State Power #823 ch2 7/2/02 4:00 PM Page 19

Chapter 2

The Economics of International Trade

Introduction

OBJECTIVES
■ Understand the three elements of microeconomic theory that serve
as the building blocks for trade theory: consumption indifference
curves, production possibilities frontiers, and optimized market
equilibrium.
■ Explore the Ricardian (classical), constant-costs model of compara-
tive advantage and mutually advantageous trade.
■ Explore the neoclassical, increasing-costs model of trade and the
contribution made by the Heckscher-Ohlin theorem to our under-
standing of the bases of comparative advantage.
■ Examine the Stolper-Samuelson theorem that trade, although bene-
ficial to a country as a whole, may create losses for particular groups
within that country, thus giving them a rational basis to prefer pro-
tection over trade.
■ Understand the new areas of economic research regarding the pos-
sible contributions of trade to national economic growth (trade and
endogenous growth) and changes in the composition of trade (intra-
industry vs. inter-industry trade).

19
State Power #823 ch2 7/2/02 4:00 PM Page 20

20 State Power and World Markets

Y
ou are living in Des Moines, Iowa. You wake up in a house built with
lumber from Canada. You wear clothes fabricated in India or Hon-
duras and shoes from Italy or Brazil. You prepare a breakfast that
includes orange juice from Brazil and cereal from Switzerland; you use sil-
verware made in China, South Korea, or Finland; and you sit at a breakfast
table manufactured in Thailand or Denmark. You drive to work in an auto-
mobile built in Japan or Germany or Sweden and use gasoline that was re-
fined from oil imported from Venezuela, Nigeria, or Saudi Arabia. At lunch
you have a salad consisting of tomatoes from Belgium, bell peppers from the
Netherlands, lettuce from Mexico, grapes from Chile, olives from Greece,
and cheese from France. You have to make a phone call during lunch, and to
do so you use a cell phone made in Finland or Sweden. After work you re-
turn home and watch the news on a television set made in Mexico; you
catch up on your e-mail using a laptop computer assembled in Taiwan of
components made in Singapore and China; and you listen to some music by
placing a German-made CD into a Japanese-made CD player. Before you
retire for the evening, you turn off the lights in your house, which were made
in Mexico. You reflect, as you drift off to sleep, on your upcoming visits to
Prague and Budapest, with a brief layover in London.1
International trade affects what we eat, what we wear, what we watch and
listen to, how we move about, where we go, and how we earn a living. But
how exactly is it possible for consumers in Iowa to obtain tomatoes from Bel-
gium? How do people in Finland know that people in Des Moines wish to
buy cell phones? How, in other words, does trade come about? What deter-
mines who sells what, and who buys what? And is all this trade a good idea?
We will see in Chapters 4 and 5 that politics determines the answers to
these questions to a remarkably large degree. However, the discipline of eco-
nomics during the past two centuries also has developed a powerful under-
standing of the sources, mechanics, and effects of international trade. By
understanding international trade theory we can identify some of the most
important and interesting political issues relating to the world political econ-
omy.
Hence, in this chapter we present the main elements of international
trade theory. We begin our review with a brief consideration of the building
blocks for trade theory that are taken from microeconomics. Using these an-
alytical building blocks, we introduce the two basic models of trade, the Ri-
cardian or classical model, and the more contemporary neoclassical model,
emphasizing both what they have in common and where they diverge. Both
State Power #823 ch2 7/2/02 4:00 PM Page 21

State Power and World Markets 21

models, we shall see, reach the same fundamentally important result: trade
improves the overall welfare of nations by allowing them to make the best
use of their scarce productive resources, and to improve their overall con-
sumption by producing certain things themselves and obtaining other goods
and services from other nations. In light of the tremendous gains that trade
holds for nations, economists are skeptical of most arguments that are made
against international commerce. However, although trade may benefit a na-
tion as a whole, groups within a nation may in some circumstances lose from
freer trade and therefore may have a rational reason to resist more open
trade. Finally, we shall explore two frontiers of research in international trade
theory, one relating trade to national economic growth, and another focusing
on the tendency of many countries to trade similar rather than dissimilar
goods. Both of these new lines of inquiry, we shall see, have led economists
to rethink some elements of their theory of trade, and both point to impor-
tant new developments in the world trading system.

Analytical Building Blocks

Economists rely on three analytical tools in their exploration of the bases for
and benefits of trade: consumption indifference curves, production possibili-
ties frontiers, and an analysis of optimized production-consumption equilib-
rium in the absence of trade.

Consumption Indifference Curves

Economists assume that individuals derive happiness, or utility, from their

consumption of goods. Different combinations of goods, or what economists
term “market baskets,” may each provide an individual with the same level of
utility. For example, an individual may be exactly as happy consuming a mar-
ket basket containing five pairs of shoes and two computers as he or she
would be consuming a market basket containing six pairs of shoes and one
computer. As a result, if presented with a choice between two such baskets,
the individual would be indifferent as to which of these two market baskets is
preferable, for each would provide the same level of utility. A graphical rep-
resentation of the different combinations of shoes and computers that would
State Power #823 ch2 7/2/02 4:00 PM Page 22

22 State Power and World Markets

provide the individual with a constant level of satisfaction would be that

person’s consumption indifference curve with respect to shoes and com-
puters.
Economists have employed the concept of an individual’s consumption
indifference curves in their analyses of trade between countries. They as-
sume that it is possible to aggregate the satisfaction levels that all residents
of one country attain from the consumption of different baskets of goods,
and therefore that it is possible to create national consumption indifference
curves. In this book, when we refer to consumption indifference curves, we
are referring to the consumption indifference curves for countries taken as a
whole.
Examples of hypothetical consumption indifference curves for the United
States are presented in Essential Economics 2.1(a). Let us begin with indif-
ference curve U0. The different combinations of computers and shoes along
that curve—for example, those represented by points A, B, C, and D—pro-
vide the country with exactly the same level of aggregate satisfaction.
U0 is convex with respect to the origin, not straight or concave. The prem-
ise underlying this characteristic of the indifference curve is that of declin-
ing marginal utility from consumption of additional increments of any
given good. That is, as an individual consumes more and more increments of
one good, each increment delivers less and less satisfaction than the preced-
ing increment, and therefore the individual becomes less and less willing to
forgo other goods to obtain that next increment. By consequence, if the
United States is at point A on U0 and is asked how many pairs of shoes it is
willing to forgo to obtain 2 million more computers, all the while retaining
the level of satisfaction it enjoys at point A, its response would be 40 million
pairs (this new market basket is represented by point B on U0). However, if
the United States is at point C on U0, in response to the same question the
answer would be not 40 million pairs, but only 10 million pairs (shown by
point D on U0).
Economists therefore suggest that a curve like U0 reveals a declining
marginal rate of substitution of computers for shoes on the part of the
United States. The marginal rate of substitution at any point along an indif-
ference curve is the absolute value of the slope of the indifference curve at
that point. Because the slope of U0 declines as we move along the curve
from left to right, so does the marginal rate of substitution decline.
One other key observation can be derived from Essential Economics
2.1(a). What happens if the United States is no longer presented with the
combinations of shoes and computers depicted by U0, but instead can
State Power #823 ch2 7/2/02 4:00 PM Page 23

State Power and World Markets 23

¬ ESSENTIAL ECONOMICS 2.1

Three Building Blocks for Economic Analysis

(a) (b)
300 Consumption Indifference Curves 300 Production Possibilities Frontiers
280 280
260 260
240 240
Shoes (millions of pairs)

Shoes (millions of pairs)

220 E 220
A
200 200 P C Increasing
180 180 opportunity
160 B 160 costs
140 F 140 A
120 120
100 U1 100 D
80 C 80
60 D 60 Constant
40 40 opportunity
U0 costs B
20 20 Q
0 0
2 6 10 14 18 22 26 30 2 6 10 14 18 22 26 30
Computers (millions) Computers (millions)

(c)
300 Production-Consumption Equilibrium
280 in Autarky
260
240
Shoes (millions of pairs)

220
200 Ea
180
160
140 Ec Ed
120
100
80
Eb U2
60
40 U1
20 U0
0
2 6 10 14 18 22 26 30
Computers (millions)

Consumption indifference curves (panel a) graphically represent the different combinations of two
goods that would provide a nation with a constant level of satisfaction. Production possibilities fron-
tiers (panel b) show the trade-offs that countries make in the production of two different goods. The
optimization of both consumption and production in autarky is shown by the point of tangency be-
tween the production possibilities frontier and the highest possible consumption indifference curve
(panel c).

Core Principle
Consumption indifference curves, production possibilities frontiers, and
production-consumption equilibrium in autarky all show a country’s production
and consumption options in the absence of international trade.
State Power #823 ch2 7/2/02 4:00 PM Page 24

24 State Power and World Markets

choose among a new set of market baskets, each of which contains more
computers, more pairs of shoes, or more of both? Such market baskets are
represented by the curve labeled U1. Since points on U1, such as E and F,
represent combinations of goods that mark an increase, as compared with
points on U0, in American consumption of shoes or computers or both, the
United States must be enjoying a higher level of satisfaction along each point
of U1 than it is at any point along U0.
Hence, and this concept will be crucially important in our later discussion
of the effects of trade, anything that permits the United States to increase its
consumption of at least one good must be improving overall American satis-
faction and must therefore be producing a movement by the United States to
a higher indifference curve. Trade, as we will see later in this chapter, is pre-
cisely such a mechanism; it provides a way for countries to improve their
overall levels of consumption and thereby to move to higher indifference
curves, signifying higher levels of overall societal satisfaction.

Production Possibilities Frontiers

The second analytical device we need to acquire for our discussion of trade
theory is the concept of a production possibilities frontier. A production
possibilities frontier represents the different combinations of goods that a
country can produce during some period of time (in our illustrative discus-
sion below, one year), given the full exploitation of the productive resources
available in the country during that period of time. Two basic types of pro-
duction possibilities frontier are presented in Essential Economics 2.1(b):
the first is built on the premise of constant opportunity costs, and the second
is built on the premise of increasing opportunity costs.

Production Possibilities Frontiers with Constant Opportunity Costs

Let us begin with the production possibilities frontier in Essential Econom-

ics 2.1(b) closest to the origin. Notice that this frontier has a negative slope,
and that it is a straight line. The first characteristic of the frontier—that its
slope is negative—indicates that opportunity costs are present in produc-
tion. As can be observed of the inner frontier in Essential Economics 2.1(b),
if the United States puts all its productive resources into shoe manufactur-
ing, it can reach point P on that frontier and produce 200 million pairs of
shoes in one year. If it decides to produce not just shoes but also computers,
State Power #823 ch2 7/2/02 4:00 PM Page 25

State Power and World Markets 25

it would need to take some of its resources out of shoe manufacturing and
reallocate them to the production of computers. For example, if the United
States elects to build 4 million computers, that is, to go from point P to point
A on the inner frontier, it needs to reduce its production of shoes by 40 mil-
lion pairs. In this example, then, the cost of building one computer in the
United States is the opportunity to produce ten pairs of shoes. In the lan-
guage of economics, the opportunity cost of producing one computer is the
production forgone of ten pairs of shoes. This notion of opportunity costs is
absolutely central to our upcoming discussion of the bases for and dynamics
of trade.
As we have noted, the inner production possibilities frontier in Essential
Economics 2.1(b) is drawn as a straight line. This second characteristic of
the frontier indicates that the opportunity cost of producing computers in
terms of shoes forgone remains constant (as we have drawn the frontier, con-
stant at ten pairs of shoes forgone for each computer built) no matter how
many additional increments of computers the United States chooses to
build. By virtue of the assumption of constant opportunity costs, if the
United States wants to go from point B to point Q on the frontier, that is,
from 16 million to 20 million computers produced, then it needs to forgo the
production of 40 million pairs of shoes, the same amount it must forgo when
moving from point A to point B on the frontier.

Production Possibilities Frontiers with Increasing Opportunity Costs

We will be using a production possibilities frontier that is based on constant

opportunity costs later in this chapter when we discuss the first major eco-
nomic theory of trade. However, in more recent years economists have built
their models of trade, which we will also discuss later, on the basis of the as-
sumption that production possibilities frontiers should represent increasing
opportunity costs. In other words, they assume that successive, equal in-
creases in the production of one good will require forgoing the production of
larger and larger increments of the other good. The rationale for this switch
to an assumption of increasing opportunity costs is that it is usually more
reasonable to assume that, although the first redeployment of an increment
of productive resources from one use to another might produce very large re-
sults, eventually, as more and more increments of resources are so rede-
ployed, decreasing returns to scale for the resource increments are likely to
come into effect.
To see how decreasing returns might arise, and how decreasing returns
State Power #823 ch2 7/2/02 4:00 PM Page 26

26 State Power and World Markets

create increasing opportunity costs, let us assume the following about a

country:

■ It has land on which it can grow either grapes or wheat;

■ Some of the land is better suited for grapes than for wheat, whereas
some of the land is better suited for wheat than for grapes; and
■ At first, the country is using all available land to produce grapes.

If this country were to decide to grow some wheat, it would need to select
plots of land to take out of grape production, and then use those plots to
plant wheat. In order to maximize the production of wheat and to minimize
the reduction in the production of grapes, the country would select the plots
of land that are the least suited for grapes and the most suited for wheat. In
doing so, it would find at first that forgoing a relatively small amount of
grapes results in the production of large amounts of wheat. But what if the
country wanted to produce more and more wheat? Eventually, it would have
to use plots of land that were progressively less suited for wheat and more
suited for grapes. As a result, the amount of additional wheat harvested per
plot of additional land so allocated would decline. In the face of such de-
creasing returns in wheat production with respect to land, an effort by the
country to produce successive equal increments of wheat would require for-
going the production of ever larger volumes of grapes—that is, there would
be increasing opportunity costs for wheat in terms of grapes forgone.
The outer frontier in Essential Economics 2.1 (b) is an example of a pro-
duction possibilities frontier for computers and shoes that is characterized by
such increasing opportunity costs. Note that movement from point P to
point C on the outer frontier, that is, increasing U.S. computer production
from 0 to 4 million computers, requires a reduction in U.S. shoe production
by 10 million pairs. However, because we are assuming decreasing returns as
America shifts resources from shoes to computers, if we instead start at
point D and increase computer production by the same amount, 4 million
computers, to reach point Q, we must accept a vastly greater drop in shoe
production, about 100 million pairs.
More generally, and in contrast to the inner production possibilities fron-
tier in Essential Economics 2.1(b), the outer frontier reflects increasing
opportunity costs for computers in terms of shoes forgone, and what econo-
mists term an increasing marginal rate of product transformation be-
tween computers and shoes. The marginal rate of product transformation at
any point along a production possibilities frontier is the absolute value of the
State Power #823 ch2 7/2/02 4:00 PM Page 27

State Power and World Markets 27

slope of the curve at that point. Insofar as the outer frontier in Essential Eco-
nomics 2.1(b) depicts a requirement that more and more shoes must be for-
gone in order to build each additional fixed increment of computers, the
slope of the outer curve is increasing, indicating an increasing marginal rate
of product transformation between shoes and computers.

Optimized Production and Consumption in Autarky

Essential Economics 2.1(c), in which consumption indifference curves and

production possibilities frontiers are superimposed on one another, provides
a final key analytical building block for our understanding of the sources and
benefits of trade. This panel presents economists’ conception of the manner
in which a country achieves its highest level of satisfaction in the context of
autarky, that is, in the absence of trade. This analysis of optimization in au-
tarky, we will see in the following discussion, allows us to appreciate how a
country can enjoy greater satisfaction if it moves from autarky to participa-
tion in the international trading system, and therefore why countries choose
to participate in world trade.
We begin with another hypothetical production possibilities frontier,
which we assume is characterized by increasing opportunity costs (i.e., its
slope is increasing as we move from left to right). As with the curves in Es-
sentialized Economics 2.1(b), this frontier depicts the different combina-
tions of computers and shoes that the United States can produce in one year
with its current resources. Because we are assuming that absolutely no trade
is taking place, the production possibilities frontier also sets the limits on
what the United States can consume: Americans are able to consume only
the combinations of the two goods that are on or inside the frontier. Assum-
ing that all goods that are produced are consumed, the effective range of
choice for the United States in terms of optimum consumption is the locus
of points that constitutes the production possibilities frontier.
In these circumstances, the United States produces the combination of
shoes and computers that provides the maximum level of satisfaction possi-
ble.
Consider the country’s reasoning if it finds itself at point Ea. How much
would it have to give up in terms of shoe production forgone in order to con-
sume 10 million rather than 4 million computers—that is, what is the drop
in shoe production that is associated with a move from Ea to Ec? The answer,
if we move along the production possibilities frontier, is about 30 million
State Power #823 ch2 7/2/02 4:00 PM Page 28

28 State Power and World Markets

pairs of shoes. Yet how many pairs of shoes would the United States be
willing to forgo to obtain those extra 6 million computers while being no
worse off in terms of its level of satisfaction? The answer, as we move down
along the consumption indifference curve U0, is about 90 million pairs. By
shifting resources from shoes to computers and moving from Ea to Ec, the
United States could have all the satisfaction it enjoys at Ea, plus the satisfac-
tion of the 60 million pairs of shoes that it would be willing to forgo, but
would not have to forgo, in order to obtain the 6 million extra computers. Re-
calling our discussion about consumption indifference curves in Essential
Economics 2.1(a), it is clear that, by moving from Ea to Ec, the United States
is enjoying a higher level of satisfaction and therefore must have shifted to a
higher indifference curve—specifically, curve U1.
Clearly, then, the combination represented by point Ea, at which the
United States is producing and consuming 4 million computers and 190 mil-
lion pairs of shoes, is not an optimum outcome. At point Ea, the United
States would be willing to forgo consuming many more shoes in exchange for
computers than it would actually have to forgo producing in order to make
those computers. More theoretically, at Ea, the country’s marginal rate of
substitution exceeds its marginal rate of product transformation; equiva-
lently, at Ea, the slope of the country’s consumption indifference curve is
greater than the slope of its production possibilities frontier. Likewise, if the
United States found itself at point Eb, it would find it in its interest to shift
resources out of computers and into shoes. At point Eb, the slope of the pro-
duction possibilities frontier is greater than the slope of the consumption in-
difference curve, which means its marginal rate of product transformation
exceeds its marginal rate of substitution.
Where, then, does the country maximize its happiness in light of its pro-
duction possibilities frontier? Maximum satisfaction is found at point Ed,
which is associated with the indifference curve U2. At that point, the mar-
ginal rate of substitution of computers for shoes is exactly equal to the mar-
ginal rate of product transformation between the two goods. That is, at point
Ed, given its preferences for the two goods and its ability to make them, the
United States cannot improve its satisfaction further by changing its produc-
tion or consumption choices. This situation can change, however, if we in-
troduce the opportunity for the United States to engage in international
trade.
State Power #823 ch2 7/2/02 4:00 PM Page 29

State Power and World Markets 29

Why Do Nations Engage in Trade?

The Ricardian Model: Two Countries, Two Goods,

and Constant Opportunity Costs

Using these basic building blocks, economists have demonstrated that coun-
tries can gain from trade and thereby have an incentive to engage in such
transactions across their borders.
To make this issue concrete, consider the following questions:

■ From 1996 to 2000, the United States bought about $400 million
per year in coffee and coffee products from Brazil. Coffee, however,
was not the largest product category of Brazilian exports to the
United States during this period; the highest it ranked was second,
in 1997. On average, the biggest category of Brazilian export prod-
ucts going to the United States during 1996–2000 was footwear,
which averaged $1.1 billion per year during that period.2 Now, it
might be obvious why the United States buys coffee from Brazil: it
cannot be grown readily in the U.S. climate and it can be in Brazil.
But how, in light of the fact that the United States is an industrial
powerhouse, could the United States possibly benefit from import-
ing from Brazil something as simple to make as shoes?
■ The biggest single category of U.S. exports to Brazil between 1996
and 2000 was telecommunications equipment: it averaged about
$1.2 billion per year during this period. Yet the biggest overall area
of U.S. exports to Brazil during the late 1990s, at an average level of
$1.8 billion per year, was information technology—that is, comput-
ers, computer parts, and parts for other office equipment. What is
remarkable about these U.S. information-technology hardware ex-
ports to Brazil during the late 1990s is that it followed a 20-year
period during which Brazil had been seeking and had made some
progress in nurturing a domestic Brazilian computer industry.3
Given this apparent Brazilian national interest in promoting an in-
digenous computer industry, why didn’t Brazil simply prohibit com-
puter imports and thereby create a market for local computer
producers? How, in other words, does Brazil gain from buying U.S.
computers rather than building them at home?
State Power #823 ch2 7/2/02 4:00 PM Page 30

30 State Power and World Markets

➣ TIMELINE 2.1

The Development of International Trade Theory

1776 Adam Smith argues his free-market critique of mercantilism

1817 David Ricardo develops Ricardian (classical) model

1879 Alfred Marshall proposes offer curves and trade

1919
1930
Eli Heckscher
Bertil Ohlin } develop what comes to be known as the
Heckscher-Ohlin theorem

}
1930 Gottfried Haberler explores increasing
opportunity costs and trade
1932 A. P. Lerner explores increasing opportunity
The foundations of the
costs and trade
neoclassical model
1933 Wassily Leontief explores increasing opportunity
costs and trade
1939 Paul Samuelson advocates the gains from trade

1941 Wolfgang Stolper and Paul Samuelson propound the Stolper-Samuelson

theorem

1975 H. G. Grubel and P. L. Lloyd examine intra-industry trade

1981 Paul Krugman and Elhanan Helpman theorize on increasing economies of scale
and trade

1986 Paul Romer develops endogenous growth theory

To understand why the United States and Brazil exchange shoes for com-
puters today, we can usefully employ the logic of comparative advantage.
The benefits of freer trade were highlighted by Adam Smith in his founda-
tional 1776 work, The Wealth of Nations, and the logic of comparative ad-
vantage as the underpinning for this view was presented by David Ricardo in
1817.4
Let us, for the following discussion, stipulate these assumptions:

■ There are only two countries, Brazil and the United States;
■ Brazil and the United States produce only two goods, computers
and shoes;
■ There are no transportation costs for goods shipped between Brazil
and the United States;
State Power #823 ch2 7/2/02 4:00 PM Page 31

State Power and World Markets 31

¬ ESSENTIAL ECONOMICS 2.2

Hypothetical Output Levels, United States and Brazil

OUTPUT PER WORKER PER YEAR OPPORTUNITY COST

Computers Pairs of shoes 1 computer 1 pair of shoes

United States 50 200 4 pairs shoes .25 computer

Brazil 5 175 35 pairs shoes .03 computer

Core Principle
An absolute advantage in manufacturing does not necessarily imply a compara-
tive advantage.

■ Only one input, labor, is required for the production of either com-
puters or shoes;
■ There are constant opportunity costs between the two goods in each
country, or, in graphical terms, the production possibilities frontiers
for both countries for computers and shoes, while possessing differ-
ent slopes, are straight lines;
■ The United States and Brazil each have one million labor-years of
total labor supply; and
■ Any one worker in the United States and Brazil is able to produce in
one year the number of computers or pairs of shoes depicted in the
first two columns of Essential Economics 2.2.

Any one worker in the United States, in this scenario, can produce more
computers than can any one worker in Brazil (50 in America as opposed to 5
in Brazil), and any American worker can also produce more pairs of shoes
than can a Brazilian worker (200 as opposed to 175 pairs). The United
States, in other words, has an absolute advantage over Brazil in both comput-
ers and shoes. In these circumstances, it might appear to be highly unlikely
that the United States could gain anything from trade with Brazil.
The key to appreciating the potential basis for mutually profitable trade
between these two countries lies in the differences in opportunity costs each
faces with respect to shoes and computers. As we discussed earlier, if the
United States wants to produce more shoes, it can do so only by moving
workers out of computer manufacturing. Over the course of one year, for
State Power #823 ch2 7/2/02 4:00 PM Page 32

32 State Power and World Markets

each worker it shifts from computers to shoes, 50 fewer computers are built
and 200 pairs of shoes are produced. For each additional pair of shoes pro-
duced, then, the United States must forgo the production of 1/4 of a com-
puter (50 computers whose production is forgone ÷ 200 pairs of shoes
thereby produced = 1/4 computer forgone per additional pair of shoes pro-
duced). Hence, the opportunity cost of each additional pair of shoes pro-
duced in the United States is the production forgone of 1/4 of a computer. If
the United States prefers to have more computers and thus shifts workers
from shoe making to computer manufacturing, then, for each worker so
shifted, computer production goes up by 50 units over the year while shoe
production goes down by 200 pairs. Therefore, the opportunity cost of 1 ad-
ditional computer is the production forgone of 4 pairs of shoes (200 pairs of
shoes whose production is forgone ÷ 50 additional computers thereby pro-
duced = 4 pairs of shoes forgone per additional computer produced).
By the same token, in Brazil, each worker shifted from computers to
shoes causes production of the former to go down by 5 computers while al-
lowing production of the latter to go up by 175 pairs; the opportunity cost of
1 additional pair of shoes made in Brazil is the production forgone of about
.03 computer. Each Brazilian worker shifted from shoe production to com-
puter manufacturing causes the former to go down by 175 pairs while allow-
ing the latter to go up by 5 computers; the opportunity cost of 1 additional
computer produced in Brazil is the production forgone of 35 pairs of shoes.
These opportunity costs for the United States and Brazil are shown in the
third and fourth columns of Essential Economics 2.2.
Now comes the critical point. If we ask where it is relatively cheaper to
build additional computers in terms of pairs of shoes forgone, we see in the
third column of Essential Economics 2.2 that the answer is in the United
States, where only 4 pairs of shoes must be forgone to build each additional
computer, as opposed to the 35 pairs that must be forgone in Brazil. The
United States, then, has a comparative advantage over Brazil in the making
of computers. By the same token, if we ask where it is relatively cheaper to
manufacture additional pairs of shoes in terms of computers forgone, we see
in the fourth column of Essential Economics 2.2 that the answer is in Brazil,
for there only .03 computers must be forgone to produce each additional
pair of shoes, whereas in the United States, .25 computers must be forgone.
Brazil, therefore, has a comparative advantage over the United States in the
production of shoes.
We can now employ the analytical tools developed in the preceding sec-
tion to appreciate how the logic of comparative advantage produces opportu-
State Power #823 ch2 7/2/02 4:00 PM Page 33

State Power and World Markets 33

¬ E SS E N T I A L ECO N O M I C S 2 .3

The Ricardian Model of Trade

(a) (b)
The United States Brazil
240 240
220 220
200 200
Shoes (millions of pairs)

Shoes (millions of pairs)

180 180 PtB
160 160
140 140
120 120
CtA
100 100
EaA CtB
80 80 EaB
U1A
60 60
40 U0A 40
U1B
20 20 U0B
PtA
0 2 6 10 14 18 22 26 30 34 38 42 46 50 0 2 6 10 14 18 22 26 30 34 38 42 46 50
Computers (millions) Computers (millions)

Core Principle
Trade between two countries allows each of them to specialize and increases
each country’s level of overall satisfaction.

nities for mutually beneficial trade between countries. Consider the two
graphs in Essential Economics 2.3: the solid straight lines in panels (a) and
(b) represent the production possibilities frontiers for the United States and
Brazil, respectively. At the moment we are assuming constant opportunity
costs between computers and shoes in each country, so the production pos-
sibilities frontier for each country is drawn as a straight line. Furthermore,
the slopes of the two countries’ production possibilities frontiers differ from
one another, reflecting the different opportunity costs each country experi-
ences in production of the two goods.
We begin with the situations in the United States and Brazil under the as-
sumption of autarky. In autarky, as we discussed earlier, the only consump-
tion opportunities available to each country are those defined by their
respective production possibilities frontiers. Let us assume that, in autarky,
demand conditions in Brazil and the United States are such that each allo-
cates 600,000 labor years to computers and 400,000 labor years to shoes. In
these circumstances, market equilibrium in the United States is at point EaA
in Essential Economics 2.3(a), at which the United States is producing and
State Power #823 ch2 7/2/02 4:00 PM Page 34

34 State Power and World Markets

consuming 30 million computers and 80 million pairs of shoes and is enjoy-

ing a level of satisfaction that is represented by consumption indifference
curve U0A. Brazil’s market equilibrium in autarky is at point EaB in Essential
Economics 2.3(b), at which it is producing and consuming 3 million com-
puters and 70 million pairs of shoes and is enjoying a level of satisfaction
that is represented by consumption indifference curve U0B.
What happens now if trade becomes possible between the United States
and Brazil? If trade becomes possible, then the United States specializes on
the basis of its comparative advantage: it stops making shoes and increases
its production of computers from 30 million to 50 million units. This is
noted in Essential Economics 2.3(a) by the movement downward along the
U.S. production possibilities frontier from EaA to PtA. Brazil also specializes
on the basis of its comparative advantage, ceasing the production of comput-
ers and, using the labor so released, increasing its shoe production from
70 million pairs to 175 million pairs. This is captured in Essential Econom-
ics 2.3(b) by the movement of Brazil upward along its production possibili-
ties frontier from EaB to PtB.
Now that the United States is producing many more computers and
Brazil is producing many more pairs of shoes, each might seek to obtain the
good it no longer produces locally by offering to enter into an exchange. The
key requirement is that both countries agree on the terms of trade—that is,
the rate at which U.S. computers will be exchanged for Brazilian shoes.
Given the domestic opportunity costs between computers and shoes in the
two countries, we know that certain offers will not be accepted:

■ If Brazil offers fewer than 4 pairs of Brazilian shoes for each U.S.
computer received, then the United States will decline to trade: it
would be cheaper for the United States to make its own shoes again
by reallocating labor from computers back to shoes, and thereby ob-
tain 4 pairs of American-made shoes for each computer forgone.
■ If the United States indicates that it will provide a computer to
Brazil only if it received more than 35 pairs of shoes in return,
then Brazil will decline the chance to trade: it would be cheaper for
Brazil to acquire computers by shifting labor out of shoes and back
into computers, since it can obtain each locally produced computer
by forgoing only 35 pairs of shoes.

Thus, for each partner voluntarily to accept the opportunity to specialize and
to trade, the terms of trade between the two countries need to fall some-
State Power #823 ch2 7/2/02 4:00 PM Page 35

State Power and World Markets 35

 PRIMARY DOCUMENT 2.1

David Ricardo and International Trade

Under a system of perfectly free commerce, each country naturally devotes its
capital and labour to such employments as are most beneficial to each. This pur-
suit of individual advantage is admirably connected with the universal good of the
whole. By stimulating industry, by rewarding ingenuity, and by using most effi-
caciously the peculiar powers bestowed by nature, it distributes labour most
effectively and most economically: while, by increasing the general mass of pro-
ductions, it diffuses general benefit, and binds together by one common tie of in-
terest and intercourse, the universal society of nations throughout the civilized
world. It is this principle which determines that wine shall be made in France and
Portugal, that corn shall be grown in America and Poland, and that hardware and
other goods shall be manufactured in England.

Source: David Ricardo, On the Principles of Political Economy and Taxation, Chapter 7, Intelex Past
Masters series, available at pastmasters2000.nlx.com/display.cfm?&clientID=1332338&depth=2&
infobase=pmbritphil.nfo&softpage=GetClient42&view=browse.

Core Principle
David Ricardo believed that trade, when combined with specialization on the
basis of comparative advantage, can provide benefits to all participants, and
that it does so through the allocation of labor to its most productive uses and
thus through the production and consumption of a greater amount of goods
globally than would be possible in the absence of trade. Ricardo also suggests in
this passage that freer trade yields other benefits, including the fostering of
technological innovations and the forging of common interests among nations.

where between 1 U.S. computer for 4 pairs of Brazilian shoes and 1 U.S.
computer for 35 pairs of Brazilian shoes. For the purposes of analysis, let us
assume that the demand for computers and shoes in both the United States
and Brazil is such that they negotiate terms of trade setting 1 computer
equivalent to 10 pairs of shoes (this is noted by the dashed lines in Essential
Economics 2.3(a) and 2.3(b)), and they agree to exchange 10 million U.S.-
made computers for 100 million Brazilian-made pairs of shoes.
Americans’ consumption therefore grows from 30 million computers in
autarky to 40 million computers with trade, and from 80 million pairs of
State Power #823 ch2 7/2/02 4:00 PM Page 36

36 State Power and World Markets

shoes without trade to 100 million pairs of shoes with it. Graphically, we see
in Essential Economics 2.3(a) that U.S. consumption moves off the U.S.
production possibilities frontier: specifically, it moves from EaA to CtA. Given
that the United States is consuming more of both computers and shoes after
it has specialized and engaged in trade than it had in autarky, it must be en-
joying greater satisfaction than it had during autarky, and this is represented
by its attainment of a higher consumption indifference curve, U1A rather
than U0A. Similarly, Brazil’s consumption of computers grows from 3 million
to 10 million, and its consumption of shoes rises from 70 million to 75 mil-
lion pairs. Graphically, we see in Essential Economics 2.3(b) that Brazil’s
consumption point also moves off its production possibilities frontier, going
from EaB to CtB. Given that Brazil with trade is now consuming more of both
computers and shoes, it must be enjoying greater satisfaction than it had in
autarky, and this is represented in Essential Economics 2.3(b) by Brazil’s
movement from consumption indifference curve U0B to U1B. Thus, the com-
bination of specialization and trade allows both countries to improve their
welfare.
David Ricardo’s model of comparative advantage and trade is of profound
historical and contemporary importance, and the power and persuasiveness
of his thinking can be appreciated even in the very brief extract from his
presentation of that model that is presented in Primary Document 2.1. The
model’s brilliance and endurance over time have resulted from its capacity to
demonstrate that countries that choose specialization and trade can escape
the seemingly tyrannical limits imposed on their consumption levels by their
individual production possibilities frontiers. In other words, countries can
use specialization and trade to achieve greater consumption and thus greater
satisfaction than was possible in autarky.
Yet Ricardo’s model, as powerful as it is, leaves two key problems unre-
solved:

■ If specialization on the basis of comparative advantage is the path to

enhanced national welfare, why do we not observe countries fully
specializing in the good(s) in which they have a comparative advan-
tage?
■ What causes one country to have a comparative advantage in one
good, and another country to have a comparative advantage in an-
other good? Why, for example, does the United States have a com-
parative advantage in computers and Brazil have a comparative
advantage in shoes?
State Power #823 ch2 7/2/02 4:00 PM Page 37

State Power and World Markets 37

The Neoclassical Model of Trade: Two Countries, Two Goods,

and Increasing Opportunity Costs

In addressing the first of these unanswered questions, we may recall that, in

our discussion so far of the Ricardian model, we have assumed that the
United States, as it shifts labor from shoes to computers, can produce each
additional computer at the opportunity cost of 4 pairs of shoes, and that this
opportunity cost remains the same no matter how many more computers the
United States produces. Neoclassical economic theory (in contrast to the
“classical” approach articulated by such economists as Ricardo) argues per-
suasively that constant opportunity costs between goods are not likely to
hold as a country progressively specializes in one or another good. Instead, as
we noted earlier in this chapter, such specialization is likely to encounter in-
creasing opportunity costs. To make this case, the neoclassical approach be-
gins with two key adjustments to the Ricardian model’s assumptions about
the production of either computers or shoes:

■ Rather than being produced with only one factor of production,

shoes and computers are produced with at least two factors of pro-
duction: in the discussion that follows, we will assume that these
two factors are labor and capital (machinery and buildings).5
■ Technical knowledge allows both computers and shoes to be made
with different amounts of labor and capital. However, compared to
shoes, the most efficient ways of making computers all use rela-
tively more capital than labor, whereas, compared to computers, the
most efficient ways of making shoes all use relatively more labor
than capital.

With these modified assumptions, the neoclassical model would expect

that, although at first the United States may be able to build one more com-
puter by shifting the labor and capital previously used to make four pairs of
shoes, eventually it would need to forgo progressively larger numbers of
shoes to make each additional computer. This is because, in comparison to
the optimal mixture of capital and labor to make a computer, the reduction
in shoe production releases too much labor and too little capital. The com-
puter industry would respond by turning to relatively more labor-intensive
manufacturing processes, but, given that computers are optimally made us-
ing capital-intensive techniques, the productivity of the computer industry as
a whole would at some point begin to decline. As a result, to keep building
State Power #823 ch2 7/2/02 4:00 PM Page 38

38 State Power and World Markets

additional increments of a given number of computers, shoe production

would have to drop by progressively larger and larger amounts. Returning to
our example, rather than being able to produce one more computer by shift-
ing the labor and capital that had been used to make four pairs of shoes, it
would become necessary at some point to shift the labor that otherwise
would yield five pairs of shoes, and then seven pairs, and so on.
The result, according to neoclassical economic theory, is that the opportu-
nity cost of computers in terms of forgone pairs of shoes increases as the
United States makes more and more computers and fewer and fewer shoes.
By the same logic, the opportunity cost of pairs of shoes in terms of forgone
computers increases as Brazil makes more and more shoes and fewer and
fewer computers. Eventually, as the United States and Brazil undertake spe-
cialization and trade, additional U.S. computer production becomes so ex-
pensive in terms of forgone shoes, and extra Brazilian shoe production
becomes so costly in terms of forgone computers, that the opportunity costs
between computers and shoes become equal across the two nations. At that
point, which in all likelihood would occur prior to full specialization, neither
of the two countries would have an incentive to specialize further. Hence, in
the neoclassical model of increasing opportunity costs, the dual process of
specialization on the basis of comparative advantage plus trade is mutually
beneficial to the nations involved, but it is not likely to lead to full special-
ization by either partner in the good in which it has a comparative advantage.
We can observe the implications of increasing opportunity costs for trade
in Essential Economics 2.4. As can be observed in the two graphs in the fig-
ure, increasing opportunity costs characterize the production possibilities
frontiers for both the United States and Brazil: both are concave with respect
to the origin. In autarky, as shown in Essential Economics 2.4(a), the United
States is at the production-consumption equilibrium point EaA, at which it
produces and consumes 14 million computers and 150 million pairs of shoes
and enjoys a satisfaction level represented by the consumption indifference
curve U0A. Brazil, as we can see in Essential Economics 2.4(b), finds itself in
autarky at EaB, at which it produces and consumes 12 million computers and
70 million pairs of shoes and enjoys a level of satisfaction that is associated
with the consumption indifference curve U0B.
With the possibility of trade, the United States begins to shift productive
resources out of shoes and into computers, the good for which it has a com-
parative advantage, while Brazil begins to shift productive resources out of
computers and into shoes, the good for which it is has a comparative advan-
tage. However, as we observed in connection to the outer frontier in Essen-
State Power #823 ch2 7/2/02 4:00 PM Page 39

State Power and World Markets 39

¬ ESSENTIAL ECONOMICS 2.4

The Neoclassical Model of Trade

(a) (b)
The United States Brazil

220 220
200 200
Shoes (millions of pairs)

Shoes (millions of pairs)

180 180
CtA
160 160
PtB
140 EaA U1A 140
120 U0A 120
100 PtA 100
CtB
80 80
60 60 EaB
40 40 U1B
20 20
U0B
0 0
2 6 10 14 18 22 26 30 34 38 42 46 50 2 6 10 14 18 22 26 30 34 38 42 46 50
Computers (millions) Computers (millions)

Core Principle
Increasing opportunity costs for trade cause specialization to end where oppor-
tunity costs and market clearance converge. Even partial specialization, when
combined with trade, produces a consumption outcome for both countries that
is superior to what is optimally possible in autarky.

tial Economics 2.1(b), we see in the panels of Essential Economics 2.4 that,
as the United States produces more and more computers, its opportunity
cost for one more computer in terms of pairs of shoes forgone increases,
and as Brazil produces more and more shoes, its opportunity cost for one
more pair of shoes in terms of computers forgone also rises. As a result, op-
portunity costs between shoes and computers come to be similar in each
country.
The point at which specialization ends in both countries reflects both this
tendency toward a convergence in opportunity costs and the clearing of the
markets in both countries for computers and shoes (both domestically pro-
duced and demanded from the trading partner). Let us assume that markets
clear at the hypothetical exchange ratio, both domestically and internation-
ally, of 1 computer for 6 pairs of shoes. This exchange ratio is depicted by the
faintly-drawn lines in the two panels in Essential Economics 2.4. With these
State Power #823 ch2 7/2/02 4:00 PM Page 40

40 State Power and World Markets

mutually agreed-upon terms of trade, the United States moves its production
from EaA to PtA in Essential Economics 2.4(a), reducing its domestic produc-
tion of shoes from 150 million to 100 million pairs and, with the productive
resources thus released, increasing its computer output from 14 million to
26 million units. It then exports 10 million computers to Brazil in exchange
for 60 million pairs of shoes; rather than consuming at EaA, the United States
is now consuming at CtA.
If we compare CtA to EaA in Essential Economics 2.4(a), we find that even
partial specialization, when combined with trade, produces a consumption
outcome for the United States that is superior to what is optimally possible
in autarky. Prior to trade, the United States is able to consume 14 million
computers and 150 million pairs of shoes; with partial specialization and
trade, it is able to consume 16 million computers and 160 million pairs of
shoes. With partial specialization and trade, the United States, with no in-
crease in resources, increases its consumption of both computers and shoes,
and this is reflected in its movement from consumption indifference curve
U0A to the higher curve U1A.
Brazil also faces in these circumstances the opportunity for an improve-
ment in its consumption and overall satisfaction. If trade becomes possible,
Brazil moves along its production possibilities frontier from EaB to PtB in
Essential Economics 2.4(b): it produces 6 million computers rather than
12 million, and with the freed-up resources it increases shoe production
from 70 million to 140 million pairs. As noted in the preceding paragraph, it
then sends 60 million of those pairs to the United States in exchange for
10 million computers. This brings Brazil to the consumption point CtB, at
16 million computers and 80 million pairs of shoes. Brazil, then, also in-
creases its consumption of both computers and shoes as a result of partial
specialization and trade. This increased satisfaction is reflected in Brazil’s
movement from U0B to the higher indifference curve UlB in Essential Eco-
nomics 2.4(b). Hence, even in the face of increasing costs, Brazil, like the
United States, reaches a higher indifference curve by abandoning autarky
and embracing specialization and trade.

Accounting for Differences in Comparative Advantage:

Relative Factor Abundance and the Heckscher-Ohlin Theorem

In response, then, to the first of the questions that emerged from our explo-
ration of the Ricardian model of trade—why we do not witness full special-
State Power #823 ch2 7/2/02 4:00 PM Page 41

State Power and World Markets 41

¬ ESSENTIAL ECONOMICS 2.5

Similarities and Differences between the Ricardian and Neoclassical Models of Trade

RICARDIAN MODEL NEOCLASSICAL MODEL

Similarities in assumptions

Number of countries 2 2
Number of goods 2 2
Technology Similar across countries Similar across countries
Transportation costs 0 0

Differences in assumptions

Factors of production 1 (labor) 2 or more (land, labor, capital)

Opportunity costs Constant Increasing

Differences in results of trade

Specialization in production Complete Partial

Improved consumption Yes Yes

Core Principle
The Ricardian and neoclassical models of trade differ in their views of opportu-
nity costs and specialization. The neoclassical model answers the Ricardian
model’s question about full specialization by highlighting the importance of in-
creasing opportunity costs.

ization in the presence of international trade—neoclassical theory argues the

following: countries can mutually benefit if they undertake specialization and
trade because in the absence of trade they differ with regard to opportunity
costs between goods, but specialization and trade eventually eliminate these
differences and thus eliminate the incentive to trade or to specialize further.
The key similarities and differences between the Ricardian and neoclassical
models of trade are summarized in Essential Economics 2.5.
But what about the second question? Why is it that the United States has
a comparative advantage in computers rather than shoes, whereas Brazil has
a comparative advantage in shoes rather than computers? The Swedish
economists Eli Heckscher in 1919 and Bertil Ohlin in 1930 provided a help-
ful response to this question by investigating the implications of adding to
the neoclassical model the following assumptions:
State Power #823 ch2 7/2/02 4:00 PM Page 42

42 State Power and World Markets

■ Compared to the availability of capital and labor in one country,

another country will have relatively more or less capital than labor
(for example, the United States is a capital-abundant country,
whereas Brazil is a labor-abundant country).
■ Preferences or tastes for one commodity over another are similar
across countries. Differences in the relative amounts of two goods
that are produced and consumed in two countries prior to special-
ization and trade are therefore a reflection not of relative demand
for those two goods but of differences in the relative costs of pro-
ducing them in the two countries.

Working with these assumptions and those listed earlier under the neo-
classical theory, Heckscher and Ohlin developed a crucially important line of
analysis about trade. What has come to be known as the Heckscher-Ohlin
theorem of the bases of trade suggests that a country will have a compara-
tive advantage in, and thus will tend to export, those goods whose production
requires the intensive use of the factor of production that it has in relative
abundance.
In the Heckscher-Ohlin framework, because the United States, com-
pared to Brazil, is abundant in capital rather than labor, the cost of capital,
relative to wages, is likely to be lower in the United States than in Brazil.
Given the earlier assumption that economically efficient production tech-
niques are capital-intensive for computers and labor-intensive for shoes, the
United States can produce computers at a lower cost and therefore at a
lower price relative to shoes than can Brazil, and Brazil can produce shoes at
a lower cost and price in comparison to computers than can the United
States. This is equivalent to saying that the opportunity cost for producing an
additional computer in terms of shoes forgone is lower in the United States
than in Brazil.
The Heckscher-Ohlin line of inquiry thus suggests that differences across
countries in their endowments of the factors of production, plus differences
in the mixtures of factors of production with which different goods are opti-
mally made, can explain why Brazil has a comparative advantage in shoes
and the United States has a comparative advantage in computers. Given that
the United States and Brazil face different relative prices of computers and
shoes (and therefore different opportunity costs of one in terms of the other)
as a result of the differences in their factor endowments and the factor in-
tensities associated with the two products, the United States best improves
its prospects for increasing its overall consumption of both computers and
State Power #823 ch2 7/2/02 4:00 PM Page 43

State Power and World Markets 43

shoes by specializing in computers and exchanging some for Brazilian shoes,

whereas Brazil best improves its consumption opportunities by specializing
in shoes and exporting some to the United States in exchange for comput-
ers.6

Why Protectionism? The Problem of the Domestic

Distribution of the Gains from Trade

In our discussion so far, we have found that both the Ricardian and the neo-
classical theories of trade yield the same basic argument about why nations
engage in trade: nations are better off trading than remaining in autarky.
Trade and specialization on the basis of comparative advantage are clear ex-
amples of the power of human ingenuity and rationality, for the increases in
welfare in the two trading states result not from increases in capital or labor
in the two countries, but rather from the more efficient use by each of its ex-
isting capital and labor resources and the construction of a new human insti-
tution: voluntary exchange across national boundaries.
Given these core findings, economists have generally supported freer
trade among nations and have been highly skeptical of arguments favoring
restrictions on trade, generally referred to as protection. Indeed, given the
big gains that nations may attain from trade, it is hard to understand how
anyone could rationally question the wisdom of efforts by nations to facilitate
international commerce, or how someone could rationally support national
policies that impair such exchange.
However, as we will see in this section, modern trade theory has identified
and explored at least one dynamic associated with specialization and trade
that could lead some groups within a country to have a perfectly rational ba-
sis for opposing freer trade, even though the country as a whole would bene-
fit from more open trade.
To appreciate how economists have come to the view that there might in
fact be a rational basis for some groups within a country to prefer protection
over liberalization of trade, we proceed in three steps: first, we specify the
main types and effects of protection; second, we review arguments for pro-
tectionism that economists find highly unpersuasive; and third, we present
the argument for protection that, from the viewpoint of economic theory,
may be persuasive or even compelling.
State Power #823 ch2 7/2/02 4:00 PM Page 44

44 State Power and World Markets

Types of Protection and Their Impact on National Welfare

Instruments of Trade Protection

Governments have three basic classes of protectionist measures at their dis-

posal: tariffs, quotas, and non-tariff barriers.

TA R I F F S
A tariff is a tax imposed by government on imported goods or services.
Tariffs can be of two types. Specific tariffs impose a tax of a fixed amount on
each unit of an item being imported. For example, the United States might
charge $5 for each and every pair of shoes imported into its territory, regard-
less of the import prices of different types of shoes. Ad valorem tariffs, by
contrast, are based on the value of the items being imported. So, for exam-
ple, the United States could impose a 25 percent tariff on each and every
imported pair of shoes.

Q U OTA S
As an alternative to a tariff, a government might impose a quota or a
quantitative limit on the amount of a good or service that may be imported
during a period of time. For example, at a given price for shoes in the United
States, imports might total ten million pairs in a given year, but the United
States might impose a quota of five million pairs on the importation of shoes
per year, thus creating a shortage of five million pairs of shoes compared to
what would have been supplied had the quota not been imposed.

N O N TA R I F F B A R R I E R S
A third option available to a government is to put into place policies that
have the effect of increasing the cost of importing goods into the country;
such policies are called nontariff barriers. For example, the United States
might put into place a law that requires the U.S. military to purchase boots
and other footwear only from U.S. sources, or the U.S. government could re-
quire health and safety certifications for imported shoes that are more oner-
ous than such requirements for domestically produced shoes, or it could
require burdensome and expensive paperwork for importers who seek to
bring shoes through customs.
State Power #823 ch2 7/2/02 4:00 PM Page 45

State Power and World Markets 45

The Effects of Protection on National Welfare

Tariffs, quotas, and nontariff barriers raise the domestic prices of imported
goods and thus reduce the price advantages in the domestic market that
might otherwise be enjoyed by suppliers from countries with a comparative
advantage in the good against which protection is being imposed. As a result,
protection makes it less profitable for foreign producers to shift resources
into the production of the protected good even though they have a compara-
tive advantage in that good. At the same time, protection reduces the incen-
tive for local producers to shift resources out of the production of goods for
which their country does not have a comparative advantage and into goods
for which the country does have a comparative advantage.
So, for example, if by virtue of imposing some form of extreme protection
the U.S. government were to cause the relative prices of shoes and computers
to be equal in both the United States and Brazil, U.S. consumers would no
longer demand Brazilian shoes. As a result, the price of shoes in Brazil relative
to computers would not go up, and Brazilian manufacturers would have no in-
centive to shift resources out of computers and into shoes. At the same time,
with no additional supply of shoes from Brazil, or no Brazilian demand for U.S.
computers, American producers would have no interest in shifting resources
from shoes to computers. The imposition by the United States of extreme pro-
tection against shoe imports would, therefore, unwind the entire sequence of
steps we have explored whereby the United States and Brazil specialize in the
goods for which each has a comparative advantage and then engage in mutu-
ally attractive exchange. Both countries, as a result of protection, would lose
the substantial increases in satisfaction that can be attained by specialization
and exchange, even in the face of increasing opportunity costs.

Arguments for Protection Strongly Resisted by Economists

Economists, looking at the sort of welfare gains that are forgone when pro-
tection is imposed against imports, are unreceptive to most arguments in fa-
vor of protection. Two arguments are particularly unappealing to economists.

The “Unfair” Cheap Foreign Labor Argument

One argument in favor of protection that is often put forward in the United
States and in Europe is that protection is legitimate and necessary because
State Power #823 ch2 7/2/02 4:00 PM Page 46

46 State Power and World Markets

foreign countries compete on the basis of very cheap labor. Economists

respond to this argument by emphasizing that absolute differences in la-
bor costs by themselves do not create differences in comparative advantage
across countries. Workers in a poor country may be paid 15 cents an hour
while workers in the United States are paid $15 an hour. However, if, as a
result of having little in the way of shoe machinery, each worker in the
poor country can produce only 1 pair of shoes per hour while an American
worker using efficient machinery can produce 100 pairs an hour, then, from
the viewpoint of labor cost per pair of shoes, shoe production would be as
costly in the poor country (15 cents per pair) as in the United States (again,
15 cents per pair). It is not differences in the cost of any single factor of pro-
duction, economists emphasize, but differences in the cross-national relative
abundance and thus the cross-national relative costs of factors of production
that drive comparative advantage.

The Infant-Industry Argument

Producers and workers in an industry that might be opened to trade or that

is already under pressure from less-expensive imports sometimes bring for-
ward the argument that, if their industry were given temporary protection, it
would in fact become (or would again become) as competitive as foreign
suppliers. Just as a cruel and unfair world imperils children, who therefore
need the protection of their parents when they are young or they might not
survive, so too, these producers and workers suggest, can trade prevent their
“infant” industry from reaching its true potential as a world competitor. Sim-
ilarly, just as children need less assistance from their parents as they grow
up, so too, these claimants for protection suggest, would their industry even-
tually outgrow that temporary protection from foreign competition.
Economists emphasize that one key flaw with this infant-industry argu-
ment is that a government is likely to be unable to decide which among the
many industries putting forward requests for protection on the basis of
infant-industry claims are in fact worthy of protection. In addition, once an
industry is afforded protection on infant-industry grounds, that industry has
few incentives to become more competitive (after all, it has a protected mar-
ket). As a result, the government may soon find itself confronting an uncom-
petitive industry and a hard choice between allowing that industry to remain
in place, at the opportunity cost of the welfare gains that trade would afford,
and permitting trade but forcing a larger and perhaps more painful realloca-
State Power #823 ch2 7/2/02 4:00 PM Page 47

State Power and World Markets 47

tion of national resources across sectors than would have been necessary had
it not afforded protection to the infant industry in the first place.

The Stolper-Samuelson Theorem and a Rational

Preference for Protection

From the viewpoint of modern trade theory, cheap-labor and infant-industry

arguments for protection are usually ill founded. However, from the view-
point of that theory, at least one possible set of circumstances could provide
some groups within a country with a rational basis for opposing freer trade.
The line of analysis that identified and explained that set of consequences
was put forward by economists Wolfgang Stolper and Paul Samuelson (the
latter a Nobel laureate) in 1941, and the important result of their analysis is
known in economics as the Stolper-Samuelson theorem.
This theorem explores the effects of the opening of a country to trade
on the returns and hence the incomes of factor owners (for example, work-
ers, capital owners, or landowners) as their countries specialize on the basis
of comparative advantage as envisioned in the Heckscher-Ohlin frame-
work—that is, as the country specializes in those goods whose most efficient
production employs intensively the factor of production that is in abundance
in that country. The Stolper-Samuelson theorem suggests that such special-
ization can cause the owners of the relatively abundant factor to experience
a gain in their returns and real incomes, but can cause the owners of the rel-
atively scarce factor to sustain a drop in their returns and real incomes. This
dichotomy will result from changes in the demand for factors of production
as specialization proceeds on the basis of comparative advantage.
For example, in our case involving the United States and Brazil, when
trade commences, U.S. computer manufacturers will encounter an increase
in demand for their product, and to meet that increased demand they will
want to expand production. To do so they will need some additional labor
and a lot of additional capital equipment, for computers are a capital-
intensive product. In our earlier discussion, we identified the source from
which both sets of additional inputs would come: the shoe industry. But we
did not reflect on why capital will move from the shoe industry to the com-
puter industry. The answer is that computer manufacturers, in dire need of
capital, will be willing to pay higher prices for it, and it is the prospect of re-
ceiving such higher returns that induces some of the owners of capital
State Power #823 ch2 7/2/02 4:00 PM Page 48

48 State Power and World Markets

equipment in the shoe industry to provide it to the computer makers. Only

the payment of that higher price by shoe manufacturers will prevent other
capital owners from also abandoning the shoe industry.
Hence, higher demand for computers will lead to higher demand for capital
equipment, and, in the face of that higher demand for capital, capital owners in
both the computer and the shoe industries will enjoy a higher return and thus
an increase in their incomes. Prices of computers in the United States would
also go up as computer demand escalated in the face of the new export market
in Brazil, but the Stolper-Samuelson model suggests that the percentage in-
crease in the return to capital would be greater than the percentage increase in
the price of computers. Therefore, capital owners would enjoy not just an in-
crease in the returns to their factor of production, but also an increase in their
real income (that is, an increase in their income after taking into account
price changes) even if they purchased only computers. Thus, owners of the rel-
atively abundant factor of production gain from an opening of trade.
Owners of the relatively scarce factor—in the present discussion, labor—
face a different situation when specialization and trade go forward. As we
have just seen, the computer industry will expand its production by drawing
both capital and labor from the shoe industry. Let us also recall that, by as-
sumption, compared to each pair of shoes, each computer is made using a
great deal of capital and not much labor. Therefore, compared to the optimal
combination of capital and labor needed to make computers, too much labor
is being released relative to the amount of capital that is becoming available
as shoe production contracts. As a result, more workers are available than are
required for employment in the computer industry and the now-reduced
shoe industry, given the current production methods using capital and labor in
the two industries. Workers laid off from shoe factories who were unable to
get jobs in the expanding (but capital-intensive) computer industry would re-
spond to their situation by offering to accept lower wages if offered a posi-
tion in either industry, and, in the face of decreasing costs for labor, both
industries would hire these less costly workers and maintain overall output
by moving to more labor-intensive manufacturing processes. Hence, in the
face of specialization and trade, owners of the relatively scarce resource in
the United States (labor) would experience a decline in their returns (that is,
their wages), and, according to the Stolper-Samuelson theorem, this decline
in wages would not be offset by the percentage decline in the prices of shoes
as the latter became more plentiful through imports from Brazil. Workers
would, in sum, experience a decline in their real incomes.
In sum, although specialization and exchange with Brazil would make the
State Power #823 ch2 7/2/02 4:00 PM Page 49

State Power and World Markets 49

United States as a whole better off, the Stolper-Samuelson theorem provides

a basis for finding that a particular group within the United States—U.S.
workers—would be worse off as a result of such a move to specialization and
commerce. Hence, although the United States as a national community
might have a rational incentive to pursue trade liberalization, U.S. workers
would have a rational basis for resisting such a liberalization of exchange.
Similarly, although the Stolper-Samuelson theorem would help us under-
stand why workers in Brazil, as the owners of the relatively abundant factor
of production in that country, would unambiguously gain from trade and thus
rationally would have an incentive to press for trade with the United States,
the theorem would also lead us to expect that Brazilian capital owners would
rationally oppose such trade.
The main response by modern trade theory to the implication of the
Stolper-Samuelson theorem (that some groups within a country may have a
rational preference for protection) is that the gains from exchange are so
great that those who win from trade liberalization could offer compensatory
payments to those who lose, and both groups would still be better off with
trade than with autarky. Will that occur, and if so, what will be the form of
compensation? Will the country instead respond to demands for protection
by imposing restraints on trade to some degree, even if this reduces the na-
tion’s overall gains from trade? Or will the country pursue liberalization with-
out compensating domestic losers? These questions concern the politics and
political institutions of the country considering liberalization, the subjects of
Chapters 4 and 5.

Recent Developments in International Trade Theory

Ricardo’s insight into comparative advantage and mutually advantageous

trade, the neoclassical identification of increasing costs as a constraint on spe-
cialization, the Heckscher-Ohlin understanding of the sources of comparative
advantage, and the Stolper-Samuelson theorem regarding possible problems
with the distribution of the gains from trade remain at the heart of the analy-
sis by modern economists of the sources and consequences of trade. In recent
years, economists have sought to engage and, to some degree, amend the neo-
classical model in order to press ahead on two important research frontiers re-
garding trade: trade’s relationship to economic growth, and the importance of
intra-industry rather than inter-industry exchanges across nations.7
State Power #823 ch2 7/2/02 4:00 PM Page 50

50 State Power and World Markets

The Benefits from Trade:

From Greater Consumption to Higher Economic Growth

Our exploration of economic theory has shown how trade can yield new op-
portunities for countries to achieve higher levels of consumption. In recent
years, economists have also turned their attention to the way in which trade
might also help nations enjoy faster long-term rates of economic growth.
The capacity for international trade to improve the aggregate growth rate
of a country during the period in which integration is occurring has long
been recognized by employment of the model of trade outlined in this chap-
ter.8 The fundamental insight of this model is that, as a country opens itself
to world markets, it specializes in the production of a narrower range of
goods, the precise choice of which depends on the country’s endowment of
productive factors, local and world tastes for and prices of goods, and the
state of world technology (which is assumed to be available to all nations for
any given good). This narrowing of a country’s product range may in turn cre-
ate opportunities for the enjoyment of increasing returns to scale over some
range of the production runs of these goods. For example, by making more
shoes, Brazil may enjoy economies of scale resulting from larger runs of par-
ticular types of shoe. This enjoyment of economies of scale permits an in-
crease in the country’s aggregate growth rate. However, according to the
increasing-costs model of trade described earlier, declining economies of
scale would eventually curtail the capacity of this specialization to boost the
rate of economic growth for the country as a whole.9
Yet recent economic analysis highlights opportunities for trade to increase
the long-term economic growth rates of countries well beyond that antici-
pated by the standard model of trade described in this chapter. What is often
termed “endogenous growth theory” identifies the ways in which economic
growth and especially intense inter-firm competition in the context of a
growing economy may motivate entrepreneurs to seek out technological in-
novations. Technological innovations, in turn, produce increases in the pro-
ductivity of labor and capital, which, in turn, boost the rate of national
growth.10 Openness to trade, the new theory suggests and a number of em-
pirical tests appear to confirm, can enhance the long-term growth trajectory
of nations, because trade expands market opportunities for home producers
as well as instigates greater competition for them, and thereby it both en-
courages and compels these firms to seek out and to invest in new technol-
ogy. Moreover, trade creates new opportunities for local firms to gain access
to new, superior technologies from abroad. Through these and other mecha-
State Power #823 ch2 7/2/02 4:00 PM Page 51

State Power and World Markets 51

nisms, economists find, trade has a good chance of increasing the productivity
of the factors of production and thereby enhances national growth rates.11
Economists caution, however, that although trade between advanced
industrialized countries—that is, between countries that are similarly well
endowed with capital and technological capabilities—is likely to act as de-
scribed above, it is at least possible that developing countries that are lacking
in such resources may not find that trade with industrialized countries will
impart to their developing economies the same pressures and opportunities
to innovate. Thus, it is not certain that trade with industrialized countries, by
prompting the search for and facilitating the acquisition of new technology,
will necessarily place developing countries on higher long-term growth tra-
jectories than would be possible in the absence of trade.12 For example,
while emphasizing that they do not suggest that developing countries would
necessarily be better off by closing their economies to the world, economists
central to endogenous growth theory have noted that if the products in
which developing countries are prompted to specialize as a result of trade are
not associated with substantial opportunities for technical improvements,
then trade might not contribute very much to the rate of technological ad-
vance in those countries and thus it might not deliver the positive dynamics
identified in endogenous growth theory.13

The Content of International Commerce:

From Inter-Industry to Intra-industry Trade

The neoclassical increasing-costs model of trade that we presented in this

chapter is concerned with explaining why two countries would gain from the
trade of two commodities that are constituents of very different industries
(in our examples, shoes and computers). For some time, however, econo-
mists have recognized and have sought to understand why much of the in-
ternational commerce we actually see in the world consists not of such
inter-industry trade, but of intra-industry trade.
Intra-industry trade consists of the exchange between countries of goods
in the same industry.14 For example, the biggest single category of U.S. ex-
ports to and imports from Mexico in 1998 consisted of electrical machinery,
equipment, and related parts. This category of goods constituted about
24 percent of the total dollar value of U.S. exports to Mexico that year, and it
also constituted 27 percent of the total dollar value of U.S. imports from
Mexico. More generally, in the late 1990s, intra-industry trade accounted for
State Power #823 ch2 7/2/02 4:00 PM Page 52

52 State Power and World Markets

about 57 percent of all U.S. trade, 60 percent of all trade by European coun-
tries, and 20 percent of all trade by Japan.15
This large volume of intra-industry trade has prompted economists to
reflect upon a number of features of the neoclassical model of trade. For ex-
ample, in our discussion of that model, it was assumed that goods are homo-
geneous across producers both within and among countries: for example, the
shoes made in the United States are all exactly the same as those made in
Brazil. Economists have noted that, from the viewpoint of consumers, goods
are in fact often differentiated: for example, automobiles are differentiated in
the eyes of consumers in terms of being speedy or being fuel-efficient. Hence,
one reason we might observe intra-industry trade, economists have suggested,
is that while some consumers in a country may prefer one variety of a given
good, others may prefer another variety of the same good, and whereas one
country might have a comparative advantage in the one variety, another coun-
try might have a comparative advantage in the other.
But how might this difference in comparative advantage come into be-
ing? Here again, economists have reconsidered elements of the neoclassical
model of trade. As we have seen, that model assumes increasing opportunity
costs as a country shifts from the production of one good to the production
of another. One reason for such increasing costs was emphasized in our dis-
cussion regarding U.S. specialization in computers: as the United States in-
creases computer production by drawing resources from the shoe industry, it
must make use of progressively larger amounts of the “wrong” factor for com-
puter manufacturing—that is, labor. A second reason might also explain in-
creasing opportunity costs between shoes and computers: decreasing returns
to scale. As firms in the computer industry grow bigger and bigger, each pro-
portional increment of labor and capital used in production may result in
fewer and fewer additional computers being built. A variety of circumstances
may cause decreasing returns to scale. Managers, for example, may experi-
ence “bureaucratic diseconomies of scale”—that is, they may become pro-
gressively less able to coordinate production at their firms as those firms
become larger and larger.
Yet recent economic analyses have suggested that often we see not de-
creasing but increasing returns to scale within a firm, at least up to a point. In
other words, the larger the number of computers already being built by a
firm, the cheaper it is to build the next increment of computers, at least up
to some very large number of computers produced. The consequence of in-
creasing returns to scale might be an inversion of the production possibilities
frontier facing a country: over some significant range of production possibili-
State Power #823 ch2 7/2/02 4:00 PM Page 53

State Power and World Markets 53

ties, the curve might be convex rather than concave! If this is correct, then as
the firms in a country specialize in a particular variety of a good, they will en-
joy an ever-greater comparative advantage in that particular variety. Although
it might be chance that determines which countries specialize in which vari-
eties of goods, once they begin to specialize, increasing returns to scale may
prompt them to devote ever-greater resources to those varieties in which
they started to specialize in the first place.
Finally, we noted earlier that the Stolper-Samuelson theorem provides a
basis for expecting some elements of society to lose from trade and thus to
have a rational basis for seeking to prevent or to undo efforts by countries to
pursue trade liberalization. This line of reasoning may be correct with re-
spect to the effects of inter-industry trade on domestic income distribution.
However, economists have suggested that intra-industry specialization does
not require the sort of large redeployments of resources across industries
that may be associated with inter-industry trade.
In the neoclassical model of inter-industry trade, workers in a capital-
abundant country experience a decline in their real wages as the country
moves from the production of labor-intensive to capital-intensive goods.
However, if intra-industry trade prompts the movement of labor within that
country from one segment of a given industry into another segment of the
very same industry, then wages are likely to be adversely affected to a much
smaller degree, and, given that intra-industry trade augments the variety of
goods available within the country, workers might actually experience a net
improvement in their overall welfare through such trade.16 Intra-industry
trade then, is less costly in terms of domestic adjustment, less disruptive in
terms of inducing shifts in national income distributions, and less likely to
prompt demands for or movements toward protection as a country opens it-
self to the world trading system. These characteristics of intra-industry trade
may in turn have prompted governments to liberalize trade since World
War II in a manner that has fostered and perhaps even accelerated the de-
velopment of intra-industry trade.17

Conclusion

In this chapter we have offered a thumbnail sketch of how economic theory

helps us understand why states engage in international trade. Trade occurs
between countries because it can be mutually advantageous. Two states can
State Power #823 ch2 7/2/02 4:00 PM Page 54

54 State Power and World Markets

both benefit if each specializes on the basis of comparative advantage and

engages in mutually voluntary trade. A country’s comparative advantage is
likely to reflect its endowment of such resources as labor, capital, and land.
Although a nation as a whole stands to gain from international trade,
some segments of society within a nation may lose. In particular, owners of
the relatively scarce factor of production may experience a reduction in their
incomes. Society as a whole may choose to compensate those who lose from
an opening of the country to trade. However, economic theory cannot tell
us whether that will happen, how it will happen, the degree to which it
will happen, whether the nation will instead choose economic closure, or
whether society will choose openness without compensation of those left be-
hind. Those questions require analysis of political institutions and dynamics
both within and across nations, which we will turn to in Chapters 4 and 5.

Notes

1
This paragraph is inspired by the famous economist John Maynard Keynes’s reflections on
life in 1914 for “an inhabitant in London” at the height and, as it turned out, the end of the first
“Golden Era” of global economic integration. See Primary Document 7.1 on page 206 for those
reflections.
2
U.S. Department of Commerce, International Trade Administration, “U.S. Trade by
Commodity with Brazil,” available at www.ita.doc.gov/td/industry/otea/usfth/top80cty/brazil.cp.
Brazilian aircraft exports to the United States grew in importance at the end of the 1990s: these
totaled about $1.2 billion in 1999 and $1.5 billion in 2000, surpassing footwear exports in each
of those years.
3
On efforts by Brazil to foster the emergence of an indigenous computer-hardware industry,
with only limited success, see Peter B. Evans, Embedded Autonomy: States and Industrial Trans-
formation (Princeton: Princeton University Press, 1995).
4
For a superb analysis of Ricardo’s model of trade and the overall historical development of
international trade theory, see Douglas Irwin, Against the Tide: An Intellectual History of Free
Trade (Princeton: Princeton University Press, 1996); and John S. Chipman, “International
Trade,” in John Eatwell, Murray Milgate, and Peter Newman, eds., The New Palgrave: A Dictio-
nary of Economics (London: Macmillan, 1991), esp. pp. 937–52.
5
There might be additional factors of production that are economically meaningful for some
goods, such as land with respect to agricultural goods.
6
However, if tastes in the United States were such that its consumers had a much stronger
preference for computers over shoes than did Brazilians, then it might be the case that, in light
of increasing costs, local demand-driven high production of computers in the United States
could cause the opportunity costs between (and therefore the relative prices of ) computers and
shoes to be equal to those in Brazil, thus undermining the basis for specialization by either or
trade between them. It is for this reason that the Heckscher-Ohlin theorem assumes that tastes
State Power #823 ch2 7/2/02 4:00 PM Page 55

State Power and World Markets 55

between countries are not so dissimilar as to bring about a demand-led equalization of opportu-
nity costs.
7
A third characteristic of international trade is garnering increasing attention—namely, the
large incidence of intra-firm exchanges as a component of total world trade. According to an
analysis by the United Nations Conference on Trade and Development (UNCTAD), for exam-
ple, exports by firms to their own affiliates in other countries totaled about $2.3 trillion in 1998,
which is about 34 percent of the total of $6.7 trillion in world exports recorded that year. See
UNCTAD, World Investment Report 1999 (Geneva: UNCTAD, 1999), Table 5, p. 10.
8
Very useful overviews of the theoretical and empirical studies relating economic openness
to national economic growth are provided by Sebastian Edwards, “Openness, Trade Liberaliza-
tion, and Growth in Developing Countries,” Journal of Economic Literature 31 (September
1993), pp. 1358–93; and Sebastian Edwards, “Openness, Productivity, and Growth: What Do
We Really Know?” Economic Journal 108 (March 1998), pp. 383–98.
9
For a helpful graphical treatment of how, according to the basic increasing-costs model of
trade, an outward movement of the production possibilities frontier results from the achieve-
ment of economies of scale after specialization and trade, see Richard E. Caves, Jeffrey
A. Frankel, and Ronald W. Jones, World Trade and Payments: An Introduction (Reading, MA:
Addison-Wesley, 1999), pp. 36–39.
10
In this model, the sources of technological change and thus future economic growth are
internal, or endogenous, to current economic processes, thus yielding the term “endogenous
growth theory.”
11
A key early statement of endogenous growth theory was put forward by Paul M. Romer,
“Increasing Returns and Long-Run Growth,” Journal of Political Economy 94 (October 1986),
pp. 1002–37. Romer presents an important intellectual history of the development of the en-
dogenous growth research program in “The Origins of Endogenous Growth,” Journal of Eco-
nomic Perspectives 8 (winter 1994), pp. 3–22. A useful juxtaposition of the main elements of
endogenous growth theory with that of the traditional model of growth, formulated by Robert
Solow, is presented by David M. Gould and Roy J. Ruffin, “What Determines Economic
Growth,” Economic Review of the Federal Reserve Bank of Dallas (2nd quarter 1993), pp. 25–40.
For empirical studies that link international economic openness to investments in innovations
and thus medium-term greater economic growth, see Edwards, “Openness, Productivity, and
Growth,” and Robert Z. Lawrence and David E. Weinstein, “Trade and Growth: Import-Led or
Export-Led? Evidence from Japan and Korea,” National Bureau of Economic Research Working
Paper 7264 ( July 1999), available at www.nber.org/papers/w7264.
12
As a result, Luis Rivera-Batiz and Paul Romer, for example, emphasize that their work on
the manner in which trade prompts technical advances and thus growth is restricted to the cir-
cumstance in which both countries have advanced research-and-development capabilities. See
Luis A. Rivera-Batiz and Paul M. Romer, “Economic Integration and Endogenous Growth,”
Quarterly Journal of Economics 106 (May 1991), esp. pp. 532 and 550.
13
Gene M. Grossman and Elhanan Helpman, “Endogenous Innovation in the Theory of
Growth,” Journal of Economic Perspectives 8 (winter 1994), pp. 440–41.
14
For a key early work on this subject, see H. G. Grubel and P. L. Lloyd, Intra-industry Trade:
The Theory and Measurement of International Trade in Differentiated Products (New York: Wiley,
1975).
State Power #823 ch2 7/2/02 4:00 PM Page 56

56 State Power and World Markets

15
Roy Ruffin, “The Nature and Significance of Intra-Industry Trade,” Economic and Finan-
cial Review of the Federal Reserve Bank of Dallas (4th quarter, 1999), pp. 7–8.
16
See Paul R. Krugman, “Intraindustry Specialization and the Gains from Trade,” Journal of
Political Economy 89 (October 1981), pp. 959–73.
17
For interesting discussions of the political implications of intra-industry trade, see Howard
P. Marvel and Edward John Ray, “Intraindustry Trade: Sources and Effects on Protection,” Jour-
nal of Political Economy 95 (December 1987), pp. 1278–91; Charles Lipson, “The Transfor-
mation of Trade: The Sources and Effects of Regime Change,” in Stephen D. Krasner, ed., In-
ternational Regimes (Ithaca: Cornell University Press, 1983), pp. 258–62; and Beth V. Yarbrough
and Robert M. Yarbrough, The World Economy: Trade and Finance (Fort Worth: The Dryden
Press, 1994).