- In general, in moving to the second stage
sumed importance, and the oligopolists often of import substitution, countries applying in-
aimed at the maintenance of market shares
ward-oriented development strategies de-em-
while refraining from actions that would in- phasized the role of prices. In socialist coun-
voke retaliation. In socialist countries, the ex- tries, resources were in large part allocated
istence of assured outlets and the emphasis centrally in physical terms; in capitalist
 521 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

countries, output and input prices were dis- fodder seeds, and beans, declined in absolute
torted and reliance was placed on non-price terms and slow increases in production neces-
measures of import restrictions and credit sitated the imports of cereals and meat that
allocation. were earlier major export products.
The slowdown in the growth of primary ex-
ports and the lack of emergence of manufac-
EFFECTS ON EXPORTS AND ON tured exports did not provide the foreign ex-
ECONOMIC GROWTH change necessary for rapid economic growth
in countries pursuing inward-oriented indus-
The discrimination in favor of import sub- trial development strategies. The situation
stitution and against exports did not permit was aggravated as net import savings de-
the development of manufactured exports in clined because of the increased need for for-
countries engaging in second-stage import eign materials, machinery, and technological
substitution behind high protection. There know how. As a result, economic growth was
were also adverse developments in primary increasingly constrained by limitations in the
exports as low prices for producers and for availability of foreign exchange, and inter-
consumers reduced the exportable surplus by mittent foreign exchange crises occurred as
discouraging production and encouraging attempts were made to expand the economy
consumption. at rates exceeding that permitted by the
In fact, rather than improvements in the growth of export earnings.
external terms of trade that were supposed to Also, the savings constraint became in-
result, turning the internal terms of trade creasingly binding as high-cost, capital-in-
against primary activities led to a decline in tensive production at second-stage import
export market shares in the countries in ques- substitution raised capital-output ratios, re-
tion. Decreases in market shares were espe- quiring ever-increasing savings ratios to
cially pronounced in cereals, meat, oilseeds, maintain rates of economic growth at earlier
and nonferrous metals, benefiting developed levels. At the same time, the loss of incomes
countries, in particular, the United States, due to the high cost of protection reduced the
Canada, and Australia. volume of available savings and, in capitalist
The volume of Argentina's principal pri- countries, negative interest rates contributed
mary exports, chiefly beef and wheat, re- to the outflow of funds.
mained, on the average, unchanged between In several developing countries, the cost of
1934-38 and 1964-66 while the world ex- protection is estimated to have reached 6-7
ports of these commodities doubled. In the percent of the gross national product. At the
same period, Chile's share fell from 28 per- same time, there is evidence that the rate of
cent to 22 percent in the world exports of growth of total factor productivity was lower
copper, which accounts for three-fifths of the in countries engaging in second-stage import
substitution than in the industrial countries.
country's export earnings.
Similar developments occurred in socialist Rather than reducing the economic distance
countries where the allocation of investment vis-a-vis the industrial countries that infant
favored industry at the expense of agricul- industry protection was supposed to promote,
ture. In Hungary, the exports of several ag- then, there was a tendency for this lag to in-
ricultural commodities, such as goose liver, crease over time.

Comment
Although policymakers tend to view import substitution and export promotion as alterna-
tive strategies, a rational trade policy would pursue both strategies until the last unit of do-
mestic resources devoted to each yields the same return in terms of foreign exchange saved
through import substitution or earned through exports. Export promotion does not mean that
 522 TRADE STRATEGY

there should be no import-substitution industries, and import substitution does not mean that
no new products should be produced for export. A rational policy implies no bias in favor of
either strategy.
Appropriate domestic policies, the exchange rate, and trade policies should allocate re-
sources efficiently to internal and external opportunities. As a condition of optimal resource
allocation, policies should equate the marginal domestic resource cost of saving foreign ex-
change with the marginal domestic resource cost of earning foreign exchange.
If in many countries the bias has in the past been too heavily in favor of import substitution,
why cannot the bias now become too great in favor of exports in a country that adopts an
export-promoting strategy? The possibility does exist, but excessive export bias is less likely
than excessive import bias. This is because an export promotion regime usually begins after
a preceding period of import substitution that has already been so biased against exports that
incentives for exports can be considerable before they reach a point of oversubsidizing exports.
Moreover, export-promoting measures tend to rely on price incentives while import-substi-
tution regimes are usually administered through quantitative controls that are characterized
by considerable dispersion and unpredictability. Export promotion measures are also subject
to more public scrutiny to the extent that they involve budgetary measures rather than quan-
titative restrictions. Exporting firms must also face price and quality competition in interna-
tional markets, and the exporting firms are more likely to realize economies of scale than are
import-substitution industries producing for only a narrow domestic market.
A number of country studies indicates that the economic cost of incentives distorted toward
export promotion has been less than the cost of those distorted toward import substitution.
Most notable is the research project sponsored by the National Bureau of Economic Research
(NBER), now published in ten country volumes. Two synthesis volumes are: J. Bhagwati,
Foreign Trade Regimes and Economic Development: Anatomy and Consequences of Exchange
Control Regimes (1978) and Anne O. Krueger, Liberalization Attempts and Consequences
(1978). Other relevant empirical evidence is provided by Bela Balassa in The Newly-Indus-
trializing Countries in the World Economy (1981) and Development Strategies in Semi-In-
dustrial Countries (1981).

VIII. B. 4. Trends and Prospects

for Export of Manufactures-
Note

As preceding selections indicate, developing what are the prospects for making such a
countries frequently utilize trade policy to strategy actually effective?
promote their industrialization. If the pro- This question might first be explored
motion of import-substitution industrializa- against the background of some of the statis-
tion (ISI) was the earlier attempted strategy tical trends. Although the developing coun-
of industrialization, there is now a new hope tries' share of world GNP from manufac-
that industrialization might be more success- turing was still less than 10 percent,
fully based on the processing of the LDCs' manufacturing output has been growing
primary materials for export and on the ex- faster in the LDCs than in the more devel-
port of other semimanufactures and manu- oped countries (MDCs). Growth rates have
factures in which LDCs might become com- been especially high in LDCs relative to
petitive. Arguments in favor of an export MDCs in metal products, including ma-
substitution strategy of industrialization chinery and equipment, basic metals, and
were presented in selection VIII.B.l. But clothing and footwear. Advances in supply
 523 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

capability cannot be denied. Exports of man- percent from 1960 to 1976, while the rate
ufactures have also grown faster in the de- was approximately 8.8 percent for developed
veloping countries than in the developed countries. Some rates of export growth are
countries. Real growth rates in exports of given in Table 1. Growth in exports of man-
manufactures from developing countries ex- ufactures byregions can be noted in Table 2.
ceeded an average annual growth rate of 1 2 Table 3 offers a comparison of the growth in

TABLE 1. Past and Projected Rates of Export Growth by Broad Product Groups (in constant 1975 prices)
Percent of LDC Percent Share of
Exports Increase
World LDCs World LDCs
1960-75 1960-75 1975-85 1975-85 1960 1975 1985
1960-75
42 1975-85
18
40
Fuel and energy 6.3 6.2 3.6 39
43 20
30 16
4.4 3.4
Agricultural products 4.2 2.6 3.1 27 12
4.8 4.2 437
Non-fuel minerals 3.9 7 7 6 6
Manufactures 12.3 21 5.8
12.2 11 26
8,9
7.1 100 100 100
Total merchandise 5.9 6.4 100 100 36 64
6.4

Source: World Bank, World Development Report, 1978, tables 13 and 25.

TABLE 2. Exports of Manufactures — Absolute Value, Growth, and Relationship to Exports and to Gross
Manufacturing Output

Value (billion current U.S. dollars) Real Manufactures

Exports as a
Rate of All Exports Manufacturing
Region and Group 1965 1973 1974 1975 1973
1965-74 Output 1973

Developing 4.6 24.3 32.5 16.3 21.5 10

33.2
Latin America 0.63 4.09 6.41 6.55 21.1 14.9 4
East Asia 1.75 13.65 17.52 17.83 20.9 48.9 30
Turkey and
0.63 2.14 2.86 3.11 51.3 108
Yugoslavia 10.7 49.6
South Asia 1.00 2.35 2.87 2.90 12.4
Middle East and 4.2
North Africa 0.33 2.1 2.0 15.0 6
1.47 4.8
Sub-Saharan Africa 0.25 0.6 0.8 4.9 155
0.7
289.0 419.0 71.3
Developed 84.5 386.0 10.8
Transitional 1.06 6.22 8.56 8.26 18.0 70.8 14
Other Western
Europe 244.61 268.56 76.2 25
55.23 188.22 10.4
North America 19.73 56.51 77.42 85.73 59.1 8
7.66 8.9 93.6 11
Japan 34.58 51.30 52.63 15.6
Australia, New
Zealand, South
Africa 3.43 4.05 22.3 8
0.84 3.92 11.4
Centrally planned 12.7 32.8 38.7 46.7 3
5.9 56.7
World total 101.8 346.1 457.2 498.9 10.6 60.0 12

Source: "World Trade and Output of Manufactures: Structural Trends and Developing Countries' Exports," World Bank Staff
Working Paper No. 314, January 1979, p. 12.
 no

m" -^ cn tj- r- oo
© in ON
c-i m on
sq no in m — rs) no
od — -^
CM <N —
<N —

On
r-- in © - *o n
mm© <N O tN <N m ©
<N m no <n m m'm
|
—
<N m
NO <N
in r-^
on no r-^

HSt <n m tj-

rn od — '•

— r-° r-'

ON <N
I I I <N ON
'5

u- o so r- p <n m m

% m — ©
<N <N m oo r- "* on m
m r~- OO
200 s _o
O n -I
"*. P
NO

P
15
:N © ©
■n © -^
3
C o o o
C
e«
O
— Tf o
- M
o
^
in Tt vd
en m On
—' ^
U (N <N w-n

r-^

•■* —
m in oo
q w /; p OO
NO On
NO

^- ^- ^^
oo <n r-
«-* iri
«-* '-, — —
vo in vd en in r-
od i^ O in no
00 ON — no in p
•5
rs) — —
c
3
O »~<
U m' rt vo
00 'EL 2 ^t m <n NO <N Tf
r- in in
in on o
c c <N © ©
NO
o rs r^
in >n 00
u
> m* o O
oo
to
.2
m m m r*
J5E © © in on
— on
in in in m r~ ^r ON
c
Tt o no Tt in r>.
O «n r-^ 1 m ~ r-i
—
o o
o
O
jd
NO
(N NO p
in r- in o m — m 4>
rn in ©
On in
|
rs
CN
<N
rsi —
d c

E
o
rn NO
o
a.
p ~
o m m tj- no in m m oo
od

1
m p no in on no m
sD
m
ON 1 1 1
PQ X
m <N r-' oo

r- on i; no q
—. — m r~-
'
NO NO <N m oo rN
d
CN
o © d >
3

rn m" rfr ON
ON

ON

II
3
o
^
r*
O no m
so o r~
O no m 1 o no m -s;
N3 o no m
O O no m
r-

M
^ no No.r- no
m no
O r-
no ^> X
*> mIo noI
2 no no r^-I ~s m o
b no no
no r^ ft.NO NO 1^
^ m © no <«> mIo no C m o no S3 W-i
I NOI noI —. in no no
3" vs m no no 3 in no no "3 m no no
JC ^3 ON ON ON On
ITi ON
NO On
NO ^-
°
^ *. 5
"^J ON ON ON O On On ON
£ ON ON ON •ii on on on

If
r-
I *
524
 525 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

exports of manufactures and primary prod- 1975. In 1975, 9 countries were each export-
ucts from several developing countries. ing more than $2 billion of manufactures,
Keesing lists the following reasons for the whereas a decade earlier no countries were.
especially rapid growth in exports of manu- The range of exports has also widened. In
factures from developing countries: almost any year in the 1960s through 1976,
1. Export incentives and industrial policies real growth rates have been over 10 percent,
have been improved in many developing not only in clothing and electronic assembly,
countries, shifting away from excessive but also in machinery, transport equipment,
textile yarn and fabrics, steel, chemicals, and
emphasis on import substitution and pro-
duction only for the domestic market; almost every other major group of manufac-
2. Exporting of manufactures is so new that tured exports. Although exports have been
it has expanded from a small base with built initially around labor-intensive, tech-
much benefit from learning and cumula- nologically standardized, "older" products
tive experience; such as textiles, clothing, and footwear, sev-
3. Supply capabilities have leaped forward eral countries have graduated to the subse-
through transfer of technology and bor- quent generation of exports that are more
rowing ofproduction methods from devel- skill-intensive and capital-intensive. Other
newcomers to the export of manufactures
oped countries, aided by rapid build-ups
of skilled manpower, know-how, and in- have then replaced the more advanced LDCs
vestment insome developing countries; with exports of the older labor-intensive
4. Trade liberalization in developed coun- products. Table 5 shows the increasing diver-
tries has been favorable to imports from sification in the exports of developing
developing countries; countries.
5. Businesses in developed countries — retail In examining the imports of the industrial
chains, other large buyers, trading firms, countries from the non-OPEC developing
and multinational manufacturing corpo- countries, the largest increases have been in
rations— have played an active and cru- the specialized machinery and the road
motor vehicles categories; the ratios of 1978
cial role while seeking low-cost sources of
to 1973 exports are 4.5 and 4.8, respectively,
supply based on low-cost labor. Their ef-
forts, which have greatly accelerated ex- as compared to an overall average of 2.7. The
port growth, have been spurred by slow next largest increases, with the ratio of 1978
labor force growth and rising real wages to 1973 imports being 3.7 and 3.4, took place
in the other categories of machinery and
in developed countries as well as by de-
clining transport costs, technological and transport equipment and household appli-
organizational innovations, and a favora- ances. All in all, the industrial countires' im-
ble international policy environment.1 ports of engineering products from the non-
The bulk of manufactured exports still OPEC developing countries shows above-av-
erage increases, with the ratio of 1978 to
comes from a fairly small number of rela- 1973 imports being 3.4.
tively advanced industrial LDCs, as shown in The developing countries have made little
Table 4. However, an increasing number of
progress in exporting commodities that are
LDCs have been able to participate in the highly intensive in physical capital, require
growth of manufactured exports. More than sophisticated technology, or necessitate the
40 LDCs had exports worth over $100 mil- availability of precision-engineered parts,
lion in 1979, compared with 22 in 1970. components, and accessories. Several of the
Countries with exports of manufactures over
$1 billion increased from 3 in 1965 to 12 in LDCs, however, have made progress in ex-
porting skilled, labor-intensive commodities,
such as ships and TV sets, as well as parts,
'Donald B. Keesing, "World Trade and Output of
Manufactures," World Bank Staff Working Paper
components, and accessories of engineering
No. 316, January 1979, pp. 6-7. products to the industrial countries. Growth
 526 TRADE STRATEGY

TABLE 4. Leading Developing-Economy Exporters of Manufactures in 1965 and 1975

Value (millions
current U.S. dollars)
Percent Share Cumulative
1965 1975 1965 1975 Share 1975
Country or Territory

Hong Kong (including

reexports) 989 5,590 21.5 15.8 15.8
Excluding reexports n.a. 4,464 — — —
4.1
Taiwan 187 4,303 2.3 12.2 28.0
Korea 104 4,136 11.7
47.9
39.7
Yugoslavia 617 2,903 13.4 8.2
Singapore (including 6.5 6.3
reexports) 300 2,233 54.2
Excluding reexports n.a. 1,286 — — —
Brazil 124 2,192 6.2
2.7
17.6 60.4
66.0
India 5.6
809 l,961a 71.9
Mexico (including border) 166 2,090 3.6 5.9
2.0 73.9
Argentina 84 723 1.8
75.8
Malaysia 68 664 1.5 1.9
Pakistan 190a 571 1.6 77.4
All developing countries 4,590 35,280 100.0 100.0 100.0

Sources: United Nations and national trade statistics.

"Including what is now Bangladesh (exported 178 in 1975).

TABLE 5. Composition of LDC Trade in Manufactures by Destination, 1975 (percent)

4.1"
Share of Inter-LDC
Trade among LDC Exports LDC Imports Trade in Total
to MDCs from MDCs LDC Imports

Machinery and transport LDCsa

18 55 286
equipment 31
Textiles 14 10 4
Clothing 5 19 121 39
13
Chemicals 7 11
Iron and steel 6 415 10 7
Other manufactures 31 19 Ji
Total 100 100 100 10

Source: United Nations, Yearbook of International Trade Statistics, various issues.

"Includes significant reexports, notably in machinery and transport equipment.

has also been rapid in the exports of other Table 5 illustrates the compositional dif-
consumer goods that require average skill in- ferences between manufactures exported to
tensity and are relatively intensive in un- MDCs and to other LDCs in 1975. In the ag-
skilled labor.2 gregate, nearly two-thirds of LDC manufac-
tured exports go to MDCs, nearly one-third
2Bela Balassa, "Structural Change in Trade in to other LDCs, and only limited amounts to
Manufactured Goods Between Industrial and Devel-
oping Countries," World Bank Staff Working Paper the central planned economies. A more com-
No. 396, June 1980, pp. 19-21. prehensive view of world trade in manufac-
 527 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

tures can be gained from the matrix pre- intertemporal pattern of comparative advan-
sented inTable 6. tage proceeds roughly from unskilled, labor-
One can see that about 60 percent of world intensive to skill-intensive to physical, capi-
trade in manufactures takes place among de- tal-intensive, and so on, to the high-
veloped countries, that developing countries' technology products in more advanced
imports of manufactures are more than three economies.
times as large as their exports of manufac- Beyond the initial capacity to acquire a
tured goods, and that nearly 85 percent of comparative advantage in resource-based or
their manufactured imports come from de- labor-intensive commodities, there are other
veloped market economies. forces that may widen the basis of exports of
The important question now is: What is the manufactures from the DCs. Some of the ef-
potential for future growth in exports of man- fects of technological change, for example,
ufactures from the LDCs? The answer will may go in this direction. Technological
depend on the changing pattern of compara- change is usually interpreted as promoting
tive advantage in the process of economic de- more trade in new manufactured commodi-
velopment, protectionism in the MDCs, ties that embody a high degree of research
growth in the MDCs, and LDC policies with and development or are intensive in highly
respect to an outward-looking strategy of de- skilled labor. Although it is true that much of
velopment. We may now explore the first two the trade in manufactures is composed of
factors, leaving the other factors to be ana- products embodying advanced scientific and
lyzed in the following two selections. technical knowledge and that the LDCs lack
Once we appreciate the fact of changing the necessary basis for producing and export-
comparative advantage, we can recognize a ing these products, it must nonetheless be
greater potential for the export of manufac- recognized that technological information is
tures. Comparative advantage proceeds eventually diffused and that a breakthrough
through "stages" within a developing country achieved in one country stimulates emulation
according to which the structure of exports in other countries. More significantly, with
changes with the accumulation of physical the lapse of time, a new product will pass
and human capital.3 A developing country's through a "product cycle" in the course of
which the technology becomes more stan-
dardized and thus more readily transferable
to an LDC.
3Bela Balassa, "A 'Stages' Approach to Compara-
tive Advantage,"
Growth in Irma National
and Resources: Adelman and
(ed.), Economic
International
As a product passes through its "new" and
"growth" phases into its "mature" phase, its
Issues, Vol. 5, 1979, pp. 121-56. production process tends to become more

TABLE 6. Matrix of World Trade in Manufactures, 1976 (billion U.S. dollars)

Centrally
World
(Part to)
Europe Developing
Developed Planned Total
Exports From/To

Developed 321 (224) 121 469

(of which Western 27
Europe) 45
(226)
28 (193) (61) (307)
(20)
Developing (10) 14 3
9 (8) 7 34 50
Centrally planned 142
World total 358 (242)
564
64
Source: "World Trade and Output of Manufactures: Structural Trends and Developing Countries' Exports," World Bank Staff
Working Paper No. 316, January 1979, p. 17.
 528 TRADE STRATEGY

standardized, requiring less of skilled man- ufactured exports from newly industrializing
agement and less of scientific and engineer-
ing know-how, and thus makes more use of The potential, however, will not be realized
countries.4
relatively unskilled labor and standardized if the MDCs impose import restrictions. The
machinery, which are easily obtainable and trade policies of the MDCs are therefore im-
maintainable. The production process of a portant in determining whether exports of
mature product may therefore be more easily manufactures from the LDCs are to grow
adopted by a newly developing country more rapidly and include a wider range of
(hence the early introduction of the textile in- products.
dustry into industrializing countries). Insofar Greater access to the markets of the devel-
as the mature products are standardized oped countries requires that the policies of
manufactures, their export marketing is also MDCs condone trade liberalization. Al-
relatively easier because specifications and though the "average" nominal tariff rate on
prices for these nondifferentiated products manufactured goods in the major industrial
are easily available and their markets are countries has been reduced considerably
highly price sensitive. through successive rounds of trade negotia-
It may be expected that eventually a given tions, there is still a wide dispersion of tariffs
technological advantage will be dissipated over the range of manufactures and semi-
and will give way to conventional factor-cost manufactures ofpotential interest to the
advantages, so that the new line of produc- LDCs, and in some cases even the nominal
tion may become more accessible to devel- tariff rates are very high. There is still a
oping countries. strong degree of protectionism in the tariff
As technological conditions and factor en- structures of the major industrial nations:
dowments change in a developing country, so nominal tariff rates still tend to rise with the
too will its comparative advantage in differ- degree of labor-intensiveness; effective rates
ent manufactured goods. The country may of protection tend to be much higher than
progress up the export ladder from an early nominal tariff rates; and effective tariffs on
generation of exports to a higher generation. the products of special interest to developing
At the same time as comparative advantage countries are generally higher than those on
changes within a country, the commodity other products, so that the tariff structure in
composition of trade changes among coun- the major developed countries remains
tries. As a country graduates into exports strongly protected against imports of labor-
with greater domestic value added, its lower- intensive manufactured goods from the
value exports are taken over by another LDCs.
newly emerging exporting country that has Even more restrictive are the nontariff bar-
been in the queue. Thus, exports of manufac- riers— especially quotas — placed on imports
tures may initially consist largely of textile from LDCs. Textiles are, for example, the
yarn and fabrics and other standardized in- most important manufactured commodities
termediate goods, such as leather, plywood, now exported by the LDCs, but textiles con-
and cement. Later, consumer goods and then front severe quota limitations. The original
capital goods may contribute a rising share of negotiation of the Arrangement Regarding
exports. Ultimately, the leading developing International Trade in Cotton Textiles was
countries may export technically complex
products, such as supertankers, aircraft, and based on the concept of "market disruption."
heavy machinery. Through this continually Members of GATT may place "restraints"
evolving process of upgrading and diversifi-
cation of manufactured exports within a de- 4For an empirical study of prospective changes, see
veloping country, and the movement of de- Bela Balassa, "Prospects for Trade in Manufactured
Goods Between Industrial and Developing Countries,
veloping countries through the queue, the 1978-1990," Journal of Policy Modeling, Vol. 2,
potential is raised for the expansion of man- No. 3, 1980, pp. 437-55.
 529 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

on particular categories of cotton textile im- access on a broader front through trade lib-
ports ifincreased imports constitute "market eralization can be more significant for
disruption," defined as: (1) a sharp and sub- LDC exports than preferences subject to
stantial increase in imports of a particular safeguards.
item; (2) import prices substantially below After prolonged but unsuccessful attempts
domestic prices in the importing country; and to secure actual effective preferential treat-
(3) serious damage to domestic producers or ment, the LDCs may be better advised to em-
threat thereof. Not only does the invocation phasize market access across the most-
of "market disruption" allow severe restric- favored-nation tariffs. Such tariff reductions
tionist measures on textiles (the major export and removal of nontariff barriers would allow
of more developing countries than any other greater access to developed markets than any
class of manufactured goods), but it also sets feasible preference system. This requires,
an undesirable precedent for other manufac- however, some liberal safeguard system and
tures that might receive similar treatment. adjustment assistance to raise the import-ab-
No policies promoting manufactured exports sorption capacity of the developed importing
will be effective if the concept of "market dis- countries. The basic problem is that of reduc-
ruption" isinvoked to restrict the entry of ing the adjustment costs entailed by a chang-
manufactures into advanced markets. This ing international division of labor in order
contradicts any attempt to grant preferences that the real sources of comparative advan-
to LDCs. If a preference system is qualified tage might operate to change the structure of
trade. Unless the MDCs can move out of
by the imposition of tariff quotas,5 then the
efficacy of preferences is clearly diminished. stagflation to high rates of growth and adopt
If developed countries are to be receptive to industrial policies and adjustment assistance
imports from LDCs, they must provide ad- programs that will facilitate the movement
justment assistance measures to move re- out of declining industries, the forces of pro-
sources out of displaced industries in which tectionism might react against imports from
the comparative advantage of developed LDCs.
countries has been lost to new, more efficient, At the same time that the MDCs must re-
higher technology product lines. Market move their trade restrictions, the LDCs must
pursue policies to take advantage of wider
5Under a system of tariff quotas, the importing opportunities. If new exports are to be sup-
country specifies the quantity of imports of a product
it is prepared to import at the preferential duty rate, ported, the LDCs will have to undertake ex-
with all imports greater than this amount being sub- port promotion policies. These policies are
ject to the full most-favored-nation rate. considered in the next selection.

Comment
Given their concern about the labor absorption problems in the LDCs, policymakers have
given special attention to the employment impact of a more rational trade policy. Some stud-
ies examine the employment effect through intercountry or intertemporal comparison of trade
patterns and the transformation of the industrial structure of the labor force. See, for exam-
ple, Hollis B. Chenery and Moises Syrquin, Patterns of Development 1960-1970(1975); Hol-
lis B. Chenery, "Interactions between Industrialization and Exports," American Economic
Review (May 1980).
Other studies compare the labor intensity of import-substituting and export industries at
each point in time. Under an export promotion strategy, export industries grow faster, as do
import-competing industries under import substitution. If employment per unit of output and
value added are greater in one set of industries than the other, then employment growth will
be faster under the trade strategy that allows the labor-intensive industries to grow relatively
faster.
 530 TRADE STRATEGY

EXHIBIT VIII. 1. Growth of Merchandise Trade

100)
Average Annual Growth Rate (percent)
Ten T1S Of

Merchand ise Trade Trade (1975

(millions of dollars) Imports
=
Exports

Exports Imports -0.7 m 1978

101
1981 1981 1981
1960-70 1970-81 1960-70 1970-81
-0.8 m 87 m
Low-income economies 42,444 f 60,117 t 5.3 ma 2.4 m 109 m
China and India 29,624 t — — — —-3.8 — —
36,567 / 1.9 m 88 m
Other low-income 12,820 t 5.0 m -7.2 5.4 m 110m
23,550 t
1 Kampuchea, Dem. — — —4.9 m* -—0.7 — — — —
2 Bhutan — — — -—0.8 — — — —
9 85 -2.8
3 Lao, PDR -5.7
4 Chad 141 Ill 79
137 5.9 7.1
5.1
5 Bangladesh 791 6.5 99
2,594 -0.4 5.1
-11.6
6 Ethiopia 374
738
195
3.6
6.2
cr 158
7 Nepal 63 69
123
455 1.3 -9.1 110
8 Burma 373 -3.1 112
263 484 -1.5
1.8 -5.3
9 0.8 -11.9 107
9 Afghanistan 7.1.8 7.4
8.9
10 Mali 154 370 2.9 110 102
7.6 108 74
11 Malawi 284 359 11.6 5.9 2.4 82
12 Zaire 662 672 5.5
-0.9 100 106
4.9
13 Uganda 395 144
317
71 6.2 98
14 Burundi -1.2
7.6
75 167 7.3 106
338 14.6 -8.1 8.0
15 Upper Volta
16 Rwanda 147 . 191 15.8 8.0 180
4.6
0.4 10.7 107
17 India 15,001 3.2 2.6 3.2 108 66
18 Somalia 8,064 199 2.4 3.2 109 98
200 113
19 Tanzania 566 6.7 6.0 121
1,410
791 3.4 -0.2
20 Viet Nam 153 -1.5
-7.0
21 China 21,560 21,566
22 Guinea 428 351 — — — —-1.9 — —
23 Haiti 333 587
— — — —-1.6 — —
4.6
24 Sri Lanka 1,036 -4.4 -3.1
1.4 151 80
1,803
886 7.5
25 Benin 36 5.0 -2.5 5.0 101
26 Central African 84
4.5 73
Rep. 136 88 1.9 116
8.1 104
27 Sierra Leone 277 238 1.9 75
0.4 112
28 Madagascar 335 494 -5.2 4.1 -16.7
449 5.4 -15. 0 -5.0 106
29 Niger 297 6.0 23.4 11.9 4.0 117 87
-7.1 -1.5 13.4
30 Pakistan 2,880 8.3 3.0 5.3 88
5,342
457 774 6.0 7.9 97 77
31 Mozambique 4.2
32 Sudan 658 2.2 0.6 86
84 88
1,529 75
63
33 Togo 344 597 10.5 1.5 8.5 10.3
34 Ghana 878 0.2 193
97
1,184
Middle-income economies 337,172 r 403,729 t 5.4 m 4.1 m 6.4 m 4.8 m 98 m 87 m
Oil exporters 146,227 t 148,221 / 3.9 m 2.5 m 2.8 m 9.5 m 95 m 133 m
7.0 m 4.3 m 7.-m
Oil importers 198,945/ 257,508 t 3.2 m 101 m 73 m
 531 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

EXHIBIT VIII. 1. Growth of Merchandise Trade (Continued)

Average Annual Growth Rate (percent) 100)

Ter ms of
Merchandise Trade Trade (1975
(millions of dollars) =
Exports Imports -1.8
-1.9
Exports Imports 1981
1981 1981 - 1.4 1960-70 1978
1960-70 1970-81 1970-81
-0.3 77 m
5.2 m 3.0 m 4.1 m 98 rr
Lower middle-income 98,497 t 122,588 t 6.5 m
7.2
35 Kenya 6.5
2.3 -1.5 144 99
1,144
416 1,946 1.2 2.5 72
68
36 Senegal 1,035 4.6
37 Mauritania 259 265 50.6 81
97
39 3.2
38 Yemen Arab Rep.
421 1,699
39 Yemen, PDR -1.9
1,096
448 18.4 2.9 -6.8 88 63
40 Liberia 531
1.2 11.9
41 Indonesia 22,259 -0.2
6.5 2.0 95
— 13,271
— 3.4
— — -—0.9 — — —
154
42 Lesotho 7.0 153
909 825 9.6 8.0 129 75
43 Bolivia
44 Honduras 760 949 10.7 4.2 11.6 1.9 102
45 Zambia 2.3 9.8 89
1,044 1,032 3.2 83 86
46 Egypt 3,233 67
8,839 0.4 9.4 129 80
47 El Salvador 792 986 5.5 2.3
4.9
0.7 6.3 -1.3 62
48 Thailand 5.2 -111.8
2.7 11.4
6,918 10,014
49 Philippines 7.7 7.2 2.6 98
87 68
5,722 7,946 2.2
50 Angola 9.0 11.5 0.2 103 152
1,744 1,640 74 63
51 Papua New Guinea 851 1,116 2.5 76
51 Morocco 113
2,242 4,356
731 2.2 3.4
53 Nicaragua 529 10.5 5.4
9.7 0.2 -1.0 17.8
54 Nigeria 18,727 18,776 6.5 102 190
0.5 1.7
55 Zimbabwe 663 704 — — — — 81
7.0 168
56 Cameroon 4.9 9.3 72 90
94
1,079 1,428 4.0 1.2 6.9
1.6 65
57 Cuba -6.5
1,128 791 5.1 5.4 82 96
1,897 16.8
58 Congo, People's Rep. 1,040 -6.7 7.1 77
59 Guatemala 9.0 72
1,281 5.0 6.7
5.2 134
1,774
4.6 90
60 Peru
3,255 -2.0
2.3 3.8 0.5 75
61 Ecuador 3,803 9.3 136
2,562 2.9 11.6 78
62 Jamaica 2,332 4.6 5.7 107
974 8.1
1,473 150
107
63 Ivory Coast 2,586 8.7 5.1 9.7
2,434 10.0 2.2 49 49
64 Dominican Rep. 1,188 1,450 3.8 5.7
2.2 1.6 2.5 145
64 Mongolia 3,190 4.2 4.0 6.5
66 Colombia 5,181 81 127
2,209 4.0 9.2 104
67 Tunsia 3,924 4.2 2.3 9.2 81
2,209 4.0 104
68 Costa Rica 968 3,924 9.5 10.0 2.2 125
— —
1,198 — — —
2.3 — — —
69 Korea, Dem. Rep.

70 Turkey
— — 2.0 95 87
4,703 1.2 72
8,911 5.2 4.1 12.6 99 130
71 Syrian Arab Rep. 2,103 3.3
732 4,663 74 61
72 Jordan 3.149 10.1 21.2 7.6 67
3.6 110
73 Paraguay 296 506 5.4 6.8 6.6
13.9
 532 TRADE STRATEGY

EXHIBIT VIII. 1. Growth of Merchandise Trade (Continued)

Average t\nnual Growth Rate (percent) 100)

Ter ms of
Merchandise Trade Trade (1975
(millions of dollars) Imports =
Exports

Exports Imports 1978

1981 1981 1981
1960-70 1970-81
-2.1 1960-70 1970-81
-4
5.4 m -13.
7.04 m 5.9 m 4.7 .m3 97 m
89 m
Upper middle-income 238,675 t 283,141 t
-1.9 20.6 10.9 105
74 Korea, Rep. of 21,254 26,131 33.4 22.0
12.5 11.6 10.5 94
75 Iran, Islamic Rep. of 10,169 12,634 -1.1 67
76 Iraq 1.4 23.6 217
209
9,372 18,907 5.4 7.1 101
6.8 2.3 109 93
72 Malaysia 12,884 13,132 5.8 94
78 Panama 10.2 10.5 93
315 1,540
79 Lebanon 14.4 1.9 5.1 3.3 101 88
1,107 3,946 12.0
80 Algeria 14,056 11,505 3.5 1.0 4.9 96 196
81 Brazil 5.0 8.7 -1.4
2.9 108 56
23,172 24,007 9.5
82 Mexico 20,033 24,168 2.8 15.3 6.4 92 89
9.6 — — — —71
83 Portugal 14.2
4,147 9,799 -2.9
84 Argentina 3.5 9.4 0.4
4.8 77
6,304 9,425 9.8 3.2 73
85 Chile -7.0 3.5 88
3,952 6,364
86 South Africa 22,670 21,485 5.4
0.7 7.2 8.2 4.6 80 61
87 Yugoslavia 10,929 7.7 4.5 8.8 104 99
15,817 4.3
88 Uruguay 2.2 3.6 124
92
1,215 1,599
89 Venezuela 20,959 10,645 1.1 4.4 9.6
-5.8 87
212
90 Greece 10.8 10.8 98 88
4,292 10.8 103 100
8,677 -4.9 9.2 4.7
12.1
91 Hong Kong 21,737 24,680 12.7 9.7 2.2 101 68
92 Israel 10.8 9.6
5,416 7,777 8.8 —
93 Singapore 20,967 27,608 4.2 12.0 9.9 102
94 Trinidad and 5.9
4.9 -7.5
Tobago 3.115 -1.5 m 3.2 96 129
3,725
-9.4
High-income oil
exporters 174,131 t 11.0m 11.0m 20.8 m 94 m 208 m
68,249 t
213
95 Libya 16,391 15,414 67.5 15.4 15.7
4.5
96 Saudi Arabia 120,240 35,244 11.0 11.2 94 205
33.5
97 Kuwait 16,561 5.2 16.0 92
94 210
8,042 10.7
98 United Arab 4.5
Emirates 20,939 — 25.6 96 189
9,549 5.5
Industrial market
economies 8.5 m 5.4 m 9.5 m 4.4 m 100 m 90 m
1,210,104 t 1,290,415/
99 Ireland 7,706 10,603 7.1 8.3 90
8.4
— —
6.4 104
99
100 Spain 20,337 33,159 11.5 18.5
101 Italy 72,215 91,022 13.6 3.6 101 86
87
6.7 9.7
2.9
102 New Zealand 4.6
5,563 5,684 4.8 3.9 1.7 114 107
103 United Kingdom 102,807 101,991 5.0 3.6 105
79
6.6 107
104 Japan 152,016 143,287 17.2 9.0 13.7 112
9.6 7.3 3.9
105 Austria 15,845 21,048 9.6 88
6.7
97
 533 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

EXHIBIT VIII. 1. Growth of Merchandise Trade (Continued)

Average Annual Growth Rate (percent) 100)

Ter ms of
Merchandise Trade Trade (1975
(millions of dollars) =
Exports Imports

Exports Imports
1981 1981 1960-70 1978 1981
1970-81 1960-70 1970-81
4.8 7.0 2.5
106 Finland 14,015 6.8 92 84
14,202 95
107 Australia 21,767 23,768 6.5 3.8 7.2 92 92
4.2 5.2
108 Canada 69,907 66.010 10.0 9.1 5.5 93
9.5
109 Netherlands 9.9 5.0 3.5 96
100 96
68,732 65,921 4.6 10.3 5.3 90
110 Belgium 55,705 62,464 10.9
111 France 100,497 120,924 8.2 6.6 11.0 6.5 101 90
112 United States 233,739 273,352 6.0 9.8 4.4 95 86
7.1 6.5
1 13 Denmark 16,317 4.9 2.5
17,874 8.2 101
163,934 5.8 10.0 5.5 101 87
86
114 Germany, Fed. Rep. 176,043 10.1
9.1 7.0 4.3 93
92 129
1 15 Norway 18,220 15,652 4.2
2.2 9.7
7.2
116 Sweden 28,630 28,824 7.7 9.0 4.5 102
117 Switzerland 27,043 4.2 108 87
30,696 8.5
East European
nonmarket economies 150,270/ 146,968 t 9.4 m 6.7 m 8.6 m 6.1 m — —
-0.4
118 Albania — — -—0.3 — — — — —96
119 Hungary 8.2 9.1 6.1 98
8,893 9.7 — — — —
1 20 Romania 12,610 8,854
12,458 9.4 8.8
121 Bulgaria 12.9 — —
1,848 14.4 11.6
2,633 8.7
6.0 101 98
122 Poland 13,182 15.224 6.7
123 USSR 79,003 72,960 9.7 5.6 7.1 8.3 — —
7.0
1 24 Czechoslovakia 5.1 — —
14,876 14,658 6.7
8.3 —
6.4 8.6 — — —
125 German Dem. Rep. 19,858 20,181

Source: World Bank, World Development Report, 1983, pp. 164-65.

indicates a total.
^Indicates a median value.
Indicates a value substantially less than one.

A number of studies indicate that the creation of employment is greater from export ex-
pansion than import substitution. Westphal found that the labor-capital ratio for manufac-
tured exports in Korea in 1966 was 3.24 compared with 2.67 for domestic production and
1.98 for import substitution; see Larry E. Westphal, "The Republic of Korea's Experience
with Export-Led Industrial Development," World Development, Vol. 6, No. 3 (1978). Krue-
ger's calculations of the ratio of labor per unit of domestic value added in exportable and
import-competing industries for several countries also indicated that exportables are in gen-
eral more labor-intensive than import substitutes; see Anne O. Krueger, "Alternative Trade
Strategies and Employment in LDCs," American Economic Review (May 1978).
When resources are switched from import substitution to export promotion, employment is
created for two reasons: (1) because of the greater degree of labor intensity in exportables,
and (2) because of the higher overall rate of return to investment in the economy and the
 534 TRADE STRATEGY

EXHIBIT VIII. 2. Structure of Merchandise Exports

Percentage Share of Merchandise Exports

Machinery
Fuels, Other and
Primary Transport
Minerals Textiles
Clothingand Other
Metals, and Commodities Manufactures
Equipment
1960 1980 1960 1980 1960 1980 1960 1980
1960 1980

Low-income economies 70 w 18w 4 w

18 w
China and India — — 37 w —15 w — —6 w 23 w
20 w 30 w 18w 27 w
Other low-income 8 w 9 w 83 w 62 w 4 w 21 w 5w
2w
9wa 6 w
1 Kampuchea, )*> —
Dem. 0 — 100 — 0 — 0(. 0 —
2 Bhutan
3 Lao, PDR C)w 5h>
4 Chad 3 — 94 — 0 — 0 — 3 —
5 Bangladesh — — — 49 — (•) — 17
34
6 Ethiopia 0 8 100 0 (•) 0 0
7 Nepal — (•) — 92 — 24 — (•)0 — (•)7
8 Burma 4 — 95 —69 0 — 0 — 1 —
9 Afghanistan (•) — 82 — — 3 — 1 —
96 —
— — 141 — 1 — 2
10 Mali 0
(.)*
11 Malawi — — 90 — 5 — 4 — 1
12 Zaire 42
(•)
— — 0 — 0 — 1 —
13 Uganda 8 — 57
92 — 0 — 0 — (•) —
14 Burundi — (•) — 99 — (•) — — 1
1 5 Upper Volta 0 (•) 100 89 0 2 0 (•)2 (•) 7

16 Rwanda 45
17 India 10 7 35 22 1 7 9
98
34 (•) C) 134
18 Somalia 0 1 88 74 0 8 301
19 Tanzania (.) 10 0 8 0 1 7
20 Viet Nam 87
16
— 25 — — — —
21 China 28 5 26
22 Guinea
42 — — 0 — 0 — 0 —
23 Haiti 0 — 58
100 — 0 — 0 — 0 —
34 Sri Lanka (•) 16 99 65 0 11 0 1 0 7
25 Benin 10 — 80 — 7 — (.) — 3 —
26 Central African
Rep. 12 (•) 86 74 (•) (•) 1 (•) 1 26
27 Sierra Leone 15 — 20 — 0 — 0 — 65 —
28 Madagascar 4 9 90 1 2 1 2 4 3
29 Niger — — 100 —43
84
230 — 0 — 0 —
73
30 Pakistan 0 7 1 2 3 11

31 Mozambique 0 — 100 —96 0 —37 0 — 0 —

32 Sudan 0 1 100 0 1 0 2 0 (•)
33 Togo 3 58 3 4 0 3 5 3
34 Ghana 7 — 89
83 —32 0 — 0 — 10 —
 535 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

EXHIBIT VIII. 2. Structure of Merchandise Exports (Continued)

Percentage Share of Merchandise Exports

Machinery
Fuels,
Other
Primary and
Minerals, and Textiles
Clothingand Transport
Other
Metals Commodities Manufactures
Equipment
1980 1980 1960 1980
1960 1980 1960 1980 1960 1960

Middle-income
economies 30 w 7 w
36 w 59 w 27 w 10w
Oil exporters 48 w 78 w 48 w 15 w 3w 9 w 1w 18 w
2w 2w
Oil importers 1w 14 w
15 w 12w 68 w 34 w 13 w 103 ww 273ww
5w 2w
Lower middle-income 20w 44 w 76 w 38 w 1w (.)w 2w 3w 11 w
50 0 51 w 0 3 12 12
35 Kenya 1 46
34 87
36 Senegal 3 94 1 1 1 3 1 11
37 Mauirtania 4 —39 — 1 — (.)w
20 — 6 —
69 20
38 Yemen Arab
Rep. — (•) — 49 — 6 — 25 —
— 75 — — — —
39 Yemen, PDR 25 (•) (•) (•)
45 38 0 0 1 0 2
40 Liberia 59
76
55 (•)
41 Indonesia 33 22 0 1 (•) (•) 1
42 Lesotho 67 11 (•)
43 Bolivia — 86 — — (•) — 1 — 2
93
44 Honduras 5 7 81 0 2 0 2 10
(•)
45 Zambia
46 Egypt 4 22 0 0 3 2
67
47 El Salvador 0 5 84 59 3 13 (•) (•)3 3 20
94
48 Thailand 7 14 91 0 9 (■)0 6 2
49 Philippines 10 21 86 42 1 6 0 2 3 29
14
57
50 Angola
5 1 Papua New 46
Guinea 0 92 52 0 0 8 2
54 Nigeria 8 95 89 4 0 (•)
(•) 0 () 3 1

55 Zimbabwe 71 — 25 — 1 — —(•) 3 —
56 Cameroon 19 33 77 90 0 1 (•)2 2 2
57 Cuba 2 5 93 64 1 0 (•)0 4 5
(•)
58 Congo, People's 4 7
Rep. 7 86 707 (•) 5
84 (•)1
59 Guatemala 2 6 95 6 0 (•)1 2
49 50 178
60 Peru 20 0 6 0 2 1
61 Ecuador 0 56
64 41 0 1 0 1 1 1
45
99
62 Jamaica 50 23 2 1 0 3 3
73
14 2 1 233
63 Ivory Coast 1 5 98 0 3 (.) 59
6 3 92 87 0 (•) 0 1 2
64 Dominican Rep.
12
65 Mongolia
66 Colombia 19 3 79 0 186 (■) 2 2
56 778 1 1 2 8 16
67 Tunisia 24 66
 536 TRADE STRATEGY

EXHIBIT VIII. 2. Structure of Merchandise Exports (Continued)

Percentage Share of Merchandise Exports

Machinery
Fuels, and
Other
Primary
Minerals, and Textiles
Clothingand Transport Other
Metals Commodities Manufactures
Equipment
1960 1960 1980 1960 1980
1960 1980 1980 1960 1980
25
95
68 Costa Rica 0 1 65 0 5 0 4 5
69 Korea, Dem.
Rep.
16
70 Turkey 8 8 89 65 0 0 3 3 8
71 Syrian Arab 74
Rep. 0 81 18 2 4 0 1 17 233
72 Jordan 0 29 96 35 0 4 0 9 4
73 Paraguay 0 (•) 100 0 0 (•) 0 12
88 (•)
Upper middle-income 38 w 32 w 46 w 10 w 10 w 41
23 w 4 w 13w 22 w
56 2w
74 Korea, Rep. of 30 1 9 8 29 (•) 20 6
75 Iran, Islamic
Rep. of 88 — 9 — 0 — 0 — 3 —
76 Iraq — 3 —46 0 — 0 — 0 —
97
77 Malaysia 20 35 74 (•) 2 (•) 11 6 6
78 Panama — 24 — — 3 — (•) — 6
67
79 Lebanon
80 Algeria 12 99 1 0 1 (•) 6
81 (•)4 (•)
81 Brazil 8 11 89 0 3 (•)
18
82 Mexico 24 50
22 4 3 1 13
19
17 7
39 64
83 Portugal 8 7 37 21 18 3 17
71 27 19
1 6 95 0 2 7 344 32
14
84 Argentina
92 (•)0 48
84 Chile 59
23
4 21
23 0 (•) 1 4
86 South Africa 29 42 2 1 154 285 23
45
87 Yugoslavia 18 9 18 4 9 18 36
71
88 Uruguay — 1 61 21 16 — 4 8 18
26
89 Venezuela 74 98 0 (■) 0 (•) (•) 2
90 Greece 9
25 (•)
28 40
27
451 17 1 3 8
5 2 81
15 4 19 31
91 Hong Kong 165
92 Israel 4 2 35 8 348 2 13 61
73
93 Singapore 1 28 18 5 4 7 26 51
94 Trinidad and 14 24
Tobago 82 93 14 2 0 (•) 0 1 4 4

High-income oil
exporters — — — — —
98 w 1w
95 Libya 100 100 0 () 0 () 0 (•) 0 (•)1 w
96 Saudi Arabia 95 99 5 (•)(.)*> 0 (•) 0 0 1
97 Kuwait — — 1 — 1 — (•)3 — 6
89 (.)*>
98 United Arab
Emirates
 537 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

EXHIBIT VIII.2. Structure of Merchandise Exports (Continued)

Percentage Share of Merchandise Exports

Machinery
Fuels, Other and
Primary
Mineral! Textiles
Clothingand Transport Other
Metals>, and Commodities Manufactures
Equipment
1960 1960 1960 1960
1960 1980 1980 1980 1980 1980

Industrial market
economies 11 w 13>v 23w 15 >v 29 w 19 w
Iw 5 w 35 30 w
99 Ireland 5 3 39 6 8 4 26 18 32
31 w
41
67 20
21 8 57 7 5 2 13 41
100 Spain
8 7 19 13
101 Italy 728 11 29 33 27
102 New Zealand 7 170 3 (•) 5 3 35
(•) 979
103 United Kingdom 7 18 8 8 4 44 35 32
10 28
11 2 2 4 23 55 26
28 46
104 Japan 10
105 Austria 26 5 12 9 16 28 45
22
79 22 1 7 13 20
106 Finland 3 8 50 18 33
107 Australia 13 28 44 1 3 7 5 37
28 (•)
108 Canada 33 23 1 1 8 26 21 22
37 18
26 45
109 Netherlands 15 23 128 4 25 30
15 34
15 9 11 7 13 22
17
110 Belgium" 40
111 France 9 8 18 18 10 5 25 38
51 35
112 United States 10 9 23 3 2 35 34 13
25 28
26
27
63 38 3 5 19 24
1 13 Denmark 2 5

114 Germany, Fed. 45

Rep. 9 7 4 7 4 5 44 39 36
1 1 5 Norway 22 9 2 1 10 12 19
59 34 48
116 Sweden 10 9 29 12 1 2 40 29
32
117 Switzerland 2 5 8 4 12 6 30
31 33 50
37
East European
nonmarket
economies — — — — —
18w 33 w 34 w
3w 21 w
118 Albania
1 19 Hungary 6 9 28 25 7 7 38 21
1 20 Romania 75 32 27
121 Bulgaria 3 — — 12 — 6 — 4 —
122 Poland — 20 — 9 — 6 — 36 —26 29

123 USSR 24 — 28 — 1 — 45
21 — —
1 24 Czechoslovakia 20 7 11 9 (•) 5 50 25 29
125 German Dem.
Rep.

Source: World Bank, World Development Report, 1983, pp. 166-67.

"Indicates a weighted average.
^Indicates a value substantially less than one.
 538 TRADE STRATEGY

larger national income, the rate and level of investment will also be increased. Employment
rises with the switch to a more labor-intensive mode of production and from an expansion in
the capital stock.
The most comprehensive study is Anne O. Krueger et al., Trade and Employment in De-
veloping Countries, I. Individual Studies (1981). See also Juergen B. Donges, "A Compara-
tive Survey of Industrialization Policies in Fifteen Semi-Industrial Countries," Weltwirtschaf-
tliches Archiv (1976); C. Hsieh, "Measuring the Effects of Trade Expansion on
Employment," International Labour Review (January 1973); William G. Tyler, "Manufac-
tured Exports and Employment Creation in Developing Countries," Economic Development
and Cultural Change (January 1976); Yves Sabolo, "Industrialisation, Exports and Employ-
ment," International Labour Review (July- August 1980).
Comment
Trade policies of the MDCs are obviously of extreme importance in determining whether
the LDCs will have market access for their exports. These policies are being increasingly
determined by the industrial countries' concern with "domestic injury" from LDC imports
and the requests for "market safeguards" to reduce market penetration. In recent years, there
has been a spread of neoprotectionism in industrial countries, as evidenced by the imposition
of quantitative restrictions on imports, "voluntary export restrictions," and "orderly market-
ing agreements." In their imposition of these trade barriers, countries have bypassed the pro-
visions ofthe General Agreement on Tariffs and Trade (GATT).
A major shortcoming of the Tokyo Round of GATT trade negotiations completed in 1979
was its failure to reach agreement on a code of "market safeguards" that would bring inter-
national order to the enactment of quotas and voluntary export restrictions. See Bela Balassa,
"The Tokyo Round and the Developing Countries," Journal of World Trade Law, Vol. 14
(1980); Donald B. Keesing and Martin Wolf, Textile Quotas Against Developing Countries
(1980); G. M. Meier, "Externality Law and Market Safeguards: Applications to the GATT
Multilateral Trade Negotiations," Harvard International Law Journal (Summer 1977); G.
M. Meier, "The Tokyo Round of Multilateral Trade Negotiations and the Developing Coun-
tries," Cornell International Law Journal (Summer 1980).

VIII. B. 5. Export Promotion:

Lessons of Experience— Note

An increasing number of countries have pro- have been the Republic of Korea, Taiwan,
moted their manufactured exports and Hong Kong, Singapore, and Brazil. The re-
reached the status of newly industrializing cord of some of these "success stories" has
countries (NICs) or semi-industrialized been noted in Chapter 1 . They have achieved
countries (SICs).1 Outstanding performers the promotion of industry through an out-
'J. Bergsman, "Growth and Equity in Semi-Indus- ward-looking policy rather than through the
trialized Countries," World Bank Staff Working protection of an inward-looking industrial
Paper No. 351, 1979. Bergsman identifies 16 princi- structure. These countries exemplify the
pal SICs: Korea, Taiwan, Colombia, Turkey, Yugo- structural transformation of an open dualis-
slavia, Mexico, Israel, Egypt, Philippines, Brazil, Por- . economy. As successful export
tugal, Hong Kong, Singapore, Greece, Argentina, and * perform-
c
Spain. Marginal cases include India, Uruguay, Chile, ers, they demonstrate that this trans! orma-
Thailand, and Malaysia. tion involves a substantial rise in the share of
 539 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

manufacturing in GNP and a significant shift the sheltered and capital-intensive import-
away from dependence on primary exports substitution industries.
toward manufactured goods as a source of When government policies are biased
foreign exchange. There is considerable evi- against exports in favor of import substitu-
dence that success in promoting manufac- tion, the effective exchange rate for imports
tured exports is critical to the course of in- (EERm) is higher than the effective exchange
dustrial development.2 rate for exports (EERX). The term EERm
What are some of the lessons to be learned equals the number of units of domestic cur-
from the experience of these countries and
rency that would be paid for a dollar's worth
the manner in which they promoted the ex- of imports, taking into account tariffs, sur-
port of their manufactured goods? charges, interest on advance deposits, and
On the demand side, conditions were other measures that affect the price of im-
highly favorable during the 1950s and 1960s ports. The term EERX is defined as the num-
up to the slowing down of growth in the world ber of units of domestic currency that can be
economy after 1973. The earlier two decades obtained for a dollar's worth of exports, tak-
were unique for the high rate of growth in the ing into account export duties, subsidies, spe-
more developed countries (MDCs) — an his- cial exchange rates, input subsidies related to
torical record period — and for the growth in exports, and other financial and tax measures
world trade. The demand for imports was that affect the price of exports.
high and rising in the MDCs, and the high An outward-looking policy shift must re-
growth rate of the MDCs fostered trade lib- duce the bias against exports by changing the
eralization and weakened the case for protec- ratio of EERx/EERm from less than unity to
tion. Of course, as growth occurs in the equal to unity. This is illustrated in Figure 1,
MDCs, their pattern of demand has to match where the international price ratio changes
the actual or potential comparative advan- from PP under a foreign trade regime that is
tage of the exporting LDCs to create oppor- biased in favor of import substitution to P'P'
tunities for exports. This is likely to be more in a regime that is neutral and represents free
true for exports of manufactured goods than trade conditions. The policy objective should
primary products. be to remove the bias against exports and es-
On the supply side, however, the LDCs tablish an unbiased "free trade" regime for
varied in their international competitiveness
and in their capacity to take advantage of ex-
port markets. In large part, the differences in
performance are to be explained by differ-
ences ingovernmental policies — especially as
related to the entire set of policies that deter-
mined the foreign trade regime and the bias
between import substitution and export pro-
motion inthe different countries. To promote
exports, governmental policies have to ac-
complish an outward-oriented policy shift
that removes the bias in favor of import sub-
stitution. As long as an import-substitution
bias exists, exports will suffer from the im-
plicit taxation and quantitative restrictions,
from the cost of inputs being above free trade
prices, and from the diversion of resources to Exportable

FIGURE 1. Definition of import-substituting and

2See Hollis B. Chenery, "Interactions between In- export promoting strategies. (Source: Adapted from
dustrialization and Exports," American Economic J. Bhagwati, Anatomy and Consequences of Exchange
Review, May 1980, pp. 281-7. Control Regimes, 1978, p. 208.)
 540 TRADE STRATEGY

exports. If the regime were to raise the ratio rate.3 In 1975, for instance, Korea's official
of EERx/EERm above unity (P*P*), it would exchange rate was 485 won per dollar, but
be oversubsidizing exports and protecting when export subsidies (internal tax exemp-
exports instead of promoting exports by sim- tions, customs duty exemptions, and interest
ply removing the bias against exports and rate subsidies) were added, the EERX be-
establishing a neutral trade regime. came 566 per dollar.
There is evidence that oversubsidization of It is not sufficient, however, simply to graft
exports is less likely to occur than oversubsi- export subsidies onto the preexisting import
dization ofimport substitution (see selection structure. If protection against imports re-
VIII.B.l). Although an outward-oriented in- mains high, the incentive to sell domestically
dustrial development strategy does not mean is likely to remain greater than the incentive
favoring exports over import substitution, it to export. Those policies that bring the ratio
does mean that similar incentives are pro- of EERm and EERX closer to unity charac-
vided to production for domestic and for ex- terize what Bhagwati and Krueger term
port markets. The creation of such a free Phases III and IV in their typology of ex-
trade regime for exports provides incentives change control regimes. (See selection
VIILB.2.)
that improve export performance. A deval-
uation of the country's overvalued exchange Much of the success of Korea, Singapore,
rate will stimulate the demand for the coun- and Taiwan has been written in terms of
try's exports. If, at the same time, tariffs and Phases III and IV policy reforms. In these
quantitative restrictions are removed from countries, a free trade regime was applied to
imports, the devaluation will be compensated exports. Exporters were free to choose be-
on the import side. tween domestic and imported inputs; they
If the EERm is initially greater than EERX, were exempted from indirect taxes on their
the exchange rates can be realigned by rais- output and inputs; and they paid no duty on
ing their EERX. This can be done through a imported inputs. The same privileges were
number of policy measures that increase extended to the producers of domestic inputs
gross export receipts, reduce the cost of ex- used in export production. These incentives
ports, or reduce profit tax liability through ensured that in the manufacturing sector, on
such measures as export exchange rates, sub- average, exports and import substitution re-
sidies to export value, tax and duty conces- ceived similar treatment. Nor was there dis-
sions, foreign exchange retention schemes, crimination against primary exports and pri-
and preferential credits.
mary production in general.4
In Korea, for example, various export pro- Experience also shows that these export in-
motion schemes have been followed: tariff ex- centives will not be effective unless the poli-
emptions on imports of raw materials and cies can be expected to last for the duration
spare parts, tariff and tax exemptions of investments in the export sector. A long-
granted to domestic suppliers of exporting term commitment to these policies is essen-
firms, domestic indirect tax exemptions, re- tial; the political and economic risks can be
duction indirect taxes on income earned in reduced only if there is policy stability.
exporting, accelerated depreciation, import
entitlement linked to exports, reduced taxes 3 Anne O. Krueger, The Developmental Role of the
on public utilities, monopoly rights granted in Foreign Sector and Aid: Studies in the Moderniza-
tion of The Republic of Korea, 1945-1975, 1979,
new export markets, direct export subsidies,
chapters 4 and 5; Larry E. Westphal, "The Republic
credit subsidies through lower interest rates, of Korea's Experience with Industrial Development,"
and preferential access to loans. When these World Development, 1978.
incentives are included in the computation of 4For a comparison with other countries, see Bela
Balassa, The Process of Industrial Development and
effective exchange rates for exports, the ef- Alternative Development Strategies, Princeton Es-
fective exchange rate becomes considerably says in International Finance No. 141, December
higher than the nominal or official exchange 1980, pp. 12-16.
 541 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

Four principles of an outward-oriented de- long-run equilibrium level and to adjust tar-
velopment strategy have been summarized by iffs and export subsidies so that domestic rel-
Professor Balassa:
ative prices remain the same.6 Quotas should
be replaced by tariffs to allow price
The experience of developing countries in the incentives.
postwar period leads to certain policy prescrip-
This is the first step of rationalization. The
tions. First, while infant-industry considerations
call for the preferential treatment of manufactur- next step should be to reduce the variance in
ing activities, such treatment should be applied on effective rates of protection and to reduce
a moderate scale, both to avoid the establishment home-market bias — including additional lib-
and maintenance of inefficient industries and to eralization, variance reduction, and bias re-
ensure the continued expansion of primary produc- duction. But the transition period to the lib-
tion for domestic and foreign markets. eralized trade regime will take a number of
Second, equal treatment should be given to ex- years, and a successful transition requires a
ports and to import substitution in the manufac- whole sequence of changes over several years.
turing sector, in order to ensure resource allocation
During this transition period, the government
according to comparative advantage and the ex-
must have the determination and political
ploitation ofeconomies of scale. This is of partic-
ular importance in the case of intermediate goods support to remove existing distortions and
and producer and consumer durables, where the achieve the reallocation of resources in the
advantages of large plant size and horizontal and face of resistance by affected interest groups.
vertical specialization are considerable and where Political stability and commitment are essen-
import substitution in the framework of small do- tial to provide strong signals and strong in-
mestic markets makes the subsequent development
centives to exporters. Furthermore, the coun-
of exports difficult. The provision of equal incen- try must have some safety net that provides
tives will contribute to efficient exportation and
import substitution through specialization in par- foreign exchange to support import liberali-
ticular products and in their parts, components, zation before the export promotion becomes
and accessories. sufficient to earn foreign exchange. An inflow
Third, infant-industry considerations apart, of foreign capital, or the ability to utilize for-
variation in incentive rates within the manufactur- eign reserves, has been of critical importance
ing sector should be kept to a minimum. This in cases of successful liberalization. From a
amounts to the application of the "market princi- study of several countries, Krueger con-
ple" in allowing firms to decide on the activities to cluded that "In many of these instances, im-
be undertaken. In particular, firms should be free port liberalization could not have begun
to choose their export composition in response to
without receipt of new lines of credit. Of
changing world market conditions.
Fourth, in order to minimize uncertainty for the those liberalizations that were possible be-
firm, the system of incentives should be stable and cause of credits, many terminated when the
automatic. Uncertainty will also be reduced if the credits were exhausted."7 If import liberali-
reform of the system of incentives necessary to zation cannot be supported by a capital in-
apply the principles just described is carried out flow, then the redesign of import protection
according to a timetable made public in advance.5 will have to be delayed until export growth
The transition to a liberalized unbiased
trade regime has, however, proved difficult 6See I. M. D. Little, T. Scitovsky, and M. Scott,
for a number of countries. Economists have Industry and Trade in Some Developing Countries,
proposed that a country start with a compen- 1970, chapter 10; B. Balassa, Policy Reform in De-
sated devaluation — that is, a devaluation veloping Countries, 1977, pp. 25-27; A. O. Krueger,
Synthesis, National Bureau of Economic Research,
coupled with the removal of quotas and tar- Vol. 9. For a good discussion of the processes and
iffs in order to bring the exchange rate to its problems in several cases of transition, see Donald B.
Keesing, "Trade Policy for Developing Countries,"
5Bela Balassa, The Process of Industrial Develop- World Bank Staff Working Paper No. 353, August
ment and Alternative Development Strategies, 1979, pp. 220-52.
Princeton, p. 24. 7Krueger, chapter 8, p. 67.
 542 TRADE STRATEGY

has begun and the balance-of-payments sit- petitive if the efficiency of labor is so low that
uation has eased. the wage costs per unit of output are actually
Once a country succeeds in shifting to an high. Product competitiveness will also de-
outward-oriented policy, it must continue to pend on such nonprice determinants of export
follow monetary and fiscal policies that will success as design, quality control, punctuality
avoid inflation. For if inflation were to ensue, in delivery, and service. A close relationship
the domestic currency would again become with Western trading companies has often
overvalued in terms of foreign currency, and helped to improve models and design, and to
tariffs and quantitative restrictions would be overcome marketing problems.
reimposed. A combination of protection for Furthermore, the supply of entrepreneur-
domestic sale of final goods and a declining ship is obviously important in improving both
real effective exchange rate or the purchas- cost competitiveness and product competi-
ing-power-parity-adjusted ef ective exchange tiveness. There has not, however, been a uni-
rate (say, EERX deflated by general con- form need for foreign entrepreneurship or the
sumer prices) will make it less profitable, rel- resources of the multinational enterprise. In
atively and absolutely, to export than to sell Korea and Taiwan, for example, only a small
domestically. percentage of the exports is produced by
At the micro level, the competitiveness of firms with any foreign equity participation.
export firms will depend on their cost com- Nor has it always been necessary to first pro-
petitiveness and product or nonprice compet- duce for the home demand before being able
itiveness.8 Cost competitiveness is deter- to export. For a number of countries export-
mined by factor productivity, factor prices, ing manufactures, the new export industries
factor proportions, and location relative to have been quite different from their old im-
export markets. Especially significant is the port-substitution industries.
relationship between wages and labor pro- What can we expect in future? Can the
ductivity for the labor-intensive exports that success stories continue their success, and
tend to dominate in the early period of export can other nations also become newly indus-
promotion. Although money wages in the trializing countries via an outward-oriented
LDC may be low, exports will not be com- industrial policy? The crucial issue now is: If
the MDCs in the future do not exhibit the
rapid growth of the 1950s and 1960s, can the
8For an illustration of this in Colombia, see David experience of the successful export perform-
Morawetz, Why the NewEmperor's ers be repeated for other LDCs? The next se-
Made in Colombia, York, New
1981.Clothes Are Not lection considers this issue.

Comment
Successful as the newly industrializing countries were in promoting exports of manufac-
tures during the past two decades, there is now a resurgence of export pessimism. Can the
East Asian model of export-oriented development be generalized to other developing coun-
tries? Can developing country exports keep growing? Will slow growth in the MDCs and the
spread of protectionism limit imports from the LDCs?
A number of studies examine these questions. Some statistical projections are offered by
Bela Balassa, "Prospects for Trade in Manufactured Goods between Industrial and Devel-
oping Countries, 1978-1990," Journal of Policy Modeling, Vol. 2, No. 3 (1980). Another
simulation exercise concludes that generalization of the East Asian model of export-led de-
velopment across all developing countries would result in untenable market penetration into
industrial countries, but the author, William P.. Cline, concludes that
the findings of the study should not be interpreted to favor a closed-economy strategy of discouragement
of exports. On the contrary, in many developing countries trade liberalization still-has far to go. Several
 543 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

EXHIBIT VIII. 3. Exports of Manufactures from Developing Countries

All Low-Income
Developing Developing
Middle-Income
Developing
Countries Countries Countries

Value of exports, 1980 (1978 billion U.S. dollars) 97.4 6.0 91.4
Annual growth of exports,
12.9 6.9
1970-80 (%) 13.4
Destination, 1977 (%) n.a. 100.0 100.0
Industrial countries n.a. 51.0 58.0
Developing countries n.a. 27.0 30.0
n.a.
Centrally planned economies 12.0 6.0
10.0 6.0
Capital-surplus oil exporters n.a.

Source: World Development Report, 1981, New York, Oxford University Press, 1981.
Note: n.a. — not available

Latin American countries in particular could raise their manufactured exports substantially without even
reaching cross-country norms, let alone export dimensions of the East Asian variety, and one strongly
suspects that in many such cases liberalization of the trading and exchange rate regimes would lead to
more rapid export growth.

See William R. Cline, "Can the East Asian Model of Development Be Generalized?" World
Development, Vol. 10, No. 2 (1982).

EXHIBIT VIII.4. Growth of Industrial Output, Employment, and Exports in Selected Countries
Thailand Malaysia Korea Singapore

Average annual growth (%)

Industrial output
11.6 — 17.2
1960-70 12.5
1970-79 9.9 16.5 8.6
10.4
Industrial employment 5.0
1960-70 5.3 11.2 4.5
1970-77 7.7 5.7
6.9
Merchandise exports 11.0
12.0 6.5
1970-79 25.7
Percentage of manufactures in merchandise exports
1960 2 6 14 46
26
1978 25 12.1° 21
industry 899.3
Percentage of labor force in
1960 23.0
4.4 11.7
1970 6.0 20.1 30.3
1977 27.5 33.2

Source: World Development Report, 1981; ILO Yearbook of Labour Statistics, 1978; and ILO Labour Force Estimates and
Projections, 1970-2000, Geneva, 1977.
"1970-76.
14.3C 14.6*
*World Bank estimates.
c1976. 20.1*
 544 TRADE STRATEGY

EXHIBIT VIII. 5. Exports, Imports, Manufactured Output, and GNP in Selected Developing
Countries (percent)

00

Share of manufactured exports in manufactured output

2.6 10.8
1960 0.8 0.4 3.0 0.7 9.7 7.9 0.9 11.2 8.6
1966 0.9 1.3 4.1 3.0 9.4 12.8 13.9 2.9 20.1 19.2 13.8
42.6 49.9 16.9
1973 3.6 4.4 2.5 7.5 8.6 14.1 40.5 4.4

Incremental ratio of manufactured exports to manufactured output

1960-66 1.0 3.6 5.5 7.7 8.9 23.9 24.8 28.4 24.8 15.8
3.2 47.5 19.5
1966-73 6.5 5.6 0.0 11.4 7.7 14.9 45.7 5.5 56.4

Share of manufactured imports in total utilization of manufactured goods

1960 14.6 10.8 26.3 30.8 19.3 28.5 24.4 19.6 28.5 22.0
56.2 17.3
1966 6.3 7.5 21.6 28.0 16.5 32.8 26.5 16.2 53.2 29.3
24.0
1973 5.4 13.0 17.5 21.5 9.5 41.2 35.9 15.2 64.3 38.9

Incremental ratio of manufactured imports to utilization of manufactured goods

49.2 13.6
1960-66 -3.9 -3.0 14.1 20.5 10.4 42.5 31.9 11.7 30.5
1966-73 4.4 15.7 10.8 14.2 -0.4 45.1 40.4 14.4 67.0 42.2 29.4

Ratio of total exports to GNP 1.5

1960 8.9 6.1 12.6 11.3 4.2 9.9 9.5 22.4
8.4 6.4 17.1
1966 7.3 7.1 15.7 9.5 4.2 12.8 6.5 5.4
4.3 26.6 14.2
4.3 15.5 47.8
1973 8.1 9.8 7.6 11.8 26.1 44.6 14.5

Incremental ratio of total exports to GNP

1960-66 5.3 12.3 23.0 3.4 4.1 20.4 13.0 4.3 52.0 24.7
4.3 10.7
1966-73 9.0 11.5 3.3 14.5 17.4 34.8 3.3 52.0 63.3 14.8

Ratio of total imports to GNP 7.5 9.0

1960 10.3 7.1 12.9 12.6 20.1 16.0 18.9 32.8
65.4
1966 5.2 6.1 13.5 12.6 7.4 21.2 7.2
8.6 62.5 19.9 18.3
18.7 40.5 22.0
1973 5.5 11.1 6.8 10.6 4.5 30.8 34.3 91.5

Incremental ratio of imports to GNP 7.3 12.3

1960-66 1.3 0.9 14.9 12.4 23.0 22.2 20.8
41.3 5.1 57.9
1966-73 6.0 14.2 2.4 8.4 0.3 37.3 9.7 103.6 50.8 26.4

Scarce: Bella Balassa, "Export Incentives and Export Performance in Developing Countries: A Comparative Analysis," Welt-
wirtschaftliches Archiv 1 14 (1978).
Note: In the absence of manufactured output figures for 1960, these have been derived from the 1966 figures by utilizing
growth rates of value added in manufacturing and inflation in prices of manufactured goods for the period 1960-66. The same
method has been applied in cases where 1973 manufactured output figures are not available.
 cd >n oo O O

sal
/— v <N On O >/-> ~H © "^ — H — H OO
c
u S H w
o
'C

S3
^— '
o.
00 oo
I — ° "5
is a /i^m

r^
13ed
ON 0>

C
ca
r-
r- s
T3 -v
\3 O co m r» O m
o
*>
c
«J

c/a
so
"c3

ON o
3
o
o 2
CO 0*
o «J"
1o ££ss
*> 53 tri

c 3 3 e- *>
o T3
o
Ui 1—1
C

>>
-O
(A

u
o
D.
X
w vo no © ~- r-~
"fi

V
^<rt
22
c
«d
.e
o
'C

s
O
Cm

o
52 1 •8 | •c
c O <* g 5
o
S 8 ■o
3 oo
0 o & ■y -a
-a CA c 3 j?

XJ <L>
00 eo <u 3 * 5 s
3a. c
s o 5 § £
2 -2 5 e
O 3
o
S S .2 V-

o c5 «« *> g
3
O .M <■<■SE x>ac
'lo
e3 a
E 4J
60 "O

o U o 15 i UJ
0 E
o 5 ■§
c .2o- '^c £ •£ =1
"o :e
< <

31 JLJ§
3
c
3
§•§
« _U 5sd 5'5U
O rt
Ud CO nJ
1 Z6
U on <

«e

545
 546 TRADE STRATEGY

VIII. B. 6. The Slowing Down of

the Engine of Growth *

Let me begin by stating my problem. For the with development theory. The collapse of in-
past hundred years the rate of growth of out- ternational trade in the 1930's had seemed ir-
put in the developing world has depended on reversible, so much so that Keynes had even
the rate of growth of output in the developed declared that we didn't need much of it any-
world. When the developed grow fast, the de- way. So in the 1940's and 1950's we created
veloping grow fast, and when the developed a whole set of theories which make sense if
slow down, the developing slow down. Is this world trade is stagnant — balanced growth,
linkage inevitable? More specifically, the regional integration, the two-gap model,
world has just gone through two decades of structural inflation — but which have little
unprecedented growth, with world trade relevance in a world where trade is growing
growing twice as fast as ever before, at about at 8 percent per annum. Also many countries,
8 percent per annum in real terms, compared basing their policies on the same assumption,
with 0.9 percent between 1913 and 1939, and oriented inwards mainly towards import sub-
less than 4 percent per annum between 1873 stitution. The fact that world trade was grow-
and 1913. During these prosperous decades, ing rapidly was not universally recognized
the less developed countries (LDCs) have until the second half of the 1960's. Then
demonstrated their capacity to increase their nearly every country discovered the virtues of
total output at 6 percent per annum, and exporting. Now we are in danger of being
have indeed adopted 6 percent as the mini- caught out again. Since 1973 the growth rate
mum average target for LDCs as a whole. of world trade has halved, and nobody knows
But what is to happen if the more developed whether this is temporary or permanent. But
countries (MDCs) return to their former most of our economic writing continues to as-
growth rates, and raise their trade at only 4 sume implicitly that a return to 8 percent is
percent per annum: Is it inevitable that the only just around the corner.
growth of the LDCs will also fall significantly Let me come back to the relationship be-
below their target? My purpose is not to pre- tween MDCs and LDCs. The principal link
dict what is going to happen, but to explore through which the former control the growth
existing relationships and how they may rate of the latter is trade. As MDCs grow
change. faster, the rate of growth of their imports ac-
The extraordinary growth rates of the two celerates and LDCs export more. We can
decades before 1973 surprised everybody. measure this link. The growth rate of world
We knew that the world economy experi- trade in primary products over the period
ences long swings in activity; that world 1873 to 1913 was 0.87 times the growth rate
trade, for example, grew faster between 1830 of industrial production in the developed
and 1873 than it grew between 1873 and countries; and just about the same relation-
1913, that is to say, between 4 and 5 percent ship, about 0.87, also ruled in the two dec-
before 1873, compared with between 3 and 4 ades to 1973.1 World trade in primary prod-
percent after 1873. But a jump to 8 percent ucts is a wider concept than exports from
was inconceivable. . . .
developing countries, but the two are suffi-
The fast pace of world trade played havoc ciently closely related for it to serve as a
proxy. We need no elaborate statistical proof
*From W. Arthur Lewis, "The Slowing Down of
that trade depends on prosperity in the in-
the Engine of Growth," American Economic Review,
September 1980, pp. 555-7, 559-64. Reprinted by 'For data, see my Growth and Fluctuations 1870-
permission. 1913, pp. 175-76.
 547 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

dustrial countries. More interesting is the ev- total output? One might perhaps conceive a
idence that the relationship was quantita- lower rate of growth of MDCs. Many MDC
tively the same over a hundred years, so that voices are calling for this — the environmen-
the two-thirds increase in the rate of growth talists, the persons who fear exhaustion of ex-
of exports of primary products from LDCs haustible resources, the advocates of greater
was no more or less than could be predicted grace and leisure in our lives, and others. But
from the increased rate of growth of MDC if the MDC growth rate falls, the LDC
production. growth rate will fall too, and LDCs will get
Most interesting is that the coefficient is the worst of it, since the terms of trade will
less than one, viz. 0.87. This means that if the move against them. Given the link, it is in the
engines of growth were industrial production interest of LDCs that the MDCs should grow
in MDCs and exports of primary products in as rapidly as they can.
LDCs, then the MDC engine was beating Three questions come to mind. First, what
slightly faster than the LDC engine. The ef- is likely to happen to the speed of MDC
fects of equal beating would not necessarily growth? Secondly, can LDCs maintain fast
be exactly the same. And there are side ef- growth even in face of a decline of MDC
fects that strengthen the connection. When growth? And thirdly, is it desirable that LDC
the beat is faster the terms of trade are ex- development depend on rapid growth of ex-
pected to be more favorable to the LDCs ports to MDCs? . . .
(though that did not happen this time). The In what follows I shall assume that indus-
domestic market prospers, so LDC industrial- trial production in MDCs grows more slowly
ization for the domestic market is speeded; than it was growing before 1973, and that the
this happened. The MDCs relax their bar- imports of these countries grow only at 4 per-
riers to imports of manufactures, so this trade cent a year, over the next twenty years. This
accelerates as well. Foreign capital flows into is not a prediction; it is merely the assump-
minerals, manufactures, and infrastructure. tion whose consequences we are seeking to
And foreign countries take more migrants, so analyze.
that the homeward flow of remittances to I shall also assume that LDCs want their
LDCs is larger in prosperous times. GDP to grow at 6 percent per annum, and
Putting it all together, including the fact that this requires thier imports to grow at 6
that industrial production grew faster in percent. This linkage follows from the fur-
LDCs than in MDCs, it is not surprising that ther assumption that the individual LDC will
the rate of growth of gross domestic product not become more self-sufficient, perhaps be-
was just about the same in LDCs and in cause itis too small; though LDCs as a group
MDCs over the quarter of a century ending will have to be more self-sufficient. No im-
in 1973, namely about 5 percent per annum. portance attaches to whether the figure for
Since LDC population was growing faster imports is the same as the figure for the
than MDC population, there is a big gap in growth of gross domestic product; all that
the growth rates of output per head, about 4 matters is that the growth rate of exports
percent in MDCs, against 2.5 percent in from LDCs is assumed to be significantly
LDCs. The performance of LDCs was re- higher than the rate of growth of imports of
markable inabsolute terms, but the gap be- LDC commodities into MDCs. The LDCs
tween MDCs and LDCs in income per head will continue to pay for some of their imports
continued to widen rapidly. out of proceeds of transfers, including foreign
Now we come to our dilemma. The objec- aid and private foreign investment, but we
tive of most people who are concerned with shall assume that this still leaves LDCs need-
these matters is to narrow the per capita gap ing, say, a 6 percent growth rate for exports,
between MDCs and LDCs. But how is one to while MDCs are assumed to increase their
do this if they are linked to equal growth of imports from LDCs only at 4 percent a year.
 548 TRADE STRATEGY

The problem is how to reconcile these two standpoint of the individual LDC it matters
growth rates. not at all what the MDC growth rate may be.
There could theoretically be a simple way, Given resources and flexibility, it can always
namely for LDCs to have an ever increasing sell more to MDCs. However, it thereby dis-
share of MDC imports, but we have closed places some other LDCs trade. What one
this door. The main link between MDC and can do cannot be done by all.
LDC economies has been the MDC demand At the level of arithmetic this problem now
for LDC primary commodities. This has been has only one solution. If total sales from
a link in terms of physical volume, not much LDCs increase at 6 percent, while sales to
affected by prices. The LDCs could not sell MDCs increase at 4 percent, sales to the rest
significantly more primary produce by reduc- of the world (given weights of seven to three)
ing prices; on the contrary, they would earn must increase initially at about 11 percent
substantially less purchasing power as the per annum. Ignoring the socialist countries,
terms of trade deteriorated. The LDCs could which could help by buying much more from
earn more by reducing the volume or by join- LDCs but won't, the LDCs can solve the
ing together in raising prices. The direct ef- problem only by accelerating sharply their
fect of these actions would be to reduce out- trade with each other.
put, but this could be offset by investing the Inter- LDC trade is still rather small —
extra earnings judiciously. However none of about 19 percent of the exports of non-OPEC
this seems to be in the cards; so we shall as- LDCs.1 The percentage did not change sig-
sume that our problem cannot be solved by nificantly over the two decades to 1973 de-
accelerated or decelerated production of pri- spite all the effort that was put into creating
mary commodities normally exported to and servicing regional trade institutions. Can
MDCs. this trade take up the slack left by MDCs as
What about manufactures? These are now MDCs slow down?
nearly 40 percent of the exports of the non- The answer is in the affirmative. Currently
OPEC LDCs, and are still their fastest grow- the LDCs depend on the MDCs for food, fer-
ing export. Could the whole problem be tilizers, cement, steel, and machinery. Taken
solved simply by increasing the growth rate as a group, LDCs could quickly end their de-
of manufactured exports to MDCs, in substi- pendence for the first four, and gradually
tution for primary products? I shall assume throw off their dependence for machinery.
that this cannot be done, since if it can be They also import a considerable quantity of
done my paper ends abruptly. Also I do not light manufactures for which they are not in
think that it can be done. The MDCs are will- any sense dependent (some $31 billion in
ing to let in manufactured exports when they 1977, compared with $47 billion of engineer-
are prosperous, since they then have many ing products). They could quickly rid them-
growing industries that can take in people selves ofthese, and more gradually throw off
displaced by imports. Our assumption that their dependence for machinery.
the MDC growth rate is low rules out this The LDCs are capable of feeding them-
possibility. It would indeed be more appro- selves now, if they adopt appropriate agrar-
priate to assume that MDCs will take less ian policies and, as our eleven new interna-
manufactures from LDCs rather than more. tional tropical agricultural research institutes
Our basic assumptions therefore are that give us better varieties and improved tech-
LDCs need to have their exports grow at 6 nology, output should more than keep up
percent a year, but MDCs will increase their with population. The problem is to get
imports from LDCs only at 4 percent a year. through the period while the birth rate re-
What is to happen to the growth of LDC
output? 2Trade data in this and the next two paragraphs are
Let me concede at once that from the from GATT, International Trade 1977-78.
 549 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

mains obstinately high to the less-frightening ity translate into an effective economic
times when the birth rate will have dropped framework?
below twenty per thousand. It may be a near One way would be to follow the customs
thing, but we should make it. union route, with LDCs giving preferential
As for fertilizers, cement, and steel, these treatment to imports from other LDCs. The
are made by applying standard technology to nucleus of this exists already in the Protocol
raw materials that are widely available out- Relating to Trade Negotiations among De-
side the MDCs. Machinery is more bother- veloping Countries which came into force in
some because important parts of this trade 1973, with the blessing of GATT, and which
involve economies of scale, continually im- provides for negotiated preferential arrange-
proving technology, and patented or secret ments among sixteen of the bigger and more
knowledge. However, several LDCs are mov- advanced LDCs. The philosophy of such an
ing into this field, and already machinery is arrangement is in line with the spirit of Bret-
1 5 percent or more of the output of manufac- ton Woods, which recognized the rights of
tures in at least eight LDCs (India, Brazil, countries to impose restrictions on other
Singapore, Chile, Korea, Argentina, Mexico, countries which were tending persistently to
and Israel)3 The LDC exports of engineering run balance-of-payments surpluses, as would
products are also growing rapidly, and con- be the situation if the LDCs were growing
trary to popular belief, already exceed LDC faster than the MDCs. One may doubt how-
exports of textiles and clothing in value. ever whether such different countries will get
There is no reason why LDCs as a group very far along the route of preferential
should not become nearly self-sufficient in concessions. If they are to prefer each other's
standard types of equipment. . . . goods, this will have to be because they are
It follows, therefore, that in the situation competitive in price with those of MDCs.
we are analyzing, where world trade deceler- In the economist's model this competitive-
ates, customs unions would be more highly ness would come about automatically. The
prized and would be made more effective, es- LDCs would run a balance-of-payments def-
pecially in regard to large-scale industries icit because of the MDC slowdown, yet per-
with region-wide economies of scale. But sist in their own rapid growth instead of slow-
even so, the leading commodities that LDCs ing down themselves. Adjustment comes in
would now have to produce to a greater ex- the old gold standard version through an out-
tent for each other cannot be shared out be- flow of gold that reduces their price levels; or
tween next door neighbors on a political in the modern versions by devaluation, which
basis. Food, fertilizers, cement, steel, and has the same effect. The real world is more
fule pick their locations more in terms of raw complicated. Inflation is universal, but ag-
material availability, and machinery will gressive sellers of manufactures have to keep
come in the first instance from those LDCs their prices down; so this set of LDCs would
that already have a substaintal industrial need special emphasis on inflation controls.
base. This new LDC trade would be world- Devaluation cannot be avoided when prices
wide, just as European and U.S. trade are cease tu be competitive, but is palliative
worldwide. rather than curative in situations when it
I am therefore arguing that it is physically triggers further increases in domestic costs
feasible for LDCs to maintain a high rate that reinstate the differential that it was
of growth even if MDCs decide otherwise meant to eliminate. The LDCs have the same
for themselves. How does physical feasibil- problem as the MDCs: the domestic price
level can no longer be controlled merely by
twirling general controls such as the rate of
3Share of machinery in manufactures calculated interest, or the supply of money, or the rate
from United Nations, Yearbook of Industrial Statis-
tics 1976. of exchange for the currency. They too now
 550 TRADE STRATEGY

experience the cost-push element in price de- termining the rate of growth of LDC produc-
termination for which the only remedy is tion, itwill be the growth of LDC production
some sort of incomes policy. We expect more that determines LDC trade, and internal
from the economic system than our grand- forces that will determine the rate of growth
parents did in the nineteenth century, by way of production. Not many countries are ready
of full employment and faster growth, and to make this switch. India is an obvious pos-
should not be surprised that the economic sibility, along with some of the other sub-
system requires more from us by way of sup- scribers tothe Protocol of 1973. It is not pos-
porting institutions. sible for all LDCs to make this switch and
These new aggressive LDCs, exporting ma- neither is it necessary, for if leading LDCs
chinery to each other, may also have prob- grow fast and import heavily, they will sub-
lems in financing their trade. Nearly every stitute to some extent for the former rapid
LDC has a separate currency. We are envi- growth of MDCs. For those who use the lan-
sioning Nigeria selling cereals to India for ru- guage of center and periphery, this means
pees, with which it buys machinery from Bra- that a number of countries leave the periph-
zil. Some kind of clearing agreement may ery and join the center. Or if they are spe-
become necessary; otherwise LDC traders cially linked to each other by preferential
will tend to do business with each other in one trade and currency arrangements, one may
or more MDC currencies, and will be con- even speak of the creation of a new center
strained bythe relative scarcity of such cur- consisting of former peripheral nations that
rencies. Perhaps the International Monetary have built a new engine of growth together.
Fund would straighten this out. A more se- The shadow on this picture is what hap-
rious problem will be to finance the export of pens to those LDCs whose best option has
capital goods from one LDC to another, since been to export raw materials to MDCs. Our
the seller is expected to help finance the exercise starts from the assumption that the
buyer. It is not likely that the LDC exporters growth rate of MDC demand is reduced, so
can do this on their own. We must assume these face surpluses and unfavorable terms of
that they will be allowed to raise untied loans trade. We have provided an escape for LDCs
in the MDC financial markets, perhaps using that can turn to exporting food or manufac-
the regional development banks as interme- tures, but we have not assumed that the new
diaries to a greater extent than is now the core LDCs will substitute for the MDCs by
case. drinking more coffee or tea, or using more
But the real problem is not whether LDCs rubber and jute. This solution therefore in-
can become competitive and hold their own volves some hardships for the less adaptable
in each other's markets. Problems of pricing LDCs, constrained by climate or by the small
and foreign exchange can work themselves size of their markets. A framework for help-
out in the world market. The real problem is
ing them exists already in the IMF's compen-
whether LDCs will persist in rapid growth
despite the slowdown of the MDCs. If the satory financing, and in the EEC's STABEX
support; but these are meant for temporary
economy is still dependent, the balance-of- fluctuations. Bigger and more persistent sup-
payments weakness will pull it down; but if it port would be required.
has attained self-sustaining growth, the Transnational corporations would proba-
weakness in the foreign exchanges merely bly play some part in the establishment of
launches a drive to export to other LDCs, and this new inter- LDC trading network. The
the weakness in the balance of payments is cadre of domestic entrepreneurs is adequate
then only transitional. in the more advanced industrial LDCs to
If a sufficient number of LDCs reach self- manage most of the range of light consumer
sustaining growth, we are into a new world. goods and light engineering products. One of
For this will mean that instead of trade de- our main concerns, however, is to diminish
 551 INWARD- VERSUS OUTWARD-ORIENTED DEVELOPMENT STRATEGIES

reliance on MDCs for heavy machinery and neurial steam in the leading LDCs to make
such, and this extends into fields where ex- this transition without a supporting frame-
perience islimited. Since we are assuming work. have
I already mentioned the main in-
that the market will send out price signals fa- ternational elements of such a framework,
voring production in LDCs, whether because namely preferential tariff and currency ar-
of tariffs or of currency adjustments, trans- rangements. The domestic element consists
national corporations will be eager to pre- of the maintenance of home demand in the
serve their markets by establishing subsidi- face of stagnant world trade in primary prod-
aries behind the protective barrier. Hostility ucts, so that the economy can continue to go
to such corporations is universal and their in- forward instead of collapsing. Much of the
fluence isdiminishing in most sectors, espe- responsibility for maintaining momentum
cially inmining, public utilities, distribution, then falls on the government, given its large
and finance; but not in manufacturing, where share of the cash economy, and also the ex-
judging by advertisements in the New York tent to which it regulates or supports the pri-
Times and the financial press, LDC govern- vate sector. It has to carry the responsibility
ments are only too anxious to invite the par- for a large investment program (private and
ticipation of transnational corporations. public) in human and physical capital. This
There are plenty of restrictions — on the hir- responsibility could not be carried without
ing of expatriate staff, on percentages of eq- external aid. The MDCs would have to be in
uity owned by foreigners, on borrowing in the a mood to say: we will not give you more
local market, on technology, and so on. Also, trade; here for a while is more aid instead.
in many cases joint ownership with local cap- The recession that started in 1974 has now
italists or with government agencies is pre- lasted long enough for LDCs to consider the
scribed. A government anxious to promote possibility that MDCs intend to maintain
industrialization and a corporation anxious to rates of GNP growth which will allow world
preserve or extend its market find common trade to expand only at around 4 percent a
ground. year. This would be a major blow to LDC
The awkward part of the exercise is to sus- growth aspirations, unless new steps were
tain the momentum of 6 percent growth taken to support rising participation of LDCs
through the transition from dependence on in LDC trade. I have been trying to analyze
MDC trade to dependence on LDC markets. what these steps might be. They ought to fig-
During this transition the leading industrial ure prominently in current North-South ne-
LDCs must establish their footholds in each gotiations, but in fact they do not, since these
others' markets, as well as those of other negotiations tactily assume that high MDC
LDCs; also the agrarian changes must occur growth rates will soon be resumed. Perhaps
which both feed the urban population and this assumption will turn out to be correct;
present a growing market for its goods and economics does not fortell the future. For the
services. It is possible that some of the lead- least, LDCs should be discussing among
ing LDCs can take this in their stride, just as themselves in what directions they would
German industrialists launched their trade wish to go, prior to negotiations with the
MDCs.
drive in the 1880's, followed by the United
States after 1895, by Japan in the 1930's, Of course, the problems tackled in this
and more recently by Brazil. They do not paper would not arise at all if MDCs were
have to begin with machinery, since LDCs willing to allow LDCs a greater share of
still import so much light manufactures from MDC markets. This would be the logical evo-
MDCs. They can start here and move more lution ofa situation where LDCs grow faster
gradually into machinery. than MDCs; trade with LDCs should become
At the other extreme, it is also possible an ever-increasing portion of MDC trade. We
that there is simply not yet enough entrepre- live in a strange world. Through the 1960's
 552 TRADE STRATEGY

and 1970's, MDCs have been dismantling therefore by their prosperity help a little to
their barriers to each other's trade while in- sustain OECD prosperity. It can hardly be an
creasing their barriers to LDC trade. Since OECD interest to force the LDCs into dis-
imports of manufactures from developing criminating against OECD sources,
countries are only two percent of the con- Nor would these problems arise if the
sumption of manufactures in OECD coun- MDCs would return to the attack on mass
tries, this indicates exceptional sensitivity to poverty within their own borders which they
minor change. Lack of sensitivity, on the launched so successfully in the 1950's and
other hand, characterizes the failure of de- 1960's, and have now abandoned; since what
veloped countries to recognize that depen- we all really need is that world trade recap-
dence is mutual, in that the non-OPEC LDCs ture its growth rate of 8 percent a year. But
take 20 percent of OECD exports, and could that is a different story.
 VIII.C. TRADE IN PRIMARY
PRODUCTS
VIII C. 1 . Techniques of Price
Raising*

The traditional approach to price raising has amenable to regulation designed to prevent
been to negotiate an agreement to control the the accumulation of excessive stocks, either
quantity of a commodity coming on the world in the hands of the marketing agency or in
market. Export restriction schemes can be ef- the producing countries themselves. In the
fective inthe short run, but they also create case of an agricultural crop, if the marketing
their own problems, particularly if they are agency is obliged to purchase the entire ex-
not sufficiently flexible in allowing increasing portable output of producing countries, it
market shares for newer, or low cost, produc- would also need the power to dispose of ex-
ers. This was a major reason, for example, for cess stocks either by selling the commodity
the breakdown of the International Coffee (at relatively low prices) for non-traditional
Agreement in 1973. Market sharing conflicts uses and/or by destruction. Alternatively, a
have also caused great difficulties in the re- ceiling on the marketing agency's stocks of
negotiation ofother commodity agreements an agricultural commodity could be agreed,
based on export quotas. so that producing countries would themselves
There are, however, other possible tech- have to hold or destroy some stocks in years
niques which can be considered which avoid of good harvests.
the market sharing problem, at least overtly. A classic example of this type of operation,
One possibility for some commodities would in this case by a private company, is afforded
be the establishment of an intergovernmental by the Central Selling Agency, a subsidiary
joint marketing agency, which would act as of De Beers, which has virtually a monopoly
sole buyer from developing producing coun- of the world supply of diamonds. The tech-
tries, and as sole seller to developed import- nique isprobably applicable to a number of
ing countries. The agency could maintain its other commodities, particularly minerals.
own stock which could, in effect, be used in Further alternatives arise in the field of fis-
the same way as the traditional type of inter- cal policy. One possibility for certain com-
national buffer stock to smooth out excessive modities would be the enforcement by all
short term price variations. But, by its mo- producing countries of an agreed minimum
nopoly position, the agency would also be in price for their exports. An informal arrange-
a position to raise the level of prices as and ment among the governments of Algeria,
when appropriate. The operational principles Italy, and Spain, to take a recent example,
of such a marketing agency, and its mode of forced up the price of mercury by about one-
financing, would need to be elaborated in third between 1973 and mid 1974.
some detail. A minimum export price scheme would
Joint marketing agencies would be easier have similar effects on the aggregate revenue
to operate for minerals than for agricultural of exporting countries as would a correspond-
commodities, since mineral output is more ing export quota scheme, but the distribution
of gains among the exporting countries is
*From Alfred Maizels, "A New International likely to differ insofar as the enforcement of
Strategy for Primary Commodities," in G. K. Hellei- a price minimum results in a shift in the
ner (ed.), A World Divided: The Less Developed
Countries in the International Economy, London, pattern of demand (e.g., away from low price
varieties of the commodity).
Cambridge University Press, 1976, pp. 41-4. Re-
printed bypermission. Another fiscal approach would be the im-

553
 554 TRADE STRATEGY

position of a uniform ad valorem export tax problem. It would, of course, have important
by all producing countries. The more elastic economic effects on both the developing and
is the aggregate supply, and the more inelas- the developed countries.
tic the aggregate demand, the greater will be As regards the latter, indexation would
the proportion of the tax borne by the con- have certain feedback effects on the rate of
sumer. The export tax approach has the ad- inflation, though this is not likely to be a
vantage ofsimplicity insofar as it avoids the major element in the whole range of infla-
problem of classifying different varieties or tionary pressures in the developed market
grades, as well as that of negotiating export economy countries. There would also be a
quotas. An important recent example is the positive element insofar as these various mea-
decision of certain banana producing coun- sures to maintain an adequate unit return, in
tries to impose a tax on banana exports; real terms, for the exports of developing
though strenuously resisted by the transna- countries would be very largely balanced by
tional corporations concerned, it seems that a greater volume, and certainly a smoother
this would be a viable technique in suitable flow, of purchases of capital equipment and
circumstances for achieving a fairer share of other development goods from the industri-
the benefits of trade for the producing alised countries.
countries. As regards the impact on the developing
Where transnational corporations are pro- countries, much would depend on national
ducers as well as traders, the possibility exists action to prevent an excessive expansion in
of increasing the net return to the producing production following an increase in the world
countries from the productive activities of price of a particular commodity. Without
these corporations in developing countries by such action, a price rise could result in the
increases in local taxes and royalties. Apart accumulation of excessive stocks, as well as
from the well known OPEC case, action on being a disincentive to invest in other sectors,
these lines has recently been taken by certain such as in manufacturing industry. Where
developing countries (Jamaica, Guyana and the monetary benefits of the price rise are
Haiti) to increase their tax and royalty reve- channelled into government revenue (as with
nue from bauxite production. An interesting an export tax), it would be administratively
feature of these new tax arrangements is that easier to divorce the world price from the net
they aim to relate government tax revenue return to the domestic producer. In other
from bauxite mining to the open market cases, internal fiscal measures may be re-
value of the end product (i.e., aluminum) quired, according to circumstances, to regu-
where that is processed. This is a special case late the growth of output, to increase the ef-
of the principle of "indexation," which might ficiency ofcommodity production and/or to
well be applicable to a number of other min- finance diversification programmes as
erals also.
appropriate.
Indexation, in its more general form, re- There may thus be scope for effective pro-
lates to measures designed to maintain the ducer action to regulate the level of world
commodity export prices received by devel- prices of a number of important export com-
oping countries in terms of the prices they modities. Detailed examination of the cir-
have to pay for their imports (or their im- cumstances ofeach commodity market will
ports of manufactures) from developed coun- be required to determine the possibilities. In
tries. Such measures clearly presuppose the any event, producer action in one set of com-
evolution of effective market control tech- modity markets would not rule out joint pro-
niques: once these are established, the index- ducer-consumer action in other markets
ation of commodity prices to the prices of, should that prove feasible and desirable for
say, manufactures exported by developed both groups of countries. Cooperation with
countries would not be a difficult technical consumer countries would be important
 555 TRADE IN PRIMARY PRODUCTS

where such countries can improve the effec- sions for annual review in the light of chang-
tiveness ofan agreement, e.g., by "policing" ing market circumstances.
the quotas under an export restriction Bilateral long term agreements of this type
scheme (as in the coffee and cocoa agree- are usual in the trade between the foreign
ments), or where they are important produc- trade enterprises of socialist countries of
ers of the commodity concerned (as for Eastern Europe and national trading corpo-
wheat). rations of individual developing countries.
However, joint producer-consumer agree- Long term agreements covering commodity
ments of the traditional type are likely to in- trade between market economy countries are
volve constraints on the price objective of in- normally undertaken between private com-
ternational action, for the reasons discussed panies. The greater part of Japanese imports
earlier. Such agreements would need to in- of iron ore, for example, are purchased under
corporate, inter alia, supply commitments by long term contracts of up to twenty years' du-
exporting countries at the maximum of a ration from mining companies in Australia
price range as a quid pro quo not only for and in a number of developing countries.
purchasing commitments by importing coun- The negotiation of long term contracts, on
tries at the minimum of the price range, but a government to government basis, would
also for greater flexibility by these latter also allow for the provision of indexation
countries on the level of the price range itself. clauses for the adjustment of prices of the
There should, however, be increasing scope commodities covered in line with changes in
also for long term agreements between com- the prices of manufactured goods. Insofar as
modity importing and exporting countries, on such indexation is related to imports of com-
either a bilateral or a multilateral basis. Such modities bydeveloped countries, it would en-
agreements could specify annual quantities able developing countries to import the same
to be traded (where this can reasonably be commodities at (lower) world market prices.
done), or annual quantitative targets. To the The wider use of long term contracts, both
extent that prices can also be specified in bilateral and multilateral, would allow devel-
such long term agreements, this would help oping countries to plan investments in new
in reducing instability in one sector of the capacity with much greater assurance of prof-
commodity export trade of developing coun- itable returns from their commodity exports
tries. However, price clauses in long term than they have had hitherto.
agreements would normally require provi-

VIII.C.2. International Resource

Bargaining — Note

The control of the price of oil by the Orga- so as to create artificial scarcities and raise
nization of Petroleum Exporting Countries export prices above competitive market lev-
(OPEC) has posed in dramatic form the els? Is oil a special situation, or can effective
question of whether the exercise of "producer producer power be exercised also for bauxite,
power" in commodities can be generalized to copper, tin, phosphates, coffee, tea, rubber,
other primary producers. Can there be addi- and other primary products?
tional producer monopoly agreements that ef- The answers to these questions rest in the
fectively control future supplies or prices of outcome of a complex process of interna-
minerals, food, and other primary products tional resource bargaining. But we might
 556 TRADE STRATEGY

identify various conditions in this process bauxite, and the inclusion of Australia (which
that will determine the relative bargaining might well join the "Third World" for such pur-
strengths of different producing and consum- poses) brings the total above 90 percent. In coffee,
ing countries. First, primary production in the four major suppliers have begun to collude
(even within the framework of the International
the LDCs must be disaggregated — at the
Coffee Agreement, which includes the main con-
least, into such categories as oil, non-fuel suming countries) to boost prices. A few countries
minerals, tropical agricultural products, ce- are coming to dominate each of the regional mar-
reals. Secondly, it should be recognized that kets for timber, the closest present approximation
market intervention can assume various to a truly vanishing resource. The percentages are
forms: a cartel of exporters that takes collu- less, but still quite impressive, for several other key
sive action to set production and export quo- raw materials and agricultural products. And the
tas or to raise export prices, or one producer United States already meets an overwhelming
acting as a price leader with others following, share of its needs for most of these commodities
or by export controls over the flow of exports from imports, or will soon be doing so.
A wide range of Third World countries thus
through imposition of quantitative limita- have sizeable potential for strategic market power.
tions or export taxes. Whether a particular They could use that power against all buyers, or in
form of market intervention will be success- a discriminatory way through differential pricing
ful in raising the export price and revenue for or supply conditions — for example, to avoid higher
a particular primary product will depend es- costs to other LDCs or against the United States
sential y on the existence of three conditions: alone to favor Europe or Japan.
(1 ) a dominant position by a producer coun- Supplying countries could exercise maximum
try in export markets or the capacity for ef- leverage through withholding supplies altogether,
fective collusion by a group of producer na- at least from a single customer such as the United
tions; (2) an inelastic demand for the product States. Withholding is a feasible policy when there
in consumer countries; (3) a low elasticity of are no substitute products available on short no-
tice, and when the foreign exchange reserves of the
supply of alternative materials for consumer
countries. suppliers become sizeable enough that they have
no need for current earnings.
The economic prospects for successful The suppliers would be even more likely to use
market intervention must ultimately be de- their monopoly power to charge higher prices for
termined bydetailed econometric studies on their raw materials, directly or through such tech-
an individual commodity basis.1 Opposing niques as insisting that they process the materials
conclusions of a general nature, however,
may be summarized here. themselves.2
Considering whether oil and OPEC repre-
The "threat from the Third World" has sent a unique case which cannot be duplica-
been emphasized by Bergsten, who states:
ted, Bergsten states that
Four countries control more than 80 percent of
the exportable supply of world copper, have al- it is very doubtful that oil is different in any qual-
itative sense. Indeed, many other OPECs look
ready organized, and have already begun to use
much easier to organize and maintain. OPEC had
their oligopoly power. Two countries account for
to pool twelve countries to control 80 percent of
more than 70 percent of world tin exports, and
world oil exports, but fewer countries are usually
four countries raise the total close to 95 percent.
Four countries combine for more than 50 percent involved in production of other primary products.
Most OPEC countries are heavily dependent on
of the world supply of natural rubber. Four coun- oil, and cartelization was especially risky for them.
tries possess over one-half the world supply of Other commodity producers are more diversified.
OPEC could politicize oil and threaten the world
'Political prospects are as crucial as economic pros- economy; its successors will have an easier task be-
pects, but they are not amenable to such systematic
empirical study. That is why the problem of interna- cause their products are less important. And eco-
tional resource bargaining might be characterized
more aptly as a problem of international resource 2C. Fred Bergsten, "The Threat from the Third
diplomacy. World," Foreign Policy, Summer 1973, pp. 107-8.
 557 TRADE IN PRIMARY PRODUCTS

nomic and political differences among OPEC sugar does not lend itself to producer cartelization
countries seem much sharper than those among because of the large number of suppliers and the
other potential cartelizers. So new OPECs seem at competition from beet sugar and substitute sweet-
least as likely as OPEC itself.3 eners. For other tropical products, such as natural
rubber and fibers, attempts to rig prices would be
In contrast with Bergsten's view, a promi- self-defeating because of competition from
nent report by economists from Japan, the synthetics. . . .
European Community, and North America In the case of cereals, experience with the In-
states that ternational Grains Arrangement suggests how dif-
ficult itis for exporters, even when their number is
the frequent portrayal of a future world of primary
limited, to maintain world prices above competi-
commodities divided up into cartels of developing
tive levels for any significant length of time. Pres-
country producers, as in the Organization of Pe-
troleum Exporting Countries (OPEC), dictating sure isbuilding up for greater price stability, how-
ever, and a renewed international effort to reduce
prices and turning the supply on and off to achieve
the range of price fluctuations for grains and pos-
political or other ends, is vastly overdrawn. Pri- sibly for oil seeds as well seems bound to be given
mary products are produced by many countries, consideration. But if such a negotiation does take
imports are not everywhere a major proportion of
place, it must have as participants both exporters
total consumption, and industrial rather than de- and importers. The end result of a successful ne-
veloping countries are the leading suppliers of gotiation would be a joint commodity agreement
many internationally traded commodities. Fur-
thermore, many primary commodities can be sub- rather than a producers' cartel.5
stituted for each other, depending on price and It is also questioned whether cartels in
availability. New collusive attempts by exporters mineral industries can be effective.
to exploit markets are entirely possible — the more
so because governments now intervene more ac- First, rarely do the major producers and exporters
tively insetting the conditions for production and of a mineral adhere to the same political and eco-
sale, because the issues have become heavily poli- nomic objectives. . . . Second, many of the produc-
ticized, and because the prospects for short-term ing nations depend heavily on mineral exports for
gains are sometimes attractive. Also, following the their foreign exchange earnings, so the costs to
example of OPEC is tempting. The crucial point, them of withholding supplies are high. . . . Third,
however, is that the number of commodities on mineral production involves high fixed costs which
which collusion would be feasible or effective is continue whether production takes place or not.
small, the economic impact is likely to be limited Thus, pressure mounts to cut prices during down-
and isolated rather than pervasive as with oil, and swings in the business cycle to keep capacity uti-
the prospects for sustained success over the me- lization from falling drastically.
dium term, to say nothing for the long term, are All this strongly suggests that attempts by pro-
ducing countries to form cartels in the mineral in-
dim.4 dustries are unlikely to succeed for any extended
On the particular prospects for market period. In time members will find the demand for
manipulation in the markets for tropical ag- their exports falling as producers outside the cartel
ricultural products, the report recalls that boost their supplies and consumers switch to sub-
stitute materials. Eventually, they will have to
past attempts by exporting countries to maintain abandon artifically high prices to avoid losing their
prices of coffee and cocoa above competitive levels
have met with only limited and temporary success, markets entirely.6
as individual exporting countries either would not Another study reaches similar conclusions:
observe or could not enforce export quotas without
Even the strongest political urge, or the most
the cooperation of importing countries. . . . Cane
adroit management, cannot alter certain basic fac-
tors that, in our judgment, severely limit the pos-
3C. Fred Bergsten, "The New Era in World Com-
modity Markets," Challenge, September-October sibleminerals.
non-fuel accomplishments of producers' alliances in
1974, p. 40.
4 Trade in Primary Commodities: Conflict or Co- 5Ibid., p. 32.
operation?, ATripartite Report, The Brookings In-
stitution, Washington, D.C., 1974, pp. 1-2. 6Ibid., pp. 30-31.
 558 TRADE STRATEGY

The key economic fact is that, while demand for community by inducing the production of
most non-fuel minerals is price-inelastic in the costlier sources of supply, aggravating infla-
short run . . . , this is not necessarily true over the tionary pressures, slowing down the rate of
long run, certainly not to the extent that holds for growth in consuming countries, and inviting
oil. Calculations based on historic experience for retaliatory measures.
tin, aluminum and copper, for example, suggest
strongly that in the long run the drop in demand The deleterious effects on the "Fourth
more than offsets any price increase, so that the World" countries that are in the "resource-
total return to the producers eventually becomes poor" category — that is, agricultural coun-
less than before the price change. Although the tries without mineral resources — should be
econometric measurement of price elasticities is a of special concern to the international com-
tricky process leading to differing estimates of in- munity. The problem of acquiring minerals
dividual cases, there is little disagreement on the (and the derivatives of fertilizer and food-
broad point about short-term and long-term price stuffs) at a price they can afford became in
elasticity. the 1970s a major handicap to the realization
The reasons are threefold — stockpiles, recycling
of development plans in resource-poor devel-
possibilities, and the use of substitutes — none of oping countries. Special arrangements that
which, of course, applies to oil in anything like the
will provide these countries with the means of
same way as yet.7 financing these imports or that will provide
The foregoing clearly represent different them with additional export revenue may im-
views on the prospect of OPEC-like cartels pose additional costs on the world economy.
being effective in various primary commodity It is also noteworthy that most primary
markets. The major purpose of the cartel products are actually produced in high-in-
might be interpretedpx as the improvement in come countries. At the same time as several
the producing country's terms of trade — rich countries are substantial exporters of
more precisely, not simply the commodity primary products, some poor countries are
Pm heavy importers of foodstuffs and industrial
term of trad ~r , but rath t i
s e PXQX er he ncome raw materials. Price-raising commodity
agreements would therefore not offer a
terms of trade , where Px is an export straightforward transfer of resources from
price index, Qx is an export volume index, rich to poor countries. On the contrary, as
and Pm is an import price index. This is an Professor Cooper notes, they would generate
attempt to raise the primary country's share a quite arbitrary distribution of gains and losses
in the gains from trade and to increase its among both developed and less developed coun-
"capacity to import" (Qm) derived from a tries. Moreover, the less developed countries that
given Qx (as distinguished from the capacity would benefit tend to be those that are better off;
to import based on a capital inflow). Thus, with minor exceptions, the poorest countries would
issues of international distributional equity not benefit, and often would lose, from raising
commodity prices. Such an action would thus not
and the size of a country's development pro- contribute much to the alleviation of world
gram (as determined by the foreign exchange
constraint), become intertwined with the pro-
cess of international resource bargaining. From the standpoint of world economic
poverty.8
But to the extent that cartels and other welfare, the policies of trade liberalization,
measures to control export prices succeed, additional aid, and foreign investment (with
they impose real costs on the international a benefit-cost ratio greater than unity) are
first-best policies in comparison with cartels
7Bension Varon and Kenji Takeuchi, "Developing
Countries and Non-Fuel Minerals," Foreign Affairs, 8Richard N. Cooper, "Developed Country Reac-
April 1974, pp. 505-6. For similar views, see also tions to Calls for a NIEO," in Toward a New Strat-
Raymond F. Mikesell, "More Third World Cartels egy for Development, A Rothko Chapel Colloquium,
Ahead?" Challenge, December 1974, pp. 26-7. 1979, p. 259.
 559 TRADE IN PRIMARY PRODUCTS

and related price fixing undertakings that tion, undergo multilateral surveillance both at
have costly side effects and are hence second- their initiation and throughout their subsequent
best as policies for increasing the gains from application, have to be administered through ex-
trade and raising the capacity to import. port taxes rather than quotas whenever possible,
The discussion of the control of trade in and apply on a most-favored-nation basis. Such a
regime would seem to provide a reasonable chance
primary products has raised other issues that
that export controls could be relatively depoliti-
deserve attention. The use of export controls
cized, and avoid unilateral reactions and emula-
that impede access to supplies imposes costs tion/retaliation cycles.
on the world economy that are similar to
those of import controls. But GATT has been Any comprehensive arrangements to "assure
access to supplies" must probably encompass com-
concerned only with the latter. Bergsten has modity agreements as well as rules limiting the use
therefore proposed new rules to govern export of trade controls. Buffer stocks, agreed production
controls: levels and perhaps other arrangements to preserve
agreed volumes of trade and price ranges would be
Like those existing GATT rules which seek to gov- needed. The nature of the decision-making ma-
ern the use of import controls, a set of rules on ex- chinery on all these issues would be a central con-
port controls would have four immediate goals: cern to all negotiating parties.
1. To deter producing countries from erecting Thus, from an intellectual standpoint, commod-
export controls except in clearly denned and jus- ity agreements and new trade rules must be viewed
tified circumstances. together in dealing with the problem of access to
2. To reinforce that deterrent by providing a supplies. But they msut also be viewed in tandem
basis for concerted response by the world trading from a negotiating standpoint, because together
community. they might provide the basis for a package which
3. To limit the scope and duration of those con- served the interests of the several different groups
trols which are actually applied.
4. To provide an international framework into of countries.9
which disputes triggered by export controls can be International commodity agreements are
channeled when they are actually applied, to re- also frequently proposed to reduce price fluc-
duce the likelihood of unilateral reactions and em- tuations and stabilize foreign exchange
ulation/retaliation cycles. . . . earnings.
The proposed set of rules would permit the use Finally, the problem of international re-
of export controls for national security purposes source bargaining relates to two other issues
and "infant industry" protection, and to avoid se- discussed in previous chapters. Another ob-
rious injury (or the threat thereof) to national
economies. But they would otherwise be pro- jective ofthe exercise of "commodity power"
scribed. is to increase the domestic value added from
the primary commodity by inducing domestic
The proscription would be backed by the re-
quirement that any country adopting an unjusti- processing of the commodity (bauxite into
fied export control had to provide adequate com- alumina, crude oil into refined petroleum
pensation or, more likely, accept any of a number products, etc.). This relates to the process of
of types of economic retaliation. Nonsignatories of industrialization via export substitution, as
the agreement could be subjected to such retalia- discussed in selection VLB. 3 above. An ad-
tion, broadening the scope of the deterrent, al- ditional objective is to exercise more national
though every effort would be made to maximize control over foreign investment in primary
membership in the new regime. Retaliation would
products and to capture more of the rents
also be authorized against "reverse dumping" and from that investment. This relates to the
against import subsidies, in the form of import
problem of private foreign investment, as dis-
duties and "countervailing export controls," cussed insection V.D, above.
respectively.
When controls were justified, their application
would be subject to agreed time limits and, for
9C. Fred Bergsten, Completing the GATT: Toward
longer-term controls, the presentation of an ac- New International Rules to Govern Export Controls,
ceptable domestic adjustment program. And they British-North American Committee, Washington,
would require advance notification and consulta- D.C., 1974, pp. 23,51-2.
 560 TRADE STRATEGY

Thus, the exercise of "producer power" commodities; and (3) an increase in the ben-
raises not just one issue — that of improve- efit-cost ratio of foreign investment. For the
ment in the terms of trade — but also other longer-run period that is relevant to the out-
related issues: (1) the stability of export come of development programs, the resolu-
prices and foreign exchange receipts in order tion of these other issues may well be more
to increase the developing country's capacity crucial than the shorter-run use of "producer
to import; (2) domestic processing of primary power" to improve the terms of trade.
Comment
The feasibility of international arrangements to stabilize commodity prices is highly de-
pendent on the characteristics of individual commodity markets. For different commodities,
see Walter P. Falcon and Eric A. Monke, "International Trade in Rice," Food Research
Institute Studies, Vol. XVII, No. 3 (1979-80); International Maize and Wheat Improvement
Center, World Wheat Facts and Trends (1981). The problems associated with buffer stocks
are analyzed by Shlomo Reutlinger, "Evaluating Wheat Buffer Stocks," American Journal
of Farm Economics, (February 1976); Anne E. Peck, "Implications of Private Storage of
Grains for Buffer Stock Schemes to Stabilize Prices, Food Research Institute Studies, Vol.
XVI, No. 3 (1977-78); David M. G. Newbery and Joseph E. Stiglitz, The Theory of Com-
modity Price Stabilization: A Study in the Economics of Risk (1981).
 VIII.D. REGIONAL
INTEGRATION AND
DEVELOPMENT-NOTE

The possibilities for South-South cooperation will be concerned here with only the impli-
are now being considered more seriously — cations offree trade in goods.
whether in deference to Lewis's concern As a basis for appraising specific proposals,
about the slow growth of the MDCs, or to we consider in this Note the benefits that
other implications of dependency, delinking, might be derived from economic groupings
and collective self-reliance. A major effort at among developing countries and the difficul-
South-South cooperation could come from ties that are likely to be encountered in their
regional integration schemes. As a close stu- formation.
dent of the North-South dialogue observes, Advocates of an economic union believe
The weakness of most proposals for South-South that its formation will accelerate the devel-
cooperation has been that they tended to build opment ofthe member countries by (1) stim-
grand designs on the basis of an aggregated, myth- ulating the establishment and expansion of
ical South. ... It may be far more productive to manufacturing industries on a more rational
follow up on avenues of cooperation on a regional basis, (2) increasing the gains from trade,
or sub-regional level and in certain specific areas and (3) providing benefits from intensified
of action. The perspective for cooperation may competition.
often be functional and geopolitical, rather than We have seen that many countries have
global and "idealized."1 adopted a policy of deliberate import substi-
Various degrees of integration are possible, tution in consumer goods. When each coun-
but most interest centers on the potential role try restricts its imports, however, and at-
of customs unions and free trade areas. Al- tempts to substitute home production,
though both a customs union and free trade industrialization becomes unduly compart-
area provide for across-the-board trade lib- mentalized, and the uneconomic multiplica-
eralization for all or most products among tion of import-competing industries is waste-
the member countries, a customs union also ful. In contrast, if manufacturing industry
adopts a common external tariff. can be encouraged in the context of a cus-
At a level less general than a customs toms union or free trade area, it may attain
union or free trade association, regional in- a higher level of productivity than results
tegration might be directed simply toward from industrial protection in each country.
"sectoral integration" — that is, the removal Greater specialization within the region can
of trade restrictions on only a selected list of increase the share of exports and imports in
commodities, or the treatment of the prob- manufacturing and reduce the excessive
lems of some one industry as a whole on a number of products manufactured in an ex-
regional basis. cessive number of protected firms.
Beyond free trade in goods, a more com- To reach an efficient scale of output, a
prehensive economic union might allow for modern manufacturing plant may have to
the free movement of factors of production, a produce a larger output than the low level of
common monetary system, and the coordi- home demand in a single underdeveloped
nation ofeconomic policies among the mem- country can absorb. By pooling markets
ber countries. It is still unrealistic to expect through the removal of internal trade bar-
this of developing countries; therefore, we riers, afree trade union might thus provide a
'Mahbub ul Haq, "Beyond the Slogan of South-
sufficiently wide export market to make econ-
omies of scale realizable. Within a union,
South Cooperation," in Khadija Haq (ed.), Dialogue
for a New Order, New York, 1981. secondary industry can become more efficient

561
 562 TRADE STRATEGY

as specialization occurs in the member coun- the region by lower-cost producers, the effect
try that acquires a comparative advantage. is one of "trade creation."3 The gains from
At the same time, the other constituent coun- trade are then increased, since the interna-
tries may now replace their imported manu- tional division of labor is improved as re-
factures from outside the union and thereby sources shift into more efficient production.
be able to spend a higher proportion of their On the other hand, some of the intraunion
foreign exchange on outside imports that are trade may merely replace trade that formerly
essential but cannot be produced efficiently occurred between members and nonmem-
within the union. A more rational pattern of bers. When the formation of an economic
production and trade within the region may union has this "trade-diverting" effect, the
therefore be an important result of international division of labor will be wors-
integration. ened if the outside source of supply is ac-
It is frequently argued that because of sim- tually alow-cost source, and its product now
ilar levels and patterns of consumption in the becomes higher priced within the union be-
LDCs, there should be more scope for inter- cause of the external tariff. In this case, there
LDC trade than for trade with developed is an uneconomic diversion of output from
countries. As Frances Stewart states, the low-cost outside source to the high-cost
supplier within the union, and the gains from
Consumption patterns among South (i.e., devel- trade are diminished.
oping) countries should be much more similar to
each other than those of North (i.e., developed) In considering whether trade creation or
countries: the sort of goods — for consumption and trade diversion is likely to dominate in a par-
investment — developed for one country in the ticular union, we have to take into account
South should be more appropriate, both for pro- the preunion level of tariff rates among the
duction and consumption, to other South countries members, the level of the postunion external
than to North products they currently import. . . . tariff compared with the preunion tariff levels
A coordinated policy to encourage South-South of each member country, the elasticities of
trade would provide the South innovators with the demand for the imports on which duties are
markets they require, and hence, in the end, would
reduced, and the elasticities of supply of ex-
be likely to prove self-justifying.2 ports from the members and foreign sources.
The extension of the market, together with Conditions are more propitious for trade cre-
the inducement to get behind the external ation when each member's preunion duties
tariff wall, may also be particularly effective are high on the others' products, the mem-
in attracting direct private foreign invest- bers are initially similar in the products they
ment in manufacturing. And over time, there produce but different in the pattern of rela-
is the further possibility that new industries tive prices at which they produce them, the
can become increasingly competitive on external tariff of the union is low compared
world markets and eventually be able to ex- with the preunion tariff levels of the mem-
port manufactured goods to nonmember bers, and the production within the union of
countries. But this depends first on establish- commodities that are substitutes for outside
ing a sufficiently wide market within the imports can be undertaken at a lower cost.
union to allow operation of a manufacturing The formation of a free trade union might
industry on a large enough scale. also result in an improvement — or at least
An expansion of trade among the member the forestalling of a deterioration — in the re-
countries is also expected to result from the gion's commodity terms of trade. This is pos-
removal of trade barriers. If this takes the sible ifthere is a reduction in the supply of
form of replacing high-cost producers within exports from the union, or the demand by
members of the union is reduced for imports
2Frances Stewart, "The Direction of International
Trade: Gains and Losses for the Third World," in G.
K. Helleiner (ed.), A World Divided, New York, 3Jacob Viner, The Customs Union Issue, New
1976, pp. 98-9. York, 1950, pp. 48-52.
 563 REGIONAL INTEGRATION AND DEVELOPMENT— NOTE

from outside, or the bargaining power of the requirements may be beyond their present
members in trade negotiations is strength- capacity. Aside from the political and admin-
ened. But unless the members of the union istrative difficulties, there are also several
are the chief suppliers on the world market economic objections to a union.
or constitute a large part of the world market To begin with, it may be argued that the
for their imports, they are unlikely to be able case for an economic union is in reality weak
to exercise sufficient monopolistic or monop- when the constituent countries have not yet
sonistic power to influence their terms of established many industries. Limitations on
trade by raising duties on their trade with the the supply side may be more of a deterrent to
outside world or by inducing outsiders to sup- the creation of an industry than is the narrow
ply their goods more cheaply. Moreover, market on the side of demand. If production
when free trade is confined only to the region, conditions do not also improve, the mere ex-
there is the risk of retaliation through the for- tension of the consumer market will not be
mation ofother economic blocs. A union may sufficient to create industries. Moreover,
thereby inhibit the realization of the more ex- when manufacturing industry is only at a ru-
tensive gains from the "universal" approach dimentary stage in the member countries,
to free trade. there is not much scope for eliminating high-
Finally, regional integration might be ben- cost manufacturers within the region. Nor is
eficial inencouraging competition among the there much scope for realizing the benefits of
member countries. Technical efficiency in ex- increased competition when there are not yet
isting industries might then be improved as similar ranges of rival products, produced
marginal firms are forced to reduce their under different cost conditions, in the several
costs, resources are reallocated from less ef- member nations. A union will not cause sub-
ficient to more efficient firms, and monopolies stantial improvement in the utilization of re-
that had previously been established behind sources unless industries have already been
tariff walls are no longer in a sheltered posi- established but need wider markets than the
tion. Further, the stimulation of competition national economy can provide for the reali-
within each country may yield not only a bet- zation ofeconomies of scale, and the member
ter utilization of given resources, but may countries have been protecting the same
also raise the rate of growth of productive re- kinds of industry, but have markedly differ-
sources. This may result from stronger incen- ent ratios of factor-efficiency in these indus-
tives to adopt new methods of production, to tries to factor-efficiency in nonprotected
replace obsolete equipment more rapidly, and branches of production.
to innovate more rapidly with more and bet- It has been pointed out that the case for a
ter investment. union is strongest among countries that have
In practice, however, a number of objec- little foreign trade in proportion to their do-
tions have been raised against proposals for mestic production, but conduct a high pro-
regional integration, and actual negotiations portion of their foreign trade with one an-
have encountered serious difficulties. As is other.4 When these conditions prevail, there
true for a union among even advanced coun- is less possibility for introducing, within each
tries, political problems take precedence, na- member country, a distortion of the price re-
tions will guard against a sacrifice of their lation between goods from other member
sovereignty, and the administration of the countries and goods from outside the union,
union may be extremely complex. For under- and more of a possibility for eliminating any
developed countries, these problems tend to
be especially acute since many have only re-
cently gained political independence, newly 4R. G. Lipsey, "The Theory of Customs Unions: A
General Survey," Economic Journal, September
established national governments may be ex- 1960, pp. 507-9. This conclusion rests, however, on
cessively concerned with their own national the assumption that there are no productive econo-
interests and needs, and the administrative mies of large scale.
 564 TRADE STRATEGY

distortion by tariffs of the price-relations be- ative advantage in only primary products and
tween domestic goods and imports from other will sell to other members only goods that it
member countries. There is therefore greater could as readily export to outside countries.
likelihood that the union will improve the use At the same time, the location of manufac-
of resources and raise real income. turing industry and ancillary activities may
A union among underdeveloped countries, become localized within one member coun-
however, is unlikely to conform to these con- try, and "polarization" results. Other mem-
ditions. The ratio of foreign trade to domestic bers may then contend that if they too had
production is generally high for these coun- been able to adopt tariff protection against
tries, and the actual volume of intraregional their partners, they would have also been
trade is normally only a small proportion of able to attract industry. A nonindustrialized
the region's total foreign trade. The gain member country may further complain that
from regional integration would therefore be in buying from an industrialized partner, in-
small. The basic difficulty is that, with exist- stead of importing from the outside, it is los-
ing trade patterns, the formation of a union ing revenue equal to the duty on outside man-
is likely to cause a considerable amount of ufactures. And, with a common external
wasteful "trade diversion." Over the long tariff, member countries no longer have the
run, comparative costs and trade patterns discretionary power to use variations in the
may change, and economies of scale may give tariff for the purpose of adjusting their na-
rise to competitive advantages as develop- tional revenues to their own requirements.
The internal strains that arise from uneven
ment proceeds, so that the scope for "trade
creation" will become greater within the development among the member countries
union. But the immediate gain is small, and may thus make it extremely difficult to pre-
the longer-run prospects for the creation of serve a regional grouping. As one study states
new trade are not likely to influence current
decisions to join a union. Surprising as it may appear to some, one of the
The case for regional preferential trading main obstacles lies in the differences among the
arrangements is stronger than that for a gen- developing countries which form a union. These
eral preference scheme if the regional ar- disparities of size and stage of economic develop-
rangement allows the avoidance of trade di- ment cause problems so far as the equitable shar-
version. GATT (Article 24) insists that ing of costs and benefits are concerned. In fact,
tariffs among members of a customs union or this emphasis on "equity" might well prove the
free trade area be reduced to zero; it can be rock on which all such schemes founder. If indeed
customs unions are instituted to take advantage of
demonstrated, however, that in some cases,
less trade diversion will result if the members a larger market, to shift production to optimum lo-
reduce their internal tariffs below the exter- calities and, in general, to stimulate economic ef-
ficiency, some sectors in some geographical areas
nal tariff but not necessarily to zero.5 In this will undoubtedly be adversely affected. Theoreti-
respect, a partial preferential arrangement cally, itshould be possible, if the customs union
has merit. leads to a rate of growth higher than would oth-
Besides the possibility of "trade diversion," erwise have been the case, to compensate the losers
other undesirable consequences may result and have a net surplus. But the technical imple-
from a union. The member countries are un- mentation of compensation schemes is fraught
likely to benefit equally, and some members with practical difficulties.
may feel that others are gaining at their ex- Basically, what has to be provided for the dis-
advantaged members, if they are to stay in the
pense. A country may have a strong compar- union, is one or more of the following preferential
measures:
5Ibid., pp. 506-7; W. M. Corden, Recent Develop- 1. Balance-of-payments support through finan-
ments in the Theory of International Trade, Prince- cial institutions of the union or external institu-
ton, 1965, p. 54. tions (e.g., the IMF). A credit system organised by
 565 REGIONAL INTEGRATION AND DEVELOPMENT— NOTE

a Regional Payments Union (perhaps initially avoidance of duplication, and better location
funded externally) might be one way by which of projects might result. There are a few out-
countries could be assisted if intra-unicn trade
standing examples of multinational invest-
were to lead to balance-of-payments problems. ment projects, and there should be scope for
2. Fiscal policy support measures. Import du-
ties, particularly for small countries, often form a many more.
large part of total government revenue. These are It is also possible for more to be accom-
slated to disappear with the union and so alterna- plished byway of partial liberalization of re-
tive fiscal means must be developed for substitut- gional trade for certain products or sectors.
ing lost receipts. Fiscal policy may also have to be Countries might identify specific sectors or
used to encourage new industry to locate in the dis- individual products for which they could
advantaged member but, of course, if this means a commit themselves not to erect trade barriers
permanent subsidy, the advantage of the customs with each other. Products that are not yet
union is lost.
fabricated in a particular group — that is,
3. Commercial policy might be especially
geared to the needs of the least developed of the new products for the region — might be sin-
gled out as a particularly suitable object of
group. For example, a longer transition period to
the zero tariff situation might be provided for. such a commitment, especially if supple-
mented by some commitments regarding a
4. Credits for new investment.6
regional investment policy.
A regional agreement might also be effec-
The most important lesson to be learned tive in bargaining over the entry of foreign
from efforts at regional integration is that, if investment and the import of technology into
the potential benefits of integration are to be the region. Instead of individual countries
fully realized, the regional association must having to engage in their own bargaining
be a strong one and must be capable of co- with the foreign investor or foreign supplier
ordinating trade policies, including ex- of technology, a regional bargaining unit
change-rate policy among the member coun- may have more power and may be able to
tries, and must provide some means for an
harmonize conditions of entry to greater ad-
equitable distribution of the costs and bene- vantage for the recipient members.
fits among members. Most proposals for re- Finally, an important area of regional co-
gional integration do not yet show promise of
operation— and one that could be used to
sufficient cohesion and policy coordination.
support other areas — is that of channeling
Although a comprehensive form of free
aid through regional integration banks or de-
trade area or full customs union may not yet
velopment corporations. The more this is
be practicable for most of the developing done, the more influential might the regional
countries, there are still substantial advan- institutions be in promoting the regional
tages that can be derived from more ad hoc
investment policies and regional trade lib-
functional types of regional cooperation short eralization policies that are necessary
of comprehensive integration. Measures of to avoid uncoordinated duplicative national
"partial integration" may help to avoid the development policies.
costs of "micro-states" and national devel- Until the risks of joining a free trade union
opment along compartmentalized lines. In are diminished, this less ambitious approach
particular, the complementary development involving efforts to secure sectoral integration
of specific industries through a regional in- may be the most feasible alternative. Even
vestment policy has considerable potential. though a customs union may be the ultimate
The realization of markets of sufficient size,
objective, it will still be a sizable accomplish-
ment in the immediate future to secure the

6Derek T. Healey, Integration Schemes among De- mutually supporting measures of regional in-
veloping Countries: A Survey, Sixth Conference of vestment policies, regional trade liberaliza-
Economists, Hobart, 1977, pp. 31-2. tion, and regional aid institutions.
 566 TRADE STRATEGY

Comment
There have been several efforts to secure economic integration among developing countries.
Most notable are the Latin American Free Trade Association (LAFTA), the Central Amer-
ican Common Market (CACM), the Andean Common Market (ACM), the Caribbean Com-
munity (CARICOM), the East African Community (EAC), and the Association of Southeast
Asian Nations (ASEAN). These integration schemes, however, have generally not lived up
to expectation, and most have undergone major modifications. Problems have generally been
encountered in determining the scope for profitable specialization, the creation or strength-
ening of appropriate production structures, and the equitable distribution of the benefits of
integration among member countries.
For an appraisal of the experience with regional integration schemes, see J. B. Nugent,
Economic Integration in Central America (1974); David Morawetz, The Andean Group
(1974); Peter Robson, The Economics of International Integration (1972); Martin Carnoy
(ed.), Industrialization in a Latin American Common Market (1972); UNCTAD, Economic
Cooperation and Integration among Developing Countries (1976); Bela Balassa and E. J. Sto-
utjesdijk, "Economic Integration among Developing Countries," Journal of Common Market
Studies (September 1975); Constantine V. Vaitsos, "Crisis in Regional Cooperation among
Developing Countries: A Survey," World Development, Vol. 6 (1978); Peter Robson, Inte-
gration, Development and Equity: Economic Integration in West Africa (1983).
CHAPTER IX

Human-Resource
Development

Another strategic policy issue is the development of human resources. If the earlier faith in
development through the accumulation of material capital has waned, it has in recent years
been replaced by a new creed of investment in human capital. It is now widely believed that
improvement in the quality of people as productive agents must be a central objective of de-
velopment policies. But how are the abilities and skills of people to be improved, and their
motivations and values modified so as to be more suitable for developmental efforts? This is
clearly one of the most difficult questions we have encountered, and yet, on its answer is likely
to depend a country's success in achieving self-sustaining development.
Part of the difficulty in formulating policy to improve the "human infrastructure" within a
less developed country is that economists have only recently begun to analyze this question
systematically. Perhaps the greater part of the difficulty, however, is that an answer to the
question entails not only economic analysis but also sociological, psychological, and political
considerations. The inherently multidisciplinary character of the answer has caused each dis-
cipline to acknowledge the question but to fall short of a satisfactory answer. This chapter,
in concentrating on the economics of the problem, will not be immune to this same criticism,
but it will at least attempt to fashion the economic analysis so that it might suggest some
links with other disciplines.
Before considering policy measures related to the quality of the population we should first
gain some perspective on the quantitative problem of population pressure. Section IX. A ex:
amines the nature of the population explosion and its importance from the standpoint of its
effects on the quality of population. Population is growing much faster in the LDCs than in
the MDCs, and the urban population is growing especially rapidly. By the end of this century,
about 80 percent of the world's population will be in the developing countries, and human-
resource policies have to be considered in the context of this rapid population growth. The
exacerbation of the employment problem is direct. The effects of population growth on the

567
 568 HUMAN-RESOURCE DEVELOPMENT

quality of health facilities, nutrition, educational programs, and public services also make it
necessary to acquire a better understanding of demographic problems. It is now especially
apposite to investigate whether population growth worsens absolute poverty and the maldis-
tribution ofincome. Do absolute poverty and the maldistribution of income contribute, in
turn, to high rates of fertility? Furthermore, what are the socioeconomic correlates of declin-
ing fertility, and how might policy interventions affect these correlates in order to reduce the
rate of population growth? If public policies can reduce mortality rates, can they also reduce
fertility rates? These questions are examined in section IX.A.
Section IX. B outlines some policies to improve health and nutrition conditions among the
world's poor. Among the first to suffer from shortages or inadequacies of food are the chil-
dren— a country's investment in the future. And children, along with their mothers, are nu-
merically dominant in developing countries. In these areas one-fifth of the population is under
the age of 5 years, two-fifths are below the age of 15 years, while mothers and children to-
gether account for over two-thirds of the total population. The children are the most vulner-
able group.
Especially tragic is the ill-health brought on or aggravated by malnutrition. Malnutrition
adversely affects mental progress, physical growth, productivity, and the working life span.
In cost-effectiveness terms — let alone fundamental human rights — there can be high returns
from programs to reduce the incidence of malnutrition.
Poverty is the major cause of disease in developing countries, and more than medical fa-
cilities isneeded to improve health conditions. Health policies must be related much more to
the environment and to the ecological, cultural, and nutritional situation which permits dis-
ease to thrive in poverty areas. As with technology, so too is there a need for an "appropriate"
transfer of medical knowledge and medical technology: health programs have been only too
often biased toward a small section of the urban population, and to oversophisticated "cura-
tive" treatment rather than more basic, widespread "preventive" treatment.
As analyzed in section IX. C, formal and informal education and training in both the mod-
ern and traditional sectors are clearly necessary for the development of human resources. Any
cost-benefit analysis of the "returns" to education must incorporate the interactions between
education and the economy, giving particular attention to education as an investment, the
importance of rural education in a developing economy, and the interdependence between
education, manpower requirements, and development. Although each individual commonly
views his or her own education as a consumption good, it is more appropriately viewed from
the standpoint of the economy's development as an investment good: to the economist, human
beings can be conceptualized as human capital or embodied savings. It is then an economic
problem to determine how much the economy should invest in human capital, and of equal
importance, what the composition of that investment should be. The selections in section IX.C
therefore consider both the quantitative growth of education and the character of education
needed in a developing economy.
Although this chapter considers population growth, health, nutrition, and education each
as separate topics, it should be realized that all of these elements of human-resource devel-
opment are interrelated. And while fertility, health, nutrition, and education affect one an-
other, so too does each, in turn, affect income. Much of our earlier discussion of basic human
needs is also relevant to the topics of this chapter.
 IX.A. POPULATION
IX. A. 1 . Population and
Poverty— Note

Few developmental problems evoke as much increasing returns to scale may have dis-
pessimism as does the rapid increase in pop- pelled the shadow of Malthus from the few
ulation inpoor countries. If economists value rich industrial countries, that shadow still
a higher rate of development, they still fear, hovers over the many poor agrarian
as John Stuart Mill did, a growth in popula- countries.
tion that "treads close on the heels of agri- In these countries, per capita income has
cultural improvement, and effaces its effects remained pitifully low, and the alarming
as fast as they are produced." It is not diffi- prospects of population growth have revived
cult to find a basis for this fear — in that a neo-Malthusian fears. Once again there is
large part of the gain in aggregate income widespread concern that economic better-
has been used simply to support a larger pop- ment will be thwarted by excessive popula-
ulation at the same low per capita level of tion pressure and that, unless acceptable
income. means are found for checking population
This pessimistic outlook on population growth, the "revolution of rising expecta-
growth is as old as academic economics it- tions" must remain unfulfilled. It has become
self.1 Long ago economics was designated the common to hear the development problem
"dismal science." Originally this was because summarized as one of "increasing the fertil-
of the view of classical economists who, when ity of the soil and reducing the fertility of
considering the long-run development of an
economy, could see only the eventual advent human
Mostbeings." development economists argue
of a dismal "stationary state" in which the against the Malthusian view that population
springs of economic progress would have will eventually be limited by food production.
evaporated, growth in output would no longer They point to the fact that apart from
outstrip population growth, and wages would drought areas, population nowhere appears
be at subsistence levels. So far, however, this yet to have approached the capacity of the
pessimism has proved unfounded for indus- land to support it. Whether or not adequate
trial nations: the economic histories of these food production is achieved, however, de-
nations represent the success stories of eco- pends a great deal on government making re-
nomic progress — of the achievement of a sources available, pursuing appropriate pric-
high rate of increase in real income, so that ing policies, ensuring necessary research, and
population and per capita real income have developing an infrastructure for transport,
both been able to increase. Indeed, when storage, and marketing.
some economists thought that the Great Regarding the problem of malnutrition
Depression of the 1930s might indicate a and hunger, this is, as noted in Chapter VII,
state of secular stagnation in mature indus- more a problem of poverty and inadequate
trial nations, they were concerned lest popu- income than a matter of inadequate global
lation growth in these countries actually be food supplies. In many countries where mal-
too slow. But although capital accumulation, nutrition isextensive, food availability would
technical progress, and the phenomenon of meet the food needs of all if it were distrib-
uted according to need and not according to
'The history of thought on population growth and income. The population-food problem is
development is traced by Lord Robbins, The Theory
of Economic Development in the History of Eco-
solved when incomes suffice to buy adequate
nomic Thought, New York, 1968, Lecture II. food at prices that provide adequate incen-

569
 570 HUMAN-RESOURCE DEVELOPMENT

tives to producers. From the viewpoint of ment and inequality, students of population
global food policy, the rich countries are ca- are now also asking whether population
pable of generating surpluses of food for ex- growth intensifies the extent of absolute pov-
port. But to meet their increased demand for erty and the maldistribution of income.
food grains by increased imports, the devel- In most developing countries, the working-
oping countries would need to export more, age population has roughly doubled in the
receive foreign aid, or borrow overseas. The past 25 years. At expected growth rates, it
great majority of developing nations will will double again in the next 25 years. This
therefore have to satisfy most of their own growth clearly intensifies pressure on em-
production needs most of the time. To this ployment and the amount of investment
end, cost-reducing innovations must continue available per labor market entrant.
to occur sufficient to outweigh the limitations That these disadvantages have become
of fixed land resources. very real for many LDCs can be recognized
It is extremely important, however, to re- if we contrast the demographic patterns of
alize that the population problem is much rich and poor countries. One striking differ-
more than a food problem: it has wider ram- ence isthat a much higher percentage of the
ifications that make it a general development total population in poor countries is in
problem. A high rate of population growth younger age groups, and life expectancy is
not only has an adverse effect on improve- much lower than in rich countries. In most of
ment in food supplies, but also intensifies the the LDCs, 40 to 50 percent of the population
constraints on development of savings, for- is below fifteen years of age, whereas in de-
eign exchange, and human resources. Rapid veloped countries the corresponding percent-
population growth, which stems from high age is about 25 percent. If the economically
birth rates, tends to depress savings per cap- productive age bracket is taken as fifteen to
ita and retards growth of physical capital per sixty-five years, the percentage of population
worker. The need for social infrastructure is in this category is considerably less in poor
also broadened, and public expenditures countries than in rich. This "bottom heavy"
must be absorbed in providing these facilities age structure of population results in a high
for a larger population rather than in provid- ratio of dependents to adult workers, which
ing directly productive assets. Population means the differentiation and productive
pressure is also likely to intensify the foreign power of the labor force are limited. Even
exchange constraint by placing more pres- worse, it means greater consumption and
sure on the balance of payments. In some constitutes a major obstacle to an increase in
cases, the need to import foodstuffs will re- savings in many LDCs. This high depen-
quire the development of new industries for dency ratio requires the economy to divert a
export expansion and/or import substitution. considerable part of its resources, that might
Possibly the most serious disadvantage to a otherwise go into capital formation, to the
high rate of population growth in a poor maintenance of a high percentage of depen-
country is that it makes the human resources dents who may never become producers or, if
constraint more difficult to remove. Larger so, only for a relatively short working life.
numbers militate against an improvement in Not only do most of the people in the world
the quality of the population as productive live in the less developed areas, but the con-
agents. The rapid increase in school-age pop- centration isincreasing. In many LDCs, the
ulation and the expanding number of labor populations are of the high-growth potential
force entrants put ever-greater pressure on type. In these countries the simple applica-
educational and training facilities and retard tion of modern public health measures has al-
improvement in the quality of education. lowed the death rate to fall spectacularly to
Similarly, too dense a population aggravates the low levels of the rich country, while birth
the problem of improving the health of the rates have remained very high and resistant
population. With the concern over unemploy- to change. If, as its true for many LDCs, the
 571 POPULATION

birth rate remains at a high level of about 40 clines infertility,2 without having to wait for
per thousand while the death rate is reduced higher levels of economic and social devel-
to about 10 per thousand, then the population opment tobring about a "normal" decline in
will double within a generation. With the de- fertility. Can family planning programs be ef-
clining death rates, the rates of natural in- fective inthe face of traditional beliefs and
crease have risen in the LDCs to 2, 3, or even social institutions that have sustained fertil-
4 percent a year in regions of persistently ity at a high level? Or can an adjustment of
high fertility. While the population in the birth rates to the fall in death rates be in-
rich countries, constituting 30 percent of the duced only by long-run forces of development
world's population, is growing at an average and the resultant changes in the traditional
of only 1 percent or less, the population in the culture? Is the future of population growth
poor countries, amounting to 70 percent of essentially a question of the future of
the world's population, is growing at an av- development?
erage rate of 2.5 percent. Almost all the dou- This Note has reversed the last question
bling times for populations in poor countries and emphasized that the very future of de-
are more rapid than the world average. The velopment isitself dependent upon a reduc-
next selection outlines the demographic situ- tion in heavy population pressures, lest it be
ation for the rest of the century. increasingly difficult to remove the shortages
The population explosion that has been ex- of capital, foodstuffs, foreign exchange, and
perienced during the last two decades in most skills that now limit the rate of development.
of the LDCs is in marked contrast with the But the total discussion of this book can be
history of the presently rich industrial na- interpreted as claiming that the potential for
tions when they were in earlier phases of economic development is greater than the po-
their development. In their nineteenth-cen- tential for population growth. To realize this
tury preindustrial phase, the countries of potential as rapidly as possible, we must
western Europe had a population growth that therefore recognize the beneficial effects that
was generally less than half the rate now ex- can come from declining fertility, and con-
isting inthe poor countries. Moreover, in the sider the inclusion of policies of family plan-
past the decline of the death rate in indus- ning as a complement to other policies of de-
trialized countries was mainly due to the de- velopment planning. We must also give due
velopment process itself — through such fac- weight to the belief that rapid population
tors as improved diet, better housing, growth is a consequence as well as a cause of
sanitation. The death rates declined as part poverty. Declining fertility is highly corre-
of the general evolution of Western society — lated with a reduction in unemployment, im-
induced by improvements in economic con- proved status of women, better health care,
ditions and accompanied by changes in social more education, and greater income.
attitudes that allowed birth rates to begin Many demographers believe the time is
falling before death rates reached their low dropping rapidly for the demographic tran-
levels. Now, however, mortality rates are sition from high to low birth and death rates.
falling in poor countries not because of de- One major study of recent trends in natality
velopment, but because modern medical in the LDCs concludes the following.
knowledge and scientific techniques of death The data support the following conclusions about
control can be readily transferred from the recent changes in natality in less developed
rich countries and applied in nonindustrial
regions:
areas. Modern medical and public health ad-
vances have stimulated declines in mortality 1. A rapidly growing number of countries have
been entering the demographic transition on
independently of economic development and
the natality side especially since 1970.
social change.
A crucial question now is whether popula- 2A general fertility rate is number of births per
tion control policies can also stimulate de- 1000 women in reproductive ages 15-49.
 572 HUMAN-RESOURCE DEVELOPMENT

2. Once a sustained reduction of the birth rate has relate socioeconomic measures and fertility
begun, it proceeds at a much more rapid pace "across the board" for all less developed countries
than it did historically in Europe and among have sometimes led to confusing results.
Europeans overseas. Finally, is there indeed a new or renewed de-
3. The "new" countries may reduce birth rates mographic transition? The evidence suggests that
quite rapidly despite initially higher levels than there is. A rapidly growing number of countries of
existed historically in Western Europe. diverse cultural background have entered the na-
4. Where available, the more refined measures of tality transition since World War II and after a
fertility, standardizing for differences in age 25-year lapse in such entries. In these countries the
structure, yield results similar to those for transition is moving much faster than it did in Eu-
crude birth rates. rope. This is probably related to the fact that prog-
5. There is no direct evidence yet that current fer- ress in general is moving much faster in such
tility reductions will terminate at levels signifi- matters as urbanization, education, health, com-
cantly higher than those achieved in European munications and often per capita income. If prog-
countries and Japan. ress in modernization continues, notably in the
larger countries, the demographic transition in the
Above observations are now based on the expe- less developed world will probably be completed
rience ofmany countries, though the precise extent much more rapidly than it was in Europe.
of fertility decline is often clouded by defective It would be foolhardy, however, not to end on a
data. There is now clear evidence of fertility re- word of caution. On any assumptions concerning
duction inthe largest countries, dramatically rapid the reduction of fertility that may occur with so-
in China, and measurably in India and in Indone- cioeconomic progress, it still follows that one may
sia. Fertility reduction is now general in East Asia anticipate and must accommodate an enormous
and in Latin America. As yet there is little evi- increase in the world population and that these in-
dence of this in tropical Africa or in the Moslem creases will be greatest precisely in those countries
Middle East and Pakistan and Bangladesh. economically least well-equipped to absorb the in-
The relationship between socioeconomic vari- crease innumbers.3
ables and fertility is clearly different within the dif-
ferent major cultural regions of the less developed
world; quite different levels and kinds of develop-
ment are associated with fertility reductions, for 3Dudley Kirk, "A New Demographic Transition?"
in Study Committee, National Academy of Sciences,
example, in East Asia and in Latin America. This Rapid Population Growth, Baltimore, 1971, pp. 145—
confirms common sense and explains why efforts to 6. Revised by the author for this edition.

IX.A.2. The Demographic

Momentum *

First, we indicate the likely demographic sit- with regard to (a) the technology of fertility
uation for the remainder of the century "in regulation and (b) the programmatic inter-
the natural course of events." Then we indi- ventions on fertility, mortality, and migra-
cate the possible developments and options tion. We then conclude with some overall
observations.
What can be said with considerable cer-
tainty isthat from now to the year 2000 —
*From Bernard Berelson et al., "Population: Cur-
rent Status and Policy Options," The Population absent a great catastrophe, virtually of nu-
Council Center for Policy Studies, Working Paper clear dimensions — only more population is
No. 44, May 1979, pp. 32-6. Reprinted by ahead for the developing sector and for the
permission. world. As the reader knows, the demographic
 573 POPULATION

momentum inherent in a young age structure fore the world and its major areas will
will carry waves of population growth into achieve the comfortable state of low fertility,
the new century even if replacement fertility low mortality, and low growth that is desir-
were to be achieved everywhere tomorrow — able in the short run and required in the long
which is of course unthinkable under any re- run; and (3) rapid urbanization will continue
alistic consideration. Even if fertility contin- in most developing countries.
ues to decline moderately in the developing Now let us break down the population
world and does not rise much in the devel- growth situation for the larger developing
oped, and even if mortality declines more countries — 29 countries with 10 million or
slowly, substantial demographic momentum more population and about 85 percent of the
will still be carried over into the next cen- developing sector. How close are they likely
tury— at that time at the doubling rate of ap- to come to a target of CBR 20, or roughly a
proximately 40-50 years. two-and-one-half-child family, by the year
2000, and thus to a rate of natural increase
The latest projections of the world's popu-
lation in 2000, as against about 4.2 billion close to 1 percent (assuming the U.N. projec-
now, range from a low of 5.9 billion to a high tion of a crude death rate of 8 to 10 at that
of 6.6, or an increase in the 40-60 percent time)? A current analysis classifies the coun-
range. Most of that increase is of course in tries by their prospect of reaching that goal
the developing world: from 3 billion now to on the basis of their present demographic
from 4.5 to 5.2 billion, for an increase in the trends, their status on presumed basic deter-
50-70 percent range. The projected annual minants offertility (like literacy and educa-
growth rates at that time range from 0.9 per- tion, life expectancy and infant mortality,
cent to 1.7 percent, with crude death rates GNP and nonagricultural employment, fe-
from 8 to 10 and crude birth rates from 18 to male status and age at marriage), and their
25. programmatic efforts, into four categories:
Given the present state of the art, it does
not seem warranted to go into the details of The Certain: South Korea, Taiwan,
the matter since there is no scientific way to Chile
choose among the projections anyway: India Their CBRs are now in the low 20s, they
and China could rebound upward, some large have low mortality and infant mortality,
high-fertility countries could stay that way. they already have favorable social settings
some countries with improved health could with improvement continuing, they have
increase their fertility. Suffice it to say that no major retarding effect from ethnic tra-
the lower projections extrapolate the most re- ditions, they have the requisite family
cent declines ... or stress the impact of fam- planning program effort, they are riding
ily planning programs ... or rest on expected substantial downward trends in fertility of
social trends like entry of women into the at least 15 years' duration. As of now, this
labor market, urbanization, and shortages of group has only 2 percent of the population
commodities . . . and that the higher ones of the developing world.
take more pessimistic views in general.
The Probable: China, Brazil, Mexico,
But there are three main points on which
such work seems to agree, all of considerable the Philippines, Thailand, Turkey, Colom-
bia, Sri Lanka, Venezuela, Malaysia
importance: (1) the trend in fertility and
growth is now down, probably on an historical These countries now have CBRs in the 30s
or below, their crude death rates are in the
scale, although the age structure will be a re-
tarding burden for the rest of the century on 6-1 1 range, they have favorable social set-
the order of, say, a 5-10 percent drag on fer- tings with advances in progress almost ev-
tility reduction in countries experiencing de- erywhere, only one (Moslem Turkey) has
clines; (2) there is still a long way to go be- a high-fertility ethnic tradition and that
 574 HUMAN-RESOURCE DEVELOPMENT

country is rapidly modernizing, only six high, an unfavorable and only slowly
have the requisite program effort but Mex- changing socioeconomic setting, Moslem
ico is now vitalizing its program and Bra- and/or African high fertility traditions, no
zil, Turkey, and Venezuela have the ad- real family planning effort in recent years.
ministrative potential to do so, all appear This group, with 13 percent of the popu-
to be on a downward trend which in some lation, has no discernible chance to achieve
cases is quite sharp although their recency the stipulated target under expectable cir-
leaves a few somewhat suspect. This group cumstances, and by present indications
has 42 percent of the total developing pop- would do well to reach CBR 35 by the end
ulation. They will come close to the target, of the century.
or could if they would.
Actually, if the Certain and the Probable
were to average CBR 20 by the end of the
The Possible: India, Indonesia, Egypt, century, the Possible CBR 25, and the Un-
Peru
likely CBR 35 — with the remaining coun-
Here is a mixed picture: India and Egypt tries at CBR 30— the birth rate for the de-
are in the upper 30s in CBR, India, Indo- veloping sector would then be about CBR 25
nesia, and Egypt had impressive fertility (identical with the U.N.'s low variant) and a
declines recently, India had and Indonesia growth rate still about 1.6 percent. Natu-
has a good family planning program for its rally, inany such projection a great deal de-
setting but the former is now in a relapsed pends upon China, India, and Indonesia
position from which it may not soon re- (with 54 percent of the developing popula-
cover. On the other hand, death rates are tion), and upon a few other large countries
still high, social factors are mixed to un- (over 50 million) — Brazil, Bangladesh, Pa-
favorable, and two countries are Moslem. kistan, Nigeria, and Mexico — that all to-
This group has 28 percent of the develop- gether make up two-thirds of the developing
ing population. They have a chance of world. But as of now, CBR 20 and a popula-
reaching the target, but it will take some tion growth rate of one percent in the devel-
doing, and given the difficulties in the two oping sector by 2000 seem out of reach short
large countries, a CBR of 25 for this group of heroic efforts, and perhaps not even then.
would be a signal achievement. As a rough rule of thumb, in the modern era
it takes an average of about 20-25 years —
say, one generation — for a country to move
The Unlikely: Bangladesh, Pakistan, Ni- from CBR 35 to CBR 20. Of the Possible and
geria, Iran, Zaire, Afghanistan, Sudan,
Unlikely countries, all but the four sub-Sa-
Morocco, Algeria, Tanzania, Kenya, Nepal haran Africa countries (Nigeria, Zaire, Tan-
In these countries, all the indicators are zania, Kenya) will be at or below CBR 35 by
unfavorable: birth rates from the mid-40s
century's end, according to the U.N. medium
up and stable so that the initial break in projection. Meanwhile, the demographic mo-
fertility is still to come, death rates still mentum continues.

Comment
An excellent review of population trends in the major regions of the developing world is
presented by Robert H. Cassen, "Current Trends in Population Change and Their Causes,"
Population and Development Review, Vol. 4, No. 2 (1978). It indicates the importance of
socioeconomic, cultural, and environmental factors in explaining fertility differences not only
among countries but also within countries. It suggests that the factors commonly associated
with fertility decline — such "correlates of fertility decline" as education, urbanization, im-
proved status of and wider employment opportunities for women, mortality declines, and in-
 575 POPULATION

creased practice of family planning — contribute to fertility differentials within countries even
in those instances in which the prevailing culture appears to give a disposition to high fertility.
An additional survey is presented in Robert H. Cassen, "Population and Development: A
Survey," World Development, Vol. 4, Nos. 10/11 (1976). This considers the macro-aspects
of the effects of population growth on the economy, on real output, food, employment, income
distribution, health, education, and urban development An extensive bibliography is also
included.

EXHIBIT IX. 1. Estimated Age Structure of the Population by Region, 1975

Percentages
5-14 65 +
Age 0-4 15-64

Eastern Africa 18.5

26.4 52.4
17.0 25.6 54.5 2.7
2.9
Middle Africa0
West Africa 18.4 26.4 2.5
Southern Africa 17.5 23.5 52.7
55.1 3.9
North Africa 17.1 27.2 52.5 3.3
Middle East (incl. Turkey) 26.0 53.0
17.1
16.5 3.9
2.9
South Asia (incl. Iran) 26.4 54.1
Southeast Asia (incl. Burma, Philippines) 17.0 26.6 3.0
Japan) 53.4
East Asia (excl. China, Taiwan, 13.3 24.8 58.3 3.6
Caribbean 14.5 26.2 54.2
17.9 27.9 50.9 5.1
Central America (incl. Mexico)
16.5 26.6 3.4
Tropical South America 53.8 7.3
10.6 19.8 62.3 3.1
Temperate South America 7.9
North America 17.6 64.3 10.2
Europe 7.7 16.2 12.9
7.8
63.8
Japan 9.1 15.4 67.7

Source: United Nations, Selected World Demographic Indicators by Countries, 1950-2000, May 1975.
"Includes Angola, Central African Empire, Chad, Congo, Equatorial Guinea, Gabon, Cameroon, and
Zaire.

EXHIBIT IX.2. The Demographic Transition

/ z
50
Birth rate Death rate Developing countries
o

3 40 h-

2050
1750 1800 1850 1900 1950 2000

Rate of population increase = birth rate — death rate

"Crude birth and death rates. The projected increases in death rates after about 1980 reflect the rising
proportion of older people in the population.
^Includes industrialized countries, the USSR, and Eastern Europe.
Source: World Development Report, 1982.
 576 HUMAN-RESOURCE DEVELOPMENT

EXHIBIT IX.3. Fertility in Relation to Income: Developing Countries, 1978

Kenya
Nepal
. . Nigeria Algeria
- 1Pakistan .
• • _ ,. . • Ivory Coast
Ghana* Bolivia
r • Nicaragua
£ 6HBangladesn\. pem Mexico
2 5 India Egypt /^<^
" #»E1 Salvador
Indonesia fha.iandTurkey Norm for 87 developing countries

2 4 Colombia* "Malays
Sri Lanka
TO -3

Korea • Argentina

2 — China

$500 SI 000 $1500 $2000 $2500 $3500

$3000
GNP per person, 1978 (U.S. dollars)

!The total fertility rate indicates the average number of children that
would be born to each woman in a population if each were to live through
her childbearing years (usually considered ages 15-49) bearing children
at the same rate as women of those ages actually did in a given year.
^Estimated fertility and GNP for China are for 1979.

IX.A.3. Fertility Determinants*

Certain variables have emerged as consis- mortality on fertility, that is, countries with
tently important in their relationship to fer- high rates of infant mortality on the whole
tility. A critical few describe characteristics have high rates of fertility. (Infant mortality
of the socioeconomic environment that can be is an important component of overall mortal-
altered through public policy: infant mortal- ity, particularly in developing countries.)
ity, female education and labor force partic- There is no country with high mortality and
ipation, availability of family planning ser- low fertility, taking high mortality as annual
vices, and, possibly, income distribution. The death rates above 15 per thousand and low
strength and direction of the effect on fertil- fertility as annual birth rates below 30 per
ity of these policy variables is very different thousand.
for countries at different stages of develop- What is the causal mechanism that links
ment and with different cultural envi- high fertility and high infant mortality? On
ronments. the one hand, high infant mortality results in
high fertility because parents who experience
INFANT MORTALITY AND child loss early ultimately may more than re-
FERTILITY place lost children, and parents in high-mor-
tality communities may insure themselves
Almost all studies of the determinants of
against future child loss by having more chil-
fertility indicate a positive effect of infant dren than they would want. On the other
hand, high fertility contributes to high infant
♦From Nancy Birdsall, "Analytical Approaches to (and child) mortality because close spacing
the Relationship of Population Growth and Develop-
ment," Population and Development Review, March
of births and many births may deplete the
and June 1977, pp. 85-90. Reprinted by permission. mother's physical resources and reduce the
 577 POPULATION

family's per capita financial resources; a south Asia; most countries of Latin America
commonly noted phenomenon is the death of now have much lower mortality and fertility
a child when, on the birth of a subsequent rates.) Where political and cultural barriers
child, the mother ceases breastfeeding the to the advocacy of family planning and other
older child. fertility-reduction efforts exist, reducing in-
The major point in the present context is fant mortality can be expected to lay the
that the relationship between infant mortal- groundwork
fertility. for later efforts concentrating on
ity and fertility differs by country and by pre-
vailing economic conditions within countries.
For example, high fertility is most likely to FEMALE EDUCATION AND LABOR
contribute to high mortality among the poor- FORCE PARTICIPATION
est groups and is most likely to show up on a
country basis in the poorest countries, as in- Female education bears one of the strong-
deed itdoes in Africa. There kwashiorkor is est and most consistent negative relationships
"the disease that kills the child whose mother to fertility for a variety of reasons: through
carries another child in her womb"; it is often its effect on raising age of marriage; because
related to malnutrition exacerbated by early it may improve the likelihood that a woman
or sudden weaning. has knowledge of and can use modern contra-
Opinion differs as to whether the poorest ceptives; and because it has some intangible
countries, even with a strong replacement ef- effect on the woman's ability to plan, her in-
fect, end up with more or fewer people as terest innonfamilial activities, and so on. No
mortality falls. Studies that have dealt with need to invoke fertility reduction to justify
the timing of the relationship, however, indi- improving educational opportunities for
cate that the response of lower fertility to women: better educated women will be more
lower mortality is maximized when the inci- productive workers, better parents, and bet-
dence of infant and child mortality is lagged ter-informed citizens; however, where male/
two to four years. This tends to support the female student ratios indicate that women
opinion that reduced infant mortality will in suffer some schooling disadvantage, fertility
the long run reduce fertility enough so that effects provide additional justification for
the rate of natural increase will go down. rectifying the imbalance. The fertility-reduc-
This view would be strengthened if studies ing effect of women's education holds true
were better able to incorporate the indirect even for the highest levels of education;
effect of reduced mortality: lower death prob- women who obtain secondary and higher ed-
abilities induce parents to invest more in ucation marry later, and increasing the age
their children, which in turn leads them to at marriage has a pronounced effect on a
lower their desired fertility. They replace a country's fertility rate.
large quantity of children with fewer children But female labor-force participation ap-
of higher quality. In other words, lower mor- pears to have an independent effect on fertil-
tality not only assists parents to achieve de- ity only for those women who work in high-
sired family size with fewer births; it may prestige, modern-sector jobs. High rates of
lead parents to reduce their desired family female labor-force participation, like vir-
size. tually all other variables, are neither a nec-
Reduction of infant mortality is an impor- essary condition for fertility decline (consider
tant policy objective on its own merits. Be- Korea, Turkey) nor a sufficient condition for
cause countries with high average levels of it (consider countries of West Africa). On
infant and child mortality also have high lev- the other hand, increasing opportunities for
women to work in the modern labor force can
els of fertility, the policy objectives of reduc-
ing both mortality and fertility can be mu- accelerate a fertility decline; where women
tually reinforcing. (Countries with the may desire more children than their hus-
highest rates are primarily in Africa and bands (as is possible in the Middle East and
 578 HUMAN-RESOURCE DEVELOPMENT

Pakistan, where custom deprives most have adequately controlled for the probable
women of opportunities in other endeavors), effect of these nonprogram changes on
fertility.
offering women some other avenue of activity
than child rearing may reduce family size. Advocates of family planning programs
Good earning opportunities, like higher edu- argue that: many programs are still in their
cation, can increase the age of marriage for infancy and have been poorly run; they have
women. absorbed tiny proportions of national budgets
From the point of view of policy, an im- and of foreign aid expenditures; they provide
portant conclusion emerging from these anal- many secondary benefits — improved health
yses is the highly tentative nature of the ef- for mothers and infants, increased control by
fect of "status of women" on fertility. But women over their own bodies, and greater
"status of women" does not define an area control by families over their own future.
where public intervention is possible; educa- Moreover, support of voluntary family plan-
tion and jobs for women do. Improving wom- ning programs is one of the few widely ac-
en's opportunities for education and for mod- cepted direct interventions governments have
ern jobs, like reducing infant mortality, has made to reduce their rates of population
its own justification; piggybacking its fertil- growth; the immediate question is not
ity-reducing benefit onto programs and proj- whether support is warranted, but how much
and in what form. What is the effectiveness
ects geared to improving benefits
creases the measured women's of
livessuch
in-
of spending on family planning relative to
projects relative to their given costs. other expenditures that also have both gen-
eral development and fertility-reducing ef-
FAMILY PLANNING SERVICES fects? Only investigations on a country basis
AND FERTILITY can begin to answer such a question.

Do family planning services reduce fertil- INCOME DISTRIBUTION

ity more efficiently and more effectively than AND FERTILITY
general development programs? The wide
range of estimates of the cost of averting a Will an improvement in the distribution of
birth ($l-$400) and its benefits ($100-$900) income within a country reduce its fertility
illustrates the difficulty of answering such a rate? In an analysis of 1960-1965 data on 64
question. Estimating the costs of averted countries, Repetto reports an elasticity of the
births requires estimating the number of general fertility rate with respect to the share
births averted because of a family planning of income received by the poorest 40 percent
program, itself a complicated task; some of-.20.
households of —.36, compared to an elas-
births might have been averted anyway if ticity ofthe general fertility rate with respect
some couples are substituting publicly pro- to increase in average per capita income of
vided contraceptives for private efforts. And
"costs per acceptor" is an unreliable measure Income distribution data are notoriously
since acceptors may make repeated visits in
poor, and Repetto's result suffered in a sub-
different areas, change contraceptives, and sequent estimation in which he omitted sev-
otherwise muddy the attendance statistics. eral East European countries (which have
Even where the existence of a family plan- both low fertility and relatively equal income
ning program is correlated over time with a distributions); the income distribution varia-
decline in fertility, assurance that the pro- ble was no longer statistically significant in
gram itself caused the decline requires the explaining the fertility variable. Moreover,
systematic elimination of such other possible whether such cross-section results should be
causes as increases in income, changes in oc- used to predict the pattern of a relationship
cupational structure, and increases in edu- over time is highly questionable.
cation. Few studies of the effect of programs There is thus no real indication that indi-
 579 POPULATION

viduals' fertility behavior is affected by their to hold. Fertility-reducing effects can be

relative income per se, over and above their added to the other benefits of raising income
absolute income; the income distribution-fer- levels of the poorest groups,
tility link found is more probably the result Highlighted above are four areas of public
of a coincidence over the period since World policy intervention. All may contribute to re-
War II between improvements in the level of ducing fertility at tolerable cost; all provide
absolute income of the poorest and increases benefits in addition to fertility reduction,
in their relative share. In any event, insofar Careful design of policy in these areas can
as "income distribution" is used to mean accelerate the fertility decline that general
some nexus of programs, including income improvement in standards of living inaugu-
increases, to alleviate the poverty of the poor- rates,
est, an inverse relationship to fertility seems

Comment
The major lines of research on the multiple relationships between population growth and
economic development have been directed to the determinants and consequences of fertility.
The research can be further divided into macro- and micro-type research. An excellent survey
of several research studies is presented by Nancy Birdsall, "Analytical Approaches to the
Relationship of Population Growth and Development," Population and Development Review
(March 1977; June 1977).
Of particular interest in the macro-consequences studies is the attempt to compare costs of
a family planning program to the projected savings realized in health or education costs. Usu-
ally the savings in the latter are shown to "pay" for the former in this form of analysis. Such
comparisons, however, necessitate assumptions regarding costs of launching family planning
programs, acceptor rates, and the relationship between acceptor rates and actual births
averted, as well as assumptions about future costs in health, education, or other areas, and
about the society's welfare function, which may value children per se in addition to capital
goods.
A number of studies have considered the benefits and costs of poopulation control programs:
A. J. Coale and E. M. Hoover, Population Growth and Economic Development in Low-Income
Countries (1958); P. M. Hauser (ed.), The Population Dilemma (1963); Bernard Berelson
(ed.), Family Planning and Population Programs (1966); Goran Ohlin, Population Control
and Economic Development (1967); Dudley Kirk and Dorothy Nortman, "Population Policies
in Developing Countries," Economic Development and Cultural Change (January 1967); J.
E. Meade, "Population Explosion, The Standard of Living, and Social Conflict," Economic
Journal (June 1967); J. Spengler, "The Economist and the Population Question," American
Economic Review (March 1966); R. A. Easterlin, "Effects of Population Growth on the Eco-
nomic Development of Developing Countries," The Annals, (January 1967); Harvey Leiben-
stein, "Socio-Economic Fertility Theories and Their Relevance to Population Policy," Inter-
national Labour Review (May-June 1974); Harvey Leibenstein, "An Interpretation of the
Economic Theory of Fertility: Promising Path or Blind Alley," Journal of Economic Litera-
ture (June 1974); Harvey Leibenstein, "Pitfalls in Benefit-Cost Analysis of Birth Prevention,"
Population Studies (June 1969); Timothy King et al., Population Policies and Economic De-
velopment (1974); Philip M. Hauser (ed.), World Population and Development (1979); R. A.
Easterlin (ed.), Population and Economic Change in Developing Countries (1980).

Comment
Besides the macro-research involving benefit-cost analysis, there has been more research
on how individual fertility decisions are affected by environmental changes. The "new eco-
 580 HUMAN-RESOURCE DEVELOPMENT

nomics of the household" treats the child as both a produced (investment) and consumer good.
Fertility is the result of rational economic choice within the household.
This economic model concludes that, for poor families in developing countries, children
entail low net costs and, in the extreme case, may actually be a net benefit. In contrast, the
literature dealing with macro-consequences concludes that high fertility entails a high net
cost to poor societies. This theoretical gap between the low private and high social costs of
children has been a principal justification for government policies to reduce fertility. On the
micro-models, see Theodore W. Schultz (ed.), Economics of the Family: Marriage, Children
and Human Capital (1974).

IX. A. 4. Economics of
Population Quality*

While land per se is not a critical factor in rigorous definition of human capital, it will
being poor, the human agent is: Investment be subject to the same ambiguities that con-
in improving population quality can signifi- tinue to plague capital theory in general and
cantly enhance the economic prospects and the capital concept in economic growth
the welfare of poor people. Child care, home models in particular. Capital is two-faced,
and work experience, the acquisition of infor- and what these two faces tell us about eco-
mation and skills through schooling and in nomic growth, which is a dynamic process,
other ways consisting primarily of investment are, as a rule, inconsistent stories. It must be
in health and schooling can improve popula- so because the cost story is a tale about sunk
tion quality. Such investments in low-income investments, and the other story pertains to
countries have, as I shall show, been success- the discounted value of the stream of services
ful in improving the economic prospects that such capital renders, which changes with
wherever they have not been dissipated by the shifting sands of growth. But worse still
political instability. Poor people in low-in- is the capital homogeneity assumption under-
come countries are not prisoners of an iron- lying capital theory and the aggregation of
clad poverty equilibrium that economics is capital in growth models. As Hicks (1965)
unable to break. There are no overwhelming has taught us, the capital homogeneity as-
forces that nullify all economic improve- sumption isthe disaster of capital theory.
ments, causing poor people to abandon the This assumption is demonstrably inappro-
economic struggle. It is now well documented priate inanalyzing the dynamics of economic
that in agriculture poor people do respond to growth that is afloat on capital inequalities
better opportunities. . . . because of the differences in the rates of re-
I now turn to measurable gains in the qual- turn, whether the capital aggregation is in
ity of both farm and nonfarm people (Schultz terms of factor costs or in terms of the dis-
19796, 1979c). Quality in this context con- counted value of the lifetime services of its
sists of various forms of human capital. I many parts. Nor would a catalog of all exist-
have argued elsewhere (Schultz 1974) that, ing growth models prove that these inequali-
while a strong case can be made for using a ties are equals. But why try to square the cir-
cle? If we were unable to observe these
inequalities, we would have to invent them
*From Theodore W. Schultz, "Nobel Lecture: The because they are the mainspring of economic
Economics of Being Poor," Journal of Political Econ-
omy, Vol. 88, No. 4, 1980, pp. 643, 645-8. Reprinted growth. They are the mainspring because
by permission. they are the compelling economic signals of
 581 POPULATION

growth. Thus, one of the essential parts of ment toward quality contributes to the solu-
economic growth is concealed by such capital
tion of the population "problem."
aggregation.
The value of additional human capital de- INVESTMENT IN HEALTH
pends on the additional well-being that
human beings derive from it. Human capital
Human capital theory treats everyone's
contributes to labor productivity and to en- state of health as a stock, that is, as health
trepreneurial ability. This allocative ability is capital and its contribution as health services.
valuable in farm and nonfarm production, in Part of the quality of the initial stock is in-
household production, and in the time and herited and part is acquired. The stock de-
other resources that students allocate to their preciates over time and at an increasing rate
education. It is also valuable in migration to in later life. Gross investment in human cap-
better job opportunities and to better loca- ital entails acquisition and maintenance
tions in which to live. It contributes impor- costs. These investments include child care,
tantly to satisfactions that are an integral nutrition, clothing, housing, medical services,
part of current and future consumption. and the use of one's own time. The flow of
My approach to population quality is to services that health capital renders consists
treat quality as a scarce resource, which im- of "healthy time" or "sickness-free time,"
plies that it has an economic value and that which are inputs into work, consumption, and
its acquisition entails a cost. In analyzing leisure activities (Grossman 1972; Williams
1977).
human behavior that determines the type and
amount of quality that is acquired over time, The improvements in health revealed by
the key is the relation between the returns the longer life span of people in many low-
from additional quality and the costs of ac- income countries have undoubtedly been the
quiring it.When the returns exceed costs, the most important advance in population qual-
stock of population quality will be enhanced. ity. Since about 1 950, life expectancy at birth
This means that increases in the supply of has increased 10 percent or more in many of
any quality component are a response to a de- these countries. People of western Europe
mand for it. It is a supply-demand approach and North America never attained so large
to investment behavior because all quality an increase in life expectancy in so short a
components are here treated as durable period. The decline in mortality of infants
scarce resources that are useful over some pe- and very young children is only part of this
riod of time. achievement. The mortality of older children,
My hypothesis is that the returns to var- youths, and adults is also down.
ious quality components are increasing over Ram and Schultz (1979) deal with the eco-
time in many low-income countries: the rents nomics of these demographic developments
that entrepreneurs derive from their alloca- in India. The results correspond to those in
tive ability rise, as do the returns to child other low-income countries. In India from
care, schooling, and improvements in health. 1951 to 1971 life expectancy at birth of
Furthermore, the rates of return are en- males increased by 43 percent and that of fe-
hanced bythe reductions in the costs of ac- males by 41 percent. Life spans over the life
quiring most of these quality components. cycle after age 10, 20 and on to age 60, for
Over time the increases in the demand for both males and females in 1971, were also
quality, in children and on the part of adults decidedly longer than in 1951.
in enhancing their own quality, reduce the The favorable economic implications of
demand for quantity; that is, quality and these increases in life span are pervasive.
quantity are substitutes, and the reduction in Foremost are the satisfactions that people de-
demand for quantity favors having and rear- rive from longer life. While they are hard to
ing fewer children (Becker and Tomes 1976; measure, there is little room for doubt that
Rosenzweig and Wolpin 1978). The move- the value of life expectancy is enhanced.
 582 HUMAN-RESOURCE DEVELOPMENT

Measurement, however, is not impossible. occurs in the case of expenditures on health,

Usher (1978) devised an ingenious extension both on public and private account.
of theory to determine the utility that people
derive from increases in life expectancy. His
empirical analysis indicates that the addi- References
tional utility increases substantially the value Becker, Gary S., and Tomes, Nigel, "Child
of personal income. Endowments and the Quantity and Qual-
Longer life spans provide additional incen- ity of Children," J.P.E. Vol. 84, No. 4, Au-
tives to acquire more education as invest- gust 1976, S143-S162.
ments in future earnings. Parents invest more Government of India, Planning Commission,
in their children. More on-the-job training Draft Five Year Plan 1978-83, New Delhi,
becomes worthwhile. The additional health 1978.
capital and the other forms of human capital Grossman, Michael, The Demand for
tend to increase the productivity of workers. Health, National Bureau of Economic Re-
Longer life spans result in more years of par- search, Occasional Paper No. 119, New
ticipation inthe labor force and bring about York, Columbia University Press, 1972.
a reduction in "sick time." Better health and Hicks, John R, Capital and Growth, Oxford,
vitality of workers in turn lead to more pro- Clarendon Press, 1965.
ductivity per man hour at work. Ram, Rati, and Schultz, Theodore W., "Life
Span, Health, Savings, and Productivity,"
INVESTMENT IN EDUCATION Economic Development and Cultural
Change, Vol. 27, April 1979, pp. 399-421.
Education accounts for much of the im- Rosenzweig, Mark R., and Evenson, Robert
provements inpopulation quality. But reck- F., "Fertility, Schooling and the Economic
oning the cost of schooling, the value of the Contribution of Children in Rural India:
work that young children do for their parents
An Econometric Analysis," Econometrica,
must be included. Even for the very young Vol. 45, July 1977, pp. 1065-79.
children during their first years of school, Rosenzweig, Mark R., and Wolpin, Kenneth
most parents forego (sacrifice) the value of I. "Testing the Quantity-Quality Fertility
the work that children perform (Rosenzweig Model: The Use of Twins as a Natural Ex-
and Evenson 1977). Another distinctive at- periment," mimeographed, New Haven,
tribute of schooling is the vintage effect by Conn., Yale University, Econ. Growth
age over time. Starting from widespread illit- Center, October 1978.
eracy, as more schooling per child is achieved Schultz, Theodore W., Transforming Tradi-
the older adults continue through life with tional Agriculture, New Haven, Conn.,
little or no schooling, whereas the children on Yale University Press, 1964. (Reprint:
entering into adulthood are the beneficiaries. New York, Arno, 1976.)
The population of India grew about 50 per- , "Human Capital: Policy Issues and
cent between 1950-51 and 1970-71. School
Research Opportunities," in Human Re-
enrollment of children ages 6-14 rose over sources, New York, Columbia University
200 percent. The rate of increase in second- Press (for National Bureau of Economic
ary schools and universities was much higher Research), 1972.
(Government of India 1978). Since schooling (ed.), Economics of the Family:
is primarily an investment, it is a serious Marriage, Children and Human Capital,
error to treat all schooling outlays as current Chicago, University Chicago Press, 1974.
consumption. This error arises from the as- -, "The Value of the Ability to Deal
sumption that schooling is solely a consumer
with Disequilibria," Journal of Economic
good. It is misleading to treat public expen- Literature, Vol. 13, September 1975, pp.
827-46.
turesditures
and ason schooling
a use of as "welfare"
resources that expendi-
has the ■, "On Economics, Agriculture, and
effect of reducing "savings." The same error the Political Economy," in Theodore Dams
583 POPULATION

and Kenneth E. Hunt (eds.), Decision- , "The Value of Higher Education in

Making and Agriculture, 16th Interna- Low Income Countries: An Economist's
tional Conference of Agricultural Econo- View," Paris, International Institute of
mists, Nairobi, Kenya, Oxford, Alden, Educational Planning, 1979. {d)
1977. , "On the Economics of the Increases
'On Economics and Politics of Ag- in the Value of Human Time over Time,"
riculture," inTheodore W. Schultz (ed.), R. C. O. Matthews (ed.), in Measurement,
Distortions of Agricultural Incentives, History and Factors of Economic Growth,
Bloomington, Indiana University Press, Fifth World Conference of the Interna-
1978. (a) tional Economic Association. (London:
Macmillan, 1980.)
, "What Are We Doing to Research
Entrepreneurship?" in Transforming Usher, Dan, "An Imputation to the Measure
Knowledge into Food in a Worldwide Con- of Economic Growth for Changes in Life
text, Minneapolis, Miller, 1978. (b) Expectancy," in Milton Moss (ed.), The
'Concepts of Entrepreneurship and Measurement of Economic and Social Per-
Agricultural Research," Kaldor Memorial formance, New York, Columbia Univer-
Lecture, Agricultural Economic Work- sity Press (for National Bureau of Eco-
shop, University Chicago paper No. 79, p. nomic Research), 1978.
26, Ames, Iowa State University, 1979 (a) Welch, Finis, "Education in Production,"
'Investment in Population Quality J.P.E., Vol. 78, No. 1, January-February
throughout Low-Income Countries," in 1970, pp. 35-59.
Philip M. Hauser (ed.), World Population , "The Role of Investments in
and Development: Challenges and Pros- Human Capital and Agriculture," in
pects, Syracuse, N.Y., Syracuse University Theodore W. Schultz (ed.), Distortions of
Press, 1979. (b) Agricultural Incentives, Bloomington, In-
Reckoning the Economic Achieve- diana University Press, 1978.
ments and Prospects of Low Income Coun- Williams, Alan, "Health Service Planning,"
tries," Snyder Memorial Lecture, West in Michael J. Artis and A. R. Nobay
Lafayette, Ind., Purdue University, 1979. (eds.), Studies in Modern Economic Anal-
ysis, Edinburgh, Blackwell, 1977.
"
(c)
 IX.B. HEALTH AND NUTRITION
IX. B. 1 . Intercountry Variations
in Health*

Health conditions vary more among the less waste management problems. On the other
developed countries than between more and hand, it is relatively literate, its communities
less developed ones. Life expectancy varies are often highly organized, and its public ad-
from about 40 years (Mali, Angola, and ministration isgenerally fairly well devel-
Upper Volta) to about 70 years (Sri Lanka, oped. Wide intercountry variations are found
Singapore, and Argentina). Life expectancy not only in health but in income, education,
generally rises with per capita incomes, thus and climate. Life expectancy for the region,
lending credence to the old notion that eco- excluding China, is about 51 years — some-
nomic development is good medicine. But the what below the average for all developing
high levels of health in a few very poor areas countries. Singapore, Malaysia, Hong Kong,
(Sri Lanka and the Indian state of Kerala, and Korea are industrializing rapidly and
for example) and the low life expectancy in a have life expectancies at birth between 63-72
smaller number of relatively wealthy areas years, whereas Laos People's Democratic Re-
(Brazil and Nigeria) demonstrate that the public has a very low life expectancy at birth
relationship is neither simple nor inescapa- (42 years) and a very high crude death rate
ble. Moreover, the pattern of disease and the (2 per 1,000). [See Table 2.]
scale of morbidity both vary widely across In South Asia, Sri Lanka and the State of
geographic areas. The differences reflect not Kerala in India differ significantly from the
only variations in income levels, environmen- remaining areas of the region. Life expec-
tal sanitation, access to health care, levels of tancy is69 years in Sri Lanka as compared
education, etc., but also such uncontrollable to only 43 years in Nepal. More accessible
factors as climate. The following paragraphs health services, higher literacy, good trans-
attempt to highlight the conditions prevailing portation and communications, and more eq-
in selected regions of the world. uitable distribution of food seem to account
The highest rates of infant and early child- for these two exceptional areas.1 Kerala re-
hood mortality and the lowest life expectan- ports especially high health status, even
cies are reported by the countries of sub-Sa- though incomes are significantly lower than
haran Africa. Life expectancy ranges from Indian national averages. [See Table 3.]
39 years to 54 years (lower than the average The goal of the ten-year health plan
for India). It is estimated that at least a mil- (1970-1980) for the Americas was to in-
lion African children die each year without crease life expectancy at birth by five years
reaching the age of five. In much of Africa, in those countries where it was below 65
half of all deaths occur among children under years in 1970. More than one-third of all
five. [See Table 1.] Latin American countries had surpassed this
The countries of Southeast Asia and the goal by 1977, and all are expected to gain at
South Pacific face the second most grave least three years by 1980. Temperate South
health problems. The region is characterized America is expected to reach 70 years of life
by population pressures, borderline nutri- expectancy by 1980; tropical South America,
tional status, and difficult water supply and with the exception of Bolivia, should reach 65

*From Frederick Golladay and Bernhard Liese, 'Gwatkin, Davidson, "Nutrition Planning and
"Health Issues and Policies in the Developing Coun- Physical Well-being in Kerala and Sri Lanka," Over-
tries," World Bank Staff Working Paper No. 412, seas Development Council, Washington, D.C. Janu-
August 1980. Reprinted by permission. ary 1978.

584
 585 HEALTH AND NUTRITION

TABLE 1 . Health Indicators for Selected African Countries

Deaths per
Births per Deaths per Thousand Life
GNP per Capita Thousand Thousand Infants Expectancy
in U.S. Dollars at Birth
1978 Population Population
1978 1978 1978
Country 1978
45 Aged 0-1
Ethiopia 120 49 22 42
39
120 43
Mali 120 49
130 — 43
45
Somalia 48 20
140 138
Burundi 47 20
Chad 140 44 21 — 46
180 133 42
Rwanda 51
48 19
Upper Volta
160
46 22 — 46
Zaire 210 19 — 46
180 46 142 46
Malawi 52 20 93
Mozambique 140 19 42
162 46
Niger 220 46 22
46
Sierra Leone 210 51
48 19
16 —
Tanzania 230 49 19 — 51
46
Benin 230 —
Lesotho 280 40
45 16 116 50
Madagascar 250 19 53 46
Central African Republic 250 44 19 — 53
Kenya 330 42
Mauritania 270 51
50
45 14
22 —51
Uganda 280 45 14 — 46
53
132
Sudan 320 48 18
23
24 46
41
Angola 300 42
Cameroon 460
48 19 — 48
Ghana 390 63 48
Liberia 460 17
18 159 48
51 163 42
Nigeria 560 50
49 18
158
Senegal 340 22

Source: Infant death rates from World Bank, Health Sector Policy Paper, second edition, February 1980, annex 1; other data
from World Development Report, 1980.

years; Caribbean countries, excluding Do- 1 : 1 in North America, 1 :4 in Central Amer-

minican Republic and Haiti, have surpassed ica, and 1 :2 in South America. Thus, greater
70 years of life expectancy. achievement has been attained in reducing
Deaths per 1,000 population in Latin deaths of 1- to 4-year-olds in Latin America
America decreased from around 1 1 in 1 960
to around 8 in the early 1970s. Although during this period.2
Table 4 below presents selected health in-
there were 1 8 Latin American countries with dicators for countries in the Latin America
crude death rates greater than 10 per 1,000
region.
in 1960, only 6 countries continued to expe-
rience this rate by 1975. The ratio of percent
decreases of deaths under 1 year of age to 2Pan American Health Organization, Health Con-
present decreases of deaths 1-4 years of age ditions in the Americas, 1973-1976. Scientific Pub-
between 1968 and 1974 was approximately lication No. 364, Washington, DC. 1978.
 586 HUMAN-RESOURCE DEVELOPMENT

TABLE 2. Health Indicators for Selected Countries in Southeast Asia and the Pacific

Deaths per
Births per Deaths per Thousand Life
Thousand Thousand Infants Expectancy
GNP per Capita
in U.S. Dollars at Birth
Population
1978 Population
1978 1978 1978
Country 1978
Aged 0-1
22 42
Lao People's 45 70
Democratic Republic 90 —
Indonesia 360 —
China 230 18 176 —
490 37 47
Thailand 32 8 68
Philippines 510 35
41 169 65 60
61
Papua New Guinea 560 — 50
Korea, Republic of 1,160 21 8 63
Malaysia 29 6 31 72
1,090
Hong Kong 19 6 12
37 67
3,040

Source: World Development Report, 1980.

TABLE 3. Health Indicators for Selected Countries in South Asia

Deaths per
Births per Deaths per Life
Thousand
GNP per Capita Thousand Thousand Infants Expectancy
in U.S. Dollars Population at Birth
Population 1978
Country 1978 1978 1978 1978
46 Aged 0-1
90 18 139 43
Bangladesh 45
Nepal 120 21 — 47
India 180 — 51
45
Pakistan 230 35 14
15 — 52
Sri Lanka 190 29 6 —
69
Source: World Development Report, 1980.

HEALTH RESOURCES erage, ratios of population to hospital beds

Problems of health infrastructure devel- are 1 5 to 20 times those in developed coun-
opment and health personnel availability in tries.3 Moreover, these few facilities often
must cover wide geographic expanses and
developing countries vary among regions and
between rural and urban areas within coun- large numbers of people. Inadequate trans-
tries. The following describes, by region, the portation systems and the cost of transport
current status of physical resources in the and lodging further limit actual rural cover-
health sector in terms of both quantity and age. In Kenya, for example, 40 percent of the
distribution. outpatients coming to rural health centers
lived within 8 km; 30 percent lived 8 to 16 km
Africa
Access to health facilities in rural areas of
3World Health Organization, WHO Quarterly Re-
African countries is inadequate. On the av- ports, Geneva, 1979, p. 696.
 587 HEALTH AND NUTRITION

TABLE 4. Health Indicators for Selected Latin American Countries

Deaths per
Births per Deaths per Thousand Life
GNP per Capita Thousand Thousand Infants Expectancy
in U.S. Dollars Population at Birth
Population 1978
Country 1978 1978 1978 1978
Aged 0-1
Honduras 480 47
44 12 118
158
Bolivia 510 15 52
Colombia 850 31 8 98 62
Paraguay 850 39 9 — 57
63
Ecuador 880 44 10 66 60
41 12
Guatemala 910 55
45
Nicaragua 840 13 77
Mexico 38 8 37 70
65
1,290 57
70
Panama 31 6 60
1,290 71
Costa Rica 28 5 28
47
1,540
Brazil 36 9 46
92 62
1,570 20
Uruguay 9 71
1,610
Argentina 21 8 —40 66
1,910
Venezuela 36 7
2,910

Source: World Development Report, 1980.

away; and 30 percent lived more than 16 km are concentrated in the capital (Dakar) area,
where only 17 percent of the population
away.4
The distribution of facilities favors urban
lives.7 Yet expanding the supply of physicians
areas. In Borno State, Nigeria, for example, is prohibitively expensive for most African
where 80 percent of the population lives in nations; medical training alone costs more
villages, only 50 percent of health clinics are than $25,000 per physician exclusive of the
in rural areas.5 There are only 233 persons capital costs of medical schools and teaching
per dispensary in the semiurban areas sur- hospitals.8 Health personnel other than phy-
rounding Lagos, whereas in outlying, rural sicians are also in short supply. On the aver-
states there are between 25,000 and 60,000 age, Africa has only 4 to 8 nurses/midwives
persons per dispensary.6 per 10,000 persons, compared to 63 in North
Africa has a severe shortage of medical America.9 Such shortages impede expansion
personnel, with between 20,000 and 40,000 of health care systems to currently under-
persons per physician. Moreover, medical served populations.
personnel are poorly distributed. The over- Table 5 presents data on health facilities
whelming majority of physicians practice in and personnel for selected African nations.
urban areas where 10 percent to 20 percent
of the population lives, while rural areas are
left virtually unserved. In Senegal, for ex- 7Robin J. Menes, M.H.S., Syncrisis: The Dynam-
ics of Health: Vol. XIX, Senegal.
ample, 76 percent of the nation's physicians
8U.S. Department of Health, Education and Wel-
4Ministry of Health, Kenya, Rural Health Strat- fare, Public Health Services, Division of Program
egy, 1973. Analysis, June 1976, p. 96.
5Derived from Facilities and Manpower Survey, 'World Bank, Health Sector Policy Paper. March
Federal Ministry of Health, Lagos, Nigeria, 1978. 1975, p. 34. Deprived from World Health Organiza-
^Facilities and Manpower Surveys, Federal Min- tion, WHO Annual Statistics. 1973-76. Vol. Ill,
istry of Health, Lagos, Nigeria, 1978. table 2.5.
 588 HUMAN-RESOURCE DEVELOPMENT

TABLE 5. Health Resources in Selected African erage 455: 1 with Bangladesh and Nepal hav-
Countries
ing the greatest shortages.12 Table 6 presents
comparative infrastructure and personnel
Population data for countries of South Asia.
Per
Per Per Nurse/
Hospital
Country Bed Physician Midwife Southeast Asia and Pacific
The health facilities in South Asia and the
Ethiopia 3,080 69,340 22,320
Niger 1,200 55,420 Pacific are relatively well developed, partic-
6,790
Mauritania 2,320 15,150 1,580 ularly at the hospital level. Ratios of popu-
Burundi 810 45,430 5.420 lation per hospital bed for Thailand, the Phil-
Sudan 1,110 12,680 980 ippines, and Papua New Guinea are among
Togo 680 20,770 the lowest in the developing world outside of
2,530
Ghana 600 10,510
2,530 Latin America. Major programs of expan-
Nigeria 1,170 14,810 1,620 sion of rural health centers have been under-
taken recently by countries in this region. In
Source: World Bank, Health Sector Policy Paper, 1980.
Indonesia, for example, the major goal of the
health plans from 1969 to 1979 was to ensure
one health center per rural administrative
South Asia
unit; by 1977 this goal had been exceeded.
South Asia has more abundant health However, underutilization of public health
manpower than Africa but a similar lack of facilities is a serious problem throughout this
facilities. The major exception is Sri Lanka, region, due either to poor staffing and man-
where a determined program of development agement as in Thailand, or due to competi-
of infrastructure and personnel has resulted tion with highly complex systems of tradi-
in ratios of population to resources that are tional medicine as in Indonesia.13
among the lowest in the developing world. Countries in the Southeast Asia and Pa-
South Asia averages 3,000 to 4,000 per-
sons per hospital bed, but only about 30 per- 12Policy Paper No. 80, Nepal Project Paper, Inte-
cent of private medical facilities compensates grated Health Services, USAID, May 1976, p. 22.
to some degree for public sector maldistri- 13World Bank, "Indonesia: Health Sector Over-
bution. Health centers and clinics, or subcen- view," East Asia and Pacific Regional Office,
ters in rural areas, are often understaffed and unpublished.
so do not adequately serve their catchment
areas which often exceed 100,000 persons TABLE 6. Health Resources in Selected South
each. In India, for example, only 58 percent Asian Countries
of rural health centers operate with full staff
complement.10 Population
Although the supply of physicians is
greater in South Asian countries such as Per Per Nurse/
Country Per
India than in other developing regions, pop- Hospital
Midwife
ulation/physician ratios are about 2,000: 1 in Bed Physician
urban areas and 12,000:1 in rural regions.11 Bangladesh
Ratios of population to nursing personnel av- India 15,050 38,540
5,640
Pakistan 1,620 4,100 3,960
Sri Lanka 2,070 3,920 5,680
10Montek S. Ahluwalia et al., "India: Occasional 330
Nepal 4,010 1,300
Papers," World Bank Staff Working Paper No. 279, 36,450 17,420
May 1978, p. 198. 6,630
1 'Ahluwalia, op cit. Source: World Bank, Health Sector Policy Paper, 1980.
 589 HEALTH AND NUTRITION

cific region have pursued vigorous programs TABLE 7. Health Resources in Selected Countries
of training for health personnel. Thailand, in Southeast Asia and the Pacific
for example, nearly doubled the number of
nurses between 1968 and 1975.14 In Indone- Population
sia, the second national plan (1974-1979) Per Per Nurse/
sought to double the number of nurses and Country Hospital Per
nearly to double the number of health Bed Midwife
Physician
assistants.
Indonesia
Except in Indonesia, physicians are in rel- 1,560
18,160
4,730
Thailand 800
atively plentiful supply in the region. How- 8,460 1,530
ever, Korea, Malaysia, the Philippines, and Philippines 640
3,150 1,050
Thailand face major problems of physician Papua New
Guinea 170 —
Korea,
emigration due to greater opportunities 2,350
abroad for specialization and economic ad-
Republic
vancement. However, the distribution of phy- of
sicians remains a problem for most countries. 1,510 2,020 1,240
In Indonesia about 60 percent of physicians Source: World Bank, Health Sector Policy Paper, 1980.
are concentrated in urban areas. In Thailand
there are three times as many physicians in
Bangkok as in the remainder of the country.
In Korea there is eight times as much physi- 500 to about 360 from 1970 to 1976. The
cian time available per capita in Seoul as in population per physician decreased from
the rural provinces.15 1,800 in 1964 to 1,300 in 1976.16 Wide vari-
Thus, the region faces problems of poor ac- ations are found throughout the regions. In
cess and underutilization of health facilities, 1976, Central America had 1,400 people per
and of maldistribution and emigration of physician, whereas tropical South America
health personnel. Table 7 presents compara- had 1,600 and temperate South America had
tive physical resources data for selected 600 — a ratio similar to that of the United
countries in the region. States.17 But the urban-rural distribution of
physicians remains skewed, with two-thirds
of physicians located in large cities, where
Latin America only
Northone-third
Americaof there the population
were 248 resides.18 In
nurses and
Latin America has the most developed and
411 nursing auxiliaries for every 100 physi-
best distributed health infrastructure of any cians. In Central America there were 80
region of the developing world. Ratios of pop- nurses and 132 nursing auxiliaries per 100
ulation to infrastructure often compare fa- physicians. The relative scarcity of nurses is
vorably to those of developed countries. Ar- largely a function of their lower social status
gentina, for example, has a lower population in Latin America but also reflects the emi-
per hospital ratio than does Spain. Excep- gration ofsome 20 percent of nurses to North
tions to this relatively bright picture, how- America.
ever, are found in Bolivia, Peru, and Problems of structure, distribution, and
Ecuador.
emigration characterize health personnel in
The population per health worker fell from Latin America. Rather than a pyramid with

14Ellen Schaengold, "Preliminary Review of the

l6Pan-American Health Organization, Health
Health Sector in Thailand," in World Bank: Income Conditions in the Americas, 1973-1976, Scientific
Growth and Poverty Alleviation in Thailand: Some
Publication No. 364, Washington, DC. 1978. p. 1 18.
Special Studies, East Asia and Pacific Programs De-
partment, June 22, 1979, annex IV, p. 14, table 5. 17 Health Condtions in the Americas, p. 119.
15World Bank mission data. 18 Health Conditions in the Americas, p. 126.
 590 HUMAN-RESOURCE DEVELOPMENT

TABLE 8. Health Resources in Selected Latin tenth that in the rural provinces of Al Has-
American Countries
akeh or Al Rakka.20 In Jordan 71 percent of
the hospital beds are located in Amman,
Population
which has 57 percent of the population.21
Per
In Afghanistan, on the other hand, basic
Per Nurse/ health centers outside of urban areas are well
Hospital Per
Country Bed Physician
Midwife distributed, but fall far short of the needed

Honduras 660
numbers. Similarly, in the People's Demo-
3,300 1,170 cratic Republic of Yemen (PDRY) 95 per-
Bolivia 490
2,120 cent of health clinics and centers are located
Paraguay 690 3,520
Peru 510 1,190 1,570 outside of Aden, but total numbers are in-
1,800 2,370 adequate, resulting in insufficient rural ac-
Chile 270 420
2,320
Costa Rica 260 1,550 570 cess.22 Egypt has developed a well-distributed
Brazil 270 rural system with sufficient numbers of rural
1,660 4,070 health units to ensure that there is a health
Source: World Bank, Health Sector Policy Paper, 1980. care facility within three kilometers of every
village. However, problems of management

a broad base of auxiliary personnel, a body of and supplies limit utilization.23

Shortages and maldistribution of some cat-
technicians and an apex of professionals, the egories ofhealth personnel characterize Mid-
structure of health manpower in Latin Amer-
ica resembles a sand clock with a slip neck of 20Raymond Ueber, M. Susan, Health and Policy-
technicians and two receptacles representing making inthe Arab Middle East, Center for Contem-
a relatively small cadre of auxiliaries and a porary Arab Studies, Georgetown University, Wash-
much larger cadre of professionals. ington, D.C., 1978, appendix G, pp. 51-52.
Table 8 presents comparative data for 21 John F. Gallivan, M.P.A., Syncrisis: The Dynam-
ics of Health, Vol. XXI, The Hashemite Kingdom of
Latin America.19 Jordan, U.S. Department of Health, Education and
Welfare, May 1977, p. 44.
Middle East 22See tables 3.42 in WHO/PDRY, People's Dem-
ocratic Republic of Yemen Primary Health Care
Programme, Aden, 1979, pp. 95-96.
The level of development of health infra- 23World Bank mission information.
structure and personnel in the Middle East
varies widely due primarily to vast income
and environmental differences. The oil ex-
TABLE 9. Health Resources in Selected Middle
porting nations of the Arabian Peninsula East Countries
have per capita incomes among the highest in
the world. Some of their problems of infra-
structure and personnel development are se- Population
rious, especially as regards access to services Per Per Nurse/
by nomadic populations. Nonetheless, they Country Hospital Per
have made very significant progress in recent Bed Midwife
Physician
years. Egypt
470
Health infrastructure in the Middle East 4,630
Yemen,
Iraq PDR 660 1,870
generally is poorly distributed. In Syria, for 32,380
Jordan 940 1,940
example, 45 percent of the health clinics are 2,550
3,820
located in Damascus and Aleppo. The popu- Syria 480
1,070 3,060
2.470 3,430
lation per hospital bed in Damascus is one- Iran 650 2,130
2,570 1,630
19 Health Conditions in the Americas, p. 124. Source: World Bank, Health Sector Policy Paper, 1980.
 591 HEALTH AND NUTRITION

die Eastern countries. Nursing, laboratory manpower in the region. In Syria, 65 percent
and technician staff are generally in short of the midwives practice in Damascus, leav-
supply. Cultural and religious traditions re- ing only 300 midwives to serve the rest of the
quire that women be attended only by fe- country. In PDRY, 73 percent of physicians
males, yet all countries in the region have a practice in Aden, and in Jordan, 76 percent
critical shortage of female personnel. Since
this problem is in part a result of the lack of Table in
practice Amman.25 the variations in the re-
9 indicates
women with the prerequisite educational lev- gion in the availability of health facilities and
els, many of the countries in the region have personnel.
turned to training traditional midwives to en-
sure access to services for women and chil-
dren. These and other shortages are often 24Yemen Arab Republic, Ministry of Health, Na-
characteristic of national personnel. In the tional Health Programme, 1976/77-1981/82,
Sana'a, August 1976, p. 51.
Yemen Arab Republic, for example, 47 per-
cent of physicians and 20 percent of nurses "Primary Health Care Program, WHO/PDRY,
table 3.43, pp. 97-8; Gallivan Jordan, Syncrisis: The
expatriates.24 also characterizes health
areMaldistribution Dynamics of Health, U.S. Department of Health, Ed-
ucation and Welfare, May 1977, p. 48.

IX.B.2. Nutrition Objectives*

The mathematically precise economic put have been large compared with the in-
growth models that have been in vogue since creases ofland, man-hours, and physical re-
the 1940s seldom take explicit account of the producible capital. Investment in human
notion of investment in human beings. In- capital is probably the major explanation for
creases intomorrow's income are assumed to this difference." A significant part of eco-
result primarily from today's additions to nomic growth in the United States and West-
material capital, and since consumption dis- ern Europe, for example, has been attributed
places capital investment, it becomes an to education, and any residual growth to
enemy of growth, not a handmaiden. Con- "knowledge."
sumption inthe form of educational services, In similar attempts to begin to measure
clothing, and eating of course have an instru- economic returns to health investment, the
mental impact on productivity, but since the cost of preventing a death is compared with
effects of such consumption are difficult to the worker's future income, had he lived. Or
identify, all growth in income is imputed to the investment in human capital — the health,
those more easily measurable factors in- food, clothing, housing, education, and other
cluded in the model. Expenditures on health expenditures necessary to enable a person to
and nutrition are also classified as consump- develop his particular skills — is measured
tion and thus fail to show up as factors af- against his loss through death any time prior
fecting national growth. to retirement. Those costs can also be mea-
Recently, however, the concept of capital sured against debility, where death is not a
has been extended to human beings. Devel- factor. Whether an illness results in tempo-
opment of the new theory was prompted by rary loss of work. days or some temporary or
the discovery that "increases in national out- permanent reduction in work capacity, the
estimated loss in output added to the cost of
medical care can be compared with proposed
*From Alan Berg, The Nutrition Factor, Brook-
expenditures for preventing the occurrence of
ings Institution, Washington,D.C, 1973, pp. 16-29. the illness in the first place.
 592 HUMAN-RESOURCE DEVELOPMENT

Similar comparisons can be made of the who need it2— without it, severe (third de-
benefits to be gained from expenditures on gree) protein-calorie malnutrition is gener-
nutrition. Improved nutrition that returns an ally fatal — annual costs would be on the
absent worker to the active labor force, or order of $6.8 billion.
helps lengthen his working life span, or over- Clearly it is cheaper to prevent malnutri-
comes adebility that is reducing his produc- tion than to cure it. However, as long as the
tive capacity, or that enables a child to return elimination of, say, a case of kwashiorkor
to school or to improve his understanding or frees a bed and other medical resources for
retention of things taught, or that enables an treatment of some other sick person who was
adult to absorb more effectively in-service otherwise unable to gain entry into the sys-
training or the advice of an agricultural ex- tem, total hospital costs will not go down.
tension agent clearly raises the flow of earn- Since unsatisfied demand for curative ser-
ings above what it would hve been in absence vices is typical in low-income countries, re-
of the improvement in well-being. duction ofmalnutrition is not likely to bring
Once a person's well-being is stabilized, about either a net reduction in medical ex-
nutrition costs become a maintenance expen- penditures or a slowing of the rate of growth
diture. Increments of nutrition no longer lead in medical system investment. Adequate nu-
to increases in productivity. An improvement trition, however, would enable the medical
in nutrition thus can help to improve or main- system to increase the welfare and restore the
tain the productivity level of an active mem- productivity of all those persons in the queue
ber of the labor force, or it can take the form who would be able to replace malnutrition
of an investment — for example, helping to patients in the system.
push up the expected lifetime earnings of a
two-year-old child. Reducing Productivity Losses
Another potentially large nutrition benefit
SAVINGS ON MEDICAL COSTS for developing countries is the reduction in
productivity losses caused by the debility of a
One measure of the benefits of a nutrition substantial portion of the labor force. Unfor-
program is in the medical costs saved through tunately, medical data of the kind needed for
reduced demand for medical services. In Ca- calculating debility seldom are available for
ribbean hospitals, 20-45 percent of the pe- poor countries, and even if clinical data are
diatric beds are filled by nutrition cases; in accessible, they do not include many of the
India, 15 percent; in Guatemala, 80 percent. sick who never enter the statistics because of
Nearly half a million hospital inpatients from the excess demand on the medical system. In
thirty-seven developing countries were offi- any case, the synergistic interaction of mal-
cially registered for malnutrition in 1968 (ac- nutrition with much prevailing illness makes
tual numbers may be higher due to classifi- it difficult to pin down the exact contribution
cation of nutritional ailments under other of each factor. Often malnutrition itself does
diseases). At an average cost of $7.50 a day not put sufferers into the queues.
for ninety days for each case,1 costs for treat- An alternative approach to measuring pro-
ing malnutrition are on the order of $340 mil- ductivity losses is through use of aggregative
lion a year to the thirty-seven countries. If data on food supply and the occupational dis-
treatment were provided to the approxi- tribution ofthe labor force. From estimates
mately 10 million preschoolaged children of daily caloric need in different occupations,

'Cost of hospitalization in Guatemala is $7.31 a 2An estimated 3 percent of the 325 million pre-
day; in Uganda, $7 84 a day. This compares to na- school-aged group in developing countries need such
tional per capita health budgets in many countries of treatment; that is twenty times the number of medical
between $1 and $2 a year. facility costs available for children for all diseases.
 593 HEALTH AND NUTRITION

shortfalls in work capacity can be calculated of developing countries indicate that reduc-
for different levels of shortfall in caloric in- tions in adult mortality would not only add
take. Comparison of a country's average ca- years to income-generating lives and reduce
loric need to average national caloric con-
sumption then yields national working dependency
on education ratios, and otherbut increase the "yield"
investments society
capacity shortfalls. For low-income countries makes in workers during their formative
these shortfalls are almost always very sub- years. The advances are only potential, how-
stantial, often as high as 50 percent. While ever, because they depend on productive em-
this means of linking individual productivity ployment being available.
to national productive capacity is concep-
tually useful, it relies on such aggregative
data that its utility in estimating the cost of The Problem of Surplus Labor
malnutrition is limited.3 Restoring a worker to good health adds
nothing to national production if no job exists
Extending Working Years for him. The apparent slow growth of em-
ployment opportunities compared with the
Another cost of malnutrition is the reduced growth of the labor force in poor countries is
number of working years resulting from early
a common source of development economists'
death. For the majority of the developing skepticism that better health and nutrition
countries, increases in the life expectancy of will bring economic benefits. Many countries
adults would add years to the working lives have a substantial labor surplus in the form
(rather than retirement years) of most of seasonal idleness in agriculture, open
adults. In countries where life expectancy at urban unemployment, and part-time or work-
ten or twenty is particularly short, the added sharing employment which has been called
working years would be those when healthy disguised unemployment. Hence the case for
adults would be at the peak of their powers seeking productivity benefits from better nu-
and earning capacity. trition, especially for the masses of the un-
Other things being equal, a lengthening of skilled, would seem weak.
working life reduces the country's depen- Yet, in rural areas of low-income coun-
dency ratio — the proportion of those in the tries, labor is more commonly in short supply
population (largely the young) who produce than in surplus during harvest and other pe-
no income to those who work. . . . Lower de- riods of intense activity. Do farm workers try
pendency ratios, of course, increase per cap- to feed themselves seasonally to higher ca-
ita income and, potentially, per capita sav- pacity, like draft animals? In fact, can they
ings as family incomes are required to do so during the weeks preceding harvest,
support fewer numbers. when cash income is at the lowest annual
The age structure and life expectancy rates point and grain prices are at their peak?
What happens to the productivity of workers
3The example given here is based on the work of who undergo alternating periods of well-
Hector Correa, who recognizes that because of data being and malnourishment?
limitations many of his assumptions are heroic. A Open unemployment does not necessarily
more refined model would offer finer breakdowns of
mean that production problems can be solved
the labor force, adjustment of work factors for local
conditions and of caloric requirements by occupation merely by hiring additional workers. Many
and local conditions, and estimates of daily intake by functions impaired by a worker's malnourish-
income level and by season. It would still fail to take ment cannot be satisfactorily corrected by
account of such important factors as employment op- added hands. Most machine-paced opera-
portunities, intra-family food distribution, the impact tions have precisely defined needs for which
of cooking habits on nutrient content, the problems of
efficiency of absorption, and the productivity impact the human work input cannot be divided
of early malnutrition on mental and physical among more workers to compensate for inef-
capacity. ficiency. Ifa job is strictly machine paced,
 594 HUMAN-RESOURCE DEVELOPMENT

the worker would have narrow scope for re- and clearly this is one of the major challenges
ducing his performance below the machine's facing many countries — an economic, as dis-
automatic demand. Malnutrition might then tinct from welfare, case cannot be made for
be reflected in shoddy output, particularly if special expenditures merely to upgrade the
the work is dependent on the worker's man- potential productivity of those masses of un-
ual precision or strength. Malnutrition also is skilled, landless, adult workers who have dim
reflected in accident rates and poor work at- prospects of gainful employment. There are,
tendance. (In numerous instances, factories of course, other limiting factors in the picture
that have introduced feeding programs have besides malnutrition — poorly functioning ex-
experienced lower accident and absenteeism tension and credit services, inadequate trans-
rates.) . . . port systems, inadequate equipment (perhaps
The quality of labor. Productivity is of exacerbrated by policies encouraging capital
course more than a function of human en- rather than labor intensive technologies), il-
ergy, of numbers of workers. Energy loss is a literacy, and so on. Any presumption that im-
limited basis for calculating the effect of mal- provements innutritional status are a suffi-
nutrition on national production. As devel- cient condition for realizable improvements
opment proceeds, human energy is replaced in productivity would therefore be simplistic.
by machine energy; human quality becomes Yet, it would be equally simplistic to dis-
more important than sheer physical capacity. miss the productivity value of nutrition be-
Demands on human physical energy output cause of the existence of idle adults. To do so
decrease as the proportion of the work force is to assume that underemployed labor is
in agriculture declines. (Although this may available (or can be made available) in the
sound like a long-term description of the de- vicinity of an activity at the right time, that
velopment process, it is already happening in it possesses required skills, that it can be
some developing countries.) In agriculture, hired in fact, and that the work is technically
timely initiative, physical dexterity, and capable of being divided among more work-
comprehension of increasingly sophisticated ers than are currently employed. Often this is
techniques all become critical to the success- not the case. Much of the unemployed and
ful exploitation of new technologies. In cul- underemployed today turns out to be a di-
tivating the new high-yielding grain varieties, verse and complex group that needs to be
farmers fall short of maximum returns be- sorted out before conclusions are reached
cause they fail, in varying degree, to apply about the value of nutritional improvements.
the recommended practices. Some con- Moreover, recognizing the increasing impor-
straints, like inaccessibility or high cost of tance of skilled manpower and general labor
credit, are beyond the farmers' control. But quality for future national growth, invest-
errors of planting depth and timing, pesticide ment in large numbers of malnourished chil-
application, and fertilizer application rates dren today can improve the quality of a sig-
and timing are not economic; they may re- nificant fraction of the future labor force.
flect such factors as education, mental per- This is probably the area in which nutrition
formance level, dexterity, and attention. will prove critical to development in the long
The small farmer exemplifies the problem. run. . . .
His decision making on the use of his own re-
sources isnot divisible. If malnutrition dur-
ing his childhood limited his learning oppor- Other Economic Benefits
tunity and undernourishment as an adult is
compounding his disadvantages, his potential Nutrition programs promise a number of
efficiency in making decisions is not increased economic benefits in addition to the direct
by the presence of unemployed labor in the productivity benefits:
neighborhood. As the incidence of communicable diseases
Given the gross unemployment picture — among the adequately nourished is lowered,
 595 HEALTH AND NUTRITION

the exposure of others to these diseases will ogies. The widest and most lasting impact,
be reduced. however, probably would come from provid-
The increased income of well-nourished ing adequate nutrition to mothers in the last
workers (or well-nourished children when trimester of pregnancy — the critical period
they enter the labor force) should improve for fetal growth — and to children six months
the living standards of their dependents, through two or three years of age (most of
thereby raising both their current consump- the needs before six months can be met
tion and their future productivity. through breast feeding). The greatest phys-
Housewives, whose activities are not mea- iological need and greatest growth occur in
sured ina market economy, should when bet- the early years: 80 percent of eventual brain
ter nourished improve performance on a weight, for example, is reached by age two.
number of economically important functions, During this time children require, relative to
not least of which is the quality of care for body weight, two and one-half times as much
the young. protein as adults; without adequate nourish-
Returns may be raised on other invest- ment, they are susceptible to severe conse-
ments closely related to human well-being, quences ofchildhood infections such as mea-
particularly education. (Low-income coun- sles and whooping cough and to diarrheal
tries now spend nearly 4 percent of their diseases. Even if a child's diet is fully ade-
gross national products on education, almost quate only in utero and during the critical
a third more than in 1960. The efficiency of early years of life, he probably will be
the education systems they support may be brought closer to his growth potential. If,
reduced as much as 50 percent by the drop- during adulthood, his energy intake level falls
out and repeater rates to which malnutrition short of some desirable norm, his productiv-
contributes heavily.) ity will already have been ratcheted to a
higher level — more relevant to a modern
economy — than a life time at his current nu-
trition level could achieve.
Benefits Compared with Costs
Even where opportunities for returns to
better nutrition appear to be significant, the
Payoff on Child Nutrition Investments
costs must be weighed. Will the nutrition ex-
penditure be less than the value of the ex- An institutional feeding program designed
pected increase in production? How will it to meet all nutritional deficiencies of a child
compare with returns to alternative invest- from six months through his third year can
ments? The answers will depend on whose
cost roughly $8 a year or a total of $20.4
malnourishment is to be corrected, what in- (This annual cost to prevent malnutrition is
crements in their productivity can be ex- approximately the same as the daily cost to
pected, how much the program will cost, treat it in a number of countries.) Suppose
whether its effects will be immediate or de- that the program averts a disability in a
layed, and what discount rate is applied to child's later performance. Assume also that if
determine present value if benefits are the disability were not avoided, the child
delayed. would be able to produce (or earn) an annual
Productivity payoffs from nutrition invest-
ment can be anticipated for workers em- *The $8 would meet deficiencies of a diet that cur-
ployed in machine-paced occupations in rently satisfies three-fourths of a child's protein need
modern manufacturing sectors, students for and two-thirds of his caloiic need. The estimate is
whom malnutrition limits the potential joint based on the production and distribution costs of Bal
Ahar, used in India's child feeding programs (the
returns from education and health expendi- child feeding program is being used here only for il-
tures, and small farmers facing the more ex- lustrative purposes, and is not being suggested as the
lowest cost means of achieving the nutrition goal).
acting demands of new agricultural technol-
 596 HUMAN-RESOURCE DEVELOPMENT

income of $200 over a thirty-five-year period petitive with returns to education; in India,
starting at the age of fifteen. How much of for example, education returns have been es-
an increase above $200 would be required to timated in the range of 9 percent to 16
make the nutrition investment break even?
Of course, returns beginning twelve years percent.
It would be hard to find a more favorable
after an investment is made are remote, and investment opportunity than avoidance of vi-
they should be adjusted to account for the tamin A blindness. Poverty-stricken societies
long waiting period. If a discount rate of 10 where such blindness is considerable have
percent is used to convert future benefits to few facilities for training the blind for pro-
present value, an improvement of only 4 per- ductive occupations. The upkeep of the blind,
cent ($8) in annual earning capacity is minimal as it is, represents a total burden to
needed to break even — the annual productiv- their families or the society. Assuming that
ity increase is about the same as the annual the average blind person could be sustained
cost of the feeding. Both the time stream and at a subsistence cost as low as $25 a year, the
the discount rate are on the conservative side, cost is 1,250 times the annual ingredient cost
however. In developing societies a child often of the vitamin A needed for prevention. If vi-
turns worker at eight or ten. Though a dis- tamin delivery techniques should raise the
count rate of 10 percent is commonly used for cost several times, the arithmetic would still
developing countries, it is probably too high be highly favorable. The average blind per-
for important irreversibilities — for assets or son need only be in a position, if sight loss
opportunities that once forgone can never be were avoided, to produce a modest fraction of
restored. The irreversibility of the human this annual consumption for the investment
condition or opportunity resulting from mal- in his sight to yield an enormous return. For
nutrition would place nutrition investments India the consumption burden (at $25 a year
at the lower end of any discount range, in- per person) of one million people blind from
creasing their competitive strength as vitamin A deficiency will cumulate to $1 bil-
investments. lion over their lifetimes. . . .
Nonetheless, an increase in future produc- Many of the links between diet, perfor-
tivity of 4 percent is a modest reflection of mance potential, and economic returns are
higher levels of performance. The actual rate poorly understood and may remain so, given
of return will depend on a number of factors. the complexities of human development and
The higher the initial income of the worker, behavior. Little is known, for example, about
the smaller proportionately need be the the relative damage caused by different de-
break-even increase in productivity. The grees of malnutrition at different ages and of
larger his shortfall in working capacity due to varying durations. The major benchmark
malnutrition, the greater will be the increase provided by medical science is the concept of
in his performance from better nutrition. If "minimum daily requirements" of specified
his improvement in performance can move nutrients. Any sustained diet below the min-
him upward in occupational groups, the gains imum daily requirement implies damage, es-
may be more significant. Increased produc- pecially in environments where widespread
tivity ought also to include side benefits and diarrhea and other ailments lead to heavy nu-
enhanced returns to other investments. In a trient losses. Although it is known that the
study that compared nutrition levels and IQ extent of damage increases as the level of nu-
differences of Chilean children with IQ and tritional deprivation falls — as a child moves
productivity (or wages) of Chilean workers, from first to third degree malnutrition — the
the potential increase in earnings from child shape of the curve relating deprivation to loss
feeding that would offset protein-calorie de- of physical, motor, and mental development,
ficiencies during the critical periods of early and to the severity of nutrition-related dis-
growth was estimated to range from 19 per- eases, is unknown. The threshold beyond
cent to 25 percent. Those rates are quite com- which the extent of loss becomes serious can-
 597 HEALTH AND NUTRITION

not be defined, nor can the degree of mental causes. Before children born today in devel-
or physical shortfall or other resulting dam- oping countries reach their fifth birthdays,
age that separates the serious from the incon- approximately 75 million youngsters will die
sequential becalculated. of malnourishment and associated illnesses.
Although it may never be possible to sep- The known numbers of those whose lives will
arate the mass of the world's malnourished be marked by illnesses that seriously strike
children into neat groups whose potential the malnourished are so great, and the effects
performance can be forecast from specific in- of malnutrition on educational and produc-
vestments invarying nutritional supplements, tive capacity so apparent, that investment in
a clear line can be drawn between those who nutrition programs can almost be undertaken
will live and the significant number of those as an act of faith. . . .
who will die of malnutrition or related

IX. B. 3. Nutrition Actions*

It may be fair at this point to offer the follow- all countries, but within countries it differs
ing conclusions about the nutrition problem: significantly among various income groups,
occupations, and regions. Malnutrition in
• Malnutrition is a problem of major Northeast Brazil, for example, is as severe as
proportions.
it is in parts of South Asia. The largest prob-
• The nutritional condition of the poor is no lem in sheer numbers is the Indian subconti-
better than it was a decade ago. In many
countries it is worse. nent. The largest problem by proportion of
population in needs is in Sahelian Africa,
• The nutrition problem is not likely to be re- Bangladesh, and parts of Central America.
solved inmost countries within a generation
In most parts of the world, supply of food has
by increasing incomes and agricultural
production. kept ahead of population growth. In sub-Sa-
haran Africa, however, food production per
• The basic problem is food-energy insuffi- capita has been declining for a decade. Life
ciency, sometimes complicated by deficien- expectancy is fifteen years less than in Asia
cies of specific nutrients.
and may well become worse.
• The principal victims are the very poor, es- The need for nutritional help exists among
pecially the rural poor. Most governments
people of all ages and both sexes, not just
are not reaching them with the benefits of small children and pregnant and lactating
nutrition; few have central ministries with
the outreach to do so. women. The problem is particularly serious
among families of landless agricultural la-
Commonly expressed goals to eradicate borers, farmers with small landholdings that
malnutrition in the near future are unrealis- rural development programs do not reach,
tic. The aim should be to overcome malnutri- small-scale fisherman, and the urban unem-
tion in those areas, forms, and population ployed. Together they constitute more than
groups in which it exerts the greatest drag on half the malnourished in most countries.
development. The problem exists in almost Their nutritional condition reflects inadequa-
cies in the availability of food, in economic
and sometimes physical access to the food
♦From Alan Berg, Malnourished People: A Policy
View, Washington, D.C., 1981, pp. 47-52. Reprinted that is available, in knowledge of the best
by permission. way to use the available resources, and in
 598 HUMAN-RESOURCE DEVELOPMENT

health practices that affect biological use of PRIORITIES

the food that is consumed.
Accordingly, nutrition efforts should be Despite the high degree of variability in
designed to expand food supplies — increasing nutritional needs among countries and in the
production and reducing food losses — in causes of and appropriate responses to those
ways that will benefit the poor, with attention needs, there are certain strategies that merit
focused on what is grown, who grows it, and high priority in most countries. Accelerated
what is stored; to increase the incomes of the growth in the incomes of the poor and-with
poor, to improve their access to food by im- few exceptions — in food production continue
proving the marketing system and adjusting to be of primary importance. Attention needs
price policies in ways that benefit consumers to be given to the development of nutrition-
without creating a disincentive to producers oriented agricultural production policies and
and by setting up special feeding programs; programs. And to ensure that food reaches
to try through education to bring about those in need, food-demand programs, in-
changes in food preferences, in the distribu- cluding the strong possibility in many in-
tion of food within families, and in hygiene; stances offood subsidies, are required. Broad
to improve health and environmental condi- programs, such as general consumer food
tions— water, sanitation, immunization, and subsidies, may under some circumstances be
management of diarrhea — and to attack spe- as effective in reaching target groups as nar-
cific problems of micronutrient deficiencies row efforts, such as institutional feeding pro-
with mass-dose capsules or through fortifi- grams designed to reach children of pre-
cation of food staples. Some changes can school age. Clearly, however, subsidy
best — or perhaps only — be brought about programs should concentrate on low-income
through changes in government policies; oth- groups instead of all income groups, on re-
ers can be attacked directly through nutrition gional rather than countrywide programs,
projects. and on seasonal rather than year-round aid.
The effectiveness of different nutrition ef- Most of the larger developing countries al-
forts and the relative importance of various ready have sizable subsidy programs. Con-
determinants of malnutrition under differing centration should be on increasing their ef-
conditions are now better understood than fect on nutrition, imporoving their efficiency
they were five years ago. Although increasing in the process.
income is fundamental to increasing food These are new areas of emphasis that can
consumption, for example, it is now seen to complement the well-known direct pro-
be less efficient than lowering food prices. grams— nutrition education, fortification of
From tightly controlled field studies it has staples with micronutrients, incorporation of
been learned that, in addition to equity ben- nutrition-related actions into health services.
efits, increasing the amount of food con- The priorities that public officials assign to
sumed and improving nutrition in other ways these various actions will depend on their
can significantly increase the weights and countries' nutrition problems and the causes
heights of total populations of children and of them, the distribution of malnutrition be-
improve their nutritional status — or at least tween rural and urban areas, the extent to
prevent or retard its rate of deterioration. It which the rural malnourished are small-farm
is possible to suggest what it costs through families, the probable cost-effectiveness of
nutrition services to avert death at an early possible programs, the institutional strength
age, to avert a day of illness, to gain an extra and funding capacity to mount programs,
centimeter in growth, and to increase psycho- and political constraints.
motor development scores by a percentage The actions selected should have the aim
point. Even some general effects of broad nu- of causing a specific improvement in specific
trition programs can be predicted. nutritional deficiencies of a specific popula-
 599 HEALTH AND NUTRITION

tion within a stated period. They should be and rural development projects aimed at im-
well defined as to content, costs, timing, lo- proving the well-being of low-income groups
cation, and means of execution. When infor- whenever it is feasible. Malnutrition some-
mation todetermine all this is not sufficient, times isused as a justification for such proj-
projects can be developed for laying the ects, but nutrition goals are not explicitly in-
groundwork. Action should not be limited to cluded in project objectives, and any
the gathering of data, but every operational nutritional gains occur largely by coinci-
program should include a track for evalua- dence. Improvements in nutrition must be ac-
tion and for learning from experience. Nutri- cepted as an important objective of these
tion work in a country with an information projects and the costs of possible actions must
base and experience in nutrition programs, be taken into account. Nutrition actions
such as Costa Rica, India, or the Philippines, should not be undertaken if their negative ef-
would be very different from work in a coun- fects on other project goals would more than
try that had previously given little attention offset the gains from other actions. But mod-
to nutrition. est reorientation of project designs can some-
The complexity of the nutrition problem times have significant nutritional effects
and the multiplicity of potential nutrition ac- without causing unacceptable changes in the
tions should not be allowed to dictate com- achievement of other goals. When benefits of
plex projects. Projects should be broadly con- various goals are conflicting, the tradeoffs
ceived with regard to content, but they among the various goals should be weighed.
should not be expected to address the many In certain agricultural and urban projects
factors that affect nutritional status. Com- the addition of nutrition objectives might im-
plex projects have generally been found dif- prove project design. The importance of nu-
ficult toimplement effectively. Whenever it is trition to objectives and design will differ
feasible, there should be a sharp focus on a among projects. In Nepal, for example, a
small number of critically needed actions. rural development project was based on an
Nutrition actions should be designed in understanding of the way food consumption
ways that limit the need for managerial in the region was related to need; designation
skills, of which many countries have a short- of project components, including the selec-
age. Similarly, there needs to be a clear focal tion of crops, flowed from this understanding.
point for administration of projects — mul- A project in the Southern Highlands in
tiministerial coordinating mechanisms have Papua New Guinea, involving a shift from
not proved to be particularly promising in ac- subsistence crops to cash crops, was modified
celerating actions. Generally a single agency to provide extension services that would help
should be made responsible for nutrition increase production in family food gardens
projects — whether an agriculture, health, or and to include other assurances that the mod-
social welfare ministry, or a planning agency ernization effort would not be nutritionally
will depend on the nature of the project and negative. In Malaysia the design for a reset-
the practices and preferences of government. tlement project in south Kelantan provided
means for settlers to meet their nutritional
NUTRITION IN AGRICULTURE needs during the first seven years, before rub-
ber trees could be tapped. The government
In countries where actions involving agri- thus withheld a portion of the land for food
culture can make a significant contribution to crops, helped to build and stock community
meeting chronic needs in nutrition — urban as fish ponds, and provided nutrition education.
well as rural — specific agriculture projects Often the steps required to incorporate nu-
designed to have an effect on nutrition should tritional goals are relatively easy to plan and
be considered. Nutrition should also be in- implement — as, for example, in the choice of
serted as an explicit objective in agricultural crops to emphasize in agricultural research
 600 HUMAN-RESOURCE DEVELOPMENT

projects — and need not be administratively agencies to broaden both their perspective on
or analytically complex. . . . their nutrition work and their view of their
Projects in agriculture and rural develop- policies and lending in other sectors, partic-
ment must noi be allowed to cause a deter- ularly inagriculture. Especially within agen-
ioriation of nutritional status. Quite uninten- cies involved in general development work,
tionally they could have harmful effects on improved nutrition, like reductions in poverty
food supply, food prices, or incomes. Both levels, needs to be seen as an overall objec-
policies and projects should be routinely ex- tive. Much of the policy and project work of
amined for their nutritional effects, includ- development institutions affects the nutrition
ing, ifpossible, their effect on groups that are of low-income groups, but these agencies are
not their direct beneficiaries. Where poten- not always aware of the extent and some-
tially deleterious effects are discerned, the of- times even the direction of that effect. Fur-
fending portions of the projects should be re- thermore, they sometimes miss cost-free or
oriented or nutritional components should be low-cost opportunities for strengthening pos-
added to offset the negative effects. itive nutrition effects in their work in other
In agriculture and rural development proj- sectors. External agencies can only make a
ects that call for evaluation, nutritional sta- substantial contribution to improved nutri-
tus should, whenever it is feasible, be a key tion if they adopt it as an explicit and sus-
measure of project performance. tained objective. This does not demand mas-
sive budgets for nutrition projects, though
identifiable nutrition projects may be the best
NUTRITION IN HEALTH response to the nutrition problem in some
countries. What is needed primarily is a nu-
The interaction of malnutrition and infec- trition dimension in development-assistance
tion has a far more serious effect on individ- programs, particularly in agriculture, and the
uals than the combined effect of the two
systematic incorporation of food consump-
working independently. Consequently, the ef- tion issues in economic dialogues and sector
fects of nutrition actions and health pro- work.
grams undertaken simultaneously are greater
Development institutions and the govern-
than the sum of their effects on the same pop- ments of developing countries can work to-
ulations would be if the actions were under- gether todo this by:
taken separately. Since integration of nutri-
tion with health services is a particularly • making a substantial effort to improve un-
efficient way of using limited resources, im- derstanding ofthe nature and extent of the
proved nutrition should be considered an ex- problem, of where in the chain of nutrition
plicit objective in all relevant health work. events the weakest links are, and of ways in
Problems associated with acute forms of mal- which nutrition issues can be better inte-
nutrition and vitamin and mineral deficien- grated into operations;
cies can be much more productively attacked • incorporating nutrition concerns explicitly
through health services than can the low lev- in agricultural and rural development
els of performance associated with chronic project work by
food deprivation. • developing
work, tritionprojects that respond to the nu-
findings of economic and sector

IMPLICATIONS FOR • analyzing the nutritional consequences of

DEVELOPMENT-ASSISTANCE projects in agriculture and rural develop-
AGENCIES ment, and
• adding nutrition components to increase the
Malnutrition is a major development prob- benefits of projects or neutralize their pos-
lem that calls for development-assistance sible negative effects;
 601 HEALTH AND NUTRITION

• including improvement of nutritional status precise nature of nutritional deprivation, the

as an objective and part of the design of ap- relative importance of its causes, its conse-
propriate health projects; and quences, and the cost-effectiveness of reme-
• setting up free-standing nutrition projects dial action. A sizable gap in knowledge that
when they are the most appropriate mech- the agencies should fill is on the nutritional
anisms for achieving stated nutrition effects of their projects, especially in agricul-
objectives. ture and rural development. They need par-
External-assistance agencies should con- ticularly to work on the design and imple-
tinue to emphasize food production, but with mentation ofsimple, low-cost monitoring and
increased attention to those foods consumed evaluation systems. Increased emphasis on
nutrition is a logical extension of the effort to
by low-income groups and further support to
projects that help to strengthen the purchas- increase production of food, to assure con-
ing power of the poor. They can also help sumption bythose who need it. The food-pol-
icy approach can complement and broaden
government bodies fill gaps to their knowl-
edge. Unknowns abound in this complex other work of development-assistance
field. Much remains to be learned about the agencies.

Comment
The special problem of famines is analyzed by Amartya Sen, Poverty and Famines (1981).
Sen develops an exchange entitlement approach to explain the causation of starvation in gen-
eral and of famines in particular. For an informative commentary on this topic, see Kenneth
Arrow's review of Sen's book in New York Books, Vol. XXIX, No. 12 (1982). Also refer to
selection I.C.6 for a discussion of the relationship between entitlements and poverty, and to
T. N. Srinivasan, "Malnutrition: Some Measurement and Policy Issues," Journal of Devel-
opment Economics, Vol. 8 (March 1981).

EXHIBIT IX.4. Life Expectancy in Relation to Per Capita GNP, 1978

Hong Kong
80 r-
_ ._ , Taiwan
n Lanka »Cuba Chile
•* Costa .Argentina
Rica
Singapore
70 - China Malaysia. ^^
Thailand Korea • Brazil
60 Philippines • Algeria Norm for 89 developing countries
Egypt
Pakistan Bolivia
50 ndonesia
Indi • Nigeria
• Ivory Coast
Bangladesh"Congo"
40 Nepal 'Senegal
Ethiopia

0 2000 2500
500 1000 3000 3500
1500
GNP/person, 1978 (U.S. dollars)
Source: Nancy Birdsall, "Population Growth and Poverty in the Developing World," Pop-
ulation Bulletin, Vol. 35, No. 5, Washington, DC, Population Reference Bureau, 1980.
 m © I

I2 I

oo r- o\ oo «o SO Tfr
(-- OO (N
tu rf ^t m ^t
o
-^
— i r^
Tt </"> rt \0
^2

CO <>

51<
O S-S
a.
8
B
o 03

III
•? cd 4>
4> 4) C W5 C U

T! C C •? .2 ^ C .- 3 y t: -2 § .. A u
S 'O v. 53 «j o
.9 .5 o is e
2 8 -J
A O.
00
fS
to Ou
XD -^
3 O
PC £ 5
< <
602
 •o S3 2 2
J 04 o o o

I £2 |
jisi
I
I *H^
so oo oo II

o(N OO— —^h oOO 1 O SO <N

g J3 33 O 1 OO OO OS SO ~ I OO Tt OO |

m r—
ed ^

o 3
ed .S3
-2ed a,a. a. <u osovos I o c^-oo | r~ os

t- a-
U =>
q q q q t-^ CN sd
q q q q q OOO I© OO IOO
r-> ci od u-i
<N <N <N «
«i
— < d
(N h
<N ^
<N n
<N
j3 «
q o o

N > m (N

X cd ^^-
5
•8 5 « *-s
"c
d

3
u co
Tf Tt tt fN rj-
04

w- ©"
s*-"' 1
c
%o S~S
cd xi
rn oo q — ; <N (N ov rj rs q
3
2£ <N <N <N <N qoocnqeN
<N^-^rn<s -« ~
en r-; tN
o\ (N<N<NCsi<N N n N (N M
a
cd ON
a.
o
0H
o
V3 _
3 c
ocd 04 — rj- <■«-»© vO
r- O
OO rs so r- <N SO en © r- m »n
m Os
do — ©
♦j
3
a.
o
a.
;e ~- m
r~ m— t-»
w-> — ©
—: d

43
T3
% ued <n o I rt SO <N © —
**
C o.
ed o ?»— ' ' -^ </-> c-i
"s'o

u 0 1- 04
I I —. (N
£ r-
o
c/3
O
a.
OS
OO
OOOOO
5 I
OOO o w lir-
: O O
Oh oo
>a> Z 0\ *3 <N — < ~ -h *
rno vi
„ ci
^ nmoNr-
- -
fS — —
ed
u-
a •2
cd
Q.
g c^3
4>
o
s
U
sd
xi -J £
I1 B Oa £

© Q 5 1 — o
cd E
x: o
i6 .28-1«
II
o
^ ^ M .- a>
C w 3 «J 4>
Ced -3
3 00 U < flQ CO U U U U
I « < 00 CO
3 cd

603
 t) S3 2 2 OO I I «/->

-J x o
1112
■<fr vo en
*U g0)• 2 I>■> 00 ON VO iO (S (S w-» m on
r» r» <n m
<n m
-<i-
<N oo
vo r- Tt — < r- — © «*
On m oo
© © -H
© oo ■«*
1 -1
1 NO

•r ° =3O "£!
O

I*
1 ©

rn en
-5 O. U 3

(J 3 4) <U r>

©©©©© ©©©o© © © © © © © © © © © p © ©
en en ^h <n rs en <N
i^ ri r^ en
- en
^ <N
© <N .— < <N
r~' r»* en cs
r^ CM

§|2A§
© Tt Tt —'

VO fS <N OO CS OO VO VO ^

j 8* 8 1 °.
t- On On I

•a ■£ -* CN v>

CO « ^2 tJ;

1§
on vo en en v> ov v> r~ vj oo Ov vo </■} rn vo © r- ov vo r- O (N iO (N 0\
<N ^ rn CN <N <N CN <N (N <N <N (N fS <N <N
en <N <n «-! c-i (S (N (N OO O
3 ° £ °. *n
— i vo

>/-> rfr en
~h vo r- en en r- en «/-> ov © 0\ ^ ^ \0 OO oo -h in n
•£ °° -- >n 6 ^t >- oo vO VO
tj- «/S en
vd <N <N «/S VO
J2 r- :3 </S vd ~^ on «/">
3 on e uS d oo ^ d
VD
On CO rn
<N
vO
CN oo OO m 11 © rs rs ©
1I © en
-3 on VO rn
©
TT
CN ©
""" 1 1 1 "^ ,—' 1 1
"~ '
1

en oo
ts CN rn
8. © © © © ©
oo
© © CN
© ©
CN
© © ©
ON
o © © © o © © © ,^-
© © © © ©
— < on m oo v> OO m CN OO
m
(X CN <N en CN <N ^h m CN m CN m

a
a.
6
o
U
©■

so

>< e
oc
•a «a
3 rt * O
«d r-
H 0
O r-
T3 I-
£ CM — N O c ^ ^

x oo^jS S5SSZ 0* W
Si o 3
i73 C7D C73 oo « 9- •-
co oq U U O O
X H H D D N o o «« o-r*
* >

604
 Ill III I IS I I

£ I
I S I III I IS I I

I I O
Os II N v©
p- m II oo m
no © ©
© «/-> r-
<n r-

On
On
I
I
on <n
© NO
-*
©
«/-> oo <n
O O —
—
ON
~4
00 i ia i i

© © ©
- I P I © © © © © © © © © ©
rn r^ en
(N (N <N
I © © © © ©
no no oo no no r-'^wS — ^
I IS I I

no no -<t T ^n
I I rt | I

r- on oo r- ir> oo r-- oo o on r- -^
i r i i

I is i I

co ~^ r- u-> o en CN|-^fNrn«^ On fN © On On
— rr oo — ©
© no m en oo v-} oo
en vi en .-i <N <N i-3 ^ (S <N c4 en en en r^ <n en <n cn| en <n ~ -^

NO

— « 14
© oo r» on q >/-> «n en on NO -<t On © © On —- © no oo
d
© ©'
(N ©
NO ©
(N ©
—
i-5 r-' -3 © •-, OO
© © © © (— © On en m od oo no vi —«
© »/-> no
tn
^r NO en oo —<
Tf rn
(N ©
m »■* NO Tf
rn 11
"~l d dr^ NO ON r-' ^ m « <o
1 T ON 1 T I - ^ © CM fN ON «/">

r- NO
NO NO
© © © ©
en © © © © o o o oo
o © © o © © © m© a, © cs
ON <N
O
©'

OO oo <s
NO —
© en
ON ©^
V. r~ w^
m
ON OO oo © © NO ON On w^ rf

r-
ON

<r)
ON * I
S ■? «? &
a:
•-> a
03
5 a J 5 S
O 3 td a.0
c c
§ I 8. =5 < J.s-g £E Ci
E vj co
« fc 8 2 2 •a H *3 •§ 5 "B ■§ -g E
s ^ §
I 3< 1*1
cd =
« Cfl W. 4> O > «N< •- CD S co H ^2 a < oo £ u. O
0* C/3 CU CO W W5 NN

0-
> >•<
605
 w O ^ ^

era I2 I8 I I <N r- m

nrol
►J * o

tota
lme
^
C ^-v

dul £1=15 II 1=12 1

Ratio
o
T3

ale
3
PQ

Fem
pq w

y
mar
E •« =c >'"'• 2 1=1
ply

pri
• poos
^j

2 o £ w
"C

fl) ^~S .M ~
I I

cap
c <N — i —
cs o. t. a*
QJ 3 <t> <U
lori<

3 -^ SO
Z I Io o o o o
pr<^ -«!t
p enp
e3 <— ■ in o\ K «
ecta

S£<2~2
uire

ears
J2
H
o >-.
g.e~
X
Birt
Life

S-tSsS
w
•S«s ■a
SO OO so r- oo >n oo so
I °° I so On SO <N >n OO

v O °* w2
■33 5S h ©
OO

3
s T3
ed
C
on so on ^
rt^J^fvitN
r- m
CO
so ^
<N
r- <n
<N — «
<n ©
co CN
r- on m
ON -^ CO
^
CO ^
oo rri
On OSO ON
O
R
CJ
C
o
a
K
so
3 o
oo
o
O.
a. "53 °°
°" —E
so © so co m CN — ■ On so oo
— Tf ON CO ~
.2a.—r- rs Tf o so O rn O © © © so © © ©
o\
r» l
jf o o
a. pi m in i^
CO (N On Os OO m on on r-- m — ' oo co on
iri -I ri m vd ri <n 't d co —^ so r4
so co r~ «n "^- I I
I I
a - 2 I
o o o o o
Tt- ON so so ON
oin o-^ o\o o— > ono o o o o o
rn O ON oo oo
o o o o~i o
~h~ ON
in ^h N *0 Tf
(O so — ^ rf q so >n m «n on ^ "^- "* ^t ON on on in on oo

00

£c £s o£ g <u ed •= 2 ^
c u •3 *
o a:4) y
cd « Z .5
«
:s oo
4) « S
«= "2
I .1 1 I 1 8
O .2 o
s s * 0..-3 1WHhHH
o •£ =3 oc° IC
« -c rO <C <u cq
sj fl oa
(i) aa
o

606
 I I I

2 I
I2 I I

I I I I2 |

I m > 0\ N
I oo O O O I£ I

pon pvn pon pop IS I^

O O I p p pob pno
r-" co
I I I I pop
o © ob <N
Tt en I ci co i/S

p o
<N <N r- CN o loom
<n no no no 1 vo VO VO 1 on
no on
no r- r-~
«n no
I I

r-
«/-> on t-~ oo in >/-> on o on I r- no <n oo r-- <n no r- oo i m no on <n no o r~ oo r» no on r-

I|rrvo<N^H
t n N m oo — > r—
cn <N ro Tt
Tt on
m |i cn)
w a ^ oo
<N Tt cs m Tt
Tt r- <n
on m
o oo
m |\ Tt
v~> m
~ o\ o\
m m m <n
<n -<t m
m ~
m r--
m no o no
socmen

w-jvooor--^ vi N o\ m a q ^o ^ a oo r-- rn t— ^ m ri m <) oo r- oo o m co Tt cn m m m

0<N<S^<N ri ^ N M N m ^ O ri ^ ^f^^orn -^r<-5<Nrsi(N <N — <N <N — — Ororn
no
-->

O m V) t^ wo « — • ^h _ i oo m o -^ co no oo
Tf- rn <N
IT) OO ON OO
o </-> on d CN o OO Tt -. <N i-3 (N NO mOOOO — <N tj- O
<n o u-i r-' Tf

O o vO O*
co O
(N (N O
no p p co
<n «o on r- p ON OO v© NO O
<*}
M Tf^ OO O
iri (S O c4 co en (N o ^ «S — " O — O TT —
I
d —!—
O O O O O o o o o o o o o o o ooooo ooooo ooooo oooo
no — r- cs </-> Tt
V>
Tt
Tt
~
ON
CO
OO
NO
NO
Tt m >r> •— i no
m «n oo on v>
no OO — ->«/"> On <^TtON«^)Tt — no ro oo — — — — «">
cni «r> »/-> Tt oo rs Tt — « on <N — - oo cm oo r~ r- no no m — on «q^ on rn
on -T .-T
— T rn — T rn —T rsf <N cm' — ' rS «/-T

g.
u
o 3 .2
.o oi
C
U Jr <sl
■S •§ e Q
- 00 O
■O 3 N

.5 o 3 .8 CO « 3
oS £ o6 o y- * j >
<U O OJ3 § 1 CCJ 4> 0- C/}
r- D >
CO ffl fflUU QQtduJ £ o a a a x
» K 3 3 s»- e 5 S Z Z C/3 OO C/5 l_ »- w
E 3 C

607
 -I ™
I I I I I I I I? I I I §
J* o

4) t)
< •■ >,
r- r- 00
? I I3 I
Jill I S I IS ' I I s

I? if
•r o o -5
I? I IS I I 8 I 5J I

1
I
no
m
1
I
1
I
no
co
I 1
I
cn on n©
cn cn cn
00 m
cn —
no
co 18 11 — ©

c« o.

p p p p
O <N -ON
I S II S *l I 2 ^ r-^ cn r*

•-5 ° £ 2 o r-
no r-> so no

U3

"3 ~ 2 in
I on I I on I r- o
r- o o 00 o rj-fnu-^mONON«^ON 00 cn

si
<±)
UQ^"2
ioni I ^ 1 no r- 00 00 cn 00
— O
,j oa 05 ^^
NO CN
on no -<fr ■*!■ r~- on «/i r- co o cn m on j\ m - - m co >r> 00

?"
tIea5s
OO^^O O —nC CN O — «-3 CN O ri r<i <j Tt m on co -*fr

.2
■^ co
O C-- co 00 >-< <-H ©
a. — r-» O o Tf 0 O r-^ co cn 00
cn cn r~ O
00
(N <N OO <N

o o o o on O O CO O On
CO
<N —^ CN
O OO O
CO -<3- CN
o
a.
NO Tf <N OO OO in 00 o — cn — tJ- p r- NO tJ; r-* i>
O ^ ^ ■* m — — co co -«t wS O co co <N <N
V

On $3 OOOOO
r— cnoono«/->
-- — p no <N
OOOOO
no no cn r- on
f^ <*!, ^ -^ 0\
OOO
r-000
^ N n
0000
no 00
00 on on
II
— < >n
r-
OOO

'-*
rf
I ~
r- no cn
On
—*</>
co
*a ON
0 O
0

ON CO
■^rcNoococo vo m •*' rl — * co— Trs
-? rf no" <n
'-, cn r-~ Tt
0

2 ^c
£
a
3
-5
<3 ^
£•3 O U
■5 s"
isJJ »i " *; u
"33 <a
w •a -°

^3 r 1
li o£ a8 a
o
II
e « 8
•5 ^ « « UJ

608
 I I

2 | m os co cs o
OOsOO— <
— ' r- so m
©Os 00©
in

rs rs

m rs co oo oo © -H OS © _H
— so m m
© © © oo © OS © © © © © © O OS oo ©
•— '
OS

SO r- r-
m <N <N
Tfr SO
— co r- ©
<N —.
— rr
SO
co ^o
<n
Os

© oo © SO © © r- © © oo © r- © ^o tt r- © r- © oo © © oo
_• © — ^ © — — © — © r-3 © © © — © q © © © © — © in in ©
© —

©(N© (N © co co — O©
r- r- t— r- r-~ so so r- r- r-

oo© os© <n r-~ — os^orNoooo o - C\ N a so — « s© so — ■ co <N oo oo os os©

co cn t -^-so h m (S r- ra r- oo so so^t — — oo m m SO SO t"- (N-^-vOOsOs Os Os

© — © ©'© © — — © — © © — ©OO©© rj © — — " d d d <n <n © do
in

<*
in TT
oo in —
<n oo
co
CO
CO <N (N
SO "^- OS (N <— ' CO CO oo
■«3- oo so in SO OO (N r» — r-~ r~; — so — os
Os co — © co M OO (N
m © CO CO Os m
Os so © r- — in —'

tj" in
in — in
os
— r- in rs (N
co r~ co - — h in
os oo ros Os so co
CO* (N — CO CO (N © OS ©
CO — CN so
© "-1
— in
<N OS rr
ro
ro CO

Os tt in tt — ' rf
t»\
rs]
© O O © © © O © © © © © © © © o
o
© © © O
CO O
m
O O O C O o o
m
o o5

I,
IT) m
t> OO N CN SO CO OO Tt — OS rs CNl
oo co
os ^t
r- m
©^ — OS » CN
>n oo <^ in Tt r~
o
© 00
—
OS Os" OS \D OO co r-' © oo' Tt u~>
OS
Os QQ
^
^ E

ON
' ^ a.
■o -o 2? " e
* c/5 > I E 2 « S

« « 3 -o -> 5
9- e I -a 5 &« .ti .12
Oil
^ r~ r—

'5b 2 5 « a S c .ti
c 3 ai 3 o « fc.
13 § S .5 2 « « "« "ed 2" 3 w "^ ° £
ki<0QUUU O X "^ > "° °
ON
r-
r-

609
*■
 IX.C. EDUCATION
IX.C.1. Investment in Human
Capital— Note

Although the objective of adding to the stock tion of capital would not increase signifi-
of physical capital has dominated investment cantly the proportion of their national
discussions, it has now become evident that a incomes devoted to capital formation.
high priority must also be assigned to invest- While investment in human beings has
ment in human capital. been a major source of growth in advanced
Many studies of economic growth in ad- countries, the negligible amount of human
vanced countries confirm the importance of investment in underdeveloped countries has
nonmaterial investment. These statistical in- done little to extend the capacity of the peo-
vestigations indicate that output has in- ple to meet the challenge of accelerated de-
creased at a higher rate than can be ex- velopment. The characteristic of "economic
plained byan increase in only the inputs of backwardness" is still manifest in several
labor and physical capital. The "residual" dif- particular forms:1 low labor efficiency, factor
ference between the rate of increase in output immobility, limited specialization in occupa-
and the rate of increase in physical capital tions and in trade, a deficient supply of entre-
and labor encompasses many "unidentified preneurship, and customary values and tra-
ditional social institutions that minimize the
factors," but a prominent element is the im-
provement inthe quality of inputs. Although incentives for economic change. The slow
some of this progress may be incorporated in growth in knowledge is an especially severe
physical capital, the improvements in intan- restraint to progress. The economic quality of
gible human qualities are more significant. the population remains low when there is lit-
For purposes of measurement, capital for- tle knowledge of what natural resources are
mation isusually identified with the net in- available, the alternative production tech-
crease of land, structures, durable equip- niques that are possible, the necessary skills,
ment, commodity stocks, and foreign claims. the existing market conditions and opportu-
But the capital stock should be interpreted nities, and the institutions that might be cre-
more broadly to include the body of knowl- ated to favor economizing effort and eco-
edge possessed by the population and the ca- nomic rationality. An improvement in the
pacity and training of the population to use it
quality of the "human factor" is then as es-
effectively. Expenditures on education and sential as investment in physical capital. An
training, improvement of health, and re- advance in knowledge and the diffusion of
search contribute to productivity by raising new ideas and objectives are necessary to re-
the quality of the population, and these out- move economic backwardness and instill the
lays yield a continuing return in the future. human abilities and motivations that are
If these expenditures are considered as capi- more favorable to economic achievement. Al-
tal expenditures, then the proportion of cap- though investment in material capital may
ital formation in national income in the rich indirectly achieve some lessening of the eco-
countries would be much larger than is con- nomic backwardness of the human resources,
ventionally indicated in national accounts the direct and more decisive means is
that treat these expenditures under the flow through investment in human beings.
of goods to ultimate consumers rather than Emphasizing the weight that should be
under capital. But since poor countries do not
'Hla Myint, "An Interpretation of Economic Back-
make many such investments in the forma- wardness," Oxford Economic Papers, June 1954, pp.
tion of human capital, this broad interpreta- 132-63.

610
 611 EDUCATION

given to the growth in the quality of human tension of human capabilities has failed to
resources, Professor Schultz illustrates the keep pace with the accumulation of physical
possible implications of the quality compo-
nent as follows: While the case for investment in human
Suppose there were an economy with the land capital.3 is gaining wider acceptance, the
resources
and physical reproducible capital including the means of attaining an increase in this type of
available techniques of production that we now investment have still received only superficial
possess in the United States, but which attempted consideration compared with the intensive in-
to function under the following restraints: there vestigations that have been made of the prob-
would be no person available who had any on-the- lems of investment in physical goods.
job experience, none who had any schooling, no
It is not difficult to identify the more im-
one who had any information about the economy
except of his locality, each individual would be portant categories of activities that improve
bound to his locality, and the average life span of human capabilities. As Professor Schultz
people would be only forty years. Surely, produc- suggests, a typical list would be:
tion would fall castastrophically. It is certain that
(1) health facilities and services, broadly con-
there would be both low output and extraordinary ceived to include all expenditures that affect the
rigidity of economic organization until the capa- life expectancy, strength and stamina, and the
bilities ofthe people were raised markedly by in- vigor and vitality of a people; (2) on-the-job train-
vesting in them. Let me now take a Bunyan-like ing, including old-style apprenticeship organized
step and suppose a set of human resources with as by firms; (3) formally organized education at the
many but no more capabilities per man than ex- elementary, secondary, and higher levels; (4)
isted as of 1900 or even as of 1929 in the United
study programs for adults that are not organized
States. The adverse effects on production in either by firms, including extension programs notably in
case would undoubtedly be large. To continue the
speculations, suppose that by some miracle India, agriculture; (5) migration of individuals and fam-
or some other low-income country like India, were ilies to adjust to changing job opportunities.4
to acquire as it were overnight a set of natural re- Underlying each of these activities, how-
sources, equipment, and structures including tech- ever, are a number of questions that should
niques of production comparable per person to
be studied more seriously.5 At the outset, the
ours — what could they do with them, given the ex- problem of measurement presents several dif-
isting skills and knowledge of the people? Surely
the imbalance between the stock of human and ficulties: Isit possible to separate the con-
sumption and investment part of expendi-
non-human capital would be tremendous.2 tures on these activities? Can the particular
Recent experience with attempts to accu- resources entering into each of these compo-
mulate physical capital at a rapid rate in poor nents be identified and measured? And can
countries bears out the necessity of due atten- the rate of return on investment in education
tion to human capital. It has become evident
that the effective use of physical capital itself 3For strong arguments that "the experience of plan-
ning seems to suggest that knowledge (and certainly
is dependent upon human capital. If there is not investment resources) is the most important
underinvestment in human capital, the rate scarce factor in underdeveloped countries with oth-
at which additional physical capital can be erwise favorable social climate," see B. Horvat, "The
productively utilized will be limited since Optimum Rate of Investment," Economic Journal.
technical, professional, and administrative December 1958, pp. 751-3.
people are needed to make effective use of *T. W. Schultz, "Investment in Human Capital in
material capital. In many newly developed Poor Countries," in P. D. Zook (ed). Foreign Trade
and Human Capital. Dallas, 1962. pp. 3-4, 11-12
countries the absorptive capacity for physical 5The remainder of our discussion concentrates on
capital has proved to be low because the ex- education and training. For a consideration of human
capital formation through health services, see Selma
2T. W. Schultz, "Reflections on Investment in J. Mushkin, "Health as an Investment," Journal of
Political Economy, Supplement. October 1962, pp.
Man," Journal of Political Economy, Supplement,
October 1962, pp. 2-3. 129-57.
 612 HUMAN-RESOURCE DEVELOPMENT

be compared with the rate of return on in- or extensive systems of higher education in
vestment insome other alternative use? As newly developing nations. They contend that
yet, no completely satisfactory empirical pro- these countries do not yet have an effective
cedure for answering these questions has demand for large numbers of educated work-
been devised. Although a few studies have re- ers; itwill take considerable time to raise the
cently made noteworthy steps in the direction presently limited absorptive capacity of the
of measuring some consequences of an in- economy for educated persons; and a poor
crease intangible capital,6 no empirical study country cannot afford to pay for as much ed-
of investment in human capital is yet free ucation as can rich countries.9 Since educa-
from some arbitrary elements, and more sta- tional outlays compete for resources that
tistical evidence is needed. have an alternative use in directly productive
Another problem of particular importance investment, it is essential to determine what
to a country engaged in development pro- proportion of national income should be de-
gram ing isto determine at what phase of voted to education. And within the education
development the formation of intangible cap- system itself it is necessary to establish prior-
ital is most significant. It can be argued that ities for the various possible forms of educa-
a high rate of increase in the demand for im- tion and training.
provements inthe quality of inputs appears From the standpoint of accelerating devel-
only at a fairly advanced phase of develop- opment, the immediate requirements may
ment. The early industrialization in Western call for emphasis on vocational and technical
Europe, for example, appears to have been training and adult education rather than on
accomplished without requiring as prerequi- a greatly expanded system of formal educa-
sites marked improvements in skills and tion. Considering its high cost and the prob-
knowledge and health of workers.7 And the lems of absorption that it raises, even the case
contribution of education to American of universal primary education is questiona-
growth has been most pronounced in the ble. Professor Lewis expresses such skepti-
more recent decades, while capital invest- cism in the following comments on African
ment was more important in earlier decades.8 proposals:
Unlike the earlier historical situation, how- The limited absorptive capacity of most West
ever, itmay now be necessary to have a rel- African economies today — especially owing to the
atively high level of skill and much more backwardness of agriculture — makes frustration
knowledge to take advantage of the more and dislocation inevitable if more than 50 percent
complex equipment and techniques that may of children enter school. This, coupled with the
be obtained from advanced countries.
high cost due to the high ratio of teachers' salaries
There are additional questions to be raised to average national income, and with the time it
concerning what types of education should be takes to train large numbers of teachers properly,
emphasized, to what degree, and how soon. has taught some African countries to proceed with
Some economists have such questions in caution; to set the goal of universal schooling
twenty years ahead or more, rather than the ten
mind when they criticize — from the view- years ahead or less associated with the first flush
point of the economic, though not social or
of independence movements. Such a decision is re-
moral, value — proposals for mass education garded as highly controversial by those for whom
literacy is a universal human right irrespective of
6For example, Mary Jean Bowman, "Human Cap- cost. ... On the other hand, considering that in
ital: Concepts and Measures," in Money, Growth, most African territories less than 25 per cent of
and Methodology, Essays in Honor of Johan Aker-
man, Lund, 1961, pp, 147-68.
'These arguments are cogently presented by W.
7Cf. Schultz, "Investments in Human Capital," pp. Arthur Lewis, "Education and Economic Develop-
3-4,11-12.
ment," International Social Science Journal, Vol. 14,
8Edward F. Denison, "Education, Economic No. 4, 1962, pp. 685-99; Thomas Balogh, "Miscon-
Growth, and Gaps in Information," Journal of Polit- ceived Educational Programmes in Africa," Univer-
ical Economy, October 1962, p. 127. sities Quarterly, June 1962, pp. 243-9.
 613 EDUCATION

children aged six to fourteen are in school, a goal facilities for agriculture may also provide a
of 50 per cent within ten years may be held to con- way of encouraging rural school-leavers to
stitute revolutionary progress.10 take up work in the rural sector rather than
More immediately serious than the lack of migrating to the towns, and the special train-
universal primary education is the deficiency ing of young school-leavers may allow them
in secondary education. The most critical to act as the agents for introducing new and
manpower requirement tends to be for people improved agricultural techniques.
with a secondary education who can be man- For the broader problems of educational
agers, administrators, professional techni- requirements, the making of "manpower sur-
cians (scientists, engineers, agronomists, doc- veys" (as discussed in selection IX. C. 3) may
tors, economists, accountants, etc.), or furnish a useful basis for determining the
subprofessional technical personnel (agricul- principal skill shortages and what types of
tural assistants, technical supervisors, nurses, training activities should be emphasized. At
engineering assistants, bookkeepers, etc.). least for the short term, the provision of ag-
Lewis characterizes the products of second- ricultural extension services, training in me-
ary schools as "the officers and noncommis- chanical and technical skills, and training in
sioned officers of an economic and social sys- supervisory and administrative skills may
tem. A small percentage goes on to university contribute the most to fulfilling manpower
education, but the numbers required from requirements. After overcoming the imme-
the university are so small that the average diate bottlenecks of scarce personnel in spe-
country of up to five million inhabitants could cific key occupations, the education system
manage tolerably well without a university of should then be devised to provide a balance
its own. Absence of secondary schools, how- between general education, prevocational
ever, isan enormous handicap. . . . The mid- preparation, and vocational education and
training.
dle and upper ranks of business consists al-
most entirely of secondary school products, We may conclude that the recent attention
and these products are also the backbone of to investment in human capital should prove
public administratiion."11 salutary in cautioning against an overempha-
Also deserving of high priority is the infu- sis on physical capital to the neglect of the
sion of new skills and knowledge into the ag- more intangible factors. When considered for
ricultural sector. In order to achieve a system a poor country, however, investment in
of modern agriculture, the quality of labor in human capital calls for new approaches and
agriculture needs to be improved as an input special emphases that differ from those in ad-
in its own right and also to allow the use of vanced economies. An extensive system of
better forms of nonhuman capital (equip- formal education is a commendable objec-
ment, seeds, insecticides, etc.). In many tive— but it must necessarily be a distant ob-
countries that have experienced substantial jective. Instead of attempting to imitate the
increases in agricultural production, the key educational system of an advanced country,
factor has not been new land or land that is newly developing countries may more suita-
superior for agriculture; nor has it been bly concentrate, at least in the early phases
mainly the addition of reproducible capital. of their development programs, on methods
More importantly, the agricultural transfor- of informal education and on the objectives
mation has been based predominantly upon of functional education. These efforts are less
new skills and useful knowledge required to time-consuming, less costly, and more di-
develop a modern agriculture.12 Educational rectly related to manpower requirements
than is a formal educational system. As such,
10Lewis, "Education and Economic Development,"
p. 689. they are likely to prove most effective in im-
"Ibid., pp. 688-90. proving the economic quality of human
resources.
12Cf. Schultz, "Investment in Human Capital,"
p. 9.
 614 HUMAN-RESOURCE DEVELOPMENT

Comment
References to the literature on human capital theory are included in a number of selections
in this chapter. For a comprehensive survey of the subject, see Mark Blaug, "Human Capital
Theory: A Slightly Jaundiced Survey," Journal of Economic Literature (September 1976).
Also significant is M. J. Bowman, "The Human Investment Revolution in Economic
Thought," Sociology of Education, Vol. 39 (1966).
For a comprehensive survey of the state of knowledge about the effects of education on
incomes, see Timothy King (ed.), "Education and Income," World Bank Staff Working Paper
No. 402 (July 1980).
A critique of social cost-benefit analysis in educational planning is offered by G. S. Fields,
"Assessing Educational Progress and Commitment," Report for the U.S. Agency for Inter-
national Development (October 1978). See also World Bank, Education (1980); Ronald Dore,
The Diploma Disease (1976); M. Selowsky, "On the Measurement of Education's Contri-
bution to Growth," Quarterly Journal of Economics (August 1969); Dean T. Jamison and
Lawrence J. Lau, Farmer Education and Farm Efficiency (1981).

IX. C. 2. Educational Investment

The educational product, in the context of tive investment projects and of deciding
economic development. . . not only includes which one is to be included within a given
the components of education usually distin- level of overall capital formation. There is no
guished as consumption (i.e., enjoyment of reason why investment in human resources
the fuller life permitted by education) and as by education should not be included in such
direct investment (with the gains accruing an analysis. However, education investment
"internally" in the form of increased earn- has certain characteristics — quite apart from
ings to the educated person), but also edu- the previously noted factor of externality —
cation as investment in the functioning of the which pose special problems and should be
economic and social system at large. These noted at the outset.
latter gains accrue "externally," not only to
those in whom the educational input is in-
vested, but also to other members of the
CHARACTERISTICS OF
community.
EDUCATIONAL INVESTMENT
The theory of investment planning may be
looked at from the micro or macro level. In The product of education outlays, to begin
macro terms, the problem is to determine the with, carries joint features of consumption
alternative growth path available to the econ- and investment. For this reason, the share of
omy, assuming the best structure of capital resources allocated to education cannot be
formation to apply in each case, and then to considered wholly an investment outlay. The
choose the optimum path on the basis of the consumption component has to compete with
community's time preference. In micro alternative forms of consumption, while the
terms, the problem is one of rating alterna- investment component must compete with al-
ternative forms of capital formation. To the
*From Richard A. Musgrave, "Notes on Educa- extent that the two parts are inseparable, the
tional Investment in Developing Nations," in OECD
Study Group in the Economics of Education, Financ- proper allocation of resources to education
ing of Education for Economic Growth, Paris, 1966, should leave the rate of return thereon (com-
pp. 31-9. Reprinted by permission. puted as ratio of the present value of addi-
 615 EDUCATION

tional earnings to investment cost) below that is carried, and even longer spans must be al-
of alternative investments which do not carry lowed for if teacher-training is taken into
joint consumption components. consideration. Even though certain skills may
This distinction between the consumption be acquired fairly rapidly, especially if a pre-
and capital formation aspects of education vious foundation is laid, the educational cap-
outlays, however, is somewhat misleading. ital stock cannot be changed quickly, partic-
The consumption product of education may ularly for the more advanced type of
be divided into current consumption (the de- education. This introduces a constraint in in-
lights of attending school) and the future vestment planning and demands a corre-
consumption (the ability to appreciate life spondingly longer planning horizon which in
more fully later on). Since the latter is much turn points to the need for public policy guid-
the major element, the consumption compo- ance seen in the context of a long-term de-
nent islargely in the nature of a durable con- velopment perspective, if not development
sumer good and hence investment. The essen-
tial distinction, thus, is not between the A similar consideration relates to a further
consumption and investment aspects of edu- feature of investment in education, i.e., the
plan.
cation output, but between education invest- relatively long useful life of the education
ment which generates imputed income (the asset. Consideration of returns over, say, a
fuller life later on) and education investment thirty-year period, lends great weight to the
which generates increased factor earnings to importance of the discount factor in assessing
the labour supplied by the educated person. the relative productivity of investment in ed-
What weight is to be given to the two com- ucation. Since the useful life for competing
ponents inthe development context, and how investments tends frequently to be shorter,
is this to be reflected in the pattern of the ed- the relative case for investment in education
ucation programme? Recent writers have is low if the appropriate rate of discount is
pointed to the extension of secondary educa- high. Thus, the selection of the appropriate
tion as being the primary goal of education rate of discount is of particular importance in
policy in countries with a low level of educa- assessing the proper share for education in
tional capital stock, with extension of ele- total capital formation. There being no de-
mentary education and technical training at veloped capital markets which provide a clear
the subsequent level of capital stock, and ex- indicator of this rate, its determination be-
pansion of higher education at a more ad- comes essentially a matter of public policy.
vanced stage.1 While this priority is derived Investments should be ranked by a present
from the projected needs for various types of value rather than an internal rate of return
skill and training, it also suggests that the im- criteria. Moreover, there may well be a dif-
puted-income component of the education ference between the government's and the
mix tends to be of particularly great impor- private investor's evaluation of present, rela-
tance at the early stages. tive to future, needs. To the extent that pub-
Secondly, investment in education is char- lic policy takes a longer view, it will also tend
acterized bya gestation period which is sub- to require a larger share for education. in in-
stantially longer than that of many other vestment outlays.
types of capital formation. Indeed, education The relatively long, useful life, moreover,
seems the time-consuming, Boehm-Bawer- makes it necessary that the type of education
kian type of investment par excellence. Pe- be chosen in order to meet future demands in
riods of ten to twenty years may be involved, particular skills. This applies less to general
depending on how far the education process and elementary education, which lays a more
flexible basis, but becomes of great impor-
tance for specialized and technical types of
'See Frederick Harbison and Charles A. Myers,
Education, Manpower and Economic Growth. New training that do not permit easy conversion.
York, 1964, Chapters 4-6. As already noted, educational planning in the
 616 HUMAN-RESOURCE DEVELOPMENT

context of a longer-term development view is ing the value of education for a developed
essential. economy, but hardly for the underdeveloped
Finally, a word regarding the resource cost country, where external benefits constitute a
of education. Recent discussions of the eco- substantial part of the total gain.
nomics of education2 emphasize, and rightly Perhaps the most important aspect of the
external benefits of education lies in the
so, that this cost not only includes teachers'
salaries, buildings and equipment, but also change in the social and cultural climate, in-
the opportunity cost of lost income on the cident to the widening of horizons, which ed-
part of the student. Depending on the struc- ucation entails. As has been pointed out
ture of the developing country, this latter many times, such a change is an essential
component may be of varying significance. condition of success for many developing na-
Where there is a general surplus of labour tions. At the same time, this benefit result is
supply, the opportunity cost of forgone earn- not an automatic consequence of education
ings will be small or non-existent. Under at large, but only of the proper type, quality
other conditions, customary use of child la- and quantity of education. Supply of profes-
bour may produce the opposite situation. sional people who cannot be absorbed into
While the former is the more typical case, appropriate positions may readily become an
other components of education cost (school external dis-economy and source of
instability.
teachers' salaries in particular) tend to be
relatively high in underdeveloped countries.3 As noted previously, different types of in-
Even though the income stream from a given vestments inthe early stages of development
factor input into education will be large (ed- tend to be highly complementary, and this
ucated workers are needed to take advantage holds par excellence for the proper combina-
of modern techniques, and education is tion of investment in education and capital
highly complementary to other types of cap- equipment. Without a capable labour force
ital formation) the rate of return on educa- modern capital equipment cannot be oper-
tional investments (relative to that on other ated, and it is precisely the access to superior
investments) is therefore not as high as sug- techniques that is the one hopeful factor in
gested bythe income stream alone. the development picture. Now it is true that
the existence of such complementarity need
EXTERNALITIES not, per se, constitute an externality which
would fail to be recorded in a perfectly func-
The standard procedure for determining tioning pricing system. If tractors can be sub-
the value of investment in education is to es- stituted for oxen only if trained drivers are
timate the future stream of incremental earn- available, this will be reflected in correspond-
ings which accrue to the student, and to dis- ing higher wages for the driver. The trouble,
count it to obtain the present value. This however, lies in the fact that in an economy
present value is then related to the cost of in- where there are as yet no tractors, the supply
vestment inobtaining the rate of return. This of drivers is unlikely to be forthcoming, and
procedure excludes additions to output which vice versa. The generation of growth
accrue externally and to the benefit of others (whether "balanced" in an overall Nurkse
rather than to the educated person alone. sense or not) requires a concerted effort to
Such oversight may be acceptable in assess- provide a chain of investments, thereby re-
ducing the risks of individual investments
2For a survey of this literature, see Chapter I in
William J. Bowen, Economic Aspects of Education,
and making it possible for an investment pro-
Princeton, 1964. gramme to succeed where individual invest-
ments would fail.
3See W. Arthur Lewis, "Priorities for Educational
Expansion," Policy Conference on Economic Growth This necessity for investment planning ex-
and Investment in Education, OECD, Washington, ists even for investments which, given the
1961, Part III, p. 37. necessary supply of entrepreneurial talent,
 617 EDUCATION

would be appropriately undertaken privately. developed country need not mean that the
It exists par excellence for investment in ed- absorptive capacity, as we understand it, is
ucation. Left to household decisions, neither low. The rate of return on education equals
the market knowledge, foresight or financial the ratio of present value of wages, or earn-
requirements are present which are needed to ings due to education, to the cost of produc-
secure adequate supplies. This is especially ing it. The high relative wage of the educated
the case in underdeveloped countries where person may merely express the extreme scar-
the whole attitude towards education has to city of the education factor, and hence its
overcome conventional barriers and become ability to command a high return at the given
reoriented to the development process. . . . cost of producing it. The high wage might be
Sensible education targets, therefore, must an indicator of under- rather than over-sup-
be developed by considering the needs of the ply, as demand may prove elastic if only in-
particular economy and the demands posed creased supplies could be made available. At
by its specific plans. This raises a question of the same time, the cost of producing educa-
the rate at which the system needs to, or can tion may be high at early stages of develop-
absorb additional supplies of educated man- ment, relative to that of producing other cap-
power. Arthur Lewis stresses the fact that ital goods, say tractors. If so, the rate of
absorption capacity is limited by the high return on education (relative to that on trac-
cost of education in developing countries. To tors) will not be as high as suggested by the
quote: high income stream (wages) accruing to the
educated factor alone. The cost of producing
The main limitation on the absorption of the edu- education must be lowered before increased
cated in poor countries is their high price, relative
investment in education is profitable. This,
to average national output per head. ... In conse-
quence, all production . . . which depends on using however, is subject to the previously noted
educated people is much more expensive, in rela- condition that education is highly comple-
tion to national income, in poor than in rich coun- mentary to other forms of capital formation.
tries. The poor countries may need the educated In addition to genuine scarcity, the high
more than the rich, but they can even less afford salaries demanded for work in positions re-
to pay for or absorb large numbers. ... In the long quiring education reflect conventional factors
run, the situation adjusts itself because the pre- and the desire to absorb the status previously
mium for education diminishes as the number of
held by the colonial official. As a result, there
educated increases. ... As the premium for edu-
cation falls, the market for the educated may is frequently a "minimum wage structure"
widen enormously. for high-level work which exceeds the eco-
nomic return on such work, and prices edu-
And he further notes that
cated services out of the market. Such rigid-
ities need to be broken down to permit an
to give eight years of primary education to every
child would cost at current prices about 0.8 per effective education policy, and to the extent
cent of national income in the USA, 1.7 per cent that they are permitted to prevail, appropri-
in Jamaica, 2.8 per cent in Ghana and 4.0 per cent ate allowance must be made in formulating
in Nigeria. The main reason for this difference is realistic education targets. A policy which re-
that, while the average salary of a primary school sults in a supply of education that cannot be
teacher is less than one and a half times per capita absorbed is obviously inefficient.
national income in the USA, a primary school A further aspect of the cost problem which
teacher gets three times per capita national income deserves special attention is the extent to
in Jamaica, five times in Ghana, and seven times which the educational effort involves a need
in Nigeria.4 for foreign exchange. Capital formation in
The fact that school teachers' wages are the developing countries is frequently limited
high relative to average wages in the under- by the fact that capital goods cannot be pro-
duced at home, and that the foreign ex-
4W. A. Lewis, op. cit., pp. 37-38. change available for acquiring them abroad
 618 HUMAN-RESOURCE DEVELOPMENT

is scarce. The question arises whether invest- PRIORITIES

ment in education is more or less capital in-
tensive than are other, and to some extent al- The matter of educational priorities is ob-
viously of vital importance. Unless the right
ternative, forms of capital formation. The
kind of education is provided, setting overall
answer depends on the level of education.
targets has little meaning. Educated people
While the exchange component of elemen- who are unable to find suitable jobs, not only
tary education cost tends to be relatively low,
fail to add to the national product but be-
it rises at the middle level, and for higher ed- come a source of political instability. Since
ucation itbecomes extremely high. At later
the cost of various types of education (pri-
stages, this pattern may fit the needs and re-
sources of the country but at the earlier mary versus secondary versus advanced; lib-
eral arts versus technical, and so forth) dif-
stages, where emphasis is on middle-level ed- fers greatly, the very setting of overall targets
ucation, the match tends to be an unfortu- has to be derived from the structural com-
nate one and points to the need for an all-out position ofthe education supply.
effort in domestic teacher-training.

Comment
Posing the question — "Why Isn't the Whole World Developed?" — Richard A. Easterlin
answers that the worldwide spread of modern economic growth has depended chiefly on the
diffusion of a body of knowledge concerning new production techniques. The acquisition and
application of this knowledge by different countries have been governed largely by whether
their populations have acquired traits and motivations associated with formal schooling. See
his paper "Why Isn't the Whole World Developed?," Journal of Economic History (March
1981).
The World Bank's Education Sector Policy Paper (1980) concludes that "Studies have
shown that economic returns on investment in education seem, in most instances, to exceed
returns on alternative kinds of investment, and that developing countries obtain higher returns
than the developed ones."
For an extensive bibliography on education and development, see Mark Blaug, (ed.), Eco-
nomics ofEducation, Vols. 1 and 2 (1968, 1969). Several writings by T. W. Schultzare highly
instructive on the role of education in development: "Capital Formation by Education," Jour-
nal ofPolitical Economy (December 1960); "Investment in Human Capital," American Eco-
nomic Review (March 1961); "Education and Economic Growth," in N. B. Henry (ed.), So-
cial Forces Influencing American Education (1961); "Investment in Human Capital in Poor
Countries," in P. D. Zook (ed.), Foreign Trade and Human Capital (1962). The special sup-
plement ofthe Journal of Political Economy (October 1962) also contains a number of per-
tinent papers covering particular aspects of the problem of "Investment in Human Beings."
Rigorous models of educational planning are presented by S. Bowles, "A Planning Model
for the Efficient Allocation of Resources in Education," Quarterly Journal of Economics
(May. 1967); G. Correa and J. Tinbergen, "Quantitative Adaptation of Education to Eco-
nomic Growth," Kyklos (1962); R. G. Davis, Planning Human Resource Development (1966);
J. Tinbergen and H. C. Bos, Econometric Models of Education (1965).
Also informative are: Marcelo Selowsky, "A Note on Preschool-Age Investment in Human
Capital in Developing Countries," Economic Development and Cultural Change, Vol. 24 (July
1976) and M. R. Rosenzweig and R. E. Evenson, "Fertility, Schooling and the Economic
Contribution of Children in Rural India," Econometrica, Vol. 45 (July 1977).
EXHIBIT IX.7. Enrollment Ratios by Region, 1960-75 (percent)
Middle East Latin Sub-Saharan
and North Africa South Asia East Asia America Africa

Male (6-11)

30 -Female (12-17)
Female (12-17) Male (12-17)
—
20 'Male (18-23) Female (18-23) Male (18-23) Female (12-17)
Male (18-23) Female (18-23)
Male (18-23)
Female (18-23) Female (18-23)
1960 65 70 75 60 65 70 75 60 65 70 75 60 65 70 75 60 65 70 75
Source: UNESCO.

EXHIBIT IX.8. Rates of Return to Education (percent)

Primary Number of
Secondary
Country Group Countries
Higher
12.3 30
All developing countries 24.2 15.4
Low income/adult literacy
27.3 17.2 12.1 11
rate under 50 percent"
Middle income/adult literacy
rate over 50 percent 22.2 14.3 12.4 19
Industrialized countries — 10.0 9.1
14
Source: World Bank.
Note: In all cases, the figures are "social" rates of return: the costs include forgone earnings (what the students could have
earned had they not been in school) as well as both public and private outlays; the benefits are measured by income before
tax. (The "private" returns to individuals exclude public costs and taxes, and are usually larger.) The studies refer to various
years between 1957 and 1978, mainly in the latter half of the period.
"In this sample of 30 developing countries, those countries with low incomes also had literacy rates below 50 percent (at
the time the studies were done). All the middle-income countries had literacy rates above 50 percent.

IX.C.3. Approaches to Human-

Resource Development*

Most modernising economies are confronted se of persons with critical skills in the mod-
simultaneously with two persistent, yet seem- ernising sector and surplus labour in both the
ingly diverse, manpower problems: the short- modernising and traditional sectors. Thus,

*From Frederick H. Harbison, "Human Resources sis Approach to Human-Resource Development Plan-
Development Planning in Modernising Economies," ning," in UNESCO, Manpower Aspects of Educa-
International Labour Review, Vol. 85, No. 5, May tional Planning, Paris, 1968, pp. 57-8. 59-64. Re-
1962, pp. 435-38, 440-1, 456-58; "A Systems Analy- printed bypermission.
 620 HUMAN-RESOURCE DEVELOPMENT

the strategy of human resources development who are qualified to enter a technical insti-
is concerned with the twofold objective of tute may also be qualified to enter a univer-
building skills and providing productive em- sity, and they prefer the latter because of the
ployment for unutilised or underutilised higher status and pay which is accorded the
manpower. The shortages and surplus of holder of a university degree. Finally, there
human resources, however, are not separate are often fewer places available in institu-
and distinct problems; they are very inti- tions providing intermediate training than in
mately related. Both have their roots in the the universities.
changes which are inherent in the develop- 3. The shortage of top-level managerial
ment process. Both are related in part to ed- and administrative personnel, in both the pri-
ucation. Both are aggravated as the tempo of vate and public sectors, is almost universal,
modernisation is quickened. And, paradoxi- as is the dearth of persons with enterpreneu-
cally, the shortage of persons with critical trial talents.
skills is one of the contributing causes of the 4. Teachers are almost always in short sup-
surplus of people without jobs. Although the ply, and their turnover is high because they
manpower problems of no two countries are tend to leave the teaching profession if and
exactly alike, there are some shortages and when more attractive jobs become available
surpluses which appear to be universal in in government, politics, or private enterprise.
modernising societies. The scarcity is generally most serious in sec-
The manpower shortages of modernising ondary education, and particularly acute in
countries are quite easy to identify, and fall the fields of science and mathematics. This
into several categories:
shortage of competent teachers is a "master
1. In all modernising countries there is bottleneck" which retards the entire process
likely to be a shortage of highly educated of human resources development.
professional manpower such as, for example, 5. In most modernising countries there are
scientists, agronomists, veterinarians, engi- also shortages of craftsmen of all kinds as
neers, and doctors. Such persons, however, well as senior clerical personnel such as book-
usually prefer to live in the major cities keepers, secretaries, stenographers, and busi-
rather than in the rural areas, where in many ness machine operators.
cases their services are most urgently needed. 6. Finally, there are usually in addition
Thus, their shortage is magnified by their rel- several other miscellaneous categories of per-
ative immobility. And, ironically, their skills sonnel in short supply, such as, for example,
are seldom used effectively. In West Africa radio and television specialists, airplane pi-
and also in many Asian and Latin American lots, accountants, economists and stat-
countries, for example, graduate engineers isticians.
may be found managing the routine opera- I shall use the term "high-level man-
tion of an electric power substation or doing
power," or alternatively "human capital," as
the work of draughtsmen. Doctors may spend a convenient designation for the persons who
long hours making the most routine medical fall into categories such as those mentioned
tests. The reason is obvious.
above. The term "human capital formation,"
2. The shortage of technicians, nurses, ag- as used in this paper, is the process of acquir-
ricultural assistants, technical supervisors, ing and increasing the numbers of persons
and other sub-professional personnel is gen- who have the skills, education, and experi-
erally even more critical than the shortage of ence which are critical for the economic and
fully qualified professionals. For this there political development of a country.
are several explanations. First, the modernis- The analysis of human capital formation is
ing countries usually fail to recognise that the thus parallel and complementary to the study
requirements for this category of manpower of the processes of savings and investment (in
exceed by many times those for senior profes- the material sense). In designing a strategy
sional personnel. Second, the few persons for development, one needs to consider the
 621 EDUCATION

total stock of human capital required, its mally short of critical skills. Indeed, as long
rates of accumulation, and its commitment to as the pace of innovation is rapid, the appe-
(or investment in) high-priority productive tite of any growing country for high-level
activities. manpower is almost insatiable.
The rate of modernisation of a country is
associated with both its stock and rate of ac- As indicated above, no two countries have
cumulation of human capital. High-level exactly the same manpower problems. Some
manpower is needed to staff new and expand- have unusually serious surpluses, and others
ing government services, to introduce new have very specialised kinds of skill bottle-
systems of land use and new methods of ag- necks. Politicians and planners, therefore,
riculture, todevelop new means of commu- need to make a systematic assessment of the
nication, to carry forward industrialisation, human resources problems in their particular
and to build the educational system. In other countries. Such assessment may be called
words, innovation, or the process of change "manpower analysis."
from a static or traditional society, requires The objectives of manpower analysis are as
very large "doses" of strategic human capi- follows: (1) the identification of the principal
tal. The countries which are making the most critical shortages of skilled manpower in each
rapid and spectacular innovations are invar- major sector of the economy, and an analysis
iably those which are under the greatest pres- of the reasons for such shortages; (2) the
sure to accumulate this kind of human capi- identification of surpluses, both of trained
tal at a fast rate. Here we may make two manpower as well as unskilled labour, and
tentative generalisations: the reasons for such surpluses; and (3) the
First, the rate of accumulation of strategic setting of forward targets for human re-
human capital must always exceed the rate sources development based upon reasonable
of increase in the labour force as a whole. In expectations of growth. Such forward targets
most countries, for example, the rate of in- are best determined by a careful examination
crease inscientific and engineering personnel and comparison, sector by sector, of the util-
may need to be at least three times that of the isation ofmanpower in a number of countries
labour force. Sub-professional personnel may which are somewhat more advanced politi-
have to increase even more rapidly. Clerical cally, socially and economically.
personnel and craftsmen may have to in- Manpower analysis cannot always be
crease at least twice as fast as the labour based on an elaborate or exhaustive survey.
force, and top managerial and administrative It is seldom possible to calculate precisely the
personnel should normally increase at a com- numbers of people needed in every occupa-
parable rate. tion at some future time. But, whether statis-
Second, in most cases, the rate of increase tics are available or not, the purpose of man-
in human capital will need to exceed the rate power analysis is to give a reasonably
of economic growth. In newly developing objective picture of a country's major human
countries which already are faced with criti- resources problems, the inter-relationships
cal shortages of highly skilled persons, the between these problems, and their causes, to-
ratio of the annual increase in high-level gether with an informed guess as to probable
manpower to the annual increase in national future trends. Manpower analysis is both
income may need to be as high as three to qualitative and quantitative, and it must be
one, or even higher in those cases where ex- based upon wise judgment as well as upon
patriates are to be replaced by citizens of the available statistics. In countries where statis-
developing countries. tics are either unavailable or clearly unrelia-
The accumulation of high-level manpower ble, moreover, the initial manpower anal
to overcome skill bottlenecks is a never-end- may be frankly impressionistic. . . .
ing process. Advanced industrial societies as Once the manpower problems of a newly
well as underdeveloped countries are nor- developing country are identified, a strategy
 622 HUMAN-RESOURCE DEVELOPMENT

must be developed to overcome them effec- strategy are interdependent, and call for a
tively. The essential components of such a well-designed and integrated attack on all
strategy are the following: (1) the building of three fronts at once. And it is imperative that
appropriate incentives; (2) the effective train- the strategy of building and utilising human
ing of employed manpower; and (3) the ra- resources be an integral part of a country's
tional development of formal education. national development programme.
These three elements are interdependent.
Progress in one area is dependent upon prog- The major thesis of this paper is that the
ress in the other two. The country's leaders manpower approach should encompass much
should not concentrate on only one or two of more than a tabulation of "heads and hands"
them at a time; they must plan an integrated in precise occupational categories. It must go
attack on all three fronts at once. far beyond the construction of purely quan-
titative forecasts, projections, or targets for
The argument has been made that invest- formal education. It should be related to a
ments in formal education alone are not broad strategy of human-resource develop-
likely to solve either critical skill shortages or ment rather than to a narrow concept of ed-
persistent labour surpluses in modernising so- ucation planning.
cieties. Investments in education are likely to Without questioning the usefulness and
contribute effectively to rapid growth only importance of the kind of quantitative anal-
(1) if there are adequate incentives to en- ysis which is characteristic of most man-
courage men and women to engage in the power surveys, I suggest that it may now be
kinds of productive activity which are needed appropriate to use in addition a systems anal-
to accelerate the modernisation process; and ysis concept. It should be possible to look at
(2) if appropriate measures are taken to shift the various constituent elements of human-
a large part of the responsibility for training resource development as a system which is
to the principal employing institutions. The somewhat analogous to a system for the gen-
building of incentives and the training of em- eration and distribution of electric power. In
ployed manpower, therefore, are necessary using this frame of reference, one can iden-
both as a means of economising on formal ed- tify skill-generating centres, such as, for ex-
ucation and as a means of making the invest- ample, universities, training institutes, and
ment in it productive. . . . employing organizations, which develop peo-
The third component of the strategy is wise ple on the job. The linkages between such
judgment and prudent investment in building centres are analogous to transmission lines.
the system of formal education. This calls for The manpower problems, such as skill short-
giving priority to investment in and develop- ages and labour surpluses, encountered by
ment of broad secondary education. It re- developing countries may be thought of as at-
quires that the costs of universal primary ed- tributable topower failures in particular gen-
ucation be kept as low as possible by applying erating centres, ineffective linkages between
new technologies which can make effective these centres, or faulty design resulting in the
use of relatively untrained teachers and failure of the total system to carry the loads
which can multiply the contribution of a very expected of it. A system of human-skill gen-
small but strategic group of highly trained eration should be designed to carry varying
professionals. Finally, in the area of higher loads; it must have built-in flexibility to meet
education, the strategy stresses the need for such loads; it must be adequate in size; and
giving priority to investment in intermediate- above all its components must be properly
level training institutions and the scientific balanced. The systems-analysis approach
and engineering faculties of universities. makes it easier to identify in operational
But this does not mean that the production terms major problem areas, and it compels
of liberally educated persons should be
neglected. the analyst to examine the critical inter-re-
lationships between various manpower, edu-
The three essential components of the cation, and economic-development pro-
 623 EDUCATION

grammes. It provides a logical starting-point other the problem would "go away," the
for building a strategy of human-resource human-resource development planner cannot
development. escape responsibility for considering ways
and means of absorbing surplus manpower
HUMAN-RESOURCE PROBLEMS IN and directing it into productive activities.
DEVELOPING ECONOMIES The evaluation of occupational needs and
skill-generating capacity has been a tradi-
The major human-resource problems in tional concern of manpower specialists. Here,
developing societies are: (a) rapidly growing unlike the situation with unemployment, it is
population; (b) mounting unemployment in possible to suggest viable solutions for rather
the modern sectors of the economy as well as clearly defined problems. Manpower require-
widespread underemployment in traditional ments can be determined; appropriate pro-
agriculture; (c) shortage of persons with the grammes offormal education and on-the-job
critical skills and knowledge required for ef- training can be devised; and progress toward
fective national development; (d) inadequate achievement of goals can be measured.
or underdeveloped organizations and institu- In setting targets for education and train-
tions for mobilizing human effort; and (e) ing programmes, the analyst is concerned
lack of incentives for persons to engage in with two related but distinct concepts —
particular activities which are vitally impor- "manpower requirements" and "absorptive
tant for national development. There are ob- capacity." "Manpower requirements" may
viously other major human-resource devel- be defined as clearly evident needs for per-
opment problems such as nutrition and sons with particular education, training, and
health, but these lie for the most part in other experience. The assumption here is that such
technical fields and are thus beyond the scope persons are necessary, if not indispensable,
of this paper. for the achievement of a programme of na-
The manpower analyst, of course, is partic- tional development. "Absorptive capacity" is
ularly interested in the present and future a looser term which refers to a country's ca-
size of the labour force, its growth rates in pacity to provide some kind of useful employ-
both the traditional and modern sectors, and ment for persons with certain educational
the factors which determine labour-force qualifications. In effect, "manpower require-
participation of various groups. Of necessity ments" should express minimum or essential
he must also be concerned with the conse- needs; "absorptive capacity" should express
quences of policies to limit population the maximum number of persons who can be
growth. For example, a reduction in birth- employed without encountering redundancy
rates will not immediately lead to a reduction or serious under-utilization of skill. The skill-
in the labour force, but at the same time it generating centres, therefore, should produce
would probably increase a country's propen- trained manpower within this range between
sity to save and to invest in productive activ- the maximum and the minimum; otherwise
ities. Population control, therefore, in addi- the skill-generation system is distorted or
tion to its other obvious benefits, may unbalanced.
contribute directly to greater labour produc- The "demand" for education or training
tivity. Certainly, the human-resource devel- must be distinguished from the allowable
opment strategist should give very close at- range between manpower requirements and
tention to population problems and assume absorptive capacity. Demand stems from so-
greater responsibility for proposing popula- cial and political pressures for various kinds
tion-control measures. of education as well as from the willingness
Mounting unemployment in urban areas is of people to pay fees to acquire it. Thus, for
probably the most serious and intractable example, the demand for university educa-
problem facing today's newly developing tion may be very high because of the status,
countries. .. . prestige, and pay enjoyed by graduates; but,
Although he might wish that somehow or in many countries, this results in the produc-
 624 HUMAN-RESOURCE DEVELOPMENT

tion of graduates who cannot be effectively the higher levels exercise authority over the
absorbed in the economy. When demand is lower levels.
clearly out of step with requirements or ab- The successful leaders of organizations, or
sorptive capacity, the country's educational more accurately the "organization builders,"
system is clearly distorted or out of balance are in any society a small but aggressive mi-
with the needs for national development. In nority committed to progress and change.
using the systems analysis approach, a major They feed the aspirations, give expression to
task of the human-resource planner is to de- the goals, and shape the destinies of peoples.
tect actual and potential distortion and to They play the principal roles on the stage of
consider measures for achieving a proper history, and they organize the march of the
balance. masses.
Another type of distortion in many coun- A major problem in many developing
tries isthe underdevelopment, if not outright countries is "organizational power failures."
neglect, of appropriate measures of training Often government ministries, commercial
persons in employment. A great deal of and industrial organizations, or educational
money is wasted in formal pre-employment institutions simply fail to "deliver the goods."
craft or technical training which could be Usually, the trouble may be traced to a
provided more efficiently and cheaply by em- dearth of "prime movers of innovation."
ploying establishments. Also the efficiency of Who then are these prime movers of inno-
skill-generating systems could be greatly im- vation? Certainly the entrepreneur who per-
proved by closer linkages between schools ceives and exploits new business ventures be-
and universities and the employing institu- longs to this group, as does the manager or
tions. For some reason, education planners top administrator in public establishments.
have been inclined to think that on-the-job He may not always have new ideas of his
development lies beyond their legitimate con- own, but his function is to organize and stim-
cern, and at the same time they appear to ulate the efforts of others. He structures or-
have ignored the task of building the neces- ganizations, and either infuses hierarchies
sary bridges between formal education and with energy and vision or fetters them with
in-service training. The systems-analysis ap- chains of conformity. But effective orgniza-
proach helps to highlight this underdeveloped tions also need other creative people. The
area of concern. agronomist who discovers better measures of
In the past, manpower analysis has centred cultivation, and the agricultural assistants
on measurement of needs for various cate- who teach the farmers to use them, belong to
gories of high-level manpower, and in doing the innovator class, as do public-health offi-
so it has usually overlooked the vital problem cers, nurses, and medical assistants. Engi-
of organization and institution building. Suc- neers are in essence designers of change, and
cessful development requires the building of engineering technicians and supervisors put
effective government organizations, private the changes to work. And last but not least,
encerprises, agricultural extension forces, re- professors, teachers, and administrators of
search institutions, producer and consumer educational institutions in many countries
co-operatives, education systems, and a host may constitute the largest group of prime
of other institutions which mobilize and di-
movers of innovation, as they are the "seed-
rect human energy into useful channels. Or- corn" from which new generations of man-
ganization ias factor of production, separate power will grow.
from labour, high-level manpower, capital, or Some innovators are "change-designers"
natural resources. The essence of organiza- who make new discoveries, suggest new
tion is the co-ordinated effort of many per- methods of organization, and plan broad new
sons toward common objectives. At the same
strategies. Others are "change-pushers" who
time the structure of organization is a hier- are able to persuade, coach and inspire peo-
archy of superiors and subordinates in which ple to put new ideas to work. Some innova-
 625 EDUCATION

tors, of course, are at the same time change- sation between the agricultural officer and
designers and change-pushers. But whether agricultural assistant are too great; and the
they are designers, pushers, or a combination earnings of scientists and engineers, in com-
of the two, the prime movers of innovation parison with administrative bureaucrats in
must have extensive knowledge and experi- government ministries, are too low. The pref-
ence. Thus, for the most part, they are drawn erences for urban living, the forces of tradi-
from the ranks of high-level manpower. But tion, and historical differentials all tend to
they need more than proven intelligence and distort the market for critical skills. It follows
thorough technical training. They should then that the demand for certain kinds of ed-
have in addition keen curiosity, a capacity for ucation, particularly at the university level, is
self-discipline, and an unquenchable desire inflated relative to the country's absorptive
for accomplishment. They should be adept at capacity. The human-resource development
asking questions. They should have the knack planner must therefore consider deliberate
of stimulating others to produce ideas and to measures to influence the allocation of man-
activate the ablest minds about them; and power into high-priority activities and occu-
they should be able to sell ideas to superiors, pations. Such measures may include major
subordinates, and associates. The prime changes in the wage and salary structure,
mover of innovation must be convinced that scholarship support for particular kinds of
change can occur as a result of individual ac- education and training, removal of barriers
tion, and he must have the drive within him against upward mobility, and in some cases
to bring it about. This may stem from a de- outright compulsion. As many developing
sire to rise in social status, to build up mate- countries have learned to their chagrin, in-
rial wealth, to acquire political influence, or vestments ineducation can be wasted unless
to preserve an already established prestige men and women have the will to prepare for
position. and engage in those activities which are most
Many of the persons holding commanding critically needed for national development.
positions in organizations are conformists or These then are the problems and tasks
even obstructors of innovation. They must be which face the human-resource development
systematically replaced by more creative in- planner — the consequences of population in-
novators. The human-resource development creases and the measures for controlling
planner should be able to locate the critical them; underemployment and unemployment
points of power loss in organizational struc- in both the traditional and modern sectors;
tures and to suggest remedial measures. skill shortages and the processes of develop-
A final problem area in human-resource ing high-level manpower to overcome them;
development is incentives. It is one thing to organizational weakness and the need to find
estimate the needs for manpower of various prime movers of innovation for institutional
qualifications but quite another to induce development; and provision of both financial
persons to prepare for and engage in occu- and non-financial incentives in order to direct
pations which are most vital for national critically needed manpower into productive
growth. In most developing countries, it is in- channels. Some of these are subject to quan-
correct to assume that relative earnings and titative analysis; others are purely qualita-
status reflect the value of the contribution of tive; and a few are subject only to intuitive
individuals to development. Pay and status judgment. But, they are all interrelated. The
are often more related to tradition, colonial systems approach forces the analyst to ex-
heritage, and political pressures than to pro- amine them simultaneously as he searches
ductivity. Characteristically, for example, for the weak spots — the points of power fail-
the rewards of subprofessional personnel and ure or the major areas of distortion — in a
technicians are far from sufficient to attract country's overall effort to effectively develop
the numbers needed — the pay of teachers is and utilize its human resources.
often inadequate; the differentials in compen-
 626 HUMAN-RESOURCE DEVELOPMENT

Comment
The activity of entrepreneurship is also essential for the development process. An interdis-
ciplinary approach — combining economic, psychological, and social factors — may best ana-
lyze the factors determining the supply of entrepreneurship. Especially significant is David
McClelland's analysis of achievement motivation as a psychological determinant of entrepre-
neurship; see The Achieving Society, 2nd ed. (1976). It may be suggested, however, that suc-
cessful entrepreneurship depends not only on an individual's motivation, but also on his or her
abilities and a permissive environment that provides incentives and opportunities for entre-
preneurs. Economists have tended to emphasize a favorable economic environment, but ac-
cording toMcClelland, whenever economic growth begins, some tiny community within the
larger society which has played the major entrepreneurial role can nearly always be identified.
The theory of achievement motivation predicts that those in the population who will take
greater advantage of increased opportunities are precisely those relatively few individuals who
have a high need to achieve. The policy implication, therefore, is that efforts must be under-
taken to instill a higher need to achieve in a larger segment of the population. This conforms
to the more general view expressed throughout this chapter that the key to development is
human resources, and that the abilities, values, and attitudes of people must be changed in
order to accelerate the process of development.
An extensive bibliography on accelerating economic development through psychological
training is presented by D. C. McClelland and D. G. Winter, Motivating Economic Achieve-
ment (1969). Critiques of the influences of achievement motivation on entrepreneurship are
offered by S. N. Eisenstadt, "The Need for Achievement," Economic Development and Cul-
tural Change (July 1963) and S. P. Schatz, "w Achievement and Economic Growth," Quar-
terly Journal of Economics (May 1965).
Other instructive studies of entrepreneurship are F. H. Harbison, "Entrepreneurial Orga-
nization as a Factor in Economic Development," Quarterly Journal of Economics (August
1956); E. Hagen, On the Theory of Social Change (1962); W. P. Glade, "Approaches to a
Theory of Entrepreneurship Formation," Explorations in Entrepreneurial History (Spring-
Summer 1967); Peter Marris, "The Social Barriers to African Entrepreneurship," Journal
of Development Studies (October 1968); G. F. Papanek, Pakistan's Development (1967),
chapter 2; Peter Kilby (ed.), Entrepreneurship and Economic Development (1971); H. Lei-
benstein, General X-Efficiency Theory and Economic Development (1976); T. W. Schultz,
"Investment in Entrepreneurial Ability," Scandinavian Journal of Economics (1980).
An important work on sociological and psychological modernization is Alex Inkeles and D.
H. Smith, Becoming Modern (1975). For an annotated bibliography, see John Brode, The
Process of Modernization (1975).

IX. C. 4. Education and

Employment*

Although the linkages between education causality, recent research results have
and employment are complex and, in the yielded new insights about the nature of these
past, often analyzed with simplistic notions of linkages.1 A number of international organi-
*From Edgar O. Edwards and Michael P. Todaro, 'For the most recent and comprehensive analysis of
"Education, Society and Development: Some Main the interrelationships between education and employ-
Themes and Suggested Strategies for International ment, see Mark Blaug, Education and the Employ-
Assistance," World Development, January 1974, pp. ment Problem in Developing Countries, Geneva,
27-30. Reprinted by permission. 1973.
 627 EDUCATION

zations, including the JLO and the Ford and The educational composition of the system's
Rockefeller Foundations, have been engaged output also will accord with the dictates of
in intensive analyses and have conducted nu- aggregate demand with the private system
merous conferences and meetings on the em- once again meeting needs unmet by the pub-
ployment topic over the past two years.2 The lic system. Since aggregate private demand
authors have participated in a number of will normally exceed job opportunities and
these activities and the following represents a since this demand will tend to be satisfied
condensed statement of our current thinking either politically (through the public system)
about the principal relationships between ed- or privately, educational systems in LDCs
ucation and employment. In the interests of are likely to overproduce in terms of the real
brevity, the argument is put forward as a se- manpower needs of the nation. Moreover,
ries of major propositions and derivative this divergence between the output of the ed-
strategies relating to the education-employ- ucational system and the number of new job
ment nexus. opportunities is likely to grow over time.
Given these inexorable pressures to overpro-
duce, major modifications in the educational
system must depend primarily on changes in
economic incentives and social constraints
SEVEN BASIC PROPOSITIONS
and not on isolated initiatives taken within
ABOUT EDUCATION AND
EMPLOYMENT the educational system itself.
Proposition 2. Education, in conjunction
Proposition 1. Pressures outside the edu- with the development of a modern sector,
cational system mil force it to overproduce in transforms various kinds of underemploy-
terms of the real manpower needs of devel- ment into open unemployment, primarily
opment, giving rise to the phenomenon of ed- through rural-urban migration.
ucated unemployment. In traditional societies it has been custom-
An educational system in any society both ary for members to share among themselves
reflects and adapts itself to the political and the work which the community considered
economic circumstances of that society. Eco- necessary to perform. Even in feudal land-
nomic signals or incentives and socio-political lord-tenant settings, the landlord often re-
constraints operative in the society are the garded itas his responsibility to share avail-
primary factors determining the employment able work among all of those who accepted
aspirations of individuals and hence the level his patronage. In most LDCs the develop-
and composition of the aggregate private ment of a modern sector has introduced very
demand for education. This demand in turn different working arrangements such as the
influences the supply, composition, and aca- discipline of a standard work-week and min-
demic standards of educational opportuni- imum wages. The modern sector has also
ties, both public and private. Political pres- been characterized by an urban, industrial
sures force the public educational system to bias and a need for a new set of specialized
attempt to meet quantitatively the aggregate occupational skills. Educational systems re-
demand for school places. In the event that sponded quickly to these modern sector
the number of these places is limited by pub- needs, focusing curricula on modern-sector
lic financial constraints or policy decisions, educational requirements and strengthening
the private sector will tend to find alternative urban schools more quickly than those in
ways of meeting unmet educational demand.3 rural areas. These several contemporaneous

2Edgar O. Edwards (ed.), Employment in Devel- 3Some examples of private-sector initiatives in re-
oping Nations, New York, 1974; and ILO, Scope. sponse to unsatisfied, aggregate demands include the
Approach and Content of Research Oriented Activi- Harambee school movement in Kenya and the prolif-
ties of the World Employment Programme. Geneva, eration of private colleges and universities in India
1972. and the Philippines.
 628 HUMAN-RESOURCE DEVELOPMENT

events have meant that many young people to provide continuing political cover for an
previously underemployed in rural areas have increasingly explosive situation.5
flocked to the cities in search of education or Proposition 4. Rising unemployment, dis-
jobs in numbers substantially in excess of the torted wage differentials, and excessive edu-
opportunities available. Many, indeed grow- cation must be taken into account if social
ing numbers, have become openly unem- rate-of-return and benefit/cost calculations
ployed, a vivid and visible transformation of are to play a role in educational planning.
widely dispersed rural underemployment into Although political factors are dominant in
the far more troublesome and potentially ex- LDC government decisions with regard to
plosive problem of open urban unemploy- the supply of educational opportunities, eco-
ment. Recent LDC research on rural-urban nomic analyses of social and private rates of
migration has demonstrated that the propen- return to investment in various levels of edu-
sity to migrate increases with increasing ed- cation remain a tool of considerable potential
ucational attainment and that the ranks of importance in assessing the wisdom of edu-
the urban unemployed are increasingly being cational investment decisions, both for local
swelled by the more educated.4 governments and donor agencies. This ana-
Proposition 3. Limited job opportunities lytical tool has been much maligned, some-
tend to be rationed by educational certifica- times justifiably so and other times as the re-
tion with the more educated replacing the less sult of a critic's misunderstanding of the
well educated, even though higher qualifica- legitimate, albeit limited, role which these
tions may not be needed to perform the task calculations can play.6 For the most part,
adequately. however, rate-of-return calculations have not
Given the chronic tendency of educational taken account of factors such as educational
systems of LDCs increasingly to overproduce displacement, rising unemployment among
in terms of available employment opportuni- the more educated, the distorted wage and
ties, the problem of rationing limited mod- price signals in the economy as a whole. We
ern-sector jobs assumes growing importance. believe that many manpower plans based on
Typically these jobs are allocated to those these calculations would be considerably
with higher levels of educational attainment modified, especially with respect to their pro-
regardless of whether that education is really jections of needed quantitative expansion
necessary for satisfactory job performance. among the highly educated, if the above fac-
The educationally less fortunate are rele- tors were taken into account. In particular,
gated to the ranks of the unemployed and un- the apparent discrepancy between rates of re-
deremployed on the fringes of the urban sec- turn to investment in education as compared
tor. This rationing device has the appealing to alternative investment in other sectors of
political merit of being apparently objective, the economy would not be nearly so wide if
relatively untainted by obvious favor, and the factors mentioned above were given ade-
patently dependent for its operation on many quate treatment.
piivate as well as public decisions. But its op-
eration does not relieve unemployment or im- 5For a more detailed discussion of the job rationing
prove the allocation of resources beyond in- by educational certification phenomenon, see E. O.
suring that those most over-educated are Edwards and M. P. Todaro, "Educational demand
and supply in the context of growing unemployment
indeed eventually employed. So the magni-
in less developed countries," World Development,
tude of the problem is left to grow and an ap- March/ April 1973.
parently fair rationing mechanism is unlikely 6For a further exposition of the role of cost-benefit
analysis in a political setting, see Edgar O. Edwards,
4See, for example, M. P. Todaro, "Education, mi- "Investment in education in developing nations: Con-
gration and fertility," paper presented at IBRD con- flict among social, economic, and political signals,"
ference on The Economics of Education: Alternative
paper presented at IBRD conference on The Econom-
Strategies for Investment, Washington, D.C., Octo- ics of Education: Alternative Strategies for Invest-
ber 1973 (mimeo). ment, Washington, D.C., October 1973 (mimeo).
 629 EDUCATION

Proposition 5. Public subsidies to educa- lie subsidies should (a) reduce the excess
tion, especially at the higher levels, contribute demand for education, (b) increase employ-
to the widening gap between "private" and ment opportunities and (c) provide develop-
"social" benefit /cost calculations and thus mental justification for a larger educational
lead to excessive levels of aggregate private system than would otherwise be the case.
demand for higher education; these subsidies Proposition 6. Constraints on upward mo-
also limit financially the ability of the public bility inthe labor market force many of those
sector both to supply educational opportuni- discriminated against to seek social and eco-
ties and to create jobs. nomic advancement through the educational
In many LDCs both the percentage and system, thus increasing the already heavy
absolute amount of educational costs borne burden carried by it.
privately tend to decrease with increasing Those who expect that without substantial
levels of educational attainment. This phe- education their employment and promotion
nomenon, incombination with the positive re- opportunities will be limited because of caste,
lationship between educational attainment race, creed, tribe, or the simple lack of over-
and expected income from employment, cre- valued exit credentials from the educational
ates an explosive private demand for higher system, may seek to circumvent these obsta-
levels of education. In order to satisfy this de- cles through further education. This implies,
mand, primary and secondary facilities must of course, that educational systems are on the
be expanded by some multiple of the demand whole characterized by fewer discriminatory
for higher education to take account of the practices than the labor market itself. In
large percentage of students who will not these circumstances, the existence of sub-
make it to the higher levels. The irony of the stantial social and political biases in hiring
situation is that the wider the gap between and promotion practices may contribute to
perceived "private" benefits and costs of sec- the overloading of the educational system. If,
ondary versus primary and tertiary versus on the other hand, such labor-market con-
secondary education and the more unprofit- straints can be reduced, some resources of the
able a given level of education becomes as a educational system will be freed for the im-
terminal point (e.g., primary education), the proved performance of its other essential
more demand for it increases as an interme- functions.
diate stage or precondition to the next level Proposition 7. Employment, in addition to
of education! This puts increased pressure on providing income, is also a major means for
the government in conjunction with donor disseminating work-related knowledge.
agencies to expand lower- level facilities in Employment is often regarded by those
order to meet the demand for higher-level seeking it in developing countries as a reward
education.7 for the successful completion of a continuous
Public subsidies not only stimulate the de- segment of the educational process. Educa-
mand for education; they also limit the abil- tion is regarded as a preparation for work;
ity of governments to create both educational work is seldom seen as a learning experience
and job opportunities. Reducing these subsi- in its own right or even as a useful basis for
dies would release government funds for job- further education. Unfortunately, this view is
creating investments, raising the economy's frequently reinforced by hiring practices of
capacity to absorb educated man-power in employers and incentive schemes supported
productive employment. Such an increase in by policy-makers, and it reflects a rather
the demand for educated man-power would common underevaluation of the learning
itself justify a more rapid growth in educa- which can be obtained through participation
tional opportunities. Thus a reduction in pub- in the labor market. Moreover, the learning
potential inherent in employment opportuni-
ties does not seem to be significantly realized
7See Edwards and Todaro, op. cit. and Blaug, op.
cit. in practice. For example, the wider sharing of
 630 HUMAN-RESOURCE DEVELOPMENT

available work among members of the labor learning experiences. If a slowdown in subsi-
force would spread the associated learning dized public higher education is off-set by ex-
experience over a larger number of benefi- pansion ofthe private sector, there is at least
ciaries; social and economic policies which the favorable incentive effect that the bene-
promoted variety in an individual's work ex- ficiaries must pay more of the cost of their
periences, whether with one or several em- education.
ployers, rather than narrow, routine special- 3. Work-sharing arrangements should be
ization would enrich his life and his potential encouraged.
In the absence of full employment, policies
for contributing to a nation's development;
and the use of information on an individual's should be explored which will encourage a
degree of success in work settings as a means wider sharing of available work as a means
for evaluating his potential to gain from fur- of dispersing income, spreading work-related
ther education would add an important di- learning experiences, and reducing the num-
mension to the criteria currently in common bers openly unemployed. It is not, however, a
use in determining who progresses in the ed- simple task to identify appropriate incentive
ucational system. mechanisms which will induce the private
sector to assume the additional managerial
SOME STRATEGIES FOR costs which might be required in order to re-
IMPROVING EDUCATION AND alize the social benefits of work sharing. This
EMPLOYMENT need not be a constraint on public-sector em-
ployment practices where social benefit cal-
In the light of the above propositions, the culations can enter directly into employment
following strategies for improving the rela-
tionship between education and employment decisions.8
4. Job-rationing by educational certifica-
are offered: tion must be modified.
1 . Minimizing imbalances, incentive distor- In order to break the vicious circle in
tions, and socio-political constraints will im- which overstated job specifications make
prove both education and employment. overeducation necessary for employment, po-
Policies which tend to remedy major eco- licies are needed which will induce or require
nomic imbalances (e.g., those between rural both public and private employers to seek re-
and urban areas), to correct distortions in in- alistic qualifications even though the task of
centives (e.g., in income and wage differen- job rationing may be made somewhat more
tials), and to alleviate social and political difficult as a result.9
constraints on upward mobility will have the 5. Subsidies for upper-level education
multiple beneficial effect of increasing job op- should be reduced.
portunities, modifying the accelerated rate of As a means of overcoming distortions in
rural-urban migration, and facilitating devel- the aggregate private demand for education
opment-related modifications of educational induced by excessive subsidies to education
systems. especially at higher levels, policies should be
2. Where politically feasible, educational promoted by which the beneficiary of educa-
budgets should grow more slowly and be tion (as opposed to his family or society as a
more oriented towards primary education. whole) would bear a larger and rising pro-
In the light of growing unemployment portion of his educational costs as he pro-
among the educated, educational budgets ceeds through the system — either directly,
should grow more slowly than in the past to
permit more funds to be used for the creation 8For some theoretical arguments in support of
of employment opportunities. Moreover, a work-sharing arrangements, see Edgar O. Edwards,
larger share of educational budgets should be "Work effort, investable surplus, and the inferiority of
allocated to the development of primary, as competition,"
1971. Southern Economic Journal, October
opposed to secondary and higher, education 9The ILO Ceylon Report, op. cit., makes extensive
as a basis for self-education and work-related recommendations along these lines.
 631 EDUCATION

through loan repayments, or by service in and educational discrimination in hiring and

rural areas.10 promotion in labor markets, (2) encourage
6. Inequities and discrimination in both greater reliance on job-related learning ex-
education and employment should be periences as a basis for occupational ad-
minimized. vancement and (3) stimulate the use of suc-
In order to reduce the burden on the edu- cessful work experience as a criterion for
cational system and to improve the perfor- advancement in the educational system.
mance of its essential functions, policies Insofar as possible, donor agencies should
should be pursued which (1) reduce social ensure that their policies of support for public
education reinforce strategies, such as those
listed above, which tend to make education
10See Edwards and Todaro, op. cit., for further dis-
cussions of the distributive implications of current appropriate, quantitatively and qualitatively,
methods of post-primary education in the context of for the employment opportunities likely to
rising levels of unemployment among the educated. emerge in the coming decade.

IX.C.5. Development as a
Generalized Process of Capital
Accumulation *

The contemporary interest in the economics tive in its implications and likely to appear
of education, and more broadly in the econ- before much more time has passed as a tran-
omies of all processes connected with the sient stage in the evolution towards a more
augmentation and application of knowledge, comprehensive formulation of economic de-
represents a confluence of interests derived velopment problems in terms of a broadly
from concerns with widely divergent prob- conceived concept of capital accumu-
lems. These problems include such matters as lation. . .
the economic value of education, the contri- Concentration on the role of human capital
bution of education to past economic devel- has already proceeded far enough to generate
opment inadvanced countries, and the role of the beginnings of a counter-revolution. The
education and expenditure on increased ed- general outlines of the counter-revolution are
ucation inthe planned development of under- indeed already apparent. On the one hand,
developed countries.1 the recent emphasis on human capital for-
The formulation of concern with the eco- mation ingrowth accountancy is based on the
nomics of education (in a broad sense) in recognition that conventional measures of la-
these particular terms, while appropriate to bour input fail to take account of improve-
the current state of economic research and ments in the quality of labour and aims pri-
thinking, is for this very reason both restric- marily at more accurate measurement of
labour inputs. Application of the same crite-
ria to inputs of capital suggests that the con-
tribution of capital may also have been
*From Harry G. Johnson, "Towards a Generalized
Capital Accumulation Approach to Economic Devel- grossly underestimated, as a result both of
opment," inOECD Study Group in the Economics of understatement of the flow of capital services
Education, The Residual Factor and Economic
into production by the conventional equation
Growth, Paris, 1964, pp. 219-25. Reprinted by of service flow with the depreciated value of
permission.
'Cf. T. W. Schultz, ed., "Reflections on Investment capital stock, and of failure to measure ac-
in Man," Journal of Political Economy, Supplement,
curately improvements in the performance
October 1962, pp. 1-8. characteristics ("quality") of capital equip-
 632 HUMAN-RESOURCE DEVELOPMENT

ment.2 On the other hand, the evidence on such a synthesis, in the form of a generalised
rates of return to educational investment in capital accumulation approach to economic
the United States does not suggest that there development, and to discuss some of its im-
has been serious general underinvestment in plications for social and economic policy.
education there, while both casual empirical The essential elements of a generalised
observation of under-developed countries and capital accumulation approach to economic
some detailed research on the relative returns development are already present in the liter-
to investments in education and material cap- ature of economics, and at least some appli-
ital in them3 suggest that at least in some cations ofthe approach (for example, the ex-
cases the proportion of resources devoted to planation of wage differentials) have been
human capital formation may be too high familiar to economists ever since economics
rather than too low.4 A rehabilitation of in- became established as a separate subject of
vestment in material capital as a potent study. The foundations of it were explicitly
source of economic growth may therefore be laid in Irving Fisher's classic work on capital
in prospect. What is more important, while and income, and carried forward by F. H.
the process of increasing economic knowl- Knight's work on the theory of capital; and
edge proceeds in phases of exaggerated con- the approach is exemplified, and its potency
centration on one or another aspect of a prob- demonstrated, in the recent research of T. W.
lem, both the effect and the intent are to Schultz, Gary Becker, and others on human
arrive at a unified and more powerful synthe- capital.5 The essence of it is to regard "capi-
sis of explanations of economic phenomena. tal" as including anything that yields a
The contemporary phase, in which the con- stream of income over time, and income as
cepts of human capital and of investment in the product of capital. From this point of
it figure as corrections of emphasis in a sys- view, as Fisher pointed out, all categories
tem of economic ideas dominated by material of income describe yields on various forms of
capital, is bound to merge into one in which capital, and can be expressed as rates of in-
human and nonhuman capital are treated as terest or return on the corresponding items of
alternative forms of capital in general. The capital. Alternatively, all forms of income-
desirability of achieving such a synthesis is yielding assets can be given an equivalent
not merely a matter of scientific economy and capital value by capitalising the income they
elegance, it is also a pre-requisite for rational yield at an appropriate rate of interest. By
discussion and formulation of policy for eco- extension, the growth of income that defines
nomic growth in both advanced and under- economic development is necessarily the re-
developed countries. The purpose of this
paper, accordingly, is to sketch the outlines of sult of the accumulation of capital, or of "in-
vestment"; but "investment" in this context
must be defined to include such diverse activ-
2Cf. Zvi Griliches, "The Sources of Measured Pro-
ities as adding to material capital, increasing
ductivity Growth: U.S. Agriculture, 1940-1960," the health, discipline, skill and education of
Journal of Political Economy, August 1960. the human population, moving labour into
3Cf. Arnold C. Harberger, Investment in Man Ver- more productive occupations and locations,
sus Investment in Machines: The Case of India, a
paper prepared for the Conference on Education and and applying existing knowledge or discov-
Economic Development, University of Chicago, April ering and applying new knowledge to in-
4-6, 1963. Harberger finds the rate of return on real crease the efficiency of productive processes.
investment in India to be substantially higher than the All such activities involve incurring costs, in
rate of return on investment in education.
the form of use of current resources, and in-
*This proposition becomes almost a truism if the vestment inthem is socially worth while if
concept of investment in human capital formation is
extended to include expenditures on improved health, the rate of return over cost exceeds the gen-
whose effects on the rate of population increase con- eral rate of interest, or the capital value of
stitute one of the major economic problems of under-
developed countries. 5See T. W. Schultz, op. cit.
 633 EDUCATION

the additional income they yield exceeds the should be regarded as no different from ma-
cost of obtaining it. From the somewhat dif- chines; on the contrary, refusal to recognise
ferent perspective of planning economic de- the investment character of a problem be-
velopment, efficient development involves al- cause people are involved may result in peo-
location ofinvestment resources according to ple receiving worse treatment than machines.
priorities set by the relative rates of return on One might, indeed, hazard the generalisation
alternative investments. that democratic free-enterprise economies
The conception of economic growth as a tend to make wasteful use of their human re-
process of accumulating capital, in all the sources, precisely because people are not suf-
manifold forms that the broad Fisherian con- ficiently regarded as socially productive
cept of capital allows, is a potent simplifica- assets.
tion of the analytical problem of growth, and Conception of economic growth as a gen-
one which facilitates the discussion of prob- eralised process of capital accumulation pro-
lems of growth policy by emphasising the rel- vides a unifying principle for the statistical
ative returns from alternative investments of explanation of past growth and the formula-
currently available resources. The Fisherian tion of policy for future growth or plans for
concept of capital, however, and the ap- economic development. It does not, how-
proach to the analysis of production and dis- ever— and cannot be expected to — dispose of
tribution problems associated with it, are not any real problems, though it does clarify un-
as yet characteristic of the work and philo- derstanding ofthem. Instead, it transforms
sophical approach of the majority of econo- these problems into problems of the special
mists, and to some the implications of the ap- characteristics of particular types of capital,
proach for policy with respect to human or of the specification of efficient investment
beings appear to be positively repugnant.
programmes.
Must economists instead employ a narrower From the point of view of economically rel-
concept of capital that identifies capital with evant differentiations, items of capital can be
material capital goods and equipment used in classified in a variety of ways. One funda-
the production process, and distinguishes it mental distinction to be drawn relates to the
sharply from labour? . . . nature of the yield or contribution to eco-
As already mentioned, the limitations of nomic welfare — the distinction between con-
accumulation of material capital as an expla- sumption capital, which yields a flow of ser-
nation of a prescription for growth have vices enjoyed directly and therefore
prompted the contemporary interest in contributing to utility, and production capi-
human capital formation, and suggest a ge- tal, which yields a flow of goods the con-
neralisation ofthe concept of capital accu- sumption ofwhich yields utility. The returns
mulation toinclude investment in all types of from production capital are directly observ-
capital formation. An important obstacle to able, and therefore more amenable to mea-
such a generalisation is that the treatment of surement than the returns on consumption
human beings as a form of capital, even if capital.
only conceptually, seems offensive to some Another fundamental distinction relates to
economists as being contrary to democratic the form in which capital is embodied — here
political philosophy. This reaction, however, it seems necessary not only to distinguish
involves a confusion of analytical approach capital embodied in human beings from cap-
and normative recommendations unfortu- ital embodied in non-human material forms,
nately only too common in discussions of eco- but also to distinguish between capital em-
nomic problems with policy connotations. To bodied inboth human and non-human phys-
recognise that important areas of socio-eco- ical forms and capital embodied in neither,
nomic policy involve decisions analytically the latter category comprising both the state
identical with decisions about investing in of the arts (the intellectual production capital
machines is not at all to imply that people of society) and the state of culture (the intel-
 634 HUMAN-RESOURCE DEVELOPMENT

lectual consumption capital of society). The richer countries.6 Social capital investment
significance of this distinction is closely re- involves a similar separation of costs of in-
lated to a third distinction — one which is par- vestment from benefits, and a similar mixture
ticularly relevant to policy problems — be- of equity and efficiency considerations. In-
tween types of capital according to whether vestment inknowledge raises the thorniest of
the returns to investment in capital accumu- all problems, since the zero marginal cost of
lation accrue to the investor or to others. knowledge to additional users implies that no
Here it seems necessary to distinguish: (a) system of recouping the cost of investment in
capital goods which render specific services knowledge-creation by charging for its use
to production or consumption by the owner; can be economically efficient. . . .
(b) human capital, the distinguishing char- The distinctions discussed above do not in-
acteristic ofwhich is that, both inherently clude adistinction between natural resources
and by legal tradition, control over the use of (natural capital) and man-made capital. For
the capital is vested in the individual em- most economic purposes, such a distinction is
bodying the capital, regardless of the source unnecessary — natural resources, like capital
of finance of the investment in it; (c) social goods, can be appropriated, transferred, and
capital or collective capital, the distinguish- invested in. Natural resources do, however,
ing characteristic of which is that for reasons raise two sorts of special problems. First
of inherent necessity or administrative con- property rights in some range of natural re-
venience its services to production or con- sources are typically vested in society or the
sumption are not charged to individual users state; this poses the problem of ensuring effi-
but are paid for by taxation of the commu- cient exploitation of these resources through
nity at large; (d) intellectual capital or appropriate accounting and charging for the
knowledge, the distinguishing characteristic use of the state's natural capital, a problem
of which is that, once created it is a free good, particularly important at the time when re-
in the sense that use of it by one individual sources are first brought into use. Secondly,
does not diminish its availability to others. some kinds of natural resources, which are
All forms of capital other than capital likely to be of particular importance to de-
goods rendering specific services to produc- veloping countries, are nonrenewable, and
tion or consumption raise serious problems pose the problems of efficient depletion and
for economic analysis measurement and pol- exhaustion — of efficient capital decumula-
icy formation. The fusion of human capital tion, rather than accumulation. The prob-
with the personality of its owner raises
among other things the problem of how far
expenditure on the creation of human capital 6Brinley Thomas has emphasised the economic ab-
surdity of the contemporary migration pattern be-
should be accounted as investment, and how tween advanced and underdeveloped countries, in
far it should be classed as consumption; while which the advanced countries cream off the profes-
the vesting of control over the use of capital sional talent of the underdeveloped countries by im-
migration and attempt to replace it by their own ex-
in the individual invested in, given the imper- perts supplied at great expense as part of development
fection of markets for personal credit, poses
aid. See Brinley Thomas, "International Factor
the problem of how far education should be Movements and Unequal Rate of Growth," The
provided at public expense. The divergence of Manchester School of Economic and Social Studies,
private and social costs and benefits inherent January 1961. The ease of migration of educated peo-
in free or subsidised education raises some ple from underdeveloped countries, especially those in
which English is the language of instruction, to ad-
particularly difficult problems in conjunction vanced countries is a serious limitation on the poten-
with the fact that educated people are espe- tialities ofachieving economic development by edu-
cational investment and suggests the social
cially mobile interregionally and internation-
desirability of devising means of obliging either the
ally, so that resources devoted to education in emigrants themselves or the countries receiving them
poor countries may run substantially to waste to repay the social capital invested in them to their
in unilateral transfers of human capital to countries of origin.
 635 EDUCATION

lems of achieving economic development vestment ofthe same capital in agricultural

through the exploitation of depleting natural improvement might yield far higher returns.
resources become particularly acute and po- On the other hand, there is a strong possibil-
litically highly charged when such exploita- ity, exemplified by the successful develop-
tion isdependent on the participation of for- ment of exports of some technologically
eign capital and enterprise. fairly advanced products from otherwise
Conception of economic development as a underdeveloped countries, that the greatest
generalised process of capital accumulation, comparative advantage for such countries
in conjunction with recognition of economi- lies in skilled-labour-intensive products, for
cally significant differences between various the reason that a generally low-wage level
types of capital, has important implications makes the cost of investment in human cap-
for the efficient programming of investment ital low (especially forgone earnings and the
of economic development. These implications cost of instruction and educational struc-
centre on the relationships of complementar- tures) bycomparison with comparable costs
ity and substitutability in both production in advanced countries. In addition, such
and consumption that may exist between countries may be able to catch up with the
types of capital provided by different invest- advanced countries far more rapidly in the
ment processes, and the consequent desira- accumulation of knowledge than in the ac-
bility ofaiming at both balanced investment cumulation ofmaterial capital.
in the production of complementary types of Apart from its implications for planning
capital and the selection of the most efficient for economic growth, a generalised capital
combinations of types of capital in the light accumulation approach to economic devel-
of the relative costs of different kinds of in- opment points to the potential fruitfulness of
vestment. The complementarity between research into and analysis of the efficiency of
modern equipment and technology, a skilled a wide range of processes and policies that in-
labour force, and social overhead capital in volve the allocation of capital but are not usu-
the transportation and distribution systems is ally thought of as concerned with investment.
by now sufficiently recognised for develop- It has, for example, been amply demon-
ment planning to aim at producing integrated strated byempirical research that rates of re-
investment programmes comprising invest- turn on investment in education vary widely
ment in education and vocational training between different levels of the education sys-
(manpower programmes) as well as invest- tem; and there is good reason for doubting
ment in industrial and social overhead. For that existing educational systems are very ef-
such comprehensive development investment ficient when considered as an industry pro-
programmes to maximise the contribution of ducing extremely long-lived capital assets.
investment to economic growth, however, The field of public health and medical care,
recognition of complementarity must be al- viewed as an industry concerned with the re-
lied with recognition of substitutability and pair aiid maintenance of human capital, also
analysis of rates of return on the total invest- offers scope for economic analyses of rates of
ment of capital in alternative programmes in- return on alternative investments. Institu-
volving investment in capital goods, human tional arrangements for supporting and re-
capital, social capital and the acquisiton of warding fundamental and applied research,
new knowledge. considered as an industry producing intellec-
Much of the literature on economic devel- tual capital, provide an even greater chal-
opment assumes far too easily that low-wage lenge to economists. Within the traditional
labour is necessarily cheap industrial labour, scope of economics, labour mobility, unem-
ignoring the magnitude of the investments in ployment policy, and policy respecting the lo-
human and social capital that may have to be cation of industry all demand the application
made to convert rural workers into skilled in- of capital theory. Perhaps the most important
dustrial labour, and the possibility that in- area requiring rationalisation in terms of a
 636 HUMAN-RESOURCE DEVELOPMENT

broadened concept of capital accumulation, come and profits taxation the costs of invest-
however, is the theory and practice of public ments inhuman capital customarily provided
finance. Not only do income tax systems typ- free or at a subsidised price to the people in-
ically make a very poor adjustment for the vested increates disincentives to the efficient
capital investment element in personal in- use and accumulation of capital of all kinds.
come, but the necessity of recouping by in-
CHAPTER X

Project Appraisal

Having considered in earlier chapters the overall contribution of capital to development, and
the various ways of financing a higher rate of investment, we now turn to the remaining prob-
lem of how to determine the allocation of investment resources. When investment resources
are as scarce as they are in a poor country, the rational allocation of capital is of the greatest
importance.
How can a developing country be assured that the investment resources used in a given
project would not have any better alternative use in terms of the country's objectives? To
answer this question is to engage in the practice of project appraisal.
National governments guide the pattern of investment either by direct investment in the
public sector or by controls on private investment. To provide rational guidance, the govern-
ment must formulate and evaluate investment projects in such a way as to be able to compare
and evaluate alternative projects in terms of their contribution to the objectives of the nation.
So too must an international lending agency such as the World Bank engage in project ap-
praisal. Even a private foreign investor may have to negotiate with a host government on the
basis of the real worth of the foreign investment to the economy as a whole.
A private firm typically invests according to commercial profitability. But national profit-
ability or social profitability may differ from commercial profitability. National benefits and
costs may differ from benefits accruing to a firm and costs incurred by the firm. The tech-
niques of project appraisal have been designed to evaluate projects when the social benefits
and costs differ from the private benefits and costs. In assessing social benefits, one may con-
sider issues of income distribution, public goods, and merit wants.
To provide an introduction to project appraisal, this chapter highlights some central issues
that must be considered when a project analyst attempts to deal systematically with the di-
vergences between private and social values. The selections concentrate on the theory of pro-
ject appraisal, not on actual computation.

637
 638 PROJECT APPRAISAL

The Note at the outset (section X.A) presents the rationale underlying project appraisal,
emphasizing that it is applied welfare economics in a world of the second best. After exam-
ining measures of private profitability (X.B.I and X.B.2), subsequent selections introduce
other considerations that are necessary for the calculation of social benefits and costs. Section
X.C explains shadow pricing, with particular attention to the key shadow prices of foreign
exchange, wages, and opportunity costs of capital. In addition, other important objectives, for
example, the designation of beneficiaries of the project and the placing of a premium on in-
vestment, are discussed in X.D.
 X.A. GENERAL RATIONALE OF
PROJECT APPRAISAL-NOTE

Project appraisal does not rest on a set of economist must be concerned with the exis-
techniques that can be applied mechanically, tence of monopolies, taxes and subsidies, ex-
but is an approach that must be interpreted ternalities, and price distortions. When mar-
analytically. Interpretation becomes neces- ket prices are not honest prices, reflecting
sary because in reality investment resources real value, or prices do not exist for public
are not allocated in a first-best world of per- goods, or do not value the social benefit of
fect competition with no externalities. In- merit wants, then the market prices will no
stances ofmarket failure and distortions are longer equate and equal the MSC and the
especially widespread in a developing econ- MSV of the commodities.
omy. The distortions that prevail in a devel- To remove price distortions, neutral fiscal
oping economy are commonly in foreign devices of lump-sum taxes and subsidy would
trade prices, factor markets, the nonoptimal constitute the best remedial policies, restor-
income distribution, and in the use of non- ing the equivalence of the MSC and the
optimal taxes and subsidies. In correcting for MSV with market price. But this is not fea-
these distortions, the practice of project ap- sible and is rarely undertaken. The problem
praisal isan exercise in the theory of the sec- then is to correct for the price distortions in
ond-best. the practice of project appraisal through
A "first-best" economy would fulfill the shadow pricing, using a cost-benefit analysis
marginal conditions of perfectly competitive based on imputed values — that is, on the real
equilibrium in all product and factor mar- economic values of costs and benefits — in-
kets, with no uncertainty, no externalities, stead of relying on the market price data
and a given income distribution. The per- used in calculating private commercial prof-
fectly competitive solution to resource allo- itability. The divergence between the MSC
cation would then be equivalent to a Pareto- and the MSV has to be taken as a constraint,
efficient allocation, reached by voluntary ex- and the shadow prices corresponding to this
change in a market price system. Market constrained or "second-best" welfare opti-
prices of goods and factors would equate and mum will need to be computed. Shadow pric-
equal the marginal social cost (MSC) of pro- ing involves efficiency prices, and the calcu-
ducing and the marginal social value (MSV) lation of a project's rate of return proceeds
of using the relevant goods and factors. For a from the internal rate of financial return to
marginal investment project, the values of the internal rate of economic return. Simi-
the outputs and inputs at market prices larly,onward calculations
would then provide the correct values to be ceed from ofthe
the discounted
project's NPV
valuepro-
of
used in calculating the net present value private net benefits (private receipts minus
(NPV) of the project. Economic costs and private costs) to the value of social net bene-
benefits would not differ from financial costs fits (social benefit minus social cost), dis-
and benefits, and calculations of private com- counted at a social rate of discount. If consid-
mercial profitability would then be equiva- eration isalso given to who the beneficiaries
lent to social economic profitability. of the project are — to valuation of the resul-
Once conditions in a developing economy tant income distribution — then "social
diverge, however, from those of the ideal prices" that weight the income distribution
first-best world, we enter a second-best are used to calculate the internal social rate
world — a world in which the development of return. The social rate of return may also
practitioner must live. The development reflect the country's valuation of the distri-

639
 640 PROJECT APPRAISAL

bution of income between consumption and only to the extent that the use of labor in this pro-
investment that results from the project.1 ject implies some sacrifice elsewhere in the econ-
A World Bank analyst has summarized omy with respect to output and other objectives of
the context of project analysis as follows: the country. Conversely, if the project has an eco-
nomic cost in this sense that does not involve a cor-
In essence, project analysis assesses the benefits responding money outflow from the project en-
and costs of a project and reduces them to a com- tity— for example, because of environmental
mon denominator. If benefits exceed costs — both effects or subsidies — this cost is not a financial
expressed in terms of this common denominator — cost. The two types of cost need not coincide. Eco-
the project is acceptable: if not, the project should nomic costs may be larger or smaller than financial
be rejected. As such, project analysis may appear costs. Similar comments apply to economic and fi-
divorced from both the fundamental objectives of nancial benefits. Economic costs and benefits are
the economy and the possible alternative uses of measured by "shadow prices," which may well dif-
resources in other projects. The definition of ben- fer from the market prices appropriate for finan-
efits and costs, however, is such that these factors cial costs and benefits.
play an integral part in the decision to accept or Shadow prices are determined by the interac-
reject. Benefits are defined relative to their effect tion of the fundamental policy objectives and the
on the fundamental objectives; costs are defined basic resource availabilities. If a particular re-
relative to their opportunity cost, which is the ben- source is very scarce (that is, many alternative
efit forgone by not using these resources in the best uses are competing for that resource), then its
of the available alternative investments that can- shadow price, or opportunity cost (the foregone
not be undertaken if the resources are used in the benefit in the best available alternative that must
project. The forgone benefits are in turn defined be sacrificed), will tend to be high. If the supply of
relative to their effect on the fundamental objec- this resource were greater, however, the demand
tives. Bydefining costs and benefits in this fashion arising from the next best uses could be satisfied
we try to ensure that acceptance of a project im- in decreasing order of importance, and its oppor-
plies that no alternative use of the resources con- tunity cost (or shadow price) would fall. Market
sumed bythis project would secure a better result prices will often reflect this scarcity correctly, but
from the perspective of the country's objectives. there is good reason to believe that in less devel-
Economic analysis of projects is similar in form oped countries imperfect markets may cause a di-
to financial analysis in that both assess the profit vergence between market and shadow prices. Such
of an investment. The concept of financial profit, divergences are thought to be particularly severe
however, is not the same as the social profit of eco- in the markets for three important resources:
nomic analysis. The financial analysis of a project labor, capital, and foreign exchange
identifies the money profit accruing to the project- Resource availabilities, however, need not be
operating entity, whereas social profit measures the only constraints operating in the economy: po-
the effect of the project on the fundamental objec- litical and social constraints may be equally bind-
tives of the whole economy. These different con- ing. The alternatives open to the government in
cepts of profit are reflected in the different items pursuing its development objectives can be limited
considered to be costs and benefits and in their val- by these noneconomic constraints to a narrower
uation. Thus, a money payment made by the pro- range than that implied by the basic resource
ject-operating entity for, say, wages is by definition availabilities. If the tools of general economic pol-
a financial cost. But it will be an economic cost
icy— that is, fiscal and monetary policy — cannot
break these constraints, project analysis should
take account of them by means of appropriate ad-
'Sometimes what we have termed the internal eco- justments inshadow prices. For example, if the
nomic rate of return is called the internal social rate
of return, and some selections that follow refer to government is unable to secure a desired redistri-
bution of income through taxation, it can use the
what -we have called efficiency prices as social prices. allocation of investment resources as an alterna-
But we wish to distinguish among the internal finan- tive method of redistributing income. If in project
cial rate of return at market prices, the internal eco-
nomic rate of return at efficiency prices, and assuming analysis higher values were to be attached to in-
all units of income are equally valuable, and the in- creases in income accruing to the poorer groups
ternal social rate of return with social prices that in- within society, investment would be biased in favor
clude the distributional aspects between consumption of these groups. In other words, all available policy
and investment and between rich and poor. tools should be working jointly toward the same
 641 GENERAL RATIONALE OF PROJECT APPRAISAL— NOTE

goals. If one instrument is inoperative or blunted, the following issues that will be considered in
other instruments may be used to achieve the same the remainder of this chapter: (1) specifica-
end. tion of costs and benefits; (2) valuation of
Project analysis is designed to permit project- costs and benefits; (3) choice and formulation
by-project decision-making on the appropriate of constraints; (4) treatment of risk and un-
choices between competing uses of resources, with
costs and benefits being defined and valued, in certainty; (5)choice of the rate of interest for
discounting future costs and benefits; and (6)
principle, so as to measure their impact on the de-
velopment objectives of the country. In many choice of a decision rule for accepting or re-
cases, however, a more direct link is necessary with jecting projects.
the sector and economy as a whole: for example, The concept of a cost or a benefit has
the merit of a project characterized by economies meaning only in relation to objectives of the
of scale cannot be judged without making an esti- country's development program. Various ob-
mate of the demand for its output, and this in turn jectives might be relevant: higher consump-
requires placing the project in its sectoral and tion or real income; better distribution of
country context.
consumption or income; employment; na-
Furthermore, in practice, many shadow prices
tional prestige or independence; changes in
(for land and natural resources, for example) are
skills, attitudes, and institutions. Because
difficult to determine independent of the project
appraisal process, because they depend on the al- project appraisal in the context of a develop-
ternative projects that have been rejected. This is ment plan is an exercise in constrained max-
the basic reason why a systematic scrutiny of plau- imization, the choice of the objective function
sible alternatives is at the heart of the appraisal and the selection of the relevant constraints
process: it is not sufficient in practice to select an are crucial. The shadow price of a given re-
acceptable project whose benefits appear to exceed source isthe increase in the maximum value
costs; it is necessary to search for alternatives with of the objective when ceteris paribus the
a larger surplus of benefits over identified costs. If economy is endowed with a unit more of the
such projects are found, it means that the oppor-
tunity cost of using, say, land in the project origi- resource. As such, it is the "marginal prod-
nally considered acceptable has been underesti- uct" of the resource, the "product" being re-
flected inthe value of the objective function.
mated or wholly neglected.2
The choice of shadow prices will therefore
If project analysis is to make its maximum
depend on the objective function and the na-
contribution to a better allocation of re- ture of the constraints that the economy
sources, itis clear that social cost-benefit faces.
analysis must underlie each stage of project In practice, the shadow prices used can
preparation — from project identification to
feasibility studies of technical possibility and only be "approximately right," given the
complexity of objectives and constraints in
economic viability, to final calculations of
any actual economy. Although the perfec-
present net social value. tionist may be left unsatisfied, the practice of
Social cost-benefit analysis in the context project appraisal is an improvement over sim-
of the development process therefore raises
ple reliance on calculations of private com-
mercial profitability. The need is clear, prog-
ress has been made, and a broad consensus on
2Lyn Squire and Herman G. van der Tak, Eco-
nomic Analysis of Projects, Baltimore and London, proper techniques of project appraisal has
(1975), pp. 15-17. emerged.

Comment
The leading sources on project appraisal techniques are UNIDO, Guidelines for Project
Evaluation (1972), reinterpreted by John R. Hansen, Guide to Practical Project Appraisal:
Social Benefit-Cost Analysis in Developing Countries (1978); and I. M. D. Little and J. A.
Mirrlees, Project Appraisal and Planning for Developing Countries (1974). Also significant
for a review of the basic issues raised by the various methods of project evaluation are Deepak
 642 PROJECT APPRAISAL

Lai, Methods of Project Analysis: A Review (1974); A. Harberger, Project Evaluation

(1976); E. V. K. Fitzgerald, Public Investment Planning for Developing Countries (1978). For
an extension to distributional issues, see L. Squire and H. G. van der Tak, Economic Analysis
of Projects (1976).
More empirical studies include the "Symposium on the Little-Mirrlees Manual," Bulletin
Oxford Institute of Economics and Statistics (February 1972); Michael Roemer and Joseph
J. Stern, The Appraisal of Development Projects (1975); I. M. D. Little and M. FG. Scott,
Using Shadow Prices (1976); M. FG. Scott et al., Project Appraisal in Practice (1976). A
useful volume focusing on benefit-cost techniques for agricultural investment is J. Price Git-
tinger, Economic Analysis of Agricultural Projects, 2nd ed. (1982).
For general background readings on benefit-cost analysis, see R. Sugen and A. Williams,
The Principles of Practical Cost-Benefit Analysis (1978); G. Irvin, Modern Cost-Benefit
Methods (1978); R. Layard (ed.), Cost-Benefit Analysis (1972); E. J. Mishan, Elements of
Cost-Benefit Analysis (1976); A. R. Prest and R. Turvey, "Cost-Benefit Analysis: A Survey,"
Economic Journal (December 1975), which includes an extensive bibliography.
It is indicative of how recent the application of cost-benefit analysis is to projects in devel-
oping countries that the first published cost-benefit analysis of any investment in an LDC was
in 1972—1. M. D. Little and D. G. Tipping, A Social Cost Benefit Analysis of the Kulai Oil
Palm Estate (OECD, 1972).
 X.B. BENEFITS AND COSTS
X.B.1. Plans and Projects*

A sound development plan requires a great there are limits to what economic analysis
deal of knowledge about existing and poten- can achieve. No matter the sector of activity,
tial projects. This is obvious enough for a costs are relatively easy to estimate on a com-
short-term operational plan (3-5 years) parable basis. But this is not always true of
which should, among other things, contain benefits. For instance, although there is by
firm and realizable plans for Government ex- now a considerable body of work concerned
penditure indifferent sectors. But it is just as with estimating the benefits of education and
true for a perspective plan, which covers a pe- medical expenditure, one cannot be proud of,
riod of, say, 10-15 years. or place much confidence in, the results as
Such a perspective plan will lay down tar- yet. But no such strong reservations need be
get rates for gross national production, con- made in the important spheres of economic
sumption, and also for investment and its fi- infrastructure, industry proper, and agricul-
nancing by both domestic and foreign ture. In these spheres, although it is of course
savings. For this to be done, it is clear that a true that formal estimates of cost and benefit
relationship has to be assumed between the always go wrong, nevertheless it would be
level of investment during the plan period highly obscurantist to suggest that one
(and also for several years prior to the plan should not try to peer into the future at all.
period) and the rate of growth of domestic And, if one is going to peer into the future, it
product which depends more or less closely is important to make sure that the manner in
on that level of investment. which it is done does not lead to biases as be-
This connection between investment and tween different sectors and different proj-
the rate of growth cannot be properly esti- ects. .. .
mated without a great deal of knowledge I have laboured the point that realistic
about actual and potential investment proj- plans cannot be formulated in the absence of
ects. .. . a great deal of project planning, and without
In principle, there are an infinite number proper economic appraisal of projects. But it
of feasible plans, only one of which is the best is also true that a good economic appraisal of
of all. One can never hope to arrive at this projects often cannot be made without a plan.
optimum plan. But unless one strives contin- Project appraisal obviously requires an es-
ually to direct one's investment to those sec- timate of the demand for the product. But
tors where it would yield the most benefits to how can one estimate the domestic product
the economy, and within sectors to projects unless one has some idea of how the economy
which yield most, one will certainly end up will develop? And how the economy develops
with a plan which is very far short of what in turn depends at least partly on the long-
could be achieved. Thus, if the division of in- range plans and policies of the Govern-
vestment between different sectors of the ment. .. .
economy is to be rational, it is essential that In presenting the twin propositions that
the costs and benefits of many different proj- "plans require projects" and "projects re-
ects in each sector should be assessed on a
quire plans," it may seem that an insoluble
comparable basis. Here, one must admit that chicken and egg dilemma has been posed. If
good plans cannot be formulated without a
♦From Ian M. D. Little, "Project Analysis in Re- proper economic appraisal of projects, and if
lation to Planning in a Mixed Economy," in OECD the real value of projects cannot be properls
Development Center, Development Problems, Sum-
mary of Papers at the Colombo Seminar, Paris, 1967, ascertained, except within the framework of
pp. 47, 49-52, 54-65. Reprinted by permission. a plan, where does one start? But the chicken

643
 644 PROJECT APPRAISAL

and egg analogy is false, as one is never to- only actual cash flows, consists of a predic-
tally devoid of knowledge. Inadequate plans tion of
are first formulated using inadequate meth- 1 . all receipts from the sale of outputs of the
ods of project appraisal. These in turn should
permit improvements in project analysis and project for each year of the life of the pro-
ject, these including the sale of any build-
appraisals, and so on. Macro-economic plan- ings and equipment remaining at the end
ning can be gradually improved in the light of the life of the project, and
of improvements in micro-economic plan- 2. all expenditures on goods and services ac-
ning, and vice versa. By such iteration and re- cording to the year in which they are
iteration, one gradually tries to come near to
made, from the date of the first expendi-
optimum planning.
tures until the end of the life of the pro-
ject. These expenditures include capital
expenditures, whether for initial equip-
PROFITABILITY ANALYSIS ment or for replacement, as well as all
current costs.
In principle, one can approach an optimum
investment plan only if all new projects are Current expenditures may be subtracted
compared with each other with a view to from current receipts. One then has two
seeing that none are rejected which are su- streams, the first consisting of net current re-
perior to any of those selected. In principle, ceipts (gross profits), and the second consist-
of course, this implies that public sector proj- ing of capital expenditures, whether for ini-
ects are compared with those in the private tial equipment or replacement. . . .
sector, both being evaluated in the same So far, it has been assumed that a single
manner. But in this section, I shall neglect well-defined project is up for consideration.
the special problems which arise from trying But, in principle, no such finalized version of
to compare public and private projects, and I the project should come forward without
shall speak as f/all projects were in the pub- much consideration having been given to al-
lic sector. ternative ways of doing more or less the same
In this section also, I shall concern myself thing. First, there is always the question of
only with the analysis of gross profits, that is, whether it would not have been better to de-
profits before depreciation is subtracted. Or sign the project on a larger or a smaller scale.
to use another terminology, with pre-tax cash Secondly, given the scale, there are usually
flows. Of course, decisions in the public sec- numerous alternative techniques which
tor should not necessarily be based on prof- might have been employed. Each of these
itability, since there are many reasons for many alternatives arising out of possible dif-
thinking that profits are not always a good ferences ofscale and techniques is in princi-
measure of the net benefit to society. Never- ple another project, which has been rejected.
theless, profitability analysis and social cost- Each of these will have had technical, man-
benefit analysis have much in common, and agerial, and economic aspects. Were these al-
it is therefore convenient, first, to set out the ternatives properly assessed from these dif-
proper principles for the former, and then to ferent angles before rejection? Naturally,
modify it in such a way as to turn it into a one cannot fully evaluate a myriad of differ-
social cost-benefit analysis. This helps to ent possibilities before one of them is put for-
bring .out the differences between the two, dif- ward for final consideration. Perhaps the
ferences which are important in that the pri- most one can say is that those responsible for
vate sector will make its own decisions on the putting forward the final project, should be as
basis of profits alone (albeit net of tax). aware as those who exercise the final choice
A. The Basic Predictions. The basic figures of the need for proper appraisal and of the
required for economic appraisal, whether or methodology which the final choosers oper-
not the appraisal limits itself to considering ate. But, also, where important and feasible
 645 BENEFITS AND COSTS

■ ■ i I I 10I I I I I 15I i I I
25
Year
20
FIGURE 1.

variants present themselves, and the eco- A closely allied, but nevertheless different,
nomic choice is not obvious, the initiating de- procedure is to experiment until one has
partments should be encouraged to consult found that rate of discount which makes the
with the central project evaluators. present value r-25
equal to zero.
v
B. The Profitability Criterion. I can now In symbols one solves the equation
turn to describing the kind of analysis which
the evaluators should conduct upon the basic
figures already described. A diagram may be ys? o + 'yY =o
helpful here — see Figure 1 .
• Above the line is measured the current cash for y. The solution (y) is called the "Internal
Rate of Return." In a riskless situation, it
flow (gross profits before depreciation). The pays to invest if the internal rate of return ex-
exception is the last year, where the positive ceeds the rate of interest (if y > r).
amount includes the disposal value of the The present value method is technically
terminal equipment. superior, since the internal rate of return can
• Below the line we have investment expen- give an incorrect result in special circum-
ditures including replacement. stances.shall
I not go into this. The only, but
• The sum of the positive and negative items not unimportant, advantage of the internal
is the cash flow.
rate of return (given that it does give the cor-
• Twenty-five years are included in the dia- rect result) is that it is more familiar to ad-
gram. This is arbitrary, but unless the dis- ministrators and businessmen. It is in fact the
count rate (see below) is extremely low, proper way to calculate what is familiarly
what happens after 25 years is very unlikely
to make a significant difference. and loosely known as the "yield."
Either of these methods is known as "dis-
Now suppose that the project can borrow count of cash flow analysis." The point of
any sums of money required at a fixed rate of using cash flow, rather than profits, is that re-
interest. Using this market rate of interest, placement expenditures are explicitly ac-
the next step is to discount all the net current counted for, and the usual rather arbitrary
receipts, and the investment xt expenditures methods of allowing for depreciation of cap-
back to the present, and add up.1 This result ital are thereby avoided. The annual figures
is known as the Present Value. In symbols, are gross of depreciation, but the final result
r-25 is a net present value or net yield. The logic
E of the method is impeccable. It is now used
ST 0 + rr by the most sophisticated firms in making
where X, is the cash flow in year t, and r is their investment decisions, although the great
the rate of interest. majority still use much more rudimentary
methods.
Ignoring risk it is logical to make any in- I have so far dealt with riskless cases only.
vestment which results in a positive present
value. There is no one correct method of allowing
for risks. Two methods of approaching the
'Discounting simply means working compound in- problem are discussed briefly below.
terest backwards. For instance, £100 cumulated for
two years at 10% interest becomes £121. £121 dis-
First, one can use a higher rate of interest
counted at 10% for two years is £100. than the market rate for discounting pur-
 646 PROJECT APPRAISAL

poses, or require that the internal rate of re- mentioned, if only because private entrepre-
turn substantially exceeds the market rate neurs certainly make some allowance, how-
before embarking on an investment. The ad- ever informal, while the subject is often not
vantage of this is that the reduction of ex- even mentioned where public sector projects
pected net receipts is greater the further in are concerned.
the future they are (and risks presumably in-
crease with time). But it is rather mechani-
cal, and leaves open the choice of how great
SOCIAL COST-BENEFIT ANALYSIS
the risk discount should be, and also the ques-
tion of whether and how this risk discount Three modifications may be made to the
should be varied between different classes of gross profitability discussed above in order to
projects. While uniformity has advantages in turn it into a social cost-benefit analysis.
preventing unconscious and irrational pref- They are as follows:
erences from biasing the result, there is the 1. Other prices than those actually reign-
corresponding disadvantage that risks do dif- ing may be used to value the inputs and out-
fer between projects, and that they can to puts of the project.
some extent be broken down for separate 2. An allowance may be made for benefits
consideration instead of being lumped to- or costs which arise elsewhere in the economy
gether inan overall risk allowance. as a result of the operation or construction of
The second method is to try to make a high the project.
and a low estimate for each expenditure and 3. The annual benefits and costs may be dif-
receipt, and so produce a high and a low es- ferently treated in a social cost-benefit anal-
timate of the present value, or the internal ysis, as compared to a profitability analysis,
rate of return. The final step consists of some in order to arrive at an investment criterion.
subjective evaluation of the result — and Thus neither "present value" nor "the inter-
choice!
nal rate of return" may be appropriate. It is
What I have said about risk is, of course, in fact necessary to introduce a new concept,
quite inadequate. But the subject had to be
the "benefit-cost ratio."

X.B.2. Measures of
Profitability*
P,

The profits of an enterprise equal the differ- with a positive figure standing for net profits
ence between its earnings and its costs. For a and a negative figure for a net loss, or the net
project the stream of future profits and losses
may be calculated period by period. The expenditure, the "present value" of the
stream is simply the discounted sum of this
main complications in the concept of profit- stream, the discounting being done at the ap-
ability arise from the necessity to convert this propriate rate of interest, I. Representing the
stream of profits and losses into some simple present value of the project at interest rate i
measure expressed as a number, e.g., the as V(i), we get:
"rate of return" or "the present value" of the V{i) = Z
project.
If the stream of profits is P0, Pu P2, . . . ,Pn, (1)

What kind of measure of profitability is

*From UN I DO, Guidelines for Project Evalua-
tion, United Nations, 1972, pp. 18-26. Reprinted by the present value? It is a fairly 0' measure,
+ good
SS 0 stream
permission. since it converts the entire of profits
 647 BENEFITS AND COSTS

into one number representing the total year. As he lends more and more he will tend
amount of profits today that would be equiv- to shift his preference in favour of money
alent to the entire stream of profits. The now; this rise in the rate of discount will con-
equivalence is defined as that given by the ap- tinue until it equals the market rate of
propriate rate of interest. For the evaluation interest.
of commercial profitability, the appropriate Thus, for a private operator, the correct
rate of interest is the rate that rules in the discount rate is the market rate of interest, at
market for borrowing and lending. If a per- least if we accept the conventional assump-
son can lend and borrow at 10 per cent, he tions ofeconomic rationality. And this he can
would have no reason to choose a project that p,
use for calculating the present value of a pro-
costs £100 today and yields only £109 next ject. Itis, of course, possible that the interest

-z
year. Similarly, he would have no reason to rate would vary from period to period, and
reject a project that yields £111 one year the calculation of the present value need not
from now and costs only £100 today. Even if be based on the assumption of an unchanging
he does not have the money, he can borrow it interest rate. Suppose that the interest rate
at 10 per cent and make a net profit of £1 by between year 0 and year 1 is iu and that be-
next year. tween year 1 and year 2 is i2 (and so on), then
What it may be asked, if the person in the present value of the project is given as:
question has a personal rate of discount that
is different from 10 per cent? What if he Wu b • • • , /„)
finds £115 next year to be equivalent to £100
today? Would he then not be justified in re- (2)
f^O + /,)... (l + /,)
jecting a project that yields £111 next year
and costs £100 this year? He would be, but It is clear that (1) is a special case of (2)
the question that arises in this case is: Why when *i = 1*2 = • • • = i^
does he have a personal rate of discount of 1 5 A relevant question to ask is the following:
per cent when he can borrow and lend at 10 At what rate of interest, constant over time,
per cent? If there were this difference, should would the present value of a project be ex-
he not proceed to borrow a large quantity of actly zero, i.e., for what / would we have V{i)
money, since he pays only £110 next year for = 0? This is not unduly complicated to cal-
£100 this year while he considers £115 next culate. An interest rate for which the present
year to be just as satisfactory as £100 this value of a project is zero is called the "inter-
year? He should thus make use of the market nal rate of return" of the project. It is per-
by borrowing, and this he should continue to fectly possible in principle that the present
do until his personal rate of discount comes value of a project may become zero at more
down to the level of the market rate of inter- than one rate of interest, e.g., the stream
est of 10 per cent. (Another possibility is that (100, -500, 600) has two internal rates of
the market rate of interest will rise, thanks to return, namely, 100 per cent and 200 per
his borrowing, but this would be unlikely in a cent. This problem, while somewhat intrigu-
large market, since any individual would tend ing, is perhaps not very important for a va-
to be a small operator compared with the en- riety of reasons. For one thing, this multiplic-
ormity ofthe entire market.) Similarly, if the ity of the internal rate of return would never
person initially has a personal rate of dis- occur if a project yielded losses up to a cer-
count lower than the market rate (e.g., if he tain point in time and thereafter yielded prof-
thinks that £100 this year is equivalent to its (e.g., after it comes into operation). In
£108 next year), then he should use the mar- such a case, the higher the interest rate the
ket as a lender. He can get £1 10 by lending lower will be the present value (as in Figure
£100 this year, and clearly it is worth his 1), and the internal rate of return must be
while to do this, since he regards £108 to be unique (0A in Figure 1).
enough compensation for losing £100 this While we cannot spend time on the qucs-
 648 PROJECT APPRAISAL

a dam on a certain river, and the construction

of one would eliminate the possibility of hav-
ing the other. Suppose that A yields a higher
internal rate of return (say, 20 per cent
rather than 10 per cent) and a lower present
value at the market rate of interest of 5 per
cent (say, £1 million rather than £2 million),
which should we choose? The conflict arises
with the question of ranking the two projects
in terms of their relative desirability. Figure
2 represents this and also illustrates that such
a conflict can arise even though the present
value of each project declines with the dis-
Discount rate count rate.
One can say that, in a significant sense, the
FIGURE 1. Present value at alternative discount
rates. present value is a better guide than the inter-
nal rate. It yields a measure of total gains
which the latter does not. If 5 per cent is the
market rate (and also the rate at which a per-
tion of multiplicity of the internal rate of re- son isready to discount future earnings), pro-
turn, the question of the relative significance ject B yields twice as much return as project
of the internal rate of return and of the pres- A, and clearly this is a good thing for B. The
ent value is an important one and should en- fact that the rate of return is higher with A
gage our attention. The present value at the than with B is not very compelling so long as
market rate of interest and the internal rate the choice is between having project A and
of return are both measures which convert having project B. To take an extreme case, a
the stream of profits into one number, and it person may get a return of 500 per cent on a
is relevant to know which one is a better penny and a return of only 20 per cent on
guide. Do they, in fact, conflict? In choosing $ 1 ,000. The latter is likely to be more attrac-
projects we may follow the rule that all proj- tive ifthe market rate of interest and the per-
ects with an internal rate of return higher son's rate of discount are 10 per cent. The
than the market rate of interest should be present value gives a measure of total gains,
chosen. Alternatively, we may recommend which the internal rate does not. In what fol-
that all projects with a positive present value lows, the present value of the commercial re-
should be selected. Does it make any differ- turns of a project at the market rate of inter-
ence which of the two rules we follow? The est will be taken as a measure of the
answer is "No, not at all," so long as the pres- profitability of a project.
ent value always goes down as the discount The present value is a good framework also
rate is raised, as in Figure 1 . If the market for public choice even though the rate of dis-
rate of interest is less than OA, the project count tobe used should not necessarily be the
represented in Figure 1 should definitely be market rate of interest but the appropriate
chosen according to both criteria, while if it social rate of discount; also, the profits should
is more than OA it should be rejected by both be social profits as measured from the point
tests. There is no conflict whatever in all this. of view of the society and not commercial
The real conflict arises when the rule of profits as normally defined. The reason for
choosing all projects satisfying either of these preferring the present-value formulation in
criteria cannot be followed. There may be this case is the same as in the case of a com-
certain specific constraints. For example, the mercial evaluation, that is, the present value
choice of one project may rule out another, gives a measure of total gains (in this case
e.g., projects A and B may be two variants of social gains and not commercial gains).
 649 BENEFITS AND COSTS

10% 20%
Discount rate

FIGURE 2. Conflict between present value and discount rate.

LIMITATIONS OF COMMERCIAL First, how much money a person is ready

PROFITABILITY to offer depends on his income level. A rich
man may offer a good deal of money for
In what respects does a measure of com- trivia, while a very poor person may find it
mercial profitability differ from that of na- difficult to spend even very small amounts of
tional economic profitability? Why is it that money on essentials. The price offered in the
commercial profits are often treated as a bad market is, thus, not a good guide to social
guide to social gains? If the framework is one welfare, for it includes the influence of in-
of "present value" in both cases, clearly the come distribution on the prices offered.
distinction must lie in the divergence between One could, of course, retort by asking why,
commercial profits and social gains in any if the guardians of public policy do not like
year and that between the market rate of in- the income distribution (e.g., if they disap-
terest and the appropriate social rate of prove of the existing inequality), they do not
discount. reform it directly. Once the distribution is re-
Commercial profits measure the difference formed, the project evaluator can simply
between the value of earnings and costs in a treat the money prices offered as guides to
certain period. The earnings are direct welfare without worrying about income dis-
money earnings of the firm at market prices tribution. This retort, while not uncommon,
and the costs are money costs, again at mar- is somewhat hollow, since there are con-
ket prices. . . . Social benefit-cost analysis straints— political, economic and social —
must go deeper and ask what is the meaning that prevent such reforms of income distri-
of market prices, that is, what do they rep- bution, and given these limitations the exer-
resent? Ifa person is ready to offer £1 for cise of project evaluation cannot be based on
something, he expects to get at least £1 worth the notion that all appropriate income redis-
of satisfaction from it. Does this mean that tributions have already been carried out.
the social value is also worth £1? Not Further, one of the simpler means of in-
necessarily. come redistribution may, in fact, be project
 650 PROJECT APPRAISAL

selection. For example, the choice may be be- satisfaction from that particular good; but he
tween project A to be located in a poor region could expect to get more, conceivably much
or project B to be placed in a rich area, or more. If we look at the market value of a
between project X, which uses a large good produced by a project, we get a measure
amount of poor, unskilled labour, which of a floor to expected satisfaction. But in fact
might otherwise be unemployed, and project the consumer may expect more and get more.
Y, which uses factors of production supplied If we try to go into the question of total sat-
by rich people. Project choice has distribu- isfaction from a project, we would have to ex-
tional implications, and sometimes it may be amine the excess of what consumers are will-
politically or socially more feasible to redis- ing to pay for its products and what they
tribute income this way rather than through actually pay. This difference is sometimes
taxes or other direct means. We have, there- called the "consumers' surplus." In Figure 3,
fore, quite a legitimate reason to consider dis- the line AB represents the maximum a con-
tributional questions in evaluating social sumer isready to pay for each unit of a good.
gains from a project. This immediately takes If the market price is BC, he will then buy
one beyond commercial profitability. OC units of it. The total expenditure he will
Second, a project may have influences that make on it is ODBC, which will represent the
work outside the market rather than through earnings of the producer from him. But the
it. For example, a particular industrial pro- value of the satisfaction that he anticipates is
ject may produce a great deal of smoke and more, namely, OABC The difference, ABD,
foul air in the town in which it is located. Or represents the surplus that he enjoys. While
a firm may train the labour force in the re- commercial profitability takes no account of
gion. While the first impact may be undesir- it, it is clearly a relevant consideration for
able and the second laudable, the firm's prof- public-project evaluation.
its may not reflect either. The cost of bad It is worth noting that for the last unit pur-
health or the unpleasant life of neighbours chased there is no surplus, since the price
may not depress commercial profits, and the paid, BC, is no lower than the price the buyer
reward for training may not go to the firm, is ready to pay. Thus, the question arises not
since, after training, the workers are free to for variations near point B, but for the choice
ieave. between bulky projects. In ascertaining the
The effects that work outside the market social gains from one sizable project rather
are called "externalities." These do not enter than another, it is relevant to know the re-
calculations of commercial profits, since
spective magnitudes of the consumers'
these are made at market prices. Externali- surplus.
ties are obviously relevant for social choice
and provide a sufficient argument for reject-
ing commercial profitability as a guide to
public policy. Externalities can arise in the
prucess of production (e.g., industries causing
water pollution), in the process of consump-
tion (e.g., additional private cars adding to
the crowding of the roads), and also in the
process of sales and distribution (e.g., garish
shop display or advertising affecting the tran-
quillity ofthe community). Externalities are
often most pervasive.
Third, even in the absence of externalities
and considerations of income distribution,
commercial profitability may still be mislead- Amount demanded
ing. If a consumer is ready to pay £1 for a
good, he expects to get at least £1 worth of FIGURE 3. Consumers' surplus.
 65 1 BENEFITS AND COSTS

Considerations of income distribution, ex- the market; the product is homogeneous, i.e.,
ternalities and consumers' surplus are among there are no variations in quality. It can be
the factors that distort commercial profits as proved that if there are no externalities (and
a measure of national gains. There are other a few other relatively minor conditions are
factors also, but the considerations discussed satisfied), an equilibrium in a perfectly com-
should have been sufficient to illustrate the
petitive market must achieve "economic effi-
difference between social and private gains. ciency," which is defined as a state where no
Another ingredient of present-value esti- one can be made better Off without making
mates isthe social rate of discount. Given any someone else worse off. Economic efficiency
series of profits (private or social) the size of is also sometimes called "Pareto optimality"
present value depends on the rates of dis- after the economist Pareto. Each firm maxi-
count. The social rates of discount may differ mizing its profits at given prices helps to
from commercial rates of interest for many achieve this type of optimality for the society.
reasons. An individual may expect to live Is this result — which is made much of in
only a certain number of years and the dis- the formal economic literature — a compel-
counting ofthe future that springs from this ling reason for recommending that public-
limitation may not be appropriate for social project evaluation be guided by profit max-
choice, since the planners may wish to take a imization at given market prices? The an-
longer view and give greater importance to swer isa decided "no." First, the absence of
the welfare levels of future generations. Even externalities is a very dubious assumption,
the general public of today (as distinct from since they are quite widespread. In the pres-
policy makers) may feel that for public proj- ence of externalities, even perfect competi-
ects, where all are forced to save simulta- tion may not achieve economic efficiency.
neously, alower rate of discount may be ap- Second, economic efficiency is a very limited
propriate than would be reflected in the requirement. It tells us nothing whatever
market behaviour of individuals. Individuals about income distribution. Some people may
may be ready to sacrifice for the future only be terribly poor and others extremely rich,
if others are ready to do the same, and while and still if no poor person can be made better
such a joint action is possible through public off without making someone else worse off,
policy, there is no way of bringing it about in the situation will still be called economically
individualistic market behaviour. Deep prob- efficient.
deep. As a criterion it does not go very
lems are involved here. . . . Suffice it to note
now that there are no compelling reasons to A no less disturbing consideration is the
believe that the market rate of interest must recognition that the result is based on all
be the appropriate rate at which to discount markets being perfect. Suppose some are and
future benefits. This is still another reason for some are not. There will then be no reason for
commercial profitability to differ from mea- believing that one would move closer to eco-
sures of social gain. nomic efficiency by maximizing profit at
given market prices, and indeed such a policy
might conceivably take one further away
from economic efficiency. The rule in ques-
PROFIT MAXIMIZATION tion works only if all units are in competitive
AND EFFICIENCY equilibrium. It does not give much guidance
to an individual enterprise if imperfections
The usual defence of commercial profit exist in the rest of the economy.
maximization as a criterion is based on the Thus, it should be clear that the relation-
implicit assumption of "perfect competition." ship between efficiency and profit maximiza-
Perfect competition is a case where there are tion under competitive conditions does not
many sellers and many buyers so that no one really establish any very strong reason for
has any monopoly power; each person has basing project selection on commercial profit
perfect knowledge; there are free entries into maximization at given market prices.
 652 PROJECT APPRAISAL

X.B.3. Social Cost-Benefit

Analysis*

In offering guidelines for the use of cost-ben- across the board, and may virtually force the
efit analysis in developing countries we pay monetary authority to increase total money
special attention to industry and agriculture, demand if a recession of activity is to be
as well as to infrastructural projects where avoided.
the output has a market price. Education, If inflation proceeded uniformly so that
health, and defence are neglected. This is not relative prices were unaffected, it would not
meant to imply that useful work is not going be a reason for prices to be a poor measure of
on in these fields. Certainly, cost-effective- real costs and benefits. But this, for institu-
ness analysis can be applied. But it is still tional and political reasons, is seldom the
very controversial whether full cost-benefit case. For example, governments in such cir-
analysis in such sectors, where benefits are cumstances will often use price controls in se-
particularly difficult to measure, is as yet suf- lected fields where they can in practice be op-
ficiently soundly based to be a good guide for erated. This makes activity in these fields
policy makers. relatively or absolutely unprofitable, without
Thus we are concerned with the applica- regard to the net benefit of such activities.
tion of cost-benefit analysis precisely in fields A particular case of such control concerns
in which it is considered unnecessary in de- the price of foreign exchange, which brings
veloped economies. The justification for this us to the next reason.
can only be that it is felt that within such sec-
tors of more advanced economies the price
mechanism works in such a way that profits Currency Overvaluation
are a reasonable measure of net benefit, but In almost all countries, the government
that this is not true of most developing "manages" the price of foreign exchange.
countries. With inflation, if the exchange rate is unal-
Why should one start with the presuppo- tered, domestic prices get out of line with
sition that actual prices are very much worse world prices. This implies that, on average,
reflectors of social cost and benefit than is the
the rupee1 prices of imports and exports are
case in advanced economies? The main rea- too low relative to those of goods which are
sons are briefly adumbrated below. not traded. So long as the currency is not de-
valued to rectify the situation, the demand
for foreign exchange for imports and other
Inflation
purposes will exceed the supply, and the gov-
Very rapid inflation is more common in de- ernment will be forced to restrict imports,
veloping countries, particularly in South often in ways which open up gaps between
America. This is no accident. The very ur- the market prices of goods and the real cost
gency of the desire to develop rapidly results of procuring them. But some governments
in a constant tendency for demand to outrun faced with a price inflation do not resort to
supply: furthermore, lagging supply in the import controls in order to maintain the do-
sectors which are most resistant to change, mestic currency overvaluation, but devalue
particularly agriculture, results in sectoral
price rises which tend to transmit themselves 'We use 'rupees' to stand for the domestic currency
unit, and 'dollars' to stand for a unit of foreign ex-
change. This is solely because it is awkward not to
*From I. M. D. Little and J. A. Mirrlees, Project have a short familiar expression for these units: forced
Appraisal and Planning for Developing Countries, to choose, we selected rupees and dollars as being the
Heinemann, London, 1974, pp. 29-37. Reprinted by units of the largest noncommunist developing and de-
permission. veloped countries, respectively.
 653 BENEFITS AND COSTS

more or less frequently. If inflation is rapid ploying farmers exercise some monopsonistic
and the government devalues periodically but
not very frequently, then it is inevitable that power.
That men by working are unable to con-
the currency will be alternatively underval- tribute as much to production as they con-
ued and overvalued. If the inflation is slow, sume iswhat is meant by underemployment.
the government usually tries to avoid deval- The extended family system permits under-
uation, and long periods of overvaluation are employment inthe towns as well as the coun-
likely. tryside. Ifrelief were given institutionally,
via unemployment benefits, the very low pro-
Wage Rates and Underemployment ductivity urban activities — petty trading,
car-watching, etc. — would largely disappear
It has been seen that the theory of com- and more people would become openly and
petition requires that the marginal product of wholly unemployed, a circumstance which
labour (the extra output resulting from the would, of course, imply that wages did not re-
employment of a small extra amount of la- flect the social cost of employment.
bour) be equal to the wage paid. The real cost of employing a man in the
Because of monopoly power, and immobil- modern sector is still a subject of controversy,
ity, there are undoubtedly serious imperfec- mainly because insufficient is known about
tions in the labour markets of many indus- the effects on the traditional sectors including
trialized countries. But these imperfections agriculture, and because these effects will
are not usually thought to cause major inter- vary widely from country to country, and
sectoral distortions of the pattern of produc- perhaps from region to region, or even town
tion (regional distortions may be an excep- to town. However, there is rather wide agree-
tion, and here wage subsidies have been ment that modern sector wages almost every-
used). On the other hand, it is often argued where overstate, perhaps greatly overstate,
that this is the case in many developing the social cost of employment.
countries.
In "modern" sectors of the economy — in-
cluding modern industry and commerce, gov- Imperfect Capital Markets
ernment, and plantations — it is common to Where risks are equal, interest rates on
find that unskilled workers earn three or four loans should be equal, if profits are to mea-
times as much as casual rural labour, a dif- sure net social benefits. Interest rates have
ference far greater than can be accounted for such an enormous range in many developing
by the difference in the cost of living; and countries, that it is implausible to suggest
therefore that the cost of employing people in that this is just a measure of differential
these sectors is apparently much greater than risks. Other factors operate, such as govern-
the loss of rural production, assuming that ment intervention, ignorance, and monopoly
such rural earnings are a fair measure of la- elements in the supply of capital, to widen the
bour's marginal contribution to production. range from low to almost astronomical rates.
It has been argued that the earnings of casual
labour overstate the marginal product of la- Large Projects
bour. This is because, in most developing
countries, the greater part of rural labour is It is more common in developing coun-
family labour. Since a dependent member of tries— especially in small countries with, as
the family cannot be sacked, he may "earn" yet, little development — that a project will be
(i.e., consume) as much as a hired man but so large as to have important repercussions
yet have a lower marginal product. As on profits elsewhere in the economy. In these
against this, in some places it is probable that circumstances, as we have seen, the profita-
the marginal product of a hired man is bility of the project itself cannot be regarded
greater than his earnings because the em- as a good measure of net social benefit.
 654 PROJECT APPRAISAL

Inelasticity of Demand for Exports system which protects it in its home market
puts it at a positive disadvantage in export
In a number of developing countries, a
large part of export receipts is accounted for markets; whereas reason suggests that if in-
by one, two, or three export commodities. dustrial production is worth special encour-
agement, then it is worth special encourage-
Where a country also accounts for a consid- ment, and not actual discouragement, in
erable part of total world production, then it
producing for export. Thus tariff protection,
can influence, within limits, the price it ob- like currency overvaluation, implies that the
tains by restricting sales — which is, of
course, an abrogation of the conditions of rupee price obtainable for an export under-
estimates the social value of that export.
perfect competititon. The free market price
cannot then correctly measure the benefit, Some developing countries have taken mea-
sures to offset this effect, but such measures
because, like any monopolist, the country
would gain if it exported less at a higher are often insufficient, and not very scientifi-
cally devised in such a way as to make the
price.
This, in turn, implies that the country country. price measure the benefit to the
rupee
would gain by devoting rather less resources
to producing these primary commodities, and Apart from the fact that protection dis-
rather more to others, or to industrialization. courages exports of both industrial and agri-
cultural products, it is also the case that dif-
This situation can be best rectified by suita- ferent industries receive enormously different
ble export taxes on the commodities, together
degrees of protection, usually for no appar-
with other policies (including use of the rev-
ently rational economic reason. This situa-
enue thus raised) which encourage the trans- tion has arisen partly because countries have
fer of resources. Some countries recognize
selected industries or plants (or have agreed
this situation and do in fact use export taxes.
But the situation has also been used as an ar- to protect private initiatives) without the
kind of economic appraisal being advocated
gument for encouraging industry by protec- here. Protection has followed the establish-
tion— which brings us to our next section. ment of industries, rather than itself being
used as a screening device.
Another reason why the relative gap be-
Protection — Import Quotas, Tariffs,
Export Disincentives tween domestic and world prices is highly di-
vergent asbetween industries is the extensive
The protection of domestic industry may use of import quotas. A country runs into bal-
be a deliberate interference with the price ance of payments problems. The situation is
mechanism designed to make it operate in a brought under control by restricting imports
manner more conducive to society's benefit and, naturally, the least essential goods are
than would a laissez-faire commercial policy. most restricted. The result may be a growth
A well-designed interference, in the shape of of domestic industry, behind protective quo-
special encouragement of industrialization, tas, which bears little relation to the long-run
may well make industrial profits a better comparative advantage of the country. If a
guide to social advantage than they otherwise wrong industry gets established it handicaps
would be, either for the reasons given above, any other industry which uses its output. For
or for other reasons. instance, steel-using industries will be hand-
The main way in which industry is spe- icapped bya high-cost local steel plant, un-
cially encouraged is by tariffs and import less the latter is subsidized so that it can sup-
quotas. Thereby, the domestic price of the ply at prices no higher than the import price.
output is kept above the import price. But the It is our belief that bad management of for-
outputs of one industry are often the inputs eign trade or foreign exchange is one of the
of another. Consequently, when an industry principal reasons why internal prices get
contemplates exporting, it finds that the very highly distorted, and hence lead to industrial
 655 BENEFITS AND COSTS

investments which are of little or no benefit at which society ought to discount the future
to the country concerned. may differ from the rate at which a firm can
We have now outlined seven important and borrow. Thus, if the government chooses a
fairly non-controversial reasons why the discount rate for projects which is lower than
price mechanism and the profit motive may the market rate of interest,2 this is in effect
not work as closely for the social advantages to say that it considers future consumption to
as in developed countries. Other more gen- be more valuable than is indicated by the ag-
eral reasons could be adduced, such as igno- gregate choices of private individuals. If the
rance of opportunities and techniques, iner- public saved more, interest rates would be
tia, short-sightedness, lack of a market lower, and the government pleased. In other
economy, and greater fragmentation of mar- words, the government considers present sav-
kets leading to local monopoly power; but ings to be more valuable than present
these have relatively little direct bearing on consumption.
project evaluation especially in the public Governments can reduce aggregate private
sector. We turn now to a further three rea- consumption, and thus increase savings, by
sons, which may be more controversial. taxation. On the other hand, taxation has ad-
ministrative and political costs. So perhaps it
is money in the hands of the government
Deficiency of Savings and
Government Income which should be considered to be more valu-
able than private consumption: this view is
Two projects may have the same net profit, strengthened by the fact that a rational gov-
but a different effect on the relative amount ernment should see to it that the value of its
of extra consumption, savings, and taxation. expenditure at the margin is equal in all lines,
As we saw, economic theory often treats whether it be defence, agricultural extension,
savings and investment as of equal value. education, or investing in industry. Many
This is really a facet of the principle of con- people will be rather unwilling to accept that
money in the hands of the government is
make no sumers' sovereignty. toIt is
difference assumed
benefit that itsome
whether can
more useful than many kinds of private ex-
extra income is consumed, or saved and penditure, especially when governments are
hence made available for investment. This is seen to waste money and promote silly in-
reasonable for an individual who freely vestments. But the project evaluator may in
chooses whether to spend or not. For him, an any case have to take a government view.
extra dollar of savings is worth the same as This is a difficult and controversial matter,
an extra dollar of consumption. But is it true which will be taken up again.
for society? Finally, it should be noted that although
To cut a long story short, if the government discussion of this problem has arisen mainly
believes that rather more savings and rather in the context of developing countries, it
less current consumption would be good for seems to us that it arises also in the case of
society, there may be a conflict. The point is rich countries.
that savings can be transformed into invest-
ment, and investment can produce extra fu- The Distribution of Wealth
ture consumption for a sacrifice of present
consumption: and the government may put a The preceding section was largely con-
relatively higher value on the consumption of cerned with the distribution of benefits, as
people in the future than do private persons. between the present and future. But there is
Furthermore, private persons may be inhib-
ited from saving by income and other taxes
which have the effect of double-taxing sav- 2,The market rate of interest' may be quite a wide
band in developing countries, even if we restrict the
ings. We have already referred to these prob- meaning of 'the market' to that for medium and large
lems above, where it was argued that the rate scale industrial borrowing.
 656 PROJECT APPRAISAL

also a problem of the distribution of benefits strong prima facie case for the use of cost-
today — the problem of inequality, to which benefit analysis.
we have already referred. There is a dilemma We have seen that the basic idea of such
here, for inequality tends to promote savings, an analysis is to use hypothetical rather than
and help future generations. This is espe- predicted actual prices when evaluating a
cially true of corporations: company profits project. The rate of discount may also not
belong mainly to the rich, but are one of the correspond to any actual interest rate. These
main sources of saving. The dilemma can be 'shadow' prices, as they are often called, are
made less acute insofar as public savings can, chosen so as to reflect better the real costs of
by increased taxation, take the place of sav- inputs to society, and the real benefits of the
ings of the rich; but there is a limit to this, outputs, than do actual prices.
and some element of dilemma remains.
The name 'shadow price' is perhaps unfor-
The extent to which project selection tunate. It suggests to many, even to some
should concern itself with different kinds of economists, that an analysis based on them is
inequality will come up again. There is the remote from reality, and therefore academic
additional important question of how far a and highbrow, and so is to be distrusted. Of
practicable criterion for project selection can course, shadow prices may be unreal in that
take proper account of inequalities. they are not the current prices of goods in a
market. But then no price in a project anal-
External Effects ysis can ever be an actual price — for every
price assumed in such an analysis necessarily
Some economists believe that external lies in the future. The whole point of a
economies are of special importance in devel- shadow price is indeed that it shall corre-
oping countries: that some industries have spond more closely to the realities of eco-
important beneficial effects on others in ways nomic scarcity and the strength of economic
which cannot be, or anyway are not, reflected needs than will guesses as to what future
in the price obtainable for the output of the prices will actually be. From now on we shall
industry, or in the price it pays for its inputs. use the term "accounting prices."
There has been much speculation and debate It is worth emphasis that if any input or
on this subject. But there is very little positive output is valued at a different price from that
evidence. Certainly there has been much actually expected to be paid or received by
naive wishful thinking — for instance, that the project, then, in our terminology, a social
the provision of electricity, steel, or transport, accounting price is being used. In this sense,
would somehow create its own demand.
most project appraisals have made use of ac-
It was already shown that many of the counting prices. For instance, it is widely ac-
more obvious external effects can be allowed cepted in project analysis that indirect taxes
for by a suitable definition of the project to on inputs should not be counted as costs. Or
be considered. But others will remain. again, for some years now, direct imports and
exports of projects have often been valued at
c.i.f. or f.o.b. prices (border prices, as we
SOCIAL OBJECTIVES AND THE shall term them) by, among others, consul-
NOTION OF ACCOUNTING PRICES tants working for the IBRD. Some evaluators
may think that they are not using shadow
A rather strong case has now been pre- prices when they make such adjustments.
sented for saying that a project's anticipated That is a matter of terminology. What we
receipts and expenditures cannot be relied want to make clear is that, in our terminol-
upon to measure social benefits and costs in ogy, they are using accounting prices.
most developing countries. It is believed that While accounting prices have been in use
this is true also of more developed economies, for some time, they have seldom been used in
but to a lesser extent. There is therefore a a comprehensive and systematic way, but
 657 BENEFITS AND COSTS

rather haphazardly. This is dangerous. Once much better reflections than actual prices for
some important prices become badly dis- many projects in many countries. Nor, of
torted— e.g., the price of labour or foreign course, is it claimed that the use of account-
exchange — the repercussions are widespread. ing prices is a very satisfactory method of
Every price is then liable to need adjustment. dealing with distortions. Many of the distor-
What we are primarily concerned with in this tions can be fully dealt with only by removing
book is to show how a whole set of accounting them — that is, by adopting policies which
prices can be systematically and logically es- lead to proper correspondence of prices and
timated and applied, yielding a practical costs and benefits. There may be yet others
method of analysis which can be expected to which, because of the difficulty of measuring
measure net social benefit better than ordi- them in a reasonably objective way, cannot
nary profitability analysis. Being practical be satisfactorily allowed for in a usable and
precludes perfectionism. We make no claim politically acceptable criterion. These have to
that accounting prices can be exact reflec- be left to the judgment of the politician and
tions of social costs and benefits — merely his advisers.
 X.C. SHADOW PRICING
X.C. 1 . Distortions in Factor
Markets*

It is appropriate to begin by stating the ob- equivalent, if the same assumptions are made
vious: cost-benefit analysis is undoubtedly the about the economic environment, though nat-
most used, and arguably the most useful, urally there are differences in emphasis as to
form of applied welfare economics. Its theo- which set of assumptions is more relevant for
retical basis as well as its limitations are
LDC's in general, and more importantly in
therefore necessarily those of its parent, the- the practical problems of estimating the rel-
oretical welfare economics. This paper is not evant values to be included in the NPV/ IRR
for those who deny any practical use for the- index, with accuracy and ease.
ory, but for those who whilst recognizing the The reason why in principle most of the
limitations of theoretical welfare economics methods are equivalent, given the same basic
nevertheless feel that in our present state of assumptions about the economic environ-
knowledge it provides the only basis for mak- ment, istheir common lineage — theoretical
ing an economic assessment of investment welfare economics. One of its basic results is
plans and proposals. that in a perfectly competitive economy (with
The purpose of any project selection pro- no uncertainty about future tastes and tech-
cedure must be to provide a decision rule for nology), allocation of resources on the basis
accepting or rejecting a project. The net pres- of market prices of goods and factors (for
ent value (NPV) or the internal rate of return which markets exist) would result in Pareto
(IRR) of the project is the index usually used. optimality for a given income distribution.1
Our chief concern in this paper will be with, Market prices of goods and factors would
first, what should be included in the time equate and equal the marginal social cost
stream of benefits and costs; secondly, what (MSC) of producing and the marginal social
are the relevant values of the various cost- value (MSV) of using the relevant goods/ fac-
benefit components; and thirdly, how the dis- tors. For a truly marginal investment project
count rate (or rates) needed for determining (in the sense that it does not alter the MSV
the NPV, or the cutoff IRR at which projects
are accepted, should be chosen. Most of the
'Pareto optimality necessitates that for a given dis-
differences in the alternative procedures re- tribution ofincome:
late to apparently differing prescriptions in (i) the marginal rates of transformation in pro-
these three respects. duction of different commodities are equal to
It will be repeatedly emphasized in this their marginal rates of substitution in con-
sumption,
paper that any substantive differences among (ii) the marginal rates of substitution between any
the alternative procedures are in large part pair of factors are the same in all the industries
dependent upon differing assumptions about in which they are used,
(iii) the marginal rates of substitution of any pair
the relevant aspects of the economic environ- of commodities is the same for all individuals
ment in which the investment decisions are
consuming both goods.
being made. One of the basic purposes of this Given that the above conditions hold, a Pareto opti-
paper will be to demonstrate that, in princi- mum will exist, such that for the given income distri-
ple, most of the suggested procedures are bution itwill not be possible to make one person better
off without making someone else worse off. Treating
the same physical commodity at different dates as
*From Deepak Lai, Methods of Project Analysis: many different commodities, equivalent intertemporal
A Review, World Bank Staff Occasional Papers No. marginal equivalences for an efficient intertemporal
16, Johns Hopkins University Press, Baltimore, 1974, program can be derived. See Dorfman, Samuelson
and Solow [21 .
pp. xiii-xvii, 24-38. Reprinted by permission.

658
 659 SHADOW PRICING

and MSCs of the output it produces and in- which are therefore generally needed in in-
puts ituses as a result of its operation), the vestment appraisal because of the divergence
values of the output and inputs at market between the MSC and MSV of the relevant
prices would provide the correct values to be commodities.
used in determining the net present value of If neutral fiscal devices (lump-sum taxes
the project.2 Market prices would be the and subsidies) are feasible, then a full Pareto
"shadow" prices to be used in project optimum could still be achieved if the govern-
selection. ment eliminates the divergence between
If the investment project being considered MSC and MSV by suitably corrective tax-
is not marginal (or if there are externalities), subsidy measures, thereby restoring the
and does affect the MSV and MSC's of its equivalence of MSC and MSKwith the mar-
output and inputs, then the relevant mea- ket price of the commodity. However, for ob-
sures of the social benefits and costs of the vious reasons it will not be possible, in most
project will be the change in the consumers' cases, to cure the divergence in this manner.
and producers' surpluses caused by the pro- In that case, the divergence between the
ject. This, in principle, will be the procedure MSC and MSV of the commodity may have
recommended by all the project selection to be taken as a datum (or a constraint) and
procedures we shall consider. In the case of the "shadow" prices corresponding to this
the perfectly competitive model, valuation of constrained (or "second-best") welfare opti-
the changes in producers' and consumers' mum will need to be computed. A large num-
surpluses, at market prices, will provide the ber, ifnot most, of the shadow prices which
correct indication of the net social benefits of
we shall consider are of this "second-best"
the project. kind.
To the extent, however, that the perfectly Second, even if the government can elimi-
competitive paradigm does not hold — for ex- nate the divergence between A/SC's and
ample due to the existence of monopolies, MSV's by suitable tax-subsidy policy, it may
taxes and subsidies, externalities, and/or in- take time for the divergence to disappear.
creasing returns — market prices will no Then current market prices will not equate
longer indicate the social costs and benefits the MSC and MSV of the relevant commod-
of using and producing different commodi- ities, but it is expected that future market
ties. The social cost to be included in the prices will. As investment takes time and its
NPV/IRR index of social profitability, prop- effects are extended into the future, it is
erly defined, will still be the marginal social
cost of the various inputs used, and the social clearly the MSCs and MSV's of the relevant
inputs/outputs appropriately dated which
benefit will be the marginal social value of
will be relevant in working out the project's
the output produced. However, the break- social profitability. If it appears likely that in
down of the perfectly competitive assump- the future an existing divergence between
tions results in market prices no longer equat- MSC and MSV will be corrected, the appro-
ing and equaling the MSC and MSV of the priately dated price which reflects the social
relevant commodities. The market price will
cost/benefit of the project will not be the cur-
not equal either the MSV or MSC — and in rent market price, nor the current MSC and
some cases of rationing may not equal either.
MSV of the commodity, but rather the "equi-
The problem then is to adjust the market librium" price which is expected to prevail in
price to obtain the relevant "shadow" prices, the future. In this sense, even when an econ-
omy is moving towards an optimal set of mar-
2The net benefits being discounted at the optimal ket prices, from a distorted current set, it
discount rate which equates the marginal rate of
transformation (mrt) in production of present into fu- may be necessary to use "shadow" prices cor-
ture consumption, to its marginal rate of indifferent responding to the future optimal market
substitution {mrs) in consumption, determined in a prices, rather than the current market or
perfect market for intertemporal consumption. shadow prices for pricing inputs and outputs
 660 PROJECT APPRAISAL

which form the time stream of benefits and question; but to the extent that future gov-
costs of the investment project. ernment policies are normally unknown, the
Third, even for a perfectly competitive element of judgment involved in deciding
economy, there will be different Pareto op- which of these alternative assumptions is rel-
tima associated with different income-distri- evant, when considering existing distortions
butions. Judging between these different in commodity and factor markets, will be of
Pareto optima will necessarily involve nor- paramount importance in deciding which is
mative judgments about the desirability of the correct "shadow" price to use. Hence it is
particular income distributions.3 Even if important to remember that differing pre-
agreement can be reached on the desired in- scriptions on alternative evaluation proce-
come-distribution, there will still be the prob- dures will most often be due to differing im-
lem of legislating this "optimal" distribution. plicit assumptions about the current and,
Again if neutral fiscal devices in the form of more importantly, the future economic
lump-sum taxes and subsidies are feasible, environment.
the government would be able to achieve a Second, though we have been discussing
Pareto optimum with the optimal distribu- the evaluation of a particular project and the
tion of income. If however, as is more likely, social valuation of its inputs and outputs in
neutral fiscal instruments are not available, what may appear to be a partial equilibrium
then the distributional effects of investment framework, in principle, any proper invest-
projects will also have to be computed, and ment criteria must take account of the total
judged against and along with their purely (direct and indirect) or what are termed the
"production" or "efficiency" effects. These general equilibrium effects of the investment
problems open up other areas where there project. Thus for instance if an industrial
may possibly be conflicting judgments, and project employs some seemingly underem-
hence prescriptions for project selection ployed labor in the urban sector, the ultimate
procedures. effects via the impact on rural-urban migra-
tion could be a significant change in total out-
Practical Problems put of the economy. The shadow wage rate
will then in this case have to incorporate both
These theoretical problems are com- the direct and indirect (via migration) effects
pounded bypractical ones. First, even though of increasing industrial employment. The
there may be agreement about the nature of MSCs and MSV's which are taken as the
the correct prices to be used in project selec- "shadow" prices in determining the social
tion, there may, nevertheless, be disagree- profitability of the investment project, must
ment as to whether or not the existing diver- therefore be the general equilibrium "shadow
gences between MSV's and MSC's which prices." This might appear to be an impossi-
affect these prices will continue into the fu- ble task, but the relative merits of alternative
ture or whether they will change. Depending investment appraisal procedures will depend
on what assumption is made about the future upon their success in taking account of the
course of the economy, the "second-best" or general equilibrium effects of projects, which
"first-best" shadow price will be the relevant will in turn, if the procedures are to be prac-
one to choose.4 In a sense, this is an empirical tical, necessitate making certain simplifying
3It being noted that investment projects affect both assumptions about the economic environ-
the intratemporal as well as the intertemporal distri- ment. Once again, these assumptions, though
bution of income; the former by the distribution of empirical in nature, require judgment, and
their net benefits amongst contemporaries at a point hence there can be disputes as to whether or
in time, and the latter by the distribution of net ben-
efits as between generations, over a period of time. not the simplifying assumptions are "realis-
tic" or relevant or both.
4The second-best shadow price is that associated
with continuing divergences, the first-best, that with For all the above reasons, even though all
no divergence, between MSV and MSC. the procedures we will consider start from
 661 SHADOW PRICING

the same theoretical foundations, and hence Note that current savings generate the time
are identical if equivalent assumptions are stream of future consumption. As consump-
made, they may nevertheless differ to the ex- tion, following the practice in theoretical wel-
tent that, in practice, they emphasize one set fare economics, is identified as the source of
of assumptions about the economic environ- economic welfare, this means that the net
ment rather than another. Hence, the contin- benefits will be dated consumption, and there
uing charges and countercharges that a par- will be the problem of making commensura-
ticular procedure has ignored or assumed ble present consumption and future con-
away an important aspect of reality, and is sumption. As long as the same relative price
hence invalid; as well as the impression con- between present and future consumption is
veyed to neutral observers of shadow pricing used to "add up" the intertemporal net ben-
on the part of different protagonists, and baf- efits of the project, it does not matter which
flement at the conflicting claims and counter- of the two relevant "commodities" (present
claims that are made for different proce- consumption or future consumption [sav-
dures. This, however, does not imply that in
ings]), we take
We turn first asto our numeraire.6
examine the adjustments
practice certain procedures are not more gen-
eral and easier to apply than others. How- necessary to take account of the distortions in
ever, itmay be more important to begin by the market for capital.
realizing that the similarities amongst the
procedures are far greater than the differ-
encs. . . Capital
Savings and investment are the means for
DISTORTIONS IN FACTOR changing the time shape of the intertemporal
MARKETS consumption stream which is feasible given
resource and technological transformation
In this part we relax the assumption made constraints. The capital market intermedi-
that factor markets are perfectly competitive. ates between those making savings and in-
We now introduce factor market distortions vestment decisions. In a perfect capital mar-
but, for simplicity and clarity, assume away ket, the social return from one unit of current
all other distortions. Thus the domestic mar- savings (the net present value of the con-
ket prices of the two primary factors will no sumption stream made possible by one unit of
longer be taken to equal their social oppor- current savings) at the margin, is equal to the
tunity costs, as we had hitherto assumed. social value of one unit of current consump-
Distortions in both the markets for labor and tion. The former will depend upon the oppor-
capital have been discussed in the project tunities open to society in production, to con-
evaluation literature. The two fully fledged vert one unit of present consumption into
evaluation procedures, the UNIDO and Lit- future consumption — that is, the social pro-
tle-Mirrlees procedures, identify the same ductivity of investment — the latter on the
distortions and, except for some differences weight society places en one unit of future
in assumptions about the likely future consumption in terms of present consump-
changes in some of the divergences, provide tion— the social rates of time preference. In
identical rules in principle, except for a dif- a perfect capital market, the rate of interest
ference in numeraire. The UNIDO proce- R will equal and equate the marginal rate of
dures take current consumption as their nu- transformation (mrt) of present into future
meraire, the Little-Mirrlees current savings.5 consumption, and the marginal rate of indif-
ferent substitution (mrs) of present and fu-
5More precisely the LM numeraire is "uncommit-
ted social income measured at border prices." It need 'The only difference will be the "cosmetic" one that
not be saved, but could be spent on uses (like admin- the appropriate discount rate to be used with savings
istration, health, law and order, et cetera) which are as the numeraire will generally be higher than that to
considered as useful as savings by the government. be used with consumption as the numeraire.
 662 PROJECT APPRAISAL

ture consumption. That is, R = mrt = mrs, Once again the first best solution would be
of the two "commodities" present and future to cure the above divergence, by appropriate
consumption. Distortions in the capital mar- tax-subsidy policy; in this case by using fiscal
ket will drive a wedge between the compo- policy to raise the savings rate in the econ-
nents of the above marginal equivalence, so omy till mrsp and mrt had become equal to
that mrt ^ mrs. Moreover, the rate of inter- mrs. Note that as the savings level is raised
est may not equal either the mrt or mrs. Fur- toward the optimal level, mrs will rise and
thermore, ithe
f capital market is segmented, mrsp and mrt will fall.
there may be a multiplicity of rates of However, the government may have im-
interest. perfect control over savings in the economy,
Two basic sources of distortion have been and may not be able to legislate the optimal
identified as causing the divergence between savings rate by direct fiscal means. In that
the mrt and mrs of present and future con- case, it will be necessary to take account of
sumption. One is due to the presence of ex- the divergence in the mrt and mrs in the cap-
ternalities, the other to the presence of mo- ital market. As long as the divergence exists,
nopolistic orfiscal distortions or both in the current savings are socially more valuable
capital market. We will consider the causes than current consumption. Hence, if the gov-
and adjustments for the former type of dis- ernment can indirectly, through its choice of
tortion, the causes of the latter type being
projects, influence the savings rate, this "sav-
self-evident, and the remedies being similar ings constraint" may be overcome over time
to those suggested for the former type. till savings and consumption are considered
The source of the externality in the capital to be equally socially valuable. The way in
market is due to the interdependence and the which the government could influence the
mortality of private savers. Being mortal, savings-consumption balance of the economy
they cannot be expected to extend their altru- through project choice is by influencing the
ism to the infinite generations which are choice of techniques and by choosing projects
properly the concern of a society, which at whose benefits tend to be saved and rein-
least in principle, is immortal. As a result, the vested rather than consumed.
savings (future consumption) generated, cet- The way in which both the UNIDO and
eris paribus, as a result of the decisions of pri- LM [Little-Mirrlees] procedures take ac-
vate savers is likely to be less than socially count of the divergence is by differentially
optimal. Furthermore, */ private savers knew
weighting
consumed the andproject's net benefits
those which which The
are saved. are
that everyone else was going to save at the
socially optimal rate, then they too would only difference between the procedures in
agree to save at this rate. Hence, the exter- principle is the difference in numeraires.
nality. The result is that the private rate of Whereas (a) the UNIDO procedures use
time preference is higher than the social. present consumption as the numeraire, and
Under laissez faire, a perfect capital market put a premium on savings, (b) the LM pro-
would insure that enough savings would be cedures use current savings as the numeraire,
invested until the social return to investment
and penalize consumption. To see this, con-
(savings) fell to the private rate of time pref- sider the following simple algebraic example.
erence. That is, the private marginal rate of Algebraic Example: First, consider the
substitution (mrsp) in consumption would be procedure adopted in the UNIDO Guide-
equated to the private and ex hypothesi, so- lines, which take consumption as its numer-
cial, marginal rate of transformation {mrt) in aire. Assume that the net benefits from a pro-
production, of present into future consump- ject in any year (t) are Bt, and that of these
tion. However, as the externality causes the 0 percent in any year are saved and rein-
marginal private rate of substitution to be vested and (1—0) percent are consumed.
higher than the marginal social rate of sub- The social rate of time preference today is d0,
stitution (mrs,) we have mrsp = mrt > mrs. and the social return to investment today is
 663 SHADOW PRICING

'"o (ro > ^o)- Moreover, over time, the diver- Confining our time horizon to T, the present
gence between r and d is likely to diminish, value of the stream of consumption made
till T years from today the divergence will possible by Rs 1 of investment today (s0) is
disappear (the level of savings will be opti- then:
mal). Finally, the project incurs capital costs *o = 0) + (2) (3)
of K in the base year, yielding the stream of
net benefits for TV years. The social opportunity costs of capital expen-
To obtain the social profitability of the pro- diture of K today are therefore s0K.
ject we first note that the opportunity cost of The benefits are Z?i in year (f), of which
the capital costs K is the present value of the OB, will be saved and invested. The value of
future consumption which would have re- this saving in terms of consumption at date /
sulted if Rs K of present investment were will be given by the social opportunity cost of
made at the current social rate of return to investment in year t, which on an analogous
investment, r0. Thus, if Rs 1 of current in- argument to that for deriving s0, will be st.
vestment, which is assumed to remain intact The rest of the benefits (1 — 6)B, will be con-
forever, leads to net output of Rs 1.1 in the sumed in year t. Hence, the present value of
next period, r0 = 1.1/1 = . 1 . Part of this re- the stream of net benefits will be given by:
turn (0) will be saved and invested. Hence,
(1 -0)Bt + sfiB,
the increase in investment next year (t = 1) (4)
will be [1 (1 + r0) - 1]0 = r06. Total in-
vestment next year will therefore be (1 + So
and the NPV of the project will be given by:
r06). The year after next (t = 2) investment
will increase by [(1 + ro0)(l + r{) - (1 + NPV = (4) - s0K
(5)
rQ6)]d = (1 + r0d)rl6. Total investment in
year t = 2 will therefore be [(1 + r0d)riy$ Thus on the "generalized" UN I DO
+ 0 + r06)] - (1 + ro0)(l + rxe). Hence, procedures7 it is necessary to know both the
changing social return to investment (the
by year T, when the savings constraint ceases r/s) as well as the changing social rates of
to operate, total investment would have ac-
cumulated to: (1 + r06)(\ + r,0). . . (1 +• discount (the d,'s). Moreover, the discount
rate used to obtain the NPV of the project
rT-X6). To get the present value of this ac-
cumulated investment which, ex hypothesi, is will be the social discount rates </,'s.
The alternative LM procedure takes sav-
as valuable as an equal amount of consump- ings as its numeraire, and uses the own rate
tion at T, we have to discount its value back
of return on investment (called the account-
to the present (/ = 0) at the changing social ing rate of interest, ARI) as the discount
rates of time preference (d0, du d2, • . . dT), rate. However, as we will show, it is identical
period by period. This present value is:
to the "generalized" UN I DO procedures, ex-
(1 + r08)(\ + r,0)...(l + rr_,0)/ cept for the change in numeraire. Following
the same argument as before, and making the
(1 + </,)(l +d2)...(l + dT) (1)
same assumptions, we found that Rs 1 of in-
In addition to this accumulated invest- vestment yielded a present value of total fu-
ment, the initial Rs 1 of investment will have ture consumption generated by the invest-
ment of s0. That is, Rs 1 of current savings
resulted in consumption of (1 — 0)(1 + r06)
in t = 1; of (1 - 0)(1 + r06)(\ + r,0) in t 7"Generalized" because, as the next section ex-
= 2; and hence in year t of (1 — 0)0 + plains, the way in which s is calculated on the
ro0)(l + r,0) ... (1 + r,_,0). The present UN I DO Guidelines assumes that its value remains
value of this stream of consumption is: constant over time. This, in practice, is likely to be an
implausible assumption. But as the above account
suggests, the UN I DO approach can be generalized,
J. (1 - fl)(l + rpfl) . . . (1 + r,_2fl)(l + r»-|g) 12) so that it is identical to the LM one. except for the
^5 (1 + </,)...(! + </f_,)(l + d.) change in numeraire.
 664 PROJECT APPRAISAL

>s0K
(investment) is worth s0 of present consump- Multiplying both sides of (8) by su and then
tion. Consumption therefore has (l/s0) the dividing both sides by 5 V s0, we get:
value if the same resources had been in-
vested. In year 1, therefore, the value of (1 Bx [(1 - 8) + 0sx] (9)
S\/s0 ' (1 + P\)
— 0)(1 + r06) consumption generated is (1
— 0)(1 + r0d)/sx. In year t, the value of the From (7) the denominator of the LHS of (9)
consumption generated from the net benefits
is equal to (1 -f- dx), hence, we get (9) equal
of the project will be B,(\ — B)/st. to:
In each year there will also be yBt savings
generated, and these will be valuable at par, Bx [(1 -0)+ 6s,] s0K (10)
as savings is our numeraire. The total value, (1 4- </,)
in terms of savings of the net benefits in any as the criterion for accepting a project on the
year, will then be:
LM procedure.
Bt{\ - 6)/ st + 6Bt But now consider the same two period case
on the UNIDO procedures; the acceptance
These total savings benefits in each year have
criterion is that the NPV given by (5) be pos-
then to be discounted back to the present at itive; and it can be seen from (4) and (5) that
the accounting rate of interest, {ARI){pt)% in this gives the identical result (10) as the cri-
each period to get the present savings value
of the project. Hence, the NPV of the project terion of acceptability. Hence, the two pro-
cedures LM and UNIDO are identical in
on the LM procedures will be given by: terms of the information needed to take ac-
NPV = Y. count of suboptimal savings. The differences
in the discount rates on the two procedures
Bt(\ - 6)1 st + 6Bt (the ARI on the LM, the social rate of dis-
K count on the UNIDO) merely reflect a
(1 - px) . . . (1 + />,_,)(! + pt) change in numeraire.
(6) LM and UNIDO in Practice: To show the
Note that, as savings is our numeraire, the equivalence in principle of the two proce-
capital costs K incurred in year 0 are valued dures, in the above algebraic example, we
at par. had assumed that the value of s is calculated
Moreover, the LM Manual derives a rela- on the UNIDO procedures on the LM as-
tionship between pt, st, and dt. It is: sumption that savings and consumption will
be equally valuable T years from today. In
*,/*,+, = (1 + p,)/(l + dt) (7) practice, however, the formula given by
Now consider a two period case, that is UNIDO to calculate the value of s assumes
from t — 0 to t — 1. For the project to be that the divergence in the relative social
acceptable on the LM criterion, the NPV value of aggregate consumption and savings,
given by (6) should be positive, that is: and hence the value of s, remains constant till
infinity. Thus the UNIDO formula for cal-
Bx [(1 - 0)(\/sx) + 6] (8) culating is:
s
(1 +Pi)
s = (1 - B)r/(d - 6r\
^hus, if say Rs 1 of investment today leads to a where s — marginal propensity to save
net return of Rs 0.1 tomorrow, of which half (.05) is r = rate of return on investment
saved and invested, and if consumption has no social d = social rate of discount of
value in terms of the numeraire savings, then the ARI
is .05/1 = .05. If on the other hand, consumption and consumption
savings are considered socially equally valuable, then
the ARI would be . 1 / 1 = 0. 1 . In general if the value
This equation will only provide meaningful
of one unit of consumption in terms of savings is \/s, values if d > Or, otherwise the social value of
the ARI in this example will be (.05 + .05/5). investment (s) will be infinite. There is no
 665 SHADOW PRICING

plausible economic reason why d must nec- is the percentage of this return saved; r0 will
essarily be greater than Br.9 Moreover, the thus be the return at accounting prices from
actual value of s given by the formula will be marginal current investments. On certain
very sensitive to the values chosen for d, 6 plausible assumptions,13 this return can be
and r, and small differences in the values of derived as a weighted average of social rates
these variables could lead to large differences of profits on existing investments in the econ-
in the value of s. omy. 0,can be estimated from data on sav-
The assumption of a constant divergence in ings propensities and tax rates.
the relative social values of aggregate con- This leaves d0, the social discount rate, and
sumption and savings, and hence a constant Tthe date when savings and consumption are
s, must therefore be rejected in favor of the expected to be equally valuable, to be
more plausible LM assumption that this di- estimated.
vergence disappears after T years and hence The social discount rates {dt) reflect the
s will typically fall over time to a value of distributional weighting given to income
unity at T.]0 (consumption) transfers between genera-
Finally, it should be emphasized that both tions. Indetermining these weights it is plau-
procedures require information on the social sible to assume that as a result of the normal
rate of discount and the social rate of return processes of growth, future generations will
to investment in the economy. We therefore in any case be richer than present ones. Just
next examine how these parameters can be how much richer will depend upon the ex-
estimated. pected rate of growth in per capita consump-
Estimating the Intertemporal Parame- tion over the future. Suppose the latter rate
ters:11 The LM Manual gives a formula12 re- is g. Further assume that the elasticity of so-
lating the ARI0 (po), and the social discount cial marginal utility (defined as the percent-
rate (d0) to s0 and T. This is: age change in social utility resulting from a
percentage change in consumption) with re-
*o= [1 + *(/><>- 4>)]r (") spect to per capita changes in consumption is
Given the definition of the ARI:
e. Then it can be shown14 that
Po - rQ [6 + (1 - 6)/s0] (12) * - 0 + giY ~ 1 (14)
Hence, substituting (12) into (11) we have: This leaves T, which is rather harder to de-
termine. However, from projections of ex-
s0 = {1 + %[r0 (6 + pected growth rates of national income and
(1 - 6)/s0) - d0] }T (13) savings, it may be possible to arrive at some
This formula succinctly expresses the various estimate of the likely date by which savings
intertemporal parameters we need to esti- are likely to be sufficient to give an adequate
mate. These are r0, d0, T and 0; and given long term growth rate. This date can then be
these s0 will be determined. taken to be T.
The social return to investment is r0 and B
13See Lai [11] for the derivation of such a rate from
a heterogenous capital dual economy model which
'Thus Maurice Scott points out that "one could avoids the capital theoretic problems arising from
argue that d should be zero in Mauritius (with per
derivations based on aggregate production functions.
capita consumption roughly constant) while 5 and r
This method is also similar to that advocated by Har-
are both positive." berger; see his Project Evaluation: Collected Papers,
10Another consequence of assuming a constant s (s, Chicago, 1972. For ah application of the method, see
= s, + \)is that from (7) above the discount rate (p, Lai [10], Appendix 11.
= dt) is the same on both LM and UN I DO
procedures. '"Assuming a constant elasticity social utility func-
tionV)
( which has per capita consumption ( C) as one
UA fuller discussion is contained in Lai [11]. of its arguments, then V = C, and d, = {U\
,2See Little and Mirrlees [14], p. 179, and Lai Ui+l)C, and g = (C,+ , - C,)/C,. Hence d, = (1 +
[10], Appendix 11.
g)'~ 1
 666 PROJECT APPRAISAL

The first term on the RHS is the output

Labor15 foregone elsewhere in the economy, which
One of the most common forms of distor-
has been traditionally identified with the
tion identified in the project evaluation liter- marginal product o labor in agriculture. In
ature isin the labor markets of surplus labor
economies, such that the wage rate does not addition, assuming that workers in both in-
equal the social opportunity cost of labor in dustry and agriculture consume all their in-
comes, the economy will be committed to
the economy.
providing them with extra consumption of (c
Two components have been traditionally
identified in the social opportunity cost of — m) as c > m. This increase in aggregate
labor in surplus labor economies. The first is consumption must be at the expense of ag-
the output foregone elsewhere in the econ- gregate savings given the well-known Keyne-
sian national income identities. But given the
omy, as a result of employing labor on the
project. The second are the costs in terms of non-optimality of savings, this increase in ag-
gregate consumption (decrease in savings)
increased aggregate consumption that may
result as more labor (which consumes most of must represent a social cost. As, ex hypoth-
esi, society values s units of consumption as
its income) is employed on the project. If, due
equal in social value to one unit of savings,
to the nonoptimality of savings (discussed
the net social cost of the increase in con-
above), present consumption is socially less
valuable than current savings, then any in- sumption (in terms of the numeraire, sav-
crease in aggregate consumption, caused by ings) will be
increasing employment as a result of the pro- (c — m) — {c — m)/s
ject, will not be as valuable as the equivalent which are the second and third terms of (15).
amount of savings. This factor will have to be
The above expression reduces to:
reflected in the measure of the social oppor-
tunity cost of labor. SWR = c - (c - m)/s (16)
To concretize this, consider a particular
We now turn to the determination of the out-
formulation of the social opportunity cost of
labor, that is, the shadow wage rate (SWR), put foregone (m), and various other compli-
cations indetermining the SWR, in the fol-
due to LM. Except for a change in numer- lowing sections.
aire, which LM take to be "savings" rather Output Foregone: In most conventional
than consumption, their analysis is similar to
other well-known ones due to Sen and Mar- analyses the output foregone m, in the above
SWR formulation, has been identified with
glin, and which have been incorporated in the
UNIDO procedures. the marginal product of the relevant labor in
its previous employment. While this would,
Assume first that the wage paid to a la-
borer in his new job, c, is above the value of given certain other assumptions to be dis-
the output foregone elsewhere by moving him cussed, be correct for labor which was previ-
from his previous employment, m. Second, ously in wage employment, it may not in gen-
eral be correct for labor which was previously
given the nonoptimality of savings, and tak-
ing savings as the numeraire, one unit of cur- self-employed. This is an important consid-
rent consumption is socially worth (1/s) eration, inview of the fact that in most de-
units of current savings. The s factor is the veloping countries a substantial portion of the
same as in the discussion of capital in the labor force is self-employed.
Moreover, in most conventional analyses it
previous section. Then, the costs of employ- was also assumed that the marginal product
ing one more person in the economy (in terms of the laborer withdrawn from the traditional
of savings) are given by:
sector, agriculture, would be zero and hence
SWR = m + (c - m) m = 0. In a definitive analysis of dualism
-(c-m)/s (15) and surplus labor within a model of family
15This section is based on Lai [9], where a fuller farms on which there is equal work and in-
treatment of the subject may be found.
come sharing, and which explicitly incorpo-
 667 SHADOW PRICING

rated leisure as an argument in the individual gence between the average (c) and marginal
peasant's utility function, Sen demonstrated (c + Ac) cost of hiring industrial labor. The
that zero marginal productivity was not a extra consumption the economy will be com-
necessary condition for the existence of sur- mitted to will then be given by the difference
plus labor. [See selection III.A.3.] The nec- between the marginal cost of hiring (c +
essary and sufficient conditions are given by Ac) and output foregone, m.
a constant disutility of effort, which implies a Hence:
constant marginal rate of substitution be-
tween income and leisure over the relevant SWR = (c + Ac)
range of hours worked per man, in the tra- - (c + Ac - m)/s (17)
ditional sector. Given this, output in the tra-
ditional sector would not fall with the with- and if the premium placed on savings is very
drawal ofworkers, and hence for them m = high (s — oo), the SWR will be higher than
0, even though the marginal productivity of the market wage (c). Note, however, that if
labor was positive in the traditional sector. there is a constant institutional wage in the
Thus, in general, for a family farm worker industrial sector, then c = 0, and the SWR
withdrawn from a farm without any hired will be given as before by (16).
labor, the change in output will not equal his Rural-Urban Migration: As certain models
marginal product. of the labor market in developing countries
Divergence Between Average and Marginal have emphasized, the impact on net output in
Costs: Certain writers have noted that the the economy cannot be deduced from the im-
conventional analysis may understate the pact effects on output in the sector from
extra consumption cost of industrial labor. which the new worker may be withdrawn.
This is due to the assumption made in these Hence, to obtain the value of m, it will be
analyses that "agricultural" workers can be necessary to trace through all the indirect ef-
hired by the "industrial" sector at a constant fects, in terms of the rural-urban migration
real wage (Wi), which is either given by a that may ensue, as the result of creating one
constant institutional wage, or else by a con- new job in, say, the industrial sector. Thus,
stant supply price of labor to the industrial if, for instance, we have, say, a laborer A
sector. Dixit [1] suggests that this assump- moving to the project and his wage in his pre-
tion may be unrealistic, especially if there are vious employment was $>v, and on his moving
terms of trade effects following a withdrawal to the project his previous job is filled by
of labor from agriculture. Then, if the indus- someone else, B, who in turn moves from a
trial labor market is competitive, the supply job which paid him $Y (w > Y), then the
price of labor to industry and hence the in- change in output (assuming that the two
dustrial wage will rise with increased indus- wage rates are determined in competitive
trial employment.16 This will create a diver- markets for hired labor) by employing A on
the project is not $w but $Y, as now the first
16In the simple closed economy two-sector model
round effect of A's migration — a fall of out-
analyzed by Dixit [1], the supply price of industrial put of $vv in his previous employment — is off-
labor is equal to the income foregone by agricultural set by a rise in output in his previous activity
family workers moving to industrial jobs. In short run by an equivalent amount when the other
equilibrium their income foregone is determined by
the average physical product of labor in agriculture
worker B replaced him, but which now re-
(assuming equal income sharing among family farm sults in the loss of output as a result of Z?'s
workers) and the relative price of agriculture output. movement from his initial job to A's previous
With the withdrawal of an agricultural worker, the
average product of labor in agriculture rises, while
jobFurthermore,
of $Y. as a result of creating one
total agricultural output (assuming no surplus labor)
falls. This last factor leads to a rise in the relative more job in the "industrial" sector, more
price of agricultural output. The net effect is to raise than one migrant may move from rural areas.
the average value product of labor in agriculture and If N people migrate, and the change in agri-
hence the supply price of labor to the industrial sector.
cultural output as a result of one person's mi-
 668 PROJECT APPRAISAL

gration is Y, then the migrant on Harris-Todaro assumptions is the

marginal product a, we have:
SWR m c-(c- NY) Is (18)
m = NY = (\/P)a = c
Harberger [5] has used one particular
model of rural-urban migration due to Har- Substitution in expressions (16) or (18)
ris-Todaro [7], to derive the SWR as always yields SWR = c, the industrial wage. How-
equal to the market wage c. This is obtained ever, as has been pointed out above, in gen-
as follows. Harris-Todaro assume that there eral itcannot be assumed that the change in
is no surplus labor in agriculture. The mi- output in the agricultural sector will equal
grants come to the cities because the ex- the marginal product of labor. In that case,
pected income in the urban sector is just the change in output m within the Harris-To-
equal to the income they forego in agricul- daro migration model will be given by Y/P,
ture. The expected urban income is deter- where Y is the change in output in agricul-
mined bythe probability {P) of finding urban ture when one worker is withdrawn. As be-
employment at the industrial wage c, which, fore we have a = Pc (where a is the income
in the Harris-Todaro model, is determined by the worker received in agriculture, which on
the equilibrium ratio of employed to the total a family farm would be equal to the average
labor force in the cities, say P.xl Further- product of the farm if we assume equal in-
more, itis assumed that agricultural workers come and work sharing on family farms).
receive their marginal product (say a). At Hence m = Y/P = Y. Then c/a and the
the margin, therefore, migrant workers will SWR given a Harris-Todaro-type migration
equate their marginal product (incomes) in function will be:
agriculture, a, to the expected wage in towns, SWR = c - c(\ - Y/a)/s
Pc (that is, a = Pc). When one more man is
hired by the industrial sector the expected From this it is obvious that, on the special
wage Pc rises as P rises. This induces rural- Harris-Todaro assumption that Y = a, the
urban migration of \/P workers, which re- SWR will equal c, the industrial wage. This
stores the probability of finding a job in the is the Harberger derivation of the SWR, in
urban sector to P, and the expected income his "Panama" example (see [6], p. 568 and
to the equilibrium level Pc — when rural- following). More important, however, the
urban migration ceases. Hence in expression Harris-Todaro-Harberger migration model is
also restrictive in many other respects, some
(18) N = \/P. Moreover, the "equilibrium"
value of P is a/c (given the migration func- of which are more serious than others. First,
tion a = Pc), and as the output foregone per it implicitly assumes that industrial wage
earners have tenure, as the rate of labor turn-
over in industry does not figure in their de-
l7See Harris and Todaro [7], p. 128. This is also the termination ofP. Empirically, this assump-
assumption made by Harberger [5], p. 570. This for- tion may not be too inaccurate, as the rate of
mulation ofP is unrealistic. A more likely determi- labor turnover does not seem to be very high
nant of the chances of a single migrant is given by the
number of vacancies occurring per unit of time di- in the industrial sector in most developing
vided by the number of candidates for those vacan- countries. Second, they consider the migra-
cies, that is, the urban unemployed. The latter in fact
was the determinant of P in the earlier Todaro for-
tion decision as a one-period decision,
mulation (see [18], p. 142). But note that while these whereas strictly it should be a multi-period
decision in which the present value of the
differing determinants of P will affect the "equilib-
rium urban unemployment" rate (which is the chief costs of migration should at the margin be
concern of other writers on rural-urban migration), equal to the present value of the benefits from
the "equilibrium" value of P will be invariant to these migration.18 If, however, as seems likely,
alternative formulations of its determinants, as it will
be determined by the rural-urban income differential. most migrants have a fairly high subjective
(In our formulation above, the equilibrium P = a/
c). For a fuller discussion see Lai [9]. 18Todaro [18], p. 143, fn. 10, notes this.
 669 SHADOW PRICING

rate of time preference (fairly short time ho- This yields the "equilibrium" value of P =
rizon), then the use of a single period migra- (a + d + u — w)/(c — w).
tion decision function may not be invalid. As before, with the creation of an extra in-
Third, Harris-Todaro do not incorporate any dustrial sector job N = \/P migrants will
of the costs of migration (real and/or move from agriculture, and as the output
"psychic")19 nor the relatively higher costs of foregone per migrant in agriculture is Y, we
urban living which the migrant would have to have the total output foregone, m - Y • (c
incur in their migration function. Finally, — w)/(a + d + u — w), and the
and most important, their migration model
fails to take account of the existence of a SWR m c - [c - Y • (c - w)/(a
+ d + u - w)]/s (19)
fairly competitive "unorganized" (services
and small industry) sector urban labor mar- and in this more general and more realistic
ket with high labor turnover and easy entry migration model, the conclusion drawn by
for new workers, which is typical of many de- Harberger that the institutionally given in-
veloping countries, and which provides some dustrial wage c, is the shadow wage, will not
income to the migrants while they are search- be valid.
ing for an "organized" (industrial) sector job Disutility of Effort: Finally, in addition to
at the high institutional wage c. changes in output, there will also be changes
Thus it is essentially the last two features in the aggregate disutility of effort (£) with
which need to be incorporated into a more increased employment. To evaluate these, as-
general migration function. To derive the sume initially that there are no imperfections
SWR for this more general migration model, in the labor market. Then, at the margin,
we continue to assume that industrial wage utility maximizing workers will equate the
earners can be taken to have tenure, and that disutility of increased effort with the utility
a one-period decision model is a fair approx- from the increased incomes (which we as-
imation to reality. However, we now assume sume are all consumed) this extra work
that in addition to the agricultural income makes possible. That is, the extra disutility of
foregone, a, the migrant has to incur migra- effort (E) must equal the change in workers'
tion costs of d, which include both the real consumption (including those left behind on
and "psychic" costs of migrating. Further- the farm) which is given by (c — m) — the
more, ifthe migrant does not succeed in ob- difference between the industrial wage (as-
taining an industrial sector job at the high in- suming the new job is in industry, and the
stitutional wage of c, he can nevertheless find worker moves to it from agriculture) and the
some labor employment in derive
the "unorganized" total output foregone by employing one more
urban market and an income w. man in the industrial sector. The value in
Finally, we assume that by living in the town terms of savings of this change in disutility of
the migrant has to incur a relatively higher effort (which so far is in terms of consump-
cost of living than in rural areas of u to main- tion equivalents) is (c — m)/s. If the value
tain the same standard of living as he enjoyed society places on the disutility of effort is X,
in the countryside. If the chances of getting then the SWR incorporating the costs of the
an "organized" (industrial) sector job are as disutility of effort will be:
before P, then at the margin the migrant will
equate the costs of migration, which are SWR = c - (c - m)/s + \(c - m)/s
= c - (c - m)(\ - \)/s (20)
given by (a + d + u) with the expected ben-
efits, [Pc + (1 - P)w]\ that is, in Next relax the assumption that all labor
equilibrium: markets are competitive, and assume that
there is an institutional wage, c, in the sector
a + d + u = Pc + (1 - P)w to which the labor is moving which is above
19Though Harris-Todaro note the existence of these the supply price of labor L. The latter term
costs, see their [7], p. 129, fn. 8. includes all the private disutilities that may
 670 PROJECT APPRAISAL

attach to the new job. Our earlier expression this to be the case, either c = m = L, or m
for the consumption equivalent of the net = 0, s -* oo, X = 0, or E = 0.
change in disutility (c — m) will now be ov- Second, there is the view associated with
erstating the true change in disutility by (c Kahn [8] and Lewis [12] that the SWR =
— L), which is the difference between the in- 0. For this to be the case: m = 0, s = 1, X
stitutional wage c, and the supply price of - 0, or E = 0.
labor L. The net change in disutilities in this = Third, for Sen [17], Marglin [15] and to
0.
more general case will therefore be given by some extent UNIDO [19], the SWR = c -
(c — m) — (c — L) = (L — m), and as c/s. For this to be valid: m = 0, X = 0, or E
before the value in terms of savings will be
(L — m)l s, and the Fourth, the LM [ 1 4] S WR is given by (1 6)
above, SWR = c — (c — m)/s. For this to
SWR = c - (c - m)/s be valid: X = 0. As they assume a positive
+ X(I - m)/s (21) marginal product in agriculture, E cannot be
If X = 0, that is, society places no value on zero.
the change in the private disutilities of effort, Finally, for Harberger [6], the SWR is the
we get the traditional SWR as in (16) above. supply price of labor L; that is, SWR = L.
If, however, it is assumed that society should For this to be valid: either X = 1, c = L, or
5=1.
value disutilities of effort at their private
costs, then X = 1, and the Part of the differences relate to empirical
matters, that is, the value of m and E. But,
SWR = L + (c - L)(\ - 1/5) (22)
in part, the differences relate to two value pa-
The first term is the supply price of labor, the rameters, sand X. The reasons why it may be
second is the value in terms of savings, of the necessary to take 5 > 1 have been given in
extra consumption generated by the excess of the section on capital above. A number of
the institutional wage over the supply price of reasons have been advanced by the present
labor. Thus, when X = 1, we get the standard author, why it may also be desirable to as-
neoclassical result, that the SWR will be the
sume that X = 0, for developing countries.21
supply price of labor, if there is no divergence However, the values assigned to these param-
between the social value of present consump- eters must be in the nature of value judg-
tion and savings, that is 5 = 1; and further- ments, and hence the possibility of conflicting
more, that if c = L, that is, if labor markets advice on the different procedures. However,
are competitive, the SWR will equal the mar- as this section has tried to show, if the same
ket wage c, no matter what the value of s, assumptions and value judgments are made,
and irrespective of any divergence between m the alternative procedures will give identical
(the output foregone elsewhere in the econ- answers, based upon the general expression
omy) and the industrial wage c. for the SWR provided by (21) above.
Alternative Formulations of the SWR: We
can now, very succinctly, compare the var-
References
ious alternative SWR's that have been sug-
gested inthe literature.20 [1] A. K. Dixit, "Short-Run Equilibrium
First, there is the view due to Galenson- and Shadow Prices in the Dual Economy."
Leibenstein [4] and Dobb [3] that the SWR Oxford Economic Papers 23 (1971):384-
is the market wage; that is, SWR = c. For 400.
[2] R. Dorfman, P. A. Samuelson, and R.
20As most writers have not included rural-urban M. Solow, Linear Programming and Eco-
migration in their models for determining the SWR,
nomic Analysis, New York: McGraw Hill
this aspect is neglected in this section. The previous
(Rand Series), 1958.
section has already dealt with the SWR's derived or 21SeeLal [9].
derivable from the Harris-Todaro-Harberger-type
models and a model developed in Lai [9].
 671 SHADOW PRICING

[3] M. H. Dobb, An Essay on Economic with Unlimited Supplies of Labour." Man-

Growth and Planning. London: Routledge chester School of Economics and Social
& Paul, 1960. Sciences 22 (\954)\ 139-91.
[4] W. Galenson and H. Leibenstein, "In- [13] I. M. D. Little and J. A. Mirrlees,
vestment Criteria, Productivity and Eco- Manual of Industrial Project Analysis in
nomic Development." Quarterly Journal Developing Countries, Volume II: Social
of Economics LXIX (1955): 343-70. Cost-Benefit Analysis. Paris: Organization
[5] A. C. Harberger, "Survey of Literature for Economic Cooperation and Develop-
on Cost-Benefit Analysis for Industrial ment, 1968.
Project Evaluation" in Evaluation of In- [14] I. M. D. Little and J. A. Mirrlees, "A
dustrial Projects. New York: United Na- Reply to Some Criticisms of the OECD
tions (UNIDO), 1968. Manual." Bulletin of Oxford University
[6] A. C. Harberger, "The Social Opportu- Institute of Economics and Statistics 34
nity Cost of Labour." International La- (1972): 153-68.
bour Review 103 (1971): 559-79. [15] S. A. Marglin, Public Investment Cri-
[7] J. R. Harris and M. P. Todaro, "Migra- teria, Cambridge, Mass.: MIT Press, 1967.
tion, Unemployment and Development: A [16] A. K. Sen, "Peasants and Dualism
Two-Sector Analysis." American Eco- With and Without Surplus Labor." Jour-
nomic Review LX (1970): 126-42. nal of Political Economy LXXIV (1966):
[8] A. K. Kahn, "Investment Criteria in De- 425-50.
velopment Programs." Quarterly Journal [17] A. K. Sen, Choice of Techniques, An
of Economics LXV (1951): 38-61. Aspect of the Theory of Planned Economic
[9] D. Lai, "Disutility of Effort, Migration Development. 3rd ed. Oxford: B. Black-
and the Shadow Wage Rate." Oxford Eco- well, 1968.
nomic Papers 25 (1973): 112-26. [18] M. P. Todaro, "A Model of Labor Mi-
[10] D. Lai, Wells and Welfare. Paris: Or- gration and Urban Development in Less
ganization for Economic Cooperation and
Developed Countries." American Eco-
Development, 1972. nomic Review LIX (1969): 138-48.
[11] D. Lai, "On Estimating Certain Inter- [19] United Nations Industrial Develop-
temporal Parameters for Project Analy- ment Organization [P. Dasgupta, A. Sen,
sis." IBRD mimeograph. Washington, and S. Marglin] , Guidelines for Project
D.C., 1973. Evaluation. New York: United Nations,
[12] W. A. Lewis, "Economic Development 1972.

X.C.2. Derivation of Shadow

Prices*

In many economies market prices do not re- nalities, foreign exchange and capital scarci-
flect the real scarcity values of the goods and ties, unemployment and underemployment,
services that are being produced. Monopo- government fixation of prices and wages, in-
lies, decreasing cost industries, taxes, exter- flation, and so on, are common in most of the
developed as well as the developing countries.
*From F. L. C. H. Helmers, "Cost-Benefit Analy-
Is it at all possible, in view of the many dif-
ficulties, todetermine all the relevant shadow
sis," paper prepared for IMF Seminar, June 9, 1980,
Washington, D.C. (processed). Reprinted by prices? It has been suggested that they
permission. should be derived from general programming
 672 PROJECT APPRAISAL

models. Such models can generate Lagran- The opportunity cost approach is necessar-
gean multipliers that represent in economic ily a detailed approach. The project analyst
terms the shadow prices, which, given the must investigate from where the resources for
constraints incorporated in the model, will re- a project will be withdrawn and what their
sult in an allocation of resources that will sat- values are in those uses. As long as resources
isfy the postulated objective function. Such with low valued uses can be transferred to
models should, in principle, be able to pro- higher valued uses, the change is beneficial.
duce the real scarcity prices of the goods and This, in a nutshell, is what the theory of pro-
services. However, although the models can ject planning is all about.
provide valuable insights concerning the Without any claim for comprehensiveness,
structural relations that exist in an economy, we may now illustrate how the shadow prices
we seriously question on practical grounds can be found in practice. When a project
whether all the shadow prices obtained from needs a certain input, then it is likely that to
the models can be used for operational work. some extent demand elsewhere will be cur-
First, the models are still highly aggregated tailed as well as that some additional quan-
so that most of the duals they generate — the tities will be produced to satisfy the addi-
shadow prices — are extremely crude. Sec- tional demand generated by the project. Thus
ond, itmust seriously be doubted whether the the opportunity cost of the input consists of
models can really depict the real world situ- the weighted average of the value foregone
ation. Itis not only that there are many dis- in the alternative use and the resource cost of
tortions but also, to formulate the models, all the additional production, the weights being
cost and demand functions should be known. the fractions of demand displaced and supply
Obviously it is impossible to collect all these induced to additional demand. What will
data. We feel, therefore, that in addition to happen in practice depends on the shape of
the models approach a more practical ap- the demand and cost functions and theoreti-
proach must be followed. cally the analyst needs therefore to investi-
In fact, this approach — the opportunity gate these functions in detail. In real life,
cost doctrine — has existed for a long time. however, it often suffices to make some rough
Consider an economy where only two estimates as the project may not be sensitive
goods — X and Y are produced. Then the cal- to the value of the input in question.
culation ofthe costs of an output expansion Foreign exchange is for many countries
would not pose a difficult problem. For in- such a scarce resource that it can be treated
stance, ifthe production of X is to be in- as a separate production factor. To determine
creased, the cost of producing the additional the shadow price of foreign exchange, we
quantity of X is to be found by measuring the apply the same analysis as for an input. We
value of the Y goods that the community will should thus consider how an additional dollar
have to give up in order to increase the pro- of foreign exchange can be obtained and then
duction ofX. Analysis of the production and determine the real resource values in the
demand functions for goods X and Y should foregone uses. In principle, there are two
readily provide the required data. The situa- ways to obtain foreign exchange: curtail im-
tion is more complicated when a multiplicity ports or increase exports.
of goods is being produced, since it will ob- In case of a curtailment of imports, the
viously be impossible to analyze the produc- shadow price of foreign exchange can be
tion functions and demand curves of all the found by comparing the domestic value of the
different goods. The opportunity cost doc- import with its c.i.f. value converted at the
trine therefore takes as a starting point the official exchange rate. Hence, if there are no
inputs which A' uses rather than the displaced quantitative restrictions, this value will be
Y goods and defines the costs of these inputs higher than the official exchange value by the
as the returns that they would earn in the amount of the import duties. Thus if the of-
next best alternative elsewhere. ficial exchange rate is U.S. $1.00 = Rs 2.00
 673 SHADOW PRICING

and the import duties are 30 percent, the do- age of skilled labor so that market wages can
mestic value of U.S. $1.00 of foreign ex- often be taken to represent the real scarcity
change isRs 2.60. values of this production factor. In several
If exports are to be increased, then the re- developing countries, however, unskilled
sources otherwise used to produce homegoods labor is available in abundant supply. The de-
will be used to produce export goods. Hence, termination ofits shadow price is distinct
the shadow price of foreign exchange consists from all other production factors because of
then of the resource value of the foregone the fact that the laborer's services are tied to
homegoods. If there are no restrictions on ex- the laborer. This means that if a laborer is
ports, the level of the export duties or subsi- withdrawn from existing employment to
dies will provide us with a good indicator of work in a new job, then not only the foregone
the domestic value. For instance, if export product of the laborer but also the disutility
duties average 10 percent of f.o.b. value of Rs of his extra effort should be taken into ac-
0.20 per U.S. $1.00 exported, then this count to determine the shadow price. Thus,
means that exports are Rs 0.20 more expen- even if the foregone product of a hired un-
sive than the homegoods that could be pro- skilled worker is negligible, the shadow price
duced with the same resources. Hence, the must be set at the price which will induce
domestic value of the exports and the shadow him to work in the new job.
price of foreign exchange would then be Rs In many cities in the developing countries,
1.80 = U.S. $1.00. we see that the influx of rural workers is
As in all cases where a resource is used, greater than the job openings in the formal
there are thus two shadow prices correspond- well-paid sector. In such cases, more than one
ing to whether the resource comes from a dis- person migrates to the cities when one person
placed use or is additionally produced. As in is hired. What this means is that in order to
reality there will be some combination of the find the shadow price of labor, we must add
two possibilities, the correct shadow price to the opportunity cost of the hired laborer
will be a weighted average of the two shadow the opportunity cost of the workers who have
prices. Now, with respect to foreign exchange migrated, but who cannot find a job in the
it is, of course, extremely difficult to deter- formal sector. The shadow price of labor in
mine by how much exports will be increased the cities may thus be a multiple of the rural
and imports reduced if additional foreign ex- shadow price of labor. It is, of course, very
change isneeded. A short-cut method is to difficult to determine precisely what the
work with a normal average of the two exact migration function is as often no relia-
shadow prices and, in our example, this ble data will be available. The best one can
would be the average of 2.60 and 1.80, so do in such cases is to assume that the shadow
that the base estimate can be set at Rs 2.20. price of urban labor lies somewhere between
Further refinements can be made by consid- the institutional and the rural wage rate ad-
ering whether one or the other possibility of justed for cost-of-living differences. The use
earning foreign exchange is the more likely of both values will show whether the project
one. For instance, if in our country import is sensitive to the shadow wage rate and in
curtailment is easier than increasing exports, cases where the shadow rate appears to be
then the foreign exchange shadow rate can be important, a range of rates of return can then
estimated at somewhere between Rs 2.20 and be calculated to help the decision-making
Rs 2.60. On the other hand, if it is easier to
increase exports, then the shadow rate could process.
Finally, a few remarks about the shadow
be set at somewhere between Rs 1.80 and Rs price of capital. As a general principle, the
2.20. Whatever value one chooses, in all cases rate of return of the project should be higher
sensitivity tests should be applied. than the rate of return which the capital re-
As regards the shadow price of labor, sources would have in their next best
many developing countries have still a short- alternative.
 674 PROJECT APPRAISAL

Let us assume that the resources consist of to determine whether the project is worth un-
displaced investment. Then there is the prob- dertaking. Two criteria have become com-
lem that the earnings of many enterprises are mon: the net present value and the internal
subject to corporation taxes and that some of rate of return criterion.
these earnings include rewards for risk tak- Under the net present value criterion all
ing. Furthermore, the dividends and interest the costs and benefits of the project are dis-
payments accruing to the investors will often counted at the opportunity cost of capital to
be subject to income taxes. Should the present values, and the project is considered
shadow price of capital be calculated gross or worthwhile if it has a positive net present
net of such taxes and risk premiums? From value. The criterion amounts thus to the cal-
the point of view of society as a whole, it is culation ofthe present value of the surplus
the foregone yield that is the relevant one. the project generates over and above the op-
Hence, the correct shadow price should be portunity cost of capital.
the rate of return gross of taxes and risk pre- The internal rate of return criterion con-
miums. That different parts of that rate of re- sists of calculating by trial and error the dis-
turn accrue to the Government is not impor- count rate at which a project has a net pres-
tant since these are mere transfer payments. ent value of zero — the project's internal rate
But what if part of all of the capital resources of return — and accepting the project only if
for the proposed project come from displaced its internal rate of return is larger than the
consumption rather than displaced invest- opportunity cost of capital.
ment? Ifthe income distribution in the coun- Is the internal rate of return criterion a
try is not optimal, then there may indeed be theoretically correct criterion? The answer is
a case to evaluate this consumption compo- negative, and the criterion may lead to wrong
nent differently from the displaced invest- results if projects are to be ranked by prior-
ment component. ... In case, however, that ity. The reason for this is that the internal
the income distribution is considered roughly rate of return criterion assumes incorrectly
optimal, consumption may be considered as that the benefits of the project under scrutiny
valuable as investment, and the shadow price will be reinvested at the internal rate of re-
of capital consists then of the weighted av- turn instead of at the opportunity cost of cap-
erage rate of return on capital in the econ- ital as assumed under the net present value
omy. In this respect it is perhaps of interest criterion. It is intuitively clear that the two
to mention that Harberger, after detailed in- criteria will therefore result in a different
vestigation, arrived at rates of return on cap- ranking of projects. If budgetary constraints
ital in India ranging from 17.2 to 21.3 per- play a role, so that a choice must be made out
cent. Lai made a similar investigation of the of a series of projects, then the only correct
opportunity cost of capital in India and found criterion is the net present value criterion.
rates of return ranging from 12 percent to 19 The way to proceed then is to calculate for
percent. It needs no elaborations that in a each project the present value of the surplus
country with an abundance of capital re- it generates per dollar of current investments,
sources, such as for instance in Saudi Arabia, and to choose those projects that have the
the shadow price of capital will be much highest surplus ratios.
lower.
The problem with the net present value cri-
terion is,however, that the opportunity cost
of capital, which serves as the discount rate,
is difficult to estimate. One may attempt to
THE CALCULATION TECHNIQUES
calculate net present values by using a mini-
After having imputed the correct scarcity mum and a maximum value for the oppor-
values to all the inputs and outputs of a pro- tunity cost of capital, but where one can only
ject, including negative values for harmful ef- estimate that the opportunity cost of capital
fects on the environment, the task at hand is lies between some very extreme values, this
 675 SHADOW PRICING

procedure is not very meaningful. In such a of acceptability — in order to sharpen the ac-
case, the internal rate of return method may cept-reject decision. This may be done for in-
be used to establish a tentative ranking order stance by calculating probability distribu-
of projects. As projects with high internal tions of the costs and benefits. We feel,
rates of return will be accepted in any case, therefore, that the internal rate of return is
one can then spend extra time on projects an important tool in practical work despite its
with low rates of return — those at the margin theoretical limitations.

X.C.3. Guide to Little-Mirrlees*

A FIRST APPROXIMATION
using shadow prices, to test for a project's net
present worth.
The Manual's central message can be sim- Thus the basic approach of the Manual is
ply stated.1 A valid assessment of an indus- not new. For those to whom the basic con-
trial project's worth to an economy often re- cepts may be hazy or unfamiliar, the Manual
quires the use of values that differ from the
values used in the normal kind of business or provides an admirable introductory explana-
tion, free of mathematics or jargon. But the
financial appraisals. Economic appraisals fre- major innovation put forward by Little and
quently must be done with "shadow" prices, Mirrlees — their big idea that has stirred up
i.e., any assumed prices that differ from those
most of the interest and most of the dust — is
that will actually be realized in the compa- their notion that all prices used in project
ny's own books ("market" prices). The par- calculation should be world prices. They are
ticular shadow prices which Little and
not satisfied with the procedure followed by
Mirrlees believe should be used are "world many institutions (including, normally, the
prices"; these represent a country's actual World Bank Group) whereby partial use is
trading opportunities. The resulting streams made of international values, e.g., applying
of annual costs and benefits should then be
weighted (discounted) to reflect the differing them to major inputs and outputs that are or
times when they will occur. The discounting easily could be traded. They argue in favor of
operation has the effect of translating future valuing all inputs at world prices, even the so-
called nontraded inputs that normally cannot
values into their present worth, allowing the
streams of future costs and benefits to be possibly be imported (i.e., electricity, con-
summed into single figures. You then sub- struction, local transport and labor). The rea-
tract the single cost figure from the benefit son Little and Mirrlees want to go "all the
figure. If there is a surplus the project is said way" in using world prices for every input
to have a favorable present social value, and and output is to avoid the distortions which
they feel creep into the calculations if only
is worth doing. Economists will at once rec-
ognize that Little and Mirrlees are relying on partial use is made of world prices. A cost-
the familiar discounted cash flow method, benefit calculation based partly on world
prices and partly on domestic prices has to be
put into a single currency through the use of
♦From G. B. Baldwin, "A Layman's Guide to Lit-
tle-Mirrlees," Finance & Development, Vol. 9, No. 1, an exchange rate. Little and Mirrlees don't
1972, pp. 16-21. Reprinted by permission. like exchange rates, not even "shadow" ex-
'The Little-Mirrlees manual is Volume II of the change rates
tic official rates. "correcting"
basedIn ontheir unrealis-
view, no exchange
Organisation for Economic Co-operation and Devel-
opment's two-volume Manual of Industrial Project
rate, no matter how good, can overcome the
Analysis in Developing Countries. distortion in relative values which arises
 676 PROJECT APPRAISAL

whenever you combine values taken from dif- SOCIAL PROFITABILITY AND
ferent sets of prices (such as world prices and PRIVATE PROFITABILITY
domestic prices). The use of one set of
Little and Mirrlees begin by pointing out
prices — world prices — bypasses this problem the many similarities between the calculation
and, by taking all prices from one common
pool, achieves a more valid ordering of the of an industrial project's private and its eco-
relative values used in constructing costs and nomic or social profitability. Both calcula-
tions start with the problem of estimating fu-
benefits.2
If by some miracle of wisdom a country ture income (i.e., "benefits") and costs
(capital and operating). The cost and benefit
had developed its economic structure under
estimates are both made up of two elements:
conditions of free trade and free exchange
(1) the number of physical units involved
rates, its investment decisions would contin-
plus (2) the price used to value each physical
uously have been made in the light of world-
wide trading opportunities; consequently, the element. "The essence of a cost-benefit anal-
relative values of home prices and world ysis is that it does not accept that actual re-
ceipts adequately measure social benefits,
prices would today stand in an undistorted
and actual expenditures social costs. But it
relationship to each other. This is hardly the
world as we find it. Protection and other does accept that actual receipts and expen-
ditures can be suitably adjusted so that the
trade restrictions are everywhere, investment difference betweeen them . . . will properly
decisions have often been made outside the
discipline and opportunities of world prices, reflect the social gain" (pp. 22-3).
A second but less crucial feature of their
and many exchange rates now serve to distort
system is the emphasis given to savings. Lit-
rather than to preserve true relationships be- tle and Mirrlees feel that resources for in-
tween domestic and world values. Little and vestment are so often a critical constraint
Mirrlees want to prevent this bad history
that they deserve to occupy a central role in
from contaminating investment decisions.
evaluating projects. Their handling of this
They want to avoid using some good prices
(world prices of traded items) and some not- problem, by converting a project's cash flow
into a pure savings flow stripped of all con-
so-good prices (the domestic prices of non-
traded inputs), which must then be merged sumption elements, is neat, and deserves dis-
cussion. But not until we look further at what
by use of a (frequently bad) exchange rate.
they have to say about the use of world
If all values can be measured in world prices,
the problem is solved. In a neat display of in-
ventiveness, Little and Mirrlees have prices.
WHY WORLD PRICES?
"solved" this problem. But some of their crit-
ics say the cure is worse than the disease. "If you can get more refrigerators by ex-
This point is far and away the most contro- porting bicycles to pay for them, than by di-
versial part of the Manual: just how much verting resources from making bicycles to
trouble is it worth to avoid using a foreign ex- making refrigerators at home, it is clearly
change rate, official or shadow? right to make and export the bicycles and im-
With this summary introduction we can port the refrigerators. But whether this is in
take a closer look at some of the main fea- fact the case, requires a knowledge both of
tures of the Little-Mirrlees approach. the relative costs of production at home, and
of world prices and market conditions" (p.
85). The reason for relating everything to
2The key value of labor is brought into this system
of world prices at its proper relative value by first giv- world prices is not because they are "more
ing it a hypothetical or shadow price in terms of its rational" than domestic prices but simply be-
domestic scarcity and then translating this into its cause "they represent the actual terms on
world-price equivalent. which the country can trade" (p. 92). So it is
 677 SHADOW PRICING

logical to argue that all internationally ships among different values and distort the
traded goods — i.e., those which the country estimate of the relationship between foreign
actually imports or exports, regardless of and domestic resources.
whether the project itself will do so — should
be valued at their c.i.f. (for imports) or f.o.b.
USING INPUT/OUTPUT ANALYSIS
(for exports) prices. But what should we do TO CHASE DOWN TRADED ITEMS
with nontraded inputs and outputs? (The
main nontraded inputs are domestic trans- Little and Mirrlees advance both prag-
portation, construction, electricity, land, and matic and theoretical reasons for wanting to
labor; other minor ones can be thought of, anchor all cost-benefit values in world prices.
e.g., water and waste disposal, telecommuni- Their pragmatic reason is that world prices
cations, advertising, banking services, main- represent actual trading opportunities, which
tenance and repair services.) heavily influence domestic investment deci-
Little and Mirrlees say that since traded sions. Every industrial investment decision
goods are valued in world prices, nontraded involves the "make or buy" decision involved
goods must be similarly valued; "only thus in the refrigerator and bicycle quotation cited
can we ensure that we are valuing everything earlier. The development process involves a
in terms of a common yardstick." Note that steady expansion of the demand for imports,
a "common yardstick" may refer either to and the only way to pay for them, in the long
use of a common currency (e.g., all values ex- run, is to produce for export only those things
pressed ineither dollars or rupees), or to the a country can produce best. To make the
use of a common source of values (e.g., world most of its opportunities a country must de-
prices, instead of a mixture of world prices ploy its resources (fundamentally, its land
and domestic prices). All methods of project and its labor force) in ways which give it the
appraisal require the use of a common cur- most for its money, i.e., the most foreign ex-
rency, but only Little-Mirrlees require the change (either earnings or savings) for do-
use of both a common currency and a com- mestic resources used. And the only way to
mon source for all values used. The problem do this in a complete and theoretically con-
of trying to express all values in a single cur- sistent manner is to chase down all the inputs
rency where they will stand in a right rela- used by a project, direct and indirect, until all
tionship toeach other can be solved either by the potentially tradable items have been val-
using "a special accounting price for foreign ued in world prices, leaving only land and
exchange" (a shadow exchange rate) or by labor. Land is dismissed as relatively unim-
revaluing domestic resources in terms of portant in most industrial projects. But in
world prices; the latter is the method recom- valuing labor, Little and Mirrlees argue that
mended in the Manual. Once all values are
even labor's own inputs (i.e., its consump-
established in terms of world prices, it does tion) should be valued in terms of world
not matter whether they are left in U.S. dol- prices (they give us some help by suggesting
lars or converted into domestic rupees (the how this refinement might be achieved).
two illustrative currencies used in the Man- It is not easy to value the nontraded inputs
ual). If a complete set of costs and benefits in world prices. The general procedure is to
are converted from dollars into rupees any take each such input (power, construction,
exchange rate can be used, since all values transport, or labor) and break it down into its
will retain a constant relationship to each own inputs; these in turn will consist of items
other. But this is not true if some values are that are traded and items that are nontraded.
taken from world prices and some are taken The latter are in turn broken down into
from domestic prices and the two sets are traded and nontraded items, etc. etc. The
then put into a single currency by using an only way this chain can be followed back very
exchange rate. This will change the relation- far is through a detailed input/output table
 678 PROJECT APPRAISAL

for the economy as a whole. Few less devel- stream stretches over many years, there is no
oped countries have in existence a table that reason for using a constant shadow price
is detailed enough to permit the needed cal- throughout the project life if there is any rea-
culations for more than a few simple indus- sonable basis for varying it).
tries. Ifno such table is available, rough-and- But Little and Mirrlees do not let the do-
ready approximations to world values can be mestic shadow price of labor determine its
used. These approximations or "conversion value in project costs. A domestic shadow
factors" are based on the ratio of domestic price reflects only labor's relative scarcity in
costs of a representative sample, of, say, con- the domestic economy; once this has been es-
struction items (wood, cement, steel, fuel, timated, this domestic value must then be
etc.) to the world price of these items. This converted into a world price. In theory this
ratio is based on using the official exchange can be done — as explained above — by de-
rate; the reciprocal of the ratio is then used composing labor's consumption into traded
to adjust the domestic cost of nontraded in- items; but in practice either a specific conver-
puts to values closer to what they would be if sion factor or the standard conversion factor
the complete input/output method had been would almost always be used.
rigorously followed. This conversion factor is
in effect a way of correcting for the overval-
uation of an exchange rate; however, it does ADJUSTING FOR A PROJECT'S
"COMMITMENT TO
it on an average basis and not with the pre-
cision theoretically achievable if all non- CONSUMPTION"
traded inputs could be decomposed into their With nations, as with individuals, there is
ultimate traded elements which could then be a neverending tug-of-war between consump-
valued in world prices. tion and savings. The ultimate purpose of all
economic activity is to raise living standards,
THE PRICE OF LABOR which means raising consumption. But it is
nonconsumption out of present income (sav-
There is nothing particularly new or dis- ings) which provides the resources for the in-
tinctive inthe Little-Mirrlees treatment of vestment necessary to assure higher con-
this much-discussed question. Most of the sumption tomorrow. Hence the battle
labor used in industry comes from a labor- between consumption today and consumption
surplus rural sector, and its departure in- tomorrow. Little-Mirrlees have an unusually
volves little or no loss of agricultural produc- clear discussion of this classical problem, al-
tion, i.e., its opportunity cost (equal to its though their operational advice seems unnec-
marginal productivity) is typically very low. essarily complicated.
However, industrial labor is usually paid con- Obviously a dollar's worth of future con-
siderably more than the agricultural subsis- sumption isnever worth a full dollar to us
tence wage. Some of this excess over what today. Future values always stand at a dis-
labor could earn in agriculture is a necessary count compared to the present, and the more
cost of making it available at industrial lo- distant the future, the greater the discount.
cations; but some of the excess is an artificial The specific rate at which future consump-
creation of trade union policies, needlessly tion is discounted is called the consumption
high minimum wages, or employer "softness" rate of interest (economists often call this the
in relation to what he could pay if he wanted social discount rate, a more ambiguous
to pay no more than a competitive wage. So term). This is not the discount rate Little and
the proper domestic price for valuing labor is Mirrlees use in discounting cost-benefit
a shadow wage that lies somewhere between streams. Why not? Because projects generate
its actual market rate and an agricultural future savings as well as future consumption.
subsistence wage. A good deal of judgment is Indeed, Little and Mirrlees believe that sav-
involved in settling on a figure (since the cost ings are so difficult to generate, and so im-
 679 SHADOW PRICING

portant to future consumption, that the main WHAT DISCOUNT RATE TO USE?
test of a project's economic merit in almost As noted, in the Little-Mirrlees system (as
all less developed countries should be its abil- in many others) the investment test used is a
ity to generate savings. There is no reason
why future savings should carry the same dis- project's present social value — the difference
count rate as future consumption. So it may between the present values of the benefit and
cost streams. These present values are, of
look at first as though future consumption
will have to be discounted at one rate and fu- course, critically dependent on the discount
rate used; different rates can change not only
ture savings at another. To avoid this, Little- the size of the present social value but can
Mirrlees revalue future consumption in terms
make a positive value turn negative. So the
of savings: this gives us a unified benefit
specific ARI is important. The Little-
stream, a "cash flow" stripped of its con- Mirrlees rule for choosing the right discount
sumption elements so that it represents only
rate is the same as that used by many others,
savings. These can then be discounted at a
single rate; the rate at which savings are dis- including the World Bank Group: "... the
accounting rate of interest should be kept as
counted iscalled the "accounting rate of in- high as possible consistent with there being
terest" (ARI). This is the discount rate to use
in as much investment as savings permit"
costcalculating
and benefitthestreams. present worth of a project's
(p. 96). ("Savings" here means domestic plus
foreign savings, the latter being net capital
The social income stream (consumption
inflow.) If a too low ARI is chosen there will
plus savings) to be generated by a project is
stripped of its consumption element by taking be excessive investment, a balance of pay-
ments deficit, and underuse of resources.
the consumption element out of both labor
income and returns to capital. On the labor Recognizing these limits and choosing a rate
that will steer the right course between them
side this adjustment is made automatically
is a matter of good judgment. Little and
by defining labor's cost in terms of its con- Mirrlees think most developing countries
sumption only. Since only this consumption
ought to use a rate around 10 per cent in real
cost is deducted from project income, any- terms, i.e., after inflation; some countries
thing labor saves remains in the net benefit
stream. The elimination of consumption from might use even 15 per cent. Rather than
the returns paid to the owners of capital is worry about the exact correctness of the
ARI, Little and Mirrlees sensibly suggest the
accomplished by applying to the project's es- trial use of three rates — high, low, and me-
timated incremental capital income an esti-
mate of the general marginal propensity to dium— to sort out the "obviously good" and
consume of those who receive interest and the "obviously bad" projects, with marginal
ones to be put off until the authorities see
dividends (for some reason no offsetting al- how large the investment program will be
lowance ismade for any extra government
and whether any clearly better projects come
consumption which may result from tax rev- along to displace the marginal candidates.
enue generated by a project).
Thus the use of a shadow price for labor
increases the benefit stream while the adjust-
ment for consumption out of profits reduces A PROJECT'S BALANCE OF
the stream: one wonders how near the truth PAYMENTS, EMPLOYMENT, AND
FUTURE CONSUMPTION EFFECTS
one would be if neither of these offsetting ad-
justments were made! But at least Little- One of the attractive features claimed for
Mirrlees give us a complete, easy-to-under- the Little-Mirrlees approach is that it pro-
stand lesson in why a "commitment to con- vides acomprehensive project evaluation test.
sumption" is"bad" and how they think it Once it is determined that a project has a
ought to be eliminated from the cost-benefit positive PSV when world prices are used, one
calculation. can be confident that it will fulfil all impor-
 680 PROJECT APPRAISAL

tant economic objectives for which projects Little and Mirrlees do not expect their
are often specifically tested. This applies to a Manual to give individual project analysts
project's impact on the balance of payments, everything they need to go out and make
on employment, and on society's claims for valid cost-benefit studies of industrial proj-
future consumption. By valuing all project in- ects. They acknowledge that they have really
puts and outputs at world prices (i.e., in written a textbook of appraisal theory, not a
terms of their foreign exchange value), "im- how-to-do-it manual for men on the firing
port-substitution andexporting is encouraged line. Furthermore, the textbook is meant for
to the maximum desirable extent." Once the the education and guidance of a small group
authorities have persuaded project appraisers of high-powered economists who, the authors
to use correct values (i.e., world prices) in hope, will be found presiding over develop-
their cost-benefit studies then "the right way ment planning at the center of things in every
to control the balance of payments is to con- country. They urge every country to prepare
centrate on high-yielding projects, and not to a much shorter manual of its own, telling
try to do more investment than saving, tax ministries and development banks how to do
policies, and foreign aid, allow." By valuing economic cost-benefit studies and giving
labor's shadow wage also in terms of foreign them the necessary accounting prices.
exchange (which may put a lower value on The weakest — and certainly the most con-
labor's consumption inputs than domestic tentious— of the various steps recommended
prices do) "producers are encouraged to use by Little-Mirrlees is the extent to which they
labor, instead of imported inputs, to the max- go in using world prices. They want us to use
imum desirable extent." Finally, as we have them for all cost and benefit values, not just
seen, the problem of balancing consumption for important items that are actually traded,
today against consumption tomorrow is han- or could be. To do this involves a lot of trou-
dled by use of a shadow wage rate — e.g., a ble for a doubtful advantage. It is a lot of
low shadow wage encourages labor-intensive trouble because of the need to use input/out-
projects, the main expression of consumption put data that usually do not exist with nearly
today over consumption tomorrow. the accuracy needed to yield accurate re-
sults— and there seems little advantage in
substituting the distortions of bad input/out-
CONTROVERSIAL POINTS put data, or the approximations of conversion
factors, for those arising from overvalued ex-
Toward the end of the Little-Mirrlees vol- change rates. It is not even true that Little
ume one comes upon this self-description: and Mirrlees get rid of exchange rates en-
The methods of project appraisal described in this tirely, since some world prices will be in U.S.
Manual, depending as they do on relatively crude dollars, some in deutsche mark, some in yen,
methods of estimating accounting prices, can be etc., and these can be merged only by using
thought of as a first step in the harnessing of the exchange rates. So part of the exchange rate
whole range of production decisions to social problem is simply pushed outside the coun-
ends. . . . The methods suggested do not depend try. In most countries, it appears far simpler
upon the prior analysis of reliable and sophisti- and sufficiently accurate to use:
cated planning models. They are practicable, and
are likely to be accurate enough to exclude all def- 1. World prices for the actually traded
initely bad projects, and allow all definitely good
ones. Small mistakes on marginal projects are less major capital and current inputs, and for
the outputs;
important (p. 188).
2. Domestic factor costs (at either shadow or
Admirable goals for any appraisal method — market prices as judged appropriate) for
reliability, simplicity, feasibility. At most the nontraded inputs; and then to
points, the Manual meets these tests; at a few 3. Convert these foreign and domestic values
points it fails them. into a single currency by resort to an ex-
 681 SHADOW PRICING

change rate (again, using any reasonable precision to which Little and Mirrlees want
rate if the official rate is felt badly out of to take us.
line). Labor operating costs, which are impor-
tant in most industrial or agricultural pro-
In a majority of industrial projects distor- jects, can be handled satisfactorily by sensi-
tions inthe values of nontraded inputs simply tivity analysis, i.e., by trying two or three
will not be important. Electricity rarely com- different values to see how much difference it
prises more than 4-5 per cent of manufac- makes to the final result. There is no point in
turing costs, so a 20 per cent distortion of its going to a lot of trouble to establish a doubt-
value will affect total costs by only 1 per cent. ful accuracy for values that do not change a
Distortions in internal transport costs for conclusion reached with more easily estab-
capital and operating costs are unlikely to lished, well-reasoned values. Little and
run more than the same order of magnitude. Mirrlees make a good case for using sensitiv-
The construction element in plant capital
ity analysis when discussing the "fuzziness"
costs is larger and may run 15-30 per cent; surrounding the ARI; they might well have
at least half of this will consist of labor costs extended its use to other cost-benefit values.
which, as with operating labor costs, can be I cannot help concluding that this particular
adjusted to an "economic" value through the feature of the Little-Mirrlees methodology —
use of a shadow wage. The distortion arising the world-pricing-of-nontraded-inputs fea-
from using domestic currency to price labor's ture that has caused so much argument — is
shadow wage is likely to be less than the mar- a tempest in a teapot. I doubt it will catch on,
gin of error inherent in deciding what shadow and it will not matter much if it doesn't.
wage to use. Thus, when one looks at the rel- There is plenty of wisdom in the basic Little-
ative unimportance of all the nontraded in- Mirrlees approach without trying to make
puts, except labor, in a majority of industrial everything depend on their controversial
projects, refinements in these values begin to method of valuing a project's nontraded
look relatively unimportant. Cost-benefit inputs.
analysis simply does not work to the order of

X.C.4. Foreign Exchange Rate*

Since the UNIDO method uses domestic cur- change rate.1 This process increases those
rency as the numeraire, the project's foreign economic-efficiency values that were mea-
exchange impact must be identified so that sured in border rupees (border prices in dol-
the project's net present economic value may lars multiplied by the market exchange rate)
be adjusted by an appropriate premium, as-
suming of course that foreign exchange is
'Foreign exchange may, of course, be less valuable
more valuable than indicated by the ex- than indicated by the market exchange rate, a situa-
tion that might prevail in a relatively open econoim
that had just undertaken a major (and essential) de-
valuation. Inthis case, the adjustment factor would
♦Frorn UNIDO, Guide to Practical Project Ap- simply assume a negative value. In practice, however,
praisal: Social Benefit-Cost Analysis in Developing in most less developed countries that are not oil pro-
Countries (John R. Hansen edition), 1978. Reprinted ducers, itis generally safe to assume that there will
by permission. be a premium on foreign exchange.
 682 PROJECT APPRAISAL

by the percentage premium on foreign ex- rately for each input or output category so
change, afactor that roughly indicates the that an adjustment can be made to the net
level of protection in the economy, i.e., the dif- present economic value of the project. . . .
ference between average market and average Whether this adjustment is done on year-by-
border prices. This adjustment makes the year values or on present values as recom-
prices established with reference to border mended above depends on the assumption
prices compatible with prices based on do- made about (a) the shadow price for foreign
mestic consumer willingness to pay in the exchange over time; and {b) the foreign ex-
protected market.2 If the foreign exchange change content over time.
impact is positive, the net present value be- In practice, it may not be unreasonable to
fore adjustment will be increased by the ad- assume that both of these values are con-
justment; conversely, if it is negative, the net stant. The first assumption is widely made;
present value will be reduced. few writings recommend that this value be
In principle, all inputs and outputs are changed over time, for there is usually little
either tradables that can be valued directly in evidence to guide such differential shadow
terms of foreign exchange or non-tradables pricing of foreign exchange.3 The second as-
whose inputs can be disaggregated in terms sumption isopen to more challenge, but gen-
of tradables, non-tradables and labour. If the erally isreasonable. Even if there is a reduc-
non-tradables identified in the first round of tion in the direct foreign exchange content of
disaggregation are further disaggregated in a a major non-tradable input (e.g., electricity
similar manner, and so on, theoretically any now produced from oil that later will be re-
non-tradable can be valued in terms of its placed by indigenous hydroelectric power),
foreign exchange, domestic labour and capi- the hydro facility will almost certainly in-
tal content. Thus, it is impossible to say that volve some direct or indirect foreign ex-
the project analyst should count only the di- change expenditures. The assumption of con-
rect foreign exchange impact of the project in stant foreign exchange content for any
this part of the UNIDO analysis; the line be- specific input or output is especially reason-
tween direct and indirect is too arbitrary to able in agricultural and industrial projects,
be of any help to the analyst. where, aside from labour, the inputs and out-
In practice, the foreign exchange premium puts are generally tradable.
needs to be applied only to those goods that Without these assumptions, it is necessary
were valued at border prices, since inputs and to calculate the foreign exchange impact of
outputs shadow priced with reference to do-
mestic consumer willingness to pay or cost of
3If there is strong evidence, however, that a given
production already implicitly include a pre- country is moving purposefully towards a free-trade
mium on foreign exchange. Goods shadow policy, the precision of the analysis could (perhaps
usefully) be increased by estimating the years until
priced "at the border" would generally in- free trade is reached, then reducing the premium on
clude (a) all major inputs and outputs; and
foreign exchange each year by the present premium
(b) any major non-traded inputs with a sub- divided by the number of years to free trade. Con-
stantial foreign exchange component (e.g., versely, ifa country has recently undergone a sub-
electricity produced from imported oil). stantial devaluation so that there is currently little or
Once an appropriate decision has been no economic premium over the official rate for foreign
made about the inputs and outputs for which exchange, but if it is expected that fundamental prob-
lems will make another devaluation necessary in the
the foreign exchange impact will be evalu- near future (not an uncommon situation), it may be
ated, these impacts must be quantified sepa- reasonable to apply a gradually rising premium. More
refined non-linear estimations would probably be
2In the Little-Mirrlees system, a reverse but equiv- pointless given the uncertainties surrounding inter-
alent approach is used. Instead of increasing border national trade and trade policies. In practice, it may
prices to the domestic-market level with the premium be sufficient to set up a limited number of periods and
on foreign exchange, domestic prices are decreased to hold the premium consistent during each period,
border-price levels by the standard conversion factor. which would simplify the calculations.
 683 SHADOW PRICING

each input and output in each year of the pro- mestic market for the good, this approach to
ject, calculate the premium, discount it back valuing foreign exchange implicitly uses the
to the present to obtain a net present eco- domestic willingness to pay a premium above
nomic adjustment value of the foreign ex- the "official" price for the foreign exchange
change impact and add it to the basic net needed to purchase these goods. If it can be
present economic value.4 The use of these as- assumed that the average premium on for-
sumptions greatly simplifies the valuation of eign exchange with an increase in economic
the foreign exchange impact in the UNIDO activity will remain about the same, an ad-
method, for then this impact can be calcu- justment factor based on a formula for aver-
lated on the basis of net present economic age shadow exchange rates will be approxi-
values. The present value of each input and mately equal to one based on a formula for a
output is multiplied by a weighted adjust- marginal shadow exchange rate.
ment factor to obtain its foreign exchange In practice, the approach outlined in the
adjustment value. These adjustment values preceding paragraph should give a usable es-
are then added up . . . and used to modify the timate of the shadow price of foreign ex-
respective net present values. The weighted change. Ifthe data are available, however, a
adjustment factors are the product of the for- more accurate indication of the shadow price
eign exchange content of each item and the of foreign exchange can be obtained. The ap-
adjustment factor for foreign exchange. All proach proposed in the Guidelines is, as in-
the year-by-year computations are thus dicated above, based on the assumption that,
avoided. in the case of a project that is a net user of
The premium to be attached to foreign ex- foreign exchange, the amount it uses may be
change isimportant. . . . The Guidelines uses taken from other users and thus falls under
a foreign exchange shadow price based on the shadow pricing category of "consumer
marginal social value as revealed by the con- willingness to pay."6 However, the use of for-
sumer willingness to pay for it in the form of eign exchange by the project may also en-
imported goods. Without going into details, courage the "production" of foreign ex-
the derivation is based on the average of the change through manufacture for export or
percentage by which the domestic market import substitution, in which case a some-
price of the last unit of each good imported what different approach is required. This
by the country exceeds its c.i.f. price ex- "production" will be stimulated if demand
pressed indomestic currency weighted by the for foreign exchange by the project changes
share of each good in the actual marginal im- the incentive to other potential producers of
port bill.5 Assuming a freely competitive do- export or import substitutes. The incentive to
undertake foreign exchange earnings or sav-
4If the year-by-year adjustments are to be made, it
would be advisable to set up the financial income ing activities depends on the "effective price"
statement and/or financial cash-flow table for the of foreign exchange, the "rupee" benefit re-
commercial-profitability analysis in stage one so that ceived bydomestic producers of goods who,
there is a double column under each year, one for the on balance, earn or save a dollar of foreign
basic price and one for the foreign exchange content
of this price. exchange. For example, if the official ex-
change rate is Rs 10/S1, the c.i.f. price of an
5A simple formula for calculating an average
shadow exchange rate (SER) similar to the UNIDO import-substituting good is $3 and the do-
approach and based on a given year's data is:
tariff equivalents should be included, and if there are
SER = OER ~(M + T,) + (X + Sx) other sources of difference between border and (do-
M + X
mestic ex works prices (e.g., transport costs and im-
where OER = official exchange rate; M = c.i.f. porting profits), these should also be considered. Fur-
value of imports; X = f.o.b. value of exports; T, = thermore, any prohibitive quantitative restrictions or
tariffs should be taken into account as well, if feasible
import tax revenues; and Sx - export subsidies. (Ex-
port taxes should be regarded as negative subsidies.) ^he converse of the following applies to projects
Where quantitative restrictions are important, their that are net producers of foreign exchange.
 684 PROJECT APPRAISAL

mestic manufacturer, because of protection, proach is oriented to the cost of producing

can charge Rs 45, the "effective" rate is Rs foreign exchange by a single project, the
45/$3, or Rs 15/$1. This incentive will re- same principles apply at the macro level.
main unchanged only if the effective foreign However, at the macro level the analysis be-
exchange price remains constant. If the effec- comes more complex in that the elasticities of
tive rate does remain constant, the demand alternative sources of foreign exchange and
for foreign exchange by the project would fall of the demand for imports become important.
on other uses, for there would be no induce- Examination of this aspect of the shadow
ment to the new production of foreign ex- price of foreign exchange8 would be a logical
change, and all would have to share the pres- extension of the UNIDO method, although
ent supply. The effective price is, however, in practice it is unlikely to have a major im-
very unlikely to remain constant, for that pact on the relative ranking of projects.
would require much more than a stable offi- Once the shadow price of foreign exchange
cial exchange rate. It would require that has been determined, the adjustment factor
there be no change in tariffs, quantitative re- for foreign exchange (AFf) should be calcu-
strictions, capital-movement regulations or lated. Itis the premium on foreign exchange
any other protective device, despite increased over the market rate and is equal to the
demand for foreign exchange by the project shadow exchange rate divided by the market
and by other users. Such a situation would exchange rate, minus one. For example, if the
almost never occur, except in the short term, shadow rate vis-a-vis the dollar is Rs 23 and
for it would deplete the country's foreign ex- the market rate is Rs 20, the AFf is 23/20 -
change reserves, which few governments 1, or 15 per cent. The AFf\s weighted by the
could allow for long. Instead, given an in- foreign exchange content of each item, and
creased demand for foreign exchange, the the weighted AFf is used to determine the
government would, for example, devalue the foreign exchange adjustment, which is then
domestic currency, increase trade restrictions added up and used to adjust the economic net
or increase export subsidies, which in turn present value as discussed above. The adjust-
would increase the incentives to produce im- ment for the value of foreign exchange puts
port substitutes and exportables to earn for- the shadow prices fully on the basis of do-
eign exchange and would reduce the incen- mestic accounting rupees.
tives to produce non-tradables. The foreign exchange adjustment factor is
Given these predictable impacts of an in- next used ... to adjust the financial cash
creased demand for foreign exchange, it be- flows to reflect the differences between what
comes important to examine the cost of pro- the project paid for foreign exchange used in
ducing additional foreign exchange and the financial transactions and what the foreign
value of such to the users. The relative im-
portance ofthese two aspects — cost and sup-
8See Bela Balassa, "Estimating the Shadow Price
ply and value and use — will depend, of
course, on the relative responses of the supply of Foreign Exchange in Project Appraisal," Oxford
Economic Papers, Vol. 26, No. 2, July 1974, pp. 147-
and demand for foreign exchange given 168. Balassa's formula for the premium on foreign ex-
changes in its effective price. change in a situation where trade restrictions are ex-
pected to continue (his second-best shadow exchange
Probably the best known approach to de- rate) is (implicitly):
termining the cost of producing foreign ex-
change- isthe domestic resource cost method Y, ejXtQ + S) + 2Z,7imM,(l + 7)
proposed by Michael Bruno.7 While this ap-
T.i efx< + H,-""^*
7Michael Bruno, "Domestic Resource Costs and Ef- AF,=AFf- = adjustment factor for foreign exchange;
where
fective Protection: Clarification and Synthesis," Jour- ef = elasticity of supply of foreign exchange; nm =
nal of Political Economy, Vol. 80, No. 1, (January- elasticity of demand for imports; X( = exports; A/, =
February 1972), pp. 16-33. imports; T = import taxes and S = export subsidies.
 685 SHADOW PRICING

exchange is really worth. Similar adjust- change the project received and the value of
ments are made with respect to foreign ex- what it gave out in return.

X.C.5. Opportunity Cost of

Capital*
The capital costs of a project can usefully be an appropriate economic adjustment is added
viewed from two perspectives for shadow to or subtracted from the net present values
pricing. For lack of better terms, let us call of the capital investments at market
them the asset and the rent components. prices. . . . In practice, capital costs with
When Rs 100,000 is invested in project X, for roughly similar adjustment factors can be
example, two things happen. First, Rs
100,000 of financial resources is converted grouped together to minimize the work.1
The second part of the capital cost, its rent
into real physical assets. Secondly, the inves- component, is the opportunity cost, or for-
tor removes this Rs 100,000 worth of finan- gone productivity of the capital in other uses.
cial resources from the national pool of sav- The analysis here is strictly parallel to that
ings that might be used for investment in for other resources:
alternative projects. Once invested in project
The economic cost of capital is the cost of
X, therefore, these assets should yield a ben- 1.
efit, or rent, at least equal to what they would generating capital resources through ad-
ditional savings;
have otherwise earned.
Its economic value is the value of addi-
The shadow pricing of capital thus pre- 2.
tional production in alternative uses.
sents two problems: (a) how to measure the
value of the physical assets per se; and (b)
how to measure the rental value or opportu- 'Some authors argue that if capital for a project
nity cost of capital — the benefits forgone by comes from abroad and is available only to the pro-
ject, the assets purchased with these funds should not
freezing investable resources as assets in pro- be recorded as a cost until the capital is repatriated.
ject X instead of using them somewhere else.
This argument assumes that, in the case of a mineral-
Pricing of the asset component is exactly exploitation project, for example, the foreign mining
the same as for any other resource. If it is a company would not have invested in the country in
fully traded good, the value is its border the absence of the project, and therefore no other in-
vestment or production is given up in order to use the
price. If it is partially traded or non-traded,
capital for the mining project. Under such circum-
its shadow price is its economic cost of pro- stances, the country would incur a real capital cost
duction ifthe project induces increased do- only when it had to repay the foreign firm's original
mestic production, or its economic value investment. This approach tends to increase the net
measured in terms of consumer willingness to present value and rate of return on projects because
at least part of the total investment cost is discounted
pay if the project takes it away from alter- over more years than if it had been counted as a cost
native users. The labour involved in the con- at the time of actual investment. While this treatment
struction ofthe physical facilities is likewise is theoretically correct, there are few instances of
valued according to the guidelines. In each projects subject to economic evaluation by national
case, the adjustment factor is calculated, and authorities in which it can truly be said that the cap-
ital would not otherwise be available to the econoim
Thus, in practice capital costs should usually be
*From UN I DO, Guide to Practical Project Ap- counted at the time the physical investment is made.
praisal: Social Benefit-Cost Analysis in Developing The most likely exception to the latter is a project in-
Countries (John R. Hansen edition), 1978. Reprinted volving a private foreign investor who would not in-
by permission. vest in the country in the absence of the project.
 686 PROJECT APPRAISAL

To the extent that capital for the project is CRI. The second is to treat the discount rate
generated from additional savings, its eco- as a budgetary device rather than as an eco-
nomic cost is the price or rent savers must be nomic reality that can be verified empirically:
paid to forgo an additional unit of present if the first estimate of the discount rate indi-
consumption, the consumption rate of inter- cates that be morefinanced
projectswith
are the
"acceptable"
est (CRI). To the extent that the capital is than can available
taken away from competing investments, its funds, the discount rate should be raised.
economic value is its marginal product at Conversely, if the investment implied by the
shadow prices in the "marginal invest- "acceptable" projects (including investing
ments,"2 the investment rate of interest.3 In the money abroad) leaves excess investment
practice, it is not necessary in the Guidelines funds, the cut-off point for the discount rate
to figure out where the capital came from, should be lowered.
and thus the blend of interest rates that Because of the problem of determining em-
should be used, because stage three in this pirical y an appropriate CRI, the Guidelines
Guide converts the value of all inputs and recommends treating it as an unknown to be
outputs into their "consumption equiva- determined later as a "switching value," that
lents." It is therefore sufficient to use the is, the value at which a project becomes ac-
CRI as the discount rate. ceptable. The graphical approach outlined
There are serious problems related to de- above is very useful in this connection, for a
termining the CRI empirically.4 In practice, switching value for one project is the value of
however, two approaches can simplify the the CRI that leaves a zero net present value
task considerably. The first is to use the "ac- for the project; it is the discount rate at which
counting rate of interest" or opportunity cost the net present value switches from positive
of capital as a crude first estimate of the to negative. Therefore, the switching value
can also be regarded as the internal rate of
return of the project. When more than one
project is being considered, the switching
2The marginal investment is the last that would be value of the CRI is the discount rate at which
undertaken if all possible investments were ranked ac- one alternative comes to have a higher net
cording to their economic profitability and available
funds were distributed accordingly. present value than the other, a point also
3If savings are more valuable than consumption, widely known as the cross-over discount rate.
the investment rate of interest theoretically should be Another major practical advantage of the
further adjusted for the additional value to the econ- graphical approach in this connection is that
omy of the part of the product that will be saved and the discount rate can be selected after the
will thus produce further rounds of additional con-
sumption and investment. discounting is completed, provided, of course,
that the rate chosen lies between 0 per cent
4A theoretical formula exists for calculating the
CRI:
and the upper "plotting point" discount rate.
CRI = ng + p
These concepts can be demonstrated with
Figure 1, which plots the net present eco-
where: nomic values of projects A and B against dis-
count rates of up to 25 per cent. If project A
g = elasticity of marginal utility of consumption is considered in isolation, its switching value
with respect to changes in per capita income. is 14 per cent, the maximum discount rate at
g = annual growth of average per capita income; which A is acceptable. Whether 14 per cent
and
is the CRI is unknown, as indicated above,
p = pure time preference.
the Guidelines suggests determining that by
In practice, because of the inherently subjective nature of n what it calls the "bottom-up" procedure. The
and p, it makes more sense to follow the "bottom-up" ap-
proach suggested in the Guidelines; there is little point in
project analyst "at the bottom" can prepare
having two areas of subjective judgment instead of one. the project appraisal indicating that the in-
ternal rate of return is 14 per cent and pres-
 687 SHADOW PRICING

10 15 25
Discount rate (percent)

FIGURE 1. Switching values.

ent it to the ministers or planners "at the If projects A and B in Figure 1 are both
under consideration as mutually exclusive al-
top."
If the ministers or planners accept the pro- ternatives, and if the planners choose project
ject, the analysts can assume that the plan- A, the analysts can assume that the CRI is
ners judge the CRI to be less than 14 per less than 11 per cent, for at discount rates
cent; if they reject the project on the basis of higher than this switching value the net pres-
the rate of return, the analysts know that the ent value of project B becomes more attrac-
CRI is higher than 14 per cent. A repetition tive. The converse conclusion is reached if
of this process, provided that the planners at project B is preferred by the ministers or
the top are consistent, will gradually narrow planners, and again the actual value can be
the estimated CRI in the country to an ac- narrowed down by repetition of the process.
ceptable range.
Comment
For a more detailed analysis of shadow exchange rates, see the following: Edmar Bacha
and Lance Taylor, "Foreign Exchange Shadow Prices: A Critical Review of Current Theo-
ries," Quarterly Journal of Economics (May 1971); Bela Balassa, "Estimating the Shadow
Price of Foreign Exchange in Project Appraisal," Oxford Economic Papers (July 1974); M.
FG. Scott, "How to Use and Estimate Shadow Exchange Rates," Oxford Economic Papers
(July 1974).
Shadow wage rates are analyzed by Deepak Lai, "Disutility of Effort, Migration and the
Shadow Wage Rate," Oxford Economic Papers (March 1973); D. Mazumdar, "The Rural-
Urban Wage Gap, Migration, and the Shadow Wage," World Bank Staff Working Paper No.
197 (1974).
Shadow interest rates are discussed by M. S. Feldstein, "The Social Time Preference Dis-
count Rate in Cost-Benefit Analysis," Economic Journal (June 1964); Charles R. Blii/cr.
"On the Social Rate of Discount and Price of Capital in Cost-Benefit Analysis," World Bank
Staff Working Paper No. 144 (1972); Deepak Lai, "On Estimating Certain Inter-temporal
Parameters for Project Analysis," World Bank (processed, 1973).
Elements of risk and uncertainty in project appraisal are examined by Shlomo Rcutlingcr,
"Techniques for Project Appraisal under Uncertainty," World Bank Staff Working Paper
 688 PROJECT APPRAISAL

No. 10 (1970); Louis Y. Pouliquen, "Risk Analysis in Project Appraisal," World Bank Staff
Working Paper "No. 11 (1970).
Also useful are the works by I. M. D. Little and M. FG. Scott, Using Shadow Prices (1976);
M. FG. Scott et al., Project Appraisal in Practice (1976); A. C. Harberger, Project Evalua-
tion (1974).
One of the most complex aspects of the exercise of project appraisal is the precise identi-
fication ofthe project evaluator's areas of control. This will affect the nature of the exercise
that the evaluator has to solve and the shadow prices that will be relevant in the evaluation.
How one assesses the government's ability to control different policies will alter markedly
the nature of the appropriate shadow prices that are selected in project appraisal. This is
illustrated, especially in the choice of shadow wage rates, by A. K. Sen, "Control Areas and
Accounting Prices," Economic Journal (March 1972). To determine the appropriate shadow
price of labor, Sen emphasizes that it is important for the project evaluator to consider the
extent of influence that the government exercises over such relevant variables as the premium
on savings over consumption, weights based on income distributional considerations, valuation
of efforts by the peasant family and laborers, and the impact of employment creation on
migration.

X.C.6. Social Prices*

The following paragraphs contain an outline view, however, argues for excluding distri-
of the general rationale for the inclusion of bution weights. If the government, through
income distribution weights in project selec- its control of fiscal policy, is able to redistrib-
tion and broadly describe the manner in ute income costlessly, no need exists to in-
which this aspect of the analysis can be intro- clude distribution weights in project selec-
duced into the standard appraisal methods. tion; project selection should under such
The proposed use of distribution weights in circumstances aim to maximize income and
project appraisal raises the question of the allow the fiscal system to redistribute it in a
circumstances in which such weights are nec- desirable fashion.
essary. If,for example, the government val- It is argued here, however, that, in general,
ues all income equally regardless of its distri- redistribution can never be costless and that,
bution— either between the public and in particular, redistribution in developing
private sector or within the private sector — countries may be so costly as to be prohibi-
the need for distribution weights disappears. tive. With regard to the general argument, all
The weights themselves, however, only ap- fiscal measures have an administrative cost
parently disappear; in reality they are still and, at least in principle, a cost resulting
there, but the implicit value judgments made from an unfavorable effect on incentives.
are such that the social cost of each resource With regard to the particular argument, the
transfer is exactly offset by the resulting so- very unequal distribution of income-con-
cial benefit. sumption in most developing countries and
Many people would consider it rather ex- the difficulty of raising additional revenue in-
treme not to assign different values to mar- dicate severe constraints on the government's
ginal increments in consumption accruing to use of the fiscal system. These constraints
different income groups. Another point of typically reflect an inability to raise sufficient
revenue because to do so is not administra-
*From Lyn Squire and Herman G. van der Tak,
Economic Analysis of Projects, Baltimore, Johns tively feasible and an inability to tax the rich
Hopkins University Press, 1975, pp. 51-56. Reprinted sufficiently because of that group's political
by permission. power. Moreover, the general fiscal system of
 689 SHADOW PRICING

most developing countries (and, in fact, most the financial measure of the increase in con-
developed countries) cannot possibly reallo- sumption, toobtain its real resource cost. Let
cate the benefits and costs of projects as var- the adjustment factor be /3, so that the pri-
ied and geographically dispersed as those vate sector enjoys an increase in real re-
usually found in these countries. If these ar- sources of Cj8, and the public sector retains
guments are accepted (rejected), distribution control over the remainder: that is, E — C(3.
weights are (are not) required for project To measure the social value of these
selection. changes reference must be made to the social
When distribution weights are considered welfare function. Assume that the increase in
necessary, they may be introduced into the social welfare resulting from a marginal in-
standard economic analysis of projects in the crease in the availability of real resources to
following fashion. Assume that a project lasts the public sector is Wg, and that the social
one year and results in a net increase of E in welfare resulting from a marginal increase in
real resources available to the economy.1 If the availability of consumption to the partic-
interest lies only in efficiency, the increase in ular income group is We. The measure of so-
real resources is an adequate measure of pro- cial benefits is then: (E - C$)Wg + CWe.
ject benefits. But if there is interest in income Although Wg is defined for real resources, We
distribution as well, the distribution of re- is defined for consumption at market prices.
sources among different groups in society This simply reflects the fact that the public
must also be examined. The resources accru- sector is concerned primarily with increases
ing to each group can then be weighted in ac- in real resources, whereas the private sector
cordance with the appropriate concept of so- derives its utility from consumption possibil-
cial welfare and summed to obtain the
ities as determined by market prices.3
measure of the project's social worth.
To determine the distribution of real re- NUMERAIRE
sources, the distribution of financial net ben-
efits must be examined, because this will de- Social benefits then should be expressed in
termine who has control over the increase in a common yardstick, or numeraire. Any com-
real resources. Obviously, the financial ben- modity or resource may be chosen as this unit
efits will accrue either to the public sector or of account, but, once it has been chosen, all
to the private sector, and, within the private values must consistently be expressed in that
sector, they will accrue either to the rich or numeraire. It is recommended that as nu-
to the poor. Assume, for example, that as a meraire the value of real resources that are
result of the project the income of one partic- freely available to the public sector be used:
ular group in the private sector is increased that is, we want to set the weight assigned to
by C and that all other net financial benefits (E — C(3) equal to unity. To do this and to
accrue to the public sector. Assume further maintain all price relativities, we divide
that the private sector allocates the entire in- throughout by Wg, so that the measure of so-
crease in income to consumption.2 The in- cial benefits (S) in the chosen numeraire is:
creased consumption will comprise various S = (£" - C(3) + C
commodities or services that will have a cost (1)
in real resources, which may or may not Increase in real
Net social = resources in
equal C. Given the many distortions in the benefits
product and factor markets of less developed public sector
countries, it may be necessary to adjust C, Social welfare
+
from increased
'For a definition of symbols, see the Glossary of
Symbols [the last section of this paper]. private sector
consu
2An increase in private savings is treated sepa-
rately: that is, for the moment it is assumed that there }IVC, however, will express the social mption
value of the
is no saving out of private income. increase in private welfare.
 690 PROJECT APPRAISAL

where WJ Wg has been replaced by a?. Thus, the grounds that all financial benefits must
a unit of private consumption expressed in accrue to the factors of production employed
domestic prices (for example, C) has to be by the project. For example, assume that the
revalued by, in this case, a> to express it in the increase in income, C, generated by the pro-
numeraire. This may appear tedious, but, if ject actually accrues as a result of increased
consumption expressed in domestic prices wage payments to labor. The efficiency cost
were used as numeraire, the reverse process of employing labor — that is, forgone mar-
would have to be gone through in order to ex- ginal product of labor — is netted out in ob-
press (E — C/3) in the consumption taining the net increase in resources, E. In
numeraire.4 addition, employing labor involves a net so-
cial cost of increased consumption. Thus, the
EFFICIENCY AND SOCIAL PRICES social cost of the labor input can be defined
as: social cost = efficiency cost + C(/3 —
For practical purposes, it is more conve- oj). And if the increase in consumption per
nient to rewrite equation (1) in the following worker is c, then the social price per worker
manner: is: social price = efficiency price + c(fi —
S = E-C(P-o) (2) (ji).5 This latter method may be useful for
some factors such as unskilled labor, for
Net social Net efficiency which it is convenient to have an all-inclusive
benefits benefits
price for purposes of decentralized decision
Net social making, and may be essential for at least one
cost of increased price, the discount rate. For other factors this
private sector may not be particularly interesting, in which
consumption case the first method could be used. Although
both methods may be used in any single pro-
This has the advantage of separately identi- ject, they must not be used for the same fac-
fying the efficiency benefits, E. Thus, the tor payment because this would involve dou-
project economist-analyst can begin by esti- ble counting. . . .
mating efficiency benefits as has been done in Some general implications of the approach
the past. The only other item he must esti- outlined above can be considered at this
mate is the increase in net income accruing stage. First, if the increase in consumption, c,
to the various income groups in the private is zero, the social price equals the efficiency
sector who are affected by the project. The price.6 This might occur in a perfect labor
other elements of equation (2) — 0 and a> — market in which the transfer of labor involves
will then either be provided by the national no change in income or consumption. Second,
planning office or else be derived from a stan- if the wage earner spends all his income on,
dard table. say, duty-free imports or, more generally, in
Thus, the mechanics of this approach are nondistorted markets, then /3 equals unity. In
fairly simple. Alternatively, but equivalently, other words, £ can be viewed as a factor that
the distributional element of projects can be corrects for market distortions, especially
included in the definition of shadow prices on those caused by trade tariffs. /? may vary for
different consumers depending on the actual
4If account is to be taken of the distribution of con-
sumption, the numeraire would have to be defined as
the value of consumption at a particular level of con- Efficiency price is used in the traditional sense of
sumption. The public income numeraire is used in I. opportunity cost: that is, the foregone marginal prod-
M. D. Little and J. A. Mirrlees, Project Appraisal uct per worker. The terms "shadow" and "account-
and Planning for the Developing Countries, 1974, the ing" are used indiscriminately to refer to both effi-
general format of which has been followed here. The ciency and social price.
consumption numeraire is used in United Nations In- throughout this paragraph comments applying to
dustrial Development Organization, Guidelines for social and efficiency prices also apply to social and ef-
Project Evaluation. ficiency benefits as defined by equation (2).
 691 SHADOW PRICING

composition of their consumption basket. discount rate will be the opportunity cost of
Third, if the government is interested in in- capital, and other factor prices will be based
come distribution, oo will tend to be low for on opportunity cost. In other words, the eval-
the rich and high for the poor, and at some uation at efficiency prices corrects for the dis-
consumption level the situation a; = (3 would tortions infactor and product markets but
obtain, so that the real resource cost incurred does not assume any constraint on the gov-
by the government, c/3, and the social benefit ernment's ability to redistribute income or to
enjoyed by the worker, co>, as a result of a invest. The third evaluation will include the
marginal increase in consumption are exactly project's distributional impact [see Equation
offsetting. This level of consumption is known (2)] if it is thought that the economy does suf-
fer from a fiscal constraint.
as the "critical consumption level": the social
price equals the efficiency price at the critical
consumption level. Fourth, as has been seen,
the government may not wish to include dis- GLOSSARY OF SYMBOLS
tribution weights in project selection, in
which case the social price always equals the E =
Net efficiency benefits
efficiency price. C =
Private sector consumption
In presenting the economic analysis of a |8 =
Consumption conversion factor
project, it will be instructive to indicate the Wg =
Marginal social value of foreign ex-
project's worth at market, at efficiency, and change inthe public sector
at social prices. The first evaluation will cor- Wc = Marginal social value of private sec-
respond to the financial appraisal of the pro- tor consumption at consumption
ject. The second will be similar to that tra- level c
ditionally used by the Bank and other S = Net social benefits
agencies: that is, all incomes will be consid- oj = Value of private sector consumption
ered equally valuable, there will be no pre- at consumption level C relative to
mium on public income or investment, the the numeraire
 X.D. OBJECTIVES AND
RELATIVE WEIGHTS

X.D. 1 . Objectives of Planning *

The evaluation of investment projects should sumers' surplus." A consumer may expect
be carried out within the broad framework of satisfaction worth $2 from a good but if the
the objectives of planning. The projects cho- market price is $1 only, he need pay no more.
sen from among the feasible alternatives So the market price does not fully measure
should be those that contribute most towards the benefits that the consumers receive from
the achievement of the planning goals. a good. Corrections for this may be impor-
Therefore, good project analysis requires a tant in some cases and may be made by the
clear statement of the planning objectives. usual method of estimating the demand
There are usually a number of objectives curve for the good.
that the planners would like to achieve. Rais-
ing the standard of living, improving the dis-
tribution ofincome, achieving economic in- FUTURE CONSUMPTION AND
dependence, reducing unemployment, etc., GROWTH RATES
may all be regarded as desirable and the
question is what relative weights to attach to The planners will be interested not merely
these different goals. If one project fulfils one in the present consumption level but also in
goal more and no goal less than another the future level. One method of bringing the
project, then we can choose the first project future in is to attach a weight to the growth
without much ado. But frequently one project rate that is achieved today. But this cannot
will fulfil one goal better and another project give an account of the future beyond the next
will achieve more in terms of another goal, period since growth rates may vary over time.
and then the conflict between the two can be A more satisfactory alternative is to estimate
the consumption levels at different points of
resolved only by the relative weighting of the
divergent achievements. time in the future, to discount them at chosen
rates of interest, and then to arrive at a total
discounted sum of consumption. The interest
CONSUMPTION GOALS rate used for discounting will reflect what rel-
ative weights the planners wish to attach to
Raising the standard of living is a basic consumption today vis-a-vis that next year,
aim of development planning. One measure the year after, and so on.
of this standard is the level of average con- What considerations are relevant for the
sumption per head evaluated at constant choice of these discount rates? That will de-
market prices. This is a rough measure but pend on the objectives of planning. Presum-
may be easier to handle than more sophisti- ably, ifthe consumption level is rising over
cated measures. If prices change over time, time and people are getting richer, the im-
the rise in the money value of consumption portance of an additional unit of consump-
will have to be corrected by the choice of an tion today may be greater than that in the
appropriate price index. future. A 5 per cent discount will mean that
Evaluation at market prices does, however, the planners regard 105 units of consumption
leave out what economists call the "con- next year to be equivalent to 100 units this
year. The planners will have to choose the
*From Amartya K. Sen, The Flow of Financial discount rates after considering the expected
Resources, United Nations Conference on Trade and
Development TD/C/C.3/94/Add.l, August 20, growth rates of total consumption and of
1971, pp. 6-10. Reprinted by permission.
population.

692
 693 OBJECTIVES AND RELATIVE WEIGHTS

DISTRIBUTION OF INCOME EMPLOYMENT OBJECTIVES

Improvement of the distribution of income Reduction of unemployment is frequently
can be an important objective of planning es- an important objective of planning. If expan-
pecially ineconomies with much inequality. sion of employment is thought to be an objec-
There are many ways of bringing income dis- tive in its own right, then a weight may be
tribution into the cost-benefit evaluation. given to the additional employment gener-
One of the simplest is to attach an additional ated. The weight should reflect the amount of
weight to the consumption gains of the poor- average consumption today that the planners
est sections of society. Depressed regions may are ready to sacrifice for generating an ad-
be given preferential treatment through a ditional unit of employment.
premium attached to consumption accruing It is worth considering whether employ-
there, e.g., consumption of region A may get ment is regarded as an objective in itself or
a weight of, say, 1.10 in comparison with a as something that is desired as a means to
general weight of 1 per unit of average other ends such as a better distribution of in-
consumption. come. If the latter is the case the planners
It is sometimes argued that improvements may prefer to give a special weight to the
in income distribution should be brought consumption of low-income groups, thereby
about not through project selection, which permitting comparisons with other means of
may interfere with production efficiency, but achieving the same end, e.g., paying doles or
by fiscal means, e.g., taxation and subsidies. providing free rations. However, if greater
There is, obviously, some advantage in ef- employment is desired for its own sake, as-
fecting redistribution through taxation, but signing aweight directly to employment may
there are limits to what can be achieved this be the better procedure.
way, given the political and administrative
constraints. If, after tax reform, the planners
still find the income distribution to be unsat-
isfactory, animprovement in the distribution BALANCE OF PAYMENTS
may quite properly figure as a specific consid- Projects are sometimes judged in terms of
eration inthe choice of investment projects.
their balance-of-payments impact. This is,
undoubtedly, an important consideration, but
its use in cost-benefit analysis requires some
MERIT WANTS caution.
Foreign exchange is not desired for its own
It is not uncommon for the government to sake. It is not a basic objective of planning.
attach a degree of importance to certain In fact, it is more useful to view the foreign
types of needs in excess of that which these exchange balance as a constraint that re-
needs get in terms of market prices. For ex- stricts the freedom of resource allocation
ample, there may be a case for attaching than as an objective in its own right. On this
greater importance to education than the view, ultimate goals such as consumption cre-
market value of education would suggest. ation, improvement of income distribution,
Other examples of such "merit wants" (as etc., should motivate planning and the bal-
they have been called) include self-reliance ance of payments should come in only as a
in modern industrial technology, wider dis- constraint that must be met in achieving
semination ofmodern science, or the employ- these other objectives.
ment of female labour in regions where wom- Even if the earning of foreign exchange is
en's emancipation is a felt need. In such not an objective in itself, it is likely to be a
cases, the planners may give an additional valuable means for achieving other objec-
weight to what they regard as especially tives. Earning $1 more in foreign exchange
"meritorious" expenditure. will then have to be valued in terms of its im-
 694 PROJECT APPRAISAL

pact on the fulfilment of the other objectives. utes to the training of the local labour force
If it is thought that having an additional dol- performs a service that is not fully reflected
lar of foreign exchange will permit policies in its profits. On the other hand, the profits of
that lead to an additional $1.50 worth of the individual firm also do not reflect any
present consumption (or its equivalent in damage done to the environment, for exam-
terms of other objectives), then $0.50 is the ple, through pollution. These non-market ef-
premium to be placed on foreign exchange fects, known as "externalities," may be very
earning. important for project selection.
Many developing countries have chronic Some of these effects are, however, not
balance-of-payments problems and the nom- easy to quantify. The value of "learning by
inal exchange rates of their currencies do not doing" in modern industrial enterprises may
reflect the real scarcity of foreign exchange. be recognized as important without being
If such a situation prevails, the case for put- susceptible of precise quantitative measure-
ting a premium on foreign exchange is a ment. The valuation of the impact on the en-
straightforward one, since the nominal value vironment isalso not easy.
requires an upward correction. There are two ways of dealing with this
An additional dollar of foreign exchange problem. One is to ignore non-quantifiable ef-
can be valued for the purpose of cost-benefit fects altogether. The other is to give some
analysis in terms of its maximum contribu- value to non-quantifiables in line with the
tion to social welfare related to the objectives general goals of planning without pretend-
of planning. One has to compare its alterna- ing to accuracy. The second approach
tive uses, viz., expansion of imports and re- seems preferable. To ignore the non-quant-
duction ofexports, and choose the feasible al- ifiables altogether is to give them the
ternative that yields the highest return in value zero. That option is in any case open
terms of social welfare in units of present under the second approach as well, but
consumption. That return represents the so- if the planners feel that they have a better
cial value of the expenditure. guess than zero, they may use that guess.
Even market value estimates, especially
EXTERNALITIES AND NON- those involving the future, involve consider-
QUANTIFIABLES able elements of guess work, and the valu-
ation of externalities does not therefore
Not all the effects of a project work differ in principle from the rest of the
through the market. A factory that contrib- exercise.

X.D.2. Choice of Technology as

a Problem of "Second Best"*

Before one can enter the debate on the com- These include the concept of technical effi-
plex question of the choice of technology in ciency, which is one of the dominant concepts
the context of development planning, certain in the field of policy-oriented economics.
preliminary issues have to be cleared up. Suppose we are considering some technical
choice i which permits the production of an
♦From A. K. Sen, "Choice of Technology: A Crit- output combination x using an input combi-
ical Survey of a Class of Debates," in UNIDO, Plan- nation y.If it is possible to produce the same
ning for Advanced Skills and Technologies, Indus-
trial and Planning Programming Series No. 3, New bundle of commodities x with less of at least
York, 1969, pp. 45-9. Reprinted by permission. one, and no more of any, of the inputs, then
 695 OBJECTIVES AND RELATIVE WEIGHTS

the technical choice i is not efficient. This is set of efficient technological possibilities, the
simply because efficiency implies producing a choice between which must be made on the
given quantity of output with as few inputs basis of some other criterion. Efficiency is
as possible. Similarly, if with that collection like a test applied in the "qualifying round,"
of inputs y, an output combination can be and it needs to be supplemented by some
produced which exceeds x in at least one, and other criterion to determine which is the win-
is no less in terms of any, of the outputs, then ner amongst those alternatives that have
again the technical choice i must be regarded
as inefficient. This is because efficiency also qualified.
This is where the notion of optimality
implies that for any given collection of inputs comes in. This is one of the basic concepts
we should try to get the maximum of outputs. used in economics: an optimum choice rep-
In fact, we can combine the two criteria to- resents the "best" among the feasible alter-
gether by treating inputs as negative out- natives. Naturally, if we are to choose the op-
puts.1 Thus denned, what efficiency requires timum combination we must have some
is that no more of any one output can be ob- criteria for discrimination between the var-
tained given the amount of the others. ious alternatives.
This takes us a certain distance, but not A preliminary point of logic may be
very far. If a certain technical choice leads to cleared up at this stage. We can distinguish
a greater output of a given commodity and a between two conditions for rational choice:
lower output of some other commodity than
the existence of either a "complete ordering"
a different technical choice would do, the cri- or a "choice set." The former requires that
terion oftechnical efficiency does not help us any two alternatives should be consistently
at all. Both these technical choices may sat- comparable with each other in terms of some
isfy the test of efficiency and yet we may be ordering relation, such as "being at least as
left with a problem still to be solved. good as . . . ." This property is sometimes
The concept of technical efficiency is often called "connectedness." Another required
applied at a given point of time, but there is condition is "transitivity," which demands
no difficulty in extending it over time. All we that if x is regarded as being at least as good
need to do is treat a certain commodity today as y, and y is regarded as being at least as
as different from the same commodity tomor- good as z, then x should be regarded as at
row. In other respects the definitions, con- least as good as z.2 When these conditions are
cepts and criteria need not be altered. The satisfied, a complete ordering exists over the
same problem of incompleteness persists, relevant conditions.
naturally, even in this extended view of tech- The existence of a choice set is somewhat
nical efficiency, embracing more than one different. This requires the existence of some
point in time. In fact, in comparing two al- alternative which is regarded as at least as
ternative technological possibilities, we might good as every alternative in the available set.
face a possibility of having more of a certain
This simply means that a "best" alternative
commodity at point of time t and less of some
commodity at a point of time (t + 1). Once 2On the logic of ordering, see Arrow and Debreu.
again the criterion of technical efficiency It may be noted that the way we have defined any two
cannot solve this problem. alternatives being comparable not only guarantees
The main usefulness of the criterion of ef- "connectedness" but also yields "reflexivity" which
ficiency isthat it permits a preliminary sort- requires that every alternative be regarded as "at
least as good as" itself. When the alternatives consid-
ing out. A number of technological possibili- ered are the same, what was defined as "connected-
ties may be eliminated on grounds of ness" isin fact a condition of "reflexivity." By and
inefficiency, and then we shall be left with a large in optimal policy decision, reflexivity is not a
major source of worry; in fact, a minimal degree of
sanity seems to be sufficient. The real problem arises
'See Debreu, p. 56. A good introduction to the with connectedness and transivity. On this, see in par-
problems of efficiency can be found also in Koopmans. ticular Arrow.
 696 PROJECT APPRAISAL

exists. The existence of a choice set may be Whether we prefer a choice set or a
regarded as sufficient for the purposes of plete ordering, we need some method of or-
choosing an optimal policy. dering, a criterion to tell whether a certain
It is important to note that the existence of alternative x is better or worse than, or indif-
a complete ordering is neither a sufficient, ferent to. another alternative v. The con.
nor a necessary condition for the existence of of technical efficiency can be used partly for
a choice set. It is not sufficient because, al- this purpose, and we may find that x is more
though we may be able to order alternatives efficient than v and simply eliminate y. How-
in a certain fashion, if there is an infinite ever, as we noted before, this does not help
number of them, it is possible that no best al- when x and y are both efficient. Much of the
ternative may exist. For example, alternative debate on the choice of techniques is con-
2 may be preferred to 1, alternative 3 to 2, cerned with supplementing the criterion of
alternative 4 to 3, and so on, ad infinitum. It technical efficiency by some other criterion
is not a necessary condition because, al- that will permit us to choose between the ef-
though we may be able to compare some al- ficient alternatives. In the discussion that fol-
ternative with all the others and find it to be lows we shall be concerned with a choice
at least as good as them all, there may, never- among a set of efficient alternatives, and shall
theless, be intransitivities, or a lack of con- assume that the inefficient ones have already
nectedness. For example x may be regarded been pruned away.
as better than y and also better than z; but We shall thus have no further use for the
we may not be able to compare y and z by concept of efficiency as such, which (it will be
the criterion we are using. Even so, we may assumed) has done its job, and the discussion
feel safe in choosing x, since it is the best al- will concentrate on some supplementary cri-
ternative, although we cannot compare its teria to take us beyond efficiency. The lively
two inferiors y and z. debate on technological choice which has
In spite of this difference between the con- taken place over the last two decades has
ditions for the existence of a choice set and been concerned with methods of supplement-
those for the existence of a complete ordering ing the relatively uncontroversial criterion of
it is clear that there is an intimate relation- technical efficiency. To this range of prob-
ship between these two aspects of rational lems we now turn.
choice. In fact, most of the discussions on op-
timality have been concerned with obtaining SUBOPTIMALITY OF SAVINGS
a criterion for a complete ordering, and it has RATE AND CHOICE OF
been supposed that this in itself will guaran- TECHNOLOGY AS A PROBLEM OF
tee the identification of a best alternative.
This presupposition makes eminently good "SECOND BEST"
sense when the number of alternatives is fi- Investment decisions can be classified into
nite, when a consequence of the existence of various types, according to whether they de-
a complete ordering is the existence of a pend on the optimum size of investment, the
choice set. When, however, the number of al- optimum capital-intensity, or the optimum
ternatives iinfinite,
s this consequence may or sectoral allocation. While it is important that
may not follow. Furthermore, even when a we recognize these investment decisions to be
complete ordering does not exist we may still different, we cannot regard them as indepen-
be able to find what is the best thing to do. dent. Indeed, much of the controversy on the
Although we shall not be concerned very choice of technology concerns the depen-
much in this paper with this contrast it is im- dence of the amount of savings on the factor
portant for us to bear in mind the difference proportions selected.
between these two requirements of rational A simple illustration may bring out the dif-
selection. Indeed, in some problems the dis- ference between some of the schools of
tinction can be extremely important. thought. It may be argued that wage earners
 697 OBJECTIVES AND RELATIVE WEIGHTS

tend to have a higher propensity to consume efficiency grounds. We should then have
than profit earners. This is likely to be spec- cases in which a higher amount of total sav-
tacularly true in a socialist economy, where ings results in a lower amount of total output.
the profits are earned by the state, but it may What our choice is in such a situation will de-
hold good even in the case of a privately- pend crucially on the additional weights to be
owned enterprise. Given this assumption, it attached to savings as against output.
appears that the proportion of additional in- At this stage we might ask: Why must we
come that is saved will depend on the distri- attach any additional weight to savings as
bution ofthat additional income between the such? After all, savings involve a certain sac-
wage earners and the profit earners. And this rifice of present in favour of future consump-
distribution, in its turn, depends on the choice tion, and what reason is there for us to believe
of technology, since a more labour-intensive that it is always better to sacrifice present
technique will (other things being equal) consumption for a corresponding future
tend to lead to a higher share of wages. amount? Indeed there is no such compelling
A special case of this has been much dis- reason in general. What the debate on choice
cussed: the assumption that the wage earners of technology did was to assume (often im-
have a propensity to consume of 1 and profit plicitly) a suboptimal rate of savings, some
earners have a propensity to consume of 0. outside constraint preventing the savings rate
This is, however, a rather limited case, and from rising to the optimal level. As a conse-
the problem with which we are concerned re- quence there was always reason to look
lates to much more general conditions, viz., kindly on any policy which led to a higher
the propensity to save of profit earners is sys- proportion of savings.
tematically higher than that of wage earners. Why this suboptimality should arise is it-
Given this assumption a direct link would be self a complex question. In the case of pri-
established between the degree of labour in- vate-enterprise economy it can certainly be
tensity chosen and the proportion of the ad- argued that the rate of savings may be con-
ditional income that will be saved.
siderably below optimal.3 In particular, it has
Situations often occur in which one tech- been argued that people might be willing to
nique will lead to a higher amount of total sign a contract forcing everyone to save a cer-
output and another technique will generate a tain amount for the future, even when they
higher amount of total savings. If we wish to may not do it individually under the market
attach additional weight to the savings gen- mechanism: a situation of this kind has been
erated, over and above the weight that is at- christened the "isolation paradox".4agreement
tached to all output (be it saved or con- There seems to be considerable
sumed), then clearly this will affect our at a practical level regarding the need for
decisions regarding which techniques to raising the rate of saving in many underde-
choose. veloped countries. Indeed one has only to look
For the purpose of this discussion total out- through the planning documents of a variety
put and total savings may be regarded as two of countries to see that one of the persistent
separate commodities, even when they are as- themes is the need for a higher rate of saving
sumed to be physically homogeneous, as in and a higher rate of growth.5 These docu-
some simple models. The question of eco- ments are clearly based on certain assump-
nomic efficiency discussed in the last section tions, usually implicit, about the objectives to
may be applied to such a case. Any technique
which generates less of either total savings or 3See Pigou, Ramsey, Baumol. Dobb, Sen (1961,
total output and no more of the other may be Feb. 1967), Marglin (Feb. 1963, May 1963), Feld-
stein, and others.
simply rejected as inefficient. But after this 4Sen (1961). See also Baumol, Marglin (Feb.
preliminary pruning operation had been car- 1963), Harberger, Lind, Phelps, and Sen (Feb. 1967).
ried out, we would be left with a set of tech- 5See R. F. Kahn for a review of some of the plan-
niques that could not be compared purely on ning documents in this context.
 698 PROJECT APPRAISAL

be achieved by the economy, in terms of In the latter case an over-all suboptimality of

which the existing rates of saving appear to savings may be taken as given, since the pro-
be below optimal. Sometimes the arguments ject in question is too small to affect the over-
are fairly sophisticated,6 sometimes not. all inoptimality of savings.
Whatever the reasons for the suboptimal- Some kind of an objective function may be
ity of the savings rate, it seems to be clear given which depends on technical choice and
that this is a persistent diagnosis for most the proportion of savings. In the absence of
under-developed economies. In the presence any constraint on savings, our choices should
of such suboptimality it is not difficult to see lead to an optimal situation with the usual
why additional weight has to be attached to marginal equalities, if the exercise is of the
the part of the additional income that is former kind: in the absence of a specific con-
saved and invested as against the part that is straint on the rate of savings, the rate of
consumed. It is in this context that much of transformation between consumption at time
the controversy on the problem of choice of t and the consumption at time (t + 1) will
technology in recent years become fully equal the rate at which we are ready to sub-
clear. stitute the one for the other.8 There will then
Essentially the problem is that of choice of be no need for a marginal preference in fa-
technology in a world of suboptimal savings. vour of future consumption, implying addi-
It can also be viewed as a problem in the the- tional weight on the savings generated. Even
ory of "second best."7 Since misallocation at in the general-equilibrium framework, if
the margin of choice between savings and some outside constraint is imposed which
consumption is due to some specific con- prevents the rate of saving from rising above
straint, this will be reflected in the choice of a certain level, a suboptimality of savings can
the degree of labour intensity implicit in the result. It will then be appropriate to attach
technological selection. The problem would additional weight to savings as against the
have been totally different if the case had part of the income immediately consumed.
been one of allocating an optimal amount of Such constraints can arise for a variety of
savings among techniques with varying de- reasons, including political difficulties in tax-
grees of labour intensity. ation. The planners may want a higher rate
A distinction should be made in this con- of saving in terms of the objectives assumed
text between by them, but fail to achieve this for fear of
1. a general equilibrium formulation where political reactions.9 Given this political con-
straint the suboptimality of the savings rate
the amounts of savings, the degrees of la- that may be generated will tend to influence
bour intensity, and the pattern of invest- the optimal technical choice in the direction
ment are to be simultaneously selected;
and of relatively more capital-intensive tech-
2. a partial equilibrium formulation where niques, asimplying a relatively higher rate of
savings.
the technical choice is confined to finding
an optimal labour intensity for a marginal This is precisely where a different school of
thought may make itself heard, arguing that
project. such political constraints do not in fact occur,
In the former case the inoptimality of the
savings rate should not be assumed although 8This is on the assumption of smooth differentiabil-
ity. When there is only a limited number of basic al-
it may result from the allocational exercise. ternatives, with resultant "kinks" in the transforma-
tion surfaces, the corresponding rule will take the
6Optimum savings models have generally tended to form of a set of inequalities. See Dorfman, Samuel-
yield extremely high rates of savings, very much in son, and Solow.
excess of the usual rates observed anywhere in the
'Whether this range of problems can arise in a fully
world. See, among others, Tinbergen (1957, 1962), socialist economy is a matter for discussion. For some
Goodwin, Chakravarty, and Sen (1967). indications that they do, see Pajestka and Marglin
7See Lancaster and Lipsey. (Jan. 1966).
 699 OBJECTIVES AND RELATIVE WEIGHTS

that the total amount of income to be saved Q - Q(L, K) (1)

can be determined by the planner in any way
he likes, and that he can then see that this We assume this to be homogeneous of the
first degree, i.e., with constant returns to
decision is executed through the machinery
scale. Let the wage rate be given by w, the
at his disposal, such as wages and incomes
propensity to consume of wage earners by cx
policy, taxation policy and monetary policies.
If this is true then the link snaps between the and the propensity to consume of profit earn-
ers by c2. The amount of the income that is
choice of techniques and the proportion of in- saved is represented as S, which is expressed
come saved. Then technical choice may be
by the following relationship:
made with the main purpose of maximizing
S = Lw(\ - d) +
the amount of output,10 and the proportion of
the output to be invested can be decided at a (Q - L w)(l - c2) (2)
separate stage.
In the context of such an assumption it will We assume that the supply of labour is un-
be right to argue that the amount of income limited.11 The object of the exercise is to
maximize a certain weighted sum of output
generated in a surplus-labour economy ought
to be maximized even at the expense of sav- and savings.12 It is to be remembered that
savings S is a part of output Q, so that the
ings. This argument could spring from the as-
weight attached to S is in the nature of a pre-
sumption either that there is no suboptimal- mium, i.e., it is an additional weight, over
ity of savings or that the proportion of income and above the weight that S receives as a part
that can be saved, even if constrained, is not
of Q. Let this premium on savings be given
dependent on the distribution of income. Var- by X, which we have taken to be positive,
ious elaborations of these arguments can be
found in economic literature. since we have assumed the savings rate to be
Having commented on what appear to be suboptimal.13 The object, therefore, is to
maximize the following welfare function
some of the major issues that divide the dif- (3)
ferent schools of thought in the debate on V = Q + \S
technical choice, we may now proceed to dis- Given the amount of capital the problem of
cuss the controversy itself in some greater de- the choice of techniques is simply to find the
tail. We develop a general framework in the
right amount L, which will determine the ap-
next section and then express the various cri- propriate degree of capital intensity (Ay I).
teria in terms of comparison and contrast.
Due to the assumption of constant returns to
This general framework uses a very simple scale, it does not matter how we choose K, for
model with one homogeneous commodity, the discussion is all in terms of ratios per unit
which nevertheless illustrates almost the en-
"For a contrast of views on the empirical accepta-
tire controversy that has taken place in re- bility of this assumption, see Nurkse, Lewis (1954,
cent years on choice of techniques for an
1955, 1958), Eckaus, Mellor and Stevens, Rosen-
under-developed economy. At a later stage, stein-Rodan, Leibenstein, Viner, Haberler, Oshima,
in the context of a model on concave pro- Fei and Ranis, Schultz, Jorgenson (1966), Marglin
gramming, we shall discard this assumption, (1966), Sen (Oct. 1966, Feb. 1967), and Mehra.
and discuss the problem in a multi-commod- 12In a general equilibrium framework the weights
ity context. should vary with the choice of techniques, and the ob-
jective function should be "non-linear." However, in
the case of a small project, the total savings and the
A GENERAL FRAMEWORK consumption for the economy as a whole may not be
much affected by the marginal choice. There the
Let there be a production function relating weights can be taken as given, much as the perfectly
competitive firm takes prices as given.
output (Q) to labour (L) and capital (K).
,3The choice discussed here is for a marginal pro-
10 The implicit framework here is that of a one- ject. A wider exercise should take X as a variable. The
commodity model, but the corresponding conditions optimality conditions, however, will remain the same
for a multi-commodity model are easy to obtain. for appropriate values of X.
 700 PROJECT APPRAISAL

of capital. It is clear that the first order con- proposed really boil down to doing this very
dition of maximization of the objective func- thing in a highly implicit manner.15 . . .
tion isgiven by the following when K is given:
^=0 References
K. J. Arrow, Social Choice and Individual
dL (4)
Values, New York, 1951.
Given the equation (1), (2) and (3) it can be W. J. Baumol, Welfare Economics and the
seen that the conditions of maximization Theory of the State, London, 1952.
given by equation (4) requires the following: S. Chakravarty, "Optimum Savings with Fi-
dQ
nite Planning Horizon," International Eco-
nomic Review, Sept. 1962.
dL [1 + X(l - c2)] = Xw(c, - c2) (5) G. Debreu, The Theory of Value, London,
1959.
As a condition on the marginal productiv- R. Dorfman, P. A. Samuelson, and R. M.
ity of labour we can re-write relationship (5) Solow, Linear Programming and Economic
as follows, defining that magnitude to which Analysis, New York, 1958.
the marginal product of labor is to be M. H. Dobb, An Essay on Economic Growth
dQ and Planning, London, 1 960.
equated as "the real cost of labour" (w*):
w (6) R. S. Eckaus, "Factor Proportions in Under-
w* = (ci ~ c2)X developed Countries," American Economic
dL 1 +(1 - c2)Xj Review, Vol. 45, Sept. 1955.
C. H. Fei and G. Ranis, Development of
Much of the controversy on the choice of
Labor Surplus Economy: Theory and Pol-
technology for an under-developed economy icy, Homewood, 1964.
with surplus labour can be seen to be varia-
tions on the theme represented by equation M. S. Feldstein, "The Social Time Prefer-
ence Discount Rate in Cost-Benefit Anal-
(6). With this general framework we can sort
out the different contributions in this contro- ysis," Economic Journal, June 1964.
versial field. R. M. Goodwin, "The Optimum Path for an
One clarifying remark should be made be- Underdeveloped Economy," Economic
Journal, Dec. 1961.
fore we proceed further. The evaluation of al-
ternative techniques depends crucially on the G. Haberler, "Critical Observations on Some
value of X, i.e., on the additional weight to be Current Notions on the Theory of Eco-
attached to investment as against consump- 1957. nomic Development," L'Industria, Vol. 2,
tion. The value of X in its turn depends on the
relative weights to be attached to consump- A. C. Harberger, "Techniques of Project Ap-
tion today as against that in the future. What praisal," National Bureau of Economic
Research, Conference on Economic Plan-
we are really attempting, therefore, is to pro-
vide a one-period model which tries to catch ning, 1964.
the essence of comparison of the relevant sets D. W. Jorgenson, "Testing Alternative The-
of time series of consumption representing al- ories of the Development of a Dual Econ-
ternative technological possibilities. omy," in I. Adelman and E. Thorbecke
That the problem of choice of techniques (eds.), The Theory and Design of Eco-
nomic Development, Baltimore, 1966.
cannot be solved except in terms of making
explicit value judgments about alternative , "Subsistence Agriculture and Eco-
nomic Growth," Oxford Economic Papers,
sets of time series has been discussed by Sen14 1967.
who also argued that the different criteria
15Explicit attempts at making these comparisons
can be found in Sen (1957, 1968). This problem has
14Sen(1957). been penetratingly studied by Marglin (Jan. 1966).
 701 OBJECTIVES AND RELATIVE WEIGHTS

R. F. Kahn, "The Pace of Development," in J. Pajestka, "Some Problems of Economic

Hebrew University, The Challenge of De- Development Planning," in Oscar Lange
velopment, Jerusalem, 1958. (ed.), Problems of Political Economy of
T. C. Koopmans, Three Essays on the State Socialism, New Delhi, 1962.
of Economic Science, New York, 1957. E. S. Phelps, Fiscal Neutrality toward Eco-
K. Lancaster and R. G. Lipsey, "The Gen- nomic Growth, New York, 1965.
eral Theory of Second Best," Review of A. C. Pigou, Economics of Welfare, 4th ed.,
Economic Studies, Vol. 24, 1956. London, 1932.
H. Leibenstein, "The Theory of Underem- F. P. Ramsey, "A Mathematical Theory of
ployment inBackward Economies," Jour- Saving," Economic Journal, Dec. 1928.
nal of Political Economy, Vol. 65, Apr. P. N. Rosenstein-Rodan, "Disguised Unem-
1957.
ployment and Underemployment in Agri-
W. A. Lewis, "Economic Development with culture," Monthly Bulletin of Agricultural
Unlimited Supplies of Labour," Manches- Economics and Statistics, Vol. 6, July-
ter School, May 1954. Aug. 1957.
, The Theory of Economic Growth, T. W. Schultz, Transforming Traditional
London, 1955. Agriculture, New Haven, 1 964.
, "Unlimited Labour: Further , "Significance of India's 1918-1919
Notes," Manchester School, Jan. 1958. Losses of Agricultural Labour — A Reply,"
R. C. Lind, "The Social Rate of Discount Economic Journal, Vol. 77, Mar. 1967.
and the Optimal Rate of Investment: Fur- A. K. Sen, "Some Notes on the Choice of
ther Comment," Quarterly Journal of Capital-Intensity in Development Plan-
Economics, May 1964.
ning," Quarterly Journal of Economics,
Nov. 1957.
S. A. Marglin, "The Social Rate of Discount
and the Optimal Rate of Investment," , "On Optimizing the Rate of Sav-
Quarterly Journal of Economics, Feb. ing," Economic Journal, Vol. 71, Sept.
1963. 1961.
, "The Opportunity Costs of Public , "Peasants and Dualism with or
Investment," Quarterly Journal of Eco- without Surplus Labour," Journal of Po-
nomics, May 1963. litical Economy, Oct. 1966.
, Industrial Development in the La- "Terminal Capital and Optimal
bour-Surplus Economy, mimeographed, Savings," in C. H. Feinstein et al. (ed.),
Jan. 1966. Capitalism, Socialism and Economic
, "Mr. Jorgenson Against the Clas- Growth, Essays in Honour of Maurice
sics," in I. Adelman and E. Thorbecke Dobb, Cambridge, 1967.
(eds.), The Theory and Design of Eco- , "Isolation, Assurance and the So-
nomic Development, Baltimore, 1966. cial Rate of Discount," Quarterly Journal
S. Mehra, "Surplus Labour in Indian Agri- of Economics, Feb. 1967.
culture," Indian Economic Review, Apr. , Choice of Techniques, Oxford,
1966. 1960; 2nd ed., 1962; 3rd ed., 1968.
J. W. Mellor and R. D. Stevens, "The Aver- J. Tinbergen, "The Optimum Rate of Sav-
age and Marginal Product of Farm Labor ing," Economic Journal, Dec. 1957.
in Underdeveloped Economies," Journal , "Optimum Savings and Utility
of Farm Economics, Vol. 38, Aug. 1956. Maximization over Time," Econometrica.
R. Nurkse, Problems of Capital Formation in
Apr. 1962.
Underdeveloped Countries, Oxford, 1953. J. Viner, "Some Reflections on the Concept
H. T. Oshima, "Underemployment in Back- of 'Disguised Unemployment/" Contribu-
ward Economies: An Empirical Com- coes a Analise do Desenvolvi 'ment o
ment," Journal of Political Economy, Vol. omico, 1957.
66, June 1958.
 702 PROJECT APPRAISAL

X.D.3. Social Importance of

Employment Creation*
It was noted earlier that one reason why em- tive importance of channelling consumption
ployment isvalued is because of its impact on to the poor. The planners' evaluation need
income distribution. An unemployed person not be concerned with the precise calculation
does not have a source of income; although it of the impact of employment on the con-
is possible to give him some income through sumption ofthe poor classes; this would be
a dole, this practice may be difficult to follow. left to the project evaluator. If, on the other
In most developing countries a dole for the hand, the central policy makers attach pre-
unemployed is not provided. Undoubtedly cise value to employment, in estimating this
part of the reason is that with a large volume value the policy makers will have to take into
of surplus labour a poor country can ill-afford account the impact of employment on the
a dole system, and productive employment consumption of the poor classes, and the
even with low output is preferable. Partly value will have to be determined by consid-
also, with disguised unemployment it is not ering this in conjunction with the innate im-
very easy to identify those to whom a dole portance ofgiving consumption opportunities
should be paid. Under these circumstances, to the poor. The second procedure is less di-
an expansion of employment in spreading the rect than the first, and there are obvious ad-
real income very widely may contribute effi- vantages infollowing the first procedure. The
ciently to the redistribution objective. As was project evaluator himself is in a better posi-
noted earlier, many of the desirable aspects tion to judge the precise impact of employ-
of employment creation, its impact on nour- ment and other factors on income distribu-
ishment, on education etc. really relate to the tion, and it seems best to do this calculation
fact that employment creates a source of live- directly and to leave the central policy mak-
lihood for the family. ers in fixing social weights unencumbered by
This being the case, it is worth considering this complex detail. By valuing redistribution
whether employmerit should be valued sepa- the division of labour between the project
rately in project selection in the light of its evaluator and the over-all policy maker will
impact on income distribution, or whether in- be more appropriate, and this is the proce-
come distribution should be given a specific dure we follow in these Guidelines.
value and employment treated as a means to A more fundamental issue is why employ-
it. Fundamentally, this is not a crucial ques- ment should be regarded as a vehicle of in-
tion, since which way we do the calculation come distribution and why income cannot be
makes little difference so long as the link be- redistributed more directly through taxation
tween employment and redistribution is and fiscal policy. In principle there is no dif-
clearly recognized and realistically calcu- ficulty inpaying a person a certain amount of
lated. However, from the point of view of the money even without employing him in a pro-
convenience of calculation this is a question ject. Employment may be quite extraneous to
of some importance. the act of payment. The objection that pay-
If, on the one hand, we treat the income ing someone without employing him will be
level of the poorer classes as the relevant item unethical need not detain us; since we are
to which value is attached, that value will considering the role of employment in income
simply reflect the planners' idea of the rela- distribution, our real concern is with getting
income to the poor person, whether or not he
is employed. In fact, to the extent that work
*From UN I DO, Guidelines for Project Evalua-
tion, Project Formulation and Evaluation Series No. is regarded as unpleasant and leisure is re-
2, New York, 1972, pp. 91-7. Reprinted by garded as valuable, paying people without
permission. making them work may well be considered a
 703 OBJECTIVES AND RELATIVE WEIGHTS

superior means of income redistribution than possibility of distributing money to nonexis-

employment. tent persons, which is not uncommon in the
The picture is, however, not so simple. context of a pure dole in a country with a de-
Payment without work may have important fective administrative system. This is another
political and social repercussions. If the Gov- reason why employment may be an impor-
ernment decides to give a number of people tant means of redistribution.
some income without work, the question Be that as it may, in considering project se-
would naturally be raised why these people lection we should have to see the redistribu-
rather than others are selected for such sup- tional impact of employment creation as a
port. This question of selection arises even for possible part of the objectives of some
employment when unemployment is wide- projects. There is nothing strange about this.
spread, and charges of favouritism in job giv- Basically, this is another reflection of a pheno-
ing are not uncommon. However, in the case menon we have outlined in several places
of employment creation, there are at any rate in these Guidelines: the best economic possi-
some possible criteria for suitability for jobs bilities are not the same as political feasibili-
and also some definite procedure for giving ties. The redistribution of income through
employment. Also, getting paid for one's job employment may be feasible when pure redis-
is considered a compensation for one's effort, tribution without employment may not be,
even though the institutional wage rate may even when the latter may be economically
be regarded as very much in excess of the un- perfectly possible.
pleasantnes ofwork in a country with un- What we are mainly concerned with is an
employment. However that may be, politi- understanding of how employment may af-
cally the question of arbitrariness in job fect income distribution. This will vary from
distribution, while important, is not likely to project to project. One has to see to what ex-
be such an explosive issue as the giving out of tent additional employment will provide in-
income without work. come to particularly depressed groups. We
In some situations income may be redistri- have already discussed earlier how the redis-
buted better through a direct payment than tribution objective can be formulated in a va-
through giving employment. In dealing with riety of ways. One of the ways is to attach an
people with some crying need, e.g., medical additional weight to the consumption of de-
facilities, it may be simpler to give income pressed groups or classes. Sometimes when a
than to give jobs. It is not being suggested whole region is known to be economically de-
that employment giving is always the best pressed, income generating in that region
means of redistributing income. It need not may be given a special weight. Quite clearly
be so at all. But often employment will be an this will include the impact of employment
important vehicle of income redistribution because in estimating the income generated
and the fact that its political feasibility is in that region, note must be taken of employ-
somewhat greater than pure distribution of ment and wages paid out. If we are con-
money, except in very special situations, can- cerned with such broad considerations, no
not be overlooked. special efforts need be made to see that the
It might be mentioned in this context that effects of employment are reflected in the in-
the possibility of corruption is perhaps also come-distribution objective; the procedure
less when income is redistributed through already suggested will do it very well. How-
employment rather than through subsidies. It ever, sometimes we may wish to attach a spe-
has been found, for example, that in giving cial importance to income accruing to de-
famine relief in countries like India, the sys- pressed groups within a specific region, and
tem of paying wages to labour in specially de- we may then wish to attach a special weight
vised work programmes is less open to misuse to the wages paid to that group. Regions very
than the system of a direct dole. There are often are internally unhomogeneous, and it
clearer records of employment and also less may be important to distinguish between the
 704 PROJECT APPRAISAL

depressed and the not-so-depressed classes in mands of perfection and the constraints of
a certain region. practicability. In principle it would be best to
While such a distinction requires a fairly determine the precise income level of each
detailed calculation, it does not really affect employed person and attach a variable
the principles. On the one hand, the work weight to his respective consumption; the
should consist of making precise estimates of weight will go up as the average income level
the income generated that will be enjoyed by goes down. This will not, however, be possible
these specific depressed groups. On the other to do in any detail. The calculation would
hand, the policy makers must indicate the ad- have to be done in terms of broad categories.
ditional value to be attached to the income
thus generated. For most of these really de- SOCIAL COST OF LABOUR
pressed groups, income and consumption are
practically identical, so that we shall not be So far we have concentrated on the benefit
far wrong if we treat the accrued income of side of employment, making only passing ref-
these groups as equivalent to their present erences to costs. In a country with full em-
consumption. In determining the value to be ployment the cost of employment of labour is
attached to the present consumption of these fairly easy to calculate. A person can be em-
poorer groups, the policy makers must note ployed ina certain project only if he is with-
that consumption to these groups will, on re- drawn from employment somewhere else.
distributional grounds, be regarded as more From the point of view of this project, there-
important than the consumption of the aver- fore, the cost of employing him may be
age citizen of the country. As part of aggre- thought to be equal to what he would have
gate consumption, consumption of these de- produced had he been employed elsewhere.
pressed groups will receive a weight in any This measure of what he would have alter-
case in the system of evaluating benefits of
natively produced is sometimes called "the
the projects. The additional weight to be at- social opportunity cost" of labour, a term
tached to the consumption of these groups that is often used in project-evaluation liter-
may be reflected through putting a positive ature. By employing the person here the so-
value on the consumption of this group over ciety isforegoing the opportunity of employ-
and above the value of aggregate con- ing him elsewhere, and thus the social
sumption. opportunity cost measures the value of the al-
For assessing the impact on employment ternative opportunity that the society is los-
and sectional income, the precise pattern of ing by putting him to work in the project
disbursement between different categories of under discussion.
expenditures would have to be examined. Defined this way, the opportunity cost of
Very often project data are provided in such labour will be positive when there is full em-
an aggregate manner that the expenditures ployment, but if there is unemployed labour
on wages are not separated, nor is it specified it should be possible to employ labour in this
where the additional people to be employed project without having to withdraw it from
would be found. In the context of the objec- elsewhere. Thus, the opportunity cost of la-
tive of redistribution in relation to employ- bour as defined above may well be zero in the
ment it would be important to obtain this context of an economy with unemployment.
breakdown, and to check what part of the Does this mean that employment of labour is
disbursement reflects the additional wage bill costless for an economy with unemployment?
and also to whom these wages are to be paid, The answer is "not at all," since along with
i.e., whether the workers come from particu- employment come other changes in the econ-
larly depressed classes on whose income we omy that may or may not involve specific
would like to place an additional weight. costs from the social point of view.
Compromises in project evaluation would, A few of the simpler considerations may be
of course, have to be struck between the de- mentioned first, followed by a more complex
 705 OBJECTIVES AND RELATIVE WEIGHTS

problem. Although labour may be unem- A more complex consideration relates to

ployed, itdoes not follow that there is no un- the impact of employment on the distribution
pleasantnes ofwork, especially since work- of the current income between consumption
ing conditions in developing countries often and investment. When an additional person
tend to be extraordinarily bad. The unpleas- is employed, drawn from the pool of the un-
antness of work for those who would have employed, and he is paid wages, some addi-
otherwise been idle cannot be dismissed. This tional purchasing power is generated, and
point assumes particular importance when this will reflect itself in an increase in con-
there is a transfer of labour from the rural sumption. Of course, increased consumption
areas to the harsh living conditions of the is desirable, and indeed, aggregate consump-
growing towns and cities in the poor coun- tion isthe first objective that we examined in
tries. The conditions of living, including the the context of cost-benefit calculation. How-
sanitary facilities and other social amenities ever, an increase in immediate consumption
in some of the urban areas of the developing may be achieved, under most circumstances,
countries, are often miserable, and it must be only through a reduction in investment. If the
assumed that there is some loss in making policy makers feel that, on balance at the
people work under such conditions. It is pos- margin, consumption and investment are
sible to feel that this is an unimportant con- equally attractive from the point of view of
sideration, since the worker in question pre- the society, it makes no difference whether
fers to take the job rather than be investment is cut a little to increase immedi-
unemployed. However, this is not a convinc- ate consumption correspondingly or whether
ing argument. The worker prefers this job be- immediate consumption is somewhat reduced
cause he is paid a certain wage, and while for the sake of expansion of a corresponding
this wage may overcompensate him for the amount of investment. If, however, we are
inconvenience of bad working and living con- dealing with an economy where the policy
ditions, this does not mean that the working makers feel that the over-all rate of invest-
and living conditions do not involve some suf- ment is deficient, a reduction in investment
fering. Therefore, just as we must calculate for the sake of an expansion of immediate
the benefits from employment in terms of consumption may be regarded as a loss. One
output creation as well as of income redistri- way of viewing this problem is to regard the
bution and other objectives, we must also price of investment in terms of immediate
take into account the social costs, if any, ol consumption to be greater than one. . . . The
additional employment, especially when it in- shadow price of investment, Fm (t), is a cru-
volves migration. cial factor in assessing the social cost of em-
From the social point of view, a further ployment, since the expansion of employment
consideration may arise because the Govern- means a shift from investment to consump-
ment may have to construct housing and tion, and the loss involved in this per unit is
other facilities in an area to which the work- equivalent to the value of [Fn\t) - 1]. If in-
ers may be moving, and these costs may not vestment and consumption are already opti-
be borne, at any rate not fully, by the workers mally distributed, i.e., there is neither under-
themselves. Because a large proportion of the investment nor overinvestment, then the
capital expenditure of a project lies in the value of Pn\t) should be equal to 1 and the
cost of townships, this could be an important loss involved would be precisely 0. However,
cost related to employment. If the benefits as is common in most developing countries, if
from new houses and good working condi- the planners believe that the level of invest-
tions are provided, they may be included ment is too low, there will be a loss, since
among the benefits of the project, but the cost Fnv(t) exceeds 1 . The social cost of labour de-
of townships and housing must be counted as pends not merely on the social opportunity
part of the cost of employment creation of the cost of labour but also on the shadow price of
project. investment.
 706 PROJECT APPRAISAL

We may seem to be contradicting ourselves group vis-a-vis that by a richer group. In

in some ways by regarding additional con- making a detailed calculation of benefits and
sumption generated through additional em- costs of a project both these considerations
ployment (a) to be a good thing because it are relevant, but they will come into our es-
leads to a better distribution of income, and timation under different objectives and will
(b) to be a bad thing because it leads to a work in different directions.
shift from investment to consumption. This Under the broad hat of aggregate con-
is, however, not a contradiction, and both sumption would come the question of the rel-
facts are correct and relevant, though they ative weight of investment vis-a-vis average
work in opposite directions. A unit of income consumption today, since the impact of in-
that accrues to a worker in a project rather vestment on future consumption is fully re-
than accruing to the project authorities can flected in our estimation of the aggregate-
be viewed in one of two ways. To the extent consumption benefit. On the other hand, the
that it reflects workers' consumption as op- special weight to be attached to the consump-
posed to average consumption of the com- tion of the poorer groups vis-a-vis average
munity, itcan be regarded as more valuable; consumption today must come under the
to the extent that it reflects consumption heading of the redistribution objective. Em-
rather than investment it must be regarded ployment creation will, therefore, come in
as less valuable if the country is suffering both the benefit and the cost side under these
from a shortage of investment vis-a-vis con- two sets of objectives; additional employment
sumption. The latter is the comparison of will involve a cost through the "aggregate-
consumption with investment and the former
consumption objective" and a benefit through
is the comparison of consumption by a poorer the "redistribution objective."
Comment
Previous selections argue that market savings tend to be below optimal, and that project
appraisal should attach a premium on investment. In putting an extra weight on savings in
the determination of the shadow price of labor, the project evaluator is raising the savings
rate through control of project selection and the choice of capital-intensive investment.
A more general analysis of intertemporal optimality can be studied using the theory of
optimal growth paths. Although the theory has been devised for advanced economies, some
aspects of the theory relate to efficient capital accumulation and may be suggestive for the
optimal allocation of investment between consumption and capital goods in a developing econ-
omy. An excellent survey article is F. H. Hahn and R. C. O. Matthews, "The Theory of
Economic Growth: A Survey," Economic Journal (December 1964). Among the many growth
models, the following may be of special interest: M. Dobb, An Essay on Economic Growth
and Planning (1960); R. Findlay, "Optimal Investment Allocation Between Consumer Goods
and Capital Goods," Economic Journal (March 1966); L. Johansen, "Some Theoretical Prop-
erties of a Two-Sector Model of Optimal Growth," Review of Economic Statistics (January
1967); L. G. Stoleru, "An Optimal Policy for Economic Growth," Econometrica (April
1965); S. Bose, "Optimal Growth and Investment Allocation," Review of Economic Statistics
(October 1968); A. K. Dixit, "Optimal Development in the Labour-Surplus Economy," ibid.
(January 1968); H. Uzawa, "Market Allocation and Optimum Growth," Australian Eco-
nomic Papers (June 1968); Jaroslav Vanek, Maximal Economic Growth (1968); R. R. Nelson
and S. G. Winter, "Neoclassical versus Evolutionary Theories of Economic Growth: Critique
and Prospectus," Economic Journal (December 1974).
CHAPTER XI

Development Planning and

Policymaking

All the preceding chapters have had some relevance for development planning, and we have
already related specific aspects of development planning to several substantive problems. But
we now want to examine the nature and scope of development planning in a more integrated
fashion. By appraising the effectiveness of development planning in light of the development
record and the analyses and policy proposals that have emerged in preceding chapters, this
final chapter may also serve as a summary evaluation of the leading issues in development
economics.
Just beneath the surface of practically all the preceding analyses has been the issue of
planning versus the price system. We now focus directly on this, with some of the materials
in this chapter arguing against the effectiveness of the price system for attaining the objectives
of development. The market mechanism is criticized as being either ineffective, unreliable, or
irrelevant for the problems now encountered by developing nations. It is contended that the
price system exists in only a rudimentary form in many of the LDCs and that market forces
are too weak to accomplish the changes needed for accelerated development. Even a fairly
well-defined price system may be considered unreliable, it is claimed, when market prices of
goods and factors are not a true reflection of the opportunity costs to society. Above all, it is
contended that the price system must be superseded insofar as the determination of the
amount and composition of investment are too important to be left to a multitude of individual
investment decisions, and the tasks of the economy entail large structural changes over a long
period ahead instead of simply marginal adjustments in the present period.
On the other hand, it can be argued that the objections to the market system are relatively
unimportant compared with the essential functions of the market and that the disadvantages
of detailed planning by a central authority are far more serious than the deficiencies of the
market system. The market is extremely valuable in the context of development as an admin-
istrative instrument that is relatively cheap to operate (XI.A.2). For this reason and others

707
 708 DEVELOPMENT PLANNING AND POLICYMAKING

outlined in XI. A, development policy might be better devoted to improving and strengthening
the market system than to supplanting the market with detailed administrative controls.
To the extent, however, that there has been a large measure of government direction over
the rate and pattern of development, we should appraise the techniques and applicability of
development planning. Taken together, the materials in sections XI. B and XI. C indicate the
nature of the available programming techniques, some operational problems, and the signif-
icance oftheir effects in practice. Although section XI. B considers general principles of model
building and provides some introduction to the theory of development planning, it does not
pretend to offer a detailed exposition of modeling techniques. Development programming has
become a specialized subject in its own right, requiring more familiarity with mathematical
and econometric techniques than can be provided here. We also slight the administrative as-
pects of development planning, leaving for more specialized courses in public administration
or development management the consideration of the organizational problems of policy
implementation.
We are, however, most interested in moving beyond the theory of development to a review
of planning practice in order to distinguish between the techniques discussed in development
literature and those that have actually been used in various countries. Although planning
models have been useful in improving the formulation of development plans, their application
is still beset with practical difficulties, and it is essential to recognize their possible abuse in
development planning.
The Note in section XI. C submits some general conclusions on the present state of devel-
opment planning and the modifications necessary to improve its future effectiveness. This
Note and other selections in this chapter may appear critical of specific plans, but it should
be realized that there can be considerable value to the planning process itself. The active
process of planning may indeed be of more benefit than the actual plan, for it can demonstrate
the need for the collection and use of more empirical information, promote the dissemination
of knowledge, clarify objectives and choices, and indicate the political and administrative
preconditions necessary for implementing policies.
What ultimately matters is the benefit that can come from a dialogue between the model-
ouilder and policymaker. While a model cannot provide a final answer, it can illuminate
choices and indicate to the policymaker the consequences of alternative decisions. The imple-
mentation ofmathematical planning is, however, dependent on the degree of maturity of non-
mathematical planning. It is only clear what aims are worth striving for when there has al-
ready been established an organized institutional, nonmathematical form of planning.
In a single model only a few hundred relationships and constraints can be considered. But people working
in the central planning agencies and lower-level institutions and enterprises "sense" hundreds of thou-
sands of further constraints and relations, and they can give expression to these in their own estimates.
Mathematical planning will develop successfully only when it develops as one element of well-prepared
and well-oriented institutional planning, connected by many threads with real economic life in developing
countries.1
This chapter closes with a Note on international policymaking and the New International
Economic Order (NIEO). Recognizing the futility of a North-South confrontation and the
unrealism of many of the South's demands, the Note proposes instead that the North-South
issues be factored into a cooperative solution of the several conflicts that the internationali-
zation process has created for rich and poor countries alike. \

'Janos Kornai, "Models and Policy: The Dialogue between Model Builder and Planner," in C. R. Blitzer et
al., Economy-Wide Models and Development Planning, New York, 1975, chapter 2.
 XI.A. PLANNING AND THE
MARKET
XI.A.1. The Flaw in the
Mechanism of Market Forces*

The free and unimpeded mechanism of mar- consumers' goods by flows of supply of these
ket forces would lead to a maximum national goods from given stocks of equipment, raw
income according to the liberal classical doc- materials, and labor.
trine. Disregarding an ethical value-judg- The price mechanism does not work in this
ment about personal income distribution and sense, however, in the third equilibrium when
special cases of increasing returns to scale the we drop the assumption of given fixed capital
maximum would also be an optimum na- and assume that the amount and composition
tional income. Any conscious deliberate ac- of investment is to be determined by a mul-
tive economic policy designed to influence the titude of individual investment decisions.
amount and the composition of investment The individual investment decision may
could not, according to this school, raise na- lead to nonoptimum allocation of resources
tional income in the long run. It is the con- for the following reasons:
tention ofthis paper that the opposite is true, a. The investor maximizes the private, not
that an economic policy designed to influence the social net marginal product. External
the amount and composition of investment economies are not sufficiently exploited.
can raise the rate of economic growth and in- Complementarity of industries is so great
crease national income. that simultaneous inducement rather than
Maximization of national income would be hope for autonomous coincidence of invest-
ment iscalled for.
reached, according to the "liberal" school, by
the working of the mechanism of supply and b. The lifetime of equipment is long (say
demand on assumption of competitive condi- ten years) so that the investor's foresight is
tions and of small changes per unit of time, likely to be more imperfect than that of the
in four stages or "equilibria": (1) allocation buyer and seller or of the producer. The in-
of given stock of consumers' goods, (2) allo- dividual investor's risk may be higher than
cation of production on assumption of given that confronting an over-all investment pro-
stock of equipment, land, and labor, (3) al- gram. The costs of an erroneous investment
location of investment, on assumption of decision are high; punishment in the form of
given stock of labor, land and capital. A loss of capital afflicts not only the investor but
fourth equilibrium condition is provided by also the national economy.
c. Because of the indivisibility (lumpiness)
Say's law.
It is true that the price mechanism works of capital, large rather than small changes
perfectly under those assumptions in the first are involved. Yet the price-mechanism works
stage, i.e., in allocation of given stocks of con- perfectly only under the assumption of small
sumers' goods. It works less perfectly, but tol- changes.
erably well, in the second stage, when we re- d. Capital markets though often well or-
place the assumption of given stocks of ganized are notoriously imperfect markets,
governed not only by prices but also by insti-
♦From Paul N. Rosenstein-Rodan, "Programming tutional or traditional rationing quotas.
in Theory and in Italian Practice," in Massachusetts
Institute of Technology, Center for International
The investment theory is indeed the weak-
Studies, Investment Criteria and Economic Growth, est link in the "liberal" theory.
Cambridge, 1955; Asia Publishing House, Bombay, It is finally recognized even by the strong-
1961, pp. 19-22. Reprinted by permission. est advocates of a free economy that an equi-

709
 710 DEVELOPMENT PLANNING AND POLICYMAKING

librium between aggregate demand and ag- sures the optimum rate of growth. In this
gregate supply (i.e., the dynamic monetary third stage programming concerns not only
equilibrium) cannot itself be ensured by the amount but also the composition of
trusting to the automatic responses of a free investment.
economy. This task can only be discharged The aim of programming is to assure the
by a deliberate policy. Without an equilib- maximum national income through time. For
rium of aggregate demand and aggregate this purpose it tries to maximize the amount
supply, however, prices cease to be reliable and to optimize the composition of invest-
parameters of choice and the price mecha- ment. The means employed may be either in-
nism breaks down. direct (monetary, fiscal and commercial pol-
The automatic responses of the market icy as well as providing information on
economy do not ensure an optimum alloca- economic trends besides other incentives and
tion in two out of four markets. They allocate disincentives) or direct (public investment).
Even if no direct means were to be employed
efficiently stocks of consumers' goods, and
supplies of these goods flowing from stocks of programming would be necessary in order to
equipment, but they do not function effi- inform investors of short- and long-run
ciently inthe fields of investment and mone- trends, notably of intersectoral demands re-
tary equilibrium. sulting from complementarity of industries
Programming is just another word for ra- (revealed by a periodically revised input-out-
tional, deliberate, consistent, and coordinated put table) and of indirect effects of invest-
economic policy. It is only spelling out explic- ment on future demand of domestic, im-
itly what was always attempted implicitly in ported, or so far exported goods. Such
any monetary (fiscal and commercial) policy. information might guide and favorably influ-
Like Mr. Jourdain who talked prose all his ence the composition of investment.
life, programming at least in the field of mon- The need for an active economic policy,
etary policy was always practiced, though it even if it were to employ only indirect means,
used shorthand rules of thumb or the "prac- can hardly be denied. Agreement on this
tical man's flairs, hunches, instincts and in- point may conceal, however, two different
sights" rather than fully spelled out "White conceptions. The "liberal" considers State in-
Papers" on output and employment. Neutral tervention as an occasionally necessary med-
Government is as unrealistic an assumption icine. "Nobody denies that clean living is the
as Neutral Money. best way to good health, but this is really not
Monetary (fiscal and commercial) policy is a sufficient reason to deny that it is some-
a form of programming using indirect means times necessary to take medicine." This view
for the achievement of its aims and targets. is based on the assumption that private in-
In its first pre-Wicksellian stage its only aim vestment decisions normally lead to an opti-
was equilibrium between aggregate demand mum position. The other view considers that
and aggregate supply without distinguishing a continuous active economic policy, beyond
too clearly between investment and consump- measures to assure an equilibrium between
tion. Itbecame more purposeful in its second aggregate demand and aggregate supply, is
post-Wicksellian stage differentiating be- necessary, since the multitude of dispersed
tween and choosing its impact effects on var- individual uninfluenced decisions will not
ious sectors of national income. In its third lead to the maximization of national income.
post-Keynesian stage it aims not only at a The real question is how far programming
monetary equilibrium but at one which as- should extend, what it should cover, what de-
sures full employment and more and more grees of "freedom" it should leave between
also one (not necessarily the same) which as- and within the various economic sectors. . . .
 71 PLANNING AND THE MARKET

XI.A.2. The Market Mechanism

as an Instrument of
Development*
In recent times, there has been a retreat both services — a mixture of bureaucracy and en-
in economic theory and in economic policy terprise, inwhich bureaucracy takes care of
from the nineteenth-century ideal of the un- the major risks of development and enter-
fettered market as a principle of economic or- prise faces and overcomes the minor ones.
ganization. But the economic pros and cons Another relevant point that Easterbrook
of this retreat have been fully debated, and makes is that an economy which succeeds in
the economist consequently has a great deal finding a formula for growth tends to repeat
to say about the relative merits of the market that pattern after it has become inappro-
as contrasted with other methods of eco- priate. For example, Britain has gone on
nomic organization, and the circumstances trying to work the internationally-oriented
appropriate to each. pattern for her nineteenth-century develop-
The subject of planning and the market in ment; Russia has been very successful in de-
economic development is, therefore, one veloping heavy industry but has not yet
which falls definitely within the field of the solved the problem of agriculture.
economist. Before I go on to discuss it, I must The alternative interpretation takes plan-
define more precisely what I mean by it. ning, inthe sense of a general direction of the
"Planning and the market" may be inter- economy, as an established principle, and
preted in two different ways. First, it may considers the market as an alternative to
refer to the contrast between direction of the other and more direct means of detailed con-
economy by Government and the policy of trol. Given the general framework of eco-
laissez-faire. This is not my subject, though nomic planning, there is still a choice be-
in a wider philosophical and historical con- tween two alternative methods of looking
text it offers much to discuss. For example, after the details. One is by direct detailed
though laissez-faire and direction are often planning by a central authority, the other is
regarded as opposites, if one looks to the his- by leaving the working out of details as far as
tory of economic development one finds (as possible to the operation of the market.
Professor Easterbrook has shown1) that eco- (There is a third alternative, in which the
nomic development is almost invariably a Government is itself the entrepreneur and
process in which planning and direction on investor, which I shall consider later.)
the one hand and freedom of enterprise on This alternative interpretation is the one I
the other play their part, and are mixed. shall be using: I shall discuss the question of
There is almost no case in which economic the market mechanism as against detailed
development has been entirely planned or en- planning as an instrument of economic de-
tirely unplanned. The usual pattern is one of velopmentshould
.I like to make it clear from
some framework of control by Government, the start that I am going to make a strong
within which the entrepreneur provides his case for the market, as the preferable instru-
ment of economic development, on two main
*From Harry G. Johnson, Money, Trade and Eco- grounds. The first is that the achievement of
nomic Growth, George Allen & Unwin Ltd., London,
1962, pp. 152-3, 156-9, 160-63. Reprinted by
the desired results by control methods is
permission. likely to be especially difficult and inefficient
'Professor Easterbrook's analysis was presented in
in an underdeveloped economy; at this point
the Marshall Lectures at Cambridge University in the I should like to remind you that a large part
spring of 1956. Unfortunately these lectures have not
been published, but some of the ideas are available in of Adam Smith's argument for laissez-faire
was the inefficiency and corruption he saw in
W. T. Easterbrook, "Long Period Comparative the Governments of his time. The second is
Study: Some Historical Cases," Journal of Economic
History, XVII, No. 4. December 1957, pp. 571-95. that the remedies for the main fault which
 7 12 DEVELOPMENT PLANNING AND POLICYMAKING

can be found with the use of the market nical improvements to profit from their ex-
mechanism, its undesirable social effects, are ploitation. Moreover, the market serves
luxuries which underdeveloped countries particularly to provide an incentive to the ac-
cannot afford to indulge in if they are really cumulation ofcapital of all kinds: first to the
serious about attaining a high rate of devel- accumulation of personal capital in the form
opment. In particular, there is likely to be a of trained skill, since such skill earns a higher
conflict between rapid growth and an equi- reward; and second to the accumulation of
table distribution of income; and a poor coun- material capital, since such capital earns an
try anxious to develop would probably be well income.
advised not to worry too much about the dis- The argument, then, is that a properly
tribution ofincome. functioning market system would tend to
I am going to make a fairly strong case for stimulate both economic efficiency and eco-
the market, because the market figures rela- nomic growth. And it is important to note
tively little in the literature of economic de- that the market does this automatically,
velopment, and the theoretical analysis which while it requires no big administrative appa-
economics has developed in relation to mar- ratus, no central decision-making, and very
kets isoften overlooked or disregarded. . . . little policing other than the provision of a
I now want to recapitulate briefly the var- legal system for the enforcement of contracts.
ious economic functions of the market and All this sounds very impressive; but it is
the price system as a method of economic or- clearly not the whole of the story. What,
ganization.shall
I be brief, as the argument then, are the objections to the market, how
is a familiar one. serious are they, and what should be done
In the first place, the market rations sup- about them in the context of economic devel-
plies of consumer goods among consumers; opment? Ishall discuss these questions in
this rationing is governed by the willingness some detail. But first I shall state briefly the
of consumers to pay, and provided the distri- central theme of my discussion. It is that in
bution ofincome is acceptable it is a socially many cases the objections to the market can
efficient process. Secondly, the market di- be overcome by reforming specific markets,
rects the allocation of production between so as to bring them closer to the ideal type of
commodities, according to the criterion of market; and that to overcome other objec-
maximum profit, which, on the same assump- tions to the market may be very expensive
tion, corresponds to social usefulness. and may not prove to be worthwhile — in
Thirdly, the market allocates the different other words, the defects of the market mech-
factors of production among their various anism may on balance be more tolerable than
uses, according to the criterion of maximiz- they look at first sight.
ing their incomes. Fourthly, it governs the Now, what are the objections to the mar-
relative quantities of specific types of labour ket? They can, I think, be classified into two
and capital equipment made available. main types. One type of objection is that the
Fifthly, it distributes income between the market does not perform its functions prop-
factors of production and therefore between erly. The other type of objection is that the
individuals. Thus it solves all the economic results produced by the functioning of the
problems of allocation of scarce means be- market are undesirable in themselves.
tween alternative ends. I begin with the first type of objection, that
These are static functions; but the market the market does not perform its functions
also serves in various ways to provide incen- properly. Here it is useful to draw a distinc-
tives to economic growth. Thus the availabil- tion between two quite different sorts of
ity of goods through the market stimulates cases — those in which the market operates
the consumer to seek to increase his income; imperfectly, and those in which a perfectly
and access to the market provides an oppor- functioning market would not produce the
tunity for inventors of new goods and tech- best results.
 713 PLANNING AND THE MARKET

Imperfect operation of the market in an cal progress or capital accumulation tend to

underdeveloped country may be attributable proceed more rapidly in industry than in ag-
to ignorance, in the sense of lack of familiar- riculture. But it raises an immediate question
ity with market mechanisms and of aware- of fact — whether divergences between social
ness of relevant information, or to the prev- and private benefit or cost are numerous and
alence of other modes of behaviour than the important or not. This is an important ques-
rational maximization of returns from effort. tion, but one on which we do not know very
In the first case, the appropriate Governmen- much for certain. The theory of increasing
tal policy would seem to me to be, not to as- returns is logically intriguing, but the influ-
sume from the market the responsibility for ence of increasing returns still has to be dis-
allocative decisions, but to disseminate the entangled from that of technical progress in
knowledge and information required to make historical growth. Again, it is a fact that few
the market work efficiently and provide the advanced countries are not industrial; but
education required to use it. The second case this by itself does not establish the wisdom of
implies a more fundamental obstacle, not a policy of forced industrialization in an
only to the use of the market but also to eco- underdeveloped country. Aside from the
nomic development itself, and suggests that question of fact, the existence of divergences
successful economic development requires a between social and private returns does not
basic change in social psychology. To my necessarily indicate a need for the govern-
mind, it raises a serious question of fact. Is it ment to replace the market mechanism; in-
really true that people in underdeveloped stead, the operation of the market can be per-
countries are strangers to the idea of maxi- fected by the use of appropriate taxes and
mizing gains? The idea that they are is very subsidies to offset any divergences between
common in the literature and policy-making social and private returns.
of economic development; one of its manifes- I now turn to the second type of objection
tations isthe implicit assumption that both to the market, the point of which is not that
supplies and demands are completely price- the market does not work in the way it
inelastic. I am very sceptical about this, should, but that the results produced are un-
desirable inthemselves. Here, I think, there
partly because of Bauer's work and partly be-
cause at least some of the actions of Govern- are two major objections to the market. The
ments in underdeveloped areas presuppose first is that the income distribution produced
that even the poorest producers are suscepti- by the market is unjust and socially undesir-
ble to price incentives. I personally do not able. The distribution of income through the
think one is justified in assuming as a general market depends on the wealth and talents of
proposition that ignorance and illiteracy nec- different individuals, and on their individual
essarily imply that men are not interested in skill in seeing a profitable opportunity of em-
making money. If it is true, there will be se- ploying their money or labour. If they make
rious difficulties in the way of economic de- a wise or lucky choice, they may obtain a
velopment; but again, the appropriate Gov- much higher income. The objection is that
ernmental policy would seem to be to educate this method of determining the distribution
the people in the practice of rational eco- of income is not just. But if you attempt to
nomic behavior. intervene in the distribution of income, you
Even if the market functions perfectly, it immediately encounter the problem that such
will not produce the best possible results by intervention interferes with the efficiency of
its own criteria if there is a difference be- the market system. If people are not allowed
tween social and private benefit or cost. This to enjoy the income they could obtain by
type of case may be particularly relevant to their decisions, their decisions in turn will be
economic development; it includes the case of affected, and the efficiency of the system will
increasing returns to scale, and can be ex- be impaired. There is, therefore, a conflict
tended to include the possibility that techni- between economic efficiencv and social jus-
 714 DEVELOPMENT PLANNING AND POLICYMAKING

tice. The extent and importance of this con- cepted, this conclusion nevertheless does not
flict islikely to vary according to the state of carry with it a number of corollaries which
economic development. The more advanced a are often attached to it. In particular, it does
country is, the more likely are its citizens to not necessarily imply that the state ought to
have consciences about the distribution of in- undertake development saving and invest-
come, and to accept the high taxation neces- ment itself. Private enterprise may be more
sary to correct it without disastrously alter- efficient than the Government in constructing
ing their behaviour; and on the other hand, and operating enterprises, so that the best
the higher the level of income reached, the policy may be to stimulate private enterprise
less serious will be any slowing down of the by tax concessions, subsidies, and the provi-
rate of growth brought about by redistribu- sion of cheap credit. Similarly, it may be
tion policies. An advanced country can afford preferable to stimulate private saving by of-
to sacrifice some growth for the sake of social fering high interest rates, rather than by
justice. But the cost of greater equality may forcing savings into the hands of the state by
be great to any economy at a low level of eco- taxation or inflation. One argument against a
nomic development that wishes to grow rap- policy of low interest rates and forced saving
idly, particularly as it is evident that histori- is that it may in the long run contribute to the
cally the great bursts of economic growth inequality of income distribution. The reason
have been associated with the prospect and is that the poor or small savers are mainly
the result of big windfall gains; it would confined to low-yielding fixed-interest inves-
therefore seem unwise for a country anxious ments, directly or indirectly in Government
to enjoy rapid growth to insist too strongly on debt, because these are safe and easily avail-
policies aimed at ensuring economic equality able, whereas the larger savers can invest
and a just income distribution. I should add their money in higher-yielding stocks and
that the problem may not be in fact as serious shares or directly in profitable enterprises.
as I have made it out to be, since in the course There is, therefore, an opportunity here for
of time rapid growth tends in various ways to Government both to stimulate saving for de-
promote a more equal distribution of velopment and to improve the distribution of
wealth. . . . income.
I have been discussing the objection to the There is another reason for being wary of
results of the market system on the grounds the proposition that the state should under-
that it produces an undesirable distribution take development investment itself — the
of income. A second objection of the same danger that if the Government undertakes in-
sort is that the free market will not produce vestment itself, especially if its administra-
as high a rate of growth as is desirable. I tors are not too clear on their objectives, the
think there is a strong case for this objection, result will be the creation of vested industrial
because people's actions in regard to saving interests inimical to further development,
and investment depend very much on their and resistant to technical change.
guesses about the future. Now people are To summarize the foregoing argument
likely to know their own current require- from the point of view of development policy,
ments better than the Government. But the it seems to me that much of development
requirements of the future have to be looked planning could usefully be devoted to the im-
at not from the individual or family point of provement and strengthening of the market
view or that of the nation as a collection of system. This does not imply the acceptance of
individuals, but from the point of view of the all the results of laissez-faire, especially with
ongoing society. The needs of society in the respect to the rate of growth; but there are
future, many economists agree, tend to be reasons for thinking that too much emphasis
underprovided for by the free market. on a fair or ethical distribution of income can
Even if the conclusion that state action is be an obstacle to rapid growth.
desirable to raise the rate of growth is ac- The argument I have presented has been
 7 15 PLANNING AND THE MARKET

concerned mainly with one side of the case amount of attention, labour mobility and im-
for the market. The other side concerns the mobility much less. Now, the process of eco-
costs and difficulties of controls, in terms of nomic development in the past, especially in
the manpower costs of the administration the nineteenth century, was characterized by
they require, and their effects in creating vast movements, not only of capital, but also
profit opportunities which bring windfall of labour, about the world. The mass move-
gains to some members of the community ment of labour between countries has now
and create incentives to evasion which in turn been more or less shut off by the growth of
require policing of the controls. I have nationalism. I believe it is important to rec-
touched on that side of the argument suffi- ognize this restriction on international com-
ciently frequently to make it unnecessary to petition, and its implications for programmes
elaborate on it further. of economic development. It means — looking
Instead, I shall comment briefly on inter- at the world economy as a whole — that the
national markets in relation to economic de- solution to the problem of maximizing world
velopment, since so far I have been implicitly output cannot be approached directly, by
concerned with internal markets. Economic bringing labour, capital, technology, and nat-
development planning inevitably has a strong ural resources together at the most efficient
autarkic bias, by reason both of its motiva- location; instead, the other productive factors
tion and of the limitation of the scope of con- have to be brought to the labour. To a large
trol to the national economy. Nevertheless, extent, "the economic development of under-
international trade can play an important developed countries" is a second-best policy,2
part in stimulating and facilitating the devel- in which gifts of capital and technical train-
opment process. Access to foreign markets ing by advanced to underdeveloped countries
for exports can permit an economy with a are a compensation for the unwillingness of
limited domestic market to exploit economies the former to consider the alternative way of
of scale, and the potentiality of such exports improving the labour to resources ratio,
can serve as a powerful attraction for foreign movement of the labour to the resources. The
capital and enterprise. Similarly, the capac- fact that development is a second-best policy
ity to import provided by exports can give a in this respect may impose severe limitations
developing economy immediate access to the on its efficiency and rapidity.
products of advanced technology, without To conclude, I have been concerned with
obliging it to go through the long and perhaps the role of the market in economic develop-
costly process of developing domestic produc- ment; and I have aimed at stressing the eco-
tion facilities. Economic nationalism and ex- nomic functions of the market, in automati-
cessive fear of the risks of international trade, cally taking decisions about various kinds of
by fostering aversion to exploiting the ad- allocations of economic resources, and the
vantages of the international market, can place in economic development programmes
therefore retard economic development of improvements in market organization and
unnecessarily. methods. I have been advocating, not a policy
One further comment on the international of laissez-faire, but recognition of the market
aspects of the market and economic devel- as an administrative instrument that is rela-
opment seems to me worth making. Discus- tively cheap to operate and may therefore be
sion of the international side of development efficient in spite of objectionable features of
has been mostly concerned with commodity
trade and commercial policy. But in fact one
of the most important ways in which the 2See J. E. Meade, The Theory of International
Economic Policy, Volume II: Trade and Welfare,
world market system is imperfect is with re- London, Oxford University Press, 1955, and R. G.
spect to the international mobility of capital
and labour. The problem of international Lipsey and Kelvin Lancaster, "The General ThcOT)
of Second Best," Review of Economic Studies, XXIV
capital movements has received a fair (1), No. 63 (1956-57), pp. 11-
 716 DEVELOPMENT PLANNING AND POLICYMAKING

its operations. The general assumption on tween the state and private enterprise, and
which I have been arguing is that economic that the problem is to devise the best possible
development is a process of co-operation be- mixture.

XI. A. 3 The Fragmented

Economy*

While economists can usefully divide their America, and Africa, primary commodity
labor as monetary theorists, tax experts, for- export enclaves were controlled by foreign-
eign trade specialists, project evaluators, and ers, and much of the general population re-
so on, a unified view of the development pro- mained outside of the market economy. In-
cess isa great analytical convenience. Why is digenous entrepreneurs had limited access to
public intervention so pervasive and generally capital, no means of acquiring advanced
so unsuccessful? Intervention is usually technologies, and little skilled labor.
prompted by the perception — sometimes cor- Thus in the determination of where in the vast
rect— that a particular market is functioning new areas of the overseas world the raw material
badly, so that authorities feel pressed to "do export industries were to be established, the pre-
something." An infant textile firm is helped existing domestic supply of labor, capital, and en-
by a tariff; or the price of an agricultural trepreneurship played a minimal role. Where they
product may be raised to permit farmers to did exist in areas of potential export production,
these factors were highly immobile and could not
use a new fertilizer-intensive technology; or a be counted upon to engage in export industry
tax exemption may be granted to a foreign
firm for automobile assembly. This pressure
for public intervention is the result of severe Newly independent governments quite
operation.1
fragmentation in the underdeveloped properly felt compelled to act as agents of
economy. change to offset economic and political colo-
nialism. In the past twenty or thirty years,
poor countries have succeeded in introducing
THE FRAGMENTED ECONOMY
some new industrial activities — particularly
the manufacture of goods previously im-
thatThefirms
economy is "fragmented"
and households in the sense
are so isolated that ported— and in mobilizing some domestic
they face different effective prices for land, factors of production. Their governments
labor, capital, and produced commodities chose to do so, however, by manipulating
and do not have access to the same technol- commodity prices in a variety of ways and by
ogies. Authorities then cannot presume that intervening directly to help some individuals
socially profitable investment opportunities or sectors of the economy at the expense of
others.
will be taken up by the private sector, be-
cause prevailing prices need not reflect true Consider the extraordinary lengths to
economic scarcity — at least not for large seg- which import tariffs have been used in Latin
ments of the population. There is historical America, with rates of several hundred per-
justification for this view in the nineteenth
1Jonathan V. Levin, The Export Economies: Their
and early twentieth centuries. In Asia, Latin Pattern of Development in Historical Perspective,
Cambridge. 1960, p. 169. For a good description of
how small farmers, who are strongly motivated to-
*From Ronald I. McKinnon, Money and Capital ward economic efficiency, can nonetheless be locked
in Economic Development, Brookings Institution, into a backward agricultural technology, see Theo-
Washington, D.C., 1973, pp. 5-8. Reprinted by dore W. Schultz, Transforming Traditional Agricul-
permission. ture, New Haven, 1964, pp. 36-48.
 717 PLANNING AND THE MARKET

cent on some goods and absolute prohibitions great income inequality between the wealthy
on the import of others, while still others few and the poverty-stricken many. This in-
enter freely. The situation on the Indian sub- come inequality has failed to induce high
continent isno different. Price and quantity rates of saving in the classical manner, but
controls on foreign trade and domestic com- governments remain reluctant to reduce the
merce make licensing and rationing com- disposable income of well-to-do investors
monplace. Byzantine patterns of industrial whose unique access to investment opportu-
taxes and subsidies complicate government nities isguaranteed by the web of official con-
budgetmaking. Consequently, the market trols and by the endemic fragmentation.
mechanism has become no better, and per-
haps even worse, as an indicator of social LIBERALIZATION AND THE
advantage. CAPITAL MARKET
Modern fragmentation, therefore, has
been largely the result of government policy How does one begin to loosen the Gordian
and goes beyond the old distinction between knot? The incredibly complex distortions in
the export enclave and the traditional subsis- commodity prices now prevailing are the un-
tence sector. One manifestation is the planned macroeconomic outcome of specific
often-noted existence of small household microeconomic interventions. But substantial
enterprises and large corporate firms — all fragmentation in the markets for land, labor,
producing similar products with different and capital provided the initial motivation for
factor proportions and very different levels of public authorities to "do something" and
technological efficiency. Continuing mecha- continues to pressure governments to inter-
nization onfarms and in factories in the pres- vene. Thus an explicit policy for improving
ence of heavy rural and urban unemployment the operation of factor markets is necessary
is another. Excess plant and equipment with to persuade authorities to cease intervening
underutilized capacity are commonly found in commodity markets. Carefully considered
in economies that are reputed to be short of liberalization in all sectors can then move
capital and that do suffer from specific bot- forward — not merely as a reaction to the
tlenecks. In rural areas, tiny landholdings more obvious mistakes of the immediate past,
may be split up into small noncontiguous par- but in ways that allay legitimate fears of pure
cels, with inadequate incentives for agricul- laissez-faire.
tural land improvements. However, the knot needs to be loosened
While tangible land and capital are badly further. To say that there are "imperfections
used, fragmentation in the growth and use of in factor markets" is distressingly vague and
human capital can be more serious and no often signals the end of formal economic
less visible. Learning-by-doing and on-the- analysis. But further systematic inquiry can
job training in the "organized" economy are proceed if the neoclassical approach of treat-
confined to narrow enclaves — export-ori- ing labor, land, and capital symmetrically as
ented in the past but now increasingly in- primary factors is dropped. It is hypothesized
ward-oriented toward "modern" manufac- here that fragmentation in the capital mar-
turing— whose employment growth may be ket— endemic in the underdeveloped envi-
less than the growth in general population. ronment without carefully considered public
Unemployment among the highly educated policy — causes the misuse of labor and land,
coexists with severe shortages in some labor suppresses entrepreneurial development, and
skills. condemns important sectors of the econonn
Indigenous entrepreneurship is narrowly to inferior technologies. Thus appropriate
based and is supported by heavy government policy in the domestic capital market is the
subsidy. Tariff protection, import licenses, key to general liberalization, and particularly
tax concessions, and low-cost bank finance to the withdrawal of unwise public interven-
commonly go to small urban elites and create tion from commodity markets. . . .
 7 18 DEVELOPMENT PLANNING AND POLICYMAKING

XI. A. 4. Planning and Cost-

Benefit Analysis*

The market is an indispensable element of stated: those who do not like the actual price
every modern society based on labor division. should try to achieve a change in the actual
Experience has shown that it will not disap- price. The problem cannot be avoided by re-
pear even after the revolutionary transfor- placing the actual price we do not like with a
mation of capitalist conditions rooted in the shadow price we prefer. Shadow prices may
private ownership of production means. The play a useful role, but only as analytical tools
market also functions in a socialist economy. in the hands of the planner. We cannot ex-
It is true that (as has already been men- pect, however, as already discussed, that mi-
tioned) the scope of the market changes in crounits (firms, local authorities, etc.) will
history. Its regularities are deeply influenced base their actual decisions on financially non-
by the actual social, economic, and political binding accounting prices.
environment: the functioning of the other In my opinion, the market should not be
components of the economic control system. left to itself. The government may intervene
Yet whatever the conditions in which it func- in it, and does in fact in every modern soci-
tions, its importance lies in that it is a real ety.1 Besides, the market may be supple-
market and not just a shadow market exist- mented by planning and other government
ing only on the paper of economic calcula- control activites of a nonmarket character.
tions. The market is a real market if the But within the sphere of market action, it
buyer spends money and the seller receives should be a real market with real prices.
money. The buyer's interest is to spend less Planning is, in the vision of the CB school,
money for the article, while the seller's inter- a faint repetition of the market: a kind of
est is to receive more for it. The real market ghost market with ghost prices and wages
functions in the framework of this con- which are more advantageous than the real
tradiction. ones.
I have no illusions about the market. I
The CB school's "planning" is a language
know very well that it is not "perfect." A real whose grammar and basic vocabulary are
market price does not express exactly "how identical with those of the market, and which
much a certain product is worth to society." differs from the market only in its accent.
The rent paid for resources does not reflect Yet the real historical role of planning is not
exactly their "relative scarcity." There may just that. Planning is a "separate language"
be not only small and temporary but also
with its own vocabulary and grammar.2 It is
large and permanent deviations between the not a faint repetition of the market, but a
real market price and the so-called social val-
uation. Ifwe think that the price wrongly ori- 'The study of shadow prices and their comparison
with actual prices may help to correct the latter and
ents those who make their choices relying on to determine government price policy.
the price signal, we may try to change the ac- 2In his classical article on externalities, Scitovsky
tual price. This may be done in various forms: (1954) underlined the complementary character of
by government price fixation, taxation and market and planning: "The proper co-ordination of
subsidies, support or hindering of cartel for- investment decisions, therefore, would require a sig-
naling device to transmit information about present
mation, etc. It is not within the scope of the
plans and future conditions as they are determined by
present study to take a stand on these ques- present plans; and the pricing system fails to provide
tions, but the following principle must be this. Hence the belief that there is need either for cen-
tralized investment planning or for some additional
*From Janos Kornai, "Appraisal of Project Ap- communication system to supplement the pricing sys-
praisal," inMichael J. Boskin (ed.), Economics and tem as a signaling device.
Human Welfare: Essays in Honor of Tibor Scitov- "It must be added that the argument of this section
sky, New York, Academic Press, 1979, pp. 91-96. applies with a special force to underdeveloped
Reprinted by permission.
countries."
 7 19 PLANNING AND THE MARKET

supplement with its own purpose. A living segment "represented" by them. This may
market must be supplemented by living seem alarming to those who believe that
planning. planning is a stricly impartial process which
The main aspect of planning is the primal deduces its figures from facts independent of
aspect, if I may express it in the language of human interests. In reality, however, plan-
mathematical programming. The targets of ning is a social process in which human
production, investment, consumption, and beings take part. Finally, the plan is obtained
domestic and foreign trade must be known out of this clash of interests.
beforehand. One must try to coordinate all
these closely related processes. In parallel THE TWO INTERPRETATIONS OF
with the physical processes, their accompa- PLANNING
nying monetary processes, i.e., the flows of
incomes and expenditures, must also be co- The various views on planning and project
ordinated. All this, of course, must be com- appraisal are summarized in Figure 1 . Figure
pleted bythe dual aspect: the planning of fu- \a shows the neoclassical (welfare econom-
ture prices and wages. But these are not ics) ideal of planning. According to this,
hypothetical prices either. The planning of planning is nothing else but the solution of
prices means that there is consideration be- the "economic problem" (i.e., allocation of
forehand ofwhat changes could be and ought scarce resources to attain maximum utility)
to be achieved in future actual prices, wages, on an economy-wide level. It presumes the
exchange rates, and interest rates. existence of a social welfare function. The
Planning is therefore a wide prognosis and objective is to reach the maximum social
organized exchange of information on the welfare.
one hand, and a preliminary coordination of While this fundamental starting point is
interests and activities on the other hand. All shared by all those holding the neoclassical
this, if well done, will improve the economic view, there are differences between them re-
efficiency and reduce the frictions of adap- garding methods. This is shown by the
tation. However, the whole thing is but an ex branching in Figure 1. According to some,
ante coordination which will later be exposed the "economic problem" can be solved only
to the test of practice. In the market sphere by an optimization mathematical model, or,
it is the actual market, and in the nonmarket at least, the solution can be approached only
sphere it is the other allocation processes in this way (Figure la, lower left rectangle).
(e.g., allocation by authorities, administra- Others hold the view that the problem must
tive selection, etc.) that will realize actual co- be solved exclusively by isolated project ap-
ordination and adaptation. They will do it praisal, i.e., by cost-benefit analysis (Figure
either by faithfully following the plan or by la, lower right rectangle). There are also in-
correcting it. It may also happen that forces termediate opinions, recognizing the useful-
active in society will resist the plan and there- ness of both procedures, but laying more
fore the goals included in the plan will not be stress on one or the other.
adhered to. The basic conception of welfare economics
In real planning, as later in the real market accepted, the various views harmonize quite
and in other real nonmarket allocation pro- well with each other. Arguments are re-
cesses, the representatives of different inter- stricted to opinions about the best means or
ests meet. It is true that this usually takes combination of means for "planning technol-
place in a very "professional" form: with eco- ogy," which is, finally, not a very fundamen-
nomic and technical argumentation. If, how- tal question.
ever, one takes a closer look at the argu- In Figure 16, I show the concept of plan-
ments, it is found that big enterprises, ning that I accept, together with many other
ministerial planners, ministers, the leaders of economists. As can be seen, it is not the
geographical regions, trade unions, etc., all "planning technology" that makes this con-
cept different from the other concept, since
try to assert in planning the "interests" of the
 720 DEVELOPMENT PLANNING AND POLICYMAKING

Planning is a strictly rational Planning is a process of cognition

and objective allocation process and reconciliation of interests

There exists a social welfare No social welfare function exists

function

Task of mathematical
The optimum plan Task of individual
Cost-benefit planning carried out
should be computed project analysis: to
analysis relying on at economy-wide and
on economy-wide sectoral level: to explore explore possibilities,
and sectoral levels: optimum shadow
consequences, and
the welfare function prices: decentralized possibilities, consequences,
linkages to other parts
attains its maximum project selection agreements, and conflicts of the national economy
of interests

(a) (b)

FIGURE 1 . Two interpretations of planning.

both mathematical planning and project The fulfillment of both functions may be
analysis also appear in this scheme, i.e., in promoted by mathematical planning on econ-
the procedures. The point of difference is in omy-wide and sectoral levels (and within it
the interpretation of the social function of by optimization models, nonoptimizing sim-
planning. It does not begin with the sterile ulation examinations, and input-output anal-
ideal of "rationality." The question here is ysis) as well as by project appraisal. The lat-
not about the just, fine, and good planning, ter is— without sectoral and economy-wide
with every participant a "Homo oeconomi- analysis — by no means sufficient for plan-
cus." Instead, it approaches planning in the ning. On the other hand, it may be a useful
realistic way and poses the empirical ques- supplement to plan calculations of a wider
tion: What is planning in practice? First, it is scope.
a cognitive process; its participants try to ex- The final question to be posed in this study
plore possibilities and to state in advance is: What then is the place of cost-benefit
what the consequences of effectuating this or analysis in planning? I think it has an impor-
that possibility would be. The task is, there- tant role to play — although much less impor-
fore, to enrich this exploration process as tant than the one assigned to it by the most
much as possible. Second, planning is a pro- ardent followers of the CB school.
cess of reconciliation of interests. It tries in
advance to coordinate activities that would
THE PLACE OF COST-BENEFIT
otherwise adjust to each other through con- ANALYSIS
flicts. Planning may help in overriding the in-
terests of particular groups to enforce more Before directly answering the question just
general goals of the political decision-makers mentioned, let us see objectively what has
and planners. Planning can fulfill these func- been the actual role of cost-benefit analysis
tions well if it clearly reveals the tensions and up to now. I think that, as has been pointed
contrasting interests to be expected. out already, it has not fundamentally influ-
 721 PLANNING AND THE MARKET

enced actual decisions. If it has exerted an Beyond this educational effect asserting it-
advantageous effect, it has not been ex- self, in the long run cost-benefit analysis has
pres ed so much in the choice of certain proj- also a direct use. If it is carried out in several
ects as in the education of planners and de- variants under alternative economic assump-
cision-makers. Ithas accustomed preparers tions and maybe alternative numerical hy-
of decisions concerned with projects, and potheses, itmay facilitate the many-sided
planners in general, as well as specialists of comparison of the advantages and disadvan-
international organizations, to examine each tages of projects to be considered. No one
project within comprehensive social interre- unique salutary cost-benefit indicator is nec-
lationships: toconsider not only direct prof- essary. Instead, a set of indicators which de-
itability counted at current prices, but to try scribe indetail the economic and social con-
to examine thoroughly the whole series of ex- sequences ofthe choice should be fixed in
pectable direct and indirect effects. What advance. The indicator vector should not be
will be the consequence of the project from too big. I shall mention only for illustration
the aspects of employment, foreign trade, that 15-30 indicators are already apt to dem-
and, in general, the economic growth of the onstrate effects widely and in detail. In a pro-
developing country? Cost-benefit analysis ject appraisal, the numerical value of these
virtually develops in those who practice it indicators must be determined. Their deter-
"conditioned reflexes" to such complexity of mination and comparison should be the main
analysis. I see its main future role in this ed- task of cost-benefit analysis (more exactly, of
ucational and disciplinary effect. A carefully the cost-benefit analysis interpreted in a
carried out cost-benefit analysis enforces the wider sense in accordance with the study).
regular collection of data in connection with Within these limits, cost-benefit analysis may
all positive and negative effects, and thus successfully promote the tasks of planning:
supplies rich material for decision prep- cognition and reconciliation of interests.
aration.

Comment
The literature on the general topic of planning and the price system is extensive, but the
following references may be cited for the particular problem of assessing the relative merits
of the market mechanism and planning in the context of development: Hugh C. J. Aitken
(ed.), The State and Economic Growth (1959); P. T. Bauer, Dissent on Development (1972),
chapter 2; A. K. Cairncross, Factors in Economic Development (1962), chapter 19; Elliot J.
Berg, "Socialism and Economic Development in Tropical Africa," Quarterly Journal of Eco-
nomics (November 1968); Maurice Dobb, An Essay on Economic Growth and Planning
(1960), chapter 1; A. O. Hirschman, "Economic Policy in Underdeveloped Countries," Eco-
nomic Development and Cultural Change (July 1957); E. S. Mason, Economic Planning in
Underdeveloped Areas (1958); R. M. Solow, "Some Problems of the Theory and Practice of
Economic Planning," Economic Development and Cultural Change (January 1962); Ray-
mond Vernon, "Comprehensive Model-Building in the Planning Process: The Case of the I ess
Developed Countries," Economic Journal (March 1966); Gustav Ranis, "Planning for Re-
sources and Planning for Strategy Change," Weltwirtschaftliches Archiv (Summer 1965); H
B. Chenery, "The Structuralist Approach to Development Policy," American Economic Re-
view, (May 1975); Charles Blitzer et al. Economy-Wide Models and Development Planning
(1975); M. Cave and P. Hare, Alternative Approaches to Economic Planning (1981 ).
Outstanding general introductions to development planning are provided by W. Arthur
Lewis, Development Planning (1966); Jan Tinbergen, Development Planning (1967); K.( 1 97 Grif-
fin and J. Enos, Planning Development (1971); M. P. Todaro, Development Planning 1 );
Mike Faber and Dudley Seers (eds.), The Crisis in Planning (197 7): G. Ranis (ed.). Govern-
 722 DEVELOPMENT PLANNING AND POLICYMAKING

merit and Economic Development (1971); G. Pyatt and E. Thorbecke, Planning Techniques
for a Better Future (1976).
An excellent survey article on development policymaking and planning is D. T. Healy, "De-
velopment Policy: New Thinking about an Interpretation," Journal of Economic Literature
(September 1972).
 XI.B. POLICY MODELS
XI.B.1. Planning Models*

Models are needed to analyse the interaction The former type provides a quantitative test
among elements in a social system. A purely of theoretical assumptions about the behav-
theoretical formulation allows some useful iour of key elements, such as the rate of sav-
deductions to be made as to the direction of ings, the requirements for capital in each sec-
changes in specified variables in response to tor, the response of agricultural producers to
particular policy measures. In this way we changes in prices, etc. The latter combines
can predict that a devaluation of the cur- these separate estimates into a coherent
rency will tend to raise exports and lower im- framework for planning. Planning models
ports of each commodity. We cannot deter- that are not based on statistical estimates
mine the magnitude of these changes without must be used with great caution. It is not the
quantitative estimates of the more important use of models per se but the tendency to take
structural relations involved in this model. the results of an inadequately tested planning
Intuitive planning is based on guesses as to model too seriously that has brought some
the orders of magnitude of the parameters in disrepute to formal planning in several
the model that most directly affect a given countries.
economic response. Such guesses are derived
either from similiar situations in the country Econometric Models
in the past or from the experience of other
countries. Intuitive planning includes the use Probably three-quarters of the overall
of statistical trends and simple accounting re- quantitative analysis of underdeveloped
lations to test the balance of supply and de- economies has been carried out in the past
mand for different commodities. This com- dozen years. The rapid increase in statistics
mon sense approach is likely to be most valid on sectoral output, employment, consump-
for short-term projections for which the un- tion, investment and national income has
derlying relations do not change greatly. made possible a growing body of comparative
Greater rigour has been introduced into studies of key aspects of the economic struc-
planning in the past decade by a more sys- ture. These studies provide a basis for testing
tematic use of statistical information both in development theories, rejecting some hy-
testing the underlying theories and in apply- potheses and refining and extending others.
ing them to particular problems. This proce- For example, capital formation has been
dure involves two different types of economic shown to have greater importance in explain-
model: ing the sources of growth in developing coun-
1 . An econometric model, which is used as a tries than in advanced countries in a number
basis for estimating past structural of recent studies. Conversely, there are no
relationships, known cases of rapid growth being achieved
2. A planning model, which is used to deter- in underdeveloped countries through techno-
mine the relationships between desired so- logical change not accompanied by substan-
tial investment. Studies have been made of
cial objectives and the policy instruments
that are proposed to achieve them. other basic elements of planning models.
such as the determinants of savings, the re-
♦From Hollis B. Chenery, "Notes on the Use of quirements for skilled labour, the price and
Models in Development Planning," in M. Faber and income elasticities of demand for major com-
Dudley Seers (eds.), The Crisis in Planning, Chatto
& Windus, 1972, p. 129-34. Reprinted by modities, etc. The next ten years will doubt-
permission. less see a great increase in econometric stud-

723
 724 DEVELOPMENT PLANNING AND POLICYMAKING

ies, which furnish the building blocks from competitors for limited resources and as sup-
which planning models can be constructed. pliers ofdesired goods and services.
The main limitation to the use of econo- The simplest planning models that can per-
metric models has been the short period of form this function are derived from Leon-
time for which usable statistical information tiefs input-output approach to describing a
is available for their estimation. There are system of general interdependence. Experi-
probably twenty underdeveloped countries ments have been carried on with this type of
for which the major series are available since model in most of the developing countries
1950 and another twenty or thirty for which having any interest in formal planning. Ini-
they were started in the following decade. In tially, these experiments were designed to
the next five to ten years, it will be possible test the possibilities for constructing such a
to carry out more extensive tests of develop- model and to train a few technicians in its
ment theories by using a mixture of cross- use. Since any comprehensive model requires
country and time series analysis covering pe- estimates of a large number of economic re-
riods long enough to have considerable statis- lations, these early experiments were often
tical validity. The availability of such studies based on rather rudimentary statistics. Solu-
will help the individual planner by providing tions to such models are of value mainly in
him with reference points to judge estimates illustrating the interdependence of develop-
for his own country. ment policies in general terms.
Despite its conceptual simplicity, the
Planning Models input-output framework has the great virtue
of forcing the analyst to consider the working
A planning model specifies the relation- of the economy as a whole. More complex
ships between the goals of the society and the formulations of individual demand and sup-
instruments that the government has avail- ply conditions can replace the initial crude
able to achieve them. The main advantage of approximations, and sectors can be further
a more formal analysis as compared to a disaggregated as data become available. In
more intuitive approach lies in the ability of this way a dozen or more countries have now
the former to determine the indirect effects of developed a sufficient empirical base and
given policies. However, the intuitive ap- enough experience in the use of intersectoral
proach can take some account of variables models so that they can be considered a prac-
such as administrative competence, political tical planning tool. In the next decade several
will and other factors that cannot be readily dozen more countries are likely to develop a
quantified. I assume, therefore, that both ap- usable inter-industry framework for their
proaches will be used and will merely illus- plans, while leaders in this field will have pro-
trate the areas in which formalization is par- gressed to the use of more complex represen-
ticularly needed. tations ofthe economic system.
The growth of poor countries depends at
least as much on their ability to reallocate re-
sources to new uses and thus to change their
economic structures as on an increase in total THE CONTRIBUTION OF
PLANNING MODELS
investment and other resource inputs. The in-
struments of development policy are more Although the past decade has been largely
likely to be specific to particular sectors of devoted to research and development on for-
the economy than is typically the case in ad- mal planning techniques, some practical re-
vanced countries, where less change is needed sults have already been achieved. The follow-
and resources move more readily in response ing examples illustrate the type of problem
to market forces. A formal model of resource for which quantitative models can provide a
allocation focuses on the interaction among notable improvement over more intuitive
different types of productive activity both as methods.
 725 POLICY MODELS

Protection and Import Substitution for future needs for specific types of skill and
One of the most common methods of stim- without comparing the return on investment
in human capital to other forms of
ulating industrial growth and reducing im- investment.
port requirements is to give tariff or quota
protection to new industries. The level of A number of countries have recently ex-
such protection is set in relation to the cost perimented with a variety of models designed
structure of the domestic manufacturer and to integrate educational policy into general
the cost of imports. Indirect effects on the planning procedures. Rates of return to dif-
ferent levels of education have been com-
rest of the economy are typically ignored in
puted from simple econometric models of
this pragmatic approach to policy.
production and used to determine the types
A number of countries have recently as- in which expansion is most desirable. In more
sessed the cumulative effects of protectionist
comprehensive studies, skilled labour is
policies by means of interindustry analysis.
treated as an input needed for production in
Since the protection of an industry increases each sector, and future needs for education
the cost of its products to its domestic cus-
are derived as part of an overall projection of
tomers, itis necessary to trace the effects of the economic structure. The results obtained
protection throughout the economy in order
to find out the total economic costs and ben- are often quite different from those of intui-
efits of this policy. This has been done by tive planning procedures, which tend to pre-
scribe the same educational structures for all
measuring the total cost of domestic re-
countries at a given income level.
sources used throughout the economy to earn
or save a dollar of foreign exchange in differ-
ent types of production. Such studies typi- Employment Versus Other Objectives
cally show that:
One area in which both planners and pol-
1. The resource cost of protection is much icy makers have failed to produce satisfac-
greater than is indicated by the nominal tory results is in the creation of new jobs.
tariff. There is notable inconsistency between the
2. There is a large and irrational variation asserted social goal of achieving full employ-
among sectors of industry and agriculture. ment and the policies that governments
3. There is usually a systematic bias against adopt. The latter are designed primarily to
exports and in favour of import substitu- increase total income within the limits of
tion in industry as a means of earning available investment and balance of pay-
foreign exchange. ments constraints. Employment creation does
Once this type of analysis has been made, not feature prominently in most development
a planning version of the model can be used strategies and is usually assumed to be a cor-
to redesign the system of tariffs and subsidies ollary of high growth rates.
so as to come closer to the ideal of equating The present generation of planning models
the productivity of marginal resource uses in does not deal with this problem very well be-
different areas. Applications of this kind have cause the basic econometric and technologi-
been made in Israel, Pakistan, India, Korea, cal information on alternative factor propor-
Chile, Argentina and elsewhere. tions is not yet available. Existing models
therefore tend to assume fixed relations be-
tween labour inputs and commodity output
Educational Planning
and to ignore the possibility of substitution
Education is another aspect of resource al- either in production or by changing the com-
location which is difficult to analyse without position of consumption. This overly rigid
taking account of repercussions throughout representation of the economy can be readily
the economy. Until recently, education policy modified, however, once the empirical basis
was typically determined with little regard for doing so has been developed.
 726 DEVELOPMENT PLANNING AND POLICYMAKING

In an economy that is threatened with ex- aggregate model and using them as a basis
cess labour as well as balance of payments dif- for input-output solutions. These are com-
ficulties, an individual investment project pared to the sectoral studies and the assump-
should be judged by its contribution to at tions of each adjusted until they are mutually
least three objectives: consistent. This procedure results in a cons-
sitent family of models, each of which is
1 . Raising the national product,
suited to a different type of policy analysis.
2. Increasing employment,
In this way, the indirect effects of alternative
3. Earning or saving foreign exchange.
policies at either the sector or the economy-
The logical framework for making this wide level can be worked out.
type of assessment is that of linear (or non- An overall model that incorporates the
linear) programming, which determines an main elements of alternative strategies can
optimal allocation of resources for the whole be particularly useful in the early stages of
economy under specified limitations. Solu- plan formulation. It should first be adjusted
tions to such a model provide a basis for to include elements of intuitive judgement as
weighing the criterion of employment crea- to the range of political and administrative
tion against other benefits and costs. The feasibility of alternative policy measures.
Feasible solutions can then be determined
"shadow prices" of a linear programming
model provide some of the needed links be- under a series of alternative assumptions as
tween sector plans and project evaluation in to the principal objectives. Tests can also be
different parts of the economy. made of the effect of alternative values of ex-
Economy-wide linear programming models ogenous variables, such as export demands or
have been constructed for this purpose weather conditions. This exploration of alter-
in at least a dozen countries. While most natives isa pragmatic aporoximation to the
of them are still in an experimental stage, formal optimization procedures of mathe-
they offer a promising approach to the recon- matical programming.
ciliation ofemployment creation, foreign ex- The calculation of feasible alternatives
change savings and other objectives of devel- provides a practical basis for the discussions
opment planning. between planners, administrators and politi-
cians that are necessary to formulate a de-
PLAN FORMULATION velopment plan. Political judgements can be
elicited much better by considering concrete
The formulation of a development plan re- alternatives rather than abstract goals. Since
quires that a number of separate studies of each policy has a cost, the specification of
individual sectors and types of policy be rec- goals and limitations can hardly be made in
onciled ina consistent framework. This can the absence of some knowledge of compre-
be done by starting from the projections of an hensive alternative allocations of resources.

XI.B.2. A Model of
Development Alternatives*

Current growth models have serious deficien- cusing on the savings-investment relationship
cies as a basis for development policy. In fo- and the possibilities of substitution between
capital and labour, they exclude questions of
*From H. B. Chenery and M. Bruno, "Develop- equal concern to policy makers, such as the
ment Alternatives in an Open Economy: The Case of
Israel," Economic Journal, Vol. 77, No. 285, 1962, changing structure of demand, the role of
pp. 79-103. Reprinted by permission. foreign trade and the allocation of resources.
 727 POLICY MODELS

As a result, formal growth theory fails to

In Tinbergen's approach the objective vari-
clarify the relations among the several instru- ables are taken as given; the problem is to
ments of development policy, which should find that best combination of values for the
be one of its major functions.
instrument variables. Theil's approach is
As Tinbergen (1956) has taught us, a pol- more general. He assumes that the social
icy model should contain variables reflecting welfare depends on values of both instrument
the economic goals of the society and the and objective variables. The optimum pro-
main instruments of government policy, and gramme isthat which produces a maximum
it should specify the more important struc- of welfare consistent with the restraints on
tural relations connecting them. The degree the system. In the absence of a quantifiable
of complexity required of a policy model for welfare function, however, the practical
a developing economy is suggested by the fol- problem of setting values for the objective
lowing list of typical objectives, instruments variables still remains.
and structural limitations. Since a single optimum programme cannot
be determined by economic analysis alone
1. Objectives: maximum income, full when there are several objective variables,
employment.
2. Instruments: capital imports, tax policy, our approach will be to establish a set of al-
ternative programmes that includes the fea-
trade policy, investment allocation. sible degree of variation in all of the relevant
3. Limitations: composition of demand, bal- variables. In cases where the Government has
ance of payments, labour supply and established a fixed policy goal, such as full
requirements, capital supply and
requirements. employment or a specified defence expendi-
ture, we follow Tinbergen in setting this as a
The omission of several of these instru- fixed objective of the programme. Otherwise
ments and limitations may make the use of we follow Theil in considering a range of val-
simpler models seriously misleading. ues for the objective variables. In this second
category are consumption, the total produc-
tive capacity of the economy and the foreign
REQUIREMENTS OF A POLICY debt. These together are assumed to deter-
MODEL mine the social welfare . . .
In order to determine the range of feasible
The functions of a policy model are to de-
termine consistent sets of economic policies programmes we include two types of con-
trolled variables. The first are instruments of
and to facilitate the choice among them. The
nature and results of development policies government policy, as defined by Tinbergen;
they are subject to more or less direct control
are represented by the variables in the model.
The model must also include in some form by the Government, as in the case of the ex-
change rate or the level of foreign borrowing.
the principal elements affecting the rate of
The second type may vary within limits sci
growth, which were discussed in the preced- by institutional factors, but may or may not
ing section.
be directly influenced by government policy
Tinbergen (1956) and Theil (1958) have The institutional limits must be included in
distinguished several types of variables in
policy models: determining realistic programmes, however.
These four categories are not mutually ex-
1. Predetermined or exogenous variables. clusive; an instrument variable ean also be an
2. Instrument variables (i.e., those subject to objective or be subject to institutional limits.
government control). for example. Several combinations are shown
3. Objective variables (i.e., those reflecting in the following classification of the polic>
the aims of policy). variables that will be included in the model
4. Other endogenous variables (i.e., those [see Table 1].
that are irrelevant for policy analysis). The institutional limits provide a erudc
 728 DEVELOPMENT PLANNING AND POLICYMAKING

TABLE 1 . Classification of Policy Variables

Policy

Objectives Institutional
Variables Fixed Variable Instruments Limits

1. Gross national product ( V) X

2. Private consumption (C) X
3. Public consumption (G) X
4. Foreign capital inflow (F) X X X
5. Unemployment rate (u) X
6. Savings rate (s) X X
7. Exchange rate (r) X
8. Rate of increase in labour productivity (/)

substitute for a more complete welfare func- by adding those relations that prove to have
tion, since they can be used to exclude values a significant effect.
of any variable that are clearly in conflict The idea of separate and conflicting limits
with welfare maximization in the particular
to growth is a basic element of Harrod's
society.1 They can also be used to allow for (1939) pioneering work. Although he is pri-
uncertainty as to the nature of a particular marily concerned with the cyclical aspects of
structural relation, as in the determinants of differences between the limits set by the sup-
increased productivity. ply of capital and the supply of labour, his
Since many factors affect growth, some of relations can be reinterpreted as a simple pol-
which can only be adequately represented in icy model of development alternatives. As is
a multi-sectoral model, it is useful to divide shown in equations (13) and (14) below, the
the analysis into two parts. To start with, an Harrod model thus interpreted contains two
aggregate model can be used to determine equations, corresponding to the supply-de-
the main development alternatives. The most mand balances for capital and labour. If all
promising of these can then be subjected to a the parameters are fixed the maximum rate
more detailed analysis, which is not feasible of growth will be determined by one of the
until the range of possibilities has been nar- two equations, and either labour or capital
rowed down. The detailed results can in turn
be used to revise the initial estimates of the will be in excess supply.2 In a policy model,
however, some of the parameters become
aggregate model. variables, and these equations determine the
The main problem in designing an aggre- value of any two variables, as, for example,
gate model for this purpose is to identify in the savings rate and growth rate at full
advance the factors that may prove to be ef- employment.
fective limits to growth. When a particular It has been argued above that there is a
restriction — such as the composition of de- third general limitation to growth on a par
mands— is omitted from a model it is implied with the two considered by Harrod: the bal-
that whatever changes take place in this ele- ance of payments. As a policy problem, the
mentwill not significantly affect the param- balance-of-payments limitation is quite sim-
eters in the model. In some cases it will be ilar to the savings-investment limitation. An
necessary to subject this assumption to a increase in the rate of growth often requires
quantitative test in order to determine its va- a change in the structure of income use in
lidity. The model can be built up in this way order to reduce the proportion going to con-
'Instead of treating the rate of unemployment as a 2In Harrod's terminology the labour-determined
fixed objective, we might assign institutional limits solution gives the "natural" rate of growth and the
to it.
capital-determined solution the "warranted" rate.
 729 POLICY MODELS

sumption and hence to increase savings. It Exogenous (Predetermined) Variables:

may also require a change in the structure of Initial values of all variables
production in order to reduce the ratio of im- t Time
ports to total output. It is not clear a priori,
either in Israel or in other countries, which of Pji Export price in sector i
KQ Initial unused capital stock
these structural relations is more likely to
/„ Final unused capital stock
limit growth or which is harder to change.
The parallelism between the two is completed In its initial form, the model consists of
by the fact that a foreign capital inflow plays twelve equations, of which seven describe the
a dual role in adding to both investment and structure of the economy, three specify the
foreign exchange resources. resource limitations and two are definitional.
The model is later reduced to four equations
STATEMENT OF THE MODEL in eight variables by eliminating the eight ir-
relevant endogenous variables — all except V
The model of development alternatives and C. A development programme can then
proposed here incorporates these three limits. be specified by assigning values to four of the
They are described by means of ten endoge- variables and determining the values of the
nous variables, including the two variable ob- remaining four from the model.
jectivVes ( and C). The variables used in the Since it is assumed that decisions on de-
model are as follows:3 velopment strategy can be based on the val-
v ues of the policy variables at the end of a
t
Endogenous (Uncontrolled) Variables: planning period of n years, solutions are only
c,I needed for a single period. The equations are
t Gross national product4 presented first in general form; estimates of
Private consumption4 the parameters are then given for the Israeli
Rt Total investment net of replace-
ment economy for a five-year planning period of
Et 1959/60 to 1964/5. The aggregate model
Replacement
Mt Exports of goods and services will be supplemented by an inter-industry
st Imports of goods and services analysis in order to take account of the com-
Kt Gross domestic savings position ofdemand and of supply limitations
Nt Total capital stock in estimating the parameters.
Labour supply The aggregate production function. Al-
u Labour demand though the Harrod-Domar model has been
criticized for its omission of substitution be-
n,
Instrument and Controlled Variables: tween labour and capital, it is generally rec-
ognized that substitution can only take place
Gt Government current expenditure4 over a period of time and depends to a large
Ft Foreign capital inflow4 extent on the installation of new equipment.
We shall therefore treat the labour-capital
u, Unemployment rate4 ratio as a function of time, but shall assume
that both inputs are required in fixed propor-
l ty tions at any moment. For simplicity, a trend
s Margina propensi to save I — — will be assiciated only with the labour input,
r Effective exchange rate which is consistent with the experience of the
/ AS
Annual increase in labour (pro-
ductivity past decade.
With complementary inputs, output is lim-
ited by whichever one is exhausted first. In
Israel, as in most of the less-developed econ-
3The subscript t refers to the year. In the base year omies, this factor is more likely to be capital.
t = 0, and in the final year of the planning period
/ = n We therefore write the production function
AVt, C„ G„ F, and u, are also objective variables. as dependent on the stock of capital and the
 730 DEVELOPMENT PLANNING AND POLICYMAKING

effectiveness of its utilization:

cient \ir therefore represents the total imports
required directly and indirectly per unit of
Vn = V0 + p(Ko ~ Kn) each type of demand. The future import pro-
+ 0(Kn - K0) (1) portions are derived from the anticipated ef-
where (3 represents the average product per fective exchange rate, r. Use of a higher ex-
unit of increase in the capital stock. The sec- change rate will result in additional import
ond term of this equation allows for the pos- substitution and a fall in the import
sibility ofreducing the level of excess capac- coefficients.6
ity, K, which is a significant factor in Israel Replacement. The replacement of capital
and elsewhere. depends on the age and estimated life of the
More important than the direct substitu- capital stock in different sectors of the econ-
tion between labour and capital is the effect omy. In aggregate form, this relationship
of a change in the composition of output on may be indicated by:
capital requirements. This will be allowed for Rt — R,(K,Kt-\Kt.
by estimating 0 from an inter-industry anal- (4)
ysis so that it becomes a weighted average of although in practice Rt is estimated on a sec-
the output-capital coefficients in each sector. tor basis.
Any substantial departure from the assumed As Domar (1957) has emphasized, the re-
composition of the increase in output will placement ofworn-out equipment is substan-
tially less than the conventional allowance for
therefore require a recalculation of @.5
Labour demand. This demand for labour depreciation in a growing economy. We
depends on the level of output and the in- therefore define net investment /, as gross in-
crease inaverage labour productivity: vestment less replacement. Part of net invest-
ment iscovered by the excess of depreciation
X0(l - 1)% (2) allowances over actual replacement.
=
where X0 is the Ltaverage labour input per unit Savings. Gross domestic savings depends
of output at the beginning of the period and on the level of per capita income, its func-
/ is the annual rate at which it decreases. The tional distribution and the Government's tax
estimate of / should also take account of the policy. In the absence of an adequate basis
anticipated composition of output. As esti- for estimating the separate effects of these
mated statistically from past trends, / in- factors, an aggregate relation of the follow-
cludes effects of both substitution and tech- ing form will be assumed:
nological change.
Sn = S0 + s(Vn- V0)
Since we assume complementarity be- (5)
tween capital and labour, equation (2) may where the marginal propensity to save out of
be restated as a production function by solv- increased GNP (s) is taken as an instrument
ing for Vt if labour should become the factor variable. It represents the combined effects of
limiting growth. tax policy, changes in income distribution
Import demand. The demand for imports and other policy measures that affect savings.
depends on the five components of total de- Labour supply. The supply of labour is de-
mand: termined from the natural increase of the ex-
isting population plus net immigration. For
M, = n'cC, + nrfit + rf(It simplicity, the combined result is assumed to
+ Rt) + i£Et (3) take an exponential form:
Each import coefficient \ir is derived from a Nt = N0(l + 7)' (6)
solution to an input-output model containing
specified proportions between domestic sup-
plies and imports in each sector. The coeffi- 6Import substitution as a result of a higher ex-
change rate may also produce a lower productivity of
^he coefficient 0 similarly depends on the sectors capital, but this can only be taken account of explic-
in which excess capacity is reduced. itly in more detailed models.
 731 POLICY MODELS

Exports. The level of total exports is the Gross national product

sum of the individual commodities and ser-
V,= Ct+Gt + /,
vices exported; each is assumed to depend on
the effective exchange rate and on foreign + R, + E, - M, (12)
prices: Reduced form of the model. The reduced
(7) form of a policy model is a set of equations
Et = Y. E^> Pei. t) involving the policy variables only, with all
The time period is also assumed to affect the irrelevant endogenous variables elimi-
achievable export levels because of the time nated.8 Inthe present case we eliminate eight
needed to penetrate new markets and to variables and eight equations, so that the
establish export organizations for new model is reduced to the following four equa-
tions in eight variables (the terms in square
products. brackets are constants):
Savings-investment equilibrium (13)
St + Ft-I,+ Rt (8) (1 ~ u)
Balance-of-payments equilibrium NoO + 7)'
Xo J (1 " 0"
Mt = Et + Ft (9) „ _ [p/0Vq + S0- R„) - sV0 + Fn
where Mt, Et and Ft are all measured at con-
(1 -Pne)En
/P --S
+ (1 vJF± (14>
stant domestic prices.7
Employment equilibrium
y + 0*c ~ »i)Gn + [JH, ~ Hc)(p/(3Vo ~ (15)
Rn)]
He + (M, ~ Hc)p/0
Lt = (1 - u)Nt (10)
where the proportion unemployed (u) is an
objective variable. C+ G = (1 - s)Vn + (s - s0)V0
Total net capital formation where
t = n-\
S0 = s0Vo and V0 = M«o ~ *„) + V0 (16)
]T
r = 0 /, = (K„-K0) (11)
Equations (13), (14) and (15) correspond
In order to express the model in terms of ini- to the three equilibrium conditions for la-
tial and final-year values only, an approxi- bour, capital and foreign exchange. When no
mation ofthe form limit is placed on F the labour-force equation
(13) provides the ultimate limit to growth be-
/„ = p(Kn - K0) (Ha) cause the other two equations can be satisfied
may be substituted for equation (11); p de- by increased foreign borrowing. With a given
pends on the rate of growth of investment F, there are three separate limits.
and the length of the planning period. Some Equation (14) corresponds to the Harrod-
assumption of this sort is needed to make /„ Domar equation, as can be seen by assuming
determinate. a one-year period, no excess capacity, and
equal marginal and average (net) savings
rates. The equation then becomes:
7The foreign capital inflow is normally given in cur- AK n nF
rent-value dollar terms. If we denote the latter by
Ft we have F, = rF',/p/, where r is the exchange (14a)
rate and pf is an implicit price index defined by the
expression
pmMJ-p1E1 or
"> = M,-e, Vo + PEn
1 - (3s (14b)
pm and pe are the respective import and export prices
(indices) on foreign markets. In the Israeli case im-
port prices are assumed constant and export prices are 8The concept isVndiscussed
= by Theil (1958, chapter
7) under somewhat different assumptions.
assumed to fall. Hence p/ > 1 and F, < rF'v
 732 DEVELOPMENT PLANNING AND POLICYMAKING

The meaning of the balance-of-payments References

limit can be clarified by~ assuming
M that all of
the import coefficients /u/ are equal and again Domar, E. D., "Depreciation, Replacement
taking a one-year period. The result is: and Growth, and Fluctuations," Economic
Journal (December 1957).
(En + Fn) (15a) Harrod, R. F., 1939, "An Essay in Dynamic
1939).
Theory," Economic Journal (March
Vn =
Theil, H., 1958, Economic Forecasts and Pol-
icy, North-Holland Publishing Co.
Tinbergen, J., 1956, Economic Policy: Prin-
ciples and Design, North-Holland Publish-
ing Co.

XI. B. 3. Policy Instruments and

Development Alternatives*

There has been little systematic analysis of lization for war. The suitability of various in-
the relative merits and defects of the policy struments for the latter purpose has been
instruments available to underdeveloped widely discussed, and the experience of the
countries. Since control of international trade United States and other countries has been
is administratively simpler than many other analysed in some detail. A similar study of
types of policy, there has been a tendency to the actual effects of development policies is
rely heavily on it as a way of influencing the needed before very firm recommendations
pattern of domestic production, without rec- can be made to the underdeveloped countries,
ognizing the drawbacks to exclusive reliance but some general comments may be in order.
on this set of instruments. Colonial areas
have been forced to devise other measures,
since protection was denied to them, but they CHARACTERISTICS OF
have rarely pursued overall development pol- INSTRUMENTS
icies. The need for a greater variety of mea-
sures to promote development has now been Policy instruments may be classified in
widely recognized, but there is still inade- various ways: by the sectors of the economy
quate consideration of the range of alterna- on which they operate, by their use of prices
tives available. or quantities as variables to be manipulated,
In its need to change the pattern of re- by the extent to which they can be effectively
source use over a relatively short period of controlled by the government, by the effect
time, the promotion of development resem- that they have on private incentives and free-
bles (in lesser degree) the problem of mobi- dom of choice, etc. In Table 1 representative
instruments are classified according to the
*From Hollis B. Chenery, "Development Policies extent of their application (general versus
and Programmes," Economic Bulletin for Latin
America, Vol. 3, No. 1, March 1958, pp. 55-60; "A specific) and their mode of operation
Model of Development Alternatives," Vol. 8, United (through prices or quantities). The general
States Papers Prepared for the United Nations Con- instruments act on broad aspects of the econ-
ference on the Application of Science and Technology
for the Benefit of the Less Developed Countries, omy— the money supply, the government
Washington, D.C., 1962, pp. 82-6. Reprinted by budget, investment, consumption — and are
permission. widely used in developed and underdeveloped
 733 POLICY MODELS

countries alike. The specific instruments are which will improve on the working of the
applied differentially to individual sectors of competitive economy without losing the ad-
the economy, as illustrated by subsidies, tar- vantages of private initiative and the auto-
iffs, or Government investment. matic adjustment of the price system.
To achieve a given effect on production, or In designing policies for specific sectors,
use of any commodity, there is a choice be- there is an argument for using price rather
tween controlling a price and controlling a than quantity instruments which is based on
quantity. In this respect, tariffs are an alter- reasoning similar to the case for general over
native to quotas, differential interest rates are specific instruments. Taxes and subsidies dis-
an alternative to capital rationing, and sub- tort the choices open to producers and users
sidies to private producers are an alternative of a commodity less than do allocation sys-
to production by the government. These mea- tems or other quantitative restrictions and
sures differ in their effects on prices and con- hence are conducive to greater flexibility and
sumer choices, in administrative convenience, over-all economic efficiency. Furthermore,
in the predictability of their results, and in the administrative requirements for price in-
other respects. A choice between quantity tervention ofthis type are generally less than
and price variables as instruments must for quantitative controls.
therefore be made by balancing the advan- Despite the general case in favour of using
tages and disadvantages in each case. the price system, there are several situations
Some of the main issues of economic policy in which quantitative measures may be
are concerned with the choice between gen- needed:
eral and specific instruments and between i. When it is necessary to limit consump-
using prices and quantities as control vari- tion of an essential commodity in short sup-
ables. There is a strong case to be made for ply (e.g., imported goods), the tax needed to
using general instruments rather than spe- bring about a given reduction in use might
cific ones. The rates of interest, taxation, and result in such high prices that the burden of
exchange are the orthodox means of exerting the reduction would fall on lower income
Government influence in a laissez-faire econ- groups. In this case, price controls and ra-
omy. Their immediate objectives are stability tioning may be preferable on welfare
in prices and the balance of payments and the
prevention of unemployment. Growth is left grounds.
ii. Where a minimum increase in produc-
to free market forces. The manipulation of tion is essential to production in other sec-
interest rates and exchange rates allow mar- tors— as in the case of power, transport and
ket forces in each sector to determine where various auxiliary facilities — the price needed
expansion or contraction of production and to ensure adequate private investment may
consumption will take place. These instru- be too high or the response of private inves-
ments therefore interfere less with the tors too uncertain. In this case, quantitative
choices of producers and consumers than do measures, such as Government investment,
measures which discriminate by sector. They may be more efficient because the cost to the
also require a less detailed analysis for their society is less or the outcome more
use and do not substitute Government judg- predictable.
ment of what is desirable for the action of iii. In general, where controls are needed
market forces. for only a short period, as in the case of tem-
The need for specific instruments to sup- porary shortages, it may be desirable to al-
plement general measures derives from the locate supplies to more essential uses rather
deficiencies in the price mechanism which than upset the general price structure and
apply primarily to specific sectors of the distort investment decisions by allowing
economy. When these factors prevent the prices to rise. Quantitative measures are also
achievement of a satisfactory rate of growth, likely to have more predictable effects in this
the problem is to devise policy measures case.
 s '5 •- 2^
<u
o 3 >>
E 3
o
o
8 "S, o X)
£3
E <l> a.
CA3 C
H 3
CA
S£
§ Oh S -C 2 m U4 U J O
00

.2
J3
60
ed T3
c
3 o
a
V)
2
1
Cfl
<3
13
J
00
o c
C/i c (A
c
3 c
1—
0
c 2
a
CO 2
Q. c
£
o b
o 0
UJ

c'£ £^

a
OQ 13
<D O o c
f O
^ ^ ' "* roj ^ N /■ N y^^S y^-^. v^^\ -x 3 «

£ -a

a c ^ SI
a a o
o c
x «

I I
3
t/5

o
c 2
00 <u
U

734
 "3
"S,

c «2 « 3
5
pro

CA <L> •- o c
u
duc

t- o >i
pro

T3 .9
•~ CA
6
tio of

o w
duc

2 «
8-8
JO It*

inv
o-
O 'C
n t

3 £
t io

to
^ — T3 >

est
onf

CA
— . o

men
o
el

B o-

t
.a >
cd
O
J
— < Q
(O o
5
of el

> CA > Oh
^ -J
cs ~^ <N <N
<u o <o
J O J 0-
a^

x
IS
2
S 5
ed C >< o
2 u
13 CA
2 8 ,S C "o
c
3 u
8 11 c3
E
I8
ca •o
cB ,tj
« -5o +- c
>i> ><L> «O i: <t> -o
Oo Oo ~ 8.-S
fa rt
<U tg

c
1e c c

1
o o u u
o S
CA
E O
C
w 't
8
•a

A
E
e
M
c3 oO
1o.
C T3
>
CA
.o >> 8 E >4>
C U

CA o ca
oa
* 6
BB
c

« s 8o
o
5 c o c
en c
x> o o C3

5 « .2 c 3cd n
<£ 8 X> u T3 (C O _ <y
B
o- o 2 2 o
c <*.
a. <*-
P >
£.5 c
Oh £
cs -J £ 4>
t/5 -f
—oO
Oh n3
/— ■- /—<N -.
— o U
ocd
«N
Oh
^-v
pi
-— ^
<N
"»—• ^— • X) rT
U

Is-
"3 .2
#o
g s
CA

X t3
_c
J 5
CA
c
3cd
•- 6 u CA

5 "o *- U
1 1/5

3
CA
C
o
c5 o
«2
3-2
x a 1 X)c3 0X
3
en ^0> H
Hcd 'C
Oh 2 ° X*

735
 736 DEVELOPMENT PLANNING AND POLICYMAKING

In these examples, it is the dynamic ele- principles given in the previous section. In
ments inthe situation and the deviation from cases of extreme shortage of foreign ex-
a desirable income distribution which provide change, tariffs (or devaluation) may be too
the principal arguments for using quantita- uncertain in their results and quotas or ex-
tive measures of control. change restrictions may be adopted as emer-
gency measures.
SPECIFIC MEASURES FOR The effect of quantitative restrictions on
INVESTMENT ALLOCATION investment in domestic substitutes for im-
ports or in sectors using imported commodi-
Although the specific measures listed in ties isgenerally less certain than that of tar-
Table 1 affect both current production and iffs. Allocations are subject to variation
the allocation of investment resources, it is according to the amount of exchange availa-
the latter aspect that is crucial for the future ble, and the profitability of domestic produc-
course of development. The various instru- tion is harder to determine than in the case
ments affect investment decisions through of a tariff.
the availability and cost of primary inputs As instruments for inducing investment in
(labour, natural resources, imported com- new types of production, subsidies may be
modities); through the supply of inputs from preferable to either quantitative restrictions
other sectors (raw materials, overhead facil- or tariffs because the price is not raised above
ities); through the demand for output (sales the level of world prices. Total demand is
taxes, export subsidies); through profits therefore greater and using sectors are not
(taxes, subsidies); and through measures di- penalized in export markets. The cost of this
rectly related to the process of investment technique in Government expenditure must
(interest rates, capital rationing, restrictions be weighed against its benefits, however.
on entry, direct Government investment). Protection from foreign competitors is only
There is therefore a considerable variety of one factor in the expansion of domestic pro-
choice between quantity and price instru- duction. Also required are entrepreneurs,
ments and among measures more or less di- capital, skilled labour, raw materials, etc.
rectly related to a particular investment. When some of these are lacking, the restric-
The a priori arguments concerning some of tion only serves to reduce imports and raise
the principal measures for influencing invest- prices to consumers. Trade restrictions are
ment decisions run somewhat as follows: therefore a rather uncertain method of di-
1. Measures of Protection. As indicated recting investment unless combined with
earlier, protective devices are perhaps the other measures affecting factor supply, and
most common instruments for influencing the they frequently have undesirable secondary
pattern of investment. For this purpose, tar- effects.
iffs are generally preferable to quantitative 2. Government Investment versus Incen-
restrictions — quotas, prohibitions, exchange tives to Private Investment. Although the ar-
controls, etc. — for reasons already indicated. guments concerning trade restrictions are
Quantitative restrictions prevent competition based mainly on economic considerations, the
with domestic producers regardless of price, choice between Government investment and
raise prices to users and limit demand, and incentives to private investors involves social
require an elaborate administrative mecha- and political factors to a large extent. In
nism and detailed economic analysis to be ef- countries that do not have strong ideological
fective. Quotas also involve a loss of revenue preferences for either private or Government
to the Government, as compared with the use enterprise, the usual approach is to rely on
of tariffs, unless the profits of importers can private investment except in cases where it
be recovered through taxes. cannot be expected to work in the public in-
The cases where quantitative measures terest (e.g., monopoly) or in which its perfor-
may nevertheless be needed derive from the mance has been demonstrably deficient.
 737 POLICY MODELS

Since the reaction of investors to various in- The possibility of attracting foreign invest-
centives (tax reduction, guaranteed markets, ment adds a further element to the problem.
low interest rates, etc.) is subject to consid- To the argument against Government invest-
erable uncertainty, such incentives are more ment must be added the loss of additional in-
likely to be adequate when a general objec- vestment resources, whiie the argument
tive isto be achieved — e.g., import substitu- against private foreign investment must in-
tion, increase of industrial employment — clude the removal of profits from the econ-
than when increases in output in specific sec- omy and the future burden on the balance of
tors are required. Because of this uncer- payments. A purely economic evaluation
tainty, the extent to which reliance on private would probably weigh the value of the addi-
investment is desirable can be determined tional investment resources and managerial
only by an actual trial of specific measures. talents more heavily than the cost of obtain-
Another alternative for securing invest- ing them (particularly where there are un-
ment in given sectors when tax incentives are employed labour and natural resources be-
thought to be inadequate or too costly to the cause of lack of these factors), but the
treasury is the intervention of a Government decision is infrequently made on purely eco-
agency as entrepreneur but not as a long- nomic grounds.
term producer. This may be done through de-
velopment corporations, which sell their in-
QUANTITATIVE ANALYSIS AND
vestments to private enterprises as they be- CHOICE OF INSTRUMENTS
come profitable, or through mixed
corporations, in which the role of the Govern- The preceding discussion has been entirely
ment declines as the enterprise becomes in qualitative terms, which at best leads to
established. the establishment of certain cases to which
The assumption underlying all these mea- particular policies apply. The identification
sures isthat it is bad for the Government to of an actual situation with the relevant case
continue permanently as a producer in most often depends on the results of quantitative
fields. There is a widespread view (shared by analysis. Such factors as the extent of the ex-
the present writer) that the lack of incentives cess demand for imports, the future amount
to efficiency in Government operations makes of unemployed labour, the magnitude of the
private operation preferable even where con- shift in resources needed in particular sec-
ditions are not favourable to the initial un- tors, and the importance to the rest of the
dertaking ofthe investment by private enter- economy of a given investment, can only be
prise. In the absence of more objective determined from such an analysis. The initial
evaluations of the experience with Govern- study of development possibilities should be
ment and private enterprise in various coun- designed to permit a choice of policy instru-
tries, itis impossible to support this conclu- ments in different fields. Once this has been
sion empirically, and it is by no means done, the long-term programme can be for-
universally held among democratic Govern- mulated in more specific terms which take
ments. In countries such as India, for exam- account of the instruments chosen.
ple, an attempt is made to ascertain the rel- The importance of a quantitative analysis
ative merits of public and private investment for the choice of policy instruments will be
in specific fields rather than starting from determined in part by the presence or ab-
this premise. Even in these cases, however, sence of the following factors:
the sectors that are chosen for Government
investment are limited in number and char- 1. Economies of scale in production;
acterized by specific structural features 2. The possibility of imports and exports;
(economies of large scale production, impor- 3. The use of the product in other sectors of
tance of the product, tendency to monopoly, production;
etc.). 4. The predictability of demand.
 738 DEVELOPMENT PLANNING AND POLICYMAKING

In the production of consumer goods, the ment — of inter-related projects of this type
main objective of the development pro- cannot be accurately determined from a par-
gramme islikely to be a certain degree of tial analysis of each investment taken sepa-
substitution of domestic production for im- rately because the profitability of one may
ports, but the choice of sector can be left to understate its contribution to the total. This
market forces. Quantitative analysis may be dynamic type of external economy (as op-
needed to determine the amount of employ- posed to the technological external economies
ment and exchange saving which should be of static analysis) can only be taken account
aimed at in the consumer goods industries, of adequately in the framework of an overall
but not to determine the choice of sector.1 analysis.
At the other extreme, the amount and dis-
tribution ofinvestment in overhead facilities
TYPES OF DEVELOPMENT
must be determined entirely from a quanti- PROGRAMMES
tative analysis of future production because
the alternative of imports is not available and A development programme is an analysis
output is needed to permit investment and which provides a basis for designing and car-
production in other sectors. In some cases the rying out development policy. There is, how-
choice between public and private investment ever, no sharp distinction between program-
will also depend on the amount of output ming and policy making, since each
required. influences the other. The main function of a
Choices among policy measures in the in- programme is to make different policies con-
termediate goods sectors are more affected sistent with each other. Ideally, it should go
by the outcome of the quantitative analysis further and help to select the best policies
than are those in consumer goods because de- and the best means of carrying them out. The
mands derive from the planned outputs of the decision to make a development programme
using sectors. Economies of scale are also does not constitute an endorsement of in-
more prevalent, and there is thus more inter- creased Government intervention, or of any
dependence among investment plans in ear- other particular set of policy instruments,
lier and later stages. While imports provide therefore.
alternative sources of supply for many inter- The nature of the analysis contained in a
mediate goods, some investments will not be development programme is determined in
undertaken unless there is a domestic supply part by the information available and in part
of materials available. To ensure the carrying by the instruments which are being consid-
out of several interconnected projects, Gov- ered. For simplicity, three general types can
ernment intervention in some form is likely to be distinguished, which I will call aggregate
be necessary because the risk to private programmes, sector programmes, and over-
investors would be too great. Investments all programmes.
centering on steel production — ore, trans- Aggregate programmes consist mainly of
port, power, iron and steel, fabricating — pro- national accounts analyses and projections of
vide a good example. Once the initial invest- other magnitudes such as industrial produc-
ments have been made, however, most of tion, labour force, average productivity, etc.
them will prove suitable for private owner- These projections are often combined with a
ship and operation. more detailed anlaysis of certain aspects of
The advantage to the economy — in terms the economy, such as the balance of pay-
of the social productivity of the total invest- ments, the sources of Government revenue,
etc
'This statement is not true where economies of
scale are important as in the case of automobile pro-
Aggregate programmes provide a fairly
duction, because then the profitability of investment adequate basis for the use of general policy
depends on an estimate of the quantity that would be instruments, but they do not furnish a check
demanded at the expected level of income. on the consistency of the results in specific
 739 POLICY MODELS

sectors nor on the balance of payments. They equilibrium must try to identify the problems
are more likely to be adequate when the com- which would exist if inflationary (or, less
position ofproduction and consumption does often, deflationary) forces were offset. The
not change too much as income increases, design of policy in such circumstances is
and when the market mechanism works well likely to call for an over-all analysis, however,
in directing investment and production whether the policy measures selected are
decisions. . . . general or specific in nature.
Sector programmes are analyses of the de-
mands and investment prospects in individual ELEMENTS OF POLICY ANALYSIS
branches of production. Their main function
is to determine the relative priority of invest- General Scope. A policy model is designed
ments within the sector. Investment pro- to determine whether proposed economic pol-
grammes for the whole economy (or for all icies are mutually consistent and to facilitate
resources controlled by the Government) are the choice among them. To fulfill these func-
sometimes constructed by merely adding up tions the model must: (a) incorporate the
the high priority projects in each sector. principal limits to achieving social objectives
The sector approach is generally recog- and (b) contain variables that indicate the
nized to be inadequate as a basis for devel- nature of the policies implied by a given so-
opment policy because it does not provide a lution. To avoid becoming unwieldy, the
test of the consistency of the decisions made model may omit variables which do not sig-
in each sector, nor a way of comparing high nificantly affect the type of problem being
priority projects in one sector with those in considered or which can be handled in a sep-
another. It has nevertheless been the princi- arate analysis.
pal basis for development policy in Latin These considerations suggest that models
America and in most under-developed coun- designed to analyze long-run policy in the
tries until quite recently. The defects in the less developed countries should be signifi-
sector approach are less serious in primary- cantly different from the models now used for
producing economies than in those which the analysis of primary short-run policies in
have reached a high degree of industrializa- the advanced countries. Certain features will
tion and hence have a greater amount of be emphasized, for example:
inter-dependence among the various sectors.
1. Long-term supply limitations, both in the
Over-all programmes combine the ele-
ments of aggregate programmes and sector aggregate and by sector, should be
programmes in varying degrees. The analysis specified.
2. Since the less-developed country is typi-
may start from over-all projections or from
sector analyses, but in the final result they cally quite dependent on foreign trade, ex-
ports and imports must be explicitly
must be reconciled. It is only by some check included.
of this kind that the consistency of the sim- 3. The model should allow for the inflow of
pler models used in the two partial ap- foreign capital as a significant element of
proaches can be tested. . . . development policy.
The need for over-all development pro-
gram es ismost acute when large structural Objectives. The objectives of economic de-
changes are required to establish or restore a velopment are not adequately describee
process of balanced growth. Large balance of the maximization of any single measure of
payments deficits, unemployment, bottle- welfare, such as per capita output or eon-
necks inover-all facilities, and lack of growth sumption. All societies recognize the need to
may each be evidence of such conditions. take account of a number of other locial ob-
These conditions may of course be merely jectives, such as reduced unemployment.
symptoms of an excess or deficiency of total greater equality of income, and redueed eco-
demand, and the diagnosis of structural dis- nomic instability. Some of these objectives
 740 DEVELOPMENT PLANNING AND POLICYMAKING

can properly be taken as limitations on the programs — i.e., combinations of the strategic
economic system by assigning a given value policy variables that are consistent with the
or target to the appropriate variable. In most social and economic structure of the econ-
cases, however, it will be desirable to consider omy. A feasible program is determined pri-
a range of alternative values in order to dis- marily bya solution to an aggregate model,
cover the opportunity cost or "trade off' be- supplemented by whatever elements of sector
tween this and other objectives. or inter-industry analysis may be necessary.
Variables. In his pioneering work on policy The range of programs to be considered
models, Tinbergen divides all variables into should cover values of the policy variables
four categories. that are politically as well as technically at-
1. Objectives, which reflect the aims of tainable. The solutions produced by the plan-
policy; ners should indicate the possibility of achiev-
ing the several social objectives in varying
2. Instruments, which measure the direct ef- degrees. In the absence of prior knowledge of
fects of policy;
the relative valuation of these objectives, the
3. Exogenous variables (data), which are economist cannot determine an optimum
taken as given;
4. Other endogenous (irrelevant) variables, program. At best, he can present the alter-
which do not directly affect either the natives insuch a way as to focus political at-
choice of policy or the social welfare. tention on the problem of determining a so-
cial choice among a limited number of
The first two types of variables — objectives relevant alternatives. This is likely to be
and instruments — are the main concern of much easier than solving the larger problem
policy analysis, and I shall call them policy of describing social preferences in general
terms.
variables. Formally, the analyst's task is to
maximize some (unknown) function of the
objective variables. As Theil has emphasized, STRUCTURE OF THE MODEL
the social welfare is also affected by the val-
ues assumed by some of the instrument vari- A policy model should specify the various
ables, such as taxes or income distribution, factors limiting growth and the ways in
which thus take on some of the welfare char- which they can be modified by the instru-
acteristics ofthe objectives. Although a few ments of government policy. The limits to be
policy variables, such as the exchange rate, included depend on the present structure of
are "pure" instruments which have no wel- the economy, the extent of the possible
fare implications per se, it will be convenient changes in the structure, and the data avail-
to assume that the desirability of a given type able for estimating economic relationships.
of policy is determined by the values of all the Since the economic structure may change
policy variables. A set of values for the policy rapidly in an underdeveloped economy, even
variables may be called a program. a general planning model must pay some at-
The Choice of Programs. In order to arrive tention to the sector composition of demand
at a political decision on the nature of the de- and supply, which will affect both trade pat-
velopment policies to be followed, it is useful terns and capital requirements. Once sector
to focus attention on a few strategic variables analyses have been made, however, their re-
having wide economic and social implica- sults can be incorporated into an aggregate
tions. These variables are primarily those model, which can then be used to study vari-
which apply to large segments of the econ- ations from the initial solution. This proce-
omy and occur in aggregate models, although dure will be followed here.
a few sector decisions may be of the same de- In designing econometric models for less
gree of importance. developed countries, it is useful to start from
The formal model contributes to the choice the experience in advanced countries, making
of policy by determining alternative feasible allowance for the differences noted above. An
 741 POLICY MODELS

examination of aggregate models by Chenery endogenous variables — i.e., those that are not
and Goldberger concluded that the long-term selected to represent policy objectives. In the
model developed for the Netherlands by the Israel case, for example, it was found conve-
Central Planning Bureau came closest to fit- nient to reduce the model to four equations in
ting the needs of the less developed econo- eight variables, of which five are instruments
mies. Features of this model that are partic- and three are objectives.
ularly relevant to the choice of development These two models both allow for the antic-
policy are: ipated composition of demand in estimating
import requirements. The coefficients m<. m*.
1. The inclusion of a number of policy in-
struments: the savings rate, import substi- Hi, \ie, represent the total import require-
tution, the relation of domestic to foreign ments per unit increase in the four compo-
nents of final demand: private consumption,
prices, migration.
2. Explicit analysis of foreign trade. government consumption, investment, and
exports. The differences in the coefficients
3. Incorporation of the results of inter-indus- show the importance of allowing for changes
try analysis into the aggregate model by
an iterative procedure. in the composition of final demand. . . .
In designing the model, the analyst has a
These elements are utilized in the policy considerable choice of variables to represent
model designed by Goldberger to analyze ex- government policy. It is sometimes feasible to
isting data available for Argentina. The use a variable that directly reflects the effects
model was further developed by Chenery and of government actions, such as the exchange
Bruno to determine development alternatives rate. More often there will be several inter-
in Israel. These models explicitly recognize mediate links between the actual policy mea-
four limits to growth: sures, which may affect individual sectors of
the economy, and the summary measure that
1. The supply of capital.
it is convenient to use in the general model.
2. The supply of labor.
3. The supply of foreign exchange. The range of choice is illustrated by the al-
4. The composition of internal and external ternatives proposed for the analysis of sav-
demand. ings in Argentina. On theoretical grounds,
the following set of equations was suggested:
The first three limits are reflected in the
basic equations of the aggregate model. The (1 ) Net domestic saving
composition of demand at different income
S = axW + a2R + T - G
levels, together with the possibilities for im-
ports and exports, determines the capital re- (2) Taxation
quirements and import demands. . . . T = tY
A policy model is characterized by having
more variables than equations. . . . This ex- (3) Income distribution
cess of variables constitutes the number of W = w(R + W)
degrees of freedom of the model. ... To solve
where
the model, the values of a number of vari-
ables equal to the number of degrees of free- W = Disposable wage income
dom must be fixed in advance. In general, the
R = Disposable non-wage income
planner can assign consistent values to any T = Taxes
combination of policy variables — objectives G = Government current expenditure
or instruments — equal to the number of de- / = Tax rate
grees of freedom and then solve for the values w = Share of wages in disposable income
of the remaining variables.
The process of solution is facilitated by This formulation shows the several effects
first eliminating some or all of the irrelevant of government policy on savingl by using
 742 DEVELOPMENT PLANNING AND POLICYMAKING

three instruments: government expenditure should cover all the consistent programs that
(G), the tax rate (/), and the share of wages might be of interest to policy makers.
(w). Since information was not available to The specified limits for each policy varia-
estimate the parameters in the savings and ble must be determined from outside sources.
tax functions, the formulation was first sim- For example, the limits to increased saving
plified to: may be set by feasible improvements in the
tax structure and administration and the
(4) S = a*Y - G probable distribution of income. An optimis-
and finally to: tic estimate of these factors would set an
(5) S = aY upper limit to the savings rate that it is worth
In the last form, a reflects all the government considering.
policies that influence the rate of savings. A This approach demands less precision and
similar range of alternatives could be shown detail in the model than would a mechanical
for other instrument variables which are less application of a formal optimizing procedure.
directly connected to the actions of the Additional information is used both in estab-
government. lishing the starting point for the analysis and
in setting the range over which the equations
in the model must hold. The use of linear ap-
proximations tonon-linear functions can be
THE DETERMINATION OF quite accurate for a limited range when it
ALTERNATIVE POLICIES would not be acceptable for the whole range
of values.
Procedures. The greater value of a formal Having specified the starting point and the
model is its ability to determine consistent allowable range for each policy variable, the
sets of alternative policies. The weakness of analyst's next step is to fix values for a num-
the model approach is its oversimplification ber of policy variables equal to the number of
of many of the individual features of the degrees of freedom in the model. If it is de-
economy in order to bring out their inter-re- sired to determine the limits to the feasible
lations. This drawback can be minimized by area of choice, this can be done by setting the
starting from a trial program in which addi- instrument variables and the objective vari-
tional elements of detailed analysis and judg- ables alternately at their minima or maxima.
ment have been introduced. The model struc- A set of solutions will determine the range of
ture should be adjusted to be consistent with values that is consistent with all the
these added elements. restrictions. . . .
Given a realistic starting point, the model The last step in the analytical procedure is
can then be used to explore a range of solu- to narrow the range of choice and finally to
tions in which the policy variables vary present a few of the leading alternative pro-
within predetermined limits. This range, grams to the political authorities for deci-
which I shall call the feasible area of choice, sion. . .

Comment
In advanced studies of planning models of development, attention is given to macroeco-
nomic, multisectoral, optimizing, and control theory models. For studies of the design and use
of econometric models and planning models, see H. B. Chenery (ed.), Studies in Development
Planning (1971); C. R. Blitzer et al., Economy-Wide Models and Development Planning
(1975). An excellent survey article is A. S. Manne, "Multi-Sector Models for Development
Planning" Journal of Development Economics Vol. 1, No. 1 (1974). Other useful references
are: S. Chakravarty, Capital and Development Planning (1969), chapters 1-5; E Malinvaud
and M. D. L. Bacharach (eds.), Activity Analysis in the Theory of Growth and Planning
(1967), chapter 7; H. B. Chenery, "The Use of Interindustry Analysis in Development Pro-
 743 POLICY MODELS

gramming," in T. Barna (ed.), Structural Interdependence and Economic Development

(1956); H. B. Chenery and K. S. Kretschmer, "Resources Allocation for Economic Devel-
opment," Econometrica (October 1956); H. B. Chenery and P. Clark, Interindustry Econom-
ics (1959), chapters 1-3; L. Johansen, A Multi-Sectoral Study of Economic Growth (1960);
Roy Radner, Notes on the Theory of Economic Planning (1963); J. K. Sengupta and G. Tint-
ner, "On Some Economic Models of Development Planning," Economia Internazionale (Feb-
ruary 1963); J. Tinbergen and H. C. Bos, Mathematical Models of Economic Growth (1962);
S. Chakravarty and L. Lefeber, "An Optimizing Planning Model," Economic Weekly (Feb-
ruary 1965); I. Adelman and E. Thorbecke (eds.), The Theory and Design of Economic De-
velopment (1965); M. Bruno, "Experiments with a Multi-Sector Programming Model," (No-
vember 1967); Russell Ackoff, "Operations Research and National Planning," Operations
Reserach (August 1957); Charles Hitch, "A Dissent," Operations Research (October 1957);
M. Bronfenbrenner, "A Simplified Mahalanobis Development Model," Economic Develop-
ment and Cultural Change (October 1960); Amartya Sen, "Interrelation Between Project,
Sectoral and Aggregate Planning," Economic Bulletin for Asia and the Far East (June-Sep-
tember 1970); G. M. Heal, Theory of Economic Planning (197 '3); G. Pyatt and E. Thorbecke,
Planning Techniques for a Better Future (1976); Leif Johansen, Lectures on Macroeconomic
Planning (1977 and 1978), parts I and II.
 XI.C. PLANNING EXPERIENCE

XI.C.1. The Crisis in Planning*

Although planning occurs in many types of Most of the time, most low-income countries
decision-making units and is often defined to have development plans which apparently en-
cover any attempt to select the best means to deavour to meet all or most of these specifi-
achieve desired ends, this paper focuses more cations; we are thus examining a highly sig-
narrowly on comprehensive development nificant aspect of applied economics, and the
planning in low-income countries (although characteristics listed are chosen to identify
the discussion is also relevant to the "special what is common to most comprehensive de-
case" of industrial countries). Advocates of velopment plans rather than to draw atten-
comprehensive development planning typi- tion to any special features or eccentricities.
cally propose that plans should meet the fol- The economic case for development plan-
lowing specifications: ning, while sometimes taken as axiomatic, is
1. Starting from the political views and goals generally made out in terms of the failings of
of the government, the plan should define an unregulated market economy. Perhaps the
policy objectives, especially as they relate chief of the arguments views planning as a
to the future development of the economy. superior means of arriving at investment and
2. It should set out a strategy by means of other decisions affecting the future, with the
which it is intended to achieve the objec- market seen as supplying information which
tives, preferably translated into specific is a poor guide for such decisions, leading to
targets. avoidable uncertainties and myopia. Thus,
3. It should present a centrally co-ordinated, Scitovsky and others drew attention to the in-
internally consistent set of principles and terdependence ofinvestment decisions and
policies, chosen as optimal means of im- alleged that aggregate investment made up
plementing the strategy and achieving the of atomistic decisions would be less than that
targets, and intended to be used as a which would result from "centralised invest-
framework to guide subsequent day-to- ment planning" providing more realistic sig-
day decisions. nals of present plans and future conditions.
4. It should cover the whole economy (hence In other ways, too, planning is seen as a
means for correcting discrepancies between
it is "comprehensive" as against "colo-
private and social valuations, for example,
nial" or "public sector" planning).
5. In order to secure optimality and consis- the market's tendency to over-value unskilled
tency, itshould employ a more-or-less for- labour. Under the influence of the "big push"
malized macro-economic model (which, school of thought, it has also been seen as the
however, may remain unpublished), em- only way to mobilize resources on the scale
ployed to project the intended future per- necessary for a successful development effort,
formance ofthe economy. and as the only practical means of weaving
6. It typically covers a period of, say, five the various threads of economic policy into a
years and finds physical expression as a consistent whole.
medium-term plan document, which may,
however, incorporate a longer-term per- THE CRISIS IN PLANNING:
spective plan and be supplemented by an- EXPLANATIONS; SOLUTIONS
nual plans.
Although it is not a matter that can be re-
duced to any simple demonstration, there
*From Tony Killick, "The Possibilities of Devel- would probably be little disagreement today
opment Planning," Oxford Economic Papers, Vol.
28, No. 2, July 1976, pp. 161-6. Reprinted by that the practice of planning has generally
permission. failed to bring many of the benefits expected
744
 745 PLANNING EXPERIENCE

from it. Waterston's study of the lessons of priately specified macro-models, to be in-
experience concluded that "there have been sufficiently specific about policies and
many more failures than successes in the im- projects, to overlook important non-eco-
plementation ofdevelopment plans";1 Seer's nomic considerations, to fail to incorpo-
keynote paper for a 1969 conference on "The rate adequate administrative provision for
Crisis in Planning" was entitled "The Prev- their own implementation.
alance of Pseudo-planning";2 and Healey is 2. Inadequate resources: incomplete and un-
surely accurate in claiming that the results reliable data; too few economists and
"have been sadly disillusioning for those who other planning personnel.
believed that planning was the only way."3 3. Unanticipated dislocations to domestic
None of this, of course, is to deny individ- economic activity: adverse movements in
ual successes nor some genuine benefits. The the terms of trade; irregular flows of de-
creation of planning agencies and prepara- velopment aid; unplanned changes in the
tion of plan documents has surely had an ed- private sector.
ucational effect among politicians and ad- 4. Institutional weaknesses: failures to locate
ministrators, helping to define, and raise the the planning agency appropriately in the
understanding of, major policy issues. Plan- machinery of government; failures of
ners do not spend all their time dressing win- communication between planners, admin-
dows and have certainly helped to raise the istrators, and their political masters; the
standard of policy decisions on matters such importation of institutional arrangements
as project selection. Nevertheless, there has unsuited to local circumstances.
been a vast gap between the theoretical ben- 5. Failings on the part of the administrative
efits and practical results of development civil service: cumbersome bureaucratic
planning. It is doubtful whether plans have procedures; excessive caution and resis-
generated more useful signals for the future tance to innovations; personal and depart-
than would otherwise have been forthcoming; mental rivalries; lack of concern with
governments have rarely, in practice, recon- economic considerations. (Finance Minis-
ciled private and social valuations except in a tries are a particularly frequent target,
piecemeal manner; because they have seldom often said to undermine the planning
been operational documents, plans have agency by resisting the co-ordination of
probably had only limited success in mobiliz- plans and budgets.)
ing resources (although they probably have
There is certainly ample evidence that
induced larger aid flows) or in co-ordinating each of these tendencies has contributed to
economic policies.
The profession cannot be criticized for the planning crisis, the precise combination
being unresponsive to this situation. Much varying over time and from country to coun-
thought has been given to the sources of poor try. But there seems to be a growing consen-
sus among economists that yet another set of
plan performance, with the most commonly factors is the most important explanation:
mentioned causes listed below:4
that "lack of government support for the
1. Deficiencies in the plans: they tend to be plans is the prime reason why most are never
over-ambitious, to be based upon inappro- carried out successfully."5 Seers, while also
finding fault with administrators and econo-
'Waterston, 1966, p. 293. mists, argues that "political forces encourage
2Seers, 1972. the production of pseudo-plans";6 Tinbergen
3Healey, 1972, p. 761. sees as one of the difficulties "that among
^o avoid a wearisome number of detailed refer-
politicians, probably as a consequence of our
ences, the reader is referred for examples of the fol-
educational system, a preference exists for
lowing points to Waterston, 1966, chapters VI to IX,
and Faber and Seers (eds.), 1972, passim. See also
5Waterston, 1966. p. 340.
Powelson, 1972; Tinbergen, 1964; and Myrdal, 1968,
chapter 15. 6Seers, 1972, p. 24.
 746 DEVELOPMENT PLANNING AND POLICYMAKING

thinking in qualitative terms only";7 and government really committed to the full ex-
Myrdal refers to "rivalries between parties or ecution ofa plan could end up making worse
ministers" as one of the major problems.8 rather than better decisions? It might simi-
In the face, presumably, of the futility of larly be asked of those advocating adminis-
advocating reformed political systems, most trative-type reforms, what makes them think
proposals for improving plan performance these solutions to be attainable through pre-
tend to the administrative or organizational. cisely those political processes which are
Frisch, for example, has developed an admin- blamed for past failings? Are not deficient in-
istrative decision model intended to secure stitutions and procedures an expression of the
"optimal implementation."9 Myrdal perhaps political system itself, not to be remedied
comes closest to advocating a political solu- without first or simultaneously instituting po-
tion in arguing for "democratic," or decen- litical changes? For example, the respective
tralized, planning.10 Helleiner also tries to roles of the planning agency and the finance
grasp the political nettle, with the prescrip- ministry reflect, in substantial part, a distri-
tion that "those engaged in planning activi- bution of political power; is it useful, then, to
ties must be sufficiently close to the seat of make proposals for raising the relative influ-
political power to be relevant to the actual ence of the planning agency while remaining
silent on the distribution of power?
process of political decision making ..." but
sees the practical application of this largely Economists have generally failed to ask
in terms of "new institutions and person- such questions and, significantly, it was a po-
nel."11 Consistent with his views on the bane- litical scientist who, on reviewing explana-
ful influence of traditional education, a U.N. tions for plan failures similar to the list given
committee headed by Tinbergen advocates above, was led to observe that it "rather
"Intensified training of many persons in- plainly adds up to the conclusion that plan-
volved";12 and Waterston's proposal for an ning is more or less bound to fail, given the
"operational approach" to planning empha- probability that many of these factors will be
sizes the use of annual plans tied into bud- present in any situation of underdevelop-
getary procedures and supplemented by ment," and to urge that "Any useful concep-
"multi-annual sector programmes."13 tualisation ofthe planning process must start
Some large questions have, however, gone from a model of politics."14 The record of
unasked in these attempts to respond to the past performance suggests the possibility that
planning crisis, leaving some doubt whether effective planning may simply not be feasible,
the resulting prescriptions have been radical so the next step is to take up Leys's point and
enough. The inclination to see politicians as examine the model of politics upon which the
the spoilers leaves one wondering why it notion of development planning appears to
would not be in their own interests to support have been built.
their planners, if planning is viewed as a way
of raising the rationality and effectiveness of
public policies. Might it be that the concept References
of development planning is one that could Faber, Mike, and Seers, Dudley (eds.), The
not, with the best will in the world, be built Crisis in Planning, London, 1972 (2 vols.).
into the process of government because "pol- Frisch, R., "Optimal di
Implementation," Re-
vista Internazionale Scienze Economiche
itics isn't like that"? Might it even be that a
e Commerciali, Milan, 1 966, No. 1 .
7Tinbergen, 1964, p. 43. Healey, Derek T., "Development Policy:
8Myrdal, 1968, p. 732. New Thinking about an Interpretation,"
9Frisch, 1966, passim.
10Myrdal, 1968, chapter 18 passim. 14Leys, 1972, pp. 56 and 60. The discussion of his
paper, summarized in the same volume, certainly
1 'Helleiner, 1972, pp. 354 and 347.
seems to justify Leys's complaint about how difficult
12Tinbergen, 1972, p. 160. other social scientists find it to communicate with
13Waterston, 1972, passim. economists (p. 79).
 747 PLANNING EXPERIENCE

Journal of Economic Literature, Septem- Seers, Dudley, "The Prevalence of Pseudo-

ber 1972.
Planning," in Faber and Seers (eds.),
Helleiner, G. K., "Beyond Growth Rates and 1972, Vol. 1, chapter 1.
Plan Volumes — Planning for Africa in the Tinbergen, Jan, Central Planning, New
1970's," Journal of Modern African Stud- Haven, 1964.
ies, October 1972.
Tinbergen, Jan, "The United Nations Devel-
Leys, Colin, "A New Conception of Plan- opment Planning Committee," in Faber
ning?" in Faber and Seers (eds.), 1972, and Seers (eds.) 1972, Vol. 1, chapter 7.
Vol. 1, chapter 3. Waterston, Albert, Development Planning:
Myrdal, Gunnar, Asian Drama, Twentieth Lessons of Experience, London, 1966.
Century Fund, New York, 1968. , "An Operational Approach to De-
Powelson, John P., Institutions of Economic velopment Planning," in Faber and Seers
Growth, Princeton, N.J., 1972. (eds.), 1972, Vol. 1, chapter 4.
Comment
The author of the foregoing selection suggests that instead of adopting a "rational actor"
model of politics, economists should view politics and political decision making from the be-
havioral standpoint. The rational actor model
would have us see governments as composed of public-spirited, knowledgeable, and role-oriented politi-
cians; clear and united in their objectives; choosing those policies which will achieve optimal results for
the national interest; willing and able to go beyond a short-term point of view. Governments are stable,
in largely undifferentiated societies; wielding a centralized concentration of power and a relatively un-
questioned authority; generally capable of achieving the results they desire from a given policy decision
(Killick, p. 171).
Opposing this view as being unrealistic and hence lacking relevance, Killick argues that
governments do not and could not function in the manner implied by the economists' concep-
tions of development planning. The behavior of governments is far from the economists' ideal-
ization of the optimization process, and the rationality of government policies may be far
different from the norms of economic rationality. (See Killick, "The Possibilities of Devel-
opment Planning," Oxford Economic Papers (July 1976), pp. 161-81.)
Also relevant on the nature of political decision making are: Graham T. Allison, Essence
of Decision (1971); Raymond A. Bauer and Kenneth J. Gergen (eds.), The Study of Policy
Formation (1968); David Braybrooke and Charles E. Lindblom, A Strategy of Decision
(1963); Herbert Simon, Models of Man (1957); Albert O. Hirschman, "Policymaking and
Policy Analysis in Latin America — A Return Journey," Policy Sciences, Vol. 6 (1975); Guil-
lermo O'Donnell, Modernization and Bureaucratic-Authoritarianism (1973); David Collier.
The New Authoritarianism in Latin America (1979).

XI.C.2. Planned Development in

India*
The structure of our economic policies in the (1) The only sensible long-term way of
last two decades of planned development was eradicating poverty must be to create more
founded on six major premises: jobs by investing and growing faster; hence.
tures. Institute of Public Enterprise, Hyderabad.
♦From Jagdish N. Bhagwati, India in the Interna-
tional Economy, Lai Bahadur Shastri Memorial Lec- 1973, pp. 3-12. Reprinted by permission.
 748 DEVELOPMENT PLANNING AND POLICYMAKING

redistributive measures which would tend to

"commanding heights" of the economy, the
use up investible resources in current con- "basic" or "key" sectors.
sumption were undesirable. (6) Finally, while these policy instruments
(2) In consequence, the Government must embraced directly the non-agricultural sector
direct its major effort at raising the domestic of our economy, though in turn influencing
savings rate and securing external aid to sup- no doubt the agricultural sector, the Plan
plement domestic savings until such time as programmes repeatedly urged land reforms
the domestic savings rate had been raised to to transform the institutional structure in the
levels adequate to do away with foreign aid rural economy and urged rural works pro-
altogether and achieve rapid growth with grammes so that expanding production
self-reliance. would go hand in hand with better income
(3) The industrial sector, which would distribution.
grow with investment and income, had to be Experience has now shown that these
planned and controlled in depth: towards this premises of our policies have been either mis-
end, very detailed targeting and regulatory li- guided orinadequate or unrealistic in our po-
censing ofindustrial establishments was con- litical framework. In short, a serious restruc-
sidered necessary and resulted in the Indus- turing of our policies is called for. Indeed,
tries (Development and Regulation) Act of many of the more dramatic measures of re-
1951. cent years, especially the nationalization of
(4) The external accounts, i.e., the balance banking and wholesale wheat trade, are evi-
of payments, had also to be regulated, not via dence that a restructuring is either being in-
exchange rate adjustments and the use of tentionally designed or being forced by the
protective tariffs, but by resort to comprehen- complex of political and economic circum-
sive exchange control so that the use of im- stances on the government. Hence a realistic
ported raw materials, capital goods and other appraisal of where our policy premises fell
supplies was to be regulated by an elaborate short is necessary to examine the optimal pol-
administrative mechanism. icy mix for our economy and to judge, in per-
(5) Alongside these four basic premises of spective thereof, the present and prospective
economic policy was the objective of an in- governmental actions and programmes. To
creasing role for the public sector in the own- this task, I shall now turn.
ership ofthe country's resources. As the suc- (1) The estimates of increasing poverty in
ceeding Industrial Policy Resolutions and the India have underlined the lesson that grow-
pronouncements of our Prime Ministers ing incomes do not necessarily trickle down
underlined, this was a political objective but to the bottom deciles. Even if they did, the
also one which reflected certain economic ax- process would be so slow that it would be in-
ioms. Thus, we thought that the public sector tolerable inthe time it would take: it is mor-
would invest where the private sector would ally difficult and politically impossible to ask
not invest (e.g., steel in the Second Plan); those in poverty to wait until perhaps the
and, in particular, the public sector would next millennium for significantly improved
generate surpluses for investment, obviating incomes. The early faith in the ability of
the need to raise savings exclusively via the growing incomes and investments to take
politically-difficult budgetary process of tax- care of the poor within the foreseeable future,
ation. These dual economic objectives in ex- in any significant fashion, is no longer with
panding the public sector went alongside us.
with the envisaged role of the expanding pub- (2) The failure of the income-generation
lic sector in conjunction with the industrial mechanism to make an impact on poverty, in
licensing policy in controlling the concentra- turn, must be traced to the inappropriateness
tion of private economic power. There was of the other policy premises I listed earlier.
also the additional political objective of se- The policy premise that domestic savings
curing public control over the so-called should be stepped up steadily and the level of
 749 PLANNING EXPERIENCE

external aid should be tapered off as the do- which no rational economic criteria have ever
mestic savings took its place has been only been defined or at least discovered. Import li-
partially fulfilled: ironically, aid has substan- censing, incombination with the principle of
tially fallen for exogenous reasons since the
indigenous availability, has conferred "auto-
mid-60's but domestic savings which in- matic protection" to domestic industry re-
creased steadily through the first three Five- gardles ofcosts. The pro-rata-to-capacity al-
Year Plans, have stagnated since then! (It is locations ofimported (and other) materials
tempting to argue, as some radical econo- to licensed units have, in combination with
mists have recently argued elsewhere, that automatic protection and licensing restric-
over the long haul the presence of foreign aid tions on domestic entry, eliminated virtually
in fact may have reduced our domestic sav- all effective foreign and domestic competi-
ings effort; but systematic economic analysis tion, and in consequence also the incentive to
does not support such a charge.) The Indo- efficiency and cost-reduction. The practi-
Pakistan War, the refugee relief burden and cally-assured markets for licensed units (in
the outflow of significant net resources to view of the elimination of foreign and domes-
Bangladesh have undoubtedly contributed tic competiton) and the guaranteed share in
their share to this continuing failure to re- the available imputs as soon as capacity is li-
sume the rate of increment in the domestic censed have also meant that, even when there
savings rate. Nonetheless, since the mid-60's was excess capacity in an industry or a prod-
even our traditional strategy of increasing the uct, there could still be adequate incentive to
rates of saving, investment and job-plus-in- add to capacity: there is little doubt that,
come-creation has not been pursued with any while we have been eager to blame aid-tying
degree of success. to project imports for much of our phenom-
(3) But the efficacy of this developmental enal underutilisation of capacity, it is very
strategy has been undoubtedly impaired also much a result of our own licensing and allo-
by the inefficiency which has resulted from cation policies which make the addition of
our framework of industrial and foreign yet more capacity to underutilized-capacity
trade policies. Even while we heroically industries ever so profitable.
raised our domestic rate of saving through
This pattern of "bureaucratic capitalism"
successive budgets, our policies which defined is reminiscent of the "bureaucratic social-
the use of these savings and accumulated ism" from which the Soviet Union is busy
capital stock have left us with intolerable inef- disengaging itself in its Liebermannist re-
ficiencies and diminished returns to our forms; but we have stuck to it, with occa-
efforts. sional Ministerial and Prime Ministerial ex-
There are so many inefficiencies inherent hortations toimprove it— exhortations which
in these policies that one could write volumes are no longer amusing — never facing up to
cataloguing them, as indeed I have! But let the stark and compelling fact that it is one of
me highlight a few of the most obvious ones, the supreme examples of counterproductive
focusing first on the industrial policy frame- and ill-conceived economic policy-making.
work and next on the foreign trade regime, Adam Smith discovered the Invisible Hand;
though the conjunction of both has dispro- and economists have correctly argued, in
portionately accentuated some of the unfor- Joan Robinson's colourful phrasing, that the
tunate consequences of either taken by itself. Invisible Hand may often work by strangu-
The detailed industrial targeting cannot lation. But the maze of meaningless controls
possibly take into account costs and benefits that we continue to work with, so that the In-
at a micro-level. Industrial licensing has visible Hand is nowhere to be seen, represents
either buttressed this targeting or made the a regression, a reductio ad absurdum that
non-targeted industrial development none- does our intelligence little credit and our
the-less go through a series of substantially economy much harm.
meaningless bureaucratic scrutinies for Indeed, one must ask why this policy
 750 DEVELOPMENT PLANNING AND POLICYMAKING

framework continues and answer that it is Compounding these inefficiencies has been
not merely a product of misguided thinking our peculiar attachment, only slightly dimin-
on the part of our intellectual Left — for sev- ished through the years, to making all import
eral of us who are on the Left have already allocations generally non-transferable. The
rejected this travesty of radicalism. It surely resulting bottle-necks and inflexibility are a
has to be explained, in the ultimate political totally gratuitous source of additional, and
analysis, as the most obvious and ideal instru- significant, waste in the system; they must
ment for creating an elaborate facade to de- also seriously impede exports where the abil-
lude the masses into the belief that socialism ity to grasp sudden opportunities and to pro-
is being practised while, in reality, the policy duce with predictable and rapid access to
merely serves to create political power and materials without red tape is surely critical.1
patronage, essentially leading to shared eco- And let us not forget the simple but critical
nomic power in the urban areas between the fact that the attempts to offset the ill-effects
private sector and the politicians, and to of over-valuation of the exchange rate on ex-
spawn ever more, in Shakespeare's phrase, port performance through greater reliance on
"caterpillars of the Commonwealth": the export incentives are also an additional
economic agents whose function is to seek out source of the corruption which has now be-
the "rents," the monopoly profits which the come almost endemic to our scene — and
system itself creates through the endless gen- again, this is totally unnecessary as the most
eration of licensed allocations, and the bu- effective and efficient incentive to export is
reaucratic agents, with whom they have a provided by eliminating the over-valuation of
symbiotic relationship, whose function is to the exchange rate as some developing, and
allocate these rent-earning privileges among many developed, countries have now come to
the competing claimants. understand and practice.
(4) Let me turn now to the foreign trade Let me stress again that the objection to
and exchange rate policies which, as I have the energetic use of flexibility in setting the
already indicated, have critically reinforced exchange rate has little justification in a rad-
this inefficient policy framework. icalism that is concerned with socialist objec-
The automatic protection mindless of the tives as distinct from so-called socialist in-
cost of import substitution and the elimina- struments of economic policy. In fact, not
tion of virtually any competitive spur to effi- merely in India, but in many other develop-
ciency, which I have already noted, were the ing countries which went through a phase of
product of the comprehensive exchange con- comprehensive exchange control in the three
trol that we settled down to as a long-term decades since the War, the exchange control
way of managing our balance of payments. regime has made a mockery of income distri-
The overvalued exchange rate has also im- butional objectives by creating profits and
plied a continuing discrimination against ex- privilege. And it is only the unthinking seg-
ports: a phenomenon that has become so ob- ment of intellectual opinion, which is me-
vious that the government has repeatedly chanically wedded to obsolete and defunct
sought to offset it by special subsidies to ex- notions, which would want to stick to over-
portation. But the several export subsidy pro- valued exchange rates as a matter of
grammes have been generally chaotic in their principle. . . .
selectivity and have reproduced on the export Finally, I must reject the widespread no-
front the irrationalities and inefficiencies that tion that somehow, in Indian conditions, ex-
have afflicted the import substitution mech- change rate flexibility would be unworkable.
anism. The multiplicity of effective exchange In this regard, the experience of the June
rates for exports, necessitated by the adher-
ence to overvalued exchange rates, is a seri- 'These and other consequences of our trade and ex-
change rate policies were extensively discussed in J.
ous source of waste and must inevitably re- Bhagwati and Padma Desai, India: Planning for In-
duce the returns to our economic sacrifices. dustrialization, London, 1970.
 751 PLANNING EXPERIENCE

1966 devaluation has been widely misunder- amply shown that we were not fully success-
stood. The 1966 devaluation turned out to ful. At the same time, the stagnation in the
have been badly planned and timed, from a rate of industrial investment, under the later
political and psychological standpoint: the policy of confining the Large Industrial
extreme pressure exerted by the Aid India Houses to the so-called "core" industries,
Consortium in favour of the devaluation pro- suggests strongly that neither the public sec-
duced a perfectly natural nationalist reaction tor nor the small-scale sector has been able to
which put the devaluation into a politically fill the vacuum: so that we have run into the
unacceptable strait-jacket; and the second additional problem that a policy which seeks
major agricultural drought that followed it to confine the growth of the Large Houses to
led to major price increases and export sup- only the supposedly socially necessary "core"
ply difficulties which swamped the effects of industries is also contributing, given the fail-
the devaluation. Careful analysis, which is ure on the part of the public and the limited
necessary to separate out the effects of the capacity of the small-scale sectors, to stag-
drought from the effects of the devaluation, nation inindustrial investments. At the same
and which Prof. T. N. Srinivasan and I have time, one cannot suppress the thought that
just concluded, strongly indicates that the there is little that is more than symbolic in a
latter were favourable; and that, in the ab- policy framework which leaves the Large In-
sence of the devaluation, the economy would dustrial Houses at their greatly-augmented
have been in yet worse shape thanks to the capital stock and accompanying politico-eco-
nomic power and seeks to convince the
drought.2 masses that somehow the prevention of a
(5) Turning next to the economic objec-
tives to which the public sector was supposed purely marginal addition to this capital stock
to contribute, I am afraid that the outcome, is tantamount to ushering in socialism.
while not as unfortunate as in the case of our Indeed, it is somewhat ironic that we have
trade and industrial policy framework, has now shifted from viewing the "core," the
been less than satisfactory. For one reason or "basic," the "key" sectors as essentially the
another, the contribution made by the public preserve of the public sector, in its exclusive
sector enterprises to public saving has fallen or dominating domain, to viewing the "core"
considerably short of our early optimism. industries as the particularly choice invest-
Brilliant economists, of whom we now have a ment field for what appear to be unsympa-
growing number, can readily construct ratio- thetic and uninvesting Large Industrial
nal reasons why a public investment may be Houses. It is almost as if the attainment of
run at a private loss and yet contribute to so- the "commanding heights of the economy.'"
cial gain. But these reasons have as little to that colourful phrase we learnt to use in pro-
do with the actual occurrence of inadequate gressive circles, had produced, not bliss but,
vertigo!
profits, and resulting savings, in the public
sector as do the sophisticated reasons for se- (6) And let me turn finally to the sixth and
lective taxing and subsidising of foreign trade last policy premise: that agricultural growth
with the situation actually obtaining and de- would be accompanied by improved income
scribed byme earlier in this Lecture. distribution via land reform and expanding
As for the role of the expanding public sec- rural works programmes. It is only an urban
tor investments, in conjunction with indus- intelligentsia which needs to be reminded
trial licensing, in restraining the growth of that, in a country such as ours, with the bulk
the concentration of economic power, the of its population and its poor on land, social-
growth of the Large Industrial Houses has ist programmes cannot be formulated mean-
ingfully except in so far as they involve the
2Cf. J. Bhagwati and T. N. Srinivasan, Foreign agricultural sector. The utmost importance
Trade Regimes and Economic Development: India,
National Bureau of Economic Research, Amsterdam, attaches therefore to our agricultural land re-
1974. forms. And yet, our progress on this front has
 752 DEVELOPMENT PLANNING AND POLICYMAKING

been slow despite volumes of legislation on length, only to conclude that there is no es-
the subject. The readiness and ease with cape from the unpleasant evidence of the
which the modern, non-agricultural sector harmful effect of our policies on industrial ef-
has been subjected to the extensive pseudo- ficiency and thence on the overall Derfor-
socialist regulation and control I have just mance and growth of our economy.
discussed contrasts starkly with the lack of These policies have also increased our ex-
genuine socialist transformation in our agri- ternal dependence — a paradox only if you
cultural economy, so that we have to-day an believe that comprehensive exchange controls
economy which is like an iceberg whose tip and licensed allocations must necessarily
has the glossy look of socialism but whose eliminate such dependence. But by adversely
bulk underneath has diverse capitalist and affecting our export performance, and by re-
quasi-capitalist elements. ducing the value added and hence increasing
Nor have our rural works programmes the consequent demand for external re-
ever become substantial or effective enough sources, ceteris paribus, corresponding to any
to make for a real impact on the agricultural investment in import substitution, we have
economy. We have repeatedly built them into compounded our dependence on foreign aid
our Plans and talked of the numerous tasks beyond what our early planners envisaged as
that the "surplus" rural labour could be mo- the period during which we could take "aid
bilized to perform. But, in the absence of or- to end aid". . . .
ganizational input and financial resources, And finally let me stress the unpleasant
we have had few results that would match but incontrovertible fact that our policies
our self-exhortations. Again, China has suc- have been inadequate to the task of bringing
ceeded admirably here by being able to use social justice to our society. By reducing ef-
its abundant labour in the communes, thus ficiency and the rate of growth, our trade and
automatically providing the organizational industrial policies have in fact cut into even
framework and obviating the need to make the small but probable improvement in the
budgetary allocations to compensate the incomes of the poor that we had hoped for.
"hired" labour. And as for their direct impact on income dis-
We thus find ourselves in a situation where tribution, they have essentially redistributed
tew of our original policy premises can be re- income from the rich to the nouveau riche:
garded as acceptable if we are to get the the poor have indeed no way of asserting
economy moving in the direction of rapid themselves in the marketplace but, in our
growth, self-reliance and social justice. policy framework, they equally cannot afford
The present trade and industrial policies to wait in queues for licences nor do they
have reduced the returns to our investments. have the position, privilege and resources to
There are no mitigating "beneficial" effects push past the line. Nor has our targeting and
such as increased savings in the economy or industrial licensing diminished the continu-
increased inducement to invest or enhanced ing and increasing use of resources to satisfy
R and D which can plausibly be cited, with the needs of the rich, and the not-so-rich.
supporting evidence. In our study of India's And, as I noted earlier, our rural pro-
foreign trade regime,3 Professor T. N. Srini- grammes and halting land reforms have left
vasan and I have examined these aspects at us with only increasing numbers below the
poverty line in the last two decades.
A restructuring of our policy framework is,
3Bhagwati and Srinivasan, ibid. therefore, called for.

Comment
In all developing countries, the management of the public sector has become a prime con-
cern. A major task of that management must relate to state-owned enterprises. As state-
owned enterprises have increased markedly in numbers, more attention is being given to pol-
 753 PLANNING EXPERIENCE

icies that might improve their efficiency. See Leroy Jones, Public Enterprise and Economic
Development (1975); A. H. Gantt and Giuseppe Dutto, "Financial Performance of Govern-
ment-Owned Corporations in Less Developed Countries," IMF Staff Papers (May 1968);
Hamlin Robinson, "Public Corporations in Malaya," Development Administration Review
(January 1973); Richard A. Musgrave et al., "Public Enterprises and the Fiscal System,"
chapter 8, in Musgrave et al., Fiscal Reform for Bolivia, vol. II (1977); A. Choksi, "State
Intervention in the Industrialization of Developing Countries: Selected Issues," World Bank
Staff Working Paper No. 341 (July 1979); William J. Baumol (ed.), Public and Private En-
terprises ina Mixed Economy (1980); Leroy Jones et al., Public Enterprises in Developing
Countries (forthcoming).

Comment
A large number of case studies are now available for an appraisal of development planning
in practice. A selected list of readings follows: I. G. Patel, "Strategy of Indian Planning," in
P. Chaudhuri (ed.), Aspects of Indian Economic Development (1971); J. N. Bhagwati and S.
Chakravarty, "Contributions to Indian Economic Analysis: A Survey," American Economic
Review (September 1969), supplement, part 2; R. S. Eckaus and K. Parikh, Planning for
Growth (1967); R. S. Eckaus, "Planning in India," in M. F. Millikan (ed.), National Eco-
nomic Planning (1967); B. S. Minhas, Planning and the Poor (1974); Pranab K. Bardhan,
"India," in Hollis Chenery et al., Redistribution with Growth (1974); Deepak Lai, Prices for
Planning: Toward the Reform of Indian Planning (1980); Ursula Hicks, "Thirty Years of
Planning: The Indian Experience," Malayan Economic Review (October 1979); Gustav F.
Papanek, Pakistan's Development (1967); W. P. Falcon and G. F. Papanek (eds.), Develop-
ment Policy: The Pakistan Experience (1971); A. MacEwan, Development Alternatives in
Pakistan (1971); G. T. Brown, Korean Pricing Policies and Economic Development in 1960s
(1973); D. C. Cole and P. N. Lyman, Korean Development: The Interplay of Politics and
Economics (1971); Wolfgang Stolper, Planning Without Facts (1966); Andrew Kamarck,
Economics of African Development, rev. ed. (1971); Paul G. Clark, Development Planning in
East-Africa (1965); R. H. Green, "Four African Development Plans," Journal of Modern
African Studies, Vol. 3, No. 2 (1966); George B. Baldwin, Planning and Development in Iran
(1967); Bertram M. Gross (ed.), Action under Planning: The Guidance of Economic Devel-
opment (1967); Albert Waterston, Development Planning: Lessons of Experience (1969);
Louis M. Goreaux, Interdependence in Planning: Multilevel Programming Studies of the Ivory
Coast (1979); Nural Islam, Development Planning in Bangladesh (1977).

XI.C.3. The Future of

Development Planning— Note

After all the discussion of the preceding in the immediate future must be given to
chapters, we are left with the ultimate ques- those countries with a dismal development
tionof why the success stories of development record even after experiencing two or more
planning have been so few and the disap- decades of planning. To this end, it is ih
pointments so many. In this Note, it may be sary first to isolate the reasons why develop-
most appropriate to concentrate on the limits ment planning has not succeeded in coping
of development planning, instead of its ac- more effectively with the strategic polic\ i»
complishments, because concerted attention sues discussed in this book. We may then
 754 DEVELOPMENT PLANNING AND POLICYMAKING

consider ways to improve the practice of de- From the foregoing chapters we can iden-
velopment planning. tify three areas about which we need funda-
Many economists would suggest that the mental understanding before there can be a
reasons for past failures are to be found more intelligent basis for development plan-
mainly in an inability or unwillingness to im- ning. One area is that of the "unexplained re-
plement development plans. Planning agen- sidual factor." To the extent that our knowl-
cies in the developing countries are now ca- edge of the sources of economic development
pable of approximating the elements of a remains inadequate, we must also remain ig-
well-formulated plan in a creditable fashion. norant of relevant policy variables. This con-
And yet we are left with the disconcerting stitutes asevere limitation on the efficacy of
question: has the ability to implement a de- development planning. Even if we were to
velopment plan improved pari passu with the succeed in disaggregating the residual into
ability to formulate a plan? It can be main- recognizable elements (such as advances in
tained that the record of development plan- technical, managerial, and organizational
ning reveals that problems of implementation knowledge), it would still be difficult to spec-
now need more attention than problems of ify exactly how these elements are to yield to
formulation. planning. Planning may operate effectively
In this vein, Sir Arthur Lewis prefaces his on the supply of inputs, but it is quite a dif-
book Development Planning with the obser- ferent matter to bring the techniques of plan-
vation that "The economics of development is ning to bear on the income-raising forces con-
not very complicated; the secret of successful stituting the residual.
planning lies more in sensible politics and Similarly, it is uncertain by what means a
good public administration."1 The implica- development plan can have an impact on pop-
tion of this statement is that the major prob- ulation growth. And yet in many countries, a
lems of development planning center upon major challenge to development planning is
implementation — that the secret of success- the control of the income-depressing forces
ful planning lies essentially in policital sta- associated with heavy population pressures.
bility and political leadership and in compe- The record of development planning in this
tent and effective public administration of area has so far been disappointing, but many
the policy instruments by which the goals of development practitioners would now con-
development can be reached. But this has ad- tend that it is of prime importance to acquire
ditional implications — namely, that econo- new meaningful sociological and scientific
mists know why poor countries have re- knowledge of the determinants of the birth
mained poor and that they also know what rate and to institute as an essential element
policy measures would accelerate their of development planning a set of policies
development. aimed at birth control.
Before placing so much emphasis on im- This brings us to the third set of factors
plementation, we should raise, however, the with which it is difficult for development
logically prior question of whether econo- planning to cope — namely, the sociocultural
mists really do know what is wrong and how and other noneconomic factors that have an
to put it right. In answering this question, we effect on the process of development. If we
might reasonably argue that a major diffi- are to improve our understanding of the two
culty with development planning has been problem areas already discussed — the resid-
the economists' inadequate understanding of ual factor and population growth — we must
the development process, and that the secret also be able to evaluate more thoroughly the
of successful planning is not only — nor even noneconomic factors in the development pro-
mainly — a matter of implementation. cess. Itis now stating the obvious to say that
economic, social, and political change are all
'W. Arthur Lewis, Development Planning, Lon- interrelated. Nonetheless, we do not yet know
don, 1966, preface. under what conditions and by what mecha-
 755 PLANNING EXPERIENCE

nisms it is possible to have the types of socio- relative neglect of other development forces
cultural and political changes that will be that are not quantifiable but are of crucial
most favorable for development. Without importance (for instance, many aspects of
such knowledge, we cannot expect develop- human-resource development and sociocul-
ment planning to be very successful. Exactly tural and political changes for which no data
how is planning to bring about cultural exist); a bias toward concentration on the for-
change? Are attitudes, motivations, and in- mulation ofa development plan without due
stitutions amenable to change by planning? regard for its implementation (even though
Or is planning actually counterproductive, formulation and implementation should be
inhibiting the emergence of the favorable so- inseparable).
ciocultural factors? If policymakers cannot The quantitative bias in the theory of de-
identify the functional relationships among velopment planning tends to reinforce the
economic and noneconomic factors, and their other biases in development planning. Inter-
quantitative significance, how can they deter- sectoral programming models and other
mine whether to operate on economic incen- planning models are generally national plan-
tives, or attitudes, or organizational struc- ning models because the locus of industrial
ture, or social relations, or any of the many policy and the sources of data are usually
other factors that connect economic and non-
economic change? Despite the countless as- Data accessibility also reinforces the
national.3
sertions that the interaction between eco- macro bias, an emphasis on physical rather
nomic and noneconomic variables is of than human capital, and a focus on a few
utmost importance, the past emphasis has large projects rather than many small proj-
been on economic planning. The planning of ects. Plans for industrial development also
social or political change has been negligible. tend to dominate insofar as the outputs of in-
But the future success of economic planning dustry are more easily measured and indus-
may depend upon an understanding of how to trial inputs are more readily specified than in
plan social and political transformations.
And to achieve this, more than "sensible pol- As for their substantive content, many de-
agriculture.4
icies and good administration" are required. velopment plans have until recently empha-
At the same time as (and possible because) sized inward-looking policies to the relative
policymakers have remained ignorant about neglect of outward-looking policies; the de-
these fundamental aspects of the develop- velopment ofthe urban industrial sector with
ment process, a number of biases have been much less concentration on rural develop-
revealed within the actual planning process.2 ment; and the simple imitation of the ad-
What has been unfortunate about these vanced countries' institutions at the expense
biases is that some of them have actually in- of innovation and adaptation.
tensified the constraints on development and In our previous discussions of the need to
others have kept the actual rate below the at- mobilize domestic and external resources,
tainable rate of development. promote agricultural development, encour-
If we may refer to a characteristic style of age foreign trade, and invest in human capi-
development planning, three types of bias tal, we have repeatedly encountered argu-
have been prominent: a bias toward macro- ments in support of those very policies that
models in plan formulation to the relative ne- have actually been ignored by many of the
glect of the microeconomic aspects of plan- past development plans. Recognizing the past
ning (such as project analysis); a bias toward
the quantitative aspects of planning to the 3See Don Humphrey, "Some Implications of Plan-
ning for Trade and Capital." in Max F. Millikan
(ed.). National Economic Planning. Nc* York. 1967.
2These are biases in the sense of both statistical 4Cf. Michaei Lipton, "Urban Bias and Agricultural
skewness and unwarranted valuations. Cf. Gunnar Planning," in Paul Streeten and Michael Lipton
Myrdal, Asian Drama, New York, 1968, appendix 4. (eds). The Crisis of Indian Planning. London, 1968.
 756 DEVELOPMENT PLANNING AND POLICYMAKING

biases in development planning, many devel- sense, more planning is actually required to
opment economists now maintain that the overcome the results of inadequate planning
time has come to rethink the nature and in the past: improved planning is now neces-
scope of development planning. It has been sary to remove the distortions caused by ar-
increasingly recognized that the comprehen- bitrary direct administrative controls that
sive "heavy-type" of central planning is still have produced a disequilibrium system and a
premature for most of the LDCs. set of trade, fiscal, financial, industrial, and
The plea is not, however, for a reversal to wage policies that are often contradictory
laissez-faire — but a plea for competent plan- and self-defeating.
ning. To achieve this, students of develop- The case for a high degree of planning,
ment must devote more research to determin- with a large amount of public investment and
ing the appropriate range and forms of deliberate industrialization constituting the
planning for a developing country (for in- core of the plan, may remain strong for coun-
stance, indicative versus controlling, or for- tries that suffer seriously from the pressure of
mal versus informal planning). The problem population on the land but have the potential
is to identify and institutionalize the most ap- for large domestic markets (such as India or
propriate way of planning for the particular China). For other countries, however, in
country at a particular period. An evolution which there are underutilized natural re-
of the planning approach itself would there- sources but only small internal markets, it
fore be expected as the country develops. can be persuasively argued that the basic
More immediately, some major revisions problem is one of creating a favorable social
of development planning can be suggested to and economic environment which will lead to
remove the biases and secure more benefits expansion of private activity, more effective
from planning in the future. These revisions use of the underutilized resources, and capi-
amount, on the one side, to making greater talization on the existing opportunities for in-
use of the market mechanism as an instru- ternational trade.
ment of development policy within the do- If the private sector is to be enlarged, how
mestic economy, while encouraging, on the is this now to be accomplished? Instead of al-
other side, an extension of multinational lowing any economic resources to be left un-
planning in the international economy. employed, governmental policies must try to
Many economists have come to believe mobilize through positive economic incen-
that in order to overcome past deficiencies in tives and inducements the latent skills and
planning and undertake policies that conform capital in the private sector. Fundamentally,
more closely to their present needs and ca- the stimulation of multiple centers of initia-
pabilities, the majority of LDCs should re- tive depends on the establishment of markets
trench to a lighter type of planning. This and encouragement of market institutions.
would rely more on decentralized decisions Economic and social overhead capital can
operating through the market mechanism, help to establish the physical conditions for a
and greater attention would be given to de- market to exist and can support the interde-
vising policies that might make private action pendence of markets. The government also
more effective. has a crucial role in building institutions such
The advocacy of greater use of the market as a banking system, a money and capital
mechanism should not be interpreted as a market, agricultural cooperatives, labor or-
call for a diminished role for the government, ganizations, rural credit institutions, and
but rather a different role. Governmental pol- training institutes. It must also be recognized
icies are needed to strengthen the market sys- that many policy measures that can affect in-
tem, and a stronger market system is in turn dividual action by altering the economic en-
needed to allow public policy to operate more vironment are not of the usual monetary or
effectively through the market. Thus, in a fiscal type of policy, but rather of a kind that
 757 PLANNING EXPERIENCE

involve the legal and institutional framework, ous specific controls. We have repeatedly
such as land tenure legislation, commercial noted criticisms that interest rates are artifi-
law, and property rights. cially low in the urban sector, exchange rates
Once market imperfections are reduced are overvalued, unskilled wages are too high
and the structure of markets improved, the in labour surplus economies, subsidization of
market mechanism can itself be used as an import-substitutes has become suboptimal,
instrument of development — promoting gov- exports and agriculture are being discrimi-
ernmental policies as well as more effective nated against. Accordingly, there are now
private activity. Instead of relying on com- more proponents advocating the adoption of
prehensive and detailed administrative con- flexible exchange-rates to avoid currency ov-
trols, the government can alter prices to ex- ervaluation, the removal of price controls on
ecute policy and can provide price and foodstuffs, and the liberalization of foreign
income stimuli for an expansion in private trade controls with the substitution of domes-
output, an increase in exports, and a widen- tic subsidy and tax schemes. It is argued that
ing of domestic markets. These price changes a more realistic price structure would remove
may extend to foreign exchange rates, inter- the need for detailed investment planning
est rates, tariffs, taxes, and subsidies. Subsidy and would induce additional private activity
and tax schemes can be especially relevant in Although there are good reasons why the
inducing firms to value inputs according to market mechanism should have a large role
their social opportunity costs, to exploit ex- within the LDC's domestic economy, there is
ternal economies, or to introduce new tech- at the same time a need for more multina-
niques ofproduction. tional planning for development within the
Of most importance is the need to remove international economy, as outlined in the
the distortions in internal price relations next Note.
which have resulted from the use of numer-

Comment
Many commentaries have been offered on the proposals for a New International Economic
Order (NIEO), advocated by the developing countries in the United Nations and its agencies
during the mid-1970s. The proposals emphasize the importance of reforming the international
environment to ensure successful development planning. The basic themes of the NIEO can
be noted in the document of UNCTAD. New Directions and Structures for Trade and De-
velopment (1976) and in the Declaration and Program of Action of the NIEO, presented at
the Sixth Special Session of the General Assembly of the United Nations in 1974.
For a discussion of the issues raised by the NIEO, see Jagdish N. Bhagwati (ed.), The IS ex
International Economic Order: The North-South Debate (1977); Edward P. Reubens (ed.),
The Challenge of the New International Economic Order (1981); Albert Fishlow et al .. Rich
and Poor Nations in the World Economy (1978); Robert W. Cox, "Ideologies and the Ne*
International Economic Order," International Organization (Spring 1979); l.ance Taylor,
"Back to Basics; Theory for the Rhetoric in the North-South Round," World Development
Vol. 10, No. 4 (1982); Hans Singer, "The New International Economic Order: an Overview/1
Journal of Modern African Studies, Vol. 16, No. 4 (1978); Mordechai E. Kreinen and .1 \1
Finger, "A Critical Survey of the NIEO," Journal of World Trade Law (1976); Jurgen B
Donges, "The Third World Demand for a NIEO," Kyklos{\911)\ W. Loehr and J.A Powelson, Selective
Threat to Development: Pitfalls of the A7£0 (1983); United Nations, The NIEO:
Bibliography (1980). Especially noteworthy is W. M. Corden, The NIEO Proposals: A Cool
Look{\919).
 XI.D. INTERNATIONAL
POLICYMAKING AND THE
NIEO-NOTE

For many developing countries, the quality of to developing countries and codes of conduct for
national economic management has dis- multinational enterprises.
tinctly improved. At the international level, Following the United Nations declaration,
however, policymaking shows little improve- some proponents of the NIEO interpreted the
ment and may actually have deteriorated in
new order as providing exemptions for LDCs
several problem areas. The world order is from established rules, thereby granting
now less responsive to the needs of the LDCs
some important short-term concessions such
than a decade ago. Although many of the
as more aid, debt relief, and trade conces-
preceding selections have argued that the
sions. Others identified a NIEO with long-
market mechanism should have a larger role
term structural changes in the international
within the LDCs domestic economy, it can
also be maintained that there is a need for system that would alter biases in income,
wealth, and power distributions between
more effective international policymaking in LDCs and MDCs. As one proponent of a
the interests of development. Improved poli- NIEO stated,
cymaking atthe international level would not
be a substitute for necessary internal re- The basic purpose of the current demand for a new
forms, but it would be complementary and international economic order is to restructure the
would extend the effectiveness of the internal prevailing market rules, largely fashioned by the
reforms. financial power of the rich nations and their multi-
nationals; toobtain a greater voice in international
To plead for improved international poli- financial institutions; and to break the age-old pat-
cymaking isnot, however, necessarily to en- terns of economic and political dependency of the
dorse the call for a New International Eco- poor nations on the goodwill of the rich nations.
nomic Order (NIEO). The NIEO is subject The main objective, however politely or skillfully
to different interpretations. The proposals for stated, is restructuring of power; whether political,
an NIEO, as adopted at the Sixth Special economic, financial, or intellectual.1
Session of the United Nations General As-
sembly, included the following: This interpretation sees in the post-colonial
power structure a continuation of domination
(a) An "integrated" program of price supports and dependence, caused not only by rules,
at levels higher than historic trends for a group of procedures, and institutions designed by the
commodities exported by developing countries.
rich countries, but also by the totality of eco-
(b) The indexation of prices of exports of devel- nomic, political, and even cultural relations
oping countries to prices of their imports from de-
veloped countries. among nations. This view stems from the ear-
(c) The attainment of the target of 0.7 percent lier analyses of Singer, Prebisch, and Myrdal
of gross national product (GNP) of developed that see polarization or backwash effects as
countries for official development assistance. the causes of failure in LDCs.2 According to
(d) The linkage, in some form, of development this interpretation, the rules of international
aid to the creation of international reserves in relations must be changed by removing
terms of special drawing rights (SDRs) on the In- biases in the present rules, by the exercise of
ternational Monetary Fund (IMF). countervailing power where at present the
(e) The so-called Lima target for shifting man-
ufacturing capacity from the developed to devel- 'Mahbub ul Haq, "Beyond the Slogan of South-
oping countries to the extent of 25 percent of world South Cooperation," in Khadija Haq (ed.), Dialogue
industrial output by the year 2000. for a New Order, New York 1981.
(f) Mechanisms for the transfer of technology 2See Chapters II, VIII.

758
 759 INTERNATIONAL POLICYMAKING AND THE NIEO— NOTE

distribution of power is felt to be unequal, vocates of better market access in the indus-
and by counteracting biases that arise not trial countries and can form strong pressure
from rules but from the nature of economic
groups.
Another view is that a NIEO is not to
processes that inflict cumulative damage on
the LDCs. Although a change in only the emerge through the benevolent generosity of
economic relations among nations may allow the rich nations but only through the efforts
scope for positive-sum games, the emphasis of the South in restructuring domestic polit-
on national power relations among sovereign ical and economic power and in organizing
states necessarily involves the relative con- greater collective bargaining at the interna-
cept of power, and hence entails a zero-sum tional level.
game. The demand for greater participation For many LDCs, however, the acquisition
in international decision making and for cor- of countervailing power remains remote, and
rections in the basis of the international the tactic still smacks of an adversary posi-
power distribution are bound to diminish the tion. A more cooperative approach might
power of the richer countries.3 build on the enlightened self-interest of the
Others have pointed out that there is really MDCs and their sharing with the LDCs of
nothing "new" to the NIEO. Some of its pro- some mutual needs for international eco-
visions were already advocated in the 1950s nomic management. Mutual needs are often
by Singer, Prebisch, and Myrdal. Moreover, analyzed in terms of an interdependent world
the NIEO is not "international," "eco- economy. Much has been written recently
nomic," nor "an order." On the contrary, about interdependence, especially from the
many of its objectives are highly nationalistic viewpoint of energy, resource balances, and
and of a zero-sum game character, favoring the global food/population equation. But in-
the LDCs instead of emphasizing mutual terdependence inot
s a new phenomenon; for
gains for all actors in a positive-sum game. centuries, world economic integration has
Furthermore, the policy objectives and policy been increasing, and the international spread
instruments are as much political as eco- of development is only a continuation of the
nomic. And, finally, given the lack of una- process. In the past two or three decades,
nimity and the willingness of some countries however, the consequences of interdepen-
for North-South confrontation, the measures dence have become more pronounced; the du-
are more likely to produce disorder than ration ofinternational economic conflicts has
order. lengthened; the extent of economic gains and
Economists may criticize the ill-designed losses has grown; and the inadequacy of pol-
measures of the NIEO, but there is still va- icy measures to resolve conflicts has
lidity insome of the objectives sought by the increased.
NIEO. The crucial question, however, is: This might be better interpreted as the re-
How can these objectives be realized? It is sult of the internationalization of the eco-
unrealistic to believe that the MDCs will nomic system. Outstanding features of this
simply grant the demands of the LDCs out of internationalization process have been the in-
a sense of altruism or humanitarianism or creases inthe international flows of commod-
morality. Nor can UNCTAD be relied on to ities, factors of production, management,
be an effective interest group for the poor technology, and financial capital, which have
countries. A more promising possibility lies become ever more responsive — or more elas-
in the enlistment of interest alignments tic— to differences between domestic and for-
among nations. Multinational and transna- eign variables. To an economist, this inter-
tional firms are, for example, important ad- nationalization ofmarkets is commendable
because it promotes efficiency, specialization,
and competition. To a national policymaker,
3Paul Streeten, "The New International Economic
Order," World Development (January 1982), pp. 1-
however, internationalization has a negative
17. side: it heightens the vulnerability of a nation
 760 DEVELOPMENT PLANNING AND POLICYMAKING

to external developments. Domestic auton- The proper response to these conflicts does
omy in policymaking is subordinated to in- not lie in succumbing to the rhetoric of the
ternational policy considerations. Interna- New International Economic Order. The
tional economics is opposed by national more realistic approach is to recognize that
politics. When the economic objectives of two the demands for a NIEO reflect the conse-
or more nations clash, international tension quences of the internationalization process
and conflict result. with its attendant tensions and conflicts. But
The conflicts can be grouped into three the adversary national interests are being
categories: (1) those that arise because a na- given a particular North-South coloration.
tion seeks to acquire a larger share of the The problems of development must therefore
gains from trade or foreign investment; (2) be factored into the solutions of the problems
those that arise when a country tries to avoid of conflict that arise from the internationali-
being damaged by developments in another zation process. This means that economic in-
country; and (3) those that arise because a dependence cannot simply be legislated in re-
country wants to maintain its domestic au- sponse to the demands of UNCTAD or the
tonomy in policymaking when confronted NIEO as if the metropolitan country were
with an international event. legislating political independence. Economic
More specifically, the major conflicts tend independence must be attained through a
to concern the following issues: structural transformation of the developing
economies in which industrial production be-
1 . Markets: The attempt of each nation to comes alarge proportion of total exports.
have greater access for its exports to the The disarray of the international economy
markets of other countries and ready ac- has had adverse effects on development. Nei-
cess to imports of needed resources from ther market forces, as would be represented
the markets of other countries. by free trade or freely floating exchange
2. Terms of trade: The attempt of each rates, nor international codes of conduct now
country to improve its terms of trade by prevail in establishing world economic order.
raising export prjces relative to import There is always the danger that the regula-
prices. tion of international economic conduct will
3. Terms of foreign investment: The at- be abandoned to either simple unilateral ac-
tempt by host governments to raise the tion or ad hoc negotiation dependent on bar-
benefit-cost ratio of foreign capital gaining power. In such an environment, the
inflows. traditional beliefs of economists in a har-
4. Adjustment costs to imports: The at- mony of interest, mutual gains from trade,
tempt of each nation to minimize market foreign investment as a nonzero-sum activ-
disruption or domestic injury from greater ity— all the beliefs that support internation-
imports. alism over nationalism — would be sub-
5. Costs of balance-of-payments adjust- merged. To avoid the dangers of nationalism
ment: The attempt of each nation to and policy competition among nations, more
minimize the cost of adjusting to a dis- international coordination of policies is
equilibrium inits balance of payments by needed. Many of the policies recommended
avoiding remedial policies or by trying to in this book depend on such coordination for
place some of the burden of adjustment on their effectiveness. The liberalization of for-
other countries.
eign trade regimes, promotion of outward-
6. Stabilization policies: The attempt of looking policies, mobilization of resources,
each nation to exercise national economic coordination of international monetary poli-
autonomy in stabilizing its own economy cies, and cooperation between IMF, World
without subjecting its policies to external Bank, and commercial banks, all require a
conditions. diminution in competition among national
 761 INTERNATIONAL POLICYMAKING AND THE NIEO— NOTE

policies and more cooperative international without the supportive existence of an inter-
action. national public sector? In the domestic econ-
Although for some problems the quality of omy, the public sector is charged with the
policymaking might be improved at the na- maintenance of full employment, redistribu-
tional level by greater use of the market sys- tion of income, and correction of market fail-
tem, the more complex problems of develop- ures. But where is the responsibility for the
ment may require for their correction an performance of analogous functions in the
extension of planning at the higher interna- world economy? The absorption of surplus
tional level. Jan Tinbergen has emphasized labor requires an international full-employ-
the series of levels at which decisions have to ment policy. An international redistribution
be taken.4 In order that the decisions regard- of income requires a more effective mecha-
ing necessary policy instruments ("action pa- nism for transferring resources from rich to
rameters") be optimal, there must not be poor countries. The pervasiveness of interna-
"external" effects — that is, the influences ex- tional market imperfections — whether ex-
erted on the well-being of groups outside the pressed as price distortions, transfer of inap-
jurisdiction of those who make the decision
should be weak. The area in which the im- propriate technology, or as "backwash
effects" — requires for their correction some
pact of the instrument will be felt determines international authority that can transcend
what decision level will be optimal. The level the limits of national jurisdiction.
should be high enough to cover the area in The task ahead is to realize the potential-
which the impact is nonnegligible. Thus, ities of more extensive international gover-
while the newly developed "nation state" is a nance. Although we can discern only the ru-
political unit and a cultural unit, it is an in- diments ofan international public sector, we
appropriate economic decision-making unit may submit that the future economic devel-
for many developmental issues. Decisions opment lies in overcoming the fragmentation
made at the national levels are often at far
in policymaking that has characterized de-
too low a level to be optimal. Trade liberali- velopment policies within the poor countries
zation and access to markets in the developed and between rich and poor countries. This ad
countries, the monitoring of export controls hoc piecemeal approach to policy has left a
and access to resources from primary-pro- distorted policy framework that is third, or
ducing countries, more multilateral aid pro- fourth — or nth best.
grams, harmonization of aid terms, compen- A student of international development
satory financing measures, international has conveniently summarized the services
stabilization policies, supplementary finan- that international agencies may be required
cial measures, coordination of international to perform; most likely, these agencies will
monetary policies — all these policies depend be:
on cooperative international action. (i) Providing a framework for the resolution of
After the development experience of the conflicts between member states.
past quarter-century, the central question for (ii) Promoting joint activities between member
the next quarter-century will be: Can na- states, which reap the advantages of scale, e.g.,
tional programs of development succeed under complementarity or production sharing
agreements.
(iii) Providing assistance in the mobilization of
resources.
4Jan Tinbergen, "Building a World Order," in J. (iv) Providing advice in fields which require spe-
N. Bhagwati (ed.), Economics and World Order cialized technical skills.
from the 1970s to the 1990s, New York 1972, pp.
(v) Providing assistance in the recruitment of
145-7; see also Martin McGuire, "Group Segrega-
personnel or organizations to perform specified
tion and Optimal Jurisdictions," Journal of Political
Economy, Vol. 82, January-February 1974, pp. 112- tasks, e.g., engineering contractors, or managing
32.
agents.
 762 DEVELOPMENT PLANNING AND POLICYMAKING

(vi) Acting on member countries' behalf in ne- pect a continued deterioration in develop-
gotiations with external organizations, e.g., foreign ment performance. So too will the potential
private investors, or aid agencies.5 to meet mutual needs be lost if in the quest
for a new international order nations assume
To promote these services more effectively,
some international economic institutions may adversary positions.
have to be reformed, and the specialized in- It is hoped that the basic economic princi-
ternational agencies will have to coordinate ples emphasized in this book, together with
their programs more fully. In particular, a the lessons of recent development history,
number of proposals have been made for might provide a basis for future progress to-
ward more effective development policymak-
changes in world trade institutions.6 Im-
provements indevelopment policy are likely ing. This book has sought to bring out the is-
to come ultimately from more cooperation sues at stake. If improved policymaking can
among the international specialized agencies, follow, there is hope for a NIEO in which the
regional agencies, and national governments "new" emphasizes the mutual interests of
of rich and poor countries alike. If instead of rich and poor countries alike, in contrast to
national interests, and provides the means of
supporting an international order for devel-
opment policy, each LDC simply pursues conflict resolution so that the "order" can
narrow national policies, we will have to ex- allay the disorderly conflicts of the interna-
tionalization process. Even if a perfect set of
5John White, "International Agencies: The Case policies cannot be formulated and imple-
for Proliferation," in G. K. Helleiner (ed.), A World mented, better policymaking can do much to
Divided, New York 1976, p. 288. reduce the disorder of the world economy,
6American Society of International Law, Remak- and to mitigate absolute poverty, to improve
ing the System of World Trade: A Proposal for In- the distribution of income, and to provide
stitutional Reform, 1976; Atlantic Council, GATT
Plus: A Proposal for Trade Reform, 1975; Miriam more productive employment — for the bene-
Camps, Collective Management, New York 1981. fit of both rich and poor countries.
INDEX

Absolute poverty, 1, 20-23, 43, 424 Balanced growth: case for, 373-76; versus
Absorptive capacity, 220, 285 unbalanced growth, 377-80
Accounting rate of interest (ARI), 665, 678-79. See Balance of payments: constraint, 728; and country
also Project appraisal risk analysis, 31 1-13; equilibrium, 731; and
Accounting prices, 652-56. See also Shadow prices international monetary system, 315-21; statistics
Achievement-oriented, 107 of, 306. See also Foreign exchange
Agriculture: as bottleneck, 85-86, 423; bias against, Balassa, Bela, 522, 541
195; contribution to development, 427-36; and Baldwin, R. E., 392, 509
exports, 410-12; interactions with industry, 406- Baran, P. A., 134, 139
13; and international trade, 466; price incentives Barber, W., 162
for, 471-83; squeeze, 430; statistics of, 384, 415- Bardhan, P. K., 463
18, 440; surplus, 432-37; and urban bias, 437-40; Basic needs, 26-30, 48-49, 114-15, 568
weakness of, 418-21. See also Dual economy; Bauer, P. T., 66, 353, 494
Industrialization; Labor surplus Beckerman, W., 13, 16
Ahluwalia, M. S., 120 Berg, A., 503
Aid, 293-309: contributions, 298-301; criticisms of, Bergsten, C. Fred, 556-57, 559
388-97; improving quality of, 301-5; statistics of, Bhagwati, Jagdish, 509, 522
289-92; through trade policy, 487; tying of, 302- "Big Push," 361-68, 632
5. See also Capital requirements Bimodal agricultural strategy, 454-56
Amin, Samir, 134-35 Bird, R. M., 284
Andean common market, 329 Boeke, J. H., 150, 167n., 170
Appropriate technology, 345-49. See also Choice of Border prices, 676-77. See also Little-Mirrlees
techniques; Technology method
Arrow, Kenneth, 601 Brandt Report, 309, 320
Autarky, 715 Brazil, 24,62-66,410-11
Bretton Woods, 2, 549
Backwardness, degree of, 101-4 Bruno, Michael, 121, 741
Baer, W., 741 Bruton, H. J., 743

763
 764 INDEX

Cairncross, A. K., 146 Development economics, as subdiscipline, 89, 145-46

Capital accumulation: generalized process, 631-36; Development, meaning of, 5-10
role of, 225-29; sources of, 234ff. See also Dual Development through trade, 485, 489-93. See also
economy; Savings; Investment Exports; Export substitution; Foreign investment
Capital-intensive bias, 194-95. See also Discount rate, 646-51. See also Accounting rate of
Employment interest; Consumption rate of interest; Project
Capital markets, 717ff.
Capital-output ratio, 199n., 215-16, 220-25, 364 appraisal of income, 15-19,47-48, 120, 128,
Distribution
Capital requirements, 223, 283-86 455-56, 655-56, 688-91, 693, 702-4; growth
Capital stretching, 349-52, 409 trade off, 36-44, 87; in Brazil, 63-66; in China,
Capital transfer, 33, 44 72-74. See also Entitlements
Capital-turnover criterion, 429 Domar, Evsey, 729-30
Cartels, in primary products, 555-60, 724 Dos Santos, Theotonio, 141
Cassen, Robert H., 574 Dual Economy, 162-67, 191-95, 214; models of,
Caves, R. E., 391 407-8, 412-13; and technical coefficients, 167-70;
Chelliah, R. J., 235 and wage differentials, 170-75. See also Dualism;
Chenery, H. B., 9, 121,509,741 Labor surplus; Lewis; Ranis-Fei model
China, 67-77, 461 Dualism, 125, 149-51, 191-95, 407ff.; technological,
Choice, logic of, 733 167-70,215
Choice of techniques, 159-61, 194, 207-10, 212-13,
352-56, 433, 694-700. See also Appropriate
technology; Technology ECLA, 502
Cline, W. R., 612 Econometric models, 723-24
Colombia, 23, 48 Education, 22, 568; and human capital, 582, 610—
Colonialism, 501-2, 716, 732. See also Dependency 31, 725; planning of, 725. See also Human
Comparative advantage, 341, 489, 492. See also resources
Development through trade; Exports Edwards, E. O., 536
Complementarities, 632 Effective exchange rate, 539-40, 542
Conditionality, 311 Effective protection, 1 1 3
Constraints, 85-87, 1 1 1 Efficiency prices, 690-91. See also Project
Consumption rate of interest (CRI), 686-87. See appraisal; Shadow prices
also Project appraisal
Control areas, 688 Egypt, 468
Eicher, Carl K., 475, 622
Cooper, Richard N., 558 Ellis, H. S., 637
Corden, W. Max, 491, 510, 775 Emmanuel, A., 502-3
Cost-benefit analysis, 643ff., 646, 652-57, 718-21. Employment, and output tradeoffs, 198-205;
See also Little-Mirrlees method; Project policies, 702-6, 725-26. See also Capital-
appraisal; Shadow prices; UNIDO method intensive bias; Unemployment; Labor surplus
Country risk analysis, 310-14. See also External Enclaves, 666
debt
Engels' law, 119
Country risk matrix, 3 1 3 Entitlements, 78-84
Cumulative causation, 499 Entrepreneurship, 548, 626
Customs Union, 772 Equity. See Distribution of income
Exchange control, 514-16
Exchange rates, 315-21, 652, 681-85. See also
Debt-servicing capacity, 351. See also Country risk Balance of payments; Foreign exchange
analysis Export-led development, conditions of, 503-9. See
Decision-taking, 378-79 also Growth through trade
Demonstration effect, 71, 504-5, 717 Export pessimism, 542-44, 546-52
Dependencia. See Dependency Exports: advantages of, 715; enclaves, 716;
Dependency, 5, 90, 111, 113, 121, 135, 139-44, promotion of, 510-12, 521-22, 538-41; statistics
502 of, 399. See also Development through trade;
Devaluation, 315, 736. See also Balance of Export pessimism; Export substitution;
payments; Foreign exchange Industrialization; Linkages
Development decades, 129, 167 Export substitution, 394-97, 522-29, 630
 765 INDEX

Extended family, 137 Growth through trade. See Comparative advantage;

External debt: and growth stages, 286-88; and debt Export-led development; Exports; Foreign
servicing, 304-5; country risk analysis of, 310-14; investment; Import substitution
statistics of, 307-8. See also Balance of
payments; Debt servicing; Foreign exchange
External economies: types of, 362-63, 738. See also Haberler, Gottfried, 496, 509
Externalities
Harrod, R. F., 728-29
Externalities, 610-18, 656, 694 Harrod-Domar model, 94, 128-29, 222
Hart, Keith, 189
Healey, Derek, 564-65, 722, 745
Factoral terms of trade, 503 Health, 581-82, 584-91; statistics of, 585-90
Factor markets, 658-71 Helleiner, G. K., 746
Factor-price equalization theorem, 492, 498-99 Hilgert, Folke, 458
Factor proportions, 208-10. See also Choice of Hirschman, A. O., 122, 332, 377-79
techniques; Technological dualism Human capital, 86, 112, 567ff., 717; investment in,
Falcon, Walter P., 474, 560, 586 213, 610-14. See also Education; Human
Family planning, 576-80. See also Population resources
Famines, 601. See also Entitlements Human resources. See Basic human needs;
Fei-Ranis model, 150-64 Education; Human capital; Malnutrition;
Fertility determinants, 576-80. See also Population Population
Finance, nominal versus real, 305 Humphrey, D. D., 494
Financial development, 216; deepening, 240-46;
repression, 216, 254-58; structure, 289. See also
Fragmented economy; Savings; Investment ILO, labor missions, 151, 190. See also Informal
First-best conditions, 639-40 sector
Fishlow, A., 82 IMF, 2, 129, 271, 277-79, 315, 318-20, 550. See
Food deficit countries, 482 also International monetary system
Food prices, 448, 476-83 Import substitution, 194, 212, 333-35, 388-98, 516-
Foreign aid. See Aid 22; in India, 749-50; and protection, 725; stages
Foreign exchange constraint, 86, 274, 286, 681-85.
of, 517-19
See also Exchange control; External debt; Two- Inappropriate policies, 193-95
gap approach; Price distortions; Trade policy Incomes policy, 240
Foreign exchange rates, 315^21. See also Balance India, 5, 334, 446ff., 749-50
of payments; Foreign exchange constraint; Price Indivisibilities, 363-64
distortions Indonesia, 446ff.
Foreign investment: and contractual arrangements, Industrialization: achievements of, 381-83; and
330; and external economies, 324-25; benefits and agriculture, 406-18; and big push, 361-68; and
costs of, 322-31; policy towards, 323; protection exports, 397-405; and linkages, 368-73; statistics
of, 327-28; terms of, 380; and terms of trade, of, 384-88; via export substitution, 394-97; via
326. See also Project appraisal import substitution, 388-94. See also External
Foreign trade. See Trade economies
Formal sector, 183-90 Inequality, 7. See also Distribution of income;
Fragmented economy, description, 716-17 Entitlements
Frank, Andre Gunder, 112, 134, 139-42 Inflation, 216-17, 244, 250, 254-74, 652. 739
Frisch, Ragnar, 142, 746 Informal sector, 114, 183-90
Information, 744
Innovations, 349-52
Galenson, W., 212 Input-output analysis, 129, 188, 398-400. 677-78,
GATT, 2, 528, 538, 549, 559, 563-64 720. 724. 726, 730
Gerschenkron, Alexander, 101-4, 106n., 253 Integrative process, 508
Ghana, 517 Internal rate of return (IRR). 310, 450. 646-51,
Gini coefficient, 25, 128 658, 674-75. See also Project appraisal
Green revolution, 126-27, 423-24, 441-46; International coffee agreement. 553
problems of, 446-52, 465-66. See also International commodity agreements, 553-60
Agriculture Internationalization process, 759-61
 766 INDEX

International monetary system, 315-21. See also Leys, Colin, 746

IMF Life expectancy, statistics, 601-2
International public sector, 781-82 Linkages, 506-8; and industrialization, 368-73; and
International trade. See Trade transnational enterprises, 332-37; generalized,
Intertemporal optimality, 696-99, 706 373; vertical, 174
Inverted-U, and inequality, 30-36, 128; and Little-Mirrlees method, 41, 652-57, 665, 675-81.
inflation, 175 See also Project appraisal
Investment criteria. See Project appraisal Lorenz curve, 15
Investment, foreign, 322-44, 737; public, 737-38, Lorenz dominance, 80
748, 751; statistics of, 206n., 306. See also
Capital accumulation; Project appraisal
Inward-looking policies, 5. See also Import Malnutrition, 568, 591-601. See also Human
substitution resources; Nutrition
Ivory Coast, 21 Malthusian trap, 150, 154
Malthus, T. R., 375, 569
Manpower training, 619-25
Japan, 37, 173, 175, 211, 253, 409, 446, 453, 459 Manufactured exports, 509, 522-29; statistics of,
523-24, 526-27, 530-37, 543-45. See also
Export substitution; Foreign investment;
Kaldor, Nicholas, 235 Multinational enterprises
Keesing, Donald B., 525, 538 Marginal social cost (MSC), 639, 658-60. See also
Kenya, 114, 176-77, 183-90 Project appraisal; Shadow prices
Keynesian analysis, 119, 145-46 Marginal social value (MSV), 639, 658-60. See
Killick, Tony, 747 also Project appraisal; Shadow prices
Kirk, Dudley, 571-72 Market, versus planning, 709-16
Korea. See South Korea Market failure, 709-11
Kornai, Janos, 146 Market safeguards, 538. See also Protection
Kravis, Irving B., 31 Marshall, Alfred, 374
Krishna, Raj, 476 Marxist analysis, 90-93, 112, 120-22, 133-39, 141.
Krueger, Anne O., 522, 533, 538 See also Dependency
Kuznets curve, 36, 40-41 Mechanization, 446
Kuznets, Simon, 30-32, 39, 91, 106, 128, 149-50, Merit wants, 693. See also Project appraisal
368, 427, 429 Migration: model of, 176-81, 211; mathematical
model, 179-81; statistics of, 182. See also Dual
economy; Labor
Labor market, 210-14. See also Dual economy; Mill, J. S., 490, 569
Disguised unemployment; Employment Monetarist approach, 216, 268-73, 277-78. See
Labor, mobility of , 7 1 5 also Inflation
Labor surplus, 125, 150, 162-67, 172, 236, 593-94, Money markets, 246-50, 254-56. See also Financial
666; meaning of, 153-59. See also Disguised development
unemployment; Labor market; Underemployment Morris, C. T., 32ff.
Ladejinsky, Wolf, 444 Multinational enterpirses, 322-44, 551; and choice
Lall, S., 142 of techniques, 352-56; and import substitution,
Land reform, 468-71, 748, 751-52. See also 333-35; and export orientation, 335-37. See also
Agriculture; Rural development Foreign investment; Technology
Latin America, 139-42, 268-73, 587, 589-90 Myint, Hla, 237-38, 509
Leading sectors, 90, 97-100, 491. See also Stages of Myrdal, Gunnar, 117, 118, 492, 746, 758
growth
Learning by doing, 717
Learning effect, 507 Natality, 487, 576-80
Leibenstein, Harvey, 169 Net present value (NPV), 639, 645-51, 658, 663.
Leontief , Wassily, 1 30 See also Project appraisal
Lewis, W. Arthur, 9, 113, 118, 124, 128, 150-51, New International Economic Order (NIEO), 116,
156-58, 162-67, 176, 191-95, 320-21, 427, 429, 708, 757-62
476,612,617 Newly industrializing countries (NICs), 538ff.
 767 INDEX

North-south. See Brandt Report; New International Project appraisal: analysis of, 637-706; benefits and
Economic Order
costs, 640-41, 652-57; and foreign investment,
Numeraire, 663-65, 689-90. See also Little- 330-31; rationale of, 639-41; measures of
Mirrlees method; UNIDO method profitability, 646-51. See also Cost-benefit
Nurkse, Ragnar, 112, 119, 157-58, 377-80, 509 analysis; First-best conditions; Second-best;
Nutrition: objectives, 591-97; policies, 597-601; Shadow prices
statistics of, 603-9. See also Human Protection, 388-94, 485-86, 654-55. See also
resources Import substitution

OPEC, 115, 143,555-58

Quotas, 733-36
Opportunity cost, 207, 672; of capital, 685-87. See
also Project appraisal
Optimization models, 720, 727, 740 Ranis-Fei model, 164, 167. See also Dual economy
Rational actor model, 747
"Optimum Population" theory, 155
Oshima, Harry T., 31, 213, 231 Regional integration, 547-52, 561-68
Residual factor, 2, 754
Ricardo, David, 427
Pakistan, 8, 446ff. Risk analysis, 687-88
Paukert, F. 33-35 Robertson, Dennis, 493
Peasant producers, 460. See also Agriculture Rosenstein-Rodan, P. N., 112-19, 366
Perroux, Francois, 506 Rostow, W. W., 90-100, 102, 104-6, 111-12
Physical quality of life index (PQLI), 10, 30 Rural development, 424, 458-63, 467. See also
Planning: biases in, 755-56; and cost-benefit Agriculture; Urban bias
analysis, 718-21; future of, 753-57; in practice, Rural markets, 463. See also Rural development
744-54; interpretations of, 719-21; models, 146- Ruttan, Vernon, 480
57, 723-42; objectives of, 692-94; policy
instruments of, 732-42; versus price system, 707,
709-16,721-22 Savings, statistics, 306
Polarization, 693, 758 Schultz,T. W., 475. 496, 611
Politics, 746-47 Scitovsky, Tibor, 744
Population: demographic trends, 572-76; and Scott, Maurice, 80-81
employment, 213-14; growth, 85, 153-55, 567- Second-best, 638, 694-700, 715
82; Marxist view of, 137-38; policy, 19, 86; and Seers, Dudley, 115, 745
poverty, 569-72; quality, 580-85; statistics of, 55, Semi-industrialized countries, 382-83. See also
181-82, 575-76. See also Human resources Industrialization
Poverty: dimensions of, 20-23; line, 1, 43. See also Sen, A. K., 165,601
Absolute poverty; Entitlements Shadow exchange rate, 673, 681-85, 687. See also
Prebisch, Raul, 123, 758 Project appraisal; Shadow prices
Preconditions for development, 102-3 Shadow prices, 178, 638, 656-57, 666-70, 671-75.
Preferences, 328 See also Cost-benefit analysis; Price distortions;
Present value: of foreign investment, 331. See also Project appraisal; Shadow exchange rate: Shadow
Net present value; Project appraisal wage rate
Price distortions, 193-95, 207-8, 243-44, 471-83, Shadow wage rate, 178, 666-70, 673, 678, 704-5.
658-71, 757. See also Factor prices; Foreign See also Project appraisal
exchange; Shadow prices; Interest; Wages Singer, Hans, 758
Price incentives: in agriculture, 471-76; and food Smith, Adam, 174
prices, 476-83. See also Price distortions; Urban Social dualism, 170
bias Social indicators: statistics, 603-9
Producer power, 486. See also International Social marginal productivity, 422. See also Shadow
commodity agreements
Productivity. See Capital accumulation; Human Social prices, 688-91. See also Shadow prices
resources; Investment; Nutrition; Production priceswelfare function, 719, 727-28, 740
Social
functions; Social marginal productivity; Sociocultural factors, 754-55
Technology South Korea, 23, 48, 56-61,78
 768 INDEX

Special drawing rights (SDRs), 315, 317-18; and 547-52; in primary products, 553-60; staple
link, 319. See also International monetary system theory of, 491. See also Comparative advantage;
Srinivasan, T. N., 463-64, 751-52, 601 Development through trade; Exports; Import
Sri Lanka, 23, 49, 461 substitution; Terms of trade
Stabilization programs, 279. See also IMF Trade policy liberalization, 510-14, 538-42. See
Stages of growth, 90-100, 102, 1 1 1-12 also Protectionist arguments; Import substitution
State-owned enterprises, 752-53 Transnational enterprises. See Multinational
Stewart, Frances, 562 enterprises
Streeten, Paul P., 108-9, 357 Trickle-down effects, 112, 114, 748
Structural: approach, 118-22, 216, 268-73; change, Tropical Africa, 176-78
397-404, 739; transformation, 456-58 Two-gap
729 analysis, 121, 129-30, 381, 384, 386, 546,
Sub-Saharan Africa, 66-68, 467
Subsidies, 713, 733, 736
Subsistence sector. See Dual economy; Agriculture
Substitution: of factors, 729-30. See also Import UNCTAD, 2, 116, 315, 333, 486, 757, 759-60
substitution Underemployment, 196-97, 209. See also
Unemployment
Unemployment: disguised, 150-52, 155-61; and
Taiwan, 80-81, 432-37, 453, 459, 461 dual economy, 191-95; equivalent, measurement
Take-off, 91,94-99, 101 n. of, 196-97; Keynesian, 192; Marxian, 168;
Tariff factories, 392-93 structural, 167. See also Labor surplus
Tariffs, 389-94, 528-29, 654, 733, 736. See also Unequal exchange, 502-3
Import substitution; Protection UNIDO method, 641, 666, 681-85. See also Project
Taxation, 194, 235-37, 239, 245, 262-63, 713,
733 appraisalagricultural strategy, 424, 453-60
Unimodal
Technical progress, 204, 227. See also Technology Urban bias, 212, 437-40
Technological dualism, 167-70
Technology: appropriate, 212, 345-48; change in,
127, 465, 471; and factor proportions, 208-9, Vent for surplus, 489
212-14; foreign, 207-8; intermediate, 115; shelf, Viner, Jacob, 8-9, 430-31
349-51. See also Choice of techniques;
Production function; Technical progress
Tenancy, 464, 469. See also Peasants; Rural Wages, 170-75, 193-94, 213-14, 323-24, 390-91,
markets; Labor markets 653; and employment, 201; minimum, 193. See
Terms of trade, 236-37, 326, 495, 502-3, 555-58, also Dual economy; Factor prices; Migration;
563 Price distortions; Unemployment
Theil, H., 727 Waterston, Albert, 745-46
Third World: term, 293 Welfare, 7-8, 16-17
Timmer, Peter, 474 Welfare economics, 638
Tinbergen, Jan, 727, 745-46, 761 Westphal, Larry E., 533
Todaro, Michael, 151,213-14 Widening gap, 1
Trade creation, 562 World Bank, 1, 129, 136, 283-84, 327, 424, 637,
Trade diversion, 562 640
Trade, foreign: and agriculture, 466; benefits of, World prices, 447, 676-77. See also Little-Mirrlees
489-92; and inequality, 498-502; inter-LDC, method
Completely revised and updated, Leading Issues in Economic Development, Fourth
Edition, presents an ordered view of development economics — without sacrificing the
variety of viewpoints and perspectives. With a broad range of theoretical, applied, and
policy materials, it illuminates the critical issues that affect the development of poor
countries and introduces students to current trends in the field.

Most of the material is new to this edition, including readings that provide detailed
coverage of the concepts of basic needs and human resources, the problems of external
debt, and the issues of project appraisal. Throughout the text, students are introduced
to new topical selections that deepen their understanding of fundamental economic
principles and empirical relationships. Extensive editorial commentary emphasizes
the thematic relations between the articles, raises key questions, and suggests further
readings; and numerous charts, tables, and exhibits illustrate the text. The articles are
written by leading development economists and include selections from government
reports, international conferences, and research studies. Leading Issues in Economic
Development, Fourth Edition, provides students with rare insights into the crucial
issues affecting development economies.

Author, Editor, and Annotator

Gerald M. Meier is Professor of International Economics at Stanford University. His
books include International Economic Reform: Collected Papers of Emile Despres
(1973); his trilogy: Problems of Trade Policy (1973), Problems of a World Monetary
Order (1982), and Problems of Cooperation for Development (1974); International
Economics: The Theory of Policy (1980); and his forthcoming Emerging From
Poverty: The Economics That Really Matters, all published by Oxford University
Press.

Oxford University Press, New York

Cover design by Egon Lauterberg ISBN 0- 19-50341 5-5