Non-Tariff Measures - A Fresh Look at Trade Policy's New Frontier

Non-Tariff Measures –
A Fresh Look at Trade
Policy’s New Frontier
edited by Olivier Cadot and Mariem Malouche
In a world where many forms of protection – including tariffs – are constrained by
WTO disciplines, non-tariff measures (NTMs) are the new frontier of trade policy. NTMs
that are poorly designed or captured by special interests can hurt competitiveness and
fragment markets; whereas well-designed ones can effectively overcome informational
and other market failures. Assisting governments in the design of NTMs is a critical
challenge for donors and development agencies. However, many issues relating, for
example, to regional harmonisation and the interaction of NTMs with market structure
are still imperfectly understood. This volume brings together recent work by young
scholars that draws on original data to shed light on some of the key analytical and
policy issues.
“Non-tariff measures are of increasing importance, and as we understand them better, we
also increasingly see their complexity. This volume provides up to date information and
analysis of this complexity, together with valuable implications for making NTMs better
serve legitimate purposes without distorting trade. It will be an essential source for trade
policy specialists as well as those concerned with domestic regulation.”
Alan V. Deardorff, John W. Sweetland Professor of International Economics and
Professor of Economics and Public Policy, University of Michigan
“Non-tariff measures (NTMs) are spreading, sometimes raising trade costs, sometimes
alleviating market failures. Disentangling these effects is urgent and necessary to inform
policymakers about the appropriate reforms to carry out. Drawing on new NTM data,
this volume leads the way showing how NTMs should be included as part of developing
countries’ domestic competiveness and regulatory reform agendas while also showing
that regional harmonisation of standards should be conducted in a flexible way.”
Jaime de Melo, Senior Fellow, FERDI and Professor, University of Geneva
“The application of non-tariff measures to international trade is expanding rapidly, but our
knowledge about them is not; at least, until now. This book offers a number of original
and instructive exercises to quantify the existence and the effects of NTMs and to see how
best to tame them as a form of protection. It will be an important guide and inspiration for
policy researchers as they seek to investigate the multifarious ways in which governments
manage trade for better or for worse.”
L Alan Winters, Professor of Economics, University of Sussex
a
a
a
THE WORLD BANK
THE WORLD BANK 9 781907 142536
ISBN 978-1-907142-53-6
Non-TariffMeasures–AFreshLookatTradePolicy’sNewFrontier

Non-Tariff Measures – A Fresh Look at Trade Policy’s New Frontier

Non-Tariff Measures – A Fresh Look at Trade Policy’s New Frontier
Copyright © 2012 by
The International Bank for Reconstruction and Development/The World Bank
1818 H Street, NW, Washington, DC 20433, USA
ISBN: 978-1-907142-53-6
All rights reserved
The findings, interpretations, and conclusions expressed herein are those of the
author(s) and do not necessarily reflect the views of the Executive Directors of
the International Bank for Reconstruction and Development/The World Bank or
the governments they represent.
The World Bank does not guarantee the accuracy of the data included in this
work. The boundaries, colors, denominations, and other information shown on
any map in this work do not imply any judgment on the part of The World Bank
concerning the legal status of any territory or the endorsement or acceptance of
such boundaries.
Rights and Permissions
The material in this publication is copyrighted. Copying and/or transmitting
portions or all of this work without permission may be a violation of applicable
law. The International Bank for Reconstruction and Development/The World
Bank encourages dissemination of its work and will normally grant permission
to reproduce portions of the work promptly.
For permission to photocopy or reprint any part of this work, please send a
request with complete information to the Copyright Clearance Center Inc., 222
Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax:
978-750-4470; Internet: www.copyright.com.
All other queries on rights and licenses, including subsidiary rights, should
be addressed to the Office of the Publisher, The World Bank, 1818 H Street
NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@
worldbank.org.

Centre for Economic Policy Research
The Centre for Economic Policy Research is a network of over 700 Research
Fellows and Affiliates, based primarily in European universities. The Centre
coordinates the research activities of its Fellows andAffiliates and communicates
the results to the public and private sectors. CEPR is an entrepreneur, developing
research initiatives with the producers, consumers and sponsors of research.
Established in 1983, CEPR is a European economics research organization with
uniquely wide-ranging scope and activities.
The Centre is pluralist and non-partisan, bringing economic research to bear
on the analysis of medium- and long-run policy questions. CEPR research may
include views on policy, but the Executive Committee of the Centre does not
give prior review to its publications, and the Centre takes no institutional policy
positions. The opinions expressed in this report are those of the authors and not
those of the Centre for Economic Policy Research.
CEPR is a registered charity (No. 287287) and a company limited by guarantee
and registered in England (No. 1727026).
Chair of the Board Guillermo de la Dehesa
President Richard Portes
Chief Executive Officer Stephen Yeo
Research Director Lucrezia Reichlin
Policy Director Richard Baldwin
The World Bank
The World Bank Group is a major source of financial and technical assistance
to developing countries around the world, providing low-interest loans, interest-
free credits and grants for investments and projects in areas such as education,
health, public administration, infrastructure, trade, financial and private sector
development, agriculture, and environmental and natural resource management.
Established in 1944 and headquartered in Washington, DC, the Group has over
100 offices worldwide. The World Bank’s mission is to fight poverty with passion
and professionalism for lasting results and to help people help themselves and
their environment by providing resources, sharing knowledge, building capacity
and forging partnerships in the public and private sectors.

Non-Tariff Measures – A Fresh
Look at Trade Policy’s New
Frontier
edited by
Olivier Cadot and Mariem Malouche

Contents
About the Authors xiii
Foreword xix
1 Overview 1
PART I: Prevalence and Pervasiveness of NTMs
2 How Much Light Do WTO Notifications Shed on NTMs? 23
Marc Bacchetta, Jürgen Richtering and Roy Santana
3 NTMs: Interpreting the New Data 57
Julien Gourdon and Alessandro Nicita
4 AVEs in TRIST 81
Olivier Cadot and Olivier Jammes
PART II: Standards and Technical Regulations: Trade
Barriers or Facilitators?
5 Reputation Matters: Spillover Effects in the Enforcement of
US SPS Measures 95
Marie-Agnès Jouanjean, Jean-Christophe Maur and Ben Shepherd
6 Sanitary Risk and Concentration in EU Food Imports 117
Mélise Jaud, Olivier Cadot and Akiko Suwa-Eisenmann
7 Do Better Standards Facilitate Exports? Evidence from China 141
Axel Mangelsdorf, Alberto Portugal-Perez and John Wilson
8 Standards Harmonisation and South–South Trade 155
Anne-Célia Disdier, Lionel Fontagné and Olivier Cadot
9 The Pro-Competitive Effect of International Harmonisation of
Product Standards 167
José-Daniel Reyes
10 Standards Harmonisation and Trade: The Case of the
East African Dairy Industry 187
Michael F Jensen and John C Keyser

PART III: Governments’Efforts to Improve NTMs
11 Improving the Quality of NTMs through Regulatory Impact
Assessment 213
12 Addressing Non-Tariff Barriers on Regional Trade in
Southern Africa 225
Ian Gillson and Nick Charalambides
13 Streamlining NTMs: How Mexico Did It 247
Ali Haddou
14 Regulatory Impact and Cost-Benefit Analyses – Tools to Evaluate
Regulatory Quality and Improve Transparency among Countries 271
Sílvia Helena G de Miranda
15 How Can Risk Management Help Enforce Technical Measures? 283
Christopher Grigoriou

List of Tables
Table A2.1 WTO notifications’coverage of NTMs by Multi-Agency Coding 50
Table A2.2 Members reporting for some key WTO notification requirements 54
Table A2.3 Precision of reporting for some key WTO notification requirements 55
Table 3.1 Use of multiple types of NTMs at different level of aggregation 70
Table 3.2 Frequency indices across economic sectors 73
Table 4.1 Illustrative case – the raw tariff data in importing country 89
Table 4.2 Ad valorem rates and pre-reform fiscal revenue 90
Table 4.3 Illustrative case: demand effect of eliminating a standard 90
Table 4.4 Illustrative case: fiscal revenue effect of eliminating a standard 91
Table 5.1 Data used in the analysis of US import refusals 105
Table 5.2 Regression results 109
Table 5A.1 Industry Class Codes for fruit and fruit products 114
Table 5A.2 Industry Class Codes for vegetables and vegetable products 115
Table 6.1 Structure of EU agri-food import (in percentage) 125
Table 6.2 Sanitary Risk Index, number of alerts and cause of rejection
by sectors 131
Table 6.3 The ten most and least risky products 132
Table 6.4 EU import pattern and Sanitary Risk Index 135
Table A6.1 Descriptive statistics of main variables, 2001–2005 140
Table 7.1 Examples of Chinese Standards 147
Table 7.2 Percentage of Chinese domestic standards 1992–2008 148
Table 7.3 Chinese agricultural exports: descriptive statistics 149
Table 7.4 Baseline results of OLS estimates 150
Table 8.1 Regression results, South–South trade 162
Table A8.1 North–South RTAs and Tradea
Coverage 166
Table 9.1 Four-digit SIC codes and descriptions for SIC code 36 171
Table 9.2 Pattern of trade in electronics between the EU-15 and its trading
partners 172
Table 9.3 Trade Costs by Industry and Year 174
Table 9.4 Import value to the EU-15, SIC4-exporter-importer-year,
1994–2004 179
Table 9.5 Probability of entering the EU market 181
Table 9.6 Change in log exports from established exporters 182
Table 10.1 Average bacterial counts in Kenyan milk samples 203
Table 12.1 NTBs that have been notified to SADC affect at least one-fifth of
regional trade 228
Table 12.2 Some progress has been made in implementing trade facilitation
instruments in Southern Africa 237
Table 12.3 NTBs cited in the NTB Monitoring Mechanism 238
Table 15.1 Contingency table for the predicted and actual status of the
declarations 291

Table 15.2 Criteria used to predict fraud – 100,000 observations 291
Table 15.3 Fraud measure on the scored criteria 293
Table 15.4 Contingency table 294

List of Figures
Figure 2.1 Types of notifications 35
Figure 3.1 Non-tariff measure classification 61
Figure 3.2 Frequency index and coverage ratios by chapter (all countries) 65
Figure 3.3 Frequency index and coverage ratios by chapter (by region) 66
Figure 3.4 Frequency indices and coverage ratios, by country 67
Figure 3.5 Number of NTMs from different chapters affecting HS 6-digit
products 69
Figure 3.6 Correlation of different 1-digit NTMs 72
Figure 3.7 Frequency indices across economic sectors, by region 75
Figure 3.8 Number of products covered by NTMs (2000 and 2010) 75
Figure 3.9 Tariffs and NTM coverage ratios 76
Figure 3.10 Correlation of NTM pervasiveness with MFN tariffs 77
Figure 3.11 Correlation of NTM pervasiveness with MFN tariffs, by product 77
Figure 4.1 How NTMs affect import volumes, values and unit values 86
Figure 5.1 Current versus lagged refusals 106
Figure 5.2 Refusals versus lagged refusals affecting similar products 106
Figure 5.3 Refusals versus lagged refusals affectiung the five closest
countries 106
Figure 6.1 The food alerts database, 2001–2005 122
Figure 6.2 EU food alerts notifications by the exporting country’s level of
development 123
Figure 6.3 Frequency of number of active suppliers 126
Figure 6.4 Distribution of point estimates of Sanitary Risk Index 130
Figure 6.5 Sanitary Risk Index and number of notifications in the RASFF 133
Figure 6.6 Evolution of EU import concentration 133
Figure 6.7 Import share distribution, for risky and safe products 135
Figure 7.1 Stock of voluntary and mandatory Chinese food standards 148
Figure 8.1 Number of North–South RTAs, by type of harmonisation clause 158
Figure 8.2 RTA configurements 160
Figure 9.1 Harmonisation and EU import value, all partner countries 176
Figure 9.2 Harmonisation and EU import value, developed partner
countries 177
Figure 9.3 Harmonisation and EU import value, developing partner
countries 177
Figure 10.1 Flow diagram of EAC trade requirements for dairy 194
Figure 11.1 A flowchart of the NTM toolkit 215
Figure 11.2 Political commitment and NTMs reform 216
Figure 15.1 Origial WCO Risk Management Framework 286
Figure 15.2 Declaration elements and infringements 292
Figure 15.3 Distribution of actual infringements by deciles of predicted
level of risk 295

List of Boxes
Box 3.1 Brief description of NTMs chapters 63
Box 10.1 Formal trade requirements as a competitiveness risk 191
Box 11.1 Institutional set-up to streamline new NTMs in Indonesia 219
Box 11.2 Two regulations hurting the competitiveness of
Mauritian firms 224
Box 12.1 Rules of origin in Southern African RTAs 230
Box 12.2 Obstacles facing the utilisation of tariff preferences in
SADC – the case of Shoprite 231
Box 12.3 The Tripartite NTB Monitoring Mechanism 238
Box 13.1 The Mexican Regulatory Impact Assessment process 255
Box 13.2 Review of courier services rules 257
Box 13.3 Review of food labelling standards 262

About the Authors
Marc Bacchetta is a counsellor in the Economic Research and Statistics
Division of the World Trade Organization which he joined in 1996. Between
2002 and 2004, he worked on trade related issues at the World Bank, on leave
from the WTO. His current research focuses on non-tariff measures and the
labour market effects of trade reforms.
Olivier Cadot is a professor of International Economics and director of the
Institute of Applied Economics at the University of Lausanne. Prior to taking
up his position at Lausanne, he was Associate Professor of Economics at
INSEAD. He has held visiting appointments at UCLA and McGill University,
New York University, Université d’Auvergne, Koç University, the Paris School
of Economics and the Institut d’Etudes Politiques de Paris. He was a Senior
Economist in the World Bank’s Trade Department between 2009 and 2011 and
has advised the French Government, the Swiss Federal Government and the
European Commission on trade policy matters. He also worked for the OECD and
the International Monetary Fund. He has contributed regularly to international
executive programmes. He is a Research Fellow of the Center for Economic
Policy Research in London, a Senior Fellow of the FERDI andAssociate Scholar
at CEPREMAP. He has published numerous scholarly papers on international
trade and economic development.
Nick Charalambides is senior associate consultant with Imani Development.
He is based in Botswana, working extensively throughout Southern and Eastern
Africa and the Caribbean and Pacific. Before this he worked for the Department
for International Development, seconded to the European Commission as trade
and poverty adviser on the Economic PartnershipAgreements and the Everything
But Arms Initiative.
Anne-Célia Disdier is a Researcher at the Paris School of Economics. Her
work focuses on the effects of tariff and non-tariff measures such as health,
phytosanitary and technical regulations on commercial trade and well-being.
Lionel Fontagné is Professor of economics at the Paris School of Economics,
Université Paris I Panthéon Sorbonne and part-time Professor at the European
University Institute (Florence). He has been the Director of the Centre d’Etudes
Prospectives et d’Informations Internationales (CEPII) from 2000 to 2006. He
is also a scientific advisor to CEPII and a consultant for the Banque de France.
Sílvia Helena Galvão de Miranda is currently a professor in the Department
of Economics, Business and Sociology at ESALQ/USP and a researcher at the
Center for Advanced Studies on Applied Economics at ESALQ, São Paolo. Her

xiv Non-Tariff Measures – A Fresh Look at Trade Policy’s New Frontie
main fields of research are international trade (particularly non-tariff barriers
to trade), trade and the environment and agribusiness. Recently, she has been
developing studies applying a cost-benefit analysis to Brazilian sanitary policies.
She was head of the Department of Economics in the Farmers Union of São Paulo
State (FAESP) from 1997-2001 and a participant in the Program of Capacity
Building in International Agricultural Trade promoted by IATRC (International
Agricultural Trade Research Consortium) and Hewlett Foundation from 2006-
2009.
Ian Gillson is a senior trade economist in the International Trade Department.
Previously he worked in the Africa Region of the World Bank as the Regional
Trade Coordinator for Southern Africa (2006-2011) after joining the Bank as a
Young Professional working in the Gender and Development Unit (2005). Prior
to this he was a Research Fellow at the Overseas Development Institute where
he developed work programmes on services trade, tariff preferences and trade
in agricultural commodities (2002-2005). He has also worked as an economist
on tax policy in the Ministry of Finance in Malawi (2000-2002). His recent
policy work focuses on regional integration, food trade, and trade/environment
linkages.
Julien Gourdon is an economist at CEPII, Paris. He holds a Ph.D. in Economics
from Université d’Auvergne (CERDI), Clermont-Ferrand. Julien is specialized in
International and Development Economics. His main topics of interest are trade
policies, exports competitiveness, the impact of international trade on income
inequalities, and the impact evaluation of trade assistance projects. Formerly, he
was Economist at the World Bank from 2006 to 2011, working for the MENA
region and at the International Trade department.
Christopher Grigoriou joined the CERDI (University of Auvergne, France)
as Associate Professor in 2007. He is also invited Professor at the University
of Lausanne and at the Swiss Graduate School of Public Administration. His
fields of interest are macroeconomic policies, international trade and applied
econometrics.
Ali B. Haddou-Ruiz joined Mexico’s Federal Competition Commission (CFC)
in 2005. In his role as Executive Secretary of the antitrust authority, he is
responsible for coordinating and supervising the work of the technical areas of
the CFC. He worked as an advisor to the Ministry of Commerce and Industrial
Development (SECOFI) in the negotiation of the North American Free Trade
Agreement during 1993, and was a member of the implementation team of the
Mexican government’s NAFTA Office in Washington, DC, from 1994 to 1996.
From 1996 to 2000, Mr. Haddou was advisor to the Head of the Economic
Deregulation Unit, also within SECOFI, and in 2000 he was named Cofemer’s
General Coordinator of Regulatory Impact Analysis, where he remained until
2004. During this period, he was voted to the OECD’s Regulatory Reform

About the Authors xv
Bureau. He has also participated in several independent regulatory reform
projects with APEC, the World Bank, Mexico’s Centro de Investigación y
Docencia Económicas, the consulting firm Jacobs and Associates, and with the
governments of Korea and Uruguay.
Olivier Jammes, a French national, holds several degrees in economics and is
an independent consultant. For the last 20 years he has been developing widely
used analytical tools for economists and policy makers, notably in international
trade. He currently lives in Laos with his family.
Mélise Jaud is an economist in the Chief Economist Office of the World
Bank’s Middle East and North Africa region. Her research interests cover firms’
dynamics and growth, international trade, trade finance and policy reforms in
developing and transition countries. Previously she worked she spent one year
in the Independent Evaluation Group. Prior to joining the World Bank she has
worked as an economic advisor to the Ministry of Agriculture in Mozambique
Marie-Agnès Jouanjean is a research fellow in the Groupe d’Economie
Mondiale at Sciences Po and LICOS Center for Institutions and Economic
Performance, KUL.
Michael Friis Jensen is an independent consultant specializing in quality and
standards issues in developing and transition countries. He holds a Ph D in
development economics from the Royal Agricultural University of Copenhagen,
Denmark and has led both research and operational work in more than twenty
countries in Africa and Central, South and South-East Asia. He led the work
of the World Bank’s International Trade Department on standards and trade
during 2007-2009 and continues to work for the World Bank and a number of
bilateral and multilateral organizations as well as private industry on quality and
standards issues.
Marie-Agnès Jouanjean is an agricultural and trade economist and international
development consultant. She has a multi-disciplinary background in conducting
academic and policy-oriented research. She has a combination of experience
both at an academic level and at an operational level working on trade and
development issues for international organisations such as the World Bank, the
OECD and the United Nations.Her specific areas of expertise include agriculture
and rural development economics and policy, Aid for Trade, product standards,
sanitary and phytosanitary measures (SPS), food security and supply side
constraints, and poverty dynamics.
John Keyser is an independent consultant in agriculture economics and rural
development. He has been living in Zambia for the past 17 years, from where
he has completed numerous assignments for World Bank and other international
organisations focused on quantitative value chain analysis, regional trade

xvi Non-Tariff Measures – A Fresh Look at Trade Policy’s New Frontie
competitiveness, rural market linkages, and dairy development. John has worked
asaconsultanttodairyprojectsinKenya,Uganda,andZambiatoassesstheimpact
of project interventions and costs and profitability of smallholder production.
Mariem Malouche in an economist in the World Bank International Trade
Department, Poverty Reduction and Economic Management. She joined the
World Bank in 2004 as an economist in the Middle East and North Africa
region. Before joining the World Bank, Mariem Malouche obtained her Ph.D.
in International Economics from University Paris-Dauphine, France. Her areas
of interest include trade policy, non-tariff measures, regional integration, export
diversification and trade finance. She has professional experience in the MENA,
SAR and AFR regions. She is co-editor of Trade Finance during the great Trade
Collapse, World Bank 2011 and co-author of Streamlining Non-tariff Measures:
A Toolkit for Policy Makers, World Bank 2012.
Axel Mangelsdorf is a postdoctoral researcher at the Federal Institute for
Materials Research and Testing in Berlin, Germany and the Berlin Institute of
Technology. He is an economist with areas of research interest covering impact
analysis of standards and standardization policy. He worked as consultant for the
World Trade Organization and the World Bank.
Jean-Christophe Maur is a senior economist in the World Bank Institute since
August 2008. His current responsibilities in WBI include leading the regional
integration and trade in services programs. Prior joining the World Bank, Jean-
Christophe was Economic Adviser in the UK Department of International
Development working on multilateral trade negotiations in trade facilitation
and non-agricultural goods market access, and also managing multilateral trade
assistance with the World Bank. Jean-Christophe has also worked for the Office
of Fair Trading, the UK Competition Authority, and the European Commission.
His research interests currently focus on regional trade integration and public
goods, trade facilitation and non-tariff barriers, intellectual property rights,
standards, and trade institutions. He recently published two edited volumes,
Preferential Trade Agreement Policies for Development: A Handbook (World
Bank) and Trade Costs and Facilitation: Open Trade and Economic Development
(E Elgar).
Alessandro Nicita is an economist at the United Nation Conference for Trade
and Development. Before his present position he worked in the Development
Economics Research Group at the World Bank. His research focuses on issues
related to international trade and development. His work has included the
measurement of the effects of trade policies, the analysis of policies improving
market access for developing countries, as well as the effect of trade policies
on poverty and inequality. He has authored several publications in the field of
international trade and economic development.

About the Authors xvii
Alberto Portugal-Perez is an economist at the World Bank. He has worked
on a variety of development and trade-related issues, such as trade costs, trade
facilitation, and aid for trade, FDI, standards, food prices, and regionalism. He
is currently working on aid effectiveness, concessional borrowing and fragile
states. He is the author of academic articles published in international refereed
journals and books.
José-Daniel Reyes is an economist working at the International Trade
Department of the World Bank. José-Daniel specialises in international trade and
development. His research investigates how changes in trade costs affect firms’
competitiveness in international markets. Formerly, he worked for the Inter-
American Development Bank (2003-2005) and for the Colombian government
(2001-2003).
Jürgen Richtering is Chief of the Integrated Data Base Section in the Economic
Research and Statistics Division of the World Trade Organization which he
joined in 1996. His current responsibilities include the management of WTO’s
tariff and import databases and the integration of these data with notifications
on non-tariff measures and services. He also contributes technical advice and
statistical services to various WTO Committees. Between 1983 and 1996, he
worked in UNCTAD, dealing initially with global modelling and then taking
charge of its Statistical Service. From 1991 to 1993, on leave from UNCTAD, he
worked in the Statistics Division of the Asian Development Bank.
Roy Santana is a counsellor in the Market Access Division of the World
Trade Organization, which he joined in 2004. His tasks include servicing the
Negotiating Group on Market Access for Non-Agricultural Products (NAMA),
and the Committees on Market Access, Customs Valuation, and the Information
Technology Agreement. Between 1998 and 2003, he worked for the Ministry of
Foreign Trade of Costa Rica where he was involved in various trade negotiations,
including the launch of the Doha Development Agenda.. He has lectured in the
IELPO program of the Barcelona University and the Summer Program of the
Graduate Institute in Geneva.
Ben Shepherd, the Principal of Developing Trade Consultants Ltd., is a
trade economist and international development consultant. He has worked on
a wide range of trade and development issues with organisations such as the
World Bank, the OECD, the Asian Development Bank, the United Nations,
and the FAO. He specialises in providing policy-relevant research, as well as
capacity-building seminars for researchers working in trade and development.
Ben’s particular areas of expertise include: trade facilitation, transport, and
infrastructure; governance, institutions, and anti-corruption; product standards,
technical barriers to trade (TBT), and sanitary/phyto-sanitary measures (SPS);
and trade in services, and the GATS.

xviii Non-Tariff Measures – A Fresh Look at Trade Policy’s New Frontier
Akiko Suwa-Eisenmann is a researcher at the French National Institute for
Agricultural Research (INRA), member of the Paris School of Economics, She
studiestheimpactoftradeliberalizationondevelopingcountries,aswellaswealth
distribution in France. She has been an administrator at the OECD development
centre and a consultant for the World Bank, the European Commission, the
French planning agency and the French national statistical institute.
John S.Wilson is a lead economist in the Development Economics Research
Group of the World Bank. He is also currently a Visiting Fellow at Columbia
University. He joined the Bank in 1999 and directs empirical and policy research
on trade facilitation, aid effectiveness, and regulatory reform issues, as they
relate to economic development. Mr. Wilson also provides expertise in Bank
operations and spent two years in the Bank’s Infrastructure Vice Presidency.
He was previously Vice President for Technology Policy at the Information
Technology Industry Council in Washington, D.C. and a Visiting Fellow at the
Institute for International Economics. He was also a Senior Staff Officer at the
U.S. National Academy of Sciences and National Research Council and Adjunct
Professor of International Affairs at Georgetown University.

Foreword
Governments continue to be subject to pressures to assist domestic firms and
industries, especially when times are bad as they are today.And incentives to use
traditional trade policies such as tariffs and quotas have changed and that other
types of non-tariff policies – such as subsidies and standards– have become
more attractive. Starting in the mid-1980s, average levels of tariff protection
in both industrialised and developing countries were gradually lowered. As of
2012, the average level of import protection has dropped to 10% or less in many
developing countries, and the average uniform tariff equivalent of merchandise
trade policies in OECD countries is now less than 4%. Imports of many
manufactures are now duty-free. Higher rates of tariff protection continue to
prevail for many agricultural products, but overall trade policies around the world
are dramatically less restrictive than they were in 1980. The ‘supply response’
to the liberalisation of trade has been enormous, with the value of global trade
expanding more than ten-fold, and foreign direct investment flows growing even
more, in part the result – and the driver – of the ever increasing fragmentation
of production and the geographic splintering of supply chains. One result of
the change in the structure of global production has been to encourage further
reduction in import tariffs, as any tariff that is imposed on any part of the supply
chain increases the cost of the final product and thus is detrimental to all the
links in any given chain.
While changes in the structure of global trade and production have implications
for the political economy of trade policy choices, analysis of these possible
impacts requires access to information on the prevalence and incidence of non-
tariff measures affecting trade. The availability of such information is rather
patchy. One lesson that has emerged from the Global Trade Alert trade policy
monitoring exercise is that it is a priority to invest greater effort in the collection
of such data. Another lesson is that it is difficult – but important – to determine
the impacts of non-tariff measures (NTMs), whether or not they are intended
to have a discriminatory effect on foreign suppliers of goods and services. In
practice, many of the NTMs that are imposed by governments have a regulatory
objective of one kind or another—safeguarding human health, plant safety,
national security, consumer protection, and so forth. Such objectives are all
legitimate reasons for government intervention – from a trade perspective the
challenge is to identify how they can be pursued in a way that minimises adverse
effects on the ability of foreign providers of goods and services to contest the
market.
Some NTMs—for instance, import or export licenses, permits and quotas—
complement or substitute for tariffs in regulating trade flows. Many others have
primarily non-trade objectives such as the protection of public health or the
environment, but affect the costs of crossing borders through the associated
procedural requirements. Countries also often maintain different standards

xx Non-Tariff Measures – A Fresh Look at Trade Policy’s New Frontier
for similar products, increasing the cost of compliance of exporters. Trade and
investment in services activities is likewise affected by numerous regulations,
many of which are essentially domestic in aim, but others are discriminatory by
design. Reducing the costs associated with regulations affecting the cross-border
movement of goods and services is a key part of any trade facilitation agenda.
Similarly, a major dimension of efforts to pursue greater economic integration of
markets revolves around actions to reduce the trade-impeding effects of NTMs.
The contributions to this volume analyse new NTM data that has been collected
in recent years by a number of international organisations working together,
and new databases on standards as well as on import refusals. In the overview
section, the editors point out how these contributions highlight the complex
interplay between regulatory design, market structure and political economy
that characterise NTMs; they also describe the importance and difficulties, in
terms of policy formulation, associated with accurate cost-benefit analysis.
We very much support their hope that the contributions in this book will be of
benefit for both academics and policymakers in contributing towards improving
our collective understanding of the trade effects of NTMs, so as to allow more
accurate and solid cost-benefit analysis.
Viv Davies Bernard Hoekman
Chief Operating Officer, Director, International Trade Department,
CEPR The World Bank

23rd
May 2012

1
Overview
A troubling phenomenon is occurring in developed and large, emerging
economies: governments, skittish about global economic trends, are introducing
new regulations to limit imports and exports. Recent World Bank analysis
(Hoekman et al, 2011) documents a trend of creeping protectionism in countries
as diverse as Argentina, Brazil and Indonesia. Instead of tariffs, other, more
indirect policies are used to hinder imports. The World Bank’s analysis, based
on World Trade Organization (WTO) monitoring reports and data from the
Global Trade Alert, a network of think tanks around the globe, shows a rising
trend in the number of non-tariff barriers (NTBs) in the first two years post-2008
followed by a sharp spike in 2011. India, China, Indonesia, Argentina, Russia
and Brazil together accounted for almost half of all the new NTBs imposed by
countries worldwide, with the EU and US accounting for the remainder.
The measures take various forms. In December 2011, amid a political shake-
up, Indonesia announced its intention to apply new technical regulations
on imported agricultural products, a move widely interpreted in the press as
signalling a tougher stance on imports. The same month, Argentina tightened
technical requirements on imports and gave officials increased discretion in
determining who was allowed to import. Brazil introduced new import licensing
requirements for car parts and other goods competing with domestic industry.
These moves are clearly aimed at restricting imports while attempting to remain
within WTO and regional commitments, with regulatory measures playing –
sometimes inefficiently – the role that tariffs might have played in the absence
of multilateral and regional disciplines.1
These examples highlight that, although the global financial crisis has not
triggered, as some feared, an explosion of protectionist measures, we are not
out of the woods yet. After all, it took 20 years for a more fragile world trading
1 In the words of Robert Baldwin (1970), ‘[t]he lowering of tariffs has, in effect, been like draining
a swamp. The lower water level has revealed all the snags and stumps of nontariff barriers that still
have to be cleared away’ (cited in Costinot, 2008). For a recent discussion of the policy substitution
hypothesis in the context of large-country terms-of-trade motives for tariffs and regulations, see
Staiger and Sykes (2009). See Limao and Tovar (2011) for an exploration in the case of Turkey.

2 Non-Tariff Measures – A Fresh Look at Trade Policy’s New Frontier
system to unravel after the crash of 1973.2
Were protectionism to creep back,
NTBs might play a central role because of their high obfuscation power and
the fact that, in spite of decades of progress, the disciplines to which they are
subjected are still imperfect.3
While NTBs are, by definition, protectionist by intent or effect, the term
‘non-tariff measures’ (NTMs) covers a broader array of regulations affecting
traded products.4
These measures can affect trade in many ways even without
protectionist objectives. As this book will show, viewing them exclusively
through the lens of protectionist games would be simplistic. They often respond
to societal demands for traceability and consumer protection against various
hazards, real or perceived – from GMOs, unhealthy foods, environmentally
damaging products, fraudulent pharmaceuticals, dangerous toys, and so on –
demands that are likely to grow over time and with consumer wealth.5
Indeed,
the WTO recognises the right of countries to regulate trade in order to achieve
legitimate non-trade objectives.6
Thus, NTMs are with us to stay – they have
already been around for a while7
– and perhaps even to proliferate, with or
without a protectionist plot.
NTMs are complex instruments; their impact on the economy being felt
through multiple channels. They are primarily regulatory instruments targeted
at particular market failures, but they affect the distribution of income in many
ways through their impact on market structure, costs and rents. Moreover, they
generate cross-border externalities not only through usual channels,8
but also
directly by segmenting markets. However, this multiplicity of effects is not
what sets them apart, as tariffs are also simultaneously fiscal and trade-policy
instruments.
Rather, what sets NTMs apart is how their effects interact with market structure.
NTMs have the power to segment markets directly, whether they be technical
standards or quantitative restrictions. Thus, they can generate market power
in a way tariffs cannot unless they are prohibitive. Their effects also depend
on the initial market structure in the importing country. For instance, if there
2 On this, see Bairoch and Kozul-Wright (1996) or Daudin et al (2008).
3 WTO (2012) provides a detailed history of the treatment of NTBs in the multilateral trading system
since World War II.
4 Our working definition excludes measures affecting services, as well as domestic production
standards, even if they affect the export competitiveness of domestic firms and hence trade flows.
That is, if a domestic regulation reduces the export potential of a domestic plant by prohibiting the
discharge of polluting effluent in a nearby river (thus raising its operating costs compared to those
of foreign competitors), by our definition it is not an NTM. By contrast, our definition includes
measures that affect equally imported and domestically produced goods; for instance, a labelling
requirement on all food containing GMOs is an NTM. See the chapter by Gourdon and Nicita in this
volume for a detailed discussion of conceptual and measurement issues.
5 See Grunert (2005) for a survey of the issues.
6 See Trebilcock and Howse (2001).
7 For instance, English laws in the 17th and 18th centuries required that all colonial trade be conducted
on British ships manned by British sailors. Also, certain goods had to be shipped to Great Britain first
before they could be sent to their final destination.
8 See Staiger and Sykes (2009).

Overview 3
is a single producing firm in the importing country, a quantitative restriction
on imports will give it market power whereas a tariff would not. Thus, ceteris
paribus, a quota of a thousand tons will lead to a higher equilibrium price in the
presence of a domestic monopoly than a tariff reducing imports to a thousand
tons.9
Some of the contributions in this book will show through careful empirics
and case studies that the interplay between regulatory design, market structure
and political economy that characterises NTMs is particularly complex.
Besides this inherent complexity, from a policy standpoint, what sets NTMs
apart is the centrality – and difficulty – of cost-benefit analysis. Balancing the
alleviation of market failures against the reduction of trade costs is a difficult
exercise for which national administrations are ill-equipped because these
considerations are under the purview of different government branches that
rarely work together. In addition, the relevant magnitudes have been scarcely
analysed or quantified simultaneously, even in the academic literature, except
perhaps in the context of legal disputes.10
We hope that the contributions in this
book will improve our understanding of the trade effects of NTMs, so as to put
such cost-benefit analyses on an increasingly solid footing.
PART I – Prevalence and pervasiveness of NTMs
In spite of over three decades of research,11
how the impact of NTMs on the
economy plays out empirically is still far from perfectly understood. Aside from
the complexities discussed above, the primary problem is one of transparency,
as data on NTMs is very fragmentary. WTO (2012) provides a good overview
of sources of information on NTMs. In Chapter 2 of this book, Marc Bacchetta,
Jürgen Richtering and Roy Santana describe the achievements and limitations of
the WTO’s key source of information, the notification system. The notification
system’s basic problem is one of incentives: because notifications are voluntary
with no sanction mechanism, countries can choose whether to notify or not.
Countries that do notify expose themselves to objections and possibly criticism,
whereas countries that don’t maintain an ambiguity – it is impossible to
know if lack of notification reflects restraint in the use of NTMs or failure to
communicate. Thus, incentives to comply are weak at best.
This voluntary nature makes it difficult to use the data for statistical analysis
purposes. Moreover, requirements have traditionally served various purposes
and the nature of the information they require Members to notify depends on
the purpose they serve. Sanitary and Phytosanitary (SPS) and Technical Barriers
to Trade (TBT) notification requirements, for example, aim at allowing other
9 This was first pointed out in Bhagwati (1968).
10 For a nice example, see Bown and Trachtman (2009). See also Beghin et al (2011). See Trachtman
(2008) and Cadot, Malouche and Saez (2012) for general guidelines.
11 Prominent early work includes, inter alia, Baldwin (1975), Feenstra (1984), Deardorff and Stern
(1985), Deardorff (1987), Baldwin (1989), Leamer (1990), or Anderson and Neary (1994). It is
difficult to do justice to the empirical literature on the trade effects of regulations and NTBs since the
mid-1990s, as it has grown at a very rapid pace.

Members to influence planned regulations.They do not require Members to notify
regulations that pre-existed the entry into force of the SPS and TBT Agreements,
nor the new measures in their final form. Thus, even if one sets the incentive
issue aside, notifications only provide a partial coverage of NTMs. Moreover,
the authors argue that the precision with which information is provided varies
considerably across notifications, calling for interpretation and assumptions if
one wants to process, summarise and present the information.
Modifying notification requirements and increasing compliance would help
improve measurability and would better serve the transparency purpose. The
WTO has attempted to go around the incentive problem by encouraging ‘reverse
notifications’,butwithlimitedsuccess.Otherpossibleavenuesincludeexpanding
specific trade concerns discussions (which seem to have largely replaced reverse
notifications) or systematically tapping other sources.
The WTO is now working closely with other agencies involved in NTM data
collection efforts to maximise the complementarity between WTO and other
sources of information on NTMs. One recent achievement is the creation of a
single classification of NTMs merging the 2009 Multi-Agency Support Team
(MAST – convened by the Secretary General of the United Nations Conference
in Trade and Development and its Group of Eminent Persons on NTBs)
nomenclature with that used by the WTO for notifications. In the long run, this
has the power to provide a boost to the creation of a much-needed worldwide
repository of NTMs. Under the impulse of UNCTAD, WTO notifications have
helped populating the World Integrated Trade Solutions (WITS) database on
NTMs for 2001/2. Researchers and the World Bank have used this information
to compute overall trade restrictiveness indices and ad-valorem equivalents to
measure their severity. These efforts have contributed to better apprehend the
impact of NTMs on trade.
In 2009, the World Bank has followed the MAST effort and funded NTM
data collection in 16 developing countries. Its model consisted of hiring local
consultants (universities, think tanks or consulting firms) and giving them
assistance and guidelines to draw up NTM inventories in collaboration with
ministries and agencies concerned. Relying on outside consultants was meant to
solve two basic problems that, as discussed in this report, plague self-notification:
(i) fragmentation of the locus of NTM issuance (which makes it more difficult
to gather information than in the case of tariffs), and (ii) perverse incentives not
to notify in order to avoid exposure. The inventories were formatted according
to the international classification, by product affected, together with information
on legal sources and enforcing agency, in order to ensure verifiability of the
information. Validation workshops were then held to ensure the data’s integrity
and seek national authorities’ approval. Finally, the data was forwarded to
UNCTAD for technical checks and posting on WITS.
Data was collected on Egypt, Lebanon, Morocco, Tunisia, Kenya, Mauritius,
Tanzania, Uganda, Senegal, Namibia, Madagascar, South Africa, Indonesia,
Cambodia, Lao PDR, and Bangladesh, in addition to the seven countries covered
by the pilot phase. Data was also collected, independently, by the EU, and a joint

Overview 5
World Bank/AfDB project is financing data collection on China. Counting 11
Latin American countries for which data collection was coordinated by ALADI
in collaboration with UNCTAD, data coverage actually stood at 31 countries
by the end of 2011 – still a far cry from universal coverage, but nevertheless a
marked improvement.12
In Chapter 3, Julien Gourdon and Alessandro Nicita offer the first statistical
analysis of the new data, exploring the incidence of various types of NTMs across
countries and industries. The incidence of NTMs varies considerably across
countries, industries and types of measures. Overall NTM coverage ratios vary
widely across countries from less than 10% to more than 90%, with no overall
pattern in terms of income levels. For instance, the EU stands out with over
85% coverage, but so do a number of East African countries including Burundi,
Kenya and Uganda. One might expect strong public concerns about traceability
and consumer protection to drive the extensive use of NTMs in the EU, but the
rationale for wide coverage in East Africa is less clear – especially in view of
efforts to streamline NTBs at the regional level.13
Gourdon and Nicita also highlight the prevalence of technical measures – SPS
and TBTs – over old-style ‘command-and-control’measures such as quantitative
restrictions and prohibitions. This may mark a welcome trend toward the
modernisation of NTMs, but it may also reflect the deflection of protectionist
pressures away from instruments subjected to stronger WTO disciplines.
Interestingly, the authors find that NTM coverage ratios correlate positively
with average tariffs across countries, suggesting policy complementarity rather
than substitution. Complementarity would be consistent with the existence of
common political-economy drivers to both NTMs and tariffs if both instruments
were subject to increasing marginal political costs.
At the time of printing, there was no published work analysing the ‘severity’
or trade-restrictiveness of NTMs using the new data, but work was ongoing.
Roughly, there are three ways of assessing the ad-valorem equivalent (AVE)
of the wedge introduced by NTMs between domestic and world prices. The
simplest method is a ‘handicraft’ one14
that consists in comparing the price of
a product affected by an NTM with its price in a similar market without the
NTM. It is essentially a case-study approach, widely used in WTO disputes such
12 As a follow-up to this and other data-collection initiatives, an ambitious multi-agency partnership,
Transparency in Trade (TNT), has been set up in 2011 by the World Bank, UNCTAD, the ITC, the
WTO and the African Development Bank (AfDB) to coordinate the collection and publication of
trade data, including on NTMs. In the area of NTMs, the TNT initiative is expected to give a ‘big
push’ to data collection with donor financing, creating a one-off global public good.
13 See UNCTAD (2012) for a description of COMESA, EAC and SADC NTB monitoring mechanisms.
See also EAC/EABC (2007).
14 The expression is from Ferrantino (2006), who provides a thorough description of the issues involved.

as the banana case.15
Indeed, it is recommended by Annex V of the WTO’s
Agricultural Agreement.16
Econometric methods fall into two categories: price- versus quantity-based
methods. Both methods yield AVEs in a broadly similar range. Price-based
methods use the variation of domestic prices at the country-product level to
identify the effect of NTMs, marked by dummy variables.17
The use of a country-
product panel makes it possible to include, in addition to tariffs, country-fixed
effects to control for transport costs, landlockedness, and other differences in
the cost of living, and product-fixed effects to control for differences in supply/
demand conditions between products. Examples of price-based econometrics
include Bradford (2003), who found average AVEs ranging between 7.8%
(Canada), 28% (UK) and 52% (Japan), and Andriamananjara et al (2004) who
found AVEs of up to 73% (for apparel) on a wide sample of prices obtained
from the Economist Intelligence Unit’s CityData database. Preliminary results
using similar methods in Cadot and Gourdon (2012a) suggest AVEs around 20%
for SPS measures in sub-Saharan Africa, where SPS methods are often more
bureaucratic hassle than anything else. Augier , Cadot, Gourdon and Malouche
(2012) found average AVEs for TBT around 13% in Tunisia and a whopping
60% for SPS in Morocco where, again, SPS measures applied on the domestic
market may be less than fully justified by consumer concerns.18
Quantity-based methods combine the variation in import values with
independent estimates of the price elasticity of import demand to identify the
effect of NTMs, also marked by dummy variables. Examples of quantity-based
estimation include Kee et al (2009) who found, across 4,545 cross-sectional
regressions at the product level, average AVEs of 9.2% (using simple averages)
and 7.8% (using trade-weighted ones). If one takes out products with no NTMs,
averages climb to 39.8% and 22.7% respectively in Kee et al (2009).
In Chapter 4, Olivier Cadot and Olivier Jammes illustrate the challenges of
using the information generated by the econometric estimation of AVEs of
NTMs and feeding it into partial equilibrium models to treat them like tariffs.
They highlight subtle difficulties that arise depending on the type of NTMs.
For example, licences or permits imposed to domestic importers generate costs
borne by consumers after the imported goods have cleared customs, in which
case their CIF import value is not directly affected. By contrast, some others,
like standards, may impose compliance costs to producers that may be passed
15 For an overview of the banana case at the WTO, see http://www.wto.org/english/tratop_e/dispu_e/
cases_e/ds27_e.htm. For an application of the ‘handicraft’ price-gap method to Nigeria’s import
prohibitions, see Cadot and Gourdon (2012b).
16 For a practical application, see eg Chemingui and Dessus (2004).
17 Thus, the estimating equation is like a difference-in-differences equation with product-country pairs
‘treated’by NTMs. If political-economy considerations are at play, the treatment can be instrumented
as in Cadot and Gourdon (2012a).
18 Paradoxically, while imposing price-raising NTMs, Morocco has food and fuel subsidies meant to
contain the cost of living, but at a staggering cost (18% of government expenditure and 4.5% of GDP
in 2008).

Overview 7
on, in full or in part, into the CIF value of imports. In such cases the treatment
must be different than for tariffs.
The authors use the World Bank Tariff Reform Impact Simulation Tool
(TRIST) model as a vehicle to illustrate some of the difficulties involved. In
particular, they show with a numerical example that eliminating a standard may
lead to reduced fiscal revenue from VAT and border taxes. However, the effect
on the tax base is less than the reduction in the compliance cost (the NTM’s
AVE) because direct effects and demand effects work at cross-purposes. The
direct effect of eliminating the compliance cost reduces the CIF value of imports
and thus the tax base, but the price reduction raises import volumes and thus the
tax base. If the elasticity of import demand is more than one, eliminating the
standard will raise the tax base. At the other extreme, if it is close to zero, the tax
base will shrink by close to the NTM’s ad-valorem equivalent.
PART II – Standards and technical regulations: Trade barriers or
facilitators?
Since the pioneering contributions of Otsuki et al (2001a, 2001b), evidence has
accumulated that standards and technical regulations,19
and in particular their
lack of harmonisation, can act as barriers to trade, in particular for developing
country exports.20
Wilson et al (2003) found that the stringency of antibiotics regulation for
bovine meat significantly reduced trade. Using their estimates to simulate the
effect of harmonisation of antibiotics regulations on the Codex Alimentarius
standard, they argued that South African, Brazilian and Argentine beef exports
would benefit substantially. The trade-inhibiting effect of SPS regulations on
developing-country agri-food exports was highlighted in a number of case
studies, including, inter alia, Henson et al (2000), Rahman (2001), Musonda and
Mbowe (2001), Henson et al (2004) or Henson and Mitullah (2004). The most
exhaustive assessment of the effect of technical regulations on agricultural trade
was carried out in Disdier et al (2008), who regressed disaggregated bilateral
trade flows in a gravity framework on coverage ratios and, in some specifications,
on AVEs as well, controlling for tariffs. In accordance with the case studies and
with the work of (inter alia) Moenius (2004), Maskus et al (2005), and Essaji
(2008), they found significant trade-inhibiting effects.
19 In everyday language, the term ‘standards’ refers to regulations affecting the technical or sanitary
characteristics of products. The WTO prefers the term ‘technical regulations’, reserving the term
‘standards’ to private ones, such as requirements imposed by supermarket chains on their suppliers.
In this book, as much as possible we will stick to the WTO terminology rule.
20 Otsuki et al (2001b) used a gravity equation to simulate the effect of the harmonisation at a stringent
level of the EU’s aflatoxin regulation on African exports. They predicted an export loss of $670
million. Their findings were later confirmed for South Africa, using the same combination of gravity
estimates and simulation, by Gebrehiwet et al (2007). However, a recent article by Xiong and Beghin
(2011) using ex-post estimation suggested that Africa’s groundnut exports to the EU were, contrary
to Otsuki et al’s grim predictions, unaffected by the new regulation, the key determinant of African
groundnut exports being domestic supply conditions.

By contrast, only a small number of papers have looked at how the enforcement
of SPS measures affects developing country exports. Alberini et al (2008)
examined the enforcement of the FDA’s seafood plant inspection programme,
while Baylis et al (2010) used the standard gravity methodology to demonstrate
the trade-reducing effect of US import refusals.
InChapter5,Marie-AgnèsJouanjean,Jean-ChristopheMaurandBenShepherd
push the analysis of US import refusals further to analyse the importance of
reputation in the enforcement of SPS measures. Their chapter reports striking
evidence that reputation effects matter in the enforcement of US SPS measures
through the import refusals system. Unsurprisingly, strong reputation effects
come from a country’s own history of compliance in a particular product;
however, the data also suggest two spillover channels. First, import refusals
are less likely if the exporting country has an established history of compliance
in relation to other goods in the same sector. Second, an established history
of compliance in relation to the same product by neighbouring countries also
helps reduce the number of import refusals. The authors also find some evidence
that these effects tend to be stronger for lower income countries.21
Incidentally,
Jouanjean et al’s chapter highlights a relatively little-known feature of the US
agri-food markets – namely, the existence of a positive list of countries allowed
to export fruit and vegetables to the US.A producer from a non-approved country
willing to enter the US market would find a closed door, until all procedures for
the country’s inclusion in the positive list have been met.22
The authors’ findings highlight the importance of a comprehensive approach
to upgrading quality-support infrastructure systems in exporting countries,
focusing on interlinked clusters of products rather than individual ones. It is
not sufficient to export a mix of compliant and non-compliant goods: reputation
matters, and the presence of the latter will make it harder to get the former into
the market as well. Consistency and reliability of production are therefore key
issues in the development of SPS capacity in poor agricultural exporters. They
also highlight potential benefits from regional cooperation in building SPS
compliance capacity and reputation. Regional approaches to the development
of standards systems are becoming more common for many reasons, such as the
ability for small, poor countries to pool technical and financial resources
In Chapter 6, Olivier Cadot, Mélise Jaud, and Akiko Suwa-Eisenmann also
find that reputation matters for exports to the EU. They examine the link between
the sanitary risk of food products and recent changes in the EU’s food-import
pattern, using a product-specific measure of sanitary risk constructed from a
new database of food alerts at the EU border. They argue that a raw count of
alerts could not measure correctly sanitary risk at the product level, as a product
imported from countries with weak quality management systems would appear
as risky even though some exporters might export it safely. Instead, they propose
21 Although the nature of the externality is different, Jouanjean et al’s results in this chapter are
consistent with those of Cadot et al (2011) which highlighted the existence of cross-firm externalities
in the survival of export products.
22 For more details, see Jouanjean (2012).

Overview 9
a measure of sanitary risk at the product level retrieved as the coefficient of a
product fixed effect in a regression of the count of alerts on exporter, importer
and bilateral control variables, including proxies for hidden protectionism.
Using their measure of sanitary risk at the product level, they find evidence
that the management of sanitary risk has indeed affected EU import patterns.
Overall, they observe a slight diversification of import sources over time;
however, this overall trend hides diversification at the extensive margin and
concentration at the intensive margin, in particular for risky products. That is, on
the one hand, the bulk of EU agro-food imports are being increasingly sourced
from a few countries; on the other hand, new suppliers enter the EU market, but
on a smaller scale than before, especially for risky products. As a result, EU
agro-food imports gradually evolve towards a two-tier distribution with a small
number of dominant suppliers and a growing fringe of marginal ones.
If product standards and technical regulations can act as trade inhibitors,
recent empirical work suggests they can also act as trade facilitators.23
For
instance, they can alleviate asymmetric-information problems (moral hazard and
adverse selection in terms of product quality) by subjecting all producers to a
common quality standard and encouraging them to invest in quality. Maertens
and Swinnen (2009) provide a detailed account of this effect in Senegal; see also
Jaffee (2003, 2005).
In Chapter 7, using a new database on Chinese standards, Axel Mangelsdorf,
Alberto Portugal-Perez and John Wilson approach the issue from a different
angle, looking at the effect of standards – voluntary and mandatory – in the
exporting country. They use a new database of Chinese food standards for seven
groups of agricultural products: meat, fish, vegetables, cereals, milk, tea, and
sugar over the period 1992–2008. Their results show that standards have a
positive effect on China’s export performance, as the benefits to standardisation
in terms of reducing potential information asymmetry and signalling enhanced
food safety and quality in foreign markets seem to surpass compliance costs.
The positive effect of Chinese standards is larger when they are harmonised
to international measures. The results suggest that there are clear benefits to
China’s steps to base their domestic standards and regulations on international
measures.
As a counterpoint, in Chapter 8, Olivier Cadot, Anne-Célia Disdier and Lionel
Fontagné highlight the potential dangers, for Southern countries, of premature
harmonisation with stringent Northern standards. For instance, as part of the
modernisation of its regulatory environment, Mauritius has recently been
considering the adoption of a technical regulation on toxic chemicals used in
paints that would closely resemble EU regulations. The benefit of harmonising
technical regulations with the EU is that Mauritius’s paint manufacturers could
have their products distributed in EU supermarkets, including in the nearby
23 For an early discussion, see Jaffee and Henson (2004).

Reunion. However, production costs would increase,24
and the increase would, in
all likelihood, price them out of East and South African markets where chemical
residues in paints are not an issue. Would enhanced access to Northern markets
compensate foreseeable losses on Southern ones?
The authors generalise on this example to ask how the harmonisation of
technical regulations in the context of North–South regional trade agreements
(RTAs) affects trade patterns. The issue is particularly pregnant for countries
in the zone of influence of large commercial powers, like Middle East and
North Africa (MENA) countries with the EU, as they may be encouraged to
converge rapidly in terms of regulatory environments. The authors explore the
issue empirically by combining a gravity model with an extension of Piermartini
and Budetta’s (2009) database on RTAs containing information on regulatory-
harmonisation clauses. Strikingly, they find that RTAs with harmonisation
clauses do reinforce hub-and-spoke trade structures (centred on the Northern
RTA partners) potentially detrimental to the integration of Southern countries in
the world economy, reinforcing previous findings by Chen and Mattoo (2008)
on the trade-diverting effects of regional harmonisation. In a world where South-
South trade grows faster than North–South (Gourdon, 2011), this may be a high
price to pay.
When they are harmonised, standards and technical regulations can also
contribute to reduce market fragmentation by overcoming coordination failures
among producers and ensuring interoperability (for instance, harmonising
electric-plug design improves the tradability of electrical appliances).25
For
instance, Shepherd (2007) showed that a country adopting international
standards significantly raised the variety of its import sources, in particular from
low-income countries, suggesting a significant market-access improvement.
In the same vein, Czubala et al (2009 ) showed that the trade-restricting effect
of the EU’s technical regulations in the textile and apparel sector on African
exports was mitigated when those technical regulations were consistent with
international standards.
However, the effects of standards harmonisation can be complex. In Chapter
9, using a rich firm-level dataset, Jose-Daniel Reyes highlights in a novel way
the interplay between standards harmonisation and market structure, using as
a natural experiment the harmonisation of EU electronics regulations and their
alignment on international standards that took place between 1991 and 1996
under the Lugano and Dresden agreements, which reduced the proportion of
EU-specific standards from half to about a quarter. Using highly detailed firm-
level data on US electronics exporters, he shows that the harmonisation of EU
technical regulations raised US exports at the ‘extensive margin’; that is, by
inducing entry of new US exporters. This suggests that barriers to entry were
reduced by the harmonisation. It also shows that harmonisation can increase
24 In an interview with the authors, a local producer estimated the cost increase at 2% to 40% of baseline
average costs, depending on the product.
25 For an interesting case study of the effect of multiple standards on trade, see Raballand and Aldaz-
Carroll’s (2005) study of the market for pallets.

Overview 11
the intensity of competition on a given market. However, Reyes also highlights
an interesting side-effect: as competition became tougher on the EU electronics
market with the entry of new US competitors (and, presumably, more intense
intra-EU competition), producers from developing countries were sidelined
and lost market share. This is paradoxical, as the lowering of trade barriers had
apparently opposite effects on US and developing country competitors, raising
profitability for the former and reducing it for the latter. Clearly, modelling the
effect of regulations in a heterogeneous-firm framework is needed to highlight
the complex effects at play.
One policy implication is clear from Reyes’ analysis – namely, that as
harmonisation and mutual recognition of technical regulations raise the intensity
of competition, donors may want to think about strengthening technical-
assistance programmes for developing country exporters to help them strive in
tough, competitive environments.
Harmonisation processes, like all regulatory decisions, can also be hijacked
by special interests, and their political economy, still relatively unexplored,
feeds on market-structure effects of the type that Reyes identified. In Chapter
10, Michael Jensen and John Keyser discuss the recent decision by the East
African Community (EAC) to harmonise member states’ technical regulations
for dairy products with international ones. The authors argue that the approach
does not make much sense from a ‘benevolent’ welfare perspective given local
consumption habits (no-one consumes the milk without first boiling it). By
contrast, harmonisation is likely to stifle intra-regional trade in dairy products, as
only the very largest operators are capable of meeting the stringent international
standards – indeed, it is a barrier to entry for smaller producers that suits the
interests of the larger ones.
PART III – Governments’ efforts to improve NTMs
The third part of the book is dedicated to the policy options for governments to
unilaterally improve the process, design and implementation of NTMs. While
WTO multilateral efforts to harmonise NTMs have not made any tangible
progress, and regional efforts tend to be lengthy processes and product or
sector specific, the unilateral approach to improve NTMs to maintain firms’
competitiveness is realisable by all governments.
The World Bank has recently developed practical guidelines based on an
innovative approach that carefully balances the reduction of business costs
against the preservation of ‘local public goods’. Cadot, Malouche and Saez
(2012) argue that seeking to eliminate NTMs altogether or to cut their number
through mechanical formulae would likely be an ill-conceived quest. Most
NTMs respond to a genuine public demand for traceability and protection
against hazards to health and the environment – a demand that can be expected
to grow over time and with wealth. Thus, except in the case of obvious non-
tariff barriers, policy toward NTMs should strive for their improvement through
better design, smarter enforcement and, ultimately, through a robust governance

framework. Moreover, NTM improvement should be viewed primarily as a
domestic issue, part of a regulatory improvement agenda driven by a concern for
enhanced competitiveness, rather than as a concession to trading partners.
The toolkit’s approach features a sequential analysis that identifies (i)
where the problem lies (if there is a ‘market failure’ that justifies government
regulation), (ii) whether the existing measure addresses it adequately and in a
cost-minimising way, and (iii) how, overall, the measure’s costs compare with
its benefits. It is adaptable, depending on (i) the scale of the problem, (ii) local
capabilities, and (iii) political traction for reform. Moreover, its implementation
can start on a small scale. Its basic assumption, though, is that the content and
depth of NTM reviews should not be watered down for the sake of adaptation
to limited-capabilities environments, lest the exercise become useless. Instead,
it should be accompanied by targeted and well-conceived technical assistance.
The toolkit also proposes an institutional setup to organise practically the
NTM review process. The setup is centered on a committee bringing together
all government agencies involved in the issuance of NTMs, the private sector,
and civil society (representatives of consumers or environmental associations).
Considerable long-term benefits can be expected from embarking in such a
process, even if starting progressively. First, the review mechanism provides
a forum for competitiveness-oriented dialogue with the private sector which
can have benefits going well beyond NTMs. Second, the review process can
progressively set the stage for a larger institutional mechanism to improve the
governance of new measures, trade-related or not. One may even envisage, in
the long run, the merger of regulatory-oversight and competition-watchdog
functions, as these functions draw on the same type of expertise (law and
industrial economics), and a single overseeing agency may have more clout.
Third, if such institutional mechanisms were to be set up in several countries in
a region or trading bloc, with similarly trained personnel, they could provide an
informal vehicle for the regional coordination of new measures at the technical
level, before uncoordinated decisions have a chance to generate political friction.
Numerous initiatives have strived to coordinate the reduction of non-tariff
barriers at the regional level, although with unequal success (see UNCTAD,
2012 for an overview). In Chapter 12, Ian Gillson and Nick Charalambides draw
lessons for developing countries from the experience of southern Africa, arguing
that there are clear benefits from closer regional coordination in regulation and
standards, particularly where national agencies face technical skill shortages or
capacity constraints. Furthering efforts to develop appropriate regional standards
(eg through SADCSTAN) can also reduce the costs of trading across borders
and reduce the scope for capture by national private sector interests (see the
discussion of dairy regulations, supra). Unlike other regions in the continent,
southern Africa has the distinct advantage that world-class testing bodies and
accreditation services are already available in South Africa. Regional integration
can leverage the capacity of these institutions to support exports and capacity
in the less advanced countries, eliminating the practice of double testing and
certification. Alternatively, countries could consider steps towards implementing

Overview 13
a regional framework for mutual recognition of the conformity-assessment
procedures qualifications/licences for service providers.
The authors also discuss a significant regional transparency initiative, the NTB
Monitoring Mechanism (NTBMM), suggesting leads to push the mechanism in
the direction of effective NTB removal. However, success has been limited so
far, highlighting the difficulty of setting up regional agendas when capabilities
are limited at the supranational level and the drive for regional integration is
weak.
At the country level, the NTM improvement agenda should be viewed
primarily as a domestic issue, part of a regulatory-improvement agenda driven
by a concern for enhanced competitiveness, rather than as a concession to
trading partners. This is the perspective taken by countries that have successfully
adopted ambitious regulatory-reform agendas, like Korea or Mexico.
In Chapter 13, Ali Haddou draws important lessons from Mexico’s regulatory-
reform efforts since the 1990s. Mexico embarked in an ambitious regulatory
reform in the middle of the December 1994 ‘tequila crisis’ when it became
clear that recovery could only come from export-led growth, requiring a
drastic improvement in the environment of business. According to Haddou,
including trade issues under the fold of a government-wide regulatory
improvement was key to make it a vehicle for improved competitiveness. He
also highlights the importance of consistency and breadth, as reforms showed
strong complementarities and could not efficiently be tackled piecemeal. Most
importantly, for Haddou what matters in the process is not so much reducing
trade costs by eliminating this or that measure, but putting in place a robust
governance framework to ensure the sustainability of regulatory improvement.
Drawing on the Mexican experience, he advocates the creation of a central
regulatory-oversight body with sufficient capabilities and clout to articulate a
vision for the country’s regulatory and trade policy beyond the vagaries of short-
term lobbying and electoral demands.
Mexico’sexperiencehighlightsthelong-termchallengesthatawaitaregulatory-
reform drive. Whereas the country’s anti-trust body has thrived and gained
political clout, COFEMER, its regulatory-oversight body, seems to have picked
fights that proved unwinnable and progressively lost its power and momentum.
This raises the issue of whether regulatory-oversight and antitrust functions
should be merged and entrusted to a single watchdog agency overseeing both
the private sector’s anti-competitive practices and the public sector’s regulatory
abuses. This would put the agency in an impartial position, leverage synergies
in its work (as antitrust and regulatory issues are deeply intertwined) and would
generate economies of scale, drawing on similar expertise (essentially law and
industrial economics).
In Chapter 14, Silvia de Miranda sheds light on Brazil’s ongoing efforts to
apply Regulatory Impact Assessment (RIA) to new SPS regulations. First-
hand experience with this process allows her to draw lessons on key success
factors, including financial commitment to support solid analytical work. Given
financial constraints in developing countries and the fact that governments do

not always apprehend, ex-ante, the benefits they may reap by applying RIA,
this is an area where donor support could be critical. Another critical success
factor when implementing RIA is to count on wide stakeholder participation.
RIA allows for a consultancy process at all steps, from the new regulation’s
definition and submission to the discussion of the RIA’s results, in order to
raise public awareness before the regulation is issued. Effective consultation not
only reduces the risk of policy mistakes, but it also ensures buy-in and reduces
adaptation costs.
Both Haddou’s and Miranda’s analyses highlight two key challenges facing
regulatory reform and the effective deployment of RIA. First, given the
fragmentation of regulatory authority across government agencies and ministries,
communication and coordination are key, although typically few structures exist
for such coordination, bureaucratic structures being often more conducive to
rivalry than to cooperation. Second, it requires strong top-down political will,
preferably from the highest level, in order to ensure that deadlock resulting from
‘horizontal’ conflicts between ministries or agencies with different mandates
and views can be broken.
Does this mean that nothing can be done if everything is not done? To some
extent, yes. International experience suggests that piecemeal regulatory reform
is inefficient and vulnerable to reversals. However, it also suggests that sweeping
political drives for efficiency, relying on radical approaches such as the so-called
‘guillotine’ (which consists in eliminating a certain number or proportion of
regulations, or all those that do not have a compelling justification), may either
quickly lose steam when serious trade-offs appear, or become ‘deregulation’
drives rather than drives for regulatory improvement, with potential costs for
local public goods such as consumer safety or the environment. Thus, in order
to prepare the ground for regulatory reform, long-term engagement is needed to
help administrations build analytical capabilities and learn to tackle technical
problems with pragmatic, technical solutions. This is the approach of the World
Bank’s NTM streamlining toolkit (Cadot et al, 2012).
This approach emphasises the need to reduce the costs incurred by the private
sector in complying with regulations while not diluting their power to protect
consumers and the environment. While the principle is clear, how to implement
it on the ground is not always self-evident. In Chapter 15, Christopher Grigoriou
discusses this key practical issue in the context of SPS measures. Drawing on the
state of the art in border management, he shows how risk management can be
used to reduce the incidence of controls while keeping the risk of non-detection
undiminished. Using econometric techniques well known to economists,
he shows that risk profiling can use the track record of operators as well as
observable risk factors to target inspections to where they have the highest power
to detect fraud or quality-control failures.

Overview 15
Key takeaways
Three main takeaways may be highlighted following this new thread of data and
analyses:
First, new effort to collect data in both developed and developing countries has
helped shed light on the prevalence and pervasiveness of NTMs. This effort calls
for more data collection and analysis of the trade restrictiveness of NTMs. The
new Transparency in Trade (TNT) initiative led by the World Bank, UNCTAD
and the International Trade Centre, with contributions from some regional
development banks and selected donors, would help make the NTM data
collection more widespread and sustainable across economies. Transparency
in the area of trade policies is essential. It plays an important role in allowing
private economic agents to make informed decisions and thereby helps markets
function more efficiently. Transparency in NTMs has become prominently
crucial with the increased use of NTMs to regulate international trade, including
during crisis times.
Second, governments should recognise the importance of addressing the NTMs
as part of their domestic competitiveness and business regulatory reform agendas.
Governments should aim to adopt good regulatory principles and improve inter-
government coordination with wide consultation processes. Governments should
set up a coordination committee among ministries that issue trade regulations
and facilitate consultations with the main stakeholders, including representatives
of consumers or environmental associations. Regulatory Impact Assessment
should be adopted to improve the quality of regulations and to avoid unnecessary
burden and nuisance for private operators. Risk management techniques can
also be used by Customs and other border agencies to reduce the incidence of
controls at the border. These instruments should be tailored to the scale of the
problem, local capabilities and political traction for reform.
Finally, regional harmonisation of standards, in particular North–South
harmonisation, should be conducted strategically and in a flexible way,
modulating the speed of convergence on the basis of export potential from
developing countries. For sectors whose primary export potential is the main
regional partner, fast harmonisation may be optimal; for sectors whose primary
export potential is in the South, slow harmonisation may be better. Small farmers
and producers may suffer the most from unnecessary harmonisation. Moreover,
as harmonisation and mutual recognition of technical regulations raise the
intensity of competition, developing country exporters would highly benefit
from technical assistance programmes to help them strive in tough competitive
environments.
References
Alberini, A, Lichtenberg, E, Mancini, D and Galinato, G (2008), ‘Was it
Something I Ate? Implementation of the FDA Seafood HACCP Program’,
American Journal of Agricultural Economics, 90(1), pp 28–41.

Anderson, J and Neary, P (1994), ‘Measuring the Restrictiveness of Trade
Policy’, World Bank Economic Review, 8, pp 151–69.
Andriamananjara, S, Dean, J, Feinberg, R, Ferrantino, M, Ludema, R and Tsigas,
M (2004), ‘The Effects of Non-Tariff Measures on Prices, Trade, and Welfare:
CGE Implementation of Policy-Based Price Comparisons’, USITC Office of
Economics Working Paper 2004-04-A.
Bairoch, P and Kozul-Wright, R (1996), ‘Globalization Myths: Some Historical
ReflectionsonIntegration,IndustrializationandGrowthintheWorldEconomy’,
paper prepared for the WIDER Conference on Transnational Corporations and
the Global Economy, King’s College, Cambridge, September 1995.
Baldwin, R (1970), Non-Tariff Distortions in International Trade, Brookings
Institution, Washington DC.
Baldwin, R (1975), Foreign Trade Regimes and Economic Development: The
Philippines, Columbia University Press, New York.
Baldwin, R (1989), ‘Measuring Nontariff Trade Policies’, National Bureau of
Economic Research, Working Paper No 2978.
Beghin, J, Disdier, A-C, Marette, S and van Tongeren, F (2011), ‘Measuring
Costs and Benefits of Non-Tariff Measures in Agri-Food Trade’, Iowa State
University Working Paper 11001.
Bhagwati, J (1968), ‘More on the Equivalence of Tariffs and Quotas’, American
Economic Review, 58(1), pp 142–6.
Bown, C P and Trachtman, J P (2009), ‘Brazil – Measures Affecting Imports
of Retreaded Tyres: A Balancing Act’, World Trade Review, 8(1), pp 85–135.
Bradford, S (2003), ‘Paying the Price: Final Goods Protection in OECD
Countries’, Review of Economic and Statistics, 85, pp 24–37.
Cadot, O, Iacovone, L, Pierola, D and Rauch, F (2011), ‘Success and Failure of
African Exporters’, The World Bank, Policy Research Working Paper 5657.
Cadot, O, Malouche, M and Saez, S (2012), Streamlining Non-Tariff Measures:
A Toolkit for Policy Makers, World Bank Group
Chemingui, M and Dessus, S (2004), ‘Assessing Non-Tariff Barriers in Syria’,
Economic Research Forum Working Paper 0425.
Chen, M and Mattoo, A (2008), ‘Regionalism in Standards: Good or Bad for
Trade?’ Canadian Journal of Economics, 41(3), pp 838–63.
Costinot, A (2008), ‘A Comparative Institutional Analysis of Agreements on
Product Standards’, Journal of International Economics, 75, pp 197–213.
Czubala, W, Shepherd, B and Wilson, J (2007), ‘Help or Hindrance? The Impact
of Harmonized Standards on African Exports’, World Bank Policy Research
Working Paper 4400.
Daudin, G, Morys, M and O’Rourke, K H (2008), ‘Globalization, 1870–1914’,
Oxford University Department of Economics Discussion Paper Series #395.
Deardorff, A V (1987), ‘Why Do Governments Prefer Nontariff Barriers?’
Carnegie-Rochester Conference Series on Public Policy, 26, pp 191–216.
Deardorff, A V and Stern, R M (1985), ‘Methods of Measurement of Nontariff
Barriers’, United Nations Conference on Trade and Development, UNCTAD/
ST/MD/28, Geneva.

Overview 17
Disdier, A-C, Fontagné, L and Mimouni, M (2008), ‘The Impact of Regulations
on Agricultural Trade: Evidence from the SPS and TBT Agreements’,
American Journal of Agricultural Economics, 90, pp 336–50.
EAC/EABC (2007), ‘Monitoring Mechanism for Elimination of Non-tariff
Barriers in EAC’, mimeo.
Feenstra, R C (1984), ‘Voluntary Export Restraints in U.S. Autos, 1980–81:
Quality, Employment, and Welfare Effects’, in R E Baldwin and A O Krueger
(eds), The Structure and Evolution of Recent U.S. Trade Policy, University of
Chicago Press, Chicago.
Ferrantino, M (2006), ‘Quantifying the Trade and Economic Effects of Non-
Tariff Measures’, OECD Trade Policy Working Paper 28.
Gebrehiwet, Y, Ngqangweni, S and Kirsten, J F (2007), ‘Quantifying the Trade
Effect of Sanitary and Phytosanitary Regulations of OECD Countries on South
African Food Exports’, Agrekon, 4(1), pp 23–39.
Gourdon, J (2011), ‘Wage Inequality in Developing Countries: South-South
Trade Matters’, International Review of Economics, 58, pp 359–83.
Grunert, K (2005), ‘Food Quality and Safety: Consumer Perception and
Demand’, European Journal of Agricultural Economics, 32, pp 369–91.
Henson, S, Brouder, A M and Mitullah, W (2000), ‘Food Safety Requirements
and Food Exports from Developing Countries: The Case of Fish Exports from
Kenya to the European Union’, American Journal of Agricultural Economics,
82(5), pp 1159–69.
Henson, S and Mitullah, W (2004), Kenyan Exports of Nile Perch: Impact of
Food Safety Standards on an Export-Oriented Supply Chain, World Bank,
Washington DC.
Henson, S, Saquib, M and Rajasenan, D (2004), ‘Impact of Sanitary Measures
on Exports of Fishery Products from India: The Case of Kerala’, The World
Bank, Agriculture and Rural Development Discussion Paper 17.
Hoekman, B, Datt, M and Malouche, M (2011), ‘Taking Stock of Trade
Protectionism Since 2008’, Economic Premise, December, The World Bank
Jaffee, S (2003), ‘From Challenge to Opportunity: Transforming Kenya’s Fresh
Vegetable Trade in the Context of Emerging Food Safety and Other Standards
in Europe’, Agricultural and Rural Development Discussion Paper, The World
Bank.
Jaffee, S (2005), ‘Delivering and Taking the Heat: Indian Spices and Evolving
Product and Process Standards’, Agricultural and Rural Development
Discussion Paper, The World Bank.
Jaffee, S and Henson, S (2004), ‘Standards and Agro-Food Exports from
Developing Countries: Rebalancing the Debate’, The World Bank, Policy
Research Working Paper 3348.
Jouanjean, Marie-Agnes (2012), ‘Market Access Food Standards: Insights
from the Implementation of US Sanitary and Phytosanitary Regulation’; Paris:
Institut d’Etudes Politiques de Paris, unpublished Ph.D dissertation.
Kee, H L, Nicita, A and Olarreaga, M (2009), ‘Estimating Trade Restrictiveness
Indices’, Economic Journal, 119, pp 172–99.

Leamer, E E (1990), ‘The Structure and Effects of Tariffs and Nontariff Barriers
in 1983’, in R W Jones and A O Krueger (eds), The Political Economy of
International Trade, Basil Blackwell, Cambridge, MA.
Limao, N and Tovar, P (2011), ‘Policy Choice: Theory and Evidence from
Commitment via International Trade Agreements’, Journal of International
Economics, 85(2), pp 186–205.
Maertens, M and Swinnen, J (2009), ‘Trade, Standards, and Poverty: Evidence
from Senegal’, World Development, 37(1), pp 161–78.
Maskus, K, Otsuki, T and Wilson, J (2005), ‘The Cost of Compliance with
Product Standards for Firms in Developing Countries:An Econometric Study’,
The World Bank, Policy Research Working Paper 3590.
Moenius, J (2004), ‘Information versus Product Adaptation: The Role of
Standards in Trade’, Northwestern University International Business and
Markets Research Center Working Paper #1.
Musonda, F M and Mbowe, W (2001), The Impact of Implementing SPS and
TBT Agreements: The Case of Fish Exports to European Union by Tanzania,
Consumer Unity Trust Society, Jaipur.
Otsuki, T, Wilson, J and Sewadeh, M (2001a), ‘What Price Precaution? European
Harmonization of Aflatoxin Regulations and African Groundnut Exports’,
European Review of Agricultural Economics, 28(3), pp 263–84.
Otsuki, T, Wilson, J and Sewadeh, M (2001b), ‘Saving Two in a Billion:
Quantifying the Trade Effect of European Food Safety Standards on African
Exports’, Food Policy, 26, pp 495–514.
Piermartini, R and Budetta, M (2009), ‘A Mapping of Regional Rules on
Technical Barriers to Trade’, in A Estevadeordal, K Suominen and R Tech
(eds), Regional Rules in the Global Trading System, Cambridge University
Press, Cambridge, pp 250–315.
Raballand, G and Aldaz-Carroll, E (2005), ‘How Do Differing Standards
Increase Trade Costs? The Case of Pallets’, The World Bank, Policy Research
Working Paper 3519.
Rahman, M (2001), EU Ban on Shrimp Imports from Bangladesh: A Case Study
on Market Access Problems Faced by the LDCs, Consumer Unity Trust
Society, Jaipur.
Shepherd, B (2007), ‘Product Standards, Harmonization, and Trade: Evidence
from the Extensive Margin’, World Bank Policy Research Working Paper
4390.
Trachtman, J P (2008), The Economic Structure of International Law, Harvard
University Press, Cambridge, MA.
Trebilcock, M and Howse, R J (2001), The Regulation of International Trade,
Routledge, London.
UNCTAD (2012), Non-tariff Measures to Trade: Economic and Policy Issues for
Developing Countries, United Nations, Geneva.
Wilson, J, Otsuki, T and Majumdsar, B (2003), ‘Balancing Food Safety and
Risk: Do Drug Residue Limits Affect International Trade in Beef?’, Journal of
International Trade Economic Development, 12(4), pp 377–402.

Overview 19
World Bank (2008a), ‘A Survey of Non-tariff Measures in the East Asia and
Pacific Region’, World Bank Policy Research Report.
– (2008b), ‘Non-tariff Measures on Goods Trade in the EAC’, Report #45708-
AFR.
– (2008c), Regulatory Transformation in the Republic of Korea, IFC/World Bank,
Washington DC.
– (2008d), Regulatory Transformation in Mexico, 1988–2000, IFC/World Bank,
Washington DC.
WTO (2012), Looking beyond International Cooperation on Tariffs: NTMs and
Services Regulations in the XXIst Century, World Trade Organization, Geneva.
Xiong, B and Beghin, J (2011), ‘Aflatoxin Redux: Does European Aflatoxin
Regulation Hurt Groundnut Exporters from Africa?’ Iowa State University
Working Paper 11026.

PART I
Prevalence and Pervasiveness of NTMs

2
How Much Light Do WTO Notifications
Shed on NTMs?
Marc Bacchetta, Jürgen Richtering and Roy Santana1
2.1 Introduction
Transparency in the area of trade policies is essential. It plays an important role in
allowing private economic agents to make informed decisions and thereby helps
markets function more efficiently. Also, Bhagwati (1988, p 85) has suggested
that protectionism can be cut back by what he calls the Dracula effect: exposing
‘evil’ to sunlight helps to destroy it. Awareness of the existence and of the cost
of certain policies, both within the country and in its trading partners, can put
pressure on governments to eliminate them. Even if transparency does not directly
‘destroy’ barriers to trade, however, it can indirectly contribute to more open
trade by facilitating trade cooperation. While the theory of international trade
cooperation has little to say about transparency, because it typically assumes that
agents are all perfectly informed, there are good reasons to believe that imperfect
information is, in practice, an important issue in the case of trade agreements. If
this is true, transparency can play a role in improving Members’ compliance to
the commitments they have made as part of a trade agreement.
One of the functions of the WTO is to facilitate the implementation,
administration and operation, and further the objectives, of the Marrakesh
Agreement and the Multilateral (and Plurilateral) Trade Agreements.2
Because
transparency at both the multilateral and the national level is essential to the
smooth operation of the agreements, and more broadly to the good functioning of
the multilateral trading system, WTO Agreements include multiple transparency
provisions. Following Wolfe (2003) and WTO (2007), these transparency
provisions can be grouped into five categories: (a) goods and services schedules
of concessions; (b) the Trade Policy Review Mechanism; (c) publication and
notification requirements; (d) internal transparency of the institution towards its
Members; (e) external transparency of the institution towards civil society. The
first three categories cover provisions that ensure the transparency of national
1 The authors are members of the WTO Secretariat. The views expressed here are their own and should
not be attributed to Members of the WTO or to the Secretariat.
2 Article III.1 of the Marrakesh Agreement Establishing the WTO.

policies, which in the area of trade in goods can be roughly divided into two
groups: tariffs and non-tariff measures (NTMs).
While considerable efforts to improve transparency on the tariff side have
been made since the end of the Uruguay Round, which have also benefited from
technological progress in the information technology area, much remains to be
done in the area of NTMs. By the end of the Uruguay Round tariff information
was only available electronically for about 40 Members and it was not available
to the general public because of very restrictive data dissemination policies. Since
then, the situation has improved dramatically, both with regard to data coverage
and data dissemination policies which are now far less restrictive.3
Thanks
to these developments, information on tariffs is now freely available through
the WTO website and covers all WTO Members’ bound commitments in the
Consolidated Tariff Schedules (CTS) database, as well as a large and increasing
percentage of Members’ applied tariffs in the Integrated Data Base (IDB). With
tariffs largely under control, attention has now turned towards NTMs, a broad
category that encompasses all trade and trade-related policy measures with
the exception of tariffs, which could restrict or distort international trade in
goods. Various initiatives aimed at enhancing transparency of NTMs have been
launched, one of which involves several multilateral agencies.4
Together with the publication obligation under Article X of the GATT 1994
and the Trade Policy Review Mechanism, notifications are one of the three
mechanisms generating information on NTMs in the WTO system. This chapter
examines the role that notifications play in the WTO and, more broadly, in the
multilateral trading system. More specifically, it asks what role notifications
have played and could play as a source of information on NTMs. It starts with
a discussion of the reasons why, and the manner in which, notifications were
developed under the GATT 1947. Section 2.3 then offers a description of the
current notifications landscape under the WTO and assesses how much of the
NTM universe is covered by such notifications. Section 2.4 discusses the quantity
and quality of notifications under the WTO, and concludes that in most cases
both compliance with the requirements and the quality of information received
are low. Section 2.5 considers what would be needed to expand their role as a
major source of transparency in NTMs. Section 2.6 concludes.
3 Two key decisions in July 2009 contributed to this improvement. The decision of the Market
Access Committee on a ‘‘Framework to enhance IDB Notifications Compliance’’ (G/MA/239 of
4 September 2009) made it easier for the Secretariat to assist Members in providing their trade and
tariff notifications by allowing the use of other relevant official sources. The second decision (G/
MA/238 of 4 September 2009) on full public disseminations of all tariff line duties created the basis
for the WTO to become a key player in the dissemination of tariff data information to the general
public.
4 See UNCTAD (2010).

How Much Light Do WTO Notifications Shed on NTMs? 25
2.2 Notifications in the GATT 1947
At the time negotiations for the establishment of an International Trade
Organization (ITO) began in 1946, the General Agreement on Tariffs and
Trade (GATT) was foreseen as a parallel and provisional ‘tariff’ agreement that
would lock in a number of tariff concessions while the broader negotiations
were concluded. The negotiating plan that was approved by the ‘Preparatory
Committee’ in February 1946 envisaged that the GATT would be legally
independent of the ITO Charter and would be brought into force as soon as
possible.5
In order to safeguard the value of the tariff concessions that would
be made in the GATT, and which would eventually become part of the ITO
framework, a number of ‘non-tariff’ clauses would be included in the General
Agreement by referencing or reproducing text from the ‘draft’ ITO Charter.6
Probably because of its provisional nature and lack of a formal structure, the
General Agreement that was signed at the end of 1947 only made scant reference
to formal notification requirements.7
As will be described in this section, the vast
array of notification requirements that exists today at the WTO is not only the
result of the provisions originally included in the GATT 1947, but also of a series
of decisions and agreements that were subsequently taken which progressively
developed new requirements, procedures, questionnaires and formats. Moreover,
the rationale behind the notification requirements also evolved to take account of
the changing needs of the multilateral trading system.
2.2.1 The Havana Charter, the GATT 1947 and its modifications
While the terms ‘notify’ and ‘notification’ only appeared in Articles XVI and
XVIII of the GATT that was signed in October 1947, there were other provisions
that required formal communications by the contracting parties. These included
three Articles which made reference to ‘notices’8
and eight which required
contracting parties to ‘consult’ when introducing certain measures.9
In contrast,
the Havana Charter (HC), which was meant to bring into life the ITO, stipulated
a higher number of ‘notifications’,10
some of which were not required by
the GATT. These included notification requirements relating to quantitative
restrictions (HC Article 20:3(b)), state monopolies (HC Article 31), and
establishment of customs unions or free trade areas (HC Article 44:3(a)). The
5 Resolution adopted on the First Session of the Preparatory Committee of the United Nations
Conference on Trade and Employment, established by the Economic and Social Council of the
United Nations on 18 February 1946. See Annexure 10 of UN doc. E/PC/T/33, Sections C and H.
6 The original plan did not work as intended and changes had to be introduced, including the adoption
of provisions revised in the context of the ITO. See, for example, GATT/CP.2/22/Rev.1 of 30 August
1948.
7 See UN document E/PC/T/214.Add.1.Rev.1.
8 Articles XIX:2 and 3(a), XXVII, XXVIII:2(a) of the GATT that was signed in 1947.
9 Articles VI:6, XII:4, XIII:4; XIV:6(b); XV:2 ; XIX; XX:II(c) of the GATT that was signed in 1947.
In addition, Article II:6(a) requires ‘concurrence’ to adjust specific duties.
10 See, for example, Havana Charter, Articles 13:6, 9 and 10; 14:1; 20:3(b); 25; 28:2; 31:3; and 44:3(a).

Havana Charter also provided a general ‘information’ obligation through Article
50:3, which required that ‘[e]ach Member shall furnish to the [ITO], as promptly
and as fully as possible, such information as is requested by the Organization for
its consideration and investigation of complaints and for its conduct of studies
under’ the trade Chapter.
NotificationsundertheGATT1947hadsimplergoals.Leavingasidethenotices
of a purely procedural nature (e.g. the ones making requests for consultations),
there were mainly two types of notifications. The first one required contracting
parties to give notice before certain measures were introduced, which would then
be followed by consultations and eventually negotiations for compensation. This
is the case, for example, of the ‘emergency actions’ under Article XIX11
– which
could nevertheless be dispensed in case of ‘critical circumstances’ – and Article
XVIII for the introduction of ‘non-discriminatory measures’ that would conflict
with concessions underArticle II of the GATT (e.g. tariff bindings).12
The second
type of notification required contracting parties to notify certain measures in
place in order to be able to preserve them. This was the case of Article XVIII:6,
which allowed contracting parties to preserve non-discriminatory measures for
the reconstruction of particular industries or branches of agriculture in force
before 1 September 1947, as long they were notified by 10 October 1947.13
Since the GATT 1947 had borrowed language from the draft ITO Charter
of 1946, and the ‘equivalent’ ITO provisions had continued to evolve as the
parallel negotiations progressed, contracting parties considered it necessary to
incorporate the revised language. This was done through a series of Protocols
amending the General Agreement, three of which were open for signature at the
same time as the Havana Charter in 1948.14
An additional ‘Protocol modifying
Part II and Article XXVI of the GATT’, was agreed shortly afterwards amending
Articles III, VI, XIII, XV, XVIII and XXVI.15
These amendments introduced
a new series of ad hoc notification requirements, including inter alia in: (1)
Article XVIII: 6, 8, 11 and 12, which resulted from a complete redraft of these
11 Article XIX:2 provides that before any contracting party can take action as per XIX:1, it shall give
notice in writing to the other contracting parties as far in advance as may be practicable. Contracting
parties having substantial interest shall be afforded an opportunity to consult in respect of the
proposed action.
12 Article XVIII:2(a), 4(c), 5(a) and (b) of the GATT that was signed in 1947.These notifications would
trigger a response by the contracting parties considering to have been ‘substantially affected’ by the
proposed measure, an ‘examination’ where contracting parties would notify if they ‘concur[red] in
principle with the proposed measure’ and, eventually negotiations to reach ‘substantial agreement’.
13 This would, however, not be a permanent exception. Within 60 days after becoming contracting
parties, relevant contracting parties should also notify the reasons in support of preserving those
measures and the period for which they wished to maintain them. Other Contracting Parties would
then examine and decide whether said measures would need to be modified or withdrawn.
14 See Press Release, ITO/228, of 24 March 1948. These included the Protocol Modifying Certain
Provisions of the GATT (entry into force: 15 April 1948), Special Protocol Modifying Article XIV of
the GATT (entry into force: 19 April 1948), and the Special Protocol Relating to Article XXIV of the
GATT (entry into force: 7 June 1948).
15 GATT/CP.2/34 of 3 September 1948. A consolidated version incorporating five of these protocols
was circulated in October 1948 in document GATT/CP/2, but some of the amendments were not
applicable to all contracting parties. See GATT/CP.4/6.

provisions; (2) Article XXIV:7(a), where contracting parties forming a customs
union or free trade area were required to promptly notify that fact, together with
a series of information that would enable contracting parties to make reports and
produce recommendations.
Once contracting parties realised that the Havana Charter would not enter
into force, a decision was taken on 24 October 1953 to review the GATT in
light of the experience gained since its provisional operation.16
The results of
such review were embodied in the 1955 ‘Protocol Amending Part II and Part
III of the GATT’17
which introduced a large number of amendments, including
new notification requirements relating to subsidies (Article XVI:5(b)) and state
trading enterprises (Article XVII:4(a)). A Part IV was subsequently added in
the 1960s to take account of the development dimension, which included in
Article XXXVII:2 the possibility to ‘report’ if commitments by developed
contracting parties under XXXVII:1 were not being given effect. Revised
reporting procedures were agreed in 1965, which were followed by guidelines
for notifications.18
2.2.2 The 1950s and 1960s
As experience in the application of the General Agreement grew, contracting
parties were faced with new situations where notifications were required. For
example, the Protocol of Provisional Application (PPA) that accompanied the
GATT allowed contracting parties to retain certain GATT inconsistent measures
pre-dating it. There was, however, no requirement to notify them.19
This lack
of transparency eventually led to disagreement in one case where a contracting
party claimed that certain restrictions on agricultural products were covered by
the PPA, but the other contracting parties disagreed.20
Contracting parties were
required thereafter, in 1960, to notify all the residual restrictions they maintained
and recognised were inconsistent with the GATT.21
This requirement eventually
evolved into a two-yearly notification of ‘residual restrictions’.22
Notifications required by Articles of the GATT also evolved through agreed
procedures and guidelines, some of which made use of questionnaires. For
example, contracting parties were required in 1958 to notify every year changes
16 G/70 of 12 November 1953. See also SR.8/20.
17 W.9/246 of 6 March 1955.
18 BISD 13S/79. See also BISD 24S/55.
19 UN document E/PC/T/214.Add.2.Rev.1. The Protocol of Provisional Application stated that
contracting parties would provisionally apply on and after 19 January 1948: ‘(a) Parts I and III of the
General Agreement on Tariffs and Trade, and (b) Part II of that Agreement to the fullest extent not
inconsistent with existing legislation’ (emphasis added).
20 Request by Federal Republic of Germany, BISD 8S/31.
21 BISD 9S/19-20 and 11S/206-210. TN.64/NTB/7 indicates that the following contracting parties
had notified import restrictions: Australia, Austria, Benelux, Canada, France, Germany, Italy, Japan,
Norway, Portugal, Federation of Rhodesia and Nyasaland, Sierra Leone, Sweden, United Kingdom,
and the United States.
22 NTM/W/6/Rev.3 of 8 August 1986.

to their rules relating to marks of origin (Article IX).23
Similarly, contracting
parties agreed in 1960 on procedures for the notification of restrictions for
balance-of-payment purposes,24
and questionnaires on state trading, subsidies
and anti-dumping.25
In 1962, new notification procedures invited contracting
parties to notify responses to questionnaires on subsidies and state trading every
three years.26
New notification requirements also resulted from informational needs
stemming from negotiations. For example, contracting parties agreed in 1958
to a ‘Programme for Expansion of International Trade’ which envisaged the
notification of agricultural policies.27
‘Committee II’was tasked with assembling
data relating to agricultural products, in collaboration with FAO, and determining
whether GATT rules had been inadequate in respect of those products.
Contracting parties were asked to submit data in connection with: (1) subsidies,
including all forms of income and price support; (2) mixing regulations; (3)
import restrictions; (4) state-trading operations; and (5) a description of national
agricultural policies.28
Contracting parties also explored the use of notifications as a tool for
trade promotion and assistance to exporters. Bearing in mind the publication
requirementsofArticleX:1,a‘GroupofExpertsintheFieldofTradeInformation’
examined in 1964 the means of establishing and operating an ‘international
trade information centre’.29
This group recommended the establishment of trade
information and trade promotion advisory services in the GATT framework,
which would rely on copies of laws, regulations, decisions, rulings and
agreements, of the kind described in Article X:1 that would be provided by
contracting parties to the Secretariat.30
Contracting parties were also expected to
indicate in their regular notifications whether such information could be made
available at large through the new services. Based on this recommendation,
contracting parties agreed to: (1) establish a ‘document centre’, (2) operate a
‘clearing house’ for trade information (today known as ‘inquiry points’), and (3)
publish a ‘Register of Sources of Trade Information’.31
However, a 1978 note by
the Secretariat notes that response to this recommendation was limited, except
in what concerned national tariffs.32
As negotiations began to deal with issues other than tariffs, contracting parties
found a new role for notifications: the identification of non-tariff barriers. The
Ministerial Decision launching the Kennedy Round provided for the large-scale
23 BISD 7S/30.
24 BISD 9S/18.
25 BISD 9S/184 (state trading) and 193 (subsidies). The anti-dumping questionnaire can be found in
L/1071 of 19 November 1959
26 BISD 11/58-59.
27 Decision of 17 November 1958, L/939.
28 See GATT BISD 7S/28, L/1192, and L/1461.
29 L/2181 of 13 March 1964.
30 Recommendation of 20 March 1964, L/2207.
31 L/2181 of 13 March 1964.
32 MTN/FR/W/17 of 1 August 1978.

negotiation of non-tariff barriers affecting trade,33
which were subsequently
based on ‘reverse-notifications’. Unlike the normal ‘self’ notifications, where
contracting parties provide information about their own trade policies, reverse
notifications seek to obtain information on barriers faced in other markets.34
Notifications received during the Kennedy Round led to a list of 18 measures to
be negotiated, which ranged from measures regulated by the General Agreement
to product-specific problems and measures maintained by a single contracting
party affecting many products.35
Reverse-notifications eventually became a
regular feature of GATT work on which a number of ‘inventories’ of non-tariff
barriers on agricultural and industrial products were developed and proved
crucial in the identification of problems and the development of new disciplines
in this area (Santana and Jackson, 2012).
2.2.3 The 1970s and 1980s
New notification requirements were introduced and refined in the 1970s as
a result of recommendations by various working parties and other regular
bodies of the GATT. For example, a notification on ‘border tax adjustments’
was introduced in 1970 as a result of a recommendation by a Working Party.36
In 1971, contracting parties introduced a questionnaire on import licensing37
that was then complemented with an annual notification of ‘changes’ to the
information contained in a consolidated compilation of the responses.38
A
similar decision required parties of the Joint Working Group to annually
notify changes in respect of the ‘residual restrictions’.39
Revised notification
procedures were agreed for the ad hoc examination of customs unions and free
trade areas under XXIV:7(a).40
In 1983, the Committee on Trade in Agriculture
adopted a work programme that invited contracting parties to supply information
on trade measures affecting these products based on an agreed format.41
The
Executive Secretariat also started playing a role by sending annual reminders of
the notification requirements.
An extensive change in the notifications landscape resulted from the conclusion
of the Tokyo Round, which embodied a number of plurilateral agreements. Some
of these aimed at tackling problems affecting specific sectors (e.g. bovine meat
and dairy products),42
and included provisions to allow for ‘information and
33 BISD 12S/47.
34 GATT/AIR/343.
35 TN.64/NTB/8 of 15 November 1964.
36 BISD 18S/108. See also L/3389.
37 L/3515 of 23 March 1971.
38 See report by the Committee on Trade in Industrial Products in L/3756, para. 76. See also SR.28/2
and L/4598.
39 This decision was reaffirmed in March 1980. See C/M/139 an L/5090.
40 BISD 19S/13.
41 BISD 30S/12. Information was compiled and circulated in the AG/FOR/ series.
42 Arrangement regarding Bovine Meat (BISD 26S/84) and International Dairy Arrangement (BISD
26S/89).

market monitoring’ based on statistics to be supplied by participants. Six other
plurilateral agreements, often referred to as the ‘Tokyo Round codes’, disciplined
the use of specific non-tariff measures.43
Three of these aimed at interpreting and
improving the disciplines that were already covered by the GATT (i.e. subsidies
and countervailing duties, anti-dumping, and customs valuation), while the other
three introduced new disciplines in previously unregulated areas (i.e. standards,
government procurement and import-licensing procedures).
All these new legal instruments incorporated specific provisions that required
the notification of certain aspects, some of which were intended to play a different
role than the other notifications thus far. For example, instead of requiring the
notification of measures that had already been adopted, the TBT Code required
the notification of ‘proposed’ technical regulation or standards not based on an
international standard, with a view to giving other contracting parties a say in
the formation of those measures. Like in the case of the agreements on bovine
meat and dairy, Article VI:9 of the Government Procurement Code required
participants to collect and provide to the Committee on an annual basis detailed
statistics on their purchases. Another new development that took place in the
context of these plurilateral agreements was the agreement of notification
‘formats’, such as the one agreed by the Committee on Technical Barriers to
Trade.44
2.2.4 The need for a systemic approach
Following the conclusion of the Tokyo Round in 1979, contracting parties felt
that a comprehensive approach was required to improve the notifications and
surveillance function of the GATT. An ‘Understanding regarding Notification,
Consultation, Dispute Settlement and Surveillance’was adopted on 28 November
1979, which invited contracting parties to notify ‘to the maximum extent possible’
the adoption of trade measures affecting the operation of the GATT. It also stated
that notification should preferably be done prior to implementation of new
measures or, if not possible, ‘promptly ex post facto’. In terms of the surveillance
function, contracting parties agreed to conduct a regular and systemic review of
developments in the trading system. To this end, they held discussions based on
a biannual publication by the Secretariat entitled ‘Developments in the Trading
System’, which was based not only on country notifications, but also on other
official and unofficial sources collected independently by the Secretariat.45
In 1980, the Director-General of the GATT put forward a proposal to improve
notifications, which listed all the requirements and a calendar in which they
were due.46
This proposal, which was adopted in March 1980, made a difference
between the notification requirements applicable to all contracting parties and
43 The Decision of 28 November 1979 stated that contracting parties’ rights and obligations of those
who were not parties to those agreements would not be affected. L/4905.
44 TBT/W/35 of 17 July 1981.
45 See, for example, L/5915, L/6205, L/6087, L/6205, L/6289, L/6366, L/6435, and L/6530.
46 C/111 of 14 March 1980.

those that were applicable only to some of them (e.g. because the obligation
was contained in protocols of accession, the enabling clause, the multifibre
agreement, waivers, the Tokyo Round Codes, etc). It is interesting to note that,
out of the approximately 21 notifications that were applicable to all contracting
parties, only five were periodic in nature (i.e. those relating to subsidies, state-
trading, marks of origin, licensing and Part IV of the GATT) and 16 were
required on an ad hoc basis (i.e. following the introduction of a measure or to
initiate specific procedures).47
In spite of these efforts, contracting parties considered that it was necessary
to keep improving and streamlining the notification procedures on the basis
of a series of reports on ‘notification and surveillance’.48
In 1984 and 1985,
the Secretariat circulated notes which identified what, in its opinion, were the
three main basic purposes of notifications: (1) to assist in the surveillance of
developments in the trading system; (2) to meet obligations under the plurilateral
agreements or to provide documentation under the GATT work programme, and
(3) to show whether individual contracting parties were fulfilling their obligations
under the GATT.49
Moreover, the Secretariat considered that the situation could
be improved if contracting parties agreed to: (1) provide for more regular and
complete notifications, which were considered the most authoritative source of
information; (2) make greater use of ‘reverse’ notifications; (3) authorise the
Secretariat to gather information from official and unofficial sources;50
and
(4) deal with measures in the ‘grey area’; i.e. measures that were disciplined
by the GATT, but lacked an express notification requirement. Although some
contracting parties endorsed these suggestions, others were concerned by the
additional burden they could represent to governments.
2.2.5 Discussions during the Uruguay Round
Based on the preparatory work that had been carried out in the previous years,51
a Negotiating Group on Functioning of the GATT was tasked in 1987 to deal
with the notifications. Discussions in the context of the preparatory committee
had concluded that: (1) There was a widely held view that there was scope for
improvement and streamlining of the notification requirements and procedures;
(2) efforts should be made to reduce and simplify the large number of existing
notifications so as to avoid unnecessary duplication and make it easier for
47 C/W/471 of 22 May 1985, Tables 1 and 2. A subsequent report by the Secretariat considered there
were only four period notifications applicable to all contracting parties. See PREP.COM(86)W/31/
Add.1.
48 See, for example, C/W/349, C/W/361, and C/W/385.
49 See C/W/446 and C/W/471.
50 Contracting parties decided in 1983 to include, in the documentation prepared by the Secretariat
for the reviews by the Council, not only actions notified by participants, but also trade measures or
agreements which had not been notified. See document C/M/169, 10 August 1983.
51 The Secretariat prepared an expanded listing of notification requirements to include areas not
previously covered by the Council’s reviews (MTN Codes, MFA, 1979 Understanding and 1982
Work Programme). See PREP.COM(86)/W/31/Add.1.

participants to comply with them; (3) a permanent body should be established
to carry out regular reviews of trade and trade-related policies of individual
participants, possibly to be called the Trade Policy Committee; and (4) there
was a need for further discussion on the question of improvement of the GATT
machinery for notifications which would need to be continued on a long-term
basis.52
During the Uruguay Round discussions, several proposals were tabled that
aimed at improving the notifications system by, inter alia, introducing a general
requirement to notify,53
developing common formats to be used, having a central
repository of notifications,54
and requiring the notification of trade measures
affecting the operation of the GATT before they were put into force. Some
contracting parties considered that improved notifications and surveillance
could allow for earlier consultations, which could in turn prevent disputes from
arising.55
In terms of the surveillance function, some considered that a separate
trade policy review mechanism could significantly improve the situation.
The launch of the Uruguay Round also led to new notifications. As part of
the mechanism for surveillance of the ‘standstill’ and ‘rollback’ commitments
that were part of the Punta del Este Declaration,56
a Surveillance Body was
established. Reverse notifications could be made to this body based on an
agreed format57
which would then be discussed. Approximately 30 alleged
violations of the standstill provision had been notified and reviewed by the
Surveillance Body as of 1993, together with more than 20 notifications relating
to the rollback commitments.58
Similarly, following a recommendation by the
Group on Quantitative Restrictions and Other Non-Tariff Measures, a study was
undertaken to assess the feasibility of developing an ‘Integrated Database’(IDB)
that would seek to establish a link between the work relating to QRs and NTMs
with tariff and import information that was notified by the contracting parties.59
The submission of tariff and import information was born out of the ‘tariff study’
that was carried out at the beginning of the 1970s to assess the results of the
Kennedy Round. This information was also used by the Secretariat during the
Tokyo Round to produce four volumes known as the ‘Basic Documentation for
the Tariff Study’.60
The Uruguay Round resulted in a massive change in the notification
requirements. Not only were many of the plurilateral Tokyo Round Codes made
part of the single undertaking, but also new agreements in areas such as services,
trade-related aspects of intellectual property, rules of origin, pre-shipments
inspection, etc became part of the WTO framework. As it will become evident in
52 Document PREP.COM(86)W/31.
53 MTN.GNG/NG14/8.
54 MTN.GNG/NG14/W/39.
55 MTN.GNG/NG13/W/14/Rev.2 of 22 June 1988, p. 6.
56 See MTN.GNG/NG14/W/3, MTN.TNC/2, and MTN.GNG/5.
57 MTN.SB/SN/.
58 MTN.SB/W/3/Rev.14 of 16 August 1993.
59 L/6073 of 27 October 1986.
60 IDB/URM/1 of 19 April 1994.

the following section, the considerable diversity of the notification requirements
under the WTO is a legacy of the wide range of objectives and methodologies
that were slowly developed in the GATT.
2.3 The nature and scope of WTO notifications
The results of the Uruguay Round included a Ministerial Decision on
Notification Procedures which contains three sections: the first affirms
Members’ commitments to their obligations regarding publication and
notification; the second one establishes a central registry of notifications within
the WTO Secretariat; and the third provides that ‘the Council for Trade in
Goods will undertake a review of notification obligations and procedures under
the Agreements in Annex 1A of the WTO Agreement’.61
A Working Group
on Notification Obligations and Procedures was established in 1995 with the
mandate to make recommendations and undertake a ‘thorough review of all
existing notification obligations of Members established under the Agreements
in Annex 1A of the WTO Agreement, with a view to simplifying, standardizing
and consolidating these obligations to the greatest extent practicable, as well
as to improving compliance with these obligations, bearing in mind the overall
objective of improving the transparency of the trade policies of Members and
the effectiveness of surveillance arrangements established to this end, and also
bearing in mind the possible need of some developing country Members for
assistance in meeting their notification obligations’.62
The WTO framework contains more than 200 different legal notification
requirements, the large majority of which relate to NTMs.63
This section
examines the elements that differentiate the current notification requirements in
order to show their diversity. It then compares the coverage of notifications with
the universe of NTMs covered by the most widely used classification of NTMs.
2.3.1 Self versus reverse notifications
While a vast majority of requirements oblige Members to provide information on
their own policies, some are ‘reverse’notifications, which allow them to identify
measures imposed by third countries. Article 5.5 of the Agreement on Import
Licensing Procedures, for example, invites ‘interested’ Members who consider
that another Member has not notified the introduction of a licensing procedure to
bring the matter to the attention of such a Member. If a notification is not made
promptly thereafter by the Member maintaining the measure, the complaining
61 G/NOP/W/1.
62 G/C/M/1.
63 Most of the notifications have been attributed a document symbol, which means that they are
circulated. One example of notifications that are not circulated is those on quantitative restrictions,
presumably for confidentiality reasons. Other notifications relate to Agreements that have been
terminated such as the Agreement on Textiles and Clothing, the plurilateral International Bovine
Meat Agreement or the International Dairy Agreement.

Member may then notify the licensing procedure itself. Another example is
Article 12:8 of the Agreement on Safeguards which states that any Member may
notify measures ‘that have not been notified by other Members that are required
by [the Agreement on Safeguards] to make such notifications’. These specific
reverse notification possibilities are complemented by the ‘Decision on Reverse
Notifications of Non-Tariff Measures’ which opens the possibility for Members
to make notifications of non-tariff measures maintained by other Members ‘in
so far as such measures are neither subject to any existing WTO notification
obligations nor to any other reverse notification possibilities under the WTO
Agreement’.64
2.3.2 Periodicity
Notifications also differ from each other with regard to their periodicity (see
Figure 2.1). Most of those covering laws and regulations are one-off requirements
with a separate obligation to notify changes thereafter. The notifications that
provide information on the measures themselves typically take two different
forms: they are either ad hoc or (semi-) annual. Depending on the objective of
the notification (see below), measures need to be notified either early on at the
project stage, or promptly after their introduction. In the latter case, they often
also need to be notified in a semi-annual or annual summary report. The case
of anti-dumping is illustrative of this sort of multiple notification requirements.
Article 16.4 of the Agreement on implementation of Article VI of GATT 1994
requires that:
Members shall report without delay to the Committee all preliminary or
final anti-dumping actions taken. Such reports shall be available in the
Secretariat for inspection by other Members. Members shall also submit,
on a semi-annual basis, reports of any anti-dumping actions taken within
the preceding six months. The semi-annual reports shall be submitted on an
agreed standard form.
2.3.3 Product-specific information
About half of the notification requirements cover NTMs that typically apply
to specific products. In those cases, notification templates generally require
Members to indicate which products are covered. The other half relates to
measures (e.g. laws and regulations) that affect, or could potentially affect, all
products (e.g. pre-shipment inspection or customs valuation).
64 Decision of 1 November 1995, G/L/60.

Figure 2.1 Types of notifications
NoƟﬁcaƟons Self
Once
Periodic
Ad hoc
Ad hoc
Reverse
(i.e. about
other
members)
Whose
measures?
Periodicity
2.3.4 NTMs covered by the notifications
Current notification requirements cover 24 broadly defined subject areas ranging
from agriculture to preferential trade agreements.65
Some of those subject areas
would not enter the definition of NTMs. There are six notification requirements
that enter in this category, including those on regional trade agreements under
GATTArticle XXIV, three on preferential trade arrangements under the Enabling
clause, four on tariff renegotiations under GATT Article XXVIII, and tariff and
trade data for the Integrated Database (IDB). The subject areas sometimes cover
more than one specific measure. Agriculture, for example, covers five NTMs:
the special safeguard, tariff quotas, export restrictions, domestic support and
export subsidies. Also, in many cases a single measure is covered by several
different notification requirements. For instance, there are five notifications on
anti-dumping: two which relate to the introduction of anti-dumping measures
(one on an ad hoc basis, and the other one in the form of a semi-annual summary)
and three which relate to anti-dumping legislation and regulations, as well as
authorities and procedures.
As mentioned, reverse notification possibilities cover, in principle, the whole
spectrum of NTMs. However, the ‘self’notification obligations only cover a large,
65 The 24 subject areas are: agriculture, anti-dumping, balance-of-payments, customs valuation,
enabling clause, government procurement, import licensing, TRIPS, import restriction (Art. XVIII),
TRIMS, maritime transport, reverse notifications, pre-shipment inspection, quantitative restrictions,
PTAs, rules of origin, safeguards, SPS, state trading, tariffs, subsidies and countervailing duties,
technical barriers to trade, textiles and clothing, services.

non-exhaustive, share of the NTMs universe. This can be seen by comparing the
list of notifications with the so-called ‘Multi-Agency’ classification (MAC) of
NTMs (see TableA2.1). The MAC comprises 16 broad categories of measures, of
which only 3 do not seem to be covered at all by WTO notification requirements.
Those are: finance measures, distribution restrictions and restrictions on post-
sales services. All the other categories are at least partly covered (i.e. a number
of subcategories are covered while others are not). This would be the case, for
example, of price control measures, of which 5 subcategories (administrative
pricing, voluntary export price restraints, anti-dumping measures, countervailing
measures, and safeguard duties) are covered by notifications while two (variable
charges and seasonal duties) are not.66
It should, in addition, be noted that where notification requirements broadly
match NTM categories, they do not necessarily cover the measures that could
be classified therein. For example, in the case of sanitary and phytosanitary
measures, Article 7 and Annex B of the SPS Agreement require governments
to notify new SPS measures which are not based on international standards and
have a significant effect on trade of other Members, and to notify those at an
early stage, that is, when amendments can still be introduced. Measures that
were in place before the entry into force of the SPS Agreement need not be
notified, nor is there an obligation to notify the final measures when they entered
into force. This means that some of the measures in place were not notified and
that some of those notified may have been amended before being implemented
or even not implemented at all.
2.4 The quantity and quality of notifications
This section examines the extent to which notifications requirements are observed
by WTO Members in practice. More precisely, it examines, on the one hand,
whether Members comply with their commitment to notify certain information
to the WTO Secretariat and, on the other hand, whether the information received
through notifications could be used to measure the incidence of NTMs in
international trade.67
2.4.1 The metrics of compliance
Answeringthecompliancequestionrequiresadetailedanalysisofnotificationsby
subject area, which has not been done in the past, or at least not systematically or
regularly. Although it is tempting to assume that compliance is a straightforward
66 Seasonal duties may be notified as tariffs.
67 This approach is not new. It was proposed by India in a communication for the General Council
in 2009 (WT/GC/W/605 of 3 July 2009) which suggested ‘the setting up of a project to enhance
the Integrated Database to include in an appropriate format non-tariff data, based on the current
notification obligations under WTO Agreements. … optimising the design and structure of present
notification systems; enhancing co-operation with related multilateral agencies; and providing
technical assistance to developing countries, in particular the LDCs.’

exercise in the case of one-off and periodic notification requirements, a closer
inspection reveals that this is not always the case. The difficulties faced in this
analysis relate to the nature of the requirements where, even if notifications
are made, it is not always possible to establish whether all measures had been
notified. Where Members are required to notify a periodic summary, one can
assume that it encompasses all relevant measures and, therefore, that it covers all
the measures falling under the requirement. For Members who do not submit a
summary report, however, the question remains as to whether they did not have
any relevant NTMs for the reporting period or if they simply did not comply
with the notification requirement. The compliance problem is more pronounced
in the case of ad hoc notification requirements which seek the reporting of new
measures, as there is no real benchmark that could be used to determine if the
Member did not fulfil its commitment and one has to fully rely on the reporting
performance of Members.
Table A2.2 in the Appendix provides an overview of the performance of
WTO Members in complying with their notification obligations. It is based
on a selection of requirements which can be linked to the NTM classification
and provide the most comprehensive information on NTMs in their respective
subject areas. The measures/requirements not covered by the table relate, in
general terms, to the notification of laws and regulations, contact points and
changes thereof. In two cases, Pre-shipment inspection and Rules of Origin, only
one time general notifications of laws and regulations are required.
There are many cases where Members have no obligation to notify if they
do not have a certain NTM in place (i.e. there is no obligation to report the
non-application of measures). This concerns all the ad hoc reporting, as well
as TRQs for agricultural products and anti-dumping and countervailing duties
notifications. In all other cases Members are required to notify both if they have
introduced and if they do not have in place the measure concerned. There are also
provisions requiring Members to make a one-off notification that no measures
exist until such time when they may be introduced (e.g. anti-dumping).
Even if one focuses on the selective list of notification requirements, the
compliance record does not look very strong, except for the one-off notifications
for which by now, 15 years after the establishment of the WTO, a fairly high
percentage of Members have provided information.68
In the case of anti-
dumping and countervailing duties one can assume that a fairly high percentage
of measures are captured because these measures are clearly not used by all
Members and proceedings in the WTO need to be undertaken to give them a
justification under WTO rules. To the extent this is accurate, one could interpret
that the lack of reporting is, in most cases, reflecting the non-application of
such measures. Other measures, in particular SPS, TBT and import-licensing
measures, are most likely more widespread than what notifications show.
68 WTO now counts 153 Members but for notifications the EU and its Member States count as one just
as Switzerland and Liechtenstein do. Full compliance therefore equals 126 notifications.

The issue of compliance was taken up in February 2009 by the Chair of the
General Council in a letter to all relevant WTO committees. The initiative was
taken up again in the Trade Policy Review Body meeting in January 2010 on the
monitoring of the recent trade developments,69
where delegations stressed the
importance of improving the record of notifications to various WTO Councils
and Committees. It was suggested that the issue of enhanced notification should
remain a focus in each of the relevant WTO bodies.70
One last point that needs to be taken into account when discussing compliance
and its measurement relates to situations where a principal NTM is implemented
through another NTM, both of which have to be notified at least once to the
WTO. For example, TRQs for agricultural products, which need to be notified
to the Committee on Agriculture, are often administered through some form
of import licensing procedure, which needs to be notified to the Committee
on Import Licensing and the QR notification. Likewise, import licensing is
often a side measure associated with SPS and TBT measures. This means that
import licensing may simply be a ‘secondary’ measure supporting the principal
measure. If such secondary measures are reported separately, as is foreseen in
the notification requirements, this creates problems of double counting which
would need to be addressed when some kind of aggregation of measures across
different subject areas is undertaken.
2.4.2 Measuring NTMs through WTO notifications
This subsection focuses on the notification formats which define to a large extent
the kind of information that is expected from Members. It should, however, be
stressed that the diverse formats of the different notification requirements make
it difficult to assess their usefulness for the measurement of NTMs.
A first question that needs to be answered in this respect is what shall be
defined as the ‘unit of measurement’ for the purposes of measuring NTMs. Only
once such a unit of measurement and associated criteria are defined can one
assess to what extent the existing notification formats satisfy them and could,
therefore, be used for a quantification of the incidence of NTMs.
2.4.3 Information needed to measure NTMs
Several key information requirements are needed in order to identify an NTM
unit of measurement that would allow us to assess the incidence of NTMs in
international trade, including:71
• The subject area of the measure.
69 WT/TPR/OV/M/7 of 3 March 2010.
70 The decision on improved tariff data collection and dissemination referred to in the introduction was
one of the outcomes of this initiative.
71 The ‘severity’ of a measure is not addressed, neither is the specific way in which it is implemented,
which in itself can have an effect on trade flows.

• The Member(s) imposing a measure.
• The Member(s) (potentially/actually) affected by the measure.
• The product(s) covered by the measure.
• The time period in which the measure is applied, potentially affecting
trade flows.
The most easily identifiable pieces of information are the Member imposing
the measure and the subject area. Unless the information relates to a ‘reverse’
notification where the imposing Member is unclear, there is never a doubt in
‘self’ notifications as to who is imposing the measure because it is the Member
reporting it. The degree of precision for the other three information requirements
can vary considerably.
The product coverage of the NTM is also a key piece of information, because
it determines whether particular tariff lines are affected by it. Most NTM
notifications provide some kind of description of the products covered by the
notification and the remaining ones can normally be considered to be applicable
to all products. While textual descriptions are often quite specific and provide
useful information on the product coverage, HS codes are of enormous analytical
value for the screening and analysis of large numbers of notifications, as well
as for estimating their impact on imports (e.g. the trade effect of the measure).
Unfortunately, notification formats for WTO NTM notifications have rarely
included the provision of HS codes as an obligatory element of information to
be provided.
The time dimension is also important. In order to assess the incidence of
NTMs over time it is crucial to know when measures entered into force (i.e.
a ‘start’ date) and when measures are discontinued (i.e. a ‘stop’ date). Ideally
one would want to track these measures with exact dates provided by Members
themselves, which thus far is possible only in the case of notifications on anti-
dumping and countervailing duties. In the absence of exact dates, it is probably
sufficient to know whether a specific NTM is in force at a given point in time.
The assumption could then be that this measure was introduced since when the
last notification was made.
2.4.4 Do existing notification requirements provide this information?
The precision with which information is provided varies considerably across
notifications and even when enough precision is required in the format, the
information is not notified by the Members. This means that it is often necessary
to interpret and to make assumptions if one wants to process, summarise and
present the information.
Anti-dumping provides an example of a very precise notification procedure
which satisfies most of these elements. Anti-dumping measures are reported

by Members in semi-annual reports with very specific information as to which
Member is affected by the measure, which products are concerned (defined
textually and at HS 6-digit level) and the exact dates when the measure was
initiated, put in force and at some point terminated, as appropriate.72
The opposite is the case of the TBT notification requirements where the
information elements are of little use for measuring NTMs. For example, the
affected Members are usually not specified, because measures are meant to apply
to all imports irrespective of their origin. The products concerned are sometimes
described vaguely (e.g. ‘textile products’), which may be at least partially
explained by the nature of the measures at issue (e.g. technical regulations,
standards, conformity assessment procedures which are often defined by
regulators who did not take into account the HS coding system in drafting the
measures). Last, but not least, the time dimension is for the most part left open.
Because these notifications submit draft proposals of future technical regulations
or conformity assessment procedures, the final published measure need not be
notified, thus making it unclear whether, and if so when and how, such draft
measure entered into force. For the measures that are put into force, there is
no requirement to notify when they were terminated. Moreover, in cases where
a ‘revised’ measure replaces fully or partially an older measure, the change is
often not traceable, which could potentially lead to double counting.
2.4.5 Can the situation be improved?
It should be evident from the description above that current TBT notifications
do not meet the basic criteria needed for measuring the incidence of NTMs.
The situation is similar in the case of SPS notifications, which becomes a
major issue if one takes into account that both of these account for the bulk of
WTO notifications. The question then arises as to whether the situation can be
improved in any way.
Changing the notification formats would, evidently, be the most straightforward
approach to redress this problem. This solution, however, would require that the
formats are discussed and approved by Members in each of the Committees
concerned, which would take time and require consensus.73
Moreover, it is
possible that the fulfilment of the main objective of some of those notifications
does not require the provision of some of the five basic elements described above,
which may translate in incomplete notifications or a delay in their preparation.
The question then arises as to whether the measurability of such ‘imperfect’
notifications could be improved by complementing and categorising the
officially submitted information of today. We believe that it could partially be
done. For example, the information gaps could be filled by making assumptions
72 The relatively precise recording of HS coding information, ‘for reference purposes only’, was only
introduced in 2008 (see G/ADP/1/Rev.1 of 3 November 2008). The original reporting format did not
mention the provision of HS codes.
73 It is also unlikely that a change to the format would be made retro-active in the sense that Members
would not be required to provide information covered by past notifications.

on the likely dates for the implementation of a measure and/or by interpreting
product information for product detail in HS codes. Another approach would be
to complement notifications with other sources of information, as for example
is done in the case of Trade Policy Reviews or the Monitoring exercises which
were launched in 2009.
Table A2.3 in the Appendix tries to depict the extent to which current
notification formats provide the five elements described above. The two main
areas to look at are the dates and the product coverage as the information on the
affected countries is often not relevant because most measures are applied on an
MFN basis.
2.5 Notifications as a source of information on NTMs
2.5.1 Challenges facing current WTO notifications
As explained in Section 2.2, the nature and role of notifications in the multilateral
trading system have evolved since the early days. In a 1984 note, the GATT
Secretariat listed what it considered to be the three main basic purposes of
notifications: (1) surveillance of developments in the trading system; (2)
implementation and operation of the agreements by meeting obligations under
the plurilateral agreements or by providing documentation under the GATT
work programme, and (3) enforcement of the commitments by showing whether
individual contracting parties were fulfilling their obligations under the GATT.
This, obviously, does not mean that new purposes have not developed over time.
Under the WTO, notification requirements are part of the system put in
place to allow Members and the organisation to fulfil its various functions.
Notifications, therefore, continue to help with the implementation and operation
of the agreements, as well as the enforcement of commitments. More precisely,
‘self’ notifications have typically been assigned one or more of the following
roles:74
• Review the implementation and operation of commitments (e.g.
agricultural tariff quotas, anti-dumping).
• Enable other Members to evaluate the trade effects and to understand
the operation of a measure (e.g. subsidy programmes).
• Trigger the initiation of certain procedures (e.g. BOP).
• Trigger negotiations or consultations with Members affected by a
measure (e.g. modification of schedules).
74 Since the WTO makes the immense majority of this information available to the general public
(Decision of 14 May 2002, ‘Procedures for the circulation and derestriction of WTO documents’,
WT/L/452), one could argue that notifications also have the role of providing relevant information to
traders thereby assisting in trade promotion.

• Provide transparency on the laws, regulations and other measures that
could affect trade, including amendments therein (e.g. PSI, import
licensing, QRs).
Certain notifications, however, also play other, different roles such as to:
• Allow other Members to participate in the formation of new regulations
(e.g. SPS, TBT).
The extent to which notifications play the role they have been assigned in a
satisfactory manner depends, among others, on compliance and quality of
notifications made by Members.As discussed in Section 2.4, compliance appears
to be generally low, except where Members have an own interest in complying
with the requirement to trigger a procedure (e.g. modification of schedules).Also
in a number of cases, the quality of notifications varies substantially, notably in
terms of completeness and/or timeliness. Where compliance and/or quality are
low, notifications cannot, by definition, play their roles in a satisfactory manner
and may distort the way in which the application of a measure is perceived. The
nature and extent of the problem may depend on the objective of the notification.
When a Member fails to notify an SPS or TBT measure or notifies it too late,
other Members are not given an opportunity to comment and influence the
drafting of the measure. In other cases, such as subsidies and other beyond-
the-border measures, notifications could significantly enhance transparency. The
effects of a subsidy that is not notified may not be easy to identify even if the
subsidy affects trade.
Low compliance may not only create a disincentive to notify, but also distort the
perceptionaboutaparticularNTM.Itisreasonabletoassumethatlowcompliance
with notification requirements may end up encouraging non-compliance (e.g.
why bother if nobody else does it?). Where there is no obligation to notify the
absence of a measure, a ‘no notification’ may be interpreted as ‘no measure’, in
which case the Members that notify may look like they are the only ones using it.
While the ideal would clearly be to have in place a system where all Members
fully comply with all notification requirements, the long history of discussions
on approaches to improve compliance strongly suggests that this is a difficult
endeavour. As mentioned above, low compliance was already seen as a problem
in the GATT (see Section 2.2).A note by the Executive Secretariat of the GATT on
the improvement of notifications procedures suggests that: ‘Successive reviews
by the Council in its special meetings have shown that some governments take
a much stricter view than others of their obligations to notify.’75
Over the years,
many of the committees have, sometimes repeatedly, attempted to raise the level
of compliance.
Part of the reason for the low compliance appears to be difficulties faced by
Members in making their notifications. In the case of subsidies, for example,
75 C/W/466 of 28 September 1984, para 6.

where the level of compliance is generally very low, a note from the Chairman
of the SCM Committee summarising the replies to a questionnaire observed that
the difficulties faced by Members were either logistical or of a more substantive
nature.76
Members identified two main types of logistical problems: resource
constraints (staff and information) and lack of internal government coordination.
On the substantive side, two main issues were mentioned: doubts concerning
the key subsidy concepts involved and difficulties in interpreting and organising
information and data to fit within the existing notification format. To the extent
this assessment is accurate, the best way to address these difficulties would be
through a combination of measures aimed at streamlining information-gathering
and compilation processes, as well as at simplifying and clarifying notification
formats and making them more flexible. Technical assistance may be necessary
for low-income countries.
Logistical and substantive difficulties, however, may not be the only reason
behind the lack of compliance. In some cases, the problem may be that
governments have no strong incentive to comply. In the absence of any sanction
for non-compliance, the only enforcement mechanisms are either self-interest
or peer-pressure. Members clearly do not consider that notifications are always
in their self-interest and tend to perceive them rather as a burden. They may
not notify for fear that the information could be held against them. Because
reverse notifications have practically disappeared in the WTO, the threat of
peer pressure may have also weakened. Moreover, as mentioned before, where
notification compliance is low, peer pressure is largely limited to raising specific
trade concerns at the committees, and some governments don’t see why they
would notify when the others don’t. Reinforcing the incentives to notify should
probably involve a combination of sticks and carrots.
2.5.2 WTO notifications could do more
Beyond the question of what could be done to improve the efficiency with
which notifications fulfil the role they are meant to play, it is interesting to
consider whether they could also play other roles. For example, the WTO
system could use information on NTMs for a number of other reasons than
those discussed so far. First, information on NTBs faced by exporters is often
needed for negotiation purposes. As mentioned in Section 2.2, under the GATT
and more recently as part of the DDA negotiations, information on NTMs was
collected specifically for negotiation purposes. For example, paragraph 16 of
the Doha Ministerial Declaration mandated Members to establish negotiations
to reduce or as appropriate eliminate non-tariff barriers (NTMs), in particular
on products of export interest to developing countries. Like in the previous
rounds of negotiations, the Negotiating Group on MarketAccess worked up until
December 2005 on a process of identification, categorisation and examination
of NTBs based on reverse notifications, with a view to ultimately negotiating
76 See G/SCM/W/546/Rev1 of 16 April 2010 and G/SCM/W/505 of 27 March 2002.

appropriate solutions for them. These reverse notifications were compiled by the
Secretariat in an inventory in which 18 Members notified the measures affecting
their exporters. Given that this first set of notifications had a number of problems,
including lack of specificity, Members requested the Secretariat to prepare a
revised format to allow for the standardisation and classification of submissions
received. Twenty-one Members submitted or resubmitted notifications under the
revised format.77
There are several reasons why regular notifications do not provide the sort
of information on NTMs that Members need for negotiations purposes. First,
self-notifications which have come to dominate the universe of notifications, are
not well suited to provide information on NTMs acting as barriers. One would
expect governments not to be very keen to provide information on measures
that they impose with the effect of restricting imports. Reverse notifications
are clearly better suited for this purpose. As discussed in Section 2, however,
reverse notifications have progressively lost their importance under the WTO.
And while WTO Agreements still include a number of provisions allowing for
reverse notifications, they remain largely unused. One reason for this may be
that at least part of their role has been taken over by the ‘special trade concerns’
mechanism used in certain WTO bodies (e.g. the SPS and TBT Committees),
which allows exporting Members to raise issues faced on foreign markets.
Second, notifications typically provide information on NTMs that are already
regulated while negotiators may be more interested in information on measures
that are not yet regulated.
AsecondpurposeforwhichMembersneedNTMinformationisforsurveillance
and monitoring. A complete description of a Member’s trade and trade-related
policies, such as the one periodically performed by the Trade Policy Review
Mechanism, would include information on a large number of NTMs. Similarly,
to document trends in trade policies, it is important to extend beyond tariffs
and other well-documented measures. Existing notifications at least partly serve
a surveillance purpose. Notifications of contingent measures for instance can
be seen as part of a surveillance mechanism. The recent initiatives aimed at
monitoring trade and trade-related policy developments in the context of the
global financial and economic crises, however, suggest that notifications are
not sufficient for monitoring and surveillance purposes. This is mostly because
notifications are often made with a relatively long lag, but also because of the
limited (or lack of) compliance affecting most of them.
Thirdly, information on NTMs is needed for analytical and research purposes.
The quality and reliability of empirical research critically depends on the quality
of data used. For example, the coverage should ideally be comprehensive. Low
or partial coverage of certain measures (e.g. SPS or TBT notifications which do
not cover pre-WTO or final measures) would be an issue if one is interested in
quantifying the effect of NTMs or in making international comparisons based
77 TN/MA/W/46 and addenda.

on that information. Confidentiality and accessibility of the data could also raise
problems.
All this means that, unless the scope of many notifications is expanded,
and compliance and quality are enhanced, WTO notifications will remain an
insufficient source of information on NTMs, which means that the trade policy
community will have to rely on other sources of information in the short and
medium term. As mentioned above, notifications serve certain specific purposes
and their information requirements are designed to serve them. They would
serve those even better if compliance and quality were improved. Compliance
with notification requirements and improved quality of the notifications may
also have positive spill-over effects, raising domestic awareness of the existence
of measures that restrict home firm competitiveness.78
Notifications, however,
are often an inappropriate source of information for negotiators simply because
they were not meant to serve that role. They are also mostly inappropriate when
it comes to monitoring and surveillance. Similarly, their contribution to the
transparency of the trading system remains limited. In this particular case, not
only compliance but also the scope of notifications and the formats which specify
exactly what Members need to notify would need to be revised if notifications
are to become the main source of information on NTMs. Such changes would
also help make the notifications data more relevant for researchers.
2.5.3 Tapping other sources of information
A number of other sources of information on NTMs exist which could be tapped
with a view to improving transparency of trade and trade-related policies and
fulfilling some of the objectives mentioned above. Some of them are internal
WTO sources. These would include, for example, the records of discussions on
‘specific trade concerns’ in certain committees. At the end of 2010, more than
270 such concerns had been discussed in the TBT committee79
and 290 in the
SPS committee.80
Another major source of information on NTMs would be the reports prepared
by the Secretariat and Members under the Trade Policy Review Mechanism.
The Secretariat’s Trade Policy Reports (TPRs) draw on notifications and
other information available in the Secretariat but also, critically, on additional
information provided by the Member under review. TPRs are prepared every
two years for the four Members with the largest share of world trade; every four
years for the next 16 Members by order of the size of their share of world trade
and every six years for the remaining Members. They provide a comprehensive
descriptionofWTOMembers’tradepolicies.Theymakeasignificantcontribution
to the transparency of the world trading system and may help identify issues for
negotiations. However, because TPRs take the form of text and are not accessible
78 It should be noted that other transparency instruments, such as trade policy reviews, which also cover
procedural obstacles, probably play this role more efficiently.
79 G/TBT/GEN/74/Rev.6
80 G/SPS/GEN/204/Rev.10 and addenda.

in a searchable database, they are not well suited for international comparisons
or for research and in particular for quantification. Moreover, the information
produced in those reports is not compiled systematically, which means that
processing information across countries is usually a burdensome endeavour. The
four- to six-years periodicity that applies to most Members could also limit the
use of TPRs for time series and comparative work (i.e. because the information
for any two given Members is most likely for different years).
Other WTO sources of information include the reports prepared by the WTO
Secretariat in the context of the global financial and economic crises. There are
two such reports. The first is a report on trade and trade-related developments
prepared periodically for the Trade Policy Review Body.81
The second series
of reports is prepared by the WTO Secretariat together with the secretariats of
the OECD and UNCTAD following a request by the G20 to monitor trade and
investment measures.82
As already mentioned, the monitoring reports do not use
information from notifications. Instead they use information collected through
questionnaires sent to selected Members, informal reverse notifications and the
press. The information collected is then submitted to the members for approval.
There are several sources outside of the WTO that could provide information
on NTMs and, even if much remains to be done, there has been considerable
progress in recent years both at the regional and at the multilateral level. In
particular, as part of a project initiated by UNCTAD in 2006, a number of
organisations have undertaken efforts to refresh the existing classification of
NTMs and to significantly upgrade data collection.83
An important contribution
of this joint project was to update the old UNCTAD Coding System of Trade
Control Measures which has been used by UNCTAD in collaboration with a
number of regional organisations to collect and classify NTMs since 1994. In
November 2009, a new classification was issued which includes new NTM
categories and sub-categories.84
From a WTO perspective, the classification is
important because it determines the WTO relevance of information collected
by other agencies. Conversely, to the extent that other agencies want to use
information collected through notifications, it will be easier for them to do so
if there is some correspondence between the classification and definitions used
at the WTO. The WTO Secretariat has been actively involved in the process
81 It is a preparatory contribution to the report by the Director-General that is called for in Paragraph
G of the TPRM mandate and that aims to assist the TPRB to undertake an annual overview of
developments in the international trading environment which are having an impact on the multilateral
trading system. See WT/TPR/OV/W/1 to WT/TPR/OV/W/3 and WT/TPR/OV/1 to 13.
82 See, for example, the Report on G20 trade and investment measures (May 2010 to October 2010)
dated 4 November 2010.
83 See UNCTAD (2010) and other contributions in this volume.
84 The November 2009 classification includes new sub-categories for ‘sanitary and phyto-sanitary
measures’ and ‘technical barriers to trade’ as well as a number of new categories including ‘export
measures’, ‘trade related investment measures’, ‘distribution restrictions’, ‘restrictions on post-sales
services’, ‘subsidies’, measures related to intellectual property rights’ or ‘rules of origin’.

of further redrafting and elaborating the classification with a view to having a
revised version.85
This new version was issued in 2012.
Several organizations have also launched NTM data collection projects.
UNCTAD and the International Trade Centre conducted a pilot project to collect
information on NTMs in seven developing countries. These pilot projects aimed
at testing the new classification as well as the data collection methodology based
on firm surveys. Since the end of the pilot project in 2009, ITC carried out large-
scale company level surveys on NTMs in more than a dozen developing and
least-developed countries on all continents. In parallel, UNCTAD and ITC have
also identified data sources and collected official data for some of the pilot project
countries as well as for leading importing countries. It should be stressed that
survey data is different in nature from data collected from official sources. Data
based on complaints and concerns also cover so-called ‘procedural obstacles’,
that is, obstacles arising from arbitrary, inefficient, costly, time-consuming, non-
transparent or corrupt administration of policies. The problem is, however, that
surveys are often not specific about the measures involved, are not systematic
and may not always be reliable. As evidenced by the papers in this volume,
the World Bank has also collected information on NTMs as part of a domestic
regulatory-improvement agenda aimed at boosting home-firm competitiveness
through improved access to inputs.86
2.6 Conclusion
As discussed, transparency in trade and trade-related policies is essential to the
smooth operation of the multilateral trading system and notifications contribute
to such transparency. Both compliance and the quality of information furnished
by Members in the current notifications scheme are not stellar. This means that
notifications may not always serve the purpose they have been designed to serve
in a satisfactory way. It also means that, at least in the short run, the WTO is
unlikely to provide a repository of information on all NTMs, that is, to be a
comprehensive if not exhaustive source of information on NTMs. This poses a
problem for the trade policy community, traders and researchers who are facing
two possibly complementary options. First, and this may be the only option in
the short term, other sources of information will need to be used to complement
notifications, which could be either internal or external to the WTO. Secondly,
Members could decide to expand and enhance the notification mechanisms. This
would involve renewed efforts to improve compliance, as well as changes in the
notification requirements and formats.
One issue that arises with the second option is of a systemic nature. Modifying
notification requirements and improving compliance with a view to improving
measurability and to better serving a transparency purpose would go a long
85 For example, some of these new categories are only defined in general terms to allow the collection
of information from the private sector through surveys. Before they can be used to collect official
information, they will require further elaboration.
86 Note that access to some of the NTM information collected by the Bank may be restricted.

way towards addressing the various needs for NTM information. The lack of
incentives associated with self-notifications schemes, however, may undermine
efforts aimed at helping Members address non-tariff barriers that affect their
exporters. To serve this objective, the WTO may need to resort to other schemes,
such as: (1) reviving the reverse notifications, (2) expanding the specific trade
concerns discussions (which seem to have largely replaced reverse notifications,
or (3) systematically tapping other sources, which do not suffer from the self-
notifications syndrome. Again, these sources could be internal to the WTO
(e.g. the reports produced in the context of the TPR mechanism and other
special reports such as the ones prepared in the context of the global financial
and economic crises), or external. In this perspective, ongoing efforts by other
agencies using business surveys to identify obstacles faced by exporters could
play a useful role.
Beyond the compliance and quality issue, WTO notification requirements
have traditionally served various purposes and the nature of the information they
require Members to notify depends on the purpose they serve. The SPS and
TBT notification requirements, for example, aim at allowing other Members
to influence the regulations that Members plan to adopt. They do not require
Members to notify regulations that pre-existed the entry into force of the SPS
and TBT Agreements, nor the new measures in their final form. They, thus,
only provide a partial coverage of these measures which may be insufficient
from a transparency perspective. In the short run, the only way to fill those gaps
is through the use of other sources of information that could complement the
notifications. Other WTO sources may help fill some gaps but, in many cases,
external sources will have to be used. From this point of view, the existing
dialogue between the WTO and other agencies involved in NTM data collection
efforts is crucial to maximise the complementarity between WTO and other
sources of information on NTMs, as well as to avoid duplication of efforts. In
this perspective, the WTO’s new ‘Integrated Trade Intelligence Portal’ (I-TIP)
will significantly facilitate interagency information sharing by providing a new
unified point of access to all WTO NTM notifications. In the longer run, it
may be possible to modify the format of certain WTO notifications and require
additional or more specific information, such as for instance date of entry into
force of SPS and TBT measures. In other cases, however, extending the scope
of notifications beyond a certain limit may not obtain the necessary consensus
among the WTO membership.
References
Bhagwati, J N (1988), Protectionism, Cambridge, MA, MIT Press.
Santana, R and Jackson, L-A (2012), ‘Identifying Non-Tariff Barriers: Evolution of
Multilateral Instruments and Evidence from the Disputes’, World Trade Review,
forthcoming.

UNCTAD (2010), ‘Non-tariff Measures: Evidence from Selected Developing
Countries and Future Research Agenda’, Developing Countries in International
Trade Studies, Geneva: UNCTAD.
Wolfe, R (2003), ‘Regulatory Transparency, Developing Countries and the WTO’,
World Trade Review 2(2), pp 157–82.
WTO (2007), World Trade Report 2007, Geneva, WTO.

Appendix
TableA2.1 WTOnotifications’coverageofNTMsbyMulti-AgencyCoding
MACcode(2009)HeadingNumberofrequirementsPeriodicityCoverage
A000Sanitaryand
phytosanitarymeasures
1(AgreementonSPS
measures)+1(Agreement
onlicensingprocedures)
AdhocChangesinSPSmeasuresnotbasedoninternational
standardandwithsignificanteffectontrade,atdraft/
projectstage
Notcovered:Pre-existingmeasures,finalmeasures.
B000Technicalbarrierstotrade13(AgreementonTBT)
ofwhich4onmeasures+
1(Agreementonlicensing
procedures)
AdhocNewtechnicalregulations(central,localand
non-governmentlevel)andconformityassessment
proceduresnotbasedoninternationalstandardsand
withsubstantialeffectontrade,atdraft/projectstage.
Notcovered:Pre-existingmeasures,finalmeasures
C000Pre-shipmentinspection
andotherformalities
Seedetailbelow
Notcovered:otherformalitiesn.e.s.
C100Pre-shipmentinspection2(AgreementonPSI)Once/adhocLawsandregulationspluschangesthereof
Notcovered:implementation(noindicationofprod.
coverage)
C200Directconsignment
requirements
Notcovered
C300Requirementstopass
throughaspecifiedcustoms
port
Notcovered
C400Importmonitoringand
surveillancerequirements
andotherautomatic
licensingmeasures
5(Agreementonimport
licensingprocedures)of
which2onmeasures
AdhocandannualAllnewlicensingproceduresandchangesthereof
D000PricecontrolmeasuresSeedetailsbelowSeedetailsbelowSeedetailbelow
Notcovered:pricecontrolmeasuresn.e.s.

D100Administrativepricing10(Agreementon
implementationofArticle
VII)
Customsvaluationlawsandregulation,
implementation
D200Voluntaryexportprice
restraints
2(Agreementon
safeguards)+1(AoA)
OnceVEPRsandothergreyareameasures
D300VariablechargesNotcovered(someinformationinIDB)
D400Anti-dumpingmeasures5(Agreementon
VI)ofwhich2onmeasures
Adhocandsemi-
annual
Allpreliminaryandfinalactions
D500Countervailingmeasures3(Agreementonsubsidies
andcountervailing
measures)ofwhich2on
measures
Adhocandsemi-
annual
Allpreliminaryandfinalactions
D600Safeguardduties12(Agreementon
safeguards)and2(AoA)of
which10onmeasures
AdhocoronceSpecialagriculturalsafeguardplusgeneralsafeguard
(atallstages)
D700SeasonaldutiesNotcovered(someinformationinIDB)
E000Licences,quotas,
prohibitionsandother
quantitycontrolmeasures
Seedetailbelow
Notcovered:otherquantitycontrolmeasuresn.e.s.
E100Non-automaticlicence5(Agreementonimport
licensingprocedures)of
which2onmeasures
AdhocandannualAllnewlicensingproceduresandchangesthereof
E200Quotas2(Decisiononnotification
proceduresforquantitative
restrictions)+2on(AoA)
(+19MFA)
Adhocand
biennial/adhoc
andannual
QuantitativerestrictionsandAGTRQs
E300ProhibitionsSameasE200
E400Quantitativesafeguard
measures
SameasD600

E500Exportrestraint
arrangements
2(Agreementon
safeguards)(alsocover
VEPRs)
OnceVERsandothergreyareameasures
F000Charges,taxesandother
para-tariffmeasures
BoundODCsareintariffschedules
F100CustomssurchargesNotcovered
F200ServicechargesNotcovered
F300Additionaltaxesand
charges
Notcovered
F400Internaltaxesandcharges
leviedonimports
Notcovered
F500Decreedcustomsvaluations10(Agreementon
VII)
Customsvaluationlawsandregulation,
implementation
coverage)
G000FinancemeasuresNotcovered
H000Anti-competitivemeasures
H100Restrictiveimportchannel1(GATTArticleXVII)AnnualStatetradingNotcovered:otherformsofrestrictions
onimportchannels
H200CompulsorynationalserviceNotcovered(possiblysomeinfonotifiedunder
GATS)
I000Trade-relatedinvestment
measures
2(AgreementonTRIMs)+
1onmaritime
Once/annualNon-conformingTRIMs,maritimevessels
J000DistributionrestrictionsNotcovered(possiblysomeinfonotifiedunder
GATS)
K000Restrictiononpost-sales
services
Notcovered(possiblysomeinfonotifiedunder
GATS)

L000Subsidies(excluding
exportsubsidies)
4(Agreementonsubsidies
andCVMs)ofwhich1on
measures+2(AoA)
Annual/adhoc
andannual
Allspecificsubsidies/domesticsupportandAMS
M000Governmentprocurement
restrictions
6(GPA)Lawsandregulationsofsignatories
Notcovered:non-signatories,implementation(no
indicationofprod.coverage)
N000Intellectualproperty18(TRIPS)Lawsandregulationspluschangesthereof
coverage)
O000Rulesoforigin4(Agreementonrulesof
origin)
Lawsandregulationspluschangesthereof
coverage)
P000Exportrelatedmeasures
P100Exportlicence,quota,
prohibition,andotherQRs
2(Decisiononnotification
proceduresforQRs)+1
(AoA)
AdhocExportrestrictionsonfoodstuffs
Notcovered:exportrestrictionsonotherproducts
P200Statetradingadministration1(GATTArticleXVII)AnnualStatetrading
P300Exportpricecontrol
measures
Notcovered
P400Measuresonre-exportNotcovered
P500ExporttaxesandchargesNotcovered
P600ExporttechnicalmeasuresNotcovered
P700Exportsubsidies3(Agreementonsubsidies
andCVMs)+1(AoA)
Adhoc/annualEliminationofexportsubsidies/AGexportsubsidies
Notes:MACcodeincludesproposeddraftrevision.
Source:WTOCentralRegistryofNotificationsandWTOlegaltexts.

TableA2.2 MembersreportingforsomekeyWTOnotificationrequirements
MACCode
(2009)
DescriptionofWTO
notification
Periodicity
NumberofMemberswithnotifications
Notificationeven
ifnoNTMs
SourceofMembercount
Once2006200720082009
AA000SPSAdhoc47405048SPSIMS
BB000TBTAdhoc63696371TBTIMS
CC100Pre-shipmentinspectionOnce109-----CRN
DD400Anti-dumpingmeasuresSemi-annual65626657G/L/223/Rev.18
D500CountervailingmeasuresSemi-annual60555652G/L/223/Rev.18
D600Safeguardduties(initiation)Adhoc95811DBonSafeguards
Specialsafeguards(AG)Annual30262317yG/L/223/Rev.18
D700
Seasonalduties(ifintariff
notification)
Annual87959293yIntegratedDataBase
EE100
Importlicensing(incl.
automatic)
Annual46373142yG/L/941
E200-
E300
Quotas,ProhibitionsBiennial95128G/L/223/Rev.18
E270TRQ(AG,importfill)Annual32272419G/L/223/Rev.18
LL000
Subsidies(excludingAG
exportsubsidies)
Biennial-71-78yG/SCM/W/546/Rev.2
NN000IntellectualpropertyAdhoc1082010CRN
OO000
Rulesoforigin(non-
preferential)
Once80-----G/L/939
Rulesoforigin(preferential)Once125-----G/L/939
PP100Exportrestrictions(AG)Adhoc1AGIMS
P200Statetrading(includesH110)Biennial36-35-yG/L/934
P700Exportsubsidies(AG)Annual61554731yG/L/223/Rev.18
Notes:ForlegalrequirementreferenceseeTableA2.3.CRN=CentralRegistryofNotifications.-=notapplicable.
Source:WTOdocumentsorinformationmanagementsystemsoftherespectivedivisionsasindicatedinlastcolumn.

TableA2.3 PrecisionofreportingforsomekeyWTOnotificationrequirements
MAC
Code
(2009)
DescriptionWTOlegalrequirement
Inforce
(nostart)
SpecificdatesProducts
InitiationStartEndDescriptionsHScodes
AA000SPSSPSArt.7AnnexBx(x)x(x)
BB000TBTTBTArt.2.95.6x(x)x(x)
CC100Pre-shipmentinspectionPSIArt.5x
DD400Anti-dumpingmeasuresGATT1994Art.VIArt.16.4xxxxx
D500CountervailingmeasuresSCMArt.25.11xxxxx
D600SafeguarddutiesSGArt.12.1xxxx(x)
Specialsafeguards(AG)AGRArt.5.718.2-MA:5xxxxx
D700Seasonalduties(ifintariffnotification)WT/L/225xxxx
EE100Importlicensing(incl.automatic)LICArt.1.4,7.3,8.2xx(x)
E200-
E300
Quotas,prohibitionsQR-(G/L/59)xxx
E270TRQ(AG,importfill)AGRArt.18.2-MA:2xxx
LL000Subsidies(excludingAGexportsubsidies)SCMArt.25.1-Art.XVI:1x
NN000IntellectualpropertyTRIPSArt.63.2-Legislationx
OO000Rulesoforigin(non-preferential)ROArt.5x
Rulesoforigin(preferential)ROArt.5§4A.IIx
PP100Exportrestrictions(AG)AGRArt.12.1(b)-ERxxxxx
P200Statetrading(includesH110)GATT1994Art.XVII:4(a)xxx
P700Exportsubsidies(AG)AGRArt.1018.2xxx
Notes:ForperiodicityseeTableA2.2.(x):Informationsometimesprovided.Forspecificdates–Initiation:thedatewhenmeasurehasbeeninitiated(anti-dumping,
countervailingduties)orputforwardforcomments(TBT,SPS);Start:thedatewhenmeasurehasbeenputinforce;End:thedatewhenmeasurehasbeenwithdrawnor
terminated.
Source:BasedoninformationprovidedintheWTOnotifications.

3
NTMs: Interpreting the New Data
Julien Gourdon and Alessandro Nicita
3.1 Introduction
Since the paper ‘Mystery of the Missing Trade’ (Trefler, 1995), many studies
have investigated the reasons why world trade is not as large as economic models
predict. One of the most compelling explanations was provided by Obstfeld and
Rogoff (2000) which suggested that large unobserved trade costs may explain
most of the discrepancies between model estimates and trade statistics. The
presence of hidden costs was supported by subsequent work, as in Anderson
and Van Wincoop (2004), whose research indicated that the costs associated
with cross-border trade even between well-integrated countries are well above
those that can be explained by geographic distance and traditional trade policies.
Although a number of studies have attempted to capture and quantify the impact
of some of the hidden costs of trading (Maskus et al, 2005; Djankov et al,
2010; Hoekman and Nicita, 2011), these attempts are greatly constrained by the
available data. The existing data on trade costs is largely related to tariffs, and
only a few databases provide information on non-tariff measures and behind
the border trade costs (eg the Doing Business database, the Trade Facilitation
Database, the Logistic Performance Index, and the UNCTAD/TRAINS NTM
database). Moreover, most of the existing data is too aggregated to be utilised
for detailed policy analysis and often provides information only on the effects of
trade impediments rather than on the impediments themselves. In practice, the
analysis must compromise in terms of policy coverage focusing on the aggregate
effects of the few countries or sectors where the data is available.
A particularly relevant issue for both researchers and policymakers is related
to the impact of non-tariff measures (NTMs) on trade. There are several reasons
to focus attention on NTMs as one of the main sources of trade costs. One reason
is that their impact on trade is still poorly understood and not easily measured.
NTMs encompass a wide set of policies that can have very diverse effects
on trade. For example, requirements on marking, labelling and packaging,
although adding to costs of production, generally are not discriminatory and
have low compliance costs, and thus have relatively unimportant trade effects.
On the other hand, quotas, voluntary export restraints and non-automatic import

authorisations often have much more significant effects. The second reason to
examine NTMs is their proliferation. While NTMs have been around for years,1
the use of NTMs to regulate trade has been rising since the 1990s both in terms
of countries adopting these measures as well as in their variety. The third reason
is that NTMs can be discriminatory. Even when NTMs are indiscriminately
applied to all imported goods, many NTMs discriminate among a country’s
trading partners because costs of compliance are often different across exporters.
Compliance costs are generally higher in low income countries, as NTM-related
production processes and export services are often more expensive, or need to
be outsourced abroad. Another reason to investigate NTMs is that they could be
protectionist. Governments are using increasingly sophisticated methods about
how they protect domestic industries. While trade barriers have historically
taken more obvious forms, such as tariffs or quotas, trade barriers are now taking
different forms which are harder to identify and quantify. A mounting concern
is that liberalisation in tariffs may be countered by the increasing number of
restrictive NTMs.
Broadly defined, NTMs include all policy-related trade costs incurred from
production to final consumer, with the exclusion of tariffs. For practical purposes
NTMs are categorised depending on their scope and/or design and are broadly
distinguished in technical measures (sanitary and phytosanitary standards (SPS);
and technical barriers to trade (TBT)) and non-technical measures. These are
further distinguished in hard measures (eg price and quantity control measures),
threat measures (eg antidumping and safeguards), and other measures such as
trade-related finance and investment measures). In practice, NTMs are measures
that have the potential to substantially distort international trade, whether their
trade effects are protectionist or not. For example, measures such as quality
standards, although generally imposed without protectionist intent, may be of
particular concern to poor countries whose producers are often ill-equipped to
comply with them.
The paucity of data on trade policy measures has been the main problem behind
the study of the effect of NTMs on trade. Seemingly simple questions such as
‘what are the policy measures countries impose’ or ‘what type of measures are
faced by particular products’ cannot be answered for most goods and countries
because of the lack of detailed information. The fact that NTMs are increasingly
used to regulate international trade makes the need to update data even more
compelling.
The reason behind the scarcity of databases on NTMs is largely related to the
difficulty in collecting the data and in assembling a consistent cross-country
database. Unlike tariffs, NTM data are not merely numbers; the relevant
information is often hidden in legal and regulatory documents. Moreover, these
1 For example, English laws in the 17th and 18th centuries required that all colonial trade be conducted
on British ships manned by British sailors. Also, certain goods had to be shipped to Great Britain first
before they could be sent to their final destination.

NTMs: Interpreting the New Data 59
documents are generally not centralised but often reside in different regulatory
agencies. All these issues make the collection of NTM data a very resource
intensive task. The first attempt to collect and categorise NTMs was conducted
by UNCTAD in the late 1990s, and the data is available in the UNCTAD Trade
Analysis and Information System database (TRAINS – accessible via WITS).2
However, the TRAINS NTMs database has not been consistently updated in the
last 10 years. To fill this gap and in response to the increased interest of both
researchers and policymakers, UNCTAD and the World Bank in collaboration
with the International Trade Centre and the African Development Bank, have
initiated a new effort on NTM data with the objectives of improving the coverage
and classification of NTMs and to update, consolidate and freely disseminate
NTM data.
As early as 2011, this joint effort produced an updated NTMs classification
as well as detailed new data for about 25 countries, with data from more
countries in the pipeline. This paper makes use of the new data to provide some
preliminary information on the incidence of NTMs across countries and by
economic sector and by type of NTMs. Given limited coverage of the new data,
the analysis is mainly descriptive in nature and employs simple indicators (an
inventory approach based on frequency and coverage ratio) rather than trying to
produce more complex measures such as price gaps or ad valorem equivalents.
In practice, the analysis focuses on identifying the relative use of various types
of NTMs, and their incidence across countries and products.
The rest of this study is organised as follows. The next section provides
some details on the definition and classification of NTM data. The bulk of the
descriptive analysis is in Section 3.3. There we provide descriptive statistics on
the incidence of NTMs in terms of frequency (number of product lines exposed
to NTMs) and coverage (share of total imports exposed to NTMs). In doing
so, we analyse differences both in terms of countries and product groups. We
also examine the evolution in the use of NTMs by using original data from the
TRAINS database versus the data collected recently. Section 3.4 explores the
relationship between NTM and traditional forms of trade policy. The last section
summarises the main findings and offers some policy conclusions.
3.2 Definition, classification and data collection
The definition of NTMs should encompass all measures altering the conditions
of international trade including policies and regulations that restrict trade as well
as those that facilitate it. NTMs are often incorrectly referred to as non-tariff
barriers (NTBs). The difference is that NTMs comprise a wider set of measures
than NTBs, which are now generally intended only as discriminatory non-tariff
measures imposed by governments to favour domestic over foreign suppliers.
The cause of this confusion is because in the past most NTMs were largely in the
form of quota or voluntary export restraints. These measures are restrictive by
2 http://wits.worldbank.org/wits.

design which explains why the word ‘barrier’ was used. In present times, policy
interventions take many more forms, and therefore it is preferable to refer to
them as ‘measures’ rather than ‘barriers’ to underline that the measure may not
be necessarily welfare or trade reducing.3
For practical purpose, the commonly
used definition of NTMs is:
‘Non-tariff measures (NTMs) are policy measures, other than ordinary
customs tariffs, that can potentially have an economic effect on international
trade in goods, changing quantities traded, or prices or both.’ (UNCTAD,
2010)
This definition is broad and to a large extent uninformative as it was in the
case of NTBs, which were defined as policies that are not tariffs. To better
identify NTMs, and distinguish among the various forms of NTMs, a detailed
classification is therefore of critical importance. To facilitate data collection
and analysis, the multitude of NTMs are often aggregated in various groups:
hard measures (eg price and quantity control measures), threat measures (eg
antidumping and safeguards), sanitary and phytosanitary standards (SPS),
technical barriers to trade (TBT), and other categories such as ‘export measures’,
‘trade-related investment measures’, ‘distribution restrictions’, ‘restrictions
on post-sales services’, ‘subsidies’, ‘measures related to intellectual property
rights’ and ‘rules of origin’. Each of these groups comprises various and often
very different forms of NTMs. The classification (see Figure 3.1) proposed by
UNCTAD and agreed by the Group of Eminent Persons on NTB4
takes this into
account and develops a tree/branch structure where measures are categorised
into chapters depending on their scope and/or design. Then each chapter is
further differentiated into several sub-groups to allow a finer classification of
the regulations affecting trade. In practice, the NTM classification encompasses
16 chapters (A to P) and each individual chapter is divided into groupings with
depth up to three levels (1, 2 and 3 digits). Although a few chapters reach the
3-digits level of disaggregation, most of them stop at 2 digits. The complete
classification can be found in UNCTAD (2010).
The classification organises NTMs into various chapters, each comprising
measures with similar purposes. All chapters reflect the requirements of the
importing country on its imports, with the exception of measures imposed
on exports (chapter P). The effect on trade of each group of measures varies
considerably. While some groups of NTMs have clear restrictive impacts, others
produce uncertain effects. For example, the subset of measures under categories
(A) through (C) has a relatively clear relationship with the market imperfections
they try to address (Beghin, 2006). These measures are largely regulatory
3 For example, NTMs such as standards and regulations may expand trade by facilitating production
and exchange of information, reducing transactions costs, guaranteeing quality, and achieving the
provision of public goods (Maskus, Wilson and Otsuki 2005). Where trade in some products would
have been difficult without clear standards, with it, trade could be created between two countries..
4 UNCTAD (2012).

policies in response to a variety of concerns raised by society in many areas
such as the environment, animal welfare, food safety and consumers’ rights.
These policies are not necessarily restrictive because these types of instruments
can also enhance consumer demand for goods by increasing quality attributes
(technical requirements) or by reducing informational asymmetries (standards).
However, many of these policies involve considerations of institutional capacity
and likely have distortionary impacts on trade. Sometimes they are imposed to
address the possible capacity failures of trade partners; and often they require an
extensive domestic institutional capacity to implement these policies. Although
different types of requirements affect different inputs and stages of production,
most of these policies also affect overall trade costs (eg certification, inspections,
etc). In addition, compliance costs often vary depending on infrastructure and
institutional capacity of the exporting country, and thus ultimately these costs do
affect trade flows.
Figure 3.1 Non-tariff measure classification (1st tier – chapters)
A SPS
B TBT
C Pre-shipment clearance and other formalities
D Price control
E Licenses, quotas, prohibition other quantity control m.
F Charges, taxes and other para-tariff measures
G Finance
H Anti-competitive
I Trade-related investment
P Export-related measures
J Distribution restrictions
K Post-sales services
L Subsidies
M Government procurement
N Intellectual property
O Rules
Technical
Measures
Importmeasures
Non-
Technical
Measures
Export
measures

Non-technical measures vary considerably by intent and scope. However, their
effect on trade is generally more understood and easier to quantify. The effects of
price control measures are relatively simple to measure, especially anti-dumping
and safeguards. Quantity control instruments have been extensively examined in
the analysis of quotas, tariff rate quotas and their administration (see Boughner et
al, 2000). Para-tariff measures can be analysed as conventional tax instruments
and their incidence is straightforward to perceive. Finance, anti-competitive, and
trade-related investment measures have indirect effects on trade, and their actual
impact is more difficult to assess. Box 3.1 provides some more details on the
measures contained in each chapter.
The classification discussed above greatly simplifies the data collection.
However, being able to classify laws and regulations into the appropriate NTM
category is only part of the challenge in assembling a database of NTMs. Besides
a proper classification, one of the problems related to data collection is that, in
most cases, there is not one sole national repository agency of NTM data as laws
and regulations affecting trade are often promulgated by different government
agencies and regulatory bodies. This makes the assembly of an exhaustive
NTM database quite a challenging task. In practice, the data has to be carefully
scrutinised for possible duplications, omissions, or any other problems in order
to minimise inaccuracies.
This chapter provides an analysis based on the newly collected NTM data
comprising 24 developing countries plus the European Union and Japan. The
data covers measures from chapters A to I, and chapter P.5
The data follows the
HS classification at the 6-digit level covering more than 5,000 different products.
5 Because of objective difficulties in the collection of data on some measures, data covering measures
from chapters J to O was not actively collected.

Box 3.1 Brief description of NTMs chapters
Chapter A, on sanitary and phytosanitary measures, refers to measures affecting areas
such as restriction for substances, restrictions for non eligible countries’ hygienic
requirements, or other measures for preventing dissemination of diseases, and others.
Chapter A also includes all conformity assessment measures related to food safety, such
as certification, testing and inspection, and quarantine.
Chapter B, on technical measures, refers to measures such as labelling, marking,
packaging, restrictions to avoid contamination or other measures protecting the
environment, standards on technical specifications, and quality requirements.
Chapter C classifies the measures related to customs formalities.
Chapter D, price control measures, includes measures that have the intention to
change the prices of imports, such as minimum prices, reference prices, antidumping or
countervailing duties.
Chapter E, licensing, quotas and other quantity control measures, groups the measures
that have the intention to limit the quantity traded, such as quotas. Chapter E also covers
licences and import prohibitions that are not SPS or TBT related.
Chapter F, on charges, taxes and other para-tariff measures, refers to taxes other than
custom tariffs. Chapter F also groups additional charges such as stamp taxes, licence fees,
statistical taxes, and also decreed customs valuation.
Chapter G, on finance measures, refers to measures restricting the payments of imports,
for example when the access and cost of foreign exchange is regulated. It also includes
measures imposing restrictions on the terms of payment.
Chapter H, on anticompetitive measures, refers mainly to monopolistic measures, such
as state trading, sole importing agencies, or compulsory national insurance or transport.
Chapter I, on trade-related investment measures, groups the measures that restrict
investment by requesting local content and thus restricting imports, or requesting that
investment should be related to export in order to balance imports.
Chapter J, on distribution restrictions, refers to restrictive measures related to the internal
distribution of imported products. These measures would hinder trade from taking place
because there would be difficulty in distributing the products once entering the country.
Chapter K, on the restriction on post-sales services, refers to difficulties in allowing
technical staff to enter the importing country to install or repair technological goods
imported.
Chapter L contains measures that relate to the subsidies that affect trade.
Chapter M, on government procurement restriction measures, refers to the restrictions
bidders may find when trying to sell their products to a foreign government.
Chapter N, on intellectual property measures, refers to the problems arising from
intellectual property rights.
Chapter O, on rules of origin, groups the measures that restrict the origin of products so
that they could benefit from reduced tariffs according to certain rules often set in multiple
simultaneous agreements with different countries.
Chapter P, on export measures, groups the measures a country applies to its exports. It
includes export taxes, export quotas or export prohibitions, etc.

3.3 The incidence of NTMs
There are various approaches for identifying the importance of trade measures
and assessing their effects on international trade. Methodologies include simple
inventory measures, computation of price gaps and the estimation of ad valorem
equivalents. As the intent is mainly to explore the collected data, the paper
utilises the simple inventory approach. This approach is based on two indices:
the frequency index and the coverage ratio. The frequency index simply captures
the percentage of products that are subject to one or more NTMs. The coverage
ratio captures the percentage of imports that is subject to one or more NTMs.
The frequency index accounts only for the presence or absence of an NTM, and
summarises the percentage of products to which one or more NTMs are applied.
In more formal terms, the frequency index of NTMs imposed by country j is
calculated as:
100⋅








=
∑
∑
i
ii
j
M
MD
F
where D is a dummy variable reflecting the presence of one or more NTMs and
M indicates whether there are imports of good i (also a dummy variable). Note
that frequency indices do not reflect the relative value of the affected products
and thus cannot give any indication of the importance of the NTMs on overall
imports.
A measure of the importance of NTMs on overall imports is given by the
coverage ratio which measures the percentage of trade subject to NTMs for the
importing country j. In formal terms the coverage ratio is given by:
100⋅








=
∑
∑
i
ii
j
V
VD
C
where the D is defined as before, and V is the value of imports in product i. One
drawback of the coverage ratio, or any other weighted average, arises from the
likely endogeneity of the weights (the fact that imports are dependent on NTMs).
This problem is best corrected by using weights fixed at trade levels that would
arise in a NTM (and tariff) free world. Otherwise, the coverage ratio would be
systematically underestimated. While one cannot get to that benchmark, it is
possible to soften the endogeneity problem (and testing for the robustness of the
results) by using trade values of past periods.
We start the descriptive analysis by aggregating all the data collected and
examining the incidence of various types of NTMs. Figure 3.2 illustrates the
distribution of NTMs across five main chapters for the 26 countries collected so
far. For each chapter both the frequency indices and coverage ratios are reported.

Figure 3.2 Frequency index and coverage ratios by chapter (all countries, simple
weighted )
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
Quantity Control
Price Control
PSI
TBT
SPS
Frequency Index Coverage Ratio
According to the newly collected data, TBTs are by far the most used regulatory
measures with the average country imposing them on about 30% of products
and trade. Countries impose SPSs on average on about 15% of trade. The large
incidence of SPS and TBT raises concerns for developing countries’ exports.
These measures impose quality and safety standards which often exceed
multilaterally accepted norms. Although these measures are not protectionist in
nature they often result in diverting trade from developing countries where the
production process and certification bodies are often inadequate. Moreover, the
cost of compliance to SPS and TBT is often higher in low income countries
as infrastructure and export services are often more expensive or need to be
outsourced abroad. In practice, SPS and TBT may erode the competitive
advantage that developing countries have in terms of labour costs and preferential
access.
Among non-technical measures, pre-shipment inspections affect, on average,
almost 20% of trade and products. Although pre-shipment inspections are often
necessary to provide some assurance on the quality/quantity of the shipment,
and thus may promote international trade, they add to the cost of trading. These
additional costs may reduce the competitiveness of countries, thus distorting
trade. Regarding price control measures (8% of trade and only 5% of products),
these constitute one of the least used forms of NTMs. Price control measures
affect only a small share of goods and are largely related to anti-dumping and
countervailing duties, as well as some form of administrative pricing for staple
food, energy and other sensitive sectors. Finally, the average country imposes
quantity controls on about 18% of products and 23% of trade. Only a small

percentage of these measures still take the form of quotas and export restrictions,
since most these quantitative restrictions are illegal under WTO rules. Some of
them, such as quotas, prohibitions and export restraints are in place, but are
largely limited to a number of sensitive products and in other cases they take
the form of non-automatic licensing used as a tool to administer the importation
of goods where SPS and TBT related issues are of particular importance. Some
quantitative restrictions such as quotas, prohibitions and export restraints are in
place, but are largely limited to a number of sensitive products.
The incidence of different forms of NTMs varies across geographic areas.
Figure 3.3 illustrates the use of NTMs by differentiating the countries in the
sample in three broad developing regions and a high income group.Although SPS
and TBT are the most used forms of NTMs independently of the region, many
countries especially in Asia and Latin America still implement a large number
of quantitative restrictions (largely in the form of licensing). African countries
appear to regulate their imports relatively more than many other developing
countries, especially in relation to PSI. The reason behind this relatively large
number of PSI is that these are often implemented to fight corruption, to
facilitate and accelerate custom procedures, and ultimately to help in the correct
evaluation of imports and their proper taxation. The large use of SPS and TBT by
African countries may result from an effort to harmonise regulations with their
main trading partner, the EU.
Figure 3.3 Frequency index and coverage ratios by chapter (by region)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Price Control
Quant. Control
PSI
SPS
TBT
Price Control
Quant. Control
PSI
SPS
TBT
Price Control
Quant. Control
PSI
SPS
TBT
Price Control
Quant. Control
PSI
SPS
TBT
LatinAmericaAsiaAfricaHighIncome

The use of NTMs varies considerably not only across regions but more so among
countries. Figure 3.4 summarises the data in terms of frequency index and the
coverage ratio for each country for all NTMs as a whole. On average, countries
apply some form of NTMs for slightly less than half of the about 5,000 products
included in the HS 6-digit classification. This figure greatly varies by country.
For example, within Africa, Tanzania and Senegal use NTMs substantially less
than Egypt, Kenya or Uganda. In Latin America, Argentina’s use of NTMs is
double that of Chile or Paraguay. In Asia, Bangladesh, Syria and the Philippines
utilise NTMs much more than Cambodia or Indonesia. Although this large
variance may be due to some extent to different primary data collection methods,
this is likely to explain only part of the differences, as a large variance is also
found for Latin American countries whose data is collected by the same agency:
the Associación Latinoamericana de Integración (ALADI).
Figure 3.4 Frequency indices and coverage ratios, by country
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Japan
European Union
Uruguay
Peru
Paraguay
Mexico
Ecuador
Colombia
Chile
Brazil
Bolivia
Argentina
Syria
Philippines
Lebanon
Lao PDR
Indonesia
Cambodia
Bangladesh
Uganda
Tunisia
Tanzania
South Africa
Senegal
Namibia
Morocco
Mauritius
Madagascar
Kenya
Egypt
Burundi
High
Income
LatinAmericaAsiaAfrica

Similar conclusions can be reached by looking at coverage ratios (the percentage
of imports subject to NTMs) as these are found to be highly correlated with
frequency indices. Although correlated, coverage ratios are often lower than
frequency indices. This is possibly due to the endogeneity issue described above
(NTMs may restrict trade and this will downwardly bias the coverage ratio).
The problem of endogeneity seems to be supported by the relatively lower
coverage ratio for Latin American countries, where measures imposing quantity
restrictions are still significant. The difference between the two incidence
measures is particularly striking for Tanzania (from 5 to 30%) and Lebanon
(from 15 to 40%). A coverage ratio relatively higher than the frequency index
can be explained by two factors. The first is import composition. Countries,
especially low income countries, often import larger volumes of products where
NTMs are more extensively used (agriculture). The second factor is a larger use
of NTM policies on products that are most traded (eg for consumer protection).
This is often the case in developed countries.
The incidence of the use of NTMs depends on both the percentage of products
(or imports) affected by NTMs, and the number of NTMs affecting each product.
Frequency and coverage ratios illustrated above do not take into account whether
more than one type of NTM is applied to the same product. In practice, a
large number of products have more than one regulatory measure applied to
them. For example, a product could be subject to a sanitary standard as well
as a technical measure on quality, and finally to some licensing. Arguably, the
greater the number of NTMs applied to the same product, the more regulated the
commerce of that product is, especially if measures are from different chapters.
The rationale is that measures within the same chapter are similar in nature and
thus often impose relatively less burden than measures from different chapters.
To better illustrate the pervasiveness of NTMs, Figure 3.5 reports the number
of products affected by one, two, three, or four types of NTMs, where types are
differentiated by chapter.
Although a large share of products affected by NTMs are subject to NTMs
from only one chapter, a substantial number of products are affected from
multiple different types of NTM. For example, among about 4,500 products
where the EU imposes NTMs, about 3,200 are subject to NTMs from only one
chapter, about 1,100 are affected by NTMs from two different chapters, and
about 250 by NTMs from three or more chapters. Although the EU frequency
index and coverage ratio are similar to that of Argentina, EU imports can be
considered relatively less regulated, as the majority of Argentina’s imports are
affected by NTMs from two or more chapters. These statistics also allow us to
verify the quality of the data. The case of Namibia (and possibly that of Kenya) is
particularly striking as all 2,900 products where NTMs are applied are subjected
to multiple NTMs from at least four different chapters. This is unlikely to be the
case and the data from Namibia would need to be further scrutinised for possible
errors in the classification procedure.

Figure 3.5 Number of NTMs from different chapters affecting HS 6-digit products
0
1000
2000
3000
4000
5000
Burundi
Egypt
Kenya
Madagascar
Mauritius
Morocco
Namibia
Senegal
SouthAfrica
Tanzania
Tunisia
Uganda
Bangladesh
Cambodia
Indonesia
LaoPDR
Lebanon
Philippines
Syria
Argentina
Bolivia
Brazil
Chile
Colombia
Ecuador
Mexico
Paraguay
Peru
Uruguay
EU
Japan
Africa Asia Latin America High
Income
NumberofProducts
1 type 2 types 3 types 4 types
It is often the case that countries apply a wide number of NTMs within each
chapter. For example, one specific good may be subject to geographical
restriction, labelling, fumigation and some conformity assessments which all
fall under the SPS chapter (A). Although some of these measures may impose
few additional costs, some others are quite distinct. A large number of measures
within a chapter could imply an even stricter regulatory framework. Thus, it is
important to provide some information on the actual number of NTMs applied
to single products. This information is given by simply calculating the average
number of NTMs applied to each HS 6-digit product. Table 3.1 reports for each
country the average number of NTMs applied to the products facing at least one
NTM at the various levels of aggregation of NTMs classification.

Table 3.1 Use of multiple types of NTMs at different level of aggregation
Name
Number of lines with
at least one NTM
Average over number of lines with NTMs
Chapter level 1-digit level
All types of
NTMs
Argentina 4035 2.47 3.00 3.16
Bangladesh 3476 1.71 1.85 1.94
Bolivia 1408 1.75 1.91 1.99
Brazil 2808 2.25 3.07 3.15
Burundi 5040 2.47 5.34 7.17
Cambodia 1687 1.14 1.46 1.86
Chile 2224 1.68 1.83 1.87
Colombia 2962 2.46 2.99 3.12
Ecuador 1935 1.68 2.21 2.27
Egypt 5014 1.72 2.09 2.34
European Union 4550 1.36 3.80 5.18
Indonesia 2353 1.65 2.05 2.84
Japan 2132 1.37 4.88 8.39
Kenya 4491 2.55 5.74 9.02
Lao PDR 4100 1.63 2.58 3.68
Lebanon 829 1.04 1.29 1.46
Madagascar 1673 1.35 1.63 1.64
Mauritius 2354 1.08 1.08 1.45
Mexico 3105 1.49 1.59 1.64
Morocco 1417 1.77 3.11 4.13
Namibia 2857 4.14 9.02 9.41
Paraguay 1399 1.41 1.69 1.70
Peru 2427 1.43 1.71 1.93
Philippines 4934 1.17 1.22 1.27
Senegal 388 1.83 2.09 2.76
South Africa 2233 1.30 1.93 2.04
Syria 4803 1.75 2.07 2.40
Tanzania 288 1.33 1.73 1.83
Thailand 146 1.21 2.62 6.46
Tunisia 1244 2.95 6.00 11.38
Uganda 5062 3.08 3.12 4.09
Uruguay 2261 1.47 1.75 1.85
Average (simple) 2676 1.80 2.76 3.61

With very few exceptions, products are rarely affected by only one type of NTM
because several regulatory measures are often applied in parallel. The average
number of NTMs affecting products facing at least one NTM is 1.82 at the chapter
level, 2.77 at the 1-digit level, and 3.61 when all possible NTMs are considered.6
These figures vary considerably across countries. For example, while Mauritius
imposes about one NTM measure at the 1-digit and chapter level for each of its
2,354 HS 6-digit products covered by NTMs, Japan imposes an average of almost
five 1-digit categories of NTMs, mainly from the same chapter on its 2,131 HS
6-digit products subject to NTMs. This suggests that Japan, although imposing
NTMs on a similar number of products than Mauritius, regulates its imports
substantially more. Similarly, Tunisian import restrictions are applied on 1,244
products. These products face an average of 3 NTMs from different chapters
and an average of 11 distinct types of NTMs. On the other hand, although Egypt
applies at least one NTM for most of the HS 6-digit lines, only about an average
of 2 NTMs are applied for each line. Although these statistics provide valuable
information, such large differences at the most disaggregated level should not be
considered proof of over-regulated import regimes. These differences could also
be due to data availability and collection procedures. In particular, differences
may be related to whether the document is detailed enough to distinguish among
several types of similar NTMs, in which case NTMs are generally classified
only under broader codes. Differences at the 1-digit level often reflect more real
differences in the use of regulatory measures for imports, and thus can provide a
better assessment of the regulatory regime. For example, both Mexico and Brazil
impose some form of NTM for about 3,000 products. However, while Mexico
applies only 1.5 1-digit NTM measure on each of these products, Brazil applies
about 3 NTMs. Arguably, Brazil’s imports can be considered on average more
regulated than those of Mexico.
As NTMs regulate different aspects of the production and trade of goods, it
is often the case that various types of NTMs are applied in parallel to the same
product. While in most cases NTMs are used complementarily and applied
simultaneously to the same product for technical or procedural reasons, there
are other cases where NTMs are applied in parallel to further insulate domestic
industries from foreign competition. To explore to what extent the various types
of NTMs are complementary, we calculate correlation coefficients among 1-digit
NTMs. Figure 3.6 shows correlation statistics where the size of dots depends on
the magnitude of the correlation.
Figure 3.6 suggests several key patterns in the concurrent use of various
types of NTMs. First, several SPS measures are often applied in parallel (for
example, tolerance limits (A2) are often used in conjunction with labelling
(A3) and hygienic requirements (A4) and also with treatment measures such
as fumigation (A5) and conformity assessments (A8). Second, SPS measures
are also often paired with TBT (B5). This is largely related to regulation on
genetically modified organisms (GMO). SPS measures are also found to be
6 These are averages only for products affected by at least one type of NTM.

correlated to direct consignment (C2), and to requirements to pass through
specified custom ports (C3). This is possibly to facilitate the inspections and
traceability of agricultural products. For similar reasons SPS measures are also
correlated with non-automatic licensing (E1).
Technical barriers to trade (TBTs) are found to be relatively less correlated
with other groups of NTMs. The only exceptions are the TBTs on production
or post production requirement (B4) which are often used simultaneously with
price controls (D4, D5 and D6) and quantity controls (E2 and E3). As there is no
clear explanation why such NTMs should be correlated, it would be interesting
to further explore this pattern. We leave this for future research. Finally, and not
surprisingly, quantity and price controls measures appear closely interrelated.
This specifically concerns quantity control: quota (E2) and prohibition (E3), and
price control measure: anti-dumping (D4) and countervailing (D5) measures.
All these measures are often used concomitantly to reinforce the protection of
specific sectors.
Figure 3.6 Correlation of different 1-digit NTMs
A1
A2
A3
A4
A5
A6
A7
A8
B1
B2
B3
B4
B5
B6
B7
B8
C1
C2
C3
C4
C9
D1
D3
D4
D5
D6
E1
E2
E3
x
A1 A2 A3 A4 A5 A6 A7 A8 B1 B2 B3 B4 B5 B6 B7 B8 C1 C2 C3 C4 C9 D1 D3 D4 D5 D6 E1 E2 E3y
We now turn to analyse the impact of NTMs across economic sectors. The use
of NTMs greatly varies across economic sectors both for technical and economic
reasons. While some products, such as agriculture, electric machinery, weapons,
are highly regulated because of consumers and environmental protection and
technical standards, some other goods are by their nature less subject to laws
and regulation. Table 3.2 reports frequencies indices of five broad categories of
NTMs for 20 economic sectors.

Table 3.2 Frequency indices across economic sectors
A: SPS
B:
TBT
C: Pre-
shipment
D:
Price
control
E:
Quantity
control
Live animals 71.3 36.2 21.3 5.7 33.4
Vegetable products 69.2 31.7 24.0 3.6 27.1
Fats and oil 51.1 26.8 12.9 8.0 20.7
Processed food 57.0 41.7 17.7 3.6 20.3
Minerals products 9.8 25.5 8.1 0.6 10.9
Chemical products 11.3 35.8 6.8 1.7 19.6
Rubber and plastics 1.2 24.1 5.7 0.8 6.3
Raw hide and skins 12.8 23.7 9.9 0.0 12.9
Wood 26.2 30.2 12.4 0.8 15.2
Paper 1.7 18.4 8.2 0.6 11.4
Textile 1.8 34.3 15.6 4.7 16.3
Footwear 0.7 38.8 16.7 3.3 17.9
Stone and cement 3.1 19.0 9.7 1.1 6.3
Base metals 1.6 21.0 9.6 1.2 12.2
Machinery electrical equipment 1.1 20.8 8.2 0.8 13.1
Motor vehicles 0.3 26.2 8.4 0.7 22.5
Optical and medical instruments 0.4 20.0 7.9 0.2 8.1
Miscellaneous goods 1.6 23.0 7.2 4.1 7.2
The use of SPS measures is largely limited to agricultural sectors and products
from animal origin, as their control is essential for ensuring the health and well
being of consumers and the protection of the environment. As a result, more
than 60% of food-related products are found to be affected by at least one form
of SPS. On the contrary, TBT can suit a much wider set of products and indeed
these are found to be more uniformly applied across economic sectors with
peaks in textiles, footwear, processed food and chemicals. Measures involving
pre-shipment requirement are widely distributed across economic sectors but
concern a more limited number of products. Pre-shipment inspections are found
to be more relevant for agricultural products, wooden products, textiles and
footwear. Price control measures such as administrative pricing, anti-dumping
and countervailing duties are trade defensive policies that by their nature are
applied only to very specific products and thus result in low frequency indices.
Like pre-shipment requirements, price control measures are more concentrated
in agricultural products, textiles and footwear. Finally, quantity control measures
are applied more or less uniformly across economic sectors with peaks on
agricultural goods, animal products, motor vehicles and chemical products.

These are sectors where particularly sensitive products are often regulated by
non-automatic licences, quotas, and sometimes outright prohibitions.
The distribution of NTMs across sectors, especially with regard to SPS and
TBT, is due more to the technical properties of products than to economic policy,
and therefore, does not vary substantially across countries. Other measures have
a more heterogeneous distribution as the choice among different measures for
the regulatory intent may be different across countries depending on various
factors such as institutional capacity, implementation costs and effectiveness.
Figure 3.7 illustrates regional averages of the frequency indices of five broad
types of NTMs across six broad economic sectors. Although SPS are similarly
applied to food products regardless of geographic region, Asian, Middle East
and North African countries do not seem to apply as many TBTs for agriculture
products, especially in comparison with Africa or Latin America. The finding
of more TBT measures in the Africa region than in Latin America and Asia is
surprising since one should expect fewer TBTs from lower income countries.
One hypothesis is that these countries are implementing EU standards, so as
to better compete in the EU market. Pre-shipment inspections are widely used
in sub-Saharan Africa while they are limited to food products, textiles, apparel
and footwear in other regions. Price control measures are limited to some food
products across all geographic regions, and to textiles and apparel in Latin
America. Finally, quantity control measures are found to have limited use in
countries in the Middle East and North Africa. These measures are instead more
widely used in Asian, sub-Saharan African and Latin American countries.
Countries are increasingly using NTMs to regulate their imports. Figure 3.8
illustrates the changes in the use of NTMs between 2000 and 2010. A caveat
with this type of analysis is that there is little comparable NTM data across time,
and most of it originates from Latin American countries. For all other countries
the collection procedures have substantially changed and original data may not
have been as complete as the data recently collected. Because of data limitation,
Figure 3.8 reports the share of four broad groups of NTMs. With the exception
of prohibitions, the number of products affected by NTMs has increased. In
particular, the category where the number of products covered by NTMs has
increased the most is SPS and TBT. As at 2010, about one-third of products in
our sample of countries were affected by one or more types of SPS and/or TBT.
Price control measures remained barely used while the use of quantity control
measures declined, possibly because of the progressive tariffication of quotas.
Finally, the use of other types of NTMs, such as pre-shipment inspection and
contingency measures, appears to have substantially increased.

Figure 3.7 Frequency indices across economic sectors, by region
A: SPS
B: TBT
C: InspecƟon
D: Price control
E: QuanƟty control
0
10
20
30
40
50
60
70
80
LaƟn America
A: SPS
B: TBT
C: InspecƟon
D: Price control
E: QuanƟty control
0
10
20
30
40
50
60
70
80
Middle East and North Africa
A: SPS
B: TBT
C: InspecƟon
D: Price control
E: QuanƟty control
0
10
20
30
40
50
60
70
80
Sub-Saharan Africa
A: SPS
B: TBT
C: InspecƟon
D: Price control
E: QuanƟty control
0
10
20
30
40
50
60
70
80
East Asia
Figure 3.8 Number of products covered by NTMs (2000 and 2010)
0%
10%
20%
30%
40%
50%
60%
Technical Measures Price Control Quantity Control Other Measures
FrequencyIndex
1999 2010

3.4 NTMs and traditional forms of trade policy
The use of multiple instruments of trade policy to regulate imports involves
not only NTMs but also traditional forms of trade policy. This section explores
whether NTMs are used as the complements or substitutes of traditional trade
policy, namely tariffs. The relationship between NTMs and tariffs can be
assessed across countries or across products. In relation to countries, the analysis
investigates whether countries applying restrictive traditional trade policies
(high tariffs) are also those where NTMs are more frequently used so as to better
protect their domestic industry from foreign competitors. If this is the case, it
would result in a positive relationship between the use of NTMs and the level of
tariffs as products may be protected not only by a large number of NTMs but also
by high tariffs. Although a large number of NTMs may result from the nature
of the product, when these are accompanied by a high tariff it may indicate the
intent to use NTMs to complement tariffs to further insulate domestic industries
from foreign competition.
The relationship between NTMs and tariffs across countries is illustrated in
Figure 3.9 where NTMs are defined by their coverage ratio.
Figure 3.9 Tariffs and NTMs coverage ratios
Argentina
Bangladesh
Bolivia
Brazil
Burundi
Cambodia
Chile
Colombia
Ecuador
Egypt
European Union
Indonesia
Japan
Kenya
Lao PDR
Lebanon
Madagascar
MauritiusMexico
Morocco
Paraguay
Peru
Philippines
Senegal
South Africa
Syria
Tanzania
Uganda
Uruguay
.2.4.6.81
CoverageRatio
0 5 10 15
Tariff-import weighed
Although Figure 3.9 shows a high degree of dispersion, it also shows a clear
positive correlation between tariffs and NTMs. The countries which apply
more restrictive traditional trade policies are also those where imports are more
affected by NTMs.
Similar conclusions are drawn by the correlation of tariffs and the number
of products affected by NTMs. Figure 3.10 shows the correlation between the
average number of NTMs at the chapter level and the tariff. The figure shows a
stronger positive relationship indicating that countries where tariffs are higher
also apply a larger number of NTMs per product.

Figure 3.10 Correlation of NTM pervasiveness with MFN tariffs (simple and
import weighted ).
Argentina
Bangladesh
Bolivia
Brazil
Burundi
Cambodia
Chile
Colombia
Ecuador Egypt
European Union
Indonesia
Japan
Kenya
Lao PDR
Lebanon
Madagascar
Mauritius
Mexico
Morocco
ParaguayPeru
Philippines
Senegal
South Africa
Syria
Tanzania
Uganda
Uruguay
11.522.53
AveragenumberofNTMsperproduct(Chapter)
0 5 10 15
Taken together, these results indicate that protectionist tariff policy is often
paired with more regulated NTMs regimes. To better explore whether NTMs
are used in addition to tariffs to protect specific sectors, one needs to assess their
relationship at the product level. Figure 3.11 illustrates the relationship between
NTMs and tariffs across economic sectors.
Figure 3.11 Correlation of NTM pervasiveness with MFN tariffs (simple and
import weighted ), by product
Live animals
Vegetable products
Fats and Oil
Processed food
Minerals products
Chemical products
Rubber and Plastics
Raw hide and skins
Wood
Paper
Textile
Footwear
Stone and Cement
Base MetalsMachinery Electrical Equipment
Motor Vehicles
Optical and Medicals instruments
Miscellaneous goods
0
2
4
6
8
AveragenumberofNTMs(1
-digit)
0 5 10 15 20

Also in this case the correlation is positive, suggesting that most regulated
economic sectors are also those where tariffs are higher.
3.5 Conclusions
This study made use of data newly collected by UNCTAD and the World Bank
to investigate the use of NTMs in about 26 countries. The analysis explored the
incidence of various types of NTMs both across countries and economic sectors.
The empirical approach consisted of simple inventory methods: frequency
indices and coverage ratios.
Although our results have to be taken as mainly descriptive, preliminary and
not to be generalised given the limited number of countries covered by the data,
they reveal some important issues. The results find that the incidence of NTMs
varies considerably across countries, across economic sectors and across types
of NTMs. Across countries, overall inventory measures range from less than
10% to more than 90% of products or trade covered by NTMs.
Regarding the incidence of technical measures (SPSs and TBTs), these are
found to be widely used. TBT are found to affect a large share (about 30%)
of international trade. SPS are also frequently used, but they are exclusively
related to agriculture and food products. Given the more limited scope for SPS,
they affect only about 15% of trade but more than 60% of agricultural products.
The large incidence of SPS and TBT raises concerns for developing countries’
exports. Although these measures are not protectionist in intent they often result
in diverting trade from developing countries where production processes and
certification bodies are inadequate, or where the cost of compliance to these
measures is higher. In practice, SPS and TBT may erode the competitive
advantage that low-income developing countries have in terms of labour costs
and preferential access.
The use of non-technical measures varies greatly across countries and
economic sectors. Among these measures the use of quantity controls has
increased but they are now largely limited to non-automatic licences while the
use of quotas has declined since most of them were made illegal by WTO rules.
As a whole, quantity control measures affect about 16% of products and 20% of
trade. Pre-shipment inspection requirements affect about 11% of trade. These are
implemented especially in low-income countries to help custom administrations
in the correct evaluation of imports and their proper taxation. Price control
measures are only rarely used and affect less than 5% of trade and only 2% of
products. Finally, the results suggest the presence of correlation between the use
of NTMs and traditional forms of trade policy. Countries that apply higher MFN
tariffs are also those that have a larger number of products and a larger extent
of imports affected by NTMs. This may indicate that NTMs have been used, at
least to some degree, as substitutes to tariffs in order to continue protecting key
economic sectors in spite of the tariff liberalisation of the last 10 years.

References
Anderson, J and Van Wincoop, E (2004), ‘Trade Costs’, Journal of Economic
Literature, 42(3), pp 691–751.
Beghin, J C (2006), ‘Nontariff Barriers’, Center for Agricultural and Rural
Development (CARD) Publications 06-wp438, Iowa State University.
Boughner, D S, de Gorter, H and Sheldon, I M (2000), ‘The Economics of Two-Tier
Tariff-Rate Import Quotas in Agriculture’, Agricultural and Resource Economics
Review, 29(1), pp 58–69.
Djankov, S, Freund, C and Pham, C S (2010), ‘Trading on Time’, The Review of
Economics and Statistics, 92(1), pp. 166–73.
Hoekman, B, and Nicita, A (2011), ‘Trade Policy, Trade Costs and Developing
Country Trade’, World Development, 39(12), pp 2069–2079.
Maskus, K, Otsuki, T and Wilson, J S (2005), ‘The Costs of Compliance with
Product Standards for Firms in Developing Countries: An Econometric Study’,
Working Paper 3590, World Bank, Washington DC.
Obstfeld, M and Rogoff, K (2000), ‘The Six Major Puzzles in Macroeconomics: Is
There a Common Cause?’ NBER Macroeconomics Annual.
Trefler, D (1995), ‘The Case of the Missing Trade and Other Mysteries’, American
Economic Review, 85(5), pp 1029–46.
UNCTAD (2010), ‘Non-Tariff Measures: Evidence from Selected Developing
Countries and Future Research Agenda’, Developing Countries in International
Trade Studies, United Nations Conference on Trade and Development, Geneva.
UNCTAD (2012), ‘Non-tariff measures to trade. Economic and policy issues for
developing countries’, Developing Countries in International Trade Studies,
United Nations Conference on Trade and Development, Geneva.

4
AVEs in TRIST
Olivier Cadot and Olivier Jammes
4.1 Introduction
Partial-equilibrium (PE) simulation models have been used in many different
contexts to provide orders of magnitudes for changes in trade flows and trade
taxes as a result of policy changes. Compared to general-equilibrium models, PE
ones ignore cross-market transmission mechanisms and feedbacks but permit
much disaggregated analysis and thus reduce aggregation biases.
Up to now, PE simulation models have covered only tariff changes. This was
very useful in the context of regional or bilateral negotiations like the Economic
Partnership Agreements (EPAs) between the EU and ACPs . However, as
substantial progress has been made in reducing tariffs, be they MFN or
preferential, tariff reductions have lost some of their salience in negotiations.
By contrast, the agenda of NTM streamlining, harmonisation and in some
cases elimination is growing in importance. In some regional blocs like the East
African Community (EAC) or the Common Market for Eastern and Southern
Africa (COMESA), specific NTM-streamlining agendas are being put in place.
At the same time, as consumers worldwide become increasingly intolerant of
hazardous products, standards proliferate. Given that NTMs are product-specific
instruments, using PE analysis to assess their effects on trade flows and border-
tax revenue is a natural extension of existing models.
The information generated by the econometric estimation of ad valorem
equivalents (AVEs) can be fed into PE models to make it possible to treat them
like tariffs. However, subtle difficulties arise depending on the type of NTMs.
Some of them, like licences or permits imposed to domestic importers, generate
costs borne by consumers after the imported goods have cleared customs, in
which case their CIF import value is not directly affected. By contrast, some
others, like standards, may impose compliance costs to producers that may be
passed on, in full or in part, into the CIF value of imports. In such cases the
treatment must be different than for tariffs.
We use the TRIST model as a vehicle to illustrate some of the difficulties
involved. In particular, we show with a numerical example that eliminating a
standard may lead to reduced fiscal revenue fromVAT and border taxes. However,

the effect on the tax base is less than the reduction in the compliance cost (the
NTM’s AVE) because direct effects and demand effects work at cross purposes.
The direct effect of eliminating the compliance cost reduces the CIF value of
imports and thus the tax base, but the price reduction raises import volumes and
thus the tax base. If the elasticity of import demand is more than one, eliminating
the standard will raise the tax base. At the other extreme, if it is close to zero, the
tax base will shrink by close to the NTM’s ad valorem equivalent.
4.2 Understanding the incidence of NTMs
4.2.1 Dissipative versus domestic-capture NTMs
NTMs are heterogeneous in their type, effect and incidence. Beyond the
descriptive categories spelled out in the 2009 MAST classification, the economist
is interested in distinguishing them in terms of their incidence (on producers or
distributors/consumers), their effects on production costs and competition, and
their impact on measured trade flows.
To lay out the issues intuitively, consider first a tolerance limit on pesticide
residuals in food that raises the marginal cost of production. Under mill pricing,
producers will pass on the extra cost onto FOB prices. Under other pricing/
competition assumptions, they may not be able to do that and may absorb part of
the cost, but let us assume that this is not the case. The FOB and CIF value of the
importing country’s imports will then go up through a direct effect, as if the NTM
had a trade-enhancing value, although it is not the case, the upward effect on
trade being a pure price one. This price increase, in turn, will trigger substitution
and demand effects. Substitution between national varieties will not be affected
if the NTM is imposed on an MFN basis, which is usually the case. However,
there may be substitution between products (at a higher level of the consumer’s
budgeting decision), a demand effect that will work at cross purposes with the
increase in trade value. Under mill pricing, the quantity demanded will go down
along the demand curve as the consumer price goes up in the importing country,
the CIF price (the one measured by trade statistics) remaining unchanged. Thus,
the direct effect on import value will be a pure price effect whereas the indirect
effect will be a pure quantity effect. The net effect depends on the elasticity of
import demand. We will call this ‘case 1’.
Consider now a licensing scheme for importers. The CIF price of imports may
or may not be affected directly, depending on the licence-allocation mechanism
and market structure. Suppose first that import licences go to the importing
country’s domestic agents who purchase at arm’s-length from exporters. Under
those circumstances, there is no direct effect on the CIF price of imports.
However, there is an indirect demand effect as the price faced by consumers in
the importing country includes the cost of licences. CIF import values will then
unambiguously go down. We will call this ‘case 2’.
Suppose now that licences go to exporters, who may have their own
distribution networks in the importing country. The situation is then akin to a

AVEs in TRIST 83
VER , with quantities going down but prices going up. The price-raising effect is
observationally equivalent to the direct effect we had in the case of a cost-raising
NTM, although it is now due to increased market power rather than higher costs.
The quantity-reducing effect is equivalent to our former indirect effect, although
it is now the mechanical consequence of the restricted number of licences.
Again, the effect on measured CIF import values depends on the elasticity of
import demand. This case being equivalent to case 1, we will squarely ignore it.
Table 4.5 proposes a breakdown of NTMs by type, corresponding to the two
polar cases identified above. A ‘one’ in the first column means the NTM falls
in case 1, a ‘one’ in the second column means it falls in case 2 (meaning that
we have assumed that licences went to importers). The overwhelming majority
of case 1 NTMs are SPS or TBT measures (codes A and B respectively in the
MAST nomenclature).
4.2.2 How effects vary by NTM type
Assume that imports of variety i (from country i) of a certain good into country
d (for destination) are affected by an NTM imposed d on an MFN basis with ad
valorem equivalent (AVE) aj
, a tariff with ad valorem rate τid
, and a domestic tax
with ad valorem rate τD
. There is no transportation cost. For now, we assume that
the AVE exists and can be estimated (more on this below).
With monopolistic competition and a constant marginal cost, producers charge
a ‘mill price’, that is, the same price to all destinations, to which trade costs
are applied multiplicatively. This generates a very simple relationship between
consumer and producer prices, but the relationship is complicated here by the
fact that NTMs can take one of the two broad types described earlier.
Let a type-1 NTM represent an SPS standard that raises the marginal cost of
production in the exporting country (case 1), and a type-2 represent a licensing
scheme that raises the cost of distribution in the importing country (case 2).
Case 1: Dissipative NTMs
Suppose that the producer of variety i can run two independent lines of
production, one satisfying d’s standard with production cost p*
i (1 + ad
), and one
not satisfying it, with production cost p*
i. Then p*
i is the world price of variety i
(when exported to destinations other than j) whereas p*
i (1 + ad
) is the CIF price
(before tariffs and domestic taxes) of that same variety when shipped to d. The
domestic price of variety i in country d will be
( )( )( )*
1 1 1 ,id i d id dp p a τ τ= + + + (1)
which is easily generalised to the case of multiple domestic taxes. Import values
reported in COMTRADE will be
( )*
*
1 when
otherwise
i i
i
i i
p a q d
M
p q
 + =
= 

 


 (2)
otherwise

where qil
is the (unobserved) volume of imports. This means that the direct
effect of the NTM’s elimination will be to reduce the observed value of exports
from i to d. However, of course demand effects will kick in and work at cross
purposes. Note that, as product standards such as SPS are typically imposed on
an MFN basis, there will be no substitution (trade-diversion) effects from their
elimination; only trade creation (or trade destruction if the NTM is put in place
rather than eliminated).
Case 2: Domestic-capture NTMs
In this case, the producer of variety i runs a single line of production and is not
concerned (because of mill pricing) by the NTM, which affects the consumer
price only. The domestic price of variety i in country d will now be
( )*
1 .id i id d dp p aτ τ= + + +
(3)
That is, the NTM’s AVE enters additively with the tariff because now, unlike in
case 1, the tariff is applied to the pre-NTM CIF price of the import. The NTM’s
AVE raises the imported good’s price after the border has been crossed, when the
distributer adds the cost of the licence cost to the good’s internal price. Internal
taxes are then applied to the price inclusive of both tariff and NTM AVE.
Unlike in case 1, the CIF value of imports is given by
*
id i idM p q=

(4)
and the NTM’s elimination will have no direct effect on observed trade values,
although it will have an indirect effect through a change in domestic demand
in country d following the change in consumer price. In that case the observed
effect of the elimination of an NTM will be like that of the elimination of a tariff
or domestic tax.
NTM combinations
The extension to several NTMs applied simultaneously to the same good is
straightforward. With n dissipative NTMs and m domestic-capture ones, the
domestic price becomes
( )( )( )*
1 1
1 1 1
n m
id i kd id kd dk k
p p a aτ τ= =
= + + + +∑ ∑
(5)
That is, relative to the tariff, all dissipative NTMs enter multiplicatively whereas
domestic-capture ones enter additively.
Suppose now that there are two NTMs, a dissipative one, a1
, and a domestic-
capture one, a2
, and that the former is eliminated in importing country d. For
simplicity of notation, let us drop the importing-country index d and introduce
superscripts 0 and 1 for before and after the change respectively. Then

AVEs in TRIST 85
( )( )
( )( )( )
1 2
0
1 2
1 1
.
1 1 1
ii
i i
a ap
p a a
τ τ
τ τ
− + + +D
=
+ + + +
(6)
We will use this formula later on when discussing how to introduce changes in
NTMs in the TRIST model.
4.3 Estimating AVEs
The econometric estimation of AVEs is described in detail elsewhere in this
book. For our purposes here, let us only briefly recall the intuition of AVEs. Let
Mid be the value of d’s imports of that good (for which we use no index in order
to avoid cluttering the notation) from i, as measured in COMTRADE statistics.
Consider case 2, which is conceptually the simplest, and assume that there is
only one NTM applied to the given good worldwide (the equations below can be
easily expanded to the case of several NTMs); let
1 if NTM is applied
0 otherwise
dI

= 

and τid
be the ad valorem rate of a tariff applied by j on imports of the good from
i. Let also εd
be the elasticity of import demand (available from Kee et al, 2009).
Assume that the import-demand function, after log-linearisation, takes the form
lnMid = a0 + α0Ci+ α2Cd + a3Id + εdln(1 + τid) + uid
where Ci
and Cd
are vectors of exporter and importer characteristics respectively
(comparative advantage and the like). Estimation issues will be discussed later
on.
The NTM’s AVE can be retrieved algebraically from estimated coefficients
as follows. Exponentiating (7), evaluating it at I = 1 and at I = 0 respectively (ie
with and without the NTM), and taking the ratio gives
31
0
id I
id I
M
e
M
a=
=
=

(8)
Similarly, exponentiating again (7), evaluating it with a tariff at rate ad
, with one
at rate zero, and taking the ratio gives
( )ln 1
0
j d d
id a a
id
M
e
M
τ ε
τ
= +
=
=

(9)

By construction, the NTM’s AVE is the value of ad
for which (8) and (9) are
equal. This gives
( )3 ln 1d daa ε= +
or
3 /
1.d
da ea ε
= −
(10)
In case 1, things are more complicated. With a direct effect of the NTM on
trade flows, α3
no longer gives a consistent estimate of the NTM’s effect on
log imports. To see this intuitively, observe that in order to estimate the effect
of a trade barrier imposed at the border, one would want to correlate variations
in import value with variations in the level or type, or simply presence of the
barrier. With a flat export supply curve, the value of imports without the barrier
is proportional to its volume, as in panel (a) of Figure 4.1, so the estimated
relationship is the true one. However, with type-1 NTMs, the value of imports
is inclusive of the compliance cost. Suppose now that the elasticity of import
demand is close to one. Then whether or not the NTM is applied, and whether or
not it reduces import volume, import value is constant, as in panel (b) of Figure
4.1. Correlating it with the NTM’s dummy will return no effect (α3
0), even if
the NTM does reduce import volumes. In other words, the relationship in (7)
should now be expressed in terms of latent, pre-compliance cost imports in order
to preserve the proportionality between volumes and values.
Figure 4.1 How NTMs affect import volumes, values and unit values
Import
quantities
Import unit
values
Import
demand
CIF unit
value, before
and after
NTM
Volume reduction
Import unit
values
Import
demand
CIF unit
value after
NTM
CIF unit
value before
NTM
Volume reduction
Import
quantities
Import value before NTM
Import value after NTM
4.4 NTMs in TRIST
4.4.1 TRIST: The basic model
TRIST is a partial-equilibrium model of the effect of tariff changes on import
values based on a three-stage budgeting process. At the broadest level, the
consumer is assumed to set the amount consumed of each good depending on
his income and their relative prices. Within each good’s allocation, the consumer
decides how much goes to the domestic variety versus imported ones. Finally,

AVEs in TRIST 87
within the allocation of foreign varieties, the consumer decides how much goes
to each national variety depending on their relative prices. The solution method
consists of working bottom up, first deciding on the share of each foreign variety
depending only on their relative prices for a given amount spent on foreign
varieties as a whole; then, deciding on the share of foreign versus domestic
varieties depending on an index of the aggregate relative price of foreign
varieties, for a given amount spent on that good; and finally setting the share of
each good.
Let q0
i and p0
i be respectively the import quantity and price of variety i before
the tariff change, q1
i and p1
i its quantity and price after the tariff change, and Dpi
= p1
i - p0
i . Let σ be the elasticity of substitution between foreign varieties (chosen
by the user) and μ the elasticity of substitution between the foreign varieties as
a whole and the domestic one. Let also pm
be an import-weighted average of
the price of foreign varieties and qm
the total imported quantity of the good (all
varieties combined), so that sm
= qi
/qm
is the share of variety i in total imports of
the good and
0
0 0
.jm
jj
m j
pp
s
p p
DD
= ∑
Finally, q is the total consumption of the good (not indexed for simplicity) in
all its varieties (imported and domestic). Using this notation, the lowest step
(substitution between foreign varieties) is given by the following equation (see
Brenton et al, 2009):
( )
0
1 0
00, constant
0
1
1 /
.
m
jj i
i iq q
ij jj
i
q p
q q
pp p
q
σ
σ
a
 D
= + 
+ D  
=
∑
∑

(11)
The intermediate step (substitution between foreign varieties and the domestic
one) is given by
1 0
0
1 m
m m
m
p
q q
p
µ
 D
= + 
 

(12)
and the change in the overall share of imports in consumption of the good is
allocated to each foreign variety in proportion to its initial share; that is
1 1 0
constant , constant
0 0
m
i i i mq q q
i i m
q q s q
q s qa
= + D
= + D

(13)

Finally, let sm
= qm
/q be the share of imports in consumption of the good, and
m mp s pD = D
be the change in the aggregate price of imports weighted by the share of imports
in consumption. In the uppermost step (overall consumption of the good),
1 0
m m mq q s q= + D
and the variation in qm due to that of the overall level of the good’s consumption
is, again, imputed to each imported variety in proportion to its initial share:
1 1 0
constant
0 0 0
.
j i mq
i i m m
q q s q
q s q s qa
= + D
= + D + D

(14)
This gives the final demand effect of a change in the price of imported variety i
on its import volume.
4.4.2 Integrating NTM AVEs
Basic issues
As we already noted, the introduction (or removal) of NTMs has both a direct
and an indirect effect on import values. The direct effect is observed only in the
case of dissipative NTMs which inflate the value of imports by the AVE of their
compliance cost. It is inexistent in the case of domestic-capture NTMs, since
they affect only domestic prices after import.
Formula (14) gives the variation in import volumes attributable to a change in
the price of imports through demand (indirect) effects. Integrating the imposition
or removal of NTMs in this formula implies no change in the model’s structure,
but a change in the construction of the exogenous import-price variation to
accommodate not only tariffs and other border taxes, as in the baseline model,
but also NTMs. This is done simply by using (6) for D pi /p0
i in the equations
above.
However, simulating changes in dissipative NTMs raises a specific problem,
namely, how to treat their direct effect on import values. For instance, the
elimination of a dissipative NTM reduces the value of imports by the amount of
the compliance cost (theAVE), as per (6). How the direct effect should be treated
in the simulation depends on its objective. If the objective is to assess the fiscal
effect of NTM elimination, then the reduction in the value of imports inclusive
of compliance costs is relevant, since this value is the tax base for tariffs and
other border taxes. If the objective is to assess the effect of NTM elimination on
import volumes, then the direct effect should be ‘filtered out’ of the simulation.
This can be done by defining as baseline imports their (unobserved) value before
compliance costs. That is, start from the case in (6) with one dissipative NTM
(with AVE a1
) and possibly a domestic-capture one, assume that the dissipative

AVEs in TRIST 89
one is eliminated, and let Mi
= pi
qi
. Baseline import value should then be the
observed value deflated by the AVE a1
:
0
0 0
1
11
i
i i iI
p
M M q
a=
= =
+

The extended model is written so as to easily accommodate this alternative way
of measuring initial imports.
Incomplete pass-through
Suppose now that there is incomplete pass-through of the AVE of NTMs. In the
case of dissipative NTMs, observed trade values will be inflated by less than
the full compliance cost. In the case of domestic-capture NTMs, they are never
inflated so nothing changes. We now define a pass-through parameter ‘bridging’
the cases of dissipative versus domestic-capture NTMs. Suppose that a fraction
λ of an NTM’s compliance cost is borne by foreign producers and passed on to
the CIF price of imports, whereas a fraction 1 – λ is borne by domestic importers
in the destination country, and hence does not affect border prices. Alternatively,
one can think of the fraction 1 – λ of the compliance cost as being absorbed by
the producers. Combining (1) and (3) with weights λ and 1 – λ respectively gives
( ) ( )*
1 1 1id i d id d dp p a aτ λ τ= + + + +  

(15)
4.4.3 An illustrative calculation
In order to get a feel for how the introduction of NTM AVEs affects fiscal-
revenue simulations in TRIST, consider the following illustrative calculation
in which a country eliminates a ‘case-1’ NTM, say a standard, imposed on all
import sources of a given product. Let the home country import $500 from each
of two exporting-country groups, labelled 1 and 2, and $1,000 from the rest of
the world. Assume that the elasticity of substitution between national varieties is
–1.5 and the price elasticity of import demand at the product level is –1, both in
algebraic values of course.
The raw data from the importing country is shown in Table 4.1, where STR
stands for import tariff, CTR for for a para-tariff charge, and excise and VAT are
self-explanatory.
Table 4.1 Illustrative case – the raw tariff data in importing country
Imports
CIF
NTM
AVE
NTM
incidence
STR CTR Excise VAT
ROW 1000.0 0.20 1 170.0 0.0 232.0
Country Group 1 500.0 0.20 1 85.0 0.0 117.0
Country Group 2 500.0 0.20 1 70.0 0.0 112.0
Totals 2000.0 325.0 0.0 461.0

From the data in Table 4.1, several transformations can be done, shown in Table
4.2. The first column ‘deflates’the CIF value of imports by the AVE of the NTM,
which is assumed estimated outside of the model (and is therefore part of the
raw data in Table 4.1). This gives lower values that we call ‘IF’, that is, including
insurance and freight but excluding the cost of complying with the standard.
Second, ad valorem rates can be calculated for all taxes by dividing receipts by
CIF import values using Table 4.1’s data. Finally, total fiscal revenue, shown
in the last column, can be calculated equivalently by summing the tax receipts
in Table 4.1 or by applying the ad valorem rates calculated in Table 4.2 to CIF
import values.
Table 4.2 Ad valorem rates and pre-reform fiscal revenue
Imports
CIF
Imports
IF
STR CTR Excise VAT
Total fiscal
revenue
ROW 1000.0 833.3 0.17 0.00 0.198 402.0
Country Group 1 500.0 416.7 0.17 0.00 0.200 202.0
Country Group 2 500.0 416.7 0.14 0.00 0.196 182.0
2000.00 1666.7 786.0
The simulation consists of eliminating the standard. This will involve no
substitution effects between national varieties since the standard was applied on
an MFN basis. However, it will reduce the CIF price of imports since it embodied
the compliance cost. The reduction in the price (and total value) of imports will
have two effects: (i) it will raise import volumes; (ii) it will reduce fiscal revenue,
since taxes are levied on the CIF value of imports. Demand effects are shown in
Table 4.3.
Table 4.3 Illustrative case: demand effect of eliminating a standard
p/p*
Price decrease Demand variation
pre post
ROW 1.682 1.402 -0.167 0.167
Country Group 1 1.685 1.404 -0.167 0.167
Country Group 2 1.637 1.364 -0.167 0.167
The ratio of domestic (post-taxes) to foreign prices goes down from the values in
the first column to those in the second, with an equivalent price decrease due, in
all three cases, to the elimination of the 20% compliance cost (the proportional
decrease is smaller for more heavily taxed varieties since the standard’s AVE
accounts for a smaller part of the total price wedge). As a result, prices decrease
by 16.7% in all three cases, so with a unit demand elasticity, volumes go up by
just 16.7%.
Fiscal revenue, by contrast, goes down as taxes are levied on the CIF value of
imports, which goes down by the amount of the eliminated AVE. This is shown
in Table 4.4.

AVEs in TRIST 91
Table 4.4 Illustrative case: fiscal revenue effect of eliminating a standard
Imports
IF = CIf
IF
NTM
AVE
NTM
incidence
STR CTR EXCISE VAT
Total
fiscal
revenue
%
variation
in fiscal
revenue
ROW 972.2 - 0 165.28 0.00 225.56 390.8 -0.0278
Country Group 1 486.1 - 0 82.64 0.00 113.75 196.4 -0.0278
Country Group 2 486.1 - 0 68.06 0.00 108.89 176.9 -0.0278
1944.4 448.19 764.2 -0.0278
The decrease in tax revenue is –2.78%. The direct effect of eliminating the
standard’s compliance cost, which reduces the tax base, is largely offset by the
demand effect shown in Table 4.3 which raises import volumes (and values). On
net, the negative effect on tax revenue is only a fraction of the direct effect, but
this is of course a result of the elasticity assumptions. With a demand elasticity
of 0.5, the tax-revenue loss jumps to –9.72%. As the price elasticity of demand
goes to zero, only the direct effect remains and the tax-revenue loss tends to the
negative of the NTM’s AVE.
In the (rare) case of non-MFN NTMs, substitution effects would magnify
revenue losses, since importers would turn to the partner exempted of the cost-
raising NTM, thus shifting the tax base from high-price sources to low-price
ones.
4.5 Conclusion
The current version of the TRIST model makes it possible to assess, ex ante, the
effect of reforms in border taxes on trade flows and fiscal revenue in a partial-
equilibrium but highly disaggregated framework. The partial-equilibrium aspect
eliminates all cross-market interactions and feedbacks but is the price to pay for
the disaggregation, which eliminates aggregation bias.
Including in the model the AVEs of NTMs is a natural extension, provided
that one uses it carefully. As discussed in the introduction, NTMs are complex
measures which serve many purposes, and focusing exclusively on their trade
effects would be at best partial and at worse misleading, if it led to policy
prescriptions for elimination disregarding the protection of local public goods.
With all the necessary caveats, the NTM-augmented version of TRIST
proposed in this chapter, which has already been developed and is available
for use, makes it possible to assess the effects – sometimes surprising – of the
elimination of non-tariff border measures. For instance, if elimination reduces
the import price of goods, the tariff base goes down, which reduces tariff revenue
for a given volume of imports through a direct effect and raises it through an
indirect effect whereby import demand goes up. TRIST’s flexibility makes it
possible to examine these effects product by product at a highly disaggregated
level, providing a flexible and simple tool for governments to assess precisely
the effects to be expected from the elimination of reform of single measures.

References
Brenton, P, Saborowski, C and von Uexkull, E (2009), ‘What explains the low
survival rate of developing country export flows?’, World Bank Policy Research
Working Paper No. 4951.
Kee, H L, Nicita, A and Olarreaga, M (2009), ‘Estimating Trade-Restrictiveness
Indices’, Economic Journal, 119, pp 172–99.

PART II
Standards and Technical Regulations:
Trade Barriers or Facilitators?

5
Reputation Matters: Spillover Effects in
the Enforcement of US SPS Measures
Marie-Agnès Jouanjean, Jean-Christophe Maur and Ben Shepherd1
5.1 Introduction
Non-tariff measures have become progressively more important as trade policy
instruments as applied tariff rates have fallen across the world in recent years.
From a development perspective, technical regulations and product standards
are a particularly important type of non-tariff measure because they highlight
the fact that the favourable market access accorded under duty and quota-free
preferential schemes remains conditional on compliance with regulations in
areas such as consumer safety. Previous research shows that product standards
and technical regulations in the large, developed markets can have two
contradictory sets of effects for developing country exporters. On the one hand,
the costs of compliance – retooling, product redesign, testing and certification
– can be substantial enough to keep many small and medium enterprises out of
international markets, thereby affecting the pattern of international specialisation
(eg Essaji, 2008). But at the same time, foreign standards can also provide the
impetus for firms and sectors to upgrade production technologies and realise
beneficial productivity gains (eg Maertens and Swinnen, 2009). The question of
which types of standards tend to promote which set of effects is clearly of vital
policy importance to developing country exporters. The issue of how best to
direct technical assistance resources so as to support the upgrading of standards
systems and development of compliance mechanisms in developing countries is
also an important part of broader Aid for Trade discussions.
Most previous work on standards and technical regulations has focused on the
rules themselves, rather than their application or enforcement through specific
at-the-border mechanisms. There are a number of recent exceptions, however.
Karov et al (2009) focus on identifying the trade impacts of US SPS regulations
at the product-country level by analysing the effects of treatment requirements
and grants of new market access. Similarly, Alberini et al (2005) examine
1 The authors are grateful to Olivier Cadot for helpful comments and suggestions on a previous
draft. The findings, interpretations, and conclusions expressed in this paper are entirely those of the
authors. They do not represent the view of the World Bank, its Executive Directors, or the countries
they represent.

implementation of the FDA’s seafood HACCP programme using a dataset
of plant inspections. Neither paper, however, deals with the import refusals
mechanism that is the focus of the present paper. Buzby et al (2008) and Buzby
and Roberts (2010) analyse similar data on US import refusals to that used in
the present paper, but only provide descriptive statistics. Baylis et al (2010) use
data on EU import refusals in a gravity model to show that they tend to decrease
trade. However, they do not examine the determinants of import refusals, and
in particular the potential for reputation effects, which is the focus of our paper.
Finally, the first stage of the empirical approach taken by Cadot et al (2010) uses
EU import alerts – closely related to refusals – as the dependent variable, but the
cross-sectional setting of their regressions means that they are unable to account
for reputation effects of the type we are interested in here.
In this paper, we use newly collected data to focus on one important example
of the de facto implementation of product standards: refusal of entry into the
US market for imported foods. The advantage of looking at measures such as
import refusals is that they are implemented on a country- and product-specific
basis, rather than being the same de jure for all exporters. Most standards and
regulations are effectively most-favoured nation (MFN) measures, which makes
it difficult to identify their effects on exporters by exploiting cross-country
variation in outcome measures, such as trade flows. Focusing on a country- and
product-specific measure, such as import refusals, provides a potentially much
richer source of data in which identification can be based on cross-country as
well as cross-product and through-time variation.
In addition to exploiting new data on US import refusals, this paper makes
a number of contributions to the existing literature. First, we provide some
of the first explicit evidence of reputation effects in the enforcement of SPS
measures. Specifically, we show that even after controlling for the size of import
flows, a history of SPS compliance is associated with fewer current import
refusals. Second, we show that it is not just reputation for a particular product
that matters, but sector-wide reputation (ie, there are cross-product spillovers
in SPS enforcement): a product tends to suffer from more import refusals if
closely related products are also subject to refusals. Third, we investigate, and
find evidence to support, the hypothesis that the reputation of neighbouring
countries also matters for SPS enforcement. Our results suggest that imports are
more likely to be refused if the same product from neighbouring countries has
also been subject to refusals. We interpret this finding as evidence in favour of
geographical spillovers in SPS enforcement.
To our knowledge, the three reputation effects we are investigating have
not been explicitly considered before in the literature. Baylis et al (2010) use
similar data to ours to analyse the determinants of US import refusals over
the period 1998–2004. Their core hypotheses are that: countries with greater
experience exporting food to the US experience fewer refusals (the ‘learning
curve’effect); and import refusals are subject to political pressure (the ‘standards
for sale’ effect; see Grossman and Helpman, 1984 for the case of tariffs). The
data indeed support the second hypothesis: standard political economy measures

Spillover Effects in the Enforcement of US SPS Measures 97
such as the level of lobbying activity, decreases in US employment, and anti-
dumping actions are significantly associated with a higher number of refusals.
However, the authors find that the data do not support their first hypothesis:
more recent exporters actually face fewer refusals than established ones, even
after controlling for export volumes. They speculate that this result may be
due to a direct reputation effect, namely that enforcement resources tend to be
concentrated on past violators. They do not pursue the point, but it is one that
we take up here, using a different empirical specification that can better identify
the effect, which we find to be highly significant. This paper therefore develops
the intuition in Baylis et al (2010) in relation to individual reputation effects, but
also extends it to include the possibility of reputational spillovers from related
products and sectors.
The paper proceeds as follows. In the next section, we provide an outline of the
US import refusals regime. Based on that description, Section 5.3 presents our
dataset, focusing on the new import refusals data. We present some preliminary
analysis that supports our hypotheses using descriptive statistical techniques,
then proceed to develop a fully specified econometric model of import refusals.
Section 5.4 presents and discusses the results from our model, and conducts
robustness checks. Section 5.5 concludes with a discussion of policy implications,
and avenues for further research.
5.2 The US import refusals regime
Imports of food into the US have significantly increased in volume, variety and
also diversity of origin. Food and agricultural imports doubled from $41 billion
in 1998 to $78 billion in 2007 (Buzby and Roberts, 2010). Fruit and vegetable
and fish product imports have followed the same trend: fruit and vegetable
imports increased more than five-fold from $811.5 million in 1989 to $4.28
billion in 2007, a 10.5% rate of annual growth (Karov et al, 2009). The value of
imports of edible fish products (including canned fish) increased 2.5 times from
$5.4 billion in 1988 to $13.7 billion in 2007. Between 2007 and 2010, imports
augmented further to reach $14.9 billion.2
Imports as a share of US consumption have been rising rapidly too. Imported
fruits and nuts represented 33.9% of total US consumption in 2004, up from
13.4% during the 1981–5 period. For vegetables, imports increased from 4.5% to
10.9% of US consumption over the same period (Buzby et al, 2008). Similarly,
while less than 50% of US fish consumption was imported products in 1980,
the market share of imported products was over 75% in 2003 (Allshouse et al,
2003).
According to Buzby and Roberts (2010), since 1990, there have been 77 new
entrants in the market for fresh and minimally processed fruits, and 40 new
entrants for fresh and minimally processed vegetables. This means for American
2 Source: Fisheries Statistics Division of the National Marine Fisheries Service (NMFS).

consumers a higher variety of products (eg, exotic fruits), and availability all
year round.
This rapid increase, which outstrips the overall growth of consumption
in the US, and most likely efficiency improvements from foreign suppliers,
demonstrates that improvements in market access have certainly played a crucial
role.
Market access conditions for fruit and vegetables and fish, however, noticeably
differ.3
Seafood trade imports into the US are generally less restricted than other
food products; there are no quotas and heavy duties on imports. Ninety per cent
of imports of fish enter duty free compared to 24% for fruits and vegetables; the
average tariff on fish products is 1.0% against 4.9% for fruits and vegetables
(WTO , 2010). Moreover, imports of fruits and vegetables have been historically
highly restricted, with strict regulation of entry eligibility for foreign products,
and for 22 categories of fruits and vegetables the presence of 31 marketing order
regulations specifying how products should be marketed.4
5.2.1 Eligibility to export to the US: Plant health controls and
Memorandum of Understanding for molluscan shellfish
The US Department of Agriculture’s Animal Plant Health and Inspection
Service (APHIS) is the agency charged with regulating entry eligibility into the
US under the authority of the Plant Protection Act. The purpose is to ‘safeguard
U.S. agriculture and natural resources from the risks associated with the entry,
establishment, or spread of animal and plant pests and noxious weeds’ (source:
APHIS website5
).
At the outset, only a share of the world’s production is eligible for export to the
US. For instance, Mexican avocados were banned until 1997. The importation
of fresh apples is permitted from only 17 countries, representing 44% of global
exports of apples, and 15% of global production (USDA ERS , 2010). For
products like fresh olives, dates and figs, none or almost no imports are allowed.6
When access to the US market is permitted, this is often under the condition
of complying with a number of pest-mitigating measures, such as treatments
(eg fumigation), the implementation of risk management systems, geographic
origin or destination restrictions, and pre-clearance requirements (Karov et al,
2009). Standards vary for each country and are recorded in the APHIS Fresh
Fruit and Vegetables Import Requirements (FAVIR),7
a positive list of fruits and
vegetables that have been approved for entry into the US and the requirements
that must be met for admission into the US. By definition, all other products are
banned, although not being on the list does not necessarily mean an outright
3 Which is one of the motivations for focusing on both sectors in our analysis.
4 See http://www.ams.usda.gov/AMSv1.0/FVMarketingOrderLandingPage.
5 See http://www.aphis.usda.gov/.
6 http://www.ers.usda.gov/Data/FruitVegPhyto/data.htm
7 Per 7 CFR 319.56 as well as the quarantine 56 streamline revision APHIS published July 2007.

import ban: it may just reflect the fact that no exporter applied for an importation
permit.8
A summary of FAVIR can be accessed through an online database.9
The control of fish product imports generally focuses on product testing
and inspection. However, raw molluscan shellfish may only be imported into
the United States from countries whose seafood regulatory authority has an
appropriate Memorandum of Understanding with the FDA (Australia, Canada,
Iceland, Japan, Mexico, New Zealand and the United Kingdom).10
5.2.2 Food safety
Food products are the source of numerous food-borne illnesses (due to pathogens,
toxins and chemicals). All food products must be unadulterated (not bear or
contain any poisonous or deleterious substances), be fit for consumption, and
not contaminated or decaying, in order to be allowed for consumption in the US.
At the federal level, there are three agencies involved in the oversight of food and
food ingredients safety: the US Department of Agriculture’s Food Safety and
Inspection Service (FSIS), the Food and Drug Administration (FDA), and the
Environmental Protection Agency (EPA). FSIS ensures the safety of imported
meats, poultry and processed egg products. FDA covers all other products.11
EPA licenses pesticide products and monitors pesticide residues in products. The
FDA enforces compliance with the limits set by FDA by testing US-produced
and imported food. To sum up, FDA supervises pesticide residues, food hygiene,
additives and contamination.
The FDA enforces the 1938 Federal Food, Drug and Cosmetics Act (FDC)
as well as other laws designed to protect consumer health, welfare, and safety.
Under s. 801 of the FDC, products are subject to inspection when imported.
Imported food products are expected to meet the same standards as domestic
products – that is, they must be pure, wholesome, safe to eat and produced under
sanitary conditions. Food imports must also contain informative and truthful
labelling in English.12
Standards mean for fruits and vegetables that FAVIR requirements have to be
met. Another important requirement is that since 1997, producers must follow
FDA’s good agricultural practices (GAP) for the control and management of
microbial food safety.
Likewise, since 1995 fish products imports must meet hazard analysis and
critical control point (HACCP) standards, as must domestic producers. HACCP
standards are also included in the Codex Alimentarius international standard.
Implementation of the standard varies internationally: in the case of the US,
HACCP applies to processors only, in contrast with the EU where it applies to
8 Applying for a permit is a complex process.
9 http://www.aphis.usda.gov/favir/info.shtml
10 http://www.fda.gov/InternationalPrograms/Agreements/MemorandaofUnderstanding/default.htm
11 This includes not only the other food products, but drugs, biologics, cosmetics, medical devices, and
electronic products that emit radiation.
12 http://www.fda.gov/ForIndustry/ImportProgram/ImportProgramOverview/default.htm

the entire supply chain. Other measures applying to seafood include traceability
requirements such as the identity preservation system for molluscans, and
labelling of origin and method of production (wild harvest or farm raised).
Other programmes concerning food products in general are also in place such
as the acidified and low acid canned foods regulations. Acidified and Low Acid
Canned Foods must be manufactured in accordance with FDA regulations. Food
canning establishments must also register with the FDA.
All the measures described above are defined on the principle of national
treatment: importers and domestic producers are subject to the exact same
requirements. There is, however, a significant difference of treatment between
domestic goods and imported products, in implementation of food safety
measures. The Act allows for refusal of imported FDA-regulated products for
‘appearing’ to be adulterated or misbranded. The law is interpreted in a broad
sense as allowing the FDA to make admissibility decisions based not only on
physical evidence such as examination, facility inspection or laboratory results,
but also based on historical data, information from other sources (eg about a
disease outbreak), labelling, and any other evidence.13
Factors such as reputation
can clearly come into play in this decision. In other words, if there is the faintest
suspicion that a product from a given origin will not meet FDA standards, it
can be detained. Therefore the standard of proof for determination of refusal
for food import products is much less strict than for domestic products, which
must be based on an actual violation. Therefore we can form the hypothesis that
refusals may be partly path-dependent (as noted by Baylis et al, 2010) since
past histories of violation from similar products and origins are criteria that may
be used to decide whether there is a suspicion of adulteration of misbranding,
which can in turn justify a refusal.
Baylis et al (2010) explain that this looser requirement may be motivated by
the fact that the FDA has less easy access to means of verification for imported
products: for instance FDA does not have extra-territorial jurisdiction and may
not be able to inspect foreign plants. This is also ultimately motivated, according
to the FDA, by limited resources for physical inspection of products.14
5.2.3 FDA process for importing products
Under the Public Health Security and Bioterrorism Preparedness and Response
Act 2002, the FDA issued regulations in December 2003 requiring two things:
(1) that food facilities (including foreign) are registered with the FDA;15
and (2)
the FDA be given advance notice of shipments of imported food. Depending on
the mode of transport, notices must be submitted between 2 and 8 hours and 5
days prior to arrival. The information required for prior notice (PN) varies, based
13 Presentation by Domenic Veneziano, Director, FDA Division of Import Operations at the Food
Agriculture Border Gateway Summit, 16 January 2008.
14 ibid footnote 10.
15 Registration is for the purpose of collecting information and is not an approval process.

on the type of entry, mode of transportation for entry, and whether the food is in
its natural state.16
The prior notice is filed with customs and the FDA. Upon reviewing the notice,
the FDA can decide to release the product, request additional information or
documents, request physical examination of the product, or recommend detention
of the product. Detention means that in the absence of petition or reconditioning
of the goods from the exporter, the product will not be released into US territory
and will either be re-exported or destroyed within approximately 90 days.
Physical examination entails verification of labelling, of the container integrity,
sampling and verification, and leads to either a recommendation of release or
detention. According to the literature, about 1–2% of all food shipments are
subject to physical examination by the FDA, and a fraction of these are subject
to sampling (Buzby et al, 2008; Baylis et al, 2010).
5.2.4 Alerts
The FDA relies on a system of alerts for particularly sensitive categories of
products in order to help it save and allocate inspection resources. Alerts are
issued when the FDA determines that there is a particular risk associated with a
product, producer/exporter, country or region of origin.
As mentioned earlier, the FDC treats imported and domestic products
differently.Importedproductsthatappeartobeinviolationofimportrequirements
can be refused admission without examination, so-called detentions without
physical examination (DWPE). In most circumstances, alerts determine that
firms and products identified will be subject to DWPE. In this case, the FDA
will automatically detain the concerned products until it is demonstrated to the
FDA that the violation has been remedied. Therefore the burden of proof falls
on the shoulders of the importer or shipper, or the manufacturer/grower of the
product. As noted by Baylis et al (2010) alerts are strikingly rarely changed:
three quarters of alerts in place in 2009 had been in place for more than 10 years,
and a significant portion of them (one quarter of all alerts) for more than 20
years.
Alerts appear to be decided on a similar legal basis to refusals, for example
the standard of ‘appearance’. According to the FDA, alerts are triggered by
historical violations at the following levels: commodities; manufacturers/
shippers; growers; sometimes importers; geographic area; and countries of origin
(Veneziano, 2008). Sources of information come from FDA’s own field offices,
but also foreign inspections and evidence from other countries.17
It is not entirely
16 See (21 CFR 1.281) for details on the required information. The PN requires additional information
to what is normally required under the import process: the country from which the article is shipped
and the crossing location within the port of arrival; revised information is also required regarding the
actual manufacturer, the registration number of the manufacturer and shipper and information about
the grower if known. http://www.fda.gov/Food/FoodDefense/Bioterrorism/ucm083245.htm
17 For instance, beyond publicly available evidence such as refusal data, the EU and US authorities
communicate on certain matters.

clear how criteria leading to an alert differ from those to determine detention
without physical examination, but it is likely that they are closely related. For
instance, previous import refusals are a factor used to justify the creation of an
alert.
Since alerts are used to allocate inspection resources and also decide automatic
detention, and that alerts are themselves decided on the basis of past history
of violation, we can assume that this acts as reinforcing the potential for path-
dependence in the refusal of foods entering the US.
5.3 Methodology and data
As the above discussion demonstrates, US border authorities exercise broad
discretions when implementing the import refusals regime. As previously
noted by Baylis et al (2010), there is a strong possibility of path dependence:
the authorities might look at past patterns of compliance in allocating scarce
enforcement resources, leading to a correlation between past and present import
refusals, even after controlling for other factors. We refer to this as the ‘own
reputation’ effect. In addition, the structure of the US import refusals system
is suggestive of two other effects that might be in operation. One is a ‘sector
reputation’ effect, by which we mean the possibility that import refusals for
a particular product are associated with past import refusals affecting closely
related products. The second is a ‘neighbour reputation’ effect, namely the
possibility that import refusals affecting a given product from one country might
be more likely if neighbouring exporters of the same product have a history of
non-compliance. In the remainder of this section, we outline the data and model
we will use to test for the existence of all three effects.
5.3.1 US import refusals data
This paper uses a new dataset of US import refusals for the period 1998–2008.
It extends the data used in Jouanjean (2011), and covers HS chapters 3 (fish and
crustaceans), 7 (vegetables), 8 (fruits), and 20 (preparations of vegetables, fruits
and nuts). Refusals in these sectors accounted for over 50% of all FDA import
violations over the 1998–2004 period (Buzby et al, 2008). We are therefore
confident that by focusing on these three sectors, we are capturing an important
part of overall import refusal activity in the US. This subsection describes the
US import refusals regime in more detail, focusing on the way in which the data
used here were collected.
Since 1998, the FDA has implemented an automatic system governing the
admission process for FDA-regulated shipments of foreign-origin products
presented for entry into the US. The system, known as the ‘Operational and
Administrative System for Import Support’ or OASIS, is designed to simplify
operations and reduce the time taken for clearance of shipments. It classifies
imported items into different risk categories by tracking historical data on
shipments that were previously refused admission into the US. Again, this

process is suggestive of path dependence and reputation effects of the type that
are the focus of this paper.
The FDA makes refusals information public in their Import Refusal Report
(IRR), which is generated from OASIS data. Reports provide information on the
manufacturer’s name and country of origin, as well as the dates and motives for
the refusal. To gain access to historical refusals data, we submitted a Freedom
of Information Act request in September 2009, which the FDA satisfied by
supplying data in May 2010.
Products in the OASIS system are identified by a specific FDA code. In order
to relate these products to reported trade flows, we constructed a correspondence
mapping FDA codes to the Harmonized System at the 4-digit level. FDA product
codes comprise five elements: Industry Code, Class, Subclass, Process Indicator
Code (PIC), and Product Code. The Industry Code describes the broadest area in
which a given product falls. In this paper, we focus on refusals related to Industry
Codes 16 (fishery and seafood products), 20–22 (fruit and fruit products), and
24–25 (vegetables and vegetable products), see Appendix)
The Class Code defines a narrower category of products that are specific to
each industry. In other words, one Class Code can have a different definition
under various Industry Codes. Class Codes provide two pieces of information.
First, they give more information on the ‘sub-group’to which a product belongs.
In addition, they help clarify the kind of transformation the product has been
subject to.
For convenience, we refer to the class sub-group as ‘Industry Type’ and to
the type of transformation as ‘Class Name’. In the event that a Class is only
assigned to a sub-group of the Industry Code but that no specific transformation
is mentioned (such as Class A in the following examples), we apply a specific
mention ‘Not Further Specified’ (NFS).
The last two codes, Subclass and Process Indicator Codes (PIC), are specific to
each food industry code. The former gives an indication as to the material type of
the container holding the product (metal, glass, etc), and the latter describes the
process used in preparing the product (raw, commercially sterile, pasteurised,
etc). The association of both the Industry and Class Codes to the Product Code
is usually enough to define the product in the HS classification.
Assumptions made in constructing the database
Exporters provide the product code relating to their shipment when they file the
prior notification required by the FDA for any food exports to the US. Exporters
build this code themselves according to FDA recommendations. However, it can
sometimes happen that exporters or importers have a different understanding
of how they should build product codes, and might sometimes lack the relevant
information. This kind of issue is highly unlikely for most of our products of
interest, since they are generally fresh or minimally processed. However, one
specific code definition from the FDA retained our attention and necessitated
more careful handling as we describe in the following.

For all Classes except NFS, the association of the Industry Code, Class, PIC
and Product is precise enough to avoid any misunderstanding by exporters.
It therefore enables an obvious correspondence with the Harmonized System
code at the HS4 level. However, the association of the specific PIC defining
‘Packaged Food’ with NFS Class can be more problematic. Indeed, exporters
tend to use this code for many types of products for which another appropriate
PIC or Class should probably be preferred. Thus, the correspondence with the
HS4 classification requires the use of the Subclass code to differentiate products
comprised under this association of PIC defining ‘Packaged Food’ and NFS
designations.
For ‘Metal’ and ‘Glass’ containers, we make the straightforward assumption
that those products were transformed, and thus fall into HS Chapter 16 for
fish and fishery products, and HS Chapter 20 for fruit and fruit products, and
vegetables and vegetable products. For containers of different materials, we
use two sources of information. First, data from 1998 to 2001 contain product
description data in which the FDA agent could fill in a precise definition of the
product. Second, we extend the research using the name of the manufacturer
from the database in order to gather information from company websites on the
type of products they are exporting. According to this sample of manufacturers,
we were able to make assumptions about the type of transformation associated
with those product codes.
5.3.2 Other data
In addition to the novel dataset on FDA import refusals discussed in the previous
subsection, we use standard data sources for the remaining variables used in
our analysis (Table 5.1). We source trade data from UN-Comtrade, accessed
via the World Bank’s WITS platform. We use US import data for 1998–2008
at the HS4 level, including all exporting countries. In light of the high quality
of US import data, we replace all missing values with zero to indicate that no
trade took place for the given exporter-product-year combination. We only
include trade data for which we have corresponding refusals data, namely HS
chapters 3, 7, 8 and 20. In addition to trade data to control for imports in levels
and first differences, we source per capita GDP data in PPP terms from the
World Development Indicators. Finally, as an additional robustness check (see
below), we estimate the model including US effectively applied import tariffs
as an additional explanatory variable. These data are sourced from UNCTAD’s
TRAINS database via the World Bank’s WITS platform.

Table 5.1 Data used in the analysis of US import refusals
Variable Description Year Source
GDPPCit
Per capita GDP of country i in year t (in PPP
terms)
1998–2008
World Development
Indicators
Importsikt
Imports of product k from country i in year t, in
quantity terms (not value)
1998–2008
UN Comtrade via
WITS
Refusalsikt
Number of import refusals affecting product k
exported from country i in year t
1998–2008 Authors
Tariffikt
Effectively applied US tariff on product k from
country i in year t
1998–2008
UNCTAD TRAINS
via WITS
5.3.3 Preliminary analysis
Before moving to a fully specified econometric model, it is useful to examine
some simple correlations in the data to see whether they support our three
hypotheses, namely the own reputation effect, the sector reputation effect, and
the neighbour reputation effect. As outlined above, we expect to see positive
associations between, on the one hand, the number of refusals for a given country-
product-year combination and, on the other, the number of refusals affecting that
country-product combination in the previous year (own reputation), the number
of refusals affecting related products – those in the same HS2 chapter – from the
same country in the previous year (sector reputation), and the number of refusals
affecting the same product from related countries – the five geographically
closest to the exporter – in the previous year (neighbour reputation). For ease of
presentation we consider data for a single year only, 2008 (the most recent year
in our database), and convert the variables to logarithms to reduce dispersion.
Inallthreecases(Figures5.1–5.3),thedataprovidesupportforourpropositions.
The positive association is strongest, as would be expected, in the case of own
reputation (Figure 5.1). Although the correlations in Figures 5.2 and 5.3, which
capture reputational spillover effects, are weaker, they are nonetheless positive
and 1% statistically significant. In terms of slope coefficients, the stronger
gradient of the line of best fit in Figure 5.3 than in Figure 5.2 provides some
preliminary evidence that neighbour reputation may be quantitatively more
important than sector reputation.
Of course, the graphical analysis we have presented is based on simple
correlations only. It does not take account of intervening influences. To address
this issue more fully, the next subsection develops an econometric model, for
which we report estimation results in the next section.

Figure 5.1 Current versus lagged refusals, 2008
0
1
2
3
4
5
0 1 2 3 4 5
Log(Refusals) (t-1)
Log(Refusals) Fitted values
Figure 5.2 Refusals versus lagged refusals affecting similar products (same HS2
chapter), 2008
0
1
2
3
4
5
0 2 4 6
Log(Refusals HS2) (t-1)
Figure 5.3 Refusals versus lagged refusals affecting the five closest countries,
2008
0
1
2
3
4
5
0 1 2 3 4 5
Log(Refusals Neighbors) (t-1)

5.3.4 Empirical model
As discussed above, we are primarily interested in assessing the impact of
reputation effects in the enforcement of US SPS regulations through import
refusals. Our dependent variable is therefore a count of the number of import
refusals affecting a particular exporter-product-year combination. Because it
takes strictly non-negative integer values, we use the Negative Binomial model
as our workhorse estimator.18
However, in additional results available on request,
we show that the choice of estimator is not critical to our main conclusions.
In particular, our results stand if a Poisson estimator is used instead. However,
we prefer the Negative Binomial model due to obvious over-dispersion in the
dependent variable, that is, its variance (27) is substantially greater than its mean
(0.5).
As independent variables, we include three measures of reputation. The first,
‘own reputation’, is simply the lagged dependent variable, that is, the number of
refusals affecting a given exporter-product combination in the previous year. The
second, ‘sector reputation’, is a lagged count of the number of refusals affecting
products in the same HS 2-digit chapter from a given exporter, but excluding the
number of refusals affecting the product in question. It is therefore a measure
of the extent to which related products are subject to import refusals. The third
variable, ‘neighbour reputation’, is a lagged count of the number of refusals
affecting the same product exported from geographically close countries. We
define ‘closeness’, using geodesic distance as the benchmark, that is, the five
closest countries to the exporter. If reputation effects are present in the data, we
expect all three of these variables to have positive and statistically significant
coefficients.
It is also important to ensure that we control for other possible influences on
the number of import refusals. The level of imports, and the speed of import
growth, are two obvious controls to include. We expect both to be positively
associated with the number of refusals. For a given probability of refusal, a
higher volume of imports leads mechanically to a greater number of refusals.
We expect import growth to be positively associated with refusals because an
import ‘surge’ is likely to lead to a concentration of enforcement resources on
the new exporter, for which there is little history of compliance. In both cases,
we use import volumes (quantities) rather than values, to ensure that we are
capturing pure quantity effects, and not mixing quantity and unit value effects,
which would be the case if we used import values. Moreover, we keep both
variables in levels, rather than taking logarithms, to ensure that observations
with zero trade are retained in the estimation sample.
In addition to import quantities in levels and first differences, we also control
for the exporting country’s per capita GDP. We use this measure as a proxy for
18 Because of numerical problems with Stata’s built-in Negative Binomial estimator, we estimate the
model using the more general GLM framework and iterative re-weighted least squares. See Santos
Silva and Tenreyro (2011) for full details of this issue as it affects the closely related Poisson
estimator.

the exporter’s level of financial and technical capacity, which is an important
determinant of its ability to comply with foreign standards. To take account
of additional country-, product-, and time-specific factors, we also include a
full set of fixed effects in those three dimensions. Product-fixed effects are of
particular importance, because they allow us to control for the inherent riskiness
of particular products, which is likely to lead to a greater rate of inspections and
refusals.
Bringing these points together, and specifying the mean of the count of refusals
conditional on a matrix of explanatory variables , gives an empirical model of
the following form:
(1)
(1) E[Refusalsikt|Xikt] =
b0 + b1Refusalsikt-1 + b2Σj=1; jβk Refusalsijt-1 + b3Σj=1; jβi Refusalsjkt-1 +
KHS2
NNeighbours
Own reputation
}
Sector reputation
}
Neighbour reputation
}b4∆Importsikt-1 + b5Importsikt-1 + b6log(GDPPCit) + fi + fk + ft
where f indicates fixed effects in the exporter (i), product (k), and time (t)
dimensions.
5.4 Estimation results and interpretation
Table 5.2 column 1 presents results for the baseline model, that is, Equation
(1) above. In line with expectations, the lag of import quantity has a positive
and 1% statistically significant coefficient; however, lagged import growth has
an unexpected negative and statistically significant coefficient.19
This finding
suggests that import volumes are obviously an important determinant of the total
number of refusals, but that it is the overall quantity of imports – not recent growth
– that matters most. Indeed, recent growth may be capturing improvements in
country competitiveness, including upgrading of the standards system, which
would tend to make refusals less likely. Finally, per capita GDP has a negative
coefficient, which is in line with expectations, but it is statistically insignificant.
Rather than interpret this result as indicating that the level of development of
national standards infrastructure is not relevant to the determination of import
refusals, we instead conclude that per capita GDP is simply a poor proxy for the
variable we are really interested in measuring. In any case, we return to the issue
of income effects below, when we split the sample into different World Bank
income group categories.
19 In additional results, available on request, we show that our findings are not sensitive to the
specification of the past import growth term. For instance, using growth over the five preceding years
does not make any qualitative difference to our results.

Table 5.2 Regression results
(1) (2) (3) (4)
Baseline Tariffs UMI+LMI+LI LMI+LI
Refusalsikt-1
Own reputation
}
0.062*** 0.047*** 0.053*** 0.057***
(0.000) (0.000) (0.000) (0.000)
Σ Refusalsijt-1j=1; jβk
KHS2
Sector reputation
}
0.004*** 0.003*** 0.006*** 0.005***
(0.000) (0.000) (0.000) (0.002)
Σ Refusalsjkt-1j=1; jβi
5
5 Neighbour reputation
}
0.011*** 0.007*** 0.013*** 0.013***
(0.000) (0.000) (0.000) (0.000)
DImportsikt-1
–0.000** –0.000 –0.000*** –0.000
(0.044) (0.622) (0.000) (0.100)
Importsikt-1
0.000*** 0.000*** 0.000*** 0.000*
(0.000) (0.002) (0.000) (0.073)
log(GDPPCit
) –0.062 0.273 –0.264 –0.455
(0.903) (0.598) (0.684) (0.560)
log(1 + tariffikt
) 3.417***
(0.001)
Constant –44.411*** –37.045*** –40.947*** 4.980
(0.000) (0.000) (0.000) (0.479)
R2 0.027 0.030 0.036 0.004
Observations 68684 15240 50952 35112
Fixed effects Exporter Exporter Exporter Exporter
HS 4 Product HS 4 Product HS 4 Product HS 4 Product
Year Year Year Year
Interestingly, even after controlling for lagged import volume, the coefficient on
the lagged number of refusals is positive and 1% statistically significant. The
refusals data are clearly quite persistent through time, which is consistent with
the existence of an own reputation effect: a history of compliance is associated
with fewer current refusals, but a history of non-compliance is associated with a
greater number of current refusals. This result lines up well with the findings of
Jouanjean (2011), and confirms the intuition of Baylis et al (2010). It is more in
line with expectations than the counterintuitive finding of Alberini et al (2005)
that seafood HACCP inspections are not based on compliance history at the
plant level.
The model in Table 5.2 column 1 also suggests that sector reputation has a
positive and statistically significant impact on the current number of refusals.

Although the coefficient is considerably smaller in magnitude than for the
own reputation variable, it is nonetheless 1% statistically significant. In
addition to a history of compliance in relation to a given product, therefore, a
history of compliance in related products is also important in determining the
current number of import refusals. It is important to stress that the count of
refusals affecting related products does not include those directly affecting a
given exporter–HS 4 digit product combination, so the effect identified by this
coefficient is quite independent of the own reputation effect.
The remaining variable in Table 5.2 column 1 is the neighbour reputation
effect, based on a ‘neighborhood’ defined as a country’s five geographically
closest neighbours. Again, the coefficient on this variable is positive and 1%
statistically significant, which is in line with expectations. In terms of magnitude,
the neighbour reputation coefficient lies between the own reputation and sector
reputation coefficients: its effect is nearly three times as strong as the sector
reputation effect, but it is still only one-sixth as strong as the own reputation
effect.
Bringing these results together, we find strong support for all three of our core
hypotheses: own reputation, sector reputation and neighbour reputation all have
a significant impact on the current number of import refusals. In the remaining
columns of Table 5.2, we present some simple robustness checks. First (column
2) we include tariffs as an additional explanatory variable. We then (columns
3–4) estimate the model separately for subsamples of countries identified by the
World Bank’s country income classification. In additional results, available on
request, we also estimate the model using alternative estimators (Poisson and,
with a dummy dependent variable for at least one refusal, Logit), as well as
separately for each World Bank geographical region.
Column 2 of Table 5.2 presents results including data on the US effectively
applied tariff rate – that is, including preferences – to assess the degree to which
SPS enforcement and protectionist measures are correlated, and to address a
potential source of omitted variables bias.20
Results are very close to the baseline
model:allthreereputationvariablesretainpositiveand1%statisticallysignificant
coefficients, although their magnitudes fall slightly in each case. Interestingly,
the tariff variable has a positive coefficient, which is suggestive of stronger SPS
enforcement in sectors that are relatively protected, perhaps due to the influence
of industry lobbies, as suggested by Baylis et al (2010). More broadly, this
result could be interpreted as indicating a degree of complementarity between
tariffs and non-tariff measures, in this case import refusals. Clearly, more work
is required before strong conclusions can be drawn, but the data at this stage are
highly suggestive of such a link.
From a policy point of view, it is important to know whether the types of
reputation effects highlighted in column 1 of Table 5.2 are common to a wide
range of countries, and in particular low- and middle-income countries. Our prior
20 The sample size in column 2 is greatly reduced due to the lack of availability of tariff data in
UNCTAD’s TRAINS database.

is that these effects should be stronger than average for low- and middle-income
countries, because compliance with standards represents more of a burden – in
terms of technical capacity and relative cost – than in high-income countries.
With this outlook in mind, Table 5.2 columns 3–4 present estimation results
for the baseline model using restricted country samples, focusing on different
per capita income groups, as defined by the World Bank. Column 3 excludes
all high-income countries (OECD and non-OECD) from the estimation sample.
Contrary to expectations, there is no general evidence of stronger reputation
effects: coefficient estimates are close to the baseline, although the magnitude
of the sector and neighbour reputation effects are somewhat stronger. Similar
findings follow from column 4, where we exclude from the sample upper
middle-income countries in addition to high-income countries: the coefficients
on sector and neighbour reputation are somewhat stronger than under the
baseline. Bringing these results together, we conclude that there is no evidence
that the own reputation effect is more important for low- and lower-middle-
income countries. On the other hand, the sector and neighbour reputation effects
do appear to be somewhat more important for those countries.21
5.5 Conclusion, policy implications and avenues for further research
This paper has produced some of the first direct evidence that reputation effects
matter in the enforcement of US SPS measures through the import refusals
system. Specifically, countries with a history of compliance tend to experience
fewer refusals, even after controlling for other factors. In addition, countries
with a history of compliance in related products also tend to experience fewer
current refusals, as do countries whose neighbours have an established history
of compliance. We interpret these last two effects as evidence of reputational
spillovers in the enforcement of SPS rules.
From a development point of view, we find some evidence that sector and
neighbour reputational effects are more important for lower-income countries.
This result sits well with findings in the broader literature on standards and trade,
which suggest that it is primarily poorer countries that encounter negative trade
effects from foreign product standards (eg Disdier et al, 2008).
Although more research is clearly needed in a number of areas – more on
this below – some important policy implications would seem to follow from
our findings. First, exporters of agricultural products seeking to break into the
US market need to focus on building SPS capacity so as to become reliable
sources. It is not sufficient to export a mix of compliant and non-compliant
goods: reputation matters, and the presence of the latter will make it harder to
get the former into the market as well. Consistency and reliability of production
are therefore key issues in the development of SPS capacity in poor agricultural
exporters.
21 We also estimated a model for low-income countries only, but the estimates failed to converge.

Second, our results strongly suggest that a comprehensive approach to SPS
compliance is likely to be more effective than a piecemeal one. Although it
might seem sensible to concentrate limited SPS capacity-building resources on
a small number of products that are individually important, such an approach
neglects the importance of the sectoral spillover effects evident in our data.
Building capacity across the sector as a whole can have important benefits for
individual products.
Similarly, the likelihood that regional reputation matters for SPS enforcement
also has important policy implications. Regional approaches to the development
of standards systems are becoming more common for many reasons, such as the
ability for small, poor countries to pool technical and financial resources (Maur
and Shepherd, 2011). Our findings suggest an additional reason for encouraging
regional standards cooperation: geographical spillovers mean that compliance
by a country’s neighbours can help it achieve more effective market access.
Currently, there is only a very small literature examining SPS measures at the
level of enforcement mechanisms, such as alerts or import refusals. Further work
in this area has the potential to bring significant insights into the workings of
product standards more generally, and in particular their effects on developing
country exporters. Baylis et al (2010) make a first attempt to assess the trade
impacts of import refusals. Extending their work to take account of the types
of reputation spillover effects we have identified here could be a fruitful avenue
for future research. Our own work highlights the need to treat import refusals as
endogenous in gravity model settings, which is an important dimension in which
the robustness of previous assessments needs to be established. Similarly, Baylis
et al (2009) provide some initial evidence suggesting that political economy
forces may be relevant in determining the application of SPS measures. Since
almost nothing is known about the political economy determinants of product
standards (cf Kono, 2006), this too would be an interesting research question to
pursue using data similar to those we have used here.
References
Alberini, A, Lichtenberg, E, Mancini, D, and Galinato, G (2005), ‘Was it Something
I Ate? Implementation of the FDA Seafood HACCP Program’, Working Paper
No 05-03, Department of Agricultural and Resource Economics, University of
Maryland.
Allshouse, J, Buzby, J C, Harvey, D and Zorn, D (2003), ‘International Trade and
Seafood Safety’, chapter 7 in J C Buzby (ed), International Trade and Food Safety:
Economic Theory and Case Studies,Agricultural Economic Report No (AER828),
USDA Economic Research Service.
Baylis, K, Martens, A, and Nogueira, L (2009), ‘What Drives Import Refusals?’,
American Journal of Agricultural Economics, 91(5): pp. 1477-83.
Baylis, K, Nogueira, L and Pace, K (2010), ‘Food Import Refusals: Evidence from
the European Union’, American Journal of Agricultural Journal of Economics,
advance access published 15 December 2010.

Buzby, J C, Unnevehr, L J and Roberts, D (2008), ‘Food Safety and Imports: An
Analysis of FDA-Related Import Refusal Reports’, Information Bulletin, 39,
USDA ERS.
Buzby, J C and Roberts, D (2010), ‘Food Trade and Food Safety Violations: What
Can We Lean from Import Refusal Data?’ American Journal of Agricultural
Journal of Economics, advance access published 17 December 2010.
Cadot, O, Jaud, M and Suwa Eisenmann,A (2010), ‘Do Food Scares Explain Supplier
Concentration? An Analysis of EU Agri-Food Imports’, CEPR Discussion Paper
7431.
on Agricultural Trade: Evidence from the SPS and TBT Agreements’, American
Journal of Agricultural Economics, 90(2), pp 336–50.
Essaji, A (2008), ‘Technical Regulations and Specialization in International Trade’,
Journal of International Economics, 76(2), pp 166–76.
Grossman, G and Helpman, E (1984), ‘Protection for Sale’, American Economic
Review, 84(4), pp 833–50.
Jouanjean, M-A (2011), ‘Standards, Reputation and Trade: Evidence from US
Horticultural Imports Refusals’, Discussion Paper Series DP 281/2011, LICOS.
Karov, V, Roberts, D and Grant, J H (2009), ‘A Preliminary Empirical Assessment of
the Effect of Phytosanitary Regulations on US Fresh Fruit and Vegetable Imports’,
Selected paper prepared for the Agricultural and Applied Economics Association’s
2009 AAEA ACCI Joint Annual Meeting, Milwaukee, 26–28 July.
Kono, D Y (2006), ‘Optimal Obfuscation: Democracy and Trade Policy
Transparency’, American Political Science Review, 100(3), pp 369–84.
Maertens, M and Swinnen, J F M (2009), ‘Trade, Standards, and Poverty: Evidence
from Senegal’, World Development, 37(1), pp 161–78.
Maur, J-C and Shepherd, B (2011), ‘Product Standards’, in J-P Chauffour and J-C
Maur (eds), Preferential TradeAgreement Policies for Development:A Handbook,
Washington DC, World Bank.
Santos Silva, J and Tenreyro, S (2011), ‘Poisson: Some Convergence Issues’, Stata
Journal, 11(2), pp 215–25.
USDA Economic Research Service (ERS) (2011), http://www.ers.usda.gov/Data/
FruitVegPhyto/data.htm.
Venezian, D (2008), ‘Import Process, Prior Notice and Import Alerts’, Presentation
at the Food and Agriculture Border Gateway Summit, January 16, 2008.

Appendix
Table 5A.1 Industry Class Codes for fruit and fruit products
Industry Industry type Class Class name
20
Berries
A Simple
B Dried or paste
C Jams, jellies, preserves, marmalades, butters and candied
D Juices or concentrates
F Purees
E Toppings and syrups
Citrus
G Simple
H Dried or paste
J Jams, jellies, preserves, marmalades, butters and candied
K Juices or concentrates
M Purees
L Toppings and syrups
Core fruit
N Simple
Q Dried or paste
R Jams, jellies, preserves, marmalades, butters and candied
S Juices or concentrates
U Purees
T Toppings and syrups
21
Pit fruit
G Simple
H Dried or paste
I Purees
Subtropical
and tropical
fruit
S Simple
T Dried or paste
U Jams, jellies, preserves, marmalades, butters and candied
X Subtropical/tropical fruit pulp
R Purees
W Toppings and syrups
V Juice, milk, creme, drink or nectar, sub/tropical fruit
Mixed
A Simple
B Dried or paste
F Purees

22
Vine fruit
A Simple
B Dried or paste
F Purees
Other fruit
G Simple
H Dried or paste
I Purees
Table 5A.2 Industry Class Codes for vegetables and vegetable products
24
Bean, pea, corn
A Simple
B Dried or paste
C with sauce
D Juice or drink
Fruit used as vegetable
F Simple
G Breaded
H Dried or paste
J Juice or drink
K with sauce
Leaf stem vegetable
T Simple
W Juice or drink
U Dried or paste
V with sauce
Sprouts from seeds peas or beans E Simple
25
Fungi products
S N.E.C
R Broken or kibbled
Q Pieces and stems, sliced
P Whole (button)
Mixed vegetables
E Simple
H with sauce
F Dried or paste
G Juice or drink
Root tuber vegetable
J Simple
L Dried or paste
N with sauce
K Breaded
M Juice or drink

6
Sanitary Risk and Concentration in EU
Food Imports1
Mélise Jaud, Olivier Cadot and Akiko Suwa-Eisenmann
6.1 Introduction
Afteraseriesofhighlypublicisedfoodscares(bovinespongiformencephalopathy
in 1992, dioxin-contaminated chickens in 1999 or bans on genetically modified
organisms), public-health concerns have started to loom large in the European
Union and resulted in an overhaul of EU food and feed legislation and enhanced
oversight. Tightened public mandatory standards are complemented by private
standards that are also de facto mandatory such as Eurepgap. In addition, public
voluntary standards such as ISO norms also apply to agri-food products. In this
study, we consider public mandatory standards on food safety.
Standards of course apply equally to all producers. Yet, they entail several
costs of compliance, not only a fixed investment cost at the outset, but also
recurring costs associated with traceability requirements (‘from farm to fork’),
certification or pre-shipment inspection. These costs could price out exporters in
poor countries and, hence, impact their access to EU markets (Garcia-Martinez
et al, 2006; Maskus et al, 2005).
Sanitary standards might all the more act as potential trade barriers for poor
countries, for two other reasons. First, the reduction in traditional trade barriers
such as tariffs and the rising number of preferential schemes result in most of
EU agro-food imports entering duty free. In contrast, the impact of remaining
barriers such as standards appears all the more significant. Second, EU imports
of farm products (which is the world’s largest market for developing countries)
has recently evolved from traditional tropical goods such as coffee, cocoa and
spices to fresh produce such as fruits, vegetables or fish. These products are
potentially more concerned by sanitary risk.
This chapter examines the link between the sanitary risk of food products
and recent changes in the EU food import pattern. We build a product-specific
measure of sanitary risk using a database that reports all food alerts at the EU
border. As will be seen, the incidence of food alerts is fairly heterogeneous
1 This is a shorter and non-technical version of Jaud et al (2010). This project was part of Jaud’s Ph.D.
dissertation at the Paris School of Economics.

across exporting countries. This implies that a raw count of alerts by product
cannot measure correctly its sanitary risk. For instance, a product imported
overwhelmingly from a country with weak quality standards would appear as
risky even though other exporters might have managed to make the product safe.
In addition, the incidence of alerts is likely to be correlated with the frequency
of controls, which might be non-random. Inspections may reflect a particular
exporter’s past performance or hidden protectionism. Hence, we rely on a two-
step procedure. In the first step, we wipe out the various factors that might affect
the incidence of food alerts independently of the characteristics of the product
and we estimate a product-specific Sanitary Risk Index. Then, in a second step,
we relate this Sanitary Risk Index to the pattern of EU food imports.
The evolution of EU foodstuff import pattern results from the interplay between
diversification and concentration forces. First, an exogenous taste for variety, or
a desire to limit monopoly positions leads to a larger number of suppliers. Over
time, the decrease in trade costs, such as transport costs and tariffs, also implies
a diversification of import sources, as more countries are able to export to the
EU. Moreover, the overall productivity growth that goes along with development
triggers entry into the EU market of poor countries which used to lie below the
required threshold of productivity (in a setting where suppliers are characterised
by heterogeneous productivity level).
However, in the case of goods with particular safety concern the importer
needs to source from reliable producers. Hence, there is a trade-off between
concentrating on top quality suppliers and keeping several suppliers in order
to test them. As new exporting countries get on the EU’s list of suppliers, they
need to build a reputation of quality. Typically, these countries will enter into the
EU markets on a small scale and increase the size of their shipments as they are
gradually revealed as reliable producers, while incumbent exporters are never
completely immune from food safety crisis. The tradeoff between these two
diverging forces evolve over time, hence the evolution of the EU food product
import pattern is open to empirical scrutiny.
We find evidence that product sanitary risk has indeed affected the European
Union’s pattern of imports. Overall, there is a slight diversification of import
sources over time. This overall trend hides diversification at the extensive margin
and concentration at the intensive margin, the more so for risky products. On
the one hand, EU agri-food imports are being increasingly sourced from a few
countries. On the other hand, additional new countries enter the EU market, on a
smaller scale than before, especially if their exports entail some risk. As a result,
EU foodstuff imports gradually evolve towards a two-tier distribution with a
small number of dominant suppliers and a growing fringe of marginal ones.
Our work relates to two distinct lines of research. First, this work contributes
to the literature on the potential impact of Sanitary and Phytosanitary (SPS)
standards as barriers to trade. EU standards in particular are often more specific

Sanitary Risk and Concentration in EU Food Imports 119
and restrictive than multilateral Codex ones2
and have been suspected of going
beyond the requirements dictated by mere sanitary concerns. An important
set of works has focused on the quantitative assessment of the trade effects of
standards. Most of them have included standards (as dummy variables signalling
their existence in a given tariff line) as explanatory variables in gravity equations.
Otsuki et al (2001) found that a 10% tighter aflatoxin standard in the EU would
be associated with an 11% reduction in edible groundnut imports. Moenius
(2006) extended the analysis to all agricultural products. His results provide
a much more balanced picture: while some country-specific standards tend to
reduce trade, others promote it, depending on the complexity of the good and
on the type of the standard itself (country specific versus harmonised standard).
Harmonisation of standards across countries may increase trade in some cases,
despite the loss in the variety of exchanged goods and regional approaches to
standards, such as the EU, can have significantly different effects on insider and
outsider countries. Overall the protective effects of standards increased during
1980 and 1995. Crucially, the impact of importing country standards varies with
the level of income of the exporting country, with negative effects being more
likely vis-à-vis developing country exporters (Disdier et al, 2008). Moreover
EU imports are more negatively affected by SPS measures than imports of other
OECD countries, suggesting a higher stringency of the EU market. Another set
of works, largely based on detailed case studies, has provided further nuance
and highlights the fact that gravity model findings tend to represent short-run,
average, or aggregate effects. In particular, short-run studies miss the catalysing
role that product standards may play in the medium run by promoting quality
upgrading and technical progress in developing countries (Diaz Rios and Jaffee,
2008; Henson and Jaffee, 2004, 2008, Jaffee 2003; World Bank, 2005).
Second, this work relates to the literature on trade patterns along the
development process. Imbs and Wacziarg (2003) have shown that the degree of
diversification of production and employment follow an inverse U shape along
development. Klinger and Lerderman (2004), Cadot et al (2011), and Dutt et
al, (2008), document a similar hump shaped pattern for exported products. As
to geographical diversification, micro studies suggest a role for proximity and
experience gained in the destination markets. A country starts exporting to one
market and gradually expands to neighbouring countries (Evenett and Venables,
2002). This stylised fact finds further support in Eaton et al’s (2007) detailed
study of Colombian exports. Recently, Amurgo-Pacheco and Pierola (2008)
explored both types of diversification, by products and partner countries. This
rapidly growing line of research also relates to the numerous works that build
on Melitz (2003) and stress the heterogeneity of firms and their differing ability
to export.
We differentiate ourselves from the existing literature in three important ways.
First, we adopt an ex-post approach and consider effective product sanitary risk
2 Codex Alimentarius Commission is a body created by the Food and Agriculture Organisation of the
United Nations (FAO) and the World Health Organisation (WHO).

based on real food alerts at the EU borders. Our ex-post approach complements
previously cited studies that are based on the existence of a standard on a
given product: the latter, by contrast, could be considered as de jure measures
of sanitary risk. Second, contrary to previous papers, we take the view point
of the importer and not of the exporter. Finally, we shift focus from gravity
modelling to an analysis of the geographical concentration of EU foodstuff
imports, using the Theil index. The separability property of the Theil index
enables us to decompose the overall geographical concentration into variations
at the extensive and intensive margins.
The rest of the chapter is structured as follows. The next section describes the
data, in particular the EU Food Alerts database. Section 6.3 details our empirical
strategy. Section 6.4 reports the main empirical results and Section 6.5 concludes.
6.2 Data
6.2.1 The Food Alerts database
We use EUROSTAT’s Rapid Alert System for Food and Feed (RASFF). The
RASFF is a system of notification and information exchange on emergency
sanitary measures taken at the border by EU Member States. It does not include
information on intra-EU trade. The system in place since 1979 was incorporated
in the EU food regulation in 2002 after major food scares. It involves the Member
States, the European Commission and the European Food Safety Authority
(created in 2001).
The database records all notifications (19,000) between 2001 and 2008 with
the identity of the importing EU Member State, exporting country, product,
hazard, type of notification, and type of measure. The database contains complete
information regarding products in verbal form and products are not coded into
the CN system. We painstakingly coded all products into CN8 categories over
the period 2001–2005 (8,895 observations). We consider imports from EU-12
Member States so as to be consistent with the trade data over a longer period.3
Thus, we created an entirely new database of food alerts.
It is important to make two remarks at the outset. First, products enter
the database only if they are exported to the EU and subject to a sanitary or
phytosanitary standard. Moreover, it is unlikely that inspections are purely
random. They could reflect the concern of inspection services with respect to
certain firms exporting a given commodity from a particular country. In the
following econometric exercise, we will try to control for these selection biases.
The firms’ targeting will be smoothed out anyway, as notifications are recorded
at the country level. Hence, despite these caveats, we believe that the RASFF
3 France, Belgium-Luxembourg, the Netherlands, Germany, Italy, Ireland, United Kingdom, Denmark,
Greece, Portugal and Spain. We use this restrictive definition for consistency of time series, as the
source of the trade data, EUROSTAT, does not provide data on Member States before their accession.

database provides reliable information on the effective degree of food imports
safety.
Notifications can be of two types: ‘information’ or ‘alerts’. In the case
of information, the hazard is deemed limited; the importing Member State
imposes a measure (e.g. destruction of the shipment) and informs the rest of the
Community of the problem, but other members do not follow suit. In the case of
notification, the hazard is deemed sufficiently serious to warrant simultaneous
action at the Community level.
Panel (a) of Figure 6.1 shows the evolution over time of the number of
notifications and its decomposition between information and alerts. The total
number of notifications shows a sharply rising trend, although somewhat
decelerating after 2003. Information notifications outnumber alerts by a ratio of
more than six to one on average over the period.
Panel (b) highlights the substantial heterogeneity among notifying EU States
in the frequency of notifications taking the average over the period 2001-2005.
Germany (23% of observations) Italy (21%) and Spain (18%) are the top notifying
countries, while Ireland accounts for only 0.39% of notifications. This may
reflect the differing agri-food imports structure of EU Member States. It could
also highlight the fact that inspections are not operated uniformly across the EU.
Thus, in the following econometric analysis, notifications are aggregated across
all EU Member States and all years, in order to smooth out these variations.
Regarding exporters’ region of origin, Asian countries are the most affected
by notifications, followed by European countries and they are steadily on the
rise after the turn of the century. Other regions are less affected and the number
of notifications seems even to slightly decrease after 2003 for Latin American
countries or stabilise after 2004 for African countries (panel c).
There is also considerable heterogeneity among products (panel e), with fruits
and nuts, fishery products, coffee and spices ranking highest. In terms of hazards,
the main cause of notifications for agricultural products is contamination by
mycotoxins (mainly aflatoxin), which alone accounts for almost 40% of
the notifications (panel f). The second cause is contamination by residues of
veterinary medicinal products (chloramphenicol, nitrofuran and tetracycline)
which together account for 14% of the notifications. Then comes the presence
of pathogenic micro-organisms (12%) and contamination by pesticide residues
(3%).
Notifications, whether alerts or information, induce very strong actions. Over
85% of notified shipments were stopped at the border in 2005. In 26% of the
cases, shipments were destroyed. The rest (45%) were re-dispatched to other
destinations; they did not penetrate the EU market but would nevertheless end
up on somebody’s plate. An additional 14% of imports were banned (panel d).

Figure 6.1 The food alerts database, 2001–2005
2000
1500
1000
500
0
2001 2002 2003 2004 2005
Notifications
Informations
Alerts
(a) Total notifications, information and alerts,
2001-2005
Germany
Italy
Spain
UK
Netherlands
France
Greece
Bel/Lux.
Portugal
Denmark
Ireland
0 5 10 15 20 25
Percent
23.28
21.61
18.09
10.82
9.99
4.68
4.54
3.97
1.74
0.96
0.39
050010001500
2001 2002 2003 2004 2005
year
Africa Asia
Europe North America
South America Oceania
(b) Notification by Member State, average
2001-2005
(c) Notifications by exporting regions,
average 2001-2005
32.56
27.42
8.51
7.46
5.18
3.42
3.35
2.98
2.86
2.70
1.59
0.95
0.82
0.68
0.20
0.17
0.09
0 5 10 15 20 25 30 35
Percent
Fruits and nuts (hs08)
Fish and fishery products (hs03)
Coffee, Tea (hs09)
Oil seeds (hs12)
Meat (hs02)
Vegetables (hs07)
Dairy products (hs04)
Animal or vegetable fats (hs15)
reparations of hs07,08 (hs20)
Miscellaneous (hs21)
Preparations of hs01,02,03 (hs16)
Preparation of hs10 (hs19)
Sugars (hs17)
Animal products (hs05)
Product milling industry (hs11)
Cocoa (hs18)
Cereals (hs10)
(d) Notifications by CN2 sectors,
average 2001-2005
39.05
14.97
12.49
6.53
5.82
4.55
4.02
3.21
1.19
0.86
0.73
0.37
0 5 10 15 20 25 30 35 40
Percent
Mycotoxins
Residues of veterinary products
Micro-organisms
Food additives
Unauthorised components
Heavy metals
Microbiological contamination
Pesticide residues
Others Contamination
Parasites
Others
Labelling Packaging
(e) Notifications by hazard category,
average 2001-2005
44.32
26.12
14.61
12.94
1.44
1.15
0 5 10 15 20 25 30 35 40
Percent
Product re-dispatched
Product seized destroyed
Ban
Others
Physical treatment
Reinforced checking
(f) Actions taken post notifications,
average 2001-2005

Figure 6.2 shows a scatter plot of exporter total number of notifications
between 2001 and 2005 against their level of development in 2001, both taken in
logs. The size of the circles is proportional to the country’s share in EU imports.
China, Turkey and Brazil are most affected by SPS notifications, but they are
also among EU’s largest suppliers. Countries like Canada, New Zealand or
the USA are large exporters, but subject to relatively few notifications; at the
other end of the spectrum, Iran and Vietnam suffer a disproportionate number of
notifications given their relatively lower import shares. This dispersion suggests
that important country-specific characteristics, on top of sales volumes, affect
the probability of getting a notification. These characteristics include obviously
the product composition of exports, as well as unobserved characteristics of
national production systems which must be taken into account when assessing
the level of product-specific sanitary risk.
Figure 6.2 EU food alerts notifications by the exporting country’s level of
development
USA
NOR
CAN
SGP
KOR
AUS
JPN
NZL
ISR
ARG
IRN
TUR
BRA
CHN
THA
TZA
MWI
ZWE
BOL
IND
VNM
IDN MAR
EGY
ECU PHL CHL
TUN
CRI
MUS
MEX
COL
GHA
NGA PAK
BGD
SEN
CIV
MDG
KEN
UGA
0
2
4
6
8
log(no.ofnotifications,2001-2005)
6 7 8 9 10 11
Log(GDPpc) 2001
Developing countries Developed countries
Scale: 0.1% EU Agri_food Import 2005
6.2.2 European agri-food imports
We use EUROSTAT agri-food import data for EU-12 at the CN8 level (the highest
level of disaggregation available, as EUROSTAT does not make 10-digit data
available to researchers). Agri-food products, excluding beverages and animal
feed, are in Chapters 1 to 21 of the CN system, which represent 3,073 potential
export lines. With 147 partner countries (exporters) including 122 developing

countries, we have a three-dimensional panel where the unit of observation is a
product imported by the EU from an extra-EU partner in a given year.4
Table 6.1 shows the evolution of the structure of EU agri-food imports between
1988 and 2005. The share of developing countries, already dominant at 66% in
1988 increased further to 75% in 2005. Interestingly, this rise is not attributable
to traditional tropical products (such as coffee, cocoa or sugar) which, as a share
of imports from developing countries, shrank from 23.9% to 15.4% over the
sample period. Rather, it is due to horticulture and fisheries products, whose share
rose from 21.7% to 29.6% of EU agri-food imports from developing countries.
Hence, the EU has been importing more fresh horticulture and fisheries products
from developing countries over time, a noticeable evolution as fisheries and
horticulture products are sensitive to sanitary and phytosanitary issues, and the
ability to meet stringent SPS standards is correlated with exporter income.
Figure 6.3 shows the distribution of EU imports by number of source countries.
The distribution is strikingly skewed, with single supplier country (there may
be several supplying firms per exporting country) accounting for over 12%
of the total number of transactions. This highlights the high degree of market
concentration among agri-food products, with a peak at one supplier country,
and an average number of 12.
The number of sourcing countries varies across sectors, however. The dairy
products sector is extremely concentrated with almost 40% of transactions
involving a unique partner. On the contrary, the vegetable plaiting material
or coffee and tea sectors exhibit no single-partner transactions. This apparent
heterogeneity across sectors requires controlling for product characteristics in
the subsequent analysis of concentration over time. In 2001–2005, the number
of potential suppliers ranged from 10 and 147 (the total number of EU partner
countries), depending on the product (defined at the CN8 level), with a mean
of 78 potential suppliers. During the same period, the average number of active
suppliers was only 12, and ranged from 1 to 96 depending on the product (Table
A1 in the Appendix).
4 At the CN8 level, reclassifications are frequent. Five types of reclassification can be distinguished: (i)
creation of a new code corresponding to a new product; (ii) creation of several new codes by splitting
a former one; (iii) creation of a new code by merging several former ones; (iv) creation of new codes
resulting from a change in the coding system (CN harmonisations in 1988, 1996 and 2002); and
finally (v) termination of old codes. Of the 3,073 CN8 codes available in our dataset, only 37.7%
are unaffected by reclassification between 1988 and 2005. Of the remainder (62.3%), 1.6% are new
products (type i), and 0.7% are terminated codes (type v). This leaves 60% of reclassifications of
‘continued’ products. Among those, half (30%) are type-ii reclassifications (splittings) and half are
type-iv (system changes). In order to reduce the inconsistencies introduced by type-iii and type-iv
reclassifications to a minimum, we used EUROSTAT’s documentation to re-reclassify new codes
into initial ones; in case (iii), where a new code was made out of several old ones, we used the first
parent’s code in the CN order. This gives us a consistent database using the initial nomenclature
throughout the sample period.

Table 6.1 Structure of EU agri-food import (in percentage)
Developing countries
(extra EU)
Developed countries
(extra EU)
Total Imports
Description 1988 2005 1988 2005 1988 2005
Traditional tropical
products
23.9 15.4 1.0 1.5 24.9 16.9
Coffee, tea, mate and
spices
15.7 7.4 0.1 0.3 15.8 7.6
Lac, resins, gums 0.5 0.5 0.2 0.2 0.7 0.7
Sugars and sugar
confectionery
2.8 2.4 0.4 0.5 3.3 2.8
Cocoa and cocoa
preparations
4.8 5.1 0.3 0.6 5.1 5.7
Tempered zone products 4.2 6.7 8.4 5.3 12.6 11.9
Live animals 0.1 0.0 0.5 0.0 0.6 0.0
Meat and edible meat offal 2.1 3.2 2.6 2.4 4.7 5.6
Dairy produce 0.2 0.3 1.9 1.0 2.2 1.4
Products of animal origin 0.9 1.1 0.5 0.2 1.4 1.4
Cereals 0.9 1.9 2.8 1.6 3.7 3.5
Fish and horticulture 21.7 29.6 12.3 12.3 33.9 41.9
Fish, crustaceans and
molluscs
5.6 11.1 6.9 7.0 12.4 18.1
Live trees and other plants 0.7 1.9 0.6 0.5 1.2 2.4
Edible vegetables 5.0 3.4 1.2 1.3 6.1 4.7
Edible fruit and nuts 10.2 13.1 3.6 3.4 13.8 16.5
Vegetable plaiting materials 0.3 0.2 0.1 0.0 0.4 0.2
Others 16.4 23.0 12.2 6.3 28.6 29.3
Products milling industry 0.1 0.1 0.1 0.0 0.1 0.1
Oil seeds and oleaginous
fruits
5.1 6.2 7.2 2.5 12.3 8.7
Animal and vegetable fats 3.7 6.7 1.1 0.3 4.8 7.0
Preparations of meat and
fish
2.9 5.3 1.7 1.0 4.6 6.3
Preparations of cereals 0.1 0.4 0.3 0.6 0.4 1.0
Preparations of vegetables,
fruit or nuts
4.1 3.6 1.3 0.3 5.5 3.9
Miscellaneous preparations 0.4 0.7 0.5 1.5 0.9 2.2
Total 66 75 34 25 100 100
Note: In 2005, traditional tropical products represented 16.9% of EU total agri-food imports, out of which
1.5% came from developed countries and 15.4% from developing countries.

Figure 6.3 Frequency of number of active suppliers (% in total number of
transactions)
0
5
10
15Percent
0 20 40 60 80 100
No. of active exporters, varying with importer product and year
6.3 Empirical strategy
6.3.1 The Sanitary Risk Index
The level of sanitary risk associated with imported agri-food products can result
from (i) intrinsic product characteristics, as some products are more vulnerable
than others to contamination; (ii) supplier characteristics, as some producers are
more able than others to apply necessary controls; or (iii) a combination of both.
Hence, a raw count of alerts by product cannot give a correct measure of product
sanitary risk. We rely thus on a two-step procedure, where in the first step, we
estimate a measure of sanitary risk at the product level based on the count of
food alerts at EU borders and controlling for the influence of confounding
factors such as the characteristics of the exporting country or some protectionist
motive of the importer country. In a second step, we relate the geographical
concentration indices to this estimated measure of risk. The sample period is
restricted to 2001–2005 where both trade and notification data are available.
This section details both steps of the procedure.
The Sanitary Risk Index (SRI) is computed using the EU food alert database,
described in Section 2. For a product k and an exporter c, the dependent variable
is the combined count of notifications from all twelve EU Member States
between 2001 and 2005.5
The unit of observation is an exporter*product pair
and the regression is cross-sectional. Our variable of interest is the coefficient on
the product dummy, dk
,in the following regression:
5 As there are consistent differences in the number of notifications among notifying EU states and
across years, all observations are pooled across all importing (notifying) countries and all years.

Alertck
= f (aImpshare
EU
ck
+ bcontrolsk
+ dk
+dk
+ eck
)
where eck
is an error term.
Because the number of notifications is a count (with over-dispersion and a large
proportion of zeros), estimation is made with a negative binomial. We include
exporter c share in EU imports of product k Impshare
EU
ck in 2000 (one year before
the sample start). Indeed, a product imported overwhelmingly from a country
with weak quality standards would appear as risky even though other exporters
might have managed to make the product safe. In addition, the incidence of alerts
is likely to be correlated with the frequency of inspections. The latter may not
be purely random: they may reflect a particular exporter’s past performance or
hidden protectionism. Hence, we include trade barriers, such as the ad-valorem
equivalent of the EU’s most favoured nation (MFN) tariff on product k in 2005,
dummy variables indicating whether product k is affected by a quota during
the sample period, or has been the object of a dispute at the WTO involving
the EU and any other country, and whether exporter c has been affected by a
ban on product k during the sample period. Indeed, a ban could result in fewer
notifications not because of a lower risk but because of its mere impact on trade
flows. We also control for the initial value of EU imports of product k in 2000,
as products imported in large volumes are likely to be inspected – and therefore,
fail inspections – more often than others. We also add an exporter fixed effect
that controls for all supplier’s characteristics that may affect the quality of the
product, such as the overall level of economic development. As a consequence,
the coefficient of the product dummy captures the part of food alerts that is due
to product characteristics, net of exporters’ characteristics and of other factors
that may affect the probability of rejection and is what we label the estimated
product’s Sanitary Risk Index.6
6.3.2 The impact of sanitary risk on EU import pattern
Theil index
In order to investigate how rising sanitary risk of agri-food products affect
the geographical pattern of EU imports, we need a summary index of the
concentration of import sources. We use the Theil entropy index (Theil, 1972),
which compares the actual distribution of market shares among suppliers to what
would be an equal distribution. The index varies from zero (perfect equality)
to one (perfect inequality). The Theil index has the property of putting a larger
weight on both extremes of the distribution. Hence, it is adequate in capturing
the action in the first tier of the distribution, where suppliers will be shown
entering the EU market small scale.
6 When estimated coefficients were not significant at the 10% level, they were set equal to zero.

Omitting the time indices, for good k, the Theil index is given by:
Tk=
1
nk
xk
c
µk
nk
c=1 ln
xk
c
µk
( )∑
where
µk=
1
nk
xk
cnk
c=1
∑
is the average value of import for good k from origin country c and nk
is the
total number of countries exporting good k. More precisely, nk
is the number of
‘potential exporters’ as not all countries can produce a specific agricultural crop
because of required type of soil and climate. Here, a potential exporter is defined
as a country that has exported good k to some destination in the world (not
necessarily an EU country) at least two years in a row over the sample period.
We impose the requirement of two consecutive years of exports in order to ensure
that the exporter is a successful one. Indeed, Besedes and Prusa (2006a, 2006b)
show that many countries export only one year and fail, while two years is the
median duration of export spells. This definition of potential exporters has the
advantage of being time invariant and allows comparing Theil indices over time.
Intensive versus extensive margins
The Theil index can be decomposed as the sum of what happens at the intensive
margin (among incumbent exporters) and what happens at the extensive margin
(allowing entry and exit of exporters). Thus we define nEU
kt as the number of
active suppliers of good k in year t, that is the countries that are effectively selling
to any EU country that good k in year t. We have seen earlier that the number
of active suppliers, which varies every year, is different from the number of
potential suppliers that is time-invariant.
A simple computation, presented in the Appendix, shows that the Theil index
at the extensive margin is a function of the inverse of the number of active
suppliers. The intuition is simple: when the number of active suppliers increases,
there is less inequality and hence, more diversification of sourcing countries:
Tkt
B
= ln
nk
nkt
EU( )
Thus, the changes in the between-groups component of the Theil index trace
exactly the percentage changes in the extensive margin defined as the number of
active suppliers (as the set of potential exporters, nk
is time invariant):
DTB
kt = TB
kt - TB
kt-1 = -DlnnEU
kt

On the other hand, the Theil index at the intensive margin is simply the index
computed over active suppliers nEU
kt :
Tkt
W
=
1
nkt
EU
xk
c
µkt
EU
nkt
EU
c=1
ln
xk
c
µkt
EU∑ ( )
Explaining concentration patterns with the Sanitary Risk Index
The product’s Sanitary Risk Index estimated previously is then thrown as an
explanatory variable of the concentration indices, in a panel regression, where
the unit of observation is a product*year pair:
Ckt = α+βControl ck+ γδk+µ. δk .δt+δt+δs+ kt
^ ^
where Ckt
is a measure of concentration (resp. Theil index or its within or between
component, or the number of active suppliers) in year t of good k belonging in
sector s. Control variables are the same as before. Sector fixed effects (at CN-4
level) control time-invariant characteristics at the sector level that may also
affect the concentration. As the product dummies used in this second step are
generated variables, standard errors are computed with a bootstrap.
6.4 Empirical results
6.4.1 The Sanitary Risk Index
The Sanitary Risk Index, estimated in the procedure presented in the previous
section ranges between 0 and 12.2, with a mean at 0.76 (Table A1). It is a rather
conservative measure of risk, because non-significant estimates are assumed
to have a zero risk. The distribution of significant point estimates is shown in
Figure 6.4. No product chapter stands out as having particularly high risk levels,
except for fisheries products, with mussels as an unsurprising outlier (sector 3).
Remarkably, traditional tropical products such as coffee (sector 9) and cocoa
(sector 18), whose share in EU foodstuff imports is as noted before, declining,
are among the safest products.

Figure 6.4 Distribution of point estimates of Sanitary Risk Index
Mussels
Shrimps Prawns
Surimi Pistachio
Swine
Ground Nuts
Nuts
Palm Oil
Frozen
Mackerel
Frozen
Trout
Chili
Dried
veg
Cashew Nuts
Dried Figs
Coffee
Anise
Soya Beans
Karite Nuts
Bamboos
Mussels
Snails
Preserved
meat
Olives
JamMeat
0
5
10
15SRI
Anim
al/vegetable
fats
C
ocoa

prepC
ofee,Tea
Fish
Fruits

nuts
M
eat
M
iscellaneousO
ilseedsPO
AO
,nes
Planting
m
aterials
Prep
of01,02,03
Prep
of07,08Prep
of10Sugars

coVegetables
Sector
10% significance 5% significance
1% significance
Table 6.2 lists the agri-food sectors ranked by the Sanitary Risk Index, with the
number of ‘risky’ products – defined as products with a positive Sanitary Risk
Index – per sector, as well as the number of alerts and the most frequent reason
for rejection. Sectors estimated with the highest sanitary risk are not necessarily
those with the highest observed number of alerts. The case of edible fruit and
nuts is a striking example. Most of the 3,210 alerts in this sector are attributable
to one single supplier, Iran’s aflatoxin-contaminated pistachios.After controlling
for that, the edible fruit and nuts sector ranks only sixth in terms of sanitary risk
while fishery products and spices emerge as the riskiest.

Table 6.2 Sanitary Risk Index, number of alerts and cause of rejection by sectors
Description
# ‘risky’
products
Sanitary Risk
Index
# Alerts
2001–5
Most frequent cause for
rejection
Coffee, tea, mate,
spices
38 2.07 934 Composition, mycotoxins
Preparations of meat
fish
32 1.29 309 Residues drugs
Oil seeds and
oleaginous fruits
25 1.04 1491 Mycotoxins
Fish, crustaceans
molluscs
Miscellaneous edible
preps
7 0.85 185 Food additives
Edible fruit and nuts 53 0.71 3210 Mycotoxins
Edible vegetables 27 0.65 441 Pesticide residues
Cocoa and cocoa
prep.
4 0.57 20 Allergens
Prepared vegetable,
fruit,nuts
44 0.54 677 Mycotoxins
Sugar 5 0.49 221
Food additives
mycotoxins
Products of animal
origin
Meat and edible
meat offal
17 0.24 498 Pathogens
Animal or vegetable
fats
7 0.18 247 Composition
Preparations of
cereals
2 0.16 167 Radiation
Dairy produce 0 0.03 367 Residues drugs
Live animals 0 0 1 Heavy metals
Live trees and other
plants
0 0 0
Cereals 0 0 158 GMO/mycotoxins
Milling industry
products
0 0 36 Food additives
Resins, gums 0 0 1 Food additives
Vegetable plaiting
materials
1 0 1 Labeling incorrect
Note: Risky products are products with a positive Sanitary Risk Index. Out of a total of 2,146 CN8
products, 373 are risky products. In column (2) we compute the Sanitary Risk Index at the CN2 level,
taking the average overall CN8 product in each CN2 sector. Column (3) reports the total number of alerts
per CN2 sector, over the period 2001–2005. The last column details the most frequent cause for an alert.

Table 6.3 goes in further detail and lists the ten most and least risky products.
It turns out that not all fish products are risky, depending on the species and the
preparation (fresh versus frozen).
Table 6.3 The ten most and least risky products
Product Risk Index
Most risky
Mussels 12.208
Shrimps 11.314
Fresh tuna fillets 10.859
Ground-nuts-shell 8.259
Seeds 8.153
Pistachios 8.063
Fish fillets 7.973
Nuts – seeds 7.951
Seaweeds – algae 7.914
Dried cuttle-fish 7.866
Least risky
Soya 0.82
Cocoa 0.876
Oil – palm 1.217
Salmon 1.612
Food preparations 1.725
Frozen cod-fillets 1.781
Nuts-fruits 2.086
Coffee 2.089
Prawns-shrimps 2.179
Hake-fillets 2.25
Figure 6.5 plots the number of notifications and the Sanitary Risk Index at the
product level. The correlation between the raw number and the estimated risk is
quite low. Hence, controlling for exporting country characteristics or bilateral
trade policies seem indeed to matter. Figure 6.5 also distinguishes between food
types, such as fresh, frozen or prepared food. Understandably, fresh produce
are in general more risky, but prepared foods such as prepared fish, chutney or
mustard also appear in the top quartile.
To our knowledge, this product-specific measure of sanitary risk is new in
the literature. It is an ex-post measure of the level of effective risk based on
real food alerts at the EU border. This index shows how food safety regulations
translate into inspections and rejections of non-compliant shipments. As such,
it complements other de jure measures that rely on the existence of public legal
standards. Disdier et al (2008) give such a de jure measure, which is the number
of standards for a given (HS6) product, notified by any country to the WTO.
The coefficient of correlation between this measure and the Sanitary Risk Index
is only -0.05. This low correlation might come from different reasons: first, it
could be the sign of a discrepancy between world practice and that of the EU.

Moreover, this correlation compares the number of WTO notifications at the HS6
level and a more disaggregated risk index which is estimated at the HS8 level. In
addition, a lot of products have at least one notification from any country in the
world, while few products, once controlled for exporting country characteristics
and the endogeneity of the inspections, are estimated as having a positive SRI.
Figure 6.5 Sanitary Risk Index and number of notifications in the RASFF
database
horses-Meat
rabbits-meat
frogs'-legs
game-meat
fish-ornamental
salmon
tunastunas
fish
hake monkfish
swordfish
fish-n.e.s.gilt-head
mackerel
fish-fillets
fish-fillets
Surimi-
meal-Fish
halibut
lobsters-Live
Mussels
Squid
Snails
Illex
Guts-stomachs
mushrooms
capsicum-fruits
vegetables-n.e.s.
mushrooms
Dried-tomatoes
vegetables-mixtures
chickpeas
beans-shelled
Sweet-potatoes
Roots-tubers
coconut
cashew
almonds
hazelnuts
walnuts
walnuts
pistachios
dates
figs
figs
pineapples
oranges
grapes grapes
'papayas'
kiwi
fruit-n.e.s.
Strawberries-uncooked
apricots
prunes
fruit-n.e.s.
Coffee Green-tea
Pepper
Pepper
peppers-Dried
capsicum-fruits
Cinnamon
Nutmeg
Coriander-
Cumin-seeds
Fennel-seeds
Ginger
Turmeric
leaves-Bay
Ground-nuts-shell
ground-nuts-Shelled
Sunflower-seeds
Sesamum
Poppy-seeds
Seeds-fruit
Liquorice-roots
Ginseng-roots
Pyrethrum
Seaweeds-algae
Apricot
Palm-oil
Sesame-oil
turkey-Preparations
poultry-preparations
Shrimps-prepared
Mussels
Gum-jelly
Cocoa
cocoa-powder
Chocolates
Sweet
Cucumbers
Fruit-capsicum
other-fruit
fruit-nuts
TomatoesVegetables-mixtures
Fruit-capsicum
Vegetables-mixtures
Jams-jellies
Nuts-seeds
Nuts-seeds
Orange-juice
Apple-juice
fruit-Juicecoffee-extracts
Tomato-ketchup
Sauces
Soups
salmon
sharks
hake
fish
swordfish
fish-livers
pacific-filletsfish-fillets
cod-fillets
sharks-fillets
saltwater-fillets
rock-lobster
lobsters
prawns-shrimps
crabs
crayfish
cuttle-fish
cuttle-fish
squid
squid
squid
octopus
molluscs
Meat-salted
Fish-salted
Prepared-meat
Prepared-salmon
Herrings-prepared
Prepared-sardines
Prepared-tuna
Prepared-mackerel
anchovies-Prepared
Prepared-fish
prepared-Crab
Crustaceans-prepared
Mango-chutney
Mushrooms-prepared
Asparagus-preserved
Mustard
0
2
4
6
8
No.notificationsinlog
-0.5 0.5 1.5 2.5
Risk index in log
45°line Fresh Frozen Prepared
Figure 6.6 Evolution of EU import concentration
3.55
3.6
3.65
3.7
Theiloverallindex
1985 1990 1995 2000 2005
year
Theil overall index
.8
.85
.9
.95
1
Theilwithinindex
1985 1990 1995 2000 2005
year
Theil within index
2.6
2.65
2.7
2.75
2.8
2.85
Theilbetweenindex
1985 1990 1995 2000 2005
year
Theil between index
9
10
11
12
13
Nberofactivesuppliers
1985 1990 1995 2000 2005
year
No. of active suppliers

6.4.2 Geographical pattern of EU agro-food imports
The average number of potential suppliers is high, at around 78 suppliers, in
contrast with the average number of actual suppliers which is just around 12
(Table A1). Concentration indices are very high, with a Theil index at 3.30 and
a Gini index at 0.74. This is consistent with our earlier observation that the
distribution of the number of active suppliers is highly skewed. It could also be
partly due to the very detailed level of disaggregation (CN-8 digit).
Figure 6.6 re-places the 2001–2005 data in a broader perspective and shows the
evolution of simple averages, over all products, of the geographical concentration
indices between 1988 and 2005. The downward trend of the Theil overall index
indicates that geographical diversification has occurred over time. It is paralleled
by a similar downward trend of the between-component of the Theil index, along
with a rising number of active suppliers; both signs suggest a diversification
at the extensive margin. However, the Theil within-group component shows an
upward trend and points to a rising concentration among active suppliers. These
trends are accentuated after 2000.
Thus, we get the following picture: on the one hand, supplier concentration is
rising at the intensive margin, meaning that the top suppliers get larger relative
to the average. On the other hand, concentration is decreasing at the extensive
margin, as more suppliers are added to the EU’s portfolio of suppliers. These
apparently diverging observations can be reconciled as follows. Suppose first a
group of three incumbent suppliers with unequal export levels to the EU, say, $4,
$3 and $2 respectively, with two more potential suppliers waiting in the wings
with zero exports to the EU. The Theil index within active suppliers is 0.04.
Suppose now that a new supplier enters the EU, with exports of only $1. As
there is one additional supplier, there is diversification at the extensive margin.
However, the group of active suppliers is now more unequal, as the largest
exporter is 4/2.5 = 1.6 times the group average. Hence, the within-Theil index
rises to 0.2. Thus, the pattern that is observed in the data is consistent with an
additional entry of small-scale suppliers into the EU market that results in both
diversification at the extensive margin and concentration at the intensive margin.
6.4.3 Product risk and geographical concentration
We now focus the analysis on the relationship between public-health concerns
and the concentration of EU suppliers. Product riskiness has a significant and
negative impact on geographical concentration of EU agro-food imports. The
coefficient of sanitary risk is indeed negatively correlated with the overall Theil
index (Table 6.4, column 1), controlling for EU protectionist motives, product
and time fixed effects. This overall trend hides two diverging evolutions. At the
intensive margin, concentration is higher for riskier products (column 2); on the
otherhand,attheextensivemargin,concentrationislowerforthoseriskyproducts
(columns 3) and the number of suppliers is higher (column 4). Coefficients in
column (2) and (4) imply that a 1% increase in the level of sanitary risk results

in a 5.8% (0.0735*0.76/0.96) increase in the level of concentration among active
suppliers and a 12.7% (2.0193*0.76/12) increase in the number of suppliers to
the EU.
Table 6.4 EU import pattern and Sanitary Risk Index
(1) (2) (3) (4)
Dependent
variable
Theil overall Theil within Theil between
No. of active
suppliers
sanitary_risk_k –0.0675*** 0.0735*** –0.1422*** 2.0193***
(–0.009) –0.008 (–0.011) (–0.209)
Observations 8,602 8,608 8,608 8,608
R-squared 0.336 0.39 0.467 0.428
Note: All regressions are estimated using OLS and control for the existence of a tariff, a quota, a ban, or a
dispute on the product considered over the period, sub-sector (CN4) and time fixed effects. Bootstrapped
standard errors are in parentheses. *, **, and *** denote statistical significance at the 10%, 5%, and 1%
levels, respectively.
Figure 6.7 illustrates further the polarisation between the top and smallest
exporters. Panels (a) and (b) show the first and last quartile import share
distribution in 2005 for risky versus safe products. Risky products are, as before,
defined as products with a positive risk coefficient. For the smallest partners
(last quartile of the import share distribution), the curve for risky products shifts
to the left compared to the curve for safe products. However, the two curves for
the top exporters (first quartile of the import share distribution) are very similar
and overlap. Thus, there is a rising polarisation between the bottom and the top
of the distribution, especially for risky products. A possible explanation for the
fact that entrants begin small scale in the presence of risk, is that they are being
tested. Alternatively, exporters would want to ‘try the market’ small scale before
taking big risks (Rauch and Watson, 2003).
Figure 6.7 Import share distribution, for risky and safe products, 2005
0
.1
.2
.3
Kernel_density
-15 -10 -5 0
ln(import share 4th quartile)
Safe Products Risky Products
0
2
4
6
8
Kernel_density
-1.5 -1 -.5 0
ln(Import Share 1st Quartile)
Safe Products Risky Products
(a) Smallest exporters, last quartile (b) Biggest exporters, first quartile

6.5 Concluding remarks
This chapter establishes a stylised fact on EU import concentration in agri-food
products and makes a link between this import pattern and rising public-health
concerns in EU food policy.
While previous empirical work have focused on a de jure definition of standards
or their ex-ante impact on trade flows, this chapter is, to our knowledge, the first
to assess their ex-post and effective impact. Using for the first time a dataset on
food alerts that provide information on the implementation of SPS standards by
EU importing countries, it contributes to the empirical debate on the evolution of
geographical concentration of EU agri-food imports across time.
The results are clear. EU foodstuff imports over the period 1988–2005
have gradually evolved towards a two-tier distribution with a small number
of increasingly dominant suppliers and a growing fringe of marginal ones.
European importers tend to buy their agri-food products from an increasingly
large portfolio of suppliers, but large orders are concentrated on a small number
of reliable suppliers. This evolution is more marked for products that entail a
sanitary risk according to the EU regulation. The explanation put forward for
such a pattern is a search for food safety from the buyer with sellers building
up reputations. However, the paper does not explicitly test for this mechanism.
Other mechanisms may also contribute to this pattern: first, our results may be
capturing the increasing vertical integration of global food supply chains, along
with the rise in foreign investment both in food processing and retailing sectors
and the dominance of large multinational food companies; second, concentration
at the intensive margin might create larger rents that in turn attract smaller
exporters. This explanation, however, would imply that there is no barrier to
entry in the agro-food sector.
The policy implications of these results are of significant interest. As EU
foodstuff distributors show growing concerns for food safety, developing
countries’ access to EU markets is increasingly constrained by stringent sanitary
requirements that are costly to implement.
References
Amurgo-Pacheco, A and Pierola, M D (2008), ‘Patterns of Export Diversification
in Developing Countries: Intensive and Extensive Margins’, Policy Research
Working Paper No. 4473, The World Bank.
Besedes, T and Prusa, T (2006a), ‘Ins, Outs and the Duration of Trade’, Canadian
Journal of Economics 104, pp 635–54.
Besedes, T and Prusa, T (2006b), ‘Production Differentiation and Duration of U.S.
Import Trade’, Journal of International Economics, 70(2), pp 339–58.
Cadot, O, Carrère, C and Strauss-Kahn, V (2011), ‘Export Diversification: What’s
behind the Hump?’ Review of Economics and Statistics 93(2), pp 590–605.

Diaz Rios, L and Jaffee, S (2008), ‘Barrier, Catalyst, or Distraction: Standards,
Competitiveness, and Africa’s Groundnut Exports to Europe’, Agriculture and
Rural Development Discussion Paper 39, The World Bank.
on Agricultural Trade : Evidence from the SPS and TBT Agreements’, American
Journal of Agricultural Economics 90(2), pp 336–50.
Dutt, P, Mihov, I and van Zandt, T (2008), ‘Trade Diversification and Economic
Development’, mimeo, INSEAD.
Eaton, J, Eslava, M, Kugler, M and Tybout, J (2008), ‘The Margins of Entry into
Export Markets: Evidence from Colombia’, in E. Helpman, D. Marin and T.
Verdier (eds), The Organization of Firms in a Global Economy, Cambridge, MA,
Harvard University Press.
Evenett, S and Venables,A (2002), ‘Export Growth in Developing Countries: Market
Entry and Bilateral Trade Flows’, mimeo.
Garcia Martínez, M, Poole, N D and Skinner, C (2006), ‘Food Safety Performance
in European Union Accession Countries: Benchmarking the Fresh Produce Import
Sector in Hungary’, Agribusiness: An International Journal 22(1), pp 69–89.
Henson, S and Jaffee, S (2004), ‘Standards and Agro-Food Exports from Developing
Countries: Rebalancing the Debate’, World Bank Policy Research Paper No. 3348.
Henson, S and Jaffee, S (2008), ‘Understanding Developing Country Strategic
Responses to the Enhancement of Food Safety Standards’, The World Economy
31(1), pp 1–15.
Imbs, J andWacziarg, R (2003), ‘The Stages of Diversification’, AmericanEconomic
Review 93(1), pp 63–86.
Vegetable Trade in the Context of Emerging Food Safety and Other Standards in
Europe’, World Bank Agriculture and Rural Development Discussion Paper.
Jaud, M, O Cadot andA Suwa-Eisenmann (2010), ‘Do Food Scares Explain Supplier
Concentration? An Analysis of EU Agri-food Imports,’ CEPREMAP Working
Papers 1015.
Klinger, B and Lederman, D (2004), ‘Discovery and Development: An Empirical
Exploration of “New” Products’, mimeo.
Maskus, K E, Otsuki, T and Wilson, J S (2005), ‘The Cost of Compliance with
World Bank Policy Research Working Paper No. 3590, Washington DC, The
World Bank.
Melitz, M J (2003), ‘The Impact of Trade on Intra-Industry Reallocations and
Aggregate Industry Productivity’, Econometrica 71(6), pp 1695–725.
Moenius, J (2006), ‘The Good, the Bad and the Ambiguous: Standards and Trade in
Agricultural Products’, mimeo.
Otsuki, T, Wilson, J S and Sewadeh, M (2001), ‘Saving Two in a Billion: Quantifying
the Trade Effect of European Food Safety Standards on African Exports’, Food
Policy 26, pp 495–514.
Rauch, J E and Watson, J (2003), ‘Starting Small in an Unfamiliar Environment’,
International Journal of Industrial Organization 21(7), pp 1021–42.

Theil, H (1972), Statistical decomposition analysis, Amsterdam,North Holland
publishing company.
World Bank (2005), ‘Food Safety andAgricultural Health Standards: Challenges and
Opportunities for Developing Country Exports’, World Bank Poverty Reduction
and Economic Management Trade Unit and Agriculture and Rural Development
Department Report No. 31207.

Appendix: Decomposition of the Theil index
Each year, the sample of EU suppliers can be divided into two groups. Group
1 (G1
) is composed of active suppliers (those actually exporting that year),
numbering n1
kt = nEU
kt . Potential exporters who do not actually send any shipment
to the EU that year are in group 0 (G0
), numbering n0
k = nk – nEU
kt . We can
decompose the Theil index into these two groups (respectively T 0
kt, and T 1
kt) and
compute the Theil index between these groups and within them (respectively T B
kt
and TW
kt).
The between-groups component of the Theil index is given by:
T
n
n
lnkt
B
j
kt
j
k
kt
j
kt
kt
j
kt
=





=
∑ ,0 1
µ
µ
µ
µ
It is a weighted average of terms involving only group means (relative to the
population mean).
By L’Hôpital’s rule: li µ
µ
µ
µ
µkt
kt
kt
kt
kt
ln0
0
0 0
0→











 = , so that:
T
n
n
lnkt
B kt
k
kt
kt
kt
kt
=






1 1 1
µ
µ
µ
µ
The mean exports of active suppliers is µkt
kt
c G kt
c
n
x
1
1
1
1
= ∈∑ ,
while the mean exports on all (potential) exporters is µkt
k kt
c
n x
1
= ∑ kn
c=1
and by
construction,
kt
c
x∑ kn
c=1c G kt
c
x1
∑ =∈
, it follows that:
T ln
n
n
kt
B k
kt
EU
=






The changes in the between-groups component of the Theil index trace exactly
the percentage changes in the extensive margin defined as the number of active
suppliers, as the set of potential exporters, nk is time invariant:
T Tkt
B
kt
B
kt
B
kt
EU
ln= − =−−1∆ ∆T n
The within-group Theil index is defined as:
T
n
n
Tkt
W
j
kt
j
k
kt
j
kt
kt
j
= =∑ ,0 1
µ
µ
The Theil index is zero when all individuals have equal shares, hence T 1
kt = 1.
For active suppliers:
T
n
x
ln
x
kt
kt c
n
k
c
kt
k
c
kt
k
1
1
1
1
=






=
∑µ µ1 1
It reduces to:
T
n
n
Tkt
W kt
EU
k
kt
EU
kt
kt=
µ
µ
1

Table A6.1 Descriptive statistics of main variables, 2001–2005
Variable Obs Mean Std. Dev. Min Max
Theil overall index 8598 3.3 0.71 0.66 4.99
Theil within (intensive margin) 8608 0.96 0.67 0 3.27
Theil between (extensive margin) 8608 2.34 1.06 -1.39 4.99
No. of active suppliers 8608 12.01 12.52 1 96
No. of potential suppliers 8608 78.56 30.09 10 147
Sanitary Risk Index 8608 0.76 1.99 0 12.21

7
Do Better Standards Facilitate Exports?
Evidence from China1
Axel Mangelsdorf, Alberto Portugal-Perez and John Wilson
7.1 Introduction
Food safety has been a major concern in China. In 2004, at least 13 infants
were reported to die from consumption of fake milk powder, and in 2008, six
children were reported to die after consuming milk adulterated with melamine,
anindustrial chemical (Gale and Buzby, 2009). Unsafe food is a serious threat
for food sales to foreign markets. Producers risk losing substantial revenue and
reputational damage when unsafe food is refused to enter foreign markets and
food products are destroyed. Problems in meeting food safety standards give rise
to health-related concerns and a loss in foreign consumers’ confidence that can
act to lower imports from China over time. According to the Asian Development
Bank (2007), a weak regulatory infrastructure and low voluntary and mandatory
standards are responsible for recurring incidences unsafe food production and the
following import refusal. Therefore, complying with best practices, international
food standards can be a powerful tool in addressing food safety problems.
Existing research on Chinese standards in manufacturing sectors has shown that
introducing new standards that are harmonized with international norms can
be trade promoting (Mangelsdorf 2011). Applying international standards is in
particular important for firms from developing countries with a reputation for
low quality or unsafe products. Hudson and Jones (2003) argue that consumers
in Western countries of associate product safety with per capita income of the
exporting country. They argue that developing countries should apply more
stringent standards to signal consumers that they produce high quality and
safe agricultural and food products. In addition, standards application increase
transparency of the product offered and reduces information asymmetries,
especially in international transactions (Blind and Jungmittag, 2005). In this
chapter, we show evidence on the impact of more stringent harmonized standards
on Chinese exports food and agricultural products following the application of
more stringent harmonized standards..
1 This chapter draws on Mangelsdorf et al (2012).

A number of government agencies are involved in regulating food safety in
China: the Ministry of Agriculture, the Administration of Quality Supervision,
Inspection and Quarantine (AQSIQ), the Food and Drug Administration,
the Ministry of Health, and the Administration for Industry and Commerce,
and the Standardization Administration of the People’s Republic of China
(SAC). SAC is the agency responsible for approving and publishing domestic
standards and adopting international standards including standards for food
and agricultural products. Standardisation in SAC takes place in 450 national
technical committees in which stakeholders from government agencies, industry
and research organisations participate. SAC also represents Chinese interests in
international organisations. China is actively involved in the Codex Alimentarius
Commission (CAC), the organisation responsible for setting international
standards for food products. For instance, SAC hosts the CAC sector committees
for pesticide residues and food additives. Setting food safety standards is a
general priority of the Chinese government. In the Eleventh Five-year Plan on
the Development of Standardization, agriculture and food safety are named
key areas for standardisation activities and the Chinese standardisation strategy
encourages the adoption of international standards (Ping et al, 2010). Despite
the participation in international standards setting bodies and the aim to adopt
international standards, the number of Chinese standards based on international
measures remains low compared to other countries. From the total number
of standards recorded in our database, only 14% are based on international
measures.
Does the introduction of mandatory and voluntary standards serves as a
quality signalling instrument for Chinese exports? Are standards for Chinese
agricultural products effective at reducing information asymmetries? The aim of
this chapter is to attempt to answer these questions by assessing econometrically
the impact of Chinese mandatory and voluntary standards on its exports of
agricultural products. We draw on recent research by Mangelsdorf et al (2012)
using a new database of Chinese food standards for seven groups of agricultural
products: meat, fish, vegetables, cereals, milk, tea and sugar.
Our focus is on standards implemented by the exporting country for its produced
goods. The new database allows differentiating between mandatory and
voluntary standards. In addition, each type of standard can be identified as being
harmonised to international norms or being a ‘purely’ domestic standard.
A collapsed gravity model was used to examine the impact of food standards in
China on Chinese agricultural exports. Estimates confirm the export-promoting
and quality signaling effect of standards. This is particularly true in regard to
mandatory standards. In particular, the analysis confirms the positive impact of
standards harmonisation. The impact of an additional standard that is based on
international measures, we find, has a larger impact than the marginal impact of
purely Chinese domestic standards.
The results discussed in this chpater provide new evidence to support the
conclusion that the application of standards in general and international standards
in particular leads to a positive trade effect of the exporting country.

Do Better Standards Facilitate Exports? Evidence from China 143
The remainder of this chapter is structured as follows. In the next section,
we screen the literature on the trade effects of standards in general and for
China in particular. We present our empirical strategy, our data and descriptive
statistics in Section 7.3. We present the empirical analysis by Mangelsdorf et
al (2012) and present the effects of standards and regulations in Section 7.4.
Finally, we summarise the results and conclude with some policy implications
and recommendations in the last section.
7.2 Literature review
7.2.1 Standards and trade
Standards can be differentiated by the freedom of choice regarding compliance
(Henson, 2004). Mandatory standards -- also called technical regulations in the
Agreement on Technical Barriers to Trade of the World Trade Organization --
are set by public regulators and compliance is obligatory. Voluntary standards
are set by national standards development organisations such as Standardization
Administration of the People’s Republic of China (SAC) or by international
organizations such as the International Organization for Standardization (ISO).
Standardization is a formal process that involves multiple stakeholders such as
industry and trade association or consumer organisations.
The literature on standards and trade differentiates the impact of standards
set in the importing countries from standards set in the exporting country. The
former argue that standards in the importing country can act as a barrier to
trade. Developing countries are concerned that domestic standards in importing
countries – especially in developed countries – increase the cost of compliance
and restrict or even prevent market access (Henson and Jaffe, 2008). Empirical
studies show that the costs of compliance can be significant. Otsuki et al (2001),
for instance, use a gravity model to show that stringent standards for maximum
allowable contamination in fruit and nuts imposed in the European Union led
to significant export losses for African exporters. At the firm level, Maskus
et al (2004) show that producers from developing countries face substantial
investment costs in order to adapt their production processes in compliance
with standards in export markets. International trade agreements encourage
harmonisation of standards across countries. Evidence on the trade increasing
effect of international harmonisation is provided, for instance, by Wilson et al
(2003) for beef products. The authors find empirical evidence that if standards
were harmonised according to Codex Alimentarius standards, international
trade value would rise by more than 50% compared to the non-harmonised
scenario and developing countries would benefit from harmonisation of
standards. Portugal-Perez et al (2010) provide similar evidence for electronic
products. Regarding the literature on the impact of standards in the exporting
country, most studies indicate a positive trade effect for the exporting country.
By providing information on safety levels and signalling quality, standards can

help to overcome incomplete or asymmetric information between producers and
consumers (Leland, 1979).
7.2.2 Standards and trade in China
The literature on trade effects of standards in China largely focuses on standards
in countries importing Chinese products. For instance, Chen et al (2008)
examines in a gravity model the impact of maximum residual standards imposed
by importing countries on Chinese exports for vegetables and aquatic products.
The author finds significant negative effects for safety standards. Yue et al
(2010) analyse the impact of maximum pesticide residue standards for tea
imposed by the European Union. The authors find that the introduction of the
stringent EU standard significantly decreases exports from developing countries
including China. They also argue that the WTO should push the EU to harmonise
their standards for tea with international CodexAlimentarius standards. Song and
Chen (2010) differentiate the impact of food regulations in short- and long-term
effects. In the short term, regulations in foreign countries, proxied as WTO SPS
notifications, have a negative effect on Chinese agriculture exports but the long-
run effect is positive. Chinese exporters can comply with foreign regulations
in the long run because they significantly invest in new testing equipment,
training and use voluntary standards to meet certification requirements in order
to stay competitive in the world market. Bai et al (2007) as well as Jin et al
(2008) confirm that Chinese food producers increasingly make use of voluntary
standards. Based on surveys of Chinese firms, the authors find that the incentives
to use voluntary standards and apply for certifications are to improve product
quality and safety and access to the foreign market. In the next section, we
briefly introduce the institutional background on standardisation in China.
7.3 Empirical model and results
7.3.1 Empirical strategy
In this section, we present the empirical model used to estimate the impact
of voluntary and mandatory standards on China’s exports. The data allows
differentiating between standards that are based on international norms and
purely Chinese domestic standards. Our data covers standards for seven product
groups: meat (HS-02), fish (HS-03), vegetables (HS-07), cereals (HS-10)), milk
(HS-0401 and HS 0402), tea (HS-0902), and sugar (HS-1701 and HS 1702).
Chinese exports in these product groups amount to about 40% of China’s total
exports in agricultural goods. These products are a frequent target of strict food
standards and heavy regulation (see, for instance, Chen and Findlay, 2008).
We base our empirical model in the micro-founded gravity model widely used
in the trade literature and adapt it to Chinese standardisation in food products
categories (Anderson and van Wincoop, 2003).
More specifically, we estimate the equation:

iktikktitikt
iktktktktktikt
Cons
TariffStVIStVDStMIStMDX
εδδδlnβ
ββββββln
6
543210
+++++
+++++=
where Xikt
is the value of Chinese exports of product k to country i in year t in
dollars; StMDkt
and StMIkt
are the stocks of mandatory purely domestic (non-
harmonised) and international harmonised standards for product k that are active
in year t; StVDkt
and StVIkt
are the stocks of voluntary domestic (non-harmonised)
and international harmonised standards for product k in year t; Tariffikt
is the tariff
imposed by country i on Chinese exports of product k in year t; lnConsikt
is the
domestic consumption of product k in country i in year t, measured in dollars;2
δit
, δkt
and δik
are interaction terms of importer-year dummies, product-year
dummies and importer-product dummies; and eikt
is an error term. Three dummy
interactions terms are incorporated in the baseline model. The interaction of
importer and product dummies (δik
) control for invariable characteristics specific
to the food sector in a given importer. Importer–year fixed effects (δit
) control
for shocks experienced by an importer a given year, such as political shocks.
Product–year fixed effects (δkt
) control for shocks specific to world markets of
a product in a given year.3
All variables are included in logarithmic form except
the standards variables. Therefore, the coefficient of the standard variables can
be interpreted as the impact on exports growth of an additional standard.
As noted before, previous empirical evidence shows that a reduction in the
number of standards in an exporting country has a positive effect on exports, as
information and signalling benefits are bigger than compliance costs. Indeed,
standards in the exporting country reduce information asymmetries regarding
quality and safety levels of food products between producers in China and
consumers in importing countries. The existence of exporter standards increases
exports compared to a situation where no standards exist (Moenius, 2006).
Therefore, we expect positive signs in the coefficients of Chinese standards
variables, with a priori, a larger impact of mandatory standards compared to
voluntary standards as compliance with the latter is optional. International
standards applied by Chinese producers are expected to have a positive effect for
Chinese exports for two reasons. First, compliance with international standards
can be a strong signal that Chinese producers meet safety and quality criteria
that are widely recognised. Second, when exporting countries use international
standards they share the same standards with the regulatory requirements in
other countries and there are no additional compliance costs.
Compared to national standards, the informational and signalling benefits of
complying with international standards are expected to be larger than complying
2 We computed the consumption of product k in country i and year t as Cikt = (Sikt + Mikt – Xikt)Pikt, where
Sikt is the domestic supply of product k in country i in year t in tonnes, Mikt (Xikt) stands for imports
(exports) of product k in country i in year t in tonnes, and Pikt is the import price of product k in
country i in year t in dollars.
3 A similar strategy of incorporating dummy interaction terms to control for unobserved effects is
adopted by Harding and Javorcik (2011) and Ferro et al (2011).

with national standards. The coefficient of tariffs is expected to be negative. As a
proxy of demand for each product group in each importing country, we include
the importing countries’ total domestic consumption and its sign is expected
to be positive as higher domestic demand is expected to increase the demand
for imports from China. To control for unobserved characteristics specific to
an importer, such as distance to China or historical linkages, and specific to a
product, we include fixed effects for importers and products.
7.3.2 Data sources
Export volumes are compiled from the COMTRADE database at the
Harmonized System (HS 1992), whereas tariffs are compiled from the TRAINS
database. Data to construct consumption variables are drawn from the Food and
Agriculture Organization of the United Nations’ statistical database FAOSTAT.4
Data on mandatory and voluntary standards are compiled from the Standards
Administration of the Peoples Republic of China (SAC), the Chinese official
national standards body.5
Chinese standards developed in SAC are accessible
through an online database, the SAC National Standards Query. To ensure
reliability and completeness, the standards have been cross checked with the
‘Chinese Bulletin of Standards’ published by AQSIQ and the German-Chinese
Standards Portal. The Chinese National Standards Query provides information
on the date the standard came into effect, and the date the standard has been
withdrawn or replaced by a newer version. Each standard is classified according
to the International Classification of Standards (ICS) nomenclature, which
allows matching standards to trade data. The database allows differentiation
between Chinese standards harmonised to international norms such as Codex
Alimentarius and standards that are unique to China.6 A concordance table
between the seven-digit ICS classes and the trade data in HS codes can be found
in Table 6A.1 in the Appendix. The database allows us to construct four different
standards variables according to two criteria: (i) mandatory versus voluntary
standards, and (ii) domestic versus international harmonised standards. Table
7.1 provides two examples to illustrate the different types of standards. The first
example, the standard GB 14939-1994 ‘Hygienic standard for canned fish’ is a
mandatory performance standard. It is equivalent with the international Codex
Standard 70. The second example, GB/T 22388-2008 is a voluntary testing
standard to determine the content of melamine which has no international
equivalent.
4 http://faostat.fao.org/
5 SAC obtained legal recognition by the Chinese government through the Standardization Law of the
People’s Republic of China from 1988 (source: ISO).
6 SAC develops mandatory and voluntary standards. The prefix ‘GB’indicates mandatory standardsand
voluntary standards are prefixed ‘GB/T’.

Table 7.1 Examples of Chinese Standards
Product
Group
Standard Standard Name
Adopted
International
Standard
Application
Degree
Fish GB 14939-1994 Hygienic standard for canned fish CAC 70:1995 NEQ
Tea GB/T 23376-2009
Determination of pesticides
residues in tea - GC/MS method
- -
Source: Authors’ calculations based on SAC National Standards Query
7.3.3 Descriptive statistics
We construct a frequency measure for the four types of standards, which counts
the number of standards directly linked to agricultural commodities7
Frequency
measures or count variables are often used to measure the impact of standards
and regulation on trade. The advantage of frequency measure is that they are
relatively easy to construct. One disadvantage is, however, that they may suffer
from the so-called ‘mixed bag’ problem, which means that each standard to
regulating food safety may differ in the level of stringency and have a different
impact on trade. For instance, the ‘hygienic-performance’ standards might have
a larger effect on exports than the ‘testing standards’, or vice versa. Due to the
absence of variables indicating the complexity or stringency of the standards
we believe that the number of standards represents a good proxy for the level of
regulatory intensity in China and has been widely used in the literature estimating
the impact of standards on trade (see, for instance, Swann et al, 1996; Blind and
Jungmittag, 2005; Moenius, 2006; Portugal-Perez et al, 2010; and Otsuki et al,
2001). Figure 7.1 depicts the evolution of the number of standards in the Chinese
food products studied here. The dotted lines in Figure 7.1 reveal that the share of
mandatory harmonised international standards decreased from 10.4% in 1992 to
8.9% in 2008 whereas the share of voluntary international harmonised standards
increased from 9.8% in 1992 to 15.0% in 2008. The transition from mandatory
standards to voluntary standards and the increased adoption of harmonised
international standards reflects China’s commitment to fulfil the obligations
under the WTO TBT and SPS Agreement.8
7 Further details on the evolution of Chinese standards are discussed in Mangelsdorf et al (2012)
8 In November 2003, China stated in the WTO Committee on Sanitary and Phytosanitary Measures
that ‘The Government of China attaches great importance to the principle that SPS measures shall
be based on international standards. In December 2001, AQSIQ issued “Regulatory Measures
Governing the Adoption of International Standards” (AQSIQ Decree No. 10), which stipulates
clearly the principles and procedures for adopting international standards’ (G/SPS/GEN 452 from
November 2003).

Figure 7.1 Stock of voluntary and mandatory Chinese food standards
0
100
200
300
400
500
600
700
800
900
1992 1994 1996 1998 2000 2002 2004 2006 2008
0%
2%
4%
6%
8%
10%
12%
14%
16%
Mandatory Standards
Voluntary Standards
Percentage of harmonized mandatory standards (right hand scale)
Percentage of harmonized voluntary standards (right hand scale)
Regarding harmonisation differences between product groups, Table 7.2 shows
in 4-year intervals the percentage of Chinese standards that are unique to the
country and not harmonised to international measures. The percentage of
Chinese domestic standards is decreasing over time or – in other words – China
increasingly bases its standards on international standards.
Table 7.2 Percentage of Chinese domestic standards 1992–2008
Product 1992 1996 2000 2004 2008
Meat 90.6 91.4 92.2 92.1 88.9
Fish 100.0 100.0 100.0 100.0 97.0
Vegetables 88.9 84.6 83.7 86.9 83.0
Cereals 100.0 98.9 99.0 98.6 88.7
Milk 73.6 75.4 75.4 75.0 80.5
Tea 100.0 100.0 100.0 75.0 74.3
Sugar 100.0 100.0 100.0 100.0 86.7
Source: Authors’ calculations based on SAC National Standards Query.
7.3.4 Results
The estimation is carried out in a panel covering Chinese agricultural exports
to 132 countries over the period 1992–2008. Table 7.3 summarises descriptive
statistics on the sample over which estimates are carried out. Notice that there
are fewer observations for consumption and tariff, as data for these variables is
sometimes not reported for some goods in an importing country. Yet, coefficient
estimates for standards do not change when these variables are excluded from
regressions, as shown below.

Table 7.3 Chinese agricultural exports: descriptive statistics
Model variable Observations Mean Standard deviation
lnXikt
7646 5.98 2.99
StMIkt
7646 1.12 1.50
StMDkt
7646 14.42 9.36
StVIkt
7646 4.77 6.22
StVDkt
7646 42.32 37.03
lnCONSikt
6848 19.82 2.88
Tariffikt 6089 0.13 0.16
Source: Authors’ calculations based on SAC National Standards Query.
Table 7.4 reports OLS estimates for our specification. Column 1 reports
estimates from the baseline model (Equation 1). All coefficients have the
expected signs. Regarding the relative impact of standards, internationally
harmonised mandatory standards seem to have the largest impact on Chinese
exports. As consumption and tariff data have missing observations for some
sectors in some countries, we exclude these variables from the baseline
specification. As shown in column 2, estimates of standards coefficients do not
change significantly to the omission of tariffs and consumption variables. In
order to control for characteristics specific to products, importer, and years, we
include the interaction of product-year dummies, importer-product dummies and
importer-year dummies. Columns 1 and 2 include all three interaction variables
and in columns 3 and 4 we only include the interaction of product-year dummies
and importer-product dummies. The coefficient for mandatory internationally
harmonised standards has the expected positive sign in all models. Voluntary
internationally harmonised standards also have the expected positive sign but the
coefficient is insignificant when excluding consumption and tariffs. Regarding
the relative impact, internationally harmonised mandatory standards seem to
have the largest impact on Chinese exports. The impact of domestic standards is
less clear. Mandatory domestic standards exert a negative influence on exports
in columns 2 to 4 but have no significant effect in column 1. Similarly, voluntary
internationally harmonised standards are positively correlated with exports in
column 3 but have no impact on exports in the other models. Thus, standards
exert a positive impact on exports when they are harmonised with international
norms. Estimates for other products are reported in Table 6A.2 in the Appendix..

Table 7.4 Baseline results of OLS estimates
Model variable
1 2 3 4
Base Base Base Base
Mandatory international standards StMIkt
1.538 1.480 0.497 1.214
[0.203]*** [0.217]*** [0.025]*** [0.462]**
Mandatory domestic standards StMDkt
-0.047 -0.310 -0.06 -0.112
[0.040] [0.103]** [0.005]*** [0.038]**
Voluntary international standards StVIkt
0.192 0.115 0.024 0.006
[0.056]** [0.121] [0.004]*** [0.021]
Voluntary domestic standards StVDkt
0.002 0.035 0.003 0.001
[0.013] [0.027] [0.001]** [0.004]
Consumption (log) lnConskt
0.122 0.127
[0.046]** [0.022]***
Tariff Tariffikt
0.083 -2.585
[0.655] [1.408]
Constant -0.847 6.862 2.834 2.661
[0.630] [1.268]*** [0.308]*** [1.391]
Observations 5555 5278 5555 7646
R² 0.85 0.86 075 0.69
Importer-product dummies Yes Yes Yes Yes
Product-year dummies Yes Yes Yes Yes
Importer-year dummies Yes Yes No No
Notes: The dependent variables are bilateral exports. Robust standard errors are clustered by product
groups and are reported in brackets. The asterisks represent the level of significance: * significant at
10%; ** significant at 5%; *** significant at 1%.
We also tested the robustness of our regression results to different specifications.
All regressions with robustness tests are reported in Mangelsdorf et al. (2012), we
present here a brief discussion. First, we aimed at measuring whether the effect of
standards is different when the importer is a high-income country and constrain
the sample to include only high-income importers. The results are qualitatively
similar to the baseline model. The impact of mandatory standards is positive and
the effect of voluntary harmonised standards. Domestic mandatory standards
exert a negative influence on exports and the effect of voluntary standards is
positive but smaller than the effect of harmonised standards. The results suggest
that high-income importers seem to value harmonised standards – mandatory in
particular – as more important than domestic standards. Second, standardisation
could also be influenced by the volume of exports in a particular sector through
lobbying from exports, which may lead to the possible endogeneity of our
standards measure due to reverse causality.However, in the case of Chinese
standards, it is unlikely that they are set in response to unexpected surges in food
imports demand from a single country in a single year, given the complexity and
the high number of Chinese agencies involved in making standards. Moreover,
generating standards is a time-consuming process. It requires the coordination

of preference from various interests groups. In international standardisation, the
time between the first preposition of a new standard and the final publication
can take more than 5 years (Blind and Jungmittag, 2005). In China, the average
time for generating a new standard is 4.7 years (Ping et al, 2010).Yet, as another
robustness check, we include alternately in our regressions two-year and three-
year lagged standards. Qualitatively, results are the same; standards exert a
positive impact on export with the highest effect for mandatory harmonised
standards.
7.5 Conclusion
This chapter presented empirical evidence on the impact of Chinese standards
for export of agricultural goods. Whereas past literature on the trade impact of
standards for developing countries has focused on the impact of standards and
standards harmonisation in the importing (developed) country, we argue that
the effect of standards application in the exporting country has to be taken into
account.Inparticular,Mangelsdorfetal(2012)showthatstandardsharmonisation
in China has a positive effect for China’s export success. The results of our
estimates for seven product categories can be summarised as follows. First,
based on new data, results confirm the trade enhancing potential of standards
outlined in previous studies (see, for instance, Swann, 2010). A larger stock of
standards in China is associated with expanding exports. Second, we can show
that the push effect of standards is larger when they are based on international
standards such as Codex Alimentarius standards. Estimates suggests that one
additional internationally harmonised standard in China is associated with an
increase in agricultural exports ranging between 0.5% and 1.54%. Third, the
trade effects of standards are different for voluntary and mandatory standards.
The impact of the mandatory standards is positive and statistically significant for
both purely domestic and international harmonised standards. The trade impact
for voluntary standards is less clear.Although voluntary domestic standards have
a positive impact on Chinese exports in most model specifications, the impact is
either smaller or statistically not significant compared to mandatory standards.
This may be due to the fact that mandatory regulations affect many more aspects
of food products along production and supply chains, among other factors.
We explain the positive impact of standards in general with the information
asymmetry reducing effect. Standards – whether voluntary or mandatory –
increase the transparency of the food products by providing information such
as maximum levels of pesticides to consumers. Obviously, standards reduce
information asymmetries between Chinese producers and foreign consumers.
The trade effect of international harmonised standards is larger than the impact
of purely domestic standards and this is particularly true for international
harmonised mandatory standards. This can be explained with the signalling
effect of standards regarding quality and food safety. Producers have an incentive
to apply standards as they are a signal for the producers’ investment in high-
quality products. In the case of Chinese food products, international mandatory

standards seem to have a larger signalling effect than purely domestic standards.
Foreign customers seem to acknowledge the producers’ decision to invest in
compliance with international standards.
We conclude the chapter with the following policy suggestions that could
also be relevant for other emerging economies. First, results show a larger trade
impact of harmonised standards compared to purely domestic standards. The
harmonization of Chinese domestic standards with international standards may
be lagging behind. Compared to countries in the European Union, for instance,
the percentage of international standards in China’s standards portfolio is
relatively low. Chinese authorities should accelerate the adoption process to
meet the requirements of internationally accepted food safety standards. In case
international standards are deemed to be too stringent for adoption in China,
Chinese authorities might consider negotiating Mutual Recognition Agreements
with other authorities from export destinations. Mutual recognition can reduce
the risk of border rejection and also avoid unnecessary duplicative testing of
food and agricultural products. Negotiating Mutual Recognition Agreements
might also be the first step of further cooperation between authorities of trading
partners. Second, it is not enough to adopt international harmonized standards
but it is also important to implement the necessary infrastructure to enforce
standards. Therefore, China should also invest capacity building measures
related to food safety supervision such as development of testing laboratories
and personnel for market surveillance. Third, China should continue to expand
participation in international standards-setting organisations such as Codex
Alimentarius Commission. Participation in international organisations allows
Chinese stakeholders to include specific national preference in international
norms and may facilitate the adoption at the national level. Fourth, under the
WTO TBT and SPS Agreement member countries should use international
standards where they exist but are free to set domestic standards to fulfil
regulatory objectives such as food safety. When developing domestic standards,
Chinese authorities can leverage incentives in WTO obligations to expand trade –
to the benefit of both China and global consumers. Existing research suggest that
developing countries can benefit from the adoption of international standards to
access foreign markets. Adoption of international standards can be facilitated
when developing countries actively participate in committees of international
standards-setting organisations. As a result, countries can influence the content
of the standards used by their trading partners. Although technical committee
work is related to participation costs, participation decreases the risk of adopting
standards that are too sophisticated or excessively costly for domestic producers.
References
Anderson, J E and van Wincoop, E (2003), ‘Gravity with Gravitas: A Solution to the
Border Puzzle’, American Economic Review, 93 (1), pp 170–92.
Asian Development Bank (2007), ADB Observations and Suggestions Policy Note,
‘Suggestions on Strengthening Food Safety in the PRC’.

Bai, L, Ma, C, Yang, Y, Zhao, S and Gong, S (2007), ‘Implementation of HACCP
System in China: A Survey of Food Enterprises Involved’, Food Control, 18, pp
1108–12.
Blind, K and Jungmittag, A (2005), ‘Trade and the Impact of Innovations and
Standards: The Case of Germany and the UK’, Applied Economics, 37 (12), pp
1385–98.
Broughton, E I and Walker, D G (2010), ‘Policies and Practices for Aquaculture
Food Safety in China’, Food Policy, 35 (471), p 478.
Chen, C, Yang, J and Findlay, C (2008), ‘Measuring the Effect of Food Safety
Standards on China’s Agricultural Exports’, Review of World Economics, 144 (1),
pp 83–106.
Czubala, W, Shepherd, B and Wilson, J S (2007), ‘Help or Hindrance? The Impact of
Harmonized Standards on African Exports’, Policy Research Working Paper 4400,
The World Bank Development Research Group Trade Team.
Gale, F and Buzby, J C (2009), ‘Imports from China and Food Safety Issues’, United
States Department of Agriculture Economic Information Bulletin No 52.
Henson, S (2004), Standards and Trade: An Overview, Guelph, ON: University of
Guelph, Canada.
Henson, S and Jaffe, S (2008), ‘Understanding Developing Country Strategic
Responses to the Enhancement of Food Safety Standards’, The World Economy,
31 (4), pp 548–68.
Hudson, J and Jones, P (2003), ‘International trade in “Quality Goods”? Signalling
Problems for Developing Countries’, Journal of International Development, 15
(8), pp 999–1013.
Jin, S, Zhou, J and Ye, J (2008), ‘Adoption of HACCP System in the Chinese Food
Industry: A Comparative Analysis’, Food Control, 19, pp 823–8.
Leland, H E (1979), ‘Quacks, Lemons, and Licensing: Theory of Minimum Quality
Standards’, Journal of Political Economy, 87 (6), pp 1328–46.
Mangelsdorf, A (2011), ‘The role of technical standards for trade between China
and the European Union’, Technology Analysis Strategic Management, 23(7),pp
725-743
Mangelsdorf, A, Portugal-Perez, A and Wilson, J S (2012), ‘Food Standards and
Exports: Evidence from China’, World Trade Review, 11(3), pp 1–20, forthcoming.
Maskus, K E, Otsuki, T and Wilson, J S (2004), ‘The Costs of Complying with
Foreign Product Standards for Firms in Developing Countries: An Econometric
Study’, Research Program on Political and Economic Change, Working Paper
PEC2004-0004.
Moenius, J (2006), ‘The Good, the Bad and the Ambiguous: Standards and Trade in
Agricultural Products’, IATRC Summer Symposium, 28–30 May, Bonn.
Otsuki, T, Wilson, J S and Sewadeh, M (2001), ‘What Price Precaution? European
Harmonisation ofAfloxin Regulations andAfrican Groundnut Exports’, European
Review of Agricultural Economics, 28 (2), pp 263–83.
Ping,W,Yiyi,WandHill,J(2010),‘StandardizationStrategyofChina–Achievements
and Challenge’, East-West Center Working Paper, Economics Series #107.

Portugal-Perez, A, Reyes, J D and Wilson, J S (2010), ‘Beyond the Information
Technology Agreement: Harmonization of Standards and Trade in Electronics’,
The World Economy, 33 (1870), p 1897.
Song, H and Chen, K (2010), ‘Trade Effects and Compliance Costs of Food Safety
Regulations: The Case of China’, Agriculture and Agricultural Science Procedia,
1, pp 429–38.
Swann, G P (2010), ‘International Standards and Trade: A Review of the Empirical
Literature’, OECD Trade Policy Working Papers, No 97.
Swann, G P, Temple, P and Shurmer, M (1996), ‘Standards and Trade Performance:
The British Experience’, Economic Journal, 106 (438), pp 1297–1313.
Wilson, J S, Otsuki, T and Majumdsar, B (2003), ‘Balancing Food Safety and
Risk: Do Drug Residue Limits Affect International Trade in Beef?’ Journal of
International Trade and Economic Development, 12 (4), pp 377–402.
Yue, N, Kuang, H, Sun, L, Wu, L and Xu, C (2010), ‘An Empirical Analysis of the
Impact of EU’s New Food Safety Standards on China’s Tea Export’, International
Journal of Food Science and Technology, 45, pp 745–50.

8
Standards Harmonisation and
South–South Trade
Anne-Célia Disdier, Lionel Fontagné and Olivier Cadot
8.1 Introduction
As part of the modernisation of its regulatory environment, Mauritius has recently
been considering the adoption of a technical regulation on toxic chemicals used
in paints that would closely resemble EU regulations. The benefit of harmonising
technical regulations with the EU is that Mauritius’s paint manufacturers could
have their products distributed in EU supermarkets, including in the nearby
Reunion. However, production costs would increase,1
and the increase would, in
all likelihood, price them out of East and South African markets where chemical
residues in paints are not an issue. Would enhanced access to Northern markets
compensate foreseeable losses on Southern ones?
In a different setting, Michael Jensen and John Keyser (2011) argue, in their
contribution to the present volume, that a recent decision by the East African
Community (EAC) to harmonise Member States’ technical regulations for dairy
products with international ones is likely to stifle intra-regional trade in dairy
products, as only the very largest operators are capable of meeting the stringent
international standards. In addition, they argue that the approach does not make
much sense given local consumption habits (no one consumes the milk without
first boiling it).
These two examples raise a common question – is there such a thing as
‘premature harmonisation’ of technical regulations by developing countries
on stringent Northern standards? The price to pay for such harmonisation is –
beyond testing and certification procedures that can be difficult to implement
in developing countries – higher production costs and ensuing losses of export
potential on low-income markets where the types of externalities addressed by
technical regulations are not primary concerns. In a world where South–South
trade grows faster than North–South (Gourdon, 2007), this may be a high price
to pay.
1 In an interview with the authors, a local producer estimated the cost increase at 2% to 40% of baseline
average costs, depending on the product.

The effect of standards on trade has been the object of a voluminous literature,
following several strands. A first strand has focused on how Northern standards
(technical regulations and private standards) affect developing-country exports,
in particular in the agri-food sector. Paarlberg and Lee (1998), Otsuki et al (2001),
Wilson and Otsuki (2004), Maskus et al (2005), Disdier et al (2008), to name
but a few, all found that food safety regulations set by Northern countries had a
depressing effect on Southern exports of foodstuffs. Beyond the particular case
of food standards, Essaji (2008) found that US technical regulations significantly
affected both the pattern of developing country exports and the probability of
export to the US. Czubala et al (2007) further showed that the trade-inhibiting
effect of Northern (EU) standards was worse when they were not harmonised
with international ones.
Some recent papers have taken a more nuanced view. From a theoretical
point of view, Jaud (2011) shows that the relationship between the stringency
of standards and the volume of imports is a complex and potentially non-
monotone one when selection and reputation effects are properly accounted for.
Empirically, some estimates of compliance costs suggest that they may not be
all that high. Two World Bank studies (Aloui and Kenny, 2004, and Cato et
al, 2005) estimated, in two very different contexts – Moroccan tomatoes and
Nicaraguan shrimps respectively – SPS compliance costs at around 3% of total
costs. Other World Bank studies have highlighted success stories of producer
adaptation to food standards, including Thai and Kenyan horticulture (Jaffee,
2003; World Bank, 2005), Thai and Nicaraguan Shrimp (World Bank 2005),
or Indian spices (Jaffee, 2005). Minten et al (2009) and Maertens and Swinnen
(2006) provided similar accounts for vegetable exports from Madagascar and
Senegal respectively. Using a sector-level gravity equation, Moenius (2004)
found that industrial-country standards raised their imports rather than reducing
them. All in all, both theory and empirics are somewhat ambiguous on how high
is the barrier created by Northern SPS standards for Southern exports.
Beyond design, a number of recent papers have looked at enforcement, with
results suggesting trade-inhibiting effects. Debaere (2010) analyses a cascade
of events affecting Asian shrimp exports to Western countries, starting with
EU enforcement, in 2001, of a zero-tolerance policy for antibiotics residues,
followed by export deflection to the US market, where they were met by a
devastating anti-dumping action. In this volume, Jouanjean, Maur and Shepherd
show that the US FDA’s enforcement of food-safety regulations at the US border
is discretionary and at least partly based on reputation effects, as past alerts on
one product imported from one country significantly raise the probability of
subsequent alerts on the same product from neighbouring countries or on similar
though distinct products.
A second strand has looked at how regional standards policies have affected
intra-regional trade and market access for outsiders, although, as highlighted by
Bourgeois et al (2007), relatively little attention has been given in the literature to
the effect of standards liberalisation in the context of Regional TradeAgreements
(RTAs). In an early contribution, Baldwin (2000) argued that mutual recognition

Standards Harmonisation and South–South Trade 157
among developed countries could well lead to a two-tier international trade
system with developing countries in the second tier. Empirically, in a paper
already discussed, Moenius (2004) found that the count of shared standards
among industrial countries significantly raised their trade. Looking at trade-
diversion effects, Amurgo-Pacheco (2006) showed that MRAs harmed third-
country exports irrespective of their level of development. Baller (2007) found
that mutual recognition of product standards in the telecom-equipment and
medical-device sectors as part of RTAs had strong effects on intra-bloc trade;
harmonisation had insignificant effects on intra-bloc trade and on the exports of
non-members if they were developing, although it had strong effects on OECD
non-member exports. Overall, the effects were stronger at the extensive margin
(the probability of positive exports). Chen and Mattoo (2008) showed that both
mutual-recognition agreements (MRAs) between industrial countries, especially
with rules of origin and, to a lesser extent, standards harmonisation, raised
significantly the probability and volume of intra-bloc trade between developed
countries, although trade with the rest of the world was reduced. MRAs without
rules of origin raised trade both within member countries and with the rest of
the world. In his contribution to this volume, Reyes finds, using firm-level data,
that the harmonisation of EU electronics standards had a pro-competitive effect
on US exporters (incumbents lost market share to entrants), but that, overall,
developing-country exporters were hurt.
Finally, two papers have shown that provisions on technical barriers to
trade (TBTs), which cover technical regulations and mandatory standards, are
widespread in regional trade agreements. In a survey of 58 RTAs with TBT
provisions, Piermartini and Budetta (2009) found more harmonisation for
technical regulations and more mutual recognition for conformity-assessment
procedures, the latter being typically preferred in US RTAs whereas the former
was preferred in EU RTAs. Piermartini and Budetta conjectured that this might
lock countries into RTAs, hampering multilateral trade liberalisation, but they
did not formally test the conjecture. In a study of 28 regional agreements signed
by the EU and the US, Horn et al (2009) found that all but two (both involving
the US) had TBT provisions. Moreover, some of those provisions (5 for the
EU and 11 for the US) were binding rather than mere recommendations. For
instance, some of the EU’s FTAs with Mediterranean countries, including the
agreements with Morocco or Tunisia, mandate the harmonisation of the partner’s
regulations on the EU’s ‘acquis communautaire’.2
In this chapter, we explore empirically the conjecture laid out at the outset
– namely, that Southern countries harmonising their standards with Northern
ones may lock themselves in a bilateral relationship, at the expense of relations
with other Southern countries. While the issue of standards harmonisation is,
in practice, largely linked to North–South RTAs, our identification strategy
2 See European Community (EC)-Morocco FTA, Article 51; European Community (EC)-Tunisia FTA,
Article 51; or EC-Palestinian Authority FTA, Article 40.

attempts to disentangle the effect of harmonisation from that of preferential trade
using RTAs with and without harmonisation.
We use a standard, theory-based panel gravity framework. In accordance with
our conjecture, we find that North–South RTAs hurt South–South trade. Our
surprisingly strong results suggest that the harmonisation of technical regulations
in North–South RTAs may well have the – presumably unintended – effect of
reinforcing hub-and-spoke patterns in world trade and even marginalising
Southern partners from the world economy.
8.2 Data and empirical specification
Our data combines Piermartini and Budetta’s annotated list of RTAs with
a panel of aggregate bilateral trade over 1990–2006 from the CEPII’s BACI
database.3
Figure 8.1 shows the number of North–South RTAs, broken down
according to the harmonisation clauses they include (or not), in three following
years: 1990, 1999 and 2006 (see Appendix Table A8.1 for detailed statistics).
The number of RTAs expanded from 4 in 1990 to 43 in 2006, covering today
19.5% of Northern imports from Southern countries. Most of the expansion in
the number of North–South RTAs is accounted for by agreements with some sort
of harmonisation clause, either for technical regulations (TR) or for conformity-
assessment procedures (CAP). Interestingly, although the figure shows that the
number of agreements promoting the use of international standards (such as ISO)
for technical regulations has shot up from 2 to 12, but their share in Northern
imports from the South has slightly declined from 12.1% to 11.8%.4
Figure 8.1 Number of North–South RTAs, by type of harmonisation clause
0
5
10
15
20
25
30
35
40
45
50
1990 1999 2006
Harmonisation of CAP
Harmonisation of TR
No harmonisation
Notes: CAP: Conformity-assessment procedures; TR: Technical regulations.
Source: Based on Piermartini and Budetta (2009).
3 See http://www.cepii.fr/anglaisgraph/bdd/baci.htm. This database uses original procedures to
harmonise the United Nations COMTRADE data (evaluation of the quality of country declarations
to average mirror flows, evaluation of cost, insurance and freight rates to reconcile import and export
declarations). See Gaulier and Zignago (2010).
4 The figure for the share of Northern imports refers to all trade between trading partners; not trade
eligible for non-zero tariff preferences (for instance, zero-MFN tariff lines are included in covered
trade).

The list of 43 RTAs considered in our exercise is provided in the appendix. It
follows the format of Piermartini and Budetta (2009), updating it with some
recent agreements. We mark bilateral South–South trade flows – on which we
focus in this chapter – according to whether they can be, directly or indirectly,
affected by North–South RTAs. For this, we define a vector of dummy variables
which we interact in order to get a string of binary codes for all possible
configurations.
That is, let i and j be two Southern countries, and consider a bilateral trade
flow where i is the exporter and j is the importer.
δ0
= { 1 if i and j are partners in the same (South-South) RTA (1)
0 otherwise
δ1
= { 1 if either i or j has an RTA with a Northern partner (2)
0 otherwise
We further define
δ2
= { 1 if only i or only j has an RTA with a Northern partner (3)
0 otherwise
δ3
= { 1 if only i has an RTA with a Northern partner (4)
0 otherwise
δ4
= { 1 if only j has an RTA with a Northern partner (5)
0 otherwise
The ensuing logical structure is shown in Figure 8.2. In order to exhaust cases of
interest, the figure shows, in columns, whether any given South–South bilateral
trade flow is governed by a South–South (SS) RTA. In lines, the figure breaks
down cases depending on whether any or both partners may also belong to a
North–South (NS) RTA.
The first column indexes regression equations. For each of these regression
equations, the second column lists dummy variables appearing on the right-hand
side (as explanatory variables), and the next two columns state conditions used
to characterise North–South (NS) RTAs to which any partner may belong.
Equation (1) distinguishes only between cases where a SS RTA exists and
cases where none exists. Conditional on either the existence or non-existence of
a SS RTA, Equation (2) distinguishes between cases where either i (the exporter)
or j (the importer) belongs to a NS RTA, from cases where neither belongs to
any. Similarly, Equation (3) distinguishes between flows whether only i or only j
belongs to a NS RTA, from cases where neither i nor j belong to any, or else both
belong to some. Finally, Equations (4), (5) and (6) distinguish between cases
where (i) only i (the exporter) belongs to a NS RTA, (ii) only j (the importer)
belongs. The figure shows the resulting strings of binary codes for δ0
(Equation

1), δ0
and δ1
(Equation 2), δ0
and δ2
(Equation 3), and δ0
, δ3
and δ4
(Equations
4, 5 and 6).
Figure 8.2 RTA configurations
Equation Dummies First condition Second condition yes no
(1) ?0 1 0
yes 11 01
no 10 00
yes 11 01
no 10 00
yes 110,111 010,011
yes 101 010
no 100 000
(2)
(3)
(4), (5), (6)
?0, ?1
?0, ?2
?0, ?3 , ?4
SS RTA
NS RTA for i or j
NS RTA for only i or only j
NS RTA for only i
NS RTA for only jno
Note: Binary-code strings in the boxes to the right represent the values of the dummies included in the
equation whose number appears in the first column. Thus, in Equation (1), ‘1’ means δ0
= 1; in Equation
(2), ‘11’ means δ0
= 1, δ1
= 1; in Equations (4), (5) or (6), ‘110,111’ means either δ0
= 1, δ3
= 1, δ4
= 0 or
δ0
= 1, δ3
= 1, δ4
= 1; and so on.
NS RTAs are further subdivided by the harmonisation clauses they include or
not. We focus on TBT provisions. Conditional on either δ3
= 1 or δ4
= 1 (NS
RTA for either the exporter or the importer respectively) we define the following
additional dummies:
δ5
= { 1 if technical regulations and harmonised in i’s NS RTA (6)
0 otherwise
and
δ6
= { 1 if technical regulations and harmonised in j’s NS RTA (7)
0 otherwise
We define similar dummies for the harmonisation of conformity-assessment
procedures.
Thus,wedistinguishbetweencaseswherethesignatoryistheimportingcountry
and cases where it is the exporting country, as the story is potentially different.
In the case of a NS RTA signed by the exporting country, the potential effect is
a cost-raising one if domestic producers are induced to adopt costlier production
processes on all their lines of production, say because of indivisibilities. The
expected effect is then a loss of competitive positions on the Southern partner’s
market, reflected in a reduced export flow. In that case the harmonisation acts
like a de facto production standard for the Southern country. In the case of a
NS RTA signed by the importing country, the potential effect is a trade barrier
one if products from the Southern partner are no longer in compliance when
technical regulations are harmonised with the North. The expected effect is then
a reduction in imports from the Southern partner. In both cases, the conjecture is
a reduction in the South–South bilateral trade flow.

We use a standard panel gravity equation with importer, exporter and time
fixed effects (see Feenstra, 2004; Baldwin and Taglioni, 2006). The dependent
variable is the FOB value of total bilateral imports from country i to country
j. GDPs are from the World Development Indicators database. Transport costs
are proxied by the bilateral distance between trading partners, also from CEPII.
We also add standard dummy variables measuring the ease of bilateral trade
including common border, common language and a common colonial history.
Finally, our estimator is the zero-inflated negative binomial model (see Burger
et al, 2009), which handles zero values and overdispersion.5
8.3 Results
Results are reported in Table 8.1. Coefficients on gravity variables have standard
signs and magnitudes. Distance reduces trade, while contiguity, common
language and past common coloniser raise it. The dummy variable controlling
for the existence of a South–South RTA is positive and significant at the 1%
level.
Columns (2)–(3) highlight a first trade-diversion effect, irrespective of
harmonisation clauses: The signature by either the exporter or the importer
of a RTA with a Northern partner tends to reduce its bilateral trade with other
Southern countries, thus, in itself, reinforcing a hub-and-spoke pattern in world
trade. Column (4) shows that this effect is stronger (larger in magnitude and
more precisely estimated) when the importer is engaged with a Northern partner.
Assuming a unit price elasticity of import demand, the ad valorem equivalent
(AVE) is 22.1%.6
That is, a Southern country signing an RTA with a Northern
partner imposes a negative trade-diversion effect on its Southern partners equal
to a 22.1% tariff, a very large effect.
Columns (5) and (6) confirm the paper’s conjecture. When the exporter is
engaged in an RTA with a Northern partner involving the harmonisation of
technical regulations, the inhibiting effect on its bilateral trade with Southern
partners is equivalent to a tariff at a whopping 35.0% under the maintained
assumption of a unit elasticity. The harmonisation of conformity-assessment
procedures has a similar effect (37.7%).
5 Poisson and negative binomial models assume that all pairs of countries have a positive probability of
trading, which may not be true for pairs of countries with scarce resources and wide bilateral distance.
The zero-inflated model deals with this problem by assuming the existence of two latent groups
within the population: (i) an ‘always zero group’ having an outcome of zero with a probability of one
and a ‘not always zero group’ having a non-zero probability of having a positive count. Robustness
tests using the Poisson estimator give very similar results. The approach is validated using over-
dispersion tests and the Vuong statistic comparing zero-inflated and non-zero-inflated models.
6 The ad valorem equivalent is exp(–β/ε)–1, where β is the estimated coefficient and ε the price
elasticity of import demand (in algebraic form, ie negative). The average price elasticity of imports,
over all goods and countries, is estimated by Kee et al (2008) at –3.12 at HS6 and –1.1 at ISIC3.
As we are dealing with aggregate flows here, the lower value estimated by Kee et al at the more
aggregate level (ISIC3) where substitution possibilities are reduced, is more plausible.

Table 8.1 Regression results, South–South trade
Dep. Var.: Bilateral imports between Southern countries
Estimator: Zero-inflated negative binomial
Equation (1) (2) (3) (4) (5) (6)
Gravity variables
Ln GDP i (exporter) 0.15* 0.15* 0.15* 0.15* 0.16* 0.16*
(0.06) (0.06) (0.06) (0.06) (0.06) (0.06)
Ln GDP j (importer) 0.37** 0.36** 0.36** 0.36** 0.36** 0.36**
(0.05) (0.05) (0.05) (0.05) (0.05) (0.05)
Ln distance -1.36** -1.36** -1.36** -1.36** -1.36** -1.36**
(0.03) (0.03) (0.03) (0.03) (0.03) (0.03)
Common border 0.94** 0.94** 0.94** 0.94** 0.94** 0.94**
(0.10) (0.10) (0.10) (0.10) (0.10) (0.10)
Common language 0.29** 0.29** 0.29** 0.29** 0.29** 0.29**
(0.06) (0.06) (0.06) (0.06) (0.06) (0.06)
Past common coloniser 0.45** 0.45** 0.45** 0.45** 0.45** 0.45**
(0.07) (0.07) (0.07) (0.07) (0.07) (0.07)
Agreements
SS RTA 0.48** 0.48** 0.48** 0.48** 0.48** 0.48**
(0.05) (0.05) (0.05) (0.05) (0.05) (0.05)
NS RTA for i, j or both -0.21**
(0.04)
NS RTA for either i or j but not both -0.15**
(0.04)
NS RTA for i (exporter) only -0.10* 0.09 0.09
(0.05) (0.08) (0.08)
NS RTA for j (importer) only -0.20** -0.16* -0.19**
(0.06) (0.08) (0.07)
Harmonisation clauses
NS RTA w/TR harm. for i only -0.30**
(0.08)
NS RTA w/TR harm. for j only -0.09
(0.08)
NS RTA w/CAP harm. for i only -0.32**
(0.08)
NS RTA w/CAP harm. for j only -0.05
-0.07
Constant 2.23 2.42 2.32 2.31 2.14 2.21
(1.74) (1.75) (1.74) (1.74) (1.74) (1.74)
Observations 296,376 296,376 296,376 296,376 296,376 296,376
Overdispersion 4.23** 4.24** 4.23** 4.23** 4.24** 4.24**
Vuong test 137.16** 137.00** 137.05** 137.05** 136.69** 136.60**
Note: Robust standard errors (importing country-exporting country clustered) in parentheses. **
significant at 1%; * significant at 5%. For overdispersion the alpha value is reported, for the Vuong test
the z-score.

8.4 Conclusion
This chapter suggests that gravity-based evidence is supportive of the conjecture
thatNorth–Southagreementsinvolvingtheharmonisationoftechnicalregulations
may well reinforce hub-and-spoke patterns in world trade by inhibiting the
expansion of South–South trade. The postulated mechanism behind this effect
is that FTAs like EU–Tunisia or EU–Morocco which stipulate that the Southern
partner harmonises its domestic regulations with those of the North may force
manufacturers in the Southern partner to position themselves in high-quality
segments where they do not necessarily have a competitive advantage. Even if
they do (or acquire it in a dynamic setting), the advantage conferred by, say, EU
standards may not carry much weight in Southern markets where those standards
do not correspond to local consumer concerns.
Our results suggest that harmonisation in North–South FTAs is a non-trivial
issue which should be approached strategically by Southern governments.
In sectors with strong expansion potential on the Northern market, rapid
harmonisation may be key to gain market access. However, in sectors with strong
expansion potential on Southern markets, progressive phase-in and softeners
might be considered to prevent the double-edged sword from cutting the wrong
way.
References
Aloui, O and Kenny, L (2004), ‘The Cost of Compliance with SPS Standards for
Morrocan Exports:A Case Study’,Agricultural and Rural Development Discussion
Paper, The World Bank.
Amurgo-Pacheco, A (2006), ‘Mutual Recognition Agreements and Trade Diversion:
Consequences for Developing Nations’, HEI Working Paper 20-2006, Geneva.
Baldwin, R (2000), ‘Regulatory Protectionism, Developing Nations, and a Two-Tier
World Trade System’, Brookings Trade Forum, pp 237–90.
Baldwin, R and Taglioni, D (2006), ‘Gravity for Dummies and Dummies for Gravity
Equations’, NBER Working Paper 12516, Cambridge, MA.
Baller, S (2007), ‘Trade Effects of Regional Standards Liberalization. A
Heterogeneous Firms Approach,’ World Bank Policy Research Working Paper
4124, Washington DC.
Bourgeois, J, Dawar, K and Evenett, S (2007), ‘A Comparative Analysis of Selected
Provisions in Free Trade Agreements’, Study Commissioned by DG Trade,
Brussels.
Burger, M J, van Oort, F G and Linders, G-J M (2009), ‘On the Specification of
the Gravity Model of Trade: Zeros, Excess Zeros and Zero-Inflated Estimation’,
Spatial Economic Analysis 4(2), pp 167–90.
Cato, J, Otwell, W and Coze, A (2005), ‘Nicaragua’s Shrimp Subsector: Developing
a Production Capacity and Export Market during Rapidly Changing Worldwide
Safety and Quality Regulations’, Agricultural and Rural Development Discussion

Chen, M and Mattoo, A (2008), ‘Regionalism in Standards: Good or Bad for Trade?’
Canadian Journal of Economics 41(3), pp 838–63.
Czubala, W, Shepherd, B and Wilson, J (2007), ‘Help or Hindrance? The Impact of
Harmonized Standards on African Exports’, The World Bank, Policy Research
Working Paper Series 4400.
Debaere, P (2010), ‘Small Fish – Big Issues: The Effect of Trade Policy on the
Global Shrimp Market’, World Trade Review 9(2), pp 353–74.
Disdier, A-C, Fontagné, L and Mimouni, M (2008), ‘The Impact of Regulations on
Agricultural Trade: Evidence from SPS and TBT Agreements’, American Journal
of Agricultural Economics 90(2), pp 336–50.
Essaji, A (2008), ‘Technical Regulations and Specialization in International Trade’,
Journal of International Economics 76, pp 166–76.
Feenstra, R (2004), Advanced InternationalTrade:Theory and Evidence, Princeton,
NJ, Princeton University Press.
Gaulier, G and Zignago, S (2010), ‘BACI: International Trade Database at the
Product-Level. The 1994-2007 Version,’ CEPII Working Papers 2010-23.
Gourdon, J (2007), ‘Trade and Wage Inequality in Developing Countries: South-
South Trade Matters’, CERDI, Etudes et Documents #E 2007.10
Horn, H, Mavroidis, P C and Sapir, A (2009), ‘Beyond the WTO? An Anatomy of
EU and US Preferential Trade Agreements’, Bruegel Blueprint 7, Brussels.
Vegetable Trade in the Context of Emerging Food Safety and Other Standards in
Europe’, Agricultural and Rural Development Discussion Paper, The World Bank.
Jaffee, S (2005), ‘Delivering and Taking the Heat: Indian Spices and Evolving
Product and Process Standards’, Agricultural and Rural Development Discussion
Jaud, M (2010), ‘Food Safety, Reputation, and Trade’, mimeo, PSE/University of
Lausanne.
Jaud, M (2011), ‘Food safety, reputation and trade’, Paris School of Economics,
Working Paper No. 2011-15
Jensen, M and Keyser, J (2011), ‘Standards Harmonization and Trade: The Case of
the East African Dairy Industry’, mimeo, The World Bank.
Jouanjean, M-A, Maur, J-C and Shepherd, B (2011), ‘Reputation Matters: Spillover
Effects in the Enforcement of US SPS Measures’, University of Munich MPRA
Paper 35270.
Kee, H L, Nicita,A and Olarreaga, M (2008), ‘Import Demand Elasticities and Trade
Distortions’, Review of Economics and Statistics 90, pp 666–82.
Maertens, M, and Swinnen, J F (2009), ‘Trade, Standards, and Poverty: Evidence
from Senegal’, World Development 37, pp 161–78.
Maskus, K, Otsuki, T and Wilson,. J (2005), ‘The Cost of Compliance with Product
Standards for Firms in Developing Countries: An Econometric Study’, World
Bank Policy Research Working Paper 3590.
Minten, B, Randrianarison, L and Swinnen, J (2009), ‘Global Retail Chains and Poor
Farmers: Evidence from Madagascar,’ World Development 37(11), pp 1728-1741.

Moenius, J (2004), ‘Information versus Product Adaptation: The Role of Standards
in Trade’, International Business and Markets Research Center Working Paper,
Northwestern University.
Otsuki, T, Wilson, J S and Sewadeh, M (2001), ‘Saving Two in a Billion: Quantifying
the Trade Effect of European Food Safety Standards on African Exports’, Food
Policy 26, pp 495–514.
Paarlberg, P and Lee, J (1998), ‘Import Restrictions in the Presence of a Health Risk:
An Illustration Using FMD’, American Journal of Agricultural Economics 80, pp
175–83.
Piermartini, R and Budetta, M (2009), ‘A Mapping of Regional Rules on Technical
Barriers to Trade’, in A Estevadeordal, K Suominen and R Tech (eds), Regional
Rules in the Global Trading System, Cambridge, Cambridge University Press, pp
250–315.
Reyes, J-D (2011), ‘The Pro-Competitive Effect of International Harmonization of
Product Standards’, mimeo, The World Bank.
Wilson, J S and Otsuki, T (2004), ‘To Spray or Not to Spray: Pesticides, Banana
Exports, and Food Safety’, Food Policy 29, pp 131–46.
World Bank (2005), ‘Food Safety and Agricultural Health Standards: Challenges
and Opportunities for Developing Country Exports’, PRMTR unit Report #31027;
Washington DC, The World Bank.

Appendix
List of North-South RTAs included in the study
Australia – Papua New Guinea (PATCRA) EFTA – Mexico
Canada – Chile EFTA – Morocco
Canada – Costa Rica EFTA – Singapore
Canada – Israel EFTA – Tunisia
Dominican Republic – Central America –
United States Free Trade Agreement (CAFTA-
DR)
EFTA – Turkey
EC – Albania Japan – Malaysia
EC – Algeria Japan – Mexico
EC – Chile Japan – Singapore
EC – Egypt New Zealand – Singapore
EC – Israel North American Free Trade Agreement
(NAFTA)
EC – Jordan Singapore – Australia
EC – Lebanon South Pacific Regional Trade and Economic
Cooperation Agreement (SPARTECA)
EC – Mexico Thailand – Australia
EC – Morocco Thailand – New Zealand
EC – South Africa Trans-Pacific Strategic Economic Partnership
EC – Syria US – Bahrain
EC – Tunisia US – Chile
EC – Turkey US – Israel
EFTA – Chile US – Jordan
EFTA – Israel US – Morocco
EFTA – Jordan US – Singapore
Table A8.1 North–South RTAs and Tradea
Coverage
1990 1999 2006
No.
Trade
coverage
(%)
No.
Trade
coverage
(%)
No.
Trade
coverage
(%)
RTAs 4 0.4 12 15.8 43 19.5
Of which RTAs with standards harmonisation
Harmonisation of technical regulations 0 0 5 14.2 21 15.7
Promotion of the use of regional standards only 0 0 2 2.1 6 3.8
Promotion of the use of international standards
(alone or in addition to regional ones)
0 0 2 12.1 12 11.8
Harmonisation of conformity assessment
procedures
0 0 4 14.1 16 14.7
Promotion of the use of regional standards only 0 0 2 2.1 7 4.0
Promotion of the use of international standards
(alone or in addition to regional ones)
0 0 1 12.1 6 10.1
Note: a Northern imports from the South.

9
The Pro-Competitive Effect of
International Harmonisation of
Product Standards
José-Daniel Reyes1
9.1 Introduction
Heterogeneity of product standards across market destination is a non-tariff
barrier (NTB) to international trade as the standards impose additional costs on
exporters and increase the time required to bring a product to market.This chapter
examines the impact of a reduction in this NTB by looking at the harmonisation
of European product standards with international norms in the electronics sector.
Merging European product standards information with international trade data,
I show that harmonisation increases imports to the EU but this impact varies
across exporting partners. While harmonisation tends to increase exports from
the developed world, it has an ambivalent impact on exports from the developing
world. I propose an explanation to this puzzle by looking at the response of US
manufacturing firms to this policy.
The divergent impact of harmonisation across type of countries was previously
noted by Chen and Mattoo (2008) for the case of regional agreements on
standards. They provide evidence that adopting a common standard in a region
boosts exports of excluding developed countries to the region but reduces
exports of excluded developing countries. My analysis is the first, to the best of
my knowledge, to extend this result for the case of harmonisation of standards
with international norms. In addition, my analysis also differs from theirs in the
proposed explanation for this feature. While they argue that harmonisation of
standards may reduce exports of excluding developing countries because they
1 The author is grateful to Rod Ludema, Brad Jensen,Andrew Bernard, Peter Schott, Mariem Malouche,
and Oliver Cadot for helpful comments and discussions. Part of the research in this chapter was
conducted while the author was a Special Sworn Status researcher at the US Census Bureau at the
Center of Economic Studies in Washington DC. Any opinion and conclusions herein are those of the
author and do not necessarily represent the views of the US Census Bureau. Results using US census
data have been reviewed to ensure that no confidential information is disclosed.

are hurt more by an increase in the stringency of standards and benefit less of
scale in integration, I argue that the driven attribute is the pro-competitive effect
in the harmonised market.2
I examine the response of US manufacturing firms to a reduction of this NTB
by linking industry-level EU product standards with US trade linked firm-level
data. I find that EU international harmonisation of standards triggers US firms’
entry into the EU – the extensive margin of trade – which fosters competition
between firms and reduces market shares – the intensive margin of trade. This
feature is what I called the pro-competitive effect of international harmonisation
of standards.3
Since the intensive margin effect impacts profits of all market
participants negatively, the key feature that determines the overall impact in
export value, at the country level, is the availability of new firms to start exporting
to the EU – the extensive margin of trade. The empirical evidence suggests that,
on average, the extensive margin effect outweighs the intensive margin effect for
developed countries whereas the reverse is true for developing countries.
Product standards are documented agreements containing technical
specifications or other precise criteria to be used consistently as rules, guidelines
or definitions of characteristics to ensure that materials, products, processes
and services are fit for their purpose. In contrast with technical regulations,
compliance is not compulsory. Product standards, on the one hand, convey
information on several product dimensions such as industrial requirements,
compatibility, consumer tastes and safety, which would be costly to collect in
their absence (informational effect).On the other hand, they impose additional
cost for exporters to alter production processes to adapt products to standards
specific to each market (cost effect). Furthermore, having country-specific or
region-specific standards makes life more difficult for exporters since they need
to adapt the production process to comply with different rules. Examples of
product standards are safety requirements for sewing machines, measures of
electromagnetic emissions from integrated circuits, specific guards for lawn
mowers, and mechanical safety of cathode ray tubes.
If product regulations are set arbitrarily they could be used as an excuse
for protectionism (‘regulatory protectionism’ in Baldwin, 2000). This may be
appealing to countries since international disputes on standards have traditionally
been regarded as matters of safety, health and environmental regulations that
fall within the primary competences of domestic governments. They are quite
different from disputes on tariffs, quotas, or intellectual property which clearly
fall within the jurisdiction of the World Trade Organisation (WTO). While the
Agreement on Technical Barriers to Trade (TBT), signed by the WTO member’s
countries in 1995, encourages countries to adopt international standards
2 In addition, Reyes (2011) presents a model featuring the role of product standards heterogeneity
across market destinations and productivity heterogeneity across firms where, independent of the
level of stringency of the harmonised standard, harmonisation always increases firm-entry into the
harmonised market, which in turn increases competition.
3 For a theoretical analysis of pro-competitive effect of trade policies and its impact on economic
welfare see Moraga-González and Viaene (2003).

International Harmonisation of Product Standards 169
whenever possible, it also recognises the rights of countries to adopt measures to
the extent they consider appropriate – for example, for human, animal or plant
life or health, for the protection of the environment, or to meet other consumer
interests such as the prevention of deceptive practices.
While the use of standards remains voluntary, the European Union has,
since the mid-1980s, made an increasing use of standards in support of its
policies and legislation. The European Commission sets compulsory regulatory
goals by means of ‘New Approach Directives’, which outline ‘essential
requirements’ associated with the manufacturing of products. The system in
place does not, however, specify how specific objectives should be achieved.
For the electronics sector, this role is fulfilled by product standards issued by
the European Committee for Electrotechnical Standardization (CENELEC).4
EU member countries are obliged to adopt these standards and withdraw any
national standard that might conflict with them. If a manufacturer chooses to
produce a product according to these standards, the product carries the CE mark,
which implies compliance with the ‘essential requirements’. On the other hand,
manufacturers may use other technical specifications when manufacturing a
product provided there is documentation certifying that the product meets the
‘essential requirements’ formulated in the Directives. Nevertheless, anecdotal
evidence suggests that the prohibitive costs of the latter option push exporting
companies to favour compliance with CENELEC standards (see, for instance,
Hanson, 2005 and United States Trade Representative (2010)). At the global
level, the International Electrotechnical Commission (IEC) is the organisation
that prepares and publishes international standards for the electronics sector.
To study the impact of European product standards harmonisation on market
competition, I use the CENELEC-IEC agreement to harmonise European product
standards to international norms as a policy experiment. The LuganoAgreement,
signed in 1991, and the Dresden Agreement, signed in 1996, sought to expedite
the adoption of international standards in the EU as well as to facilitate the
adoption of EU standards internationally. This synergy has taken the number of
purely European standards as a share of all standards published by CENELEC
from 50% in the early 1990s to 25% in 2008. In this context, the decrease in the
share of idiosyncratic standards is a reduction in a NTB to international trade.
This chapter is structured as follows. The next section describes the data
and presents summary statistics. Section 3 presents the impact of international
harmonisation of European product standards on EU imports in the electronics
sector. Section 4 shows evidence of the pro-competitive effect of international
harmonisation using US firm-level data. Section 5 concludes.
4 For example, the Low Voltage Directive (2006/95/EC) outlines ‘essential requirements’ for electrical
equipment with a voltage between 50 and 1000 V for alternating current and between 75 and 1500
V for direct current. Among other conditions, it establishes that ‘persons and animals are adequately
protected against the danger of physical injury or other harm which might be caused by direct or
indirect contact’. Consequently, CENELEC issued a standard (EN 50371:2002) to demonstrate the
compliance of low power electronic and electrical apparatus with the basic restrictions related to
human exposure to electromagnetic fields (10 MHz - 300 GHz).

9.2 Data
The empirical analysis concentrates on the Electronic sector defined by the 4-digit
industries classified under chapter 36 in the Standard Industrial Classification
(SIC). This sector was chosen because of the availability of the EU product
standards data and the level of disaggregation was selected in order to be able to
use firm-level information from the US Census of manufacturing firms for years
prior to 2002. This sector consists of 36 four-digit SIC Industries (SIC4) that
range from vehicular lighting equipment and electric lamps to semiconductors
and transformers. Table 9.1 provides a description of the relative level of detail
between industries.
The pattern of trade in electronics between the original EU-15 members5
and its
trading partners is dominated by exports from high income countries. Table 9.2
shows the evolution of export shares across income groups and main exporting
countries. Although high income countries represent the bulk of exports to the
EU, middle income countries have gained market share in the last decade. At
the country level, the USA, Japan and Korea – high income countries – have
lost market share throughout the time span while China, Philippines and Turkey
– middle income countries – have more than doubled their participation. The
market share of low income countries is almost zero.
To assess the degree of harmonisation of EU product standards with
international norms I use the World Bank EU Electrotechnical Standards
Database (EUESDB). The EUESDB provides the first catalogue of European
standards in the electrotechnical sector6
and their relationship with worldwide
standards. The database provides an inventory of the ‘stock’ of active standards7
issued by CENELEC and their link with standards published by the International
Electrotechnical Commission (IEC). Product standards are classified according
to the International Classification of Standards (ICS) and the database covers the
period 1990–2007.8
See Portugal-Perez et al (2010) for a full description of the
EUESDB and its construction. To assign product standards into SIC4 industries
I use the mapping developed by Reyes (2011).
5 Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, the
Netherlands, Portugal, Spain, Sweden and the United Kingdom.
6 This sector refers to electrical, electronic and related technologies. More information is available at
www.iec.ch.
7 The primary variable of interest is the total number of standards with which an exporter should
comply during a particular year.
8 A list of the ICS codes can be found at www.iso.org/iso/ics6-en.pdf.

Table 9.1 Four-digit SIC codes and descriptions for SIC code 36
Industry Description
36 Electronic and Other Electrical Equipment and Components,
Except Computer Equipment
3612 Transformers
3613 Switchgear and switchboard apparatus
3621 Motors and generators
3624 Carbon or graphite products
3625 Relays and industrial controls
3629 Electrical industrial apparatus
3631 Household cooking equipment
3632 Refrigerators and refrigerating equipment
3633 Household laundry equipment
3634 Electric housewares and fans
3635 Household vacuum cleaners
3639 Household appliances, nec
3641 Electric lamps
3643 Current-carrying wiring devices
3644 Noncurrent-carrying devices
3645 Residential Electric Lighting fixtures
3647 Vehicular lighting equipment
3648 Lighting equipment, nec
3651 Radio and tv receiving sets, phonographs, record players record
3652 Phonograph records; pre-recorded magnetic tapes or wires master
3661 Telephone and telegraph apparatus
3663 Radio, broadcast, and television communications equipment
3669 Other communications equipment, nec
3671 Electron tubes
3672 Printed circuit boards
3674 Semiconductors and related devices
3675 Electronic capacitors
3676 Resistors for electronic applications
3677 Electronic coils and transformers
3678 Connectors, for electronic applications
3679 Electronic components, nec
3691 Storage batteries
3692 Primary batteries
3694 Electrical starting and ignition equipment for internal combustion engines
3695 Recording media
3699 Electrical equipment and supplies, nec
Note: This table provides the codes and description of the 36 four digit SIC industries included in the
sample. Some names are truncated to reduce clutter.

Table 9.2 Pattern of trade in electronics between the EU-15 and its trading
partners
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
High income
countries
82.6 82.1 80.1 79.2 78.8 78.3 75.6 72.5 67.6 63.3 61.1
United States 26.4 26.4 28.4 28.7 28.2 27.5 26.7 26.4 20.5 15.6 13.3
Japan 26.4 24.0 20.2 18.4 18.7 18.7 18.5 15.6 14.8 14.1 13.0
Korea 6.2 6.8 5.4 4.4 4.3 5.4 5.9 5.0 5.7 6.1 7.8
Hungary 1.5 1.8 2.3 3.0 3.7 3.7 3.6 4.5 5.1 5.6 6.3
Singapore 4.9 4.6 4.6 4.6 4.7 4.3 4.2 3.3 3.0 3.2 3.7
Czech Republic 1.1 1.4 1.5 1.6 2.1 2.3 2.3 3.1 3.7 3.9 3.6
Poland 0.9 1.1 1.3 1.7 2.1 2.1 1.7 2.4 3.2 3.4 2.9
Switzerland 5.4 5.4 4.9 4.2 4.3 4.0 3.0 3.4 3.5 3.2 2.9
Hong Kong
SAR, China
3.2 3.2 3.4 3.7 3.8 3.7 2.3 2.0 1.7 1.7 1.7
Slovak Republic 0.2 0.2 0.3 0.4 0.6 0.5 0.5 0.7 0.9 1.0 1.0
Canada 1.3 1.3 1.7 1.6 1.7 1.7 2.3 1.7 1.2 0.9 0.9
Israel 0.7 0.7 0.9 1.0 1.3 1.2 1.2 1.2 0.9 0.7 0.7
Slovenia 0.8 0.8 0.8 0.8 0.8 0.7 0.6 0.7 0.9 0.9 0.7
Rest 3.6 4.4 4.4 5.1 2.5 2.5 2.8 2.5 2.5 3.0 2.6
Upper middle
income
countries
7.1 7.7 8.3 7.9 7.8 7.3 8.2 9.1 9.9 11.3 10.9
Malaysia 5.1 5.5 6.1 5.3 4.5 4.1 4.7 4.9 4.7 5.2 4.4
Turkey 0.7 0.8 0.9 1.1 1.6 1.4 1.2 1.4 2.2 2.1 2.2
Costa Rica 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.1 0.4 1.2 1.4
Romania 0.2 0.2 0.2 0.2 0.3 0.3 0.6 0.7 0.8 0.9 1.0
Mexico 0.3 0.4 0.4 0.5 0.6 0.6 0.7 1.0 0.8 0.6 0.7
Rest 0.8 0.8 0.7 0.8 0.8 0.9 0.8 1.0 1.0 1.3 1.2
Lower middle
income
countries
10.2 10.2 11.5 12.8 13.3 14.3 16.2 18.4 22.4 25.4 27.9
China 6.2 6.0 6.3 6.7 7.6 8.5 10.7 12.5 15.9 18.4 20.9
Philippines 1.0 1.1 2.0 2.2 1.5 1.7 1.9 2.0 2.0 2.5 2.5
Thailand 1.4 1.4 1.5 2.0 1.9 1.9 1.7 1.5 1.6 1.5 1.7
Indonesia 0.6 0.5 0.5 0.6 0.6 0.6 0.7 0.8 1.0 0.9 0.8
Tunisia 0.4 0.4 0.4 0.4 0.5 0.5 0.5 0.6 0.7 0.7 0.7
Rest 0.6 0.8 0.8 0.9 1.2 1.1 0.7 1.0 1.2 1.4 1.3
Low income
countries
0.1 0.0 0.0 0.0 0.1 0.1 0.0 0.0 0.1 0.0 0.0
Note: This table shows the export share by country and by income group. Numbers are the share of
export value of each cell (country-year or income-year) in total EU imports in a given year. Country
classification is taken from the World Bank Country Classification (http://data.worldbank.org/about/
country-classifications).
Source: COMTRADE

In order to see the changes in the harmonisation rate within industries and
across time, I define the non-harmonised share of standards for industry i in
year t (NHi
t ) as the number of CENELEC standards that are not ‘identical’ to
an existing IEC standard as a share of the total number of standards in each
SIC4 industry. This definition is convenient because a reduction in this variable
(higher harmonisation level) is interpreted as a reduction in a NTB and can
be compared with a reduction in tariff duties. Tariff rates are also computed
at the industry-exporter-year level (ti
jt) as the weighted average rate across all
6-digit HS products within each SIC4 industry, using the EU’s import value as
weights. For some products, tariffs were binding to zero by year 2000 due to the
Information Technology Agreement (ITA), which is a tariff-cutting mechanism
enforced by the WTO between nations accounting for at least 90% of world IT
trade. These zero-tariff bindings were on an MFN basis and thus available to
exports from any other WTO member country.
Table 9.3 reports average tariffs and non-tariff trade costs across SIC4
industries for five-year intervals from 1994 to 2004. European tariff rates decline
across a broad range of industries over time in the Electronic sector. Indeed, over
the entire period, tariffs were halved for approximately 40% of industries. The
rate of tariff declines, however, varies substantially across industries. According
to the directives laid down by the Lugano and the Dresden agreements, European
product standards have progressively been harmonised to international norms.
The decline in the non-harmonised share of standards also differs across
industries. The highest reduction is among industries producing household
appliances, including cooking equipment, refrigerators, laundry equipment and
vacuum cleaners.
The trade costs constructed here have several advantages. First, they are
derived directly from a database used by firms to document the regulation
requirements to export to the European Union. Second, they vary across
industries and time. Even with these advantages, some caveats should be noted.
First, the EUESDB does not provide information on which to base an assessment
of the relative technical complexity of individual standards. Constructing such a
measure requires highly specialised technical and commercial information that
is currently not available. Second, product standards might vary across products
within an industry. Mapping standards to products, however, is quite difficult and
would have to be done manually.9
Given the number of standards for electronic
products this option is not currently feasible.
9 A manual mapping has been implemented for the textiles sector in Shepherd (2006) and used by
Shepherd (2007) and by Czubala et al (2009).

Table 9.3 Trade Costs by Industry and Year
SIC4
Tariff rate (%) Non-harmonised share of stds(%)
1994 1999 2004 1994 1999 2004
3612 1.2 2.0 0.3 76.9 49.1 46.3
3613 0.9 0.6 0.1 60.4 55.8 40.7
3621 0.7 0.6 0.2 66.7 67.3 65.2
3624 1.9 1.5 0.2 11.1 5.9 4.6
3625 0.9 0.7 0.1 71.2 58.3 41.3
3629 2.2 0.5 0.0 58.3 67.7 58.8
3631 1.8 1.5 0.5 91.4 28.7 29.3
3632 1.0 1.3 0.3 91.4 28.7 29.3
3633 1.2 1.2 0.3 81.1 39.5 46.2
3634 1.6 2.1 0.3 90.1 22.9 29.3
3635 1.7 1.8 0.4 71.4 61.3 44.1
3639 1.7 2.4 0.4 79.6 66.7 62.0
3641 1.4 0.7 0.2 42.4 35.5 30.7
3643 1.0 0.6 0.2 50.8 40.2 40.7
3644 2.7 2.9 0.6 22.9 21.5 19.9
3647 2.0 2.0 0.4 42.4 35.5 30.7
3648 1.3 1.8 0.4 42.4 35.5 30.7
3651 1.1 2.6 1.1 14.3 23.0 25.2
3652 0.6 1.3 0.1 5.6 8.1 8.9
3661 0.9 0.6 0.0 10.5 10.5 12.7
3663 0.8 1.4 0.2 20.9 16.5 14.3
3669 0.9 1.0 0.1 5.6 8.1 8.9
3671 1.9 2.6 1.0 75.0 69.2 75.0
3672 1.3 0.7 0.0 41.0 43.4 32.3
3674 2.6 0.0 0.0 100.0 84.2 39.0
3675 2.0 0.6 0.0 58.3 67.7 58.8
3676 1.9 0.3 0.0 70.6 66.7 65.2
3677 1.0 1.4 0.3 76.9 49.1 46.3
3678 1.5 0.7 0.1 54.2 54.5 45.2
3679 1.3 0.7 0.1 100.0 90.9 62.5
3691 1.0 1.2 0.2 38.9 32.5 29.6
3692 2.5 2.7 0.6 38.9 32.5 29.6
3695 1.4 0.7 0.0 5.6 8.1 8.9
3699 1.0 0.9 0.2 47.6 47.3 43.0
Average 1.4 1.3 0.3 53.4 42.1 36.9
Note: This table summarizes tariffs and the share of non-harmonised standards across 4-digits industries.
Tariffs are weighted averages of the undelying six-digit codes, using EU import value as weights. The
final row is the unweighted average of all manufactuing industries included in the analysis.
Source: TRAINS and EUESDB.

9.3 International harmonisation of product standards and global trade
The relationship between international harmonisation of product standards and
trade has only recently started to receive attention from researchers. Otsuki et al
(2001) provides the first case study on the beneficial impact of harmonisation on
imports of groundnut products fromAfrica during 1988–98. Czubala et al (2009)
later expanded this analysis to the textiles sector to find that internationally
harmonised standards are less trade restrictive than purely European standards.
Building on this result, Shepherd (2007) finds that although standards have a
negative impact on partner country export variety – the extensive margin of
trade, international harmonisation acts as a mitigating factor. Portugal-Perez
et al (2010) offers evidence that the impact of harmonisation is also positive
in more complex goods. Specifically, they find that EU standards that are
harmonised to international norms in the electronics sector have a positive and
significant effect on EU imports. Subsequently, Reyes (2011) examines the
response of US manufacturing firms to this harmonisation process to find that it
increases the probability of observing firm entry into the European market – the
extensive margin of trade – and that new entrants are drawn mainly from the
most productive set of firms already exporting to less stringent markets before
harmonisation – the extensive margin of trade composition.
Overall, the literature indicates that harmonisation of product standards to
international norms constitutes a way to boost trade. Although standards tend
to hinder trade in simple products (including textiles and agricultural goods)
and promote trade in complex products (including machinery and electronics
goods),10
harmonisation always exerts a positive net influence on trade flows. A
graphic way to illustrate this effect for the electronics sector is depicted in Figure
9.1. Between 1994 and 2004, it shows the relationship between the log difference
in total EU-15 import value in each industry and the corresponding change in
the non-harmonised share of product standards. The graph shows a negative
correlation indicating that international harmonisation of product standards in
the electronics sector is associated with higher import volumes in Europe.
10 This divergent effect is explained by the dual impact of product standards on production and trade
costs. On the one hand, standards may impose additional costs on exporters as it may be necessary to
adapt products for specific markets (cost effect). On the other hand, standards can reduce exporter’s
information cost if they convey relevant market information, like industrial requirements or consumer
tastes, which would be costly to gather in the absence of the standards (informational effect). As it
turns out, the cost effect outweighs the informational effect for simple products while the opposite is
true for complex products. See Moenius (2004).

Figure 9.1 Harmonisation and EU import value (1994–2004), all partner
countries
3612
3613
3621
3624
3625
3629
3631 3632
3633
3634
3635
3639
3641
3643
3644
36473648
3651
3652
3661
3663
3669
3671
3672
3674
3675
3676
3677
3678
3679
36913692
3695
3699
-1
-.5
0
.5
dlog(NH)
-1 0 1 2
dlog (imports)
In this chapter, I explore the impact of harmonisation across different types
of EU partner countries. Breaking up the total import value into imports from
developed countries (high-income countries) and imports from developing
countries (middle-income countries), it is easy to see that harmonisation is
more strongly correlated with increments in total import value in the former
group (Figure 9.2) than in the latter group (Figure 9.3). Below I show that once
I control for the usual gravity variables, the correlation for developing countries
vanishes while the correlation for developed countries survives.
To formally examine these empirical trends, I use a standard gravity model
of international trade. Specifically, I regress EU-15 import value in industry i
from country k to country j in year t (xi
jkt) on distances (Djk) and my measures of
trade costs (ti
jt and NHi
t). I control for cultural relationships such as colonial links
(Dcolonyjk) and common official language (Dcomlangjk) between country-pairs
as well. An industry is a SIC4 code and importing countries are the original EU-
15 members. I include fixed effects in the year (gt) , exporter (gk), importer (gj),
and industry (gi) dimensions. Robust standards errors are adjusted for clustering
at the importer level.11
ln(xi
jkt
)=b0
+b1
ln(Djk
)+b2
ln(ti
jt
)+b3
ln(NH
i
t)+b4
Dcolonyjk+b5
Dcomlangjk+b5
DnewEUjk+gt+gk+gj+gi+ei
jkt
11 The regression is at the exporter-importer-industry-year level in order to be able to take account
of sector-year variation in harmonisation, as well as country-level variation in common trade costs
proxies such as distance, common language and colonial history.

Figure 9.2 Harmonisation and EU import value (1994–2004), developed partner
countries
3612
3613
3621
3624
3625
3629
3631 3632
3633
3634
3635
3639
3641
3643
3644
36473648
3651
3652
3661
3663
3669
3671
3672
3674
3675
3676
3677
3678
3679
36913692
3695
3699
-1
-.5
0
.5
dlog(NH)
-.5 0 .5 1 1.5 2
dlog(Imports)
Figure 9.3 Harmonisation and EU import value (1994–2004), developing partner
countries
3612
3613
3621
3624
3625
3629
3631 3632
3633
3634
3635
3639
3641
3643
3644
3647 3648
3651
3652
3661
3663
3669
3671
3672
3674
3675
3676
3677
3678
3679
36913692
3695
3699
-1
-.5
0
.5
dlog(NH)
0 1 2 3 4
dlog(Imports)

Iestimatethisequationbycontrollingforthenon-randomselectionofobservations
with positive export values using the Heckman two-stage procedure. The two
stages are separately identified by the functional form and by the instrumental
variable from the first-stage regression. An appropriate instrument is a variable
that is correlated with the probability of export but largely uncorrelated with the
export volume. I follow Portugal-Perez et al (2010) and use the lagged decision
to export as an instrument (represented by a dummy variable that is equal to one
if there were country k’s exports in industry i to country j in the preceding year).
Table 9.4 reports the results of estimating this equation from 1994 to 2004. The
first column presents the estimation using the full set of exporting countries.12
As suggested by previous literature, I confirm the important role for European
product standards harmonisation on both the export value and the propensity to
export: the negative coefficients on .and indicate that falling trade costs are
followed by an increase in the propensity to export – the extensive margin of
trade – as well as in the export value within industry-exporter-importer-year
bins. Interestingly, product standard harmonisation seems to be more important
than tariffs at the extensive margin. Columns 2 and 3 present the results of
estimating the same equation for the set of exporters from developed countries
and for the set of exporters from developing countries, respectively. Results
indicate that European harmonisation of standards in electronics is associated
with higher import value from developed countries and with lower import value
from developing countries. Meanwhile a 1% of reduction in the non-harmonised
share is associated with a 0.12% increase in export value within bins for the
average exporter from the developed world; this reduction is associated with a
0.07% decrease in export value within bins for the average exporter from the
developing world, ceteris paribus.
In order to present specific country evidence columns 4–5 report the results for
the two most important exporters from the developed world, meanwhile columns
6–10 report results for the most important exporters from the developing world.13
Results confirm the previous findings. On the one hand, the US and Japan benefit
from harmonisation both at the export value within bins and at the propensity to
export. On the other hand, there is a robust result that exports from developing
countries did not benefit from harmonisation. Results at the extensive margin
vary across developing countries; while the probability of exporting from China
has increased in response to harmonisation, it has decreased for the case of
Tunisia and has no significant impact for Turkey, Malaysia and Philippines.14
12 Since exports of electronics from low income countries is almost zero (see Table 9.2), I drop these
countries from the estimation.
13 Exporter countries that later became EU members were not considered for the country regressions.
14 The ability of firms to enter into the EU market in response to an international harmonisation of
product standards depends, in part, on the degree of market diversification of firms in countries that
use international norms (Reyes, 2011). Arguably, Chinese exporting firms are diversified enough into
these types of markets so they are better endowed to enter into the EU market as a consequence of
this harmonisation process

Table9.4 ImportvaluetotheEU-15,SIC4-exporter-importer-year,1994–2004
RegressorAll
Developed
exporters
Developing
exporters
UnitedStatesJapanTurkeyMalaysiaChinaTunisiaPhilippines
HeckamnOutcome
Ln(distance)-0.814-0.871-0.193
[0.019]***[0.022]***[0.040]***
Ln(tariffrate)-0.056-0.067-0.045-0.1270.022-0.152-0.054-0.146-0.0810.016
[0.004]***[0.006]***[0.006]***[0.042]***[0.073][0.068]**[0.025]**[0.031]***[0.051][0.027]
Ln(NHShare)-0.035-0.1230.066-0.487-0.3650.0860.3170.0440.3490.562
[0.008]***[0.011]***[0.011]***[0.061]***[0.107]***[0.059][0.072]***[0.088][0.080]***[0.078]***
Colonydummy0.2600.2600.274
[0.027]***[0.038]***[0.040]***
Commonlanguagedummy0.2120.449-0.085
[0.028]***[0.038]***[0.045]*
Lambda-1.282-1.341-1.382-2.972-5.156-2.686-3.338-4.366-1.701-2.862
[0.016]***[0.020]***[0.025]***[1.623]**[0.861]***[0.149]***[0.165]***[0.427]***[0.128]***[0.137]***
HeckamnSelection
Ln(distance)-0.393-0.366-0.41
[0.009]***[0.013]***[0.014]***
Ln(tariffrate)-0.031-0.037-0.028-0.584-0.008-0.168-0.073-0.042-0.091-0.065
[0.002]***[0.003]***[0.002]***[0.074]***[0.054][0.036]***[0.013]***[0.024]*[0.021]***[0.011]***
Ln(NHshare)-0.086-0.089-0.085-0.939-0.433-0.025-0.047-0.1770.067-0.008
[0.003]***[0.006]***[0.004]***[0.093]***[0.128]***[0.034][0.035][0.063]***[0.032]**[0.030]
Colonydummy0.2800.5200.154
[0.013]***[0.023]***[0.016]***
Commonlanguagedummy0.2420.1920.286
[0.013]***[0.022]***[0.017]***
Lagexportdecision1.4571.4811.4373.2782.0591.5791.8232.1141.6471.73
[0.006]***[0.009]***[0.007]***[0.880]***[0.168]***[0.052]***[0.053]***[0.100]***[0.049]***[0.045]***
YearfixedeffectsYesYesYesYesYesYesYesYesYesYes
ExporterfixedeffectsYesYesYes--------------
ImporterfixedeffectsYesYesYesYesYesYesYesYesYesYes
IndustryfixedeffectsYesYesYesYesYesYesYesYesYesYes
Observations7001282236524764764862486248624862486248624862
Note:Regressionsareattheexporter-importer-industry-yearlevel.Robuststandardserrorsadjustedforclusteringattheimporterlevelareinparentheses.Industryfixed
effectsareforthree-digitSICs.***significantatthe1%level;**significantatthe5%level;*significantatthe10%level.Coefficientsfortheregressionsconstantand
dummyvariablesaresuppressed.

Henceforth, I propose an explanation for the puzzle of export flows across type of
EU trading partners. Using the US linked/Longitudinal Firm Trade Transaction
Database (LFTTD), which links individual US trade transactions to US firms
in the Longitudinal Business Database (LBD),15
in conjunction with firm level
information from the Censuses of Manufactures (CM) of the Longitudinal
Research Database (LRD) of the US Census Bureau, I find that EU international
harmonisation of standards triggers US firm entry into the EU – the extensive
margin of trade – which makes competition tougher and reduces market shares
of market participants – the intensive margin of trade. Since this negative impact
applies to all exporting firms, the key feature to determine the overall impact
in export values is the availability of new firms to start exporting to the EU –
the extensive margin of trade. Results presented in Table 9.4 suggest that the
extensive margin effect dominates the intensive margin effect in developed
countries while the reverse is true for developing countries.
9.4 The pro-competitive effect of harmonisation of prouct standards
In this section, I explore the firm-level relationship between changing trade costs,
export growth and firm entry decisions using US trade linked firm-level data.
Overall, the empirical findings suggest that EU product standards harmonisation
contributes significantly to explain the export entry patterns observed among US
firms. I also find that harmonisation reduces market shares of established US
exporters to the EU. These findings are what I called the pro-competitive effect
of harmonisation of product standards: a reduction of a NTB to international
trade – harmonisation of product standards – triggers firm entry from the most
productive set of non-exporters firms to the EU, which, in turn, increases
competition and reduces market shares from established exporters.
I begin by estimating the impact of falling trade costs on the probability that
non-exporting US manufacturing firms to the EU become exporters to the EU
– the extensive margin of trade – via a logistic regression on my measure of
changing trade costs, firm productivity and other firm characteristics. I use firms’
information from the 1992, 1997, and 2002 US Censuses of Manufactures. I
define the change in trade costs for Census year t as the log difference in tariffs
over the preceding five years (Dti
t+5) and as the log difference on my measure of
product standard heterogeneity over the preceding five years (DNHi
t+5). These
regressions are given by:
(spec1)Pr(Et+5 = 1) = f(b1Dti
t+5 + b2DNHi
t+5 + gt + gi)
t+5 + b2DNHi
t+5 + b3PRt + gt + gi)
t+5 + b2DNHi
t+5 + b3PRt + b4Zt + gt + gi)
15 See Bernard et al (2009) for a complete description of the LFTTD and its construction. For an
extensive discussion of the LBD see Jarmin and Miranda (2002).

where Et+5 is a dummy variable equal to one if a firm does not export to the
EU in year t+5; PRt and becomes an exporter to the EU in year ; is the firm’s
revenue based labour productivity, and is a set of additional firm characteristics.
Additional firm controls include size, capital intensity, wage level, and multi-
plant dummies. I also include industry gi and time gt fixed effects and cluster the
standard errors at the industry level.
Results are reported across three columns in Table 9.5, with the first column
focusing on my trade costs measures and subsequent columns including
additional firm characteristics. Across all specifications, I find a positive and
statistically significant association between product standards harmonisation and
the probability that a non-exporting firm to the EU becomes an exporter to the
EU across Census years. The probability of becoming an EU exporter is higher
in industries with greater harmonisation of product standards. Surprisingly,
EU tariff changes do not affect the probability of becoming an exporter.
In specification two and three, I find, as expected, a positive and significant
association between firms’ productivity and their entry into exporting. Larger
and more capital-intensive firms are more likely to become exporters, as are
multi-plant firms and firms that pay higher wages.
Table 9.5 Probability of entering the EU market
Regressor Spec1 Spec2 Spec3
Change in tariff rate 0.122 0.127 0.199
[0.198] [0.209] [0.209]
Change in NH share -0.155 -0.14 -0.145
[0.053]*** [0.057]** [0.059]**
Ln(labour productivity) 0.469 0.256
[0.112]*** [0.103]**
Ln(employment) 0.255
[0.106]**
Ln(K/L) 0.083
[0.038]**
Ln(wage) 0.26
[0.125]**
Multiple-plant firm 0.409
[0.080]***
Year fixed effects yes yes yes
Industry fixed effects yes yes yes
Observations 4294 4294 4294
Log likelihood -1993.16 -1973.75 -1869.62
Note: Firm-level logistic regression results. Robust standard errors adjusted for clustering at the four-
digits SIC level are in parentheses. Industry fixed effects are for three-digit SICs. Regression cover two
panels: 1992 to 1997 and 1997 to 2002. *** significant at the 1% level; ** significant at the 5% level; *
significant at the 10% level. Coefficients for the regressions constant and dummy variables are suppressed.
To look at the response of established exporters to the changes in my trade costs –
the intensive margin of trade, I estimate an OLS regression of the log differences

in exports to the EU across Census years, ln(Exportst+5) - ln(Exportst), on firm
characteristics,
(spec1)Dt:t+5ln(Exports) = b1Dti
t+5 + b2DNHi
t+5 + gt + gi + ei
t)
(spec1)Dt:t+5ln(Exports) = b1Dti
t+5 + b2DNHi
t+5 + b3Zt + gt + gi + ei
t)
where the variables are defined as above. An observation in this regression is
a firm that exports to the EU in two consecutive Census years; this explains
the relatively small number of observations in the regression. As above, my
regressions include year and industry fixed effects and standard errors are
clustered at the industry level.
ResultsarereportedinTable9.6.Bothcolumnsreportapositiveandstatistically
significant relationship between changes in the non-harmonised share of
standards and changes in exports from established exporters: firms in industries
with a relatively greater degree of harmonisation experience a larger decrease in
exports across Census years. Tariffs, on the other hand, are negatively associated
with export growth. Additional results indicate no statistically significant
relationship between export growth and firm productivity, employment, wages,
and status as part of a multiple-plant firm.
Table 9.6 Change in log exports from established exporters
Regressor Spec1 Spec2
Change in tariff rate -0.063 -0.052
[0.000]*** [0.012]***
Change in NH share 0.022 0.015
[0.000]*** [0.015]*
Ln (labour productivity) 0.03
[0.065]
Ln(employment) -0.014
[0.036]
Ln(K/L) -0.023
[0.040]
Ln(wage) 0.038
[0.039]
Multiple-plant firm -0.077
[0.087]
Year fixed effects yes yes
Industry fixed effects yes yes
Observations 2279 2279
R2 0.012 0.013
Note: Firm-level OLS regression results. Robust standar errors adjusted at the four-digit SIC level are in
parentheses. Industry fixed effects are for three-digit SICs. Regressions cover two panels: 1992 to 1997
and 1997 to 2002. *** significant at the 1% level; ** significant at the 5% level; * significant at the 10%
level. Coefficients for the regressions constant and dummy variables are suppressed.

9.5 Conclusions
With the decline of traditional trade barriers, differences in product standards
across market destinations are increasingly viewed as impediments for trade.
Yet despite the growing evidence of the negative impact of product standard
heterogeneity, there is little understanding of the gains of international
harmonisation on market dynamics. This chapter documents the asymmetric
impact of harmonisation across exporting countries according to their level of
development. Based on an analysis on the response of US manufacturing firms to
this harmonisation process, I propose an explanation for the observed divergence
across partner countries based on the pro-competitive effect that international
harmonisation entails in the harmonised market.
The European harmonisation of product regulation to international norms in
the electronic sector, which arose at the beginning of the 1990s, constitutes a way
to boost imports (Portugal-Perez et al, 2010). In this chapter, I provide evidence
about the heterogeneity of the response of export flows across partner countries
to this policy: While harmonisation tends to increase exports from the developed
world, it has an ambivalent impact on exports from the developing world.
To provide an explanation for this puzzle, I use highly detailed US trade linked
firm data to examine the response of US manufacturing firms to the international
harmonisation process in the EU. I find that harmonisation increases US export
value to the EU and that this increase is due to more US firms entering the EU
market – the extensive margin of trade. I also find that harmonisation decreases
market shares of existing exporters – the intensive margin of trade. These two
impacts on the harmonised market are what I called the pro-competitive effect
of harmonisation of product standards. Since the intensive margin effect impacts
profits of all market participants negatively, the key feature that determines the
overall impact in export value, at the country level, is the availability of new firms
to start exporting to the EU – the extensive margin of trade. Results indicate that,
on average, the extensive margin effect dominates the intensive margin effect in
developed countries while the reverse is true for developing countries.
This result is important because it puts the changes in market dynamics as a
cornerstone to explain the heterogeneity in the responses of exporting countries
to international harmonisation of standards. Before going into the policy
implications of my analysis, it is useful to discuss some of its limitations. First,
the results are specific to the electronics sector which is characterised mostly by
north-north trade. Second, due to data availability, I do not consider an arguably
important venue that may also affect market dynamics: the role of quality-
upgrading in response to harmonisation of standards. An option for market
participants is to respond to the pro-competitive effect by increasing the quality
of their products. The difference responses between developed and developing
countries along the quality margin are worth investigating in future research.
This venue looks quite promising in the electronics sector in particular as these
products are ones for which differentiation and variety are very important.

The policy implications of these results are of significant interest. Besides
advocating for deepening the effort on international harmonisation in electronics
by, for example, building on the Information Technology Agreement to commit
on harmonisation of standards at the multilateral level – to the extent this is
feasible – this chapter unveils the important need of exporters from the developing
world to find ways to stay competitive in a world where competition is thriving
due to reduction in barriers to international trade.
References
Baldwin, R E (2000), ‘Regulatory Protectionism, Developing Nations, and a Two-
Tier World Trade System’, in S Collins and D Rodrik (eds), Brookings Trade
Forum 2000, Washington DC, Brookings Institution Press.
Bernard, A B, Jensen, J B and Schott, P K (2009), ‘Importers, Exporters and
Multinationals: A Portrait of Firms in the U.S. that Trade Goods’, in T Dunne, J
B Jensen and M J Roberts (eds), Producer Dynamics: New Evidence from Micro
Data, Chicago, University of Chicago Press.
Chen, M X and Mattoo, A (2008), ‘Regionalism in Standards: Good or Bad for
Trade?’, Canadian Journal of Economics, 41(3), pp 838–63.
Czubala, W, Shepherd, B and Wilson, J (2009), ‘Help or Hidrance? The Impact
of Harmonized Standards on African Exports’, Journal of African Economies,
doi:10.1093/jae/ejp003.
Hanson, D (2005), CE Marking, Product Standards andWorldTrade, Northampton,
USA, Edward Elgar.
Jarmin, R S and Miranda, J (2002). ‘The Longitudinal Business Database’, Working
Paper, US Census Bureau, Center for Economic Studies.
Moenius, J (2004), ‘Information Versus Product Adaptation: The Role of Standards
in Trade’, International Business and Markets Center Working Paper, Kellogg
School of Management, Northwestern University, Evanston, IL.
Moraga-Gonzalez, L and Viaene, J-M (2003), ‘An Example of Procompetitive Trade
Policies’, Working Paper 504, Research Seminar in International Economics,
University of Michigan.
Otsuki, T, Wilson, J S and Sewadeh, M (2001), ‘What Price Precaution? European
Harmonisation of Aflatoxin Regulations and African Groundnut Exports’,
European Review of Agriculture Economics 28(3), pp 263–84.
Portugal-Perez, A, Reyes, J-D and Wilson, J S (2010), ‘Beyond the Information
Technology Agreement: Harmonization of Standards and Trade in Electronics’,
The World Economy, 33(12), pp. 1870–97.
Reyes, J-D (2011), ‘Product Standards Harmonization and Firm Heterogeneity in
International Trade’, mimeo, Georgetown University.
Shepherd, B (2006), ‘The EU Standards Database: Overview and User Guide’,
mimeo, available at http://go.worldbank.org/6OEYNCYSD0
Shepherd, B (2007), ‘Product Standards, Harmonization, and Trade: Evidence from
the Extensive Margin’, Policy Research Working Paper No. 4390, The World
Bank.

Office of the United States Trade Representative, (2010), ‘Technical Barriers to
Trade (TBT Report)’, available at http://www.ustr.gov/about-us/press-office/
reports-and-publications/2010-0

10
Standards Harmonisation and Trade: The
Case of the East African Dairy Industry
Michael F Jensen and John C Keyser1
10.1 Introduction
Two trends in trade policy intersect in this chapter. First, as traditional trade
barriers such as tariffs and quotas are less used than before, trade policy analysts
and practitioners have turned their attention to non-tariff barriers (NTBs). The
expectation is that as traditional barriers come down, protectionist interests
will find new outlets and NTBs are particularly well suited to the needs of
protectionism. Therefore, the understanding of NTBs is a prominent field in
trade research today. Second, standards are increasingly applied in international
trade today. The use of standards have proliferated over the last two to three
decades and now serve a multiplicity of policy objectives including trade
facilitation purposes and the protection of life and health of humans, animals
and the environment. In a trade policy context, standards are both indispensable
for trade facilitation purposes and a source of NTBs and as such a growing issue
of concern.
Policymakers often address the risk of using standards as NTBs by harmonising
domestic and regional standards with their international equivalents. The idea is
that if standards are made the same across borders they cannot present trade
obstacles any more. The use of international standards have grown rapidly since
the signing of two international trade agreements on standards and technical
requirements, namely the WTO Agreement on the Application of Sanitary and
Phytosanitary Measures (the SPS Agreement) and the WTO Agreement on
Technical Barriers to Trade (the TBT agreement). Both the SPS and the TBT
agreements encourage member countries to apply international standards but
stop short of making the use of them mandatory. Since the entry into force
of the agreements in 1994, many donors and international organisations have
1 This paper is based on a longer report prepared by the authors for the World Bank (Jensen and Keyser,
2010).

supported interventions in developing countries to apply international standards
as the basis for domestic regulation.
This chapter argues that international harmonisation may erect rather than
remove NTBs in developing country trade. The chapter is an empirical study of
NTBs in the East African dairy sector. The East African Community (EAC) has
focused on the removal of NTBs in an attempt to avoid a policy reversal after the
Partner States courageously removed tariffs on intra-regional trade. In the dairy
sector, the harmonisation of EAC standards with international ones has been
financed by a donor agency and is widely supported by government agencies
in the five EAC Member States with the purpose of easing regional trade. The
chapter identifies and discusses present and future NTBs in East African dairy
trade. Few NTBs currently constrain intra-regional trade, but the newly adopted
harmonised standards may constitute a significant barrier in the future. NTBs
may arise because very little milk today abides by the new harmonised standards
which have not been implemented. In effect, the new harmonised standards set
the bar for dairy quality so high that few producers will be able to comply even
if granted a long time period (say 10 years) to adjust. The implementation of the
standards would make it possible to stop a very large share of dairy trade at the
border without benefiting social objectives such as food safety.
The problem is that the new harmonised standards are largely copied from
international standards designed for a context widely different from the East
African one. The dairy industry in East Africa is based on long atomistic supply
chains dominated by small farmers and catering largely to poor consumers who
consume milk after boiling it. Dairy products following international standards
would be prohibitively expensive for most East Africans and would not be
significantly safer from a food safety perspective because milk is boiled before
consumption anyway. The case study of the East African dairy industry is an
example of the import of inappropriate regulation. It is clear that globalisation
and increased attention to food safety and similar social objectives regulated by
standards require more and better regulation in developing countries. However,
the appropriateness of the regulation is crucial to making the regulation meet
the social objectives in an economically efficient way. The focus on economic
efficiency is essential in poor countries. The study of the East African dairy
sector discusses the role played by the reliance on foreign actors like donor
agencies and international standardisation bodies for the appropriateness of
regulation. Inappropriate regulation may lead to the waste of resources as well
as the erection of NTBs at the border.
This paper gives an overview of the size and organisation of the EAC dairy
industry as well as the role currently played by trade. International dairy trade is
currently marginal in the EAC with most traded milk sold in local markets. The
paper subsequently identifies and discusses potential NTBs that might constrain
trade today. It then discusses the potential for future NTBs through an analysis
of the newly harmonised EAC dairy standards. Finally, conclusions are offered
in the final section.

Standards Harmonisation and Trade: East African Dairy Industry 189
10.2 Background
The present EAC was established in 2000 by Kenya, Uganda and Tanzania.
Rwanda and Burundi joined in 2007. The EAC has grown into one of the most
dynamic regional trade agreements on theAfrican continent.As of January 2010,
the EAC removed all tariffs on intra-EAC trade.This marks a strong drive towards
a more efficient, market-driven economy. The EAC is committed to this drive and
well aware that enjoying the benefits of trade requires long-term commitment.
The welcoming attitude to free trade, however, can only be sustained in a
context of meaningful market access. Both scholars and practitioners fear that
the worldwide tendency to falling tariffs may be counteracted by policy reversals
whereby less transparent trade barriers take the place of falling tariffs. The EAC
has met this fear by including binding commitments for the Partner States to
work towards the elimination of NTBs between them. The Partner States are to
design a mechanism that identifies and monitors NTBs and commit themselves
to eliminating existing barriers on intra-EAC trade.
NTBs are an incredibly difficult analytical as well as policy area. Beyond
traditional barriers such as tariffs, there are no databases that may inform
policymakers about the gravity of a NTB problem and where it is located. The
diversity of potential barriers is extreme. This uncertainty makes NTBs the
perfect place to look for protectionist interests wanting an edge in the domestic
market. This report uses a value chain approach. Focus is on the dairy industry
as a representative for perishable agricultural commodities. The report is based
on desk research and a three-week fieldwork period in Rwanda, Uganda and
Kenya during December 2009. During fieldwork illustrative data was collected
primarily through stakeholder interviews. Emphasis was put on interviews with
private sector participants and officials working on an everyday basis with
practical trade issues.
Protectionist interests are not alone in creating trade barriers. The many new
regulatory areas that EAC authorities will have to deal with as international
integration continues may create trade barriers by mistake rather than design.
Food safety regulation, for instance, is a new topic that may be badly designed or
poorly implemented due to capacity problems rather than bad intentions.
Dairy production and marketing are a significant part of the agricultural
economy in East Africa. According to FAO data, EAC countries produced
around six billion litres of fresh cow milk in 2007, roughly equal to one-quarter
of the total for all of Africa. Kenya and Uganda are the largest dairy producing
nations in East Africa and account for 58% and 25% of the EAC’s total milk
supply respectively. Less than half of the total EAC production is marketed.
Yet, market oriented dairying is a major economic activity in East Africa noted
for extensive smallholder farmer involvement. In diverse settings throughout the
region, dairy has been shown to provide small farmers a regular cash income that
can be several times greater than many other types of on- and off-farm enterprise
(Staal et al, 2003; Keyser, 2003; IFAD, 2005; Parsons and Matiru, 2008). Other
recognised benefits of dairy production include growth linkages to input service

providers, milk traders and dairy processors; the ability to supply nutritious
and affordable food to the local population; and opportunities for long-term
expansion into growing domestic and regional export markets (Bennett et al,
2006).
All EAC countries have large and vibrant small-scale trading sectors in which
various kinds of milk vendors, milk shop owners, and mini-processors link
small farmers with city and town buyers. These markets generally have lower
quality requirements than formal processors and account for 80–85% of total
milk marketing in Kenya and as much as 95% of milk marketing in Uganda,
Tanzania, Rwanda and Burundi.
While seen as an industry with future trade growth potential, intra-EAC dairy
trade is currently very limited. Dairy exports presently account for less than 1% of
total marketed production in the EAC with almost 80% of these exports destined
for other EAC and COMESA Member Countries. In Kenya, for example, only
14% of total marketed milk is handled by large commercial dairy processors and
the nation’s largest processor says that exports account for no more than 10%
of its total production with most sales going to Uganda and Tanzania. Similarly,
Kenya’s second largest dairy processor says it has stopped exports for several
months from in late 2009 because of an ongoing drought and commitment to
serve domestic markets first. Of course, some dairy firms in the EAC are geared
much more toward export production, but the overall pattern whereby most milk
sales are informal and most formal sales target domestic consumers must be kept
in mind when discussing trade policy.
Currently, production capacity is largely determined by rainfall patterns which
are about the same across the EAC countries. Given the fact that consumption
patterns are the same and that each country produced the same variety of dairy
products it is natural that intra-EAC trade is low. However, continuing growth in
dairy production will likely lead to higher trade which is likely to cause conflicts
between the EAC countries. The purpose of trade is to achieve gains according
to each country’s comparative advantage rather than to maximise on foreign
exchange earnings or similar mercantilist objectives. During fieldwork in the
EAC, many stakeholders doubted the benefits of import. For some, exports were
a measure of success and imports were a sign of failure. A mercantilist view of
trade is still fashionable within many agencies. The highly sceptical views on
imports were accompanied by often overly optimistic views on exports, such
as a focus on future EU bound exports. Scholars with the EAC still have much
work to do convincing policymakers that imports play a vital economic role and
that national self-sufficiency is not a viable strategy for small economies. Policy
work must be focused on future developments and the need to keep markets open
to allow for a more dynamic trade-oriented dairy sector to emerge
10.3 Do non-tariff barriers explain low trade?
The low level of intra-EAC trade is often perceived as a problem in trade
policy discussions (World Bank, 2008). One often-offered explanation is the

proliferation of NTBs in the region. EAC countries have a long history of trade
disputes flaring up in dairy. One example is presented in Box 10.1. As tariffs
have been officially removed by January 2010, the use of NTBs would be a
logical choice for protectionist interests in the future.
Looking at the current situation, it is also apparent that formal sector traders
face many legal requirements that take time to complete and add to the total cost of
regional trade. Before an export deal can be concluded, for example, participants
on both sides of the border must obtain a permit from the national dairy board or
other relevant dairy authority to authorise the transaction. Veterinary certificates,
certificates of origin, national quality seals, and quality test results for each batch
are also required. Each of these requirements has the potential to be used as a
non-tariff barrier if ever there is a perceived benefit from preventing exports or
keeping a neighbouring country’s dairy products out.
Box 10.1 Formal trade requirements as a competitiveness risk
A good example of how quality standards and laboratory testing requirements can
militate against formal sector trade is the story of Uganda’s exports of butter to Rwanda.
In this case, Rwanda’s only licensed trader bringing butter into the country decided to
stop imports because of difficulties with recognition of Ugandan quality certificates
and demands for additional laboratory tests by the Rwanda Bureau of Standards (RBS).
According to the importer, this was because the batch numbers were not identical
throughout each consignment since the butter was manufactured on different days. This
problem also meant the trader could not obtain pre-clearance for the goods. As a result,
the importer decided it would be easier to stop trading in butter and cancelled all orders
with the Ugandan supplier.
Despite this move, Ugandan butter remained on store shelves in Kigali. As explained
by border officials, it is impractical to stop small consignments from crossing and many
dairy products are brought into Rwanda in small quantities, sometimes in cool boxes
but otherwise with no refrigeration or other kinds of quality control. Similar to how
formal sector dairy chains have a difficult time competing with informal milk vendors in
domestic markets this story shows that formal traders also have a difficult time competing
with informal traders in the regional market.
Although the efforts to regulate dairy trade and harmonise regional standards may seem
like an obvious step towards an improved trade regime, such moves can actually have
negative consequences for formal sector operators. While it is important to have a well-
regulated trade system, the system must be cost competitive and simple to use in order to
avoid creating further competitiveness constraints for formal dairy operators.

10.3.1 Recent trade disputes
As noted, the EAC region has a long history of dairy trade disputes. As recently
as the first quarter of 2009, Uganda’s largest milk processor alleged that Kenya
had introduced several measures that it believed were aimed specifically at
protecting Kenyan dairy plants. Specifically, the Ugandan processor complained
the Kenya Bureau of Standards (KEBS) was insisting on new conformity
assessments instructions for imported milk powder that required 34% protein
in full cream powder when the highest level that can be achieved from cow’s
milk is 25–26%. The Ugandan processor also complained about Kenya banning
foreign registered four-axel trucks and said that the Kenyan Revenue Authority
(KRA) was holding Ugandan long-life milk at the border for more than two
weeks, even when the consignments had all of the required documents. As a
result, transporters were refusing to carry Ugandan products and exports more
or less ground to a halt. Other dairy processors in Uganda and Tanzania made
similar complaints at the time and alleged that Kenyan authorities were using
whatever excuse they could find to protect local producers and keep foreign
dairy products out of the country.2
Equally, Kenyan exporters have complained that Uganda requires additional
laboratory analysis of all Kenyan dairy products whereby samples of each
shipment must be sent in advance to receive the required permits. According
to EAC agreements, national quality seals are meant to be sufficient proof of
conformity with EAC standards, but additional testing by Uganda (and other
EAC countries) has still been required leading to extra cost and delay. Moreover,
Kenyan processors say that Uganda maintains higher-level national dairy
standards with regard to adulteration and other quality variables than apply in
the domestic market, meaning they must produce special ‘export quality’batches
to trade with their regional neighbour.
During a previous dispute in 2002–2003, dairy trade between Kenya and
Uganda effectively came to a complete stop. According to dairy insiders in
Nairobi, Kenya first refused to grant import licences for dairy products from
Uganda and Uganda followed suit by refusing to grant import licences for
products from Kenya. To justify these moves, various objections were raised
on both sides of the border over the quality standards of the other country’s
product, labelling and packing requirements, and reliability of national quality
seals. A similar non-tariff dispute arose in 2004 between Kenya and Zambia
when a shipment of UHT milk from Kenya was rejected on quality grounds and
alleged non-compliance with domestic standards in Zambia.
2 For articles in the press on recent trade disputes, see Mwamunyange (2008), Omolo (2008) and
Michael (2009).

10.3.2 Requirements for trade
The trade regime in the EAC is complex despite the recent removal of tariffs.
This section presents a diagnostic approach by tracing the specific requirement
for EAC dairy trade throughout the value chain. By comparing the formal
requirements with current practice, this analysis aims to identify potential
bottlenecks and threats to regional trade.
Figure 10.1 provides a broad overview of current requirements for dairy
trade in the EAC region. As shown, successful international trade begins with
a competitive production and marketing system in the exporting country and
is dependent on market demand in the importing country. Even before any
potential non-tariff barriers come into play, therefore, there are many challenges
to successful dairy trade at the level of primary production and post-harvest
marketing.
As shown in Figure 10.1, various kinds of licences, permits and certificates are
required for dairy trade within the EAC region that include:
• Dairy export permit
• Dairy import permit
• Veterinary certificates
• Conformity analysis
• Certificate of origin
• Packing list and invoice
• Import declaration form
• Clearance certificate.
From a value chain perspective, trade competitiveness can be measured by
comparing the build-up of total costs with the amount customers are willing
to pay for the final product. In an open economy, this price is determined
competitively and value flows upstream from the consumer to each producer
and marketing company in the chain. All costs and profit margins taken by sector
participants before the product’s value reaches the farm level, including the costs
of complying with quality standards and trade reqiurements, therefore have a
direct bearing on the amount that can be paid to farmers, and thus rural incomes
and potential for poverty reduction.

Figure 10.1 Flow diagram of EAC trade requirements for dairy
ExporƟng country TransacƟon requirements ImporƟng country
Transport and distribuƟon Sale
Border procedures (by clearing agent)
ImporƟng country
Entry registraƟon (IDF and all
other documents)
Payment of taxes
VerificaƟon of contents (at
border or delivery point)
Quality inspecƟon
Transit rouƟng?
If yes, require third
country customs bond
ExporƟng country
Market
surveillance
Packing list
and invoice
CerƟficate of
origin
Conformity
analysis
Dairy export
permit
Licensed
trader/importer
Market
Demand?
IDF
Dairy import
permitDocuments from
buyer (importer)
Order from foreign
buyer
Transport to border
Documents from seller
(exporter)
Veterinary
cerƟficate
InternaƟonal markeƟng
CompeƟƟve processing
Good milk supply
Clearance
cerƟficate
Veterinary
cerƟficate
(by trader and final vendor)
Customs release
and seal.
(by trader/importer)
Customs entry doc
Dairy export permit. On the export side, dairy permits or ‘dairy board certificates’
as they are sometimes called, are issued by the relevant dairy authority to
authorise the shipment abroad. Very often, dairy certificates are issued together
with veterinary certificates and/or certificates of quality analysis as a kind of
umbrella document to show that the exporter and specific product being traded
meets all required conditions for trade. An important risk with dairy export
permits is that these can be used to limit trade if the exporting country’s dairy
authority ever sees a benefit in doing so. During its review of each export
application, for example, the website of the Kenya Dairy Board (KDB) says,
‘the Board will take into cognizance the prevailing dairy production situation
in the country’. The KDB maintains similar controls over dairy imports and
only grants import licences after it considers the prevailing production situation
in Kenya. In practice, the KDB was said to be extremely helpful in facilitating

exports by each of the Kenyan firms met. Nevertheless, these powers and similar
powers in other countries are a significant threat to the ideal free trade situation
based purely on comparative advantage.
Dairy import permit. Just as dairy processors are required to obtain an export
permit, so too are dairy importers required to obtain a permit from their local
dairy authority. In Uganda, the Dairy Development Authority (DDA) issues
import permits quarterly based on a declaration of how much and what kind
of dairy products the importer intends to bring in. Importers are not allowed
to go beyond the declared amount and must comply with all other quality
and laboratory testing requirements described below. In Rwanda, importers
must apply for a permit through the Rwanda Animal Resources Development
Authority (RARDA). Specifically, on collecting the application forms from
RARDA, the trader must take samples of the product to the Rwanda Bureau
of Standards (RBS) to obtain a ‘certificate of analysis’. The analysis was said
to take about two business days. When the RBS certificate has been issued, the
trader is able to return to RARDA and obtain the import permit. For large and
medium-scale traders that are known to RARDA, the import permit is issued
for 12 months. New traders, however, are required to apply for a fresh permit
for every shipment until they establish their own record of competence at which
point they may be approved for a long-term permit.
Veterinary certificate. Veterinary certificates are required from both the
importing and exporting country and must be obtained for each individual
shipment. On the export side, the intent is to show that the dairy product is made
of milk from healthy, disease free animals. Similarly, on the import side, the
certificate is to acknowledge that the local veterinary authorities do not have any
concerns for the animal health status in the exporting country.
As with other requirements, the traders met for this study did not raise any
major concerns for obtaining the veterinary certificate for dairy, except to say
that it demands a considerable amount of time and effort to collect. In Kenya,
for example, veterinary certificates are issued separately for every shipment as
part of the application for an import or export permit and takes three business
days to process on top of the time required for the KDB to review the application
and make its own recommendation. Since milk processors normally obtain
their supplies from the same or similar sources throughout the year, one way
to streamline the trade process would be to issue the veterinary certificates on
an annual or even quarterly basis rather than for each individual shipment. If a
particular disease breaks out, then the blanket no-objection letter could easily be
suspended.
Conformity analysis. Beyond the trade licences and certificates issued by the
local dairy authority and veterinary departments, each country’s national bureau
of standards also plays a role in sanctioning dairy trade. To begin with, dairy
manufacturers are required to register with their national bureau of standards
for the purpose of obtaining the national quality seal. According to formal EAC
agreements, each country’s quality mark should be accepted by all other EAC
members as sufficient proof that the dairy plant and dairy product conforms to

regional health and safety standards. A computer system has been developed that
allows border inspectors to check whether individual EAC products have been
awarded the quality mark or not.
In practice, however, the principle of mutual recognition of national quality
marks is not always applied to dairy and additional laboratory analysis is routinely
carried out as a condition for regional trade. In Uganda, for example, the DDA
has required dairy importers to send samples of every export batch to the Uganda
National Bureau of Standards (UNBS) to obtain a local quality certificate. Batch
numbers and manufacturing dates are verified at the border to ensure they match
the samples that were sent to UNBS for testing.3
Occasionally, samples are
drawn at the border even with the UNBS pre-approval documents. Exporters
say the testing normally takes two to three business days but can sometimes take
much longer depending on any backlogs at the UNBS facility.
In an apparent change of this policy, dairy importers reported that the Uganda
DDA issued a letter in December 2009 to say that it would no longer require
every foreign consignment to be analysed at UNBS. While the operational details
of the new policy were still being worked out at the time of data collection, the
essence of the letter was that dairy exporters would have the option of carrying
out the laboratory analysis in the country of origin or in Uganda. The new policy
facilitates trade but still appears to fall short of the agreed EAC principle whereby
each country’s national quality seal should be accepted as sufficient and routine
proof of conformity with regional standards.
Moreover, the problem of additional quality analysis is not unique to Uganda.
Exporters to Kenya say they must provide independent laboratory results for
each batch in addition to proof of the national quality mark. Likewise, dairy
traders in Rwanda said that the Rwanda Bureau of Standards (RBS) sometimes
insist on additional laboratory work regardless of the foreign country’s quality
mark or other foreign laboratory results, which are always required. Similarly,
in Tanzania, dairy imports must be registered with the Tanzania Food and Drug
Authority (TFDA) and Tanzania Dairy Authority (TDA) in addition to carrying
the exporting country’s quality seal.
Certificate of origin. Per the terms of the EAC Free Trade Agreement,
certificates of origin are used to verify the goods were produced within the EAC
region so that the importer can avoid paying import duty. Certificates of origin
are provided by the exporter as part of the trade deal. Although the procedure for
issuing certificates of origin varies from country to country, no concerns were
voiced for this process by any of the people met. In Uganda, exporters said the
process works extremely well and reported that all they must do upon registration
is to buy a book of blank certificates from the Uganda Export Promotion Council
and return the receipt stubs when the book is finished. In Kenya, certificates of
origin for dairy are issued through the KDB.
3 If samples were not sent to Uganda in advance (by airfreight or other means), UNBS inspectors will
draw samples at the border and take them to Kampala for analysis. The goods are allowed to proceed,
but only with a customs bond to guarantee they are not offloaded and sold before the laboratory
results are back.

Packing list and invoice. As with all internationally traded goods, exporters are
required to provide a detailed invoice and packing list for customs valuation.
Although dairy produce from within the EAC region is no longer subject to
import duty, this procedure still applies. The packing list must specify batch
numbers, packing dates, expiry dates, and other key information needed to
match the goods against the quality test results.
Import declaration form. The IDF is a simple customs document that the
importer must fill out and submit to the clearing agent to use during border
procedures. The IDF will contain information on the exporter, importer, type of
product, packing date, expiry date, batch number, and other information needed
to identify the product and trace its compliance with customs and standards
requirements.
Clearance certificate. This document is issued by the importing country’s
dairy authority as final proof that the consignment meets all trade requirements.
It is used to facilitate border procedures. Dairy imports to Kenya from outside
the EAC region are subject to a 7% KDB levy on top of normal customs duty.
10.3.3 Transport and movement across borders
Once all required documents have been compiled, regional transport and border
crossings are the next stage in the transaction. UHT milk and dry milk powder do
not require refrigeration so, in many ways, are the best-suited dairy products for
international trade and are usually shipped by road.Yoghurt, butter and cheese, on
the other hand, require refrigeration so are much more demanding and expensive
to ship. Because of the long transit times by road to some EAC markets (such as
Kenyan exports to Rwanda and Burundi), dairy products often go by air. Even
for shipments between Kenya and Uganda, airfreight is not unusual.
During transit, any number of logistical barriers may arise. The poor quality
of roads, risk of police roadblocks including demands to inspect the produce,
and excessive use of weighbridges may add time to the journey and constitute
potential threats to regional dairy trade. These constraints, of course, also apply
to domestic trade and can have particularly serious consequences for dairy
processors because of the extra time taken to move raw milk from the farm into
the dairy plant.
Although international logistics were not a major focus of investigation, it
is worth noting that many of the dairy traders met for this study specifically
remarked that the transport situation had improved in recent months. Whereas
there were previously 47 police roadblocks on the road between Mombasa and
the Malaba border crossing with Uganda, for example, there were said to be
only 14 roadblocks then. Similarly, it was reported Uganda closed all highway
weighbridges in October 2009 and that international transporters only need to
have their truck weighed once when entering the country. Beyond international
dairy trade, these moves are of obvious benefit to other sectors and may be
regarded as important steps toward improving the regional trade environment.

In terms of actual border procedures, customs clearance is always handled by
a licensed clearing agent or freight forwarder. No major complaints were voiced
over this part of the trade process and participants in the system were virtually
unanimous in saying that as long as all the paperwork is in order, everything
works fine. Sometimes there may be questions over the reported value of a
consignment, but with the new 0% tariff on EAC dairy imports this issue has
effectively gone away.
10.3.4 Market surveillance
As shown in Figure 10.1 the final step in the trade process is market surveillance.
As a matter of routine business, national bureaus of standards are meant to
draw samples of imported and domestic produce from store shelves and take
the product for testing to ensure compliance with standards. Consistent with
the agreed principle between EAC countries of mutual recognition of each
country’s national quality seal, laboratory tests should not be required for entry.
Instead, the only stage that additional tests should be carried out is during market
surveillance when domestic and imported products are tested alike.
In actual practice, however, the preceding discussion shows that laboratory
testing and other kinds of SPS analysis are almost always required for dairy
products to gain entry to another EAC country. Quality certificates with matching
batch numbers for the produce being traded must be presented regardless of
the agreement to recognise each country’s national quality mark. Because
domestic produce can be sold without further testing, this practice is a significant
disadvantage to dairy importers. Efforts to streamline the quality verification
process, therefore, could have considerable impact on the costs of regional trade
and opportunities for trade expansion.
10.3.5 Do NTBs explain low trade?
Despite anecdotal evidence of recent trade dispute and the many trade
requirements that could potentially be used to disrupt trade, EAC dairy trade
is not currently significantly affected by NTBs. The majority of private market
actors interviewed stated that they did not experience major problems with
NTBs. Actually, stakeholders report that the current trade situation has improved
considerably. Other than complaints for the additional laboratory tests required
by Uganda and large amounts of paperwork required for trade by all countries,
no major barriers to trade were identified by any of the people met for this study.
Instead, the most common remark was that as long as all the paperwork is in
place everything works fine. Some of the people interviewed attributed this
improvement to the recent drought in Kenya, which they say has increased the
country’s need for dairy imports and warned that the situation could deteriorate
next time the rains are good. Others, however, took a more favourable view and
put the improvement down to the general maturing of the dairy industry whereby

dairy processors and dairy regulators in all EAC countries have come to realise
there is much more to gain from free trade than by working against it.
From an institutional perspective, for example, one very important development
in favour of free trade was said to be the formation of the East African Dairy
Regulatory Authorities Council (EADRAC) in 2006. EADRAC is composed of
all national dairy boards or other government body responsible for the dairy
industry in EAC countries. EADRAC meets twice a year with the purpose
of providing a regular venue for dialogue on regional trade issues, non-tariff
trade disputes, and other matters of regional importance. According to industry
insiders, EADRAC was instrumental in resolving the complaints of NTBs
against Kenya in 2009. EADRAC members explained that the phone lines are
always open whenever there is an issue to discuss and emphasised their strong
belief that EADRAC will make future trade disputes less likely.
Taken together the question of standards remains an important challenge to
EAC free trade aspirations. On the one hand, EAC countries have committed
themselves to mutual recognition of each other’s quality seals, but in practice,
there is an evident lack of confidence in foreign quality marks that has led to
demands for extra laboratory analysis and delays in the trade process. According
to a senior representative of UNBS, this issue is currently being discussed
between EAC countries in an effort to harmonise certification procedures and
specify the exact competences required of each certifying body. Whether or not
EAC countries can reasonably expect to develop the full set of competences
needed to test dairy products according to the letter of all regional standards any
time soon, however, is unlikely and would almost certainly involve very high
costs that are disproportionate to the value of EAC dairy exports.
Dairy trade grew strongly during the decade prior to 2008 albeit from a very
low base level. Only an insignificant amount of total marketed milk is traded
in some form. This is highly unlikely to be the result of NTBs. Trade remains
burdened by frictional costs which traders consider an unnecessary nuisance
although the aggregate costs of these were generally believed to be small.
10.4 Will harmonised standards constrain or facilitate future trade?
10.4.1 Standards and trade in developing countries
Standards are descriptions of the characteristics of a product. They serve to
establish agreement between producer and consumer on the nature of the product
traded between them thereby making production and trade possible while
ensuring that the product meets the expectations of the consumer. Standards
serve a multiplicity of functions ranging from simple trade standards such as
standards of weight and length to meeting key social objectives like food safety.
Standards are part of the basic infrastructure of a market economy and little, if
any, production and trade would take place without them.
Standards are defined and implemented by either private or public actors and
they may be informal or formal. In developing countries much production and

trade take place with reference only to informal standards. For instance, most
food is traded in open markets with the two parties to the transaction agreeing
on the nature of the product based on traditional perceptions. Milk in the EAC
is expected to be raw, not diluted with water, and fresh. Normally, no public
agencies interfere with this trade. Naturally, one side of the transaction may
know the nature of the product better than the other and this leaves opportunities
for cheating open. To reduce such market fraud and its negative consequences
including food poisoning, various options exists. These include, for instance,
buying only from traders with a good reputation or sticking to established brands
when it comes to processed dairy products. Another alternative is the use of
government regulation.
Government regulation is widespread in developed countries but much less
effective in developing countries. Nevertheless, at least on paper, government
regulation has been widespread in the EAC dairy industry. Governments have
attempted to control quality and other parameters by the use of standards and
laws although without much success. Dairy regulations are often outdated and
in some instances date back to the days of colonisation. Governments have been
constrained both by the lack of effectiveness of their bureaucracies, the lack of
funding and the challenge of controlling the dairy industry which is dominated
by hundreds of thousands of small-scale milk producers and traders scattered
over large rural areas with poor infrastructure. Therefore, while government
regulation was extensive on paper, in reality it was very weak.
At the international level, standards have proliferated over the last two to three
decades. Food safety regulation in developed countries has been expanded as a
response to recent food safety scares and, at the same time, standards have become
the topic of many trade disputes. Many other types of standards have been more
commonly used in response to globalisation.Value chains are increasingly broken
up in geographically dispersed segments and standards are needed to ensure that
parts and inputs can be assembled in the final product. Many governments in
developing countries and the donors and international organisations that support
them have also given increased priority to food safety and other standards. The
result has been reforms of food safety regulation, standardisation bodies and
other legislation and institutions involved with standards.
Standards are increasingly seen as a source for NTBs. The global trend towards
lower tariffs is expected to lead to the use of untraditional means of protectionism
and standards are often useful in this respect as they are highly technical and can
be tailor-made to stop particular imports. Two multilateral trade agreements,
the SPS and TBT agreements, have been created within the WTO to distinguish
legitimate standards from those that are NTBs and to outlaw the latter. Both
agreements encourage, but do not stipulate, Member States to harmonise their
standards. The logic behind standards harmonisation is that when standards
become the same across borders, they can no longer be used as an excuse to stop
trade. Since the entry into force of the agreements in 1994, many donors and
international organisations have supported interventions in developing countries
to apply international standards as the basis for domestic regulation.

10.4.2 Harmonised EAC dairy standards
The EAC, supported by the donor community, has chosen to harmonise its
dairy standards with the international ones. In October 2006, the EAC dairy
standards committee held the final meeting in the harmonisation procedure
recommending the adoption of eight product standards covering butter, UHT
milk, yoghurt, powdered milk, raw milk, pasteurised milk, dairy ices and ice
cream, and sweetened and condensed milk. In addition, nine testing methods
supporting these standards were harmonised. The standards were later adopted
as common EAC standards. The Common Market for Eastern and Southern
Africa (COMESA) is currently working on adopting near identical standards
as common COMESA standards. In addition, efforts are underway to adopt the
standards in the Southern African Development Community (SADC) as well.
The basis for the EAC harmonised dairy standards is the standards of the Codex
Alimentarius. The Codex Alimentarius is created by the United Nations’ Food
and Agricultural Organization (FAO) and the World Health Organisation (WHO)
to set food standards and its standards are officially recognised as international
standards in the SPS agreement. The standards are codifications of best practice
in developed country agriculture.
The harmonised EAC dairy standards were formulated as a trade liberalisation
policy. The standards harmonisation work was largely promoted and driven by
the USAID project entitled ‘Regional Agricultural Trade Expansion Support
Program’ (RATES) based in Nairobi together with Land O’Lakes. Working
through the newly created Eastern and Southern Africa Dairy Association
(ESADA) that Land O’Lakes and RATES helped to set up and fund, a process of
working with regional governments began to develop and promote harmonised
EAC dairy standards based on the Codex Alimentarius. The strong trade
orientation of the programme is witnessed in a RATES policy paper:
A pre-requisite for increasing intra-regional trade is rationalization of
policies, procedures, regulations, rules, standards and grade that govern the
dairy sub-sectors in individual countries and their harmonisation between
countries. The process of rationalization and harmonisation will contribute
to creating an enabling institutional environment for dairy markets in the
region. (RATES, 2004: 2)
The policy paper is the main written justification of the standards harmonisation
identified during fieldwork.As witnessed above, the paper demonstrates a strong
belief in the existence of NTBs disguised as dairy standards.
The policy paper also recommends that trade needs should be carefully
balanced with other policy objectives:
The challenge of rationalization and harmonisation of policies therefore
requires a critical look at both the relevant technical and economic issues
by informed stakeholders at all levels, and a keen awareness that optimal

outcomes must be based on tradeoffs between the technical issues and the
economic interests of the stakeholders. In addition, an important underlying
consideration is the ability of individual countries to effectively implement/
enforce the agreed on standards … [A]n important issue for the dairy markets
in COMESA and EAC region is to commonly adopt standards that do not
unnecessarily hurt domestic and regional markets, especially where there
are insignificant exports outside a region and/or where sanitary conditions
are similar (RATES, 2004: 2)
However, in reality, trade liberalisation and, in particular, the view that dairy
standards constituted NTBs, became the dominating priority in the standard-
setting process. Neither the policy paper nor any other background work
made efforts to analyse the effects of the standards to be harmonised. Nor
does the paper, or subsequent policy work, make an attempt to evaluate any
gaps in implementation capacity. Such gaps would soon appear as the Codex
Alimentarius standards on which the harmonised standards were to be based,
were developed for much more advanced dairy industries than the EAC ones.
The standard-setting process was highly problematic. RATES supported the
harmonisation process through a trade specialist within the programme itself
and through a dairy technologist employed at Land O’Lakes. RATES funded a
number of regional and national workshops to discuss the standards and to allow
stakeholders to provide inputs into the standard-setting process. No publicly
available minutes exist from the meetings, but the consultants have discussed the
standard-setting process with key persons involved. These in-depth interviews
have uncovered a series of key aspects of the standard-setting process:
• The process was driven mainly by the donor and by the technical
agencies involved such as bureaus of standards.
• Private sector inputs were relatively limited during the process.Although
some of the larger dairies in East Africa participated in some meetings,
small-scale milk vendors were rarely consulted.
• Some countries participated in the standard-setting process without
good knowledge about the quality of their own milk and the challenges
they would face in quality upgrading.
• The standard-setting process was fast, lasting approximately one year.
This appears especially rapid given the status of knowledge and the
wide range of product standards harmonised across the entire sector in
EAC and COMESA countries.
• The standard-setting process mainly involved dairy technologists. No
assessment of economic and/or wider social impacts was made.

10.4.3 Justification of harmonised EAC dairy standards
Dairy standards do not give rise to significant NTBs in East Africa as evidenced
in the discussion of potential NTBs above. However, harmonised dairy standards
could still be justified by either public health or business concerns. This section
discusses the harmonised dairy standards from these two perspectives.
Public health
Public health is a major motivation behind dairy standards. Milk and milk
products may cause harm if the microbial quality is poor and if zoonoses, diseases
transferred from animals to humans, are present. A study by the Smallholder
Dairy Project analyses public health issues in EAC dairy consumption (Omore
et al, 2005). The study demonstrates that the microbial quality is generally poor
and exceeds the newly set EAC standards by a large margin. The EAC standard
for raw milk sets three qualities of milk with class C/III milk being the lowest
quality. For this class total bacterial count may not exceed 2,000,000 colony-
forming units (cfu) per millilitre (ml) and the milk may not contain more than
50,000 cfu per ml of coliforms. The total bacterial count is mainly a function of
storage temperature and time elapsed since milking, while the coliform count
indicates the level of hygiene, since coliforms are microorganisms of faecal
origin.
Most milk analysed by Omore et al exceeded the limit set for total bacterial
count by a large margin. The reasons were tropical temperatures, long delivery
times to the point of consumption, and lack of cooling technology. Milk fresh
from the udder contains very low bacterial counts (less than 1000 cfu/ml), but the
environment where milking is done, the equipment used and often poor hygiene
quickly changes this. This is clearly seen in Table 10.1. Bacterial numbers are
higher the further away from the production centre the milk is tested. In all
samples, apart from in rural Nakuru (very close to the milking) the total bacterial
counts exceed the EAC raw milk standard by a wide margin.
Table 10.1 Average bacterial counts in Kenyan milk samples
Total bacterial count Coliform bacterial count
Farmer group 7,900,000 15,000
Average 39,800,000 50,000
Nairobi 316,000,000 50,000
Urban Nakuru 20,000,000 20,000
Rural Nakuru 1,300,000 1,000
Source: Omore et al (2005).
Pasteurised milk also has difficulties complying with EAC standards. The EAC
standard stipulates that pasteurised milk should have a total bacterial count of
less than 30,000 and a coliforms count of less than 10. Over half of pasteurised
milk samples failed to meet the standard. In Nairobi, 82% and 59%, respectively,

had total bacterial and coliforms counts above the standard.A substantial number
of samples recorded total counts above 1 million cfu/ml.
The work done by Omore et al (2005) and stakeholder interviews establish that
most EAC dairy products would be technically illegal if the harmonised dairy
standards were implemented. Currently the threat of enforcement is remote, at
least on the domestic market. The challenges of enforcing such radical changes
as stopping a very large proportion of the current dairy trade are immense.
However, stopping dairy trade at the border, fully or partially, is more feasible as
the limited size of this trade is unlikely to cause severe market disruptions and as
control points are already established.
Despite the low dairy quality observed by Omore et al (2005), the consumption
habits of EAC consumers largely prevent the low quality of milk from becoming a
major health risk.All sampled urban households boiled milk (raw or pasteurised)
before consumption as did 96% of rural households. The widespread practice
of boiling milk reduces the public health risks from bacterial pathogens to
very low levels (Omore et al, 2005). It should be noted that the dairy product
mainly consumed in the EAC region, that is raw milk destined to be consumed
after boiling, is not covered by the adopted harmonised standards. The Codex
Alimentarius, the source of the adopted standards, do not cover raw milk
consumed after boiling as a separate product, due to the different consumption
habits in the Western countries for which the CodexAlimentarius standards were
originally developed.
Business concerns
Quality is a major issue in dairying, not only for public health reasons, but
also because of the need to control shelf life and improve other economic
parameters. Any market actor will need to manage quality issues and in East
Africa, all commercial dairies have a quality management system. The nature
and effectiveness of the quality management system, however, varies greatly
depending on the capacities of the company and the quality demand of the
end market it serves. Some market actors have a very rudimentary system and
experience frequent problems with shelf life and inconsistent quality of their
products, while others have much more modern systems that allow them to serve
the high-end market.
A company may benefit from trade standards. Trade standards serve important
functions both in domestic and international exchange. They transfer knowledge
to market participants about the key parameters of the product they are
exchanging. The variability inherent in dairy products necessitates a degree of
standardisation to allow market participants to choose the dairy products that fit
their production process. UHT milk, for instance, necessitates the procurement
of raw milk with very low bacterial counts because the milk is for long-term
storage during which even small bacteria amounts will grow exponentially.
Standards also promote market growth as they reassure consumers and traders
about the nature of the product exchanged. In this sense, standards make it

possible to exploit the benefits of trade like economies of scale and improved
quality and increased product variety.
The formation of product quality may happen through private or government
actions or both. In the EAC, quality management is mostly a private undertaking.
Private dairies do their own testing and set their own quality criteria. Informal
traders and consumers interact using informal understanding of what constitute
acceptable quality milk as the basis. The public regulations that do exist appear
to have a limited influence on the formation of quality criteria for milk and dairy
products.
The harmonised EAC dairy standards play no role in private quality control.
The companies will operate with their own quality requirements that may or may
not be aligned with formal standards such as standards set by national bureaus
of standards. It is therefore important to distinguish between such private
requirements, which are often determined internally in a company to optimise
its production process and marketing opportunities, and the ones coming from
outside the company, such as EAC standards.A company has no use of unrealistic
standards that, if they chose to comply with them, force it to produce products
of a quality that consumers cannot afford and for which there is no market. The
difference between the dairy markets that companies cater to in the EAC and
the markets for which the Codex Alimentarius standards were developed make
the new harmonised standards unrealistic and therefore irrelevant. The lack of
effective private sector consultations during the harmonisation process is partly
responsible for the lack of demand for the standards from EAC dairy operators.
10.4.4 The impact of harmonised EAC dairy standards
Ironically, the harmonised EAC dairy standards may in the future be used to block
the trade they were designed to facilitate. Both the analysis done in Omore et al
(2005) and stakeholder interviews confirm that most dairy products in the EAC
would be technically illegal if EAC authorities were to implement and enforce
the new harmonised standards. In the past, the regulatory frameworks for dairy
industries in the EAC have rarely been enforced as they were rarely updated to
reflect current realities and unnecessarily complicated. The regulation practised
has always differed significantly from the regulation on paper. This tradition
is likely to continue after the adoption of harmonised EAC dairy standards.
Nevertheless, the harmonised standards may be used to stop trade at the border
should protectionist interests succeed in achieving a policy reversal after the
Partner States removed tariffs on intra-regional trade.
During the harmonisation process, decision-makers relied on untested
assumptions about the role of the dairy standards that existed prior to
harmonisation.StandardswereassumedtobeasourceofNTBsandharmonisation
was assumed to be able to remove them. Both assumptions were not analysed but
have later been proved wrong. The analysis done in this chapter did not find that
dairy regulation including standards caused significant NTBs.

Harmonisation could not remove NTBs, because the Codex Alimentarius
standards adopted are unfit for the East African dairy industry. In principle,
harmonisation simply means making standards the same across borders, but
harmonisation with Codex Alimentarius standards means making standards the
same and much higher. As few dairy operators are ably to comply with the new
standards, they do not facilitate but in effect constrain trade.
Codex Alimentarius standards are the end result of a long process of
quality upgrading in developed countries. EAC decisionmakers, including
government agencies and donors, did not understand the process of quality
upgrading. Dairying in the USA and the EU began when demand increased
with urbanisation. Technological development allowed the introduction of
mass production machinery and conservation methods like pasteurisation. In
the latter half of the 19th Century dairying began to grow from an on-farm
activity to industry. Standards were used to allow for long distance exchange
and to safeguard consumer health. Standards were introduced gradually and
represented a codification of existing best practices. Standards were targets that
the industry could aim for.4
The harmonised EAC dairy standards overlook the
history of the development behind the standards. In the EAC, the introduction of
Codex Alimentarius dairy standards is not a codification of best practice. It is an
attempt to make the industry leap frog to a level much beyond current practice.
Some proponents of the harmonised EAC dairy standards admit that current
dairy quality falls significantly short of the level set in the standards, but they
argue that quality problems are a short-term adjustment problem. Further, they
maintain that the industry will have to upgrade in any case. A common reason
given for the need to upgrade is the desire to access international export markets
beyond EAC countries. This view is problematic for two reasons: first, the strong
dominance of small farmers and informal markets is not a short-term transitory
phenomenon, but a key characteristic of EAC dairy farming that will persist for
many decades. Second, there is a lack of demand for the standards about to be
put in place. The East African consumption habit of boiling raw milk removes
the public health rationale and the private market has created its own quality
levels and associated quality management systems. Both public health and
private market coordination can be improved, but not by offering essentially
irrelevant standards.
The introduction of harmonised EAC dairy standards will have negative
consequences for the rural and urban poor. First, small-scale traders have lower
costs than commercial processors and therefore pay higher farm gate prices.
Second, small-scale trading in itself is a major job market for poor people.
Third, the harmonised EAC dairy standards impose many additional costs of
compliance including investments in new equipment at the farm, bulking, and
processing levels, new cold chain facilities, and public and private laboratory
4 For an example of the use of standards and other means to safeguard public health and improve the
dairy industry, see the description in Wessel (1984) of the development of the Baltimore milk market
in the early part of the 20th century.

facilities. Higher processing costs must either raise the cost to consumers or
will result in less total value from milk being available to flow up the chain to
farmers. Simply put, inappropriate standards may mean that poor consumers
will have a higher food bill and will reduce their milk consumption thereby
foregoing the benefits from the nutritional value of milk, while farmers are also
likely to suffer from lower farm gate prices and reduced demand.
10.5 Conclusions and policy recommendations
Policymakers in many developed and developing countries often chose the
harmonisation of standards as a trade facilitation instrument. This chapter
illustrates that the design of the harmonised standard chosen will determine
whether harmonisation facilitates trade or itself becomes a source of new NTBs.
Harmonised standards must be designed for local contexts. The harmonised
standards will need to address key social objectives and be realistically achieved
by industry operators as well as to be implementable and enforceable by local
authorities given the constraints in technical, human and financial resources that
they face. Poorly designed standards may themselves become sources of new
NTBs. Standards that industry operators cannot achieve may force regulators
of domestic production and trade to ignore them in order to avoid market
disruptions. Protectionist pressures may push border regulators to enforce the
harmonised standards. If they chose to do so, they can stop a large part of trade
at will.
Working with donors, the EAC has adopted a series of dairy standards that are
not designed to meet the social objectives of the EAC countries. Dairy products
are subject to public health risks due to their highly perishable nature. This
is addressed by the common consumption habit of only consuming raw milk
after boiling. This consumption habit is not reflected in the new harmonised
standards. Furthermore, the standards do not build on the existing quality
management practices of East African dairy operators, but attempt to introduce
a series of radical new production and trading methods. Dairy operators cannot
meet the new requirements in an economically meaningful way. If followed,
the implementation of the new standards will exclude poor producers and
traders from the dairy industry due to high production costs and reduce the milk
consumption of poor rural and urban consumers severely because of high milk
prices.
The EAC just finished a year-long process of removing tariffs on intra-EAC
trade. However, support for protectionism is still widespread among government
officials and business leaders. The trade policy challenge of the coming years
is to avoid policy reversals. NTBs can be expected to surface as an alternative
to traditional forms of trade protectionism. The new harmonised EAC dairy
standards make nearly all dairy products technically illegal and can therefore be
used to stop dairy products at the border.
The new EAC dairy standards need to be reviewed in order to avoid them
becoming NTBs. A review may establish demand for another set of standards

better in tune with market needs and capabilities. Such standards could be
developed with the assistance of donors and international organisations. The
FAO and the WHO could be consulted on the development of a standard for the
unique product of the region: raw milk destined to be boiled before consumption.
Standards must be assessed vis-à-vis clearly articulated demands from their
users in the private sector in the EAC.
Methods for the assessment of regulation in developing countries need to be
developed and incorporated into policymaking. The policy process that led to the
adoption of the harmonised EAC dairy standards did not include the assessment
of economic and social effects. EAC countries face many new demands for
regulation in new areas such as food safety, animal health and environmental
protection. The process of formulating policy must be improved to ensure that
the new issues are addressed in line with the needs and capacities of the EAC
region. EAC Partner States and the donors that support them should avoid
importing policy measures designed for developed countries without adjusting
them to the realities in East Africa. In developed countries Regulatory Impact
Analysis (RIA) is increasingly being used to improve the knowledge base of
policymakers. RIA includes economic and social factors that need to be explicitly
analysed when considering policy alternatives. RIA is a relatively new concept
in developing countries (Jacobs, 2006). The use of RIA to analyse the economic
and social impact of new regulation should be considered in the EAC. Donors
and international organisations with the appropriate analytical expertise would
need to develop regulatory assessment methodologies suited to developing
country contexts.
References
Bennett, A, Lhoste, F, Crook, J and Phelan, J (2006), ‘The Future of Small Scale
Dairying’, in A McLeod (ed), Livestock Report 2006, The Food and Agriculture
Organization (FAO), Animal Production and Health Division, Rome.
IFAD (International Fund for Agricultural Development) (2005), Republic of Kenya
Smallholder Dairy Commercialization Programme, Report and Recommendation
of the President to the Executive Board on Proposed Financial Assistance, IFAD,
Rome.
Jacobs, S (2006), Current Trends in Regulatory Impact Analysis: The Challenges of
Mainstreaming RIA into Policy-Making. Washington DC, Jacobs and Associates.
Jensen, M F and Keyser, J C (2010), Non-Tariff Measures on Goods Trade in the
East African Community – Assessment of Regional Dairy Trade. Washington
DC, The World Bank.
Keyser, J C (2003), Impact Assessment: Uganda Private Sector Dairy Industry
Development, Kampala, Land O’Lakes International Development.
Michael, W (2009), ‘East Africa: Milk Trade War Spills Over Uganda and Kenya’,
IPS Inter-Press Service News Agency, 12 March, Nairobi.
Mwamunyange, J (2008), ‘Nairobi Rules Souring our Milk – Dar’, The East African,
9 August, Nairobi.

Omolo, L O (2008), ‘Trade War has Intensified between Kenya and Tanzania’, online
article posted by Africa Press International, 9 September, Nairobi.
Omore, A, Lore, T, Staal, S, Kutwa, J, Ouma, R, Arimi, S and Kang’ethe, E (2005)
‘Addressing the Public Health and Quality Concerns towards Marketed Milk in
Kenya’, SDP Research and Development Report 2, Smallholder Dairy Project.
Parsons, R L and Matiru, G N (2008), Kenya Dairy Development Program (KDDP)
Impact Assessment, Nairobi, Land O’Lakes International Development.
RATES (Regional Agricultural Trade Expansion Support Program) (2004), Regional
Dairy Trade Policy Paper, Report researched and published under the auspices
of COMESA and EAC in collaboration with the RATES centre, Nairobi and
ASARECA/ECAPAPA, Entebbe with support from USAID/REDSO, Nairobi.
Nairobi, RATES.
Staal, S J, Waithaka, M, Njoroge, L, Mwangi, D M, Njubi, D and Wokabi, A (2003),
Costs of Milk Production in Kenya, Nairobi, The Smallholder Dairy Project
(SDP).
Wessel, J E (1984), ‘Baltimore’s Dairy Industry and the Fight for Pure Milk, 1900–
1920’. Paper presented at the thirtieth annual meeting of the Business History
Conference, 8–10 March. Bureau of Economic and Business Research, College
of Commerce and Business Administration, University of Illinois at Urbana-
Champaign, IL.
World Bank (2008), Non-Tariff Measures on Goods Trade in the East African
Community: Synthesis Report. Report No. 45708-AFR. 10 October, PREM 2,
AFCRI, Africa Region, The World Bank.

PART III
Governments’Efforts to Improve NTMs

11
Improving the Quality of NTMs through
Regulatory Impact Assessment
11.1 Introduction
Whatever their form and intent, like all trade instruments, non-tariff measures
create wedges between domestic and world prices. In the case of QRs, these
wedges generate rents for distributors or holders of licences. In the case of
technical regulations, the wedges will typically reflect compliance costs, making
NTMs ‘dissipative barriers’. Conceptually, NTMs can have two main direct
effects on the home economy. When imposed on consumer goods, they affect
poverty and the distribution of income, with sometimes non-trivial effects, in
particular when domestic producers of substitute products have market power,
which is likely to be the case in small markets. When imposed on intermediate
products, NTMs affect the competitiveness of domestic firms. Recent work
(Augier et al, 2011) shows that firms benefit from imported intermediates
through higher productivity when they have the in-house skills to make good use
of them. This suggests that NTMs making foreign sourcing costlier can reduce
productivity growth.
It is critical for governments to deal effectively with NTMs for various reasons.
First, NTMs encompass a broad array of regulations including traditional
ones, like prohibitions or quantitative restrictions, but also new-type measures
like technical regulations (TBTs) and sanitary and phytosanitary measures
(SPS). While the WTO recognises the right of governments to regulate trade,
governments need to be aware that these measures can raise the prices faced by
households for key consumer products, in particular foodstuffs, reducing real
incomes and aggravating poverty. For instance, a pilot application of the World
Bank’s toolkit in Nigeria suggested that eliminating import prohibitions would
reduce the cost of living by enough to lift three million people out of poverty.
These NTMs can also penalise domestic employers dependent on critical inputs
from abroad, hurting competitiveness and employment. Another application
of the toolkit in Indonesia highlighted a case where an ill-conceived technical
regulation on imported steel penalised domestic producers of cars and household
goods, among Indonesia’s largest providers of jobs and export earnings.

NTMs can also be easily hijacked for protectionist purposes, in which case
they are called non-tariff barriers (NTBs). First, they can discriminate de facto
(eg through technical requirements designed to hurt foreign producers more than
domestic ones). Second, by their complexity and the lack of transparency that
often surrounds them, they can conveniently obfuscate protectionist intent and
effect. Thus, they represent a largely – as yet – untapped arsenal of protectionist
measures with limited and imperfect WTO disciplines.
Yet, seeking to eliminate NTMs altogether or to cut their number through
mechanical formulae would likely be an ill-conceived quest. Most NTMs respond
to a genuine public demand for traceability and protection against hazards to
health and the environment – a demand that can be expected to grow over time
and with wealth. Thus, except in the case of NTBs, policy toward NTMs should
strive for their improvement through better design, smarter enforcement and,
ultimately, through a robust governance framework.
11.2 Improving NTMs: a governance framework
NTM improvement should be viewed primarily as a domestic issue, part
of a regulatory-improvement agenda driven by a concern for enhanced
competitiveness, rather than as a concession to trading partners. This is the
perspective taken by countries which have successfully adopted ambitious
regulatory-reform agendas, like Korea or Mexico. In practice, however,
regulatory improvement is difficult. For new measures, the standard approach is
‘Regulatory Impact Assessment’(RIA). For existing measures, there has been so
far no tool to guide their review. This section will focus on the latter following
the World Bank toolkit (Cadot, Malouche and Saez, 2012).
Trade economists and lawyers have tried to get around conceptual problems
by drawing a distinction between non-tariff measures (NTMs) and non-
tariff barriers (NTBs). The latter are the evil form of the former, when trade
restrictiveness is, deliberately or not, in excess of what is needed for the measure’s
non-trade objectives. One could, conceptually, try to put the various types of
NTMs in boxes corresponding to either NTMs or NTBs, and to give them colour
codes – green, amber or red – like in agricultural negotiations. However, such
an approach is unlikely to take us very far, as what matters on the ground has
typically more to do with how measures are applied than with what measures
are applied. For instance, a technical standard may create unnecessary problems
essentially because it requires certification of foreign production facilities.
Or, an SPS measure may be highly trade-inhibiting because every shipment is
inspected, but would be much less so if customs used risk profiling.
In other words, the devil is in the detail when it comes to streamlining NTMs.
Governments need to recognise the complexity and diversity of NTMs. Problems
should be identified at the country level through consultations with the private
sector, and technical solutions should be sought through careful analysis and
private-public dialogue. The underlying philosophy is similar to what is known
as ‘regulatory improvement’ or ‘Regulatory Impact Assessment’ (RIA) (Figure

Improving the Quality of NTMs 215
11.1). However, whereas RIA is for the ex-ante analysis of measures – before
they are adopted – the approach being discussed here aims at the review of
existing measures, in response to specific demands from countries struggling
with legacies of complicated and penalising regulations. Dealing with existing
measures has the advantage of responding to an immediate need and focusing on
measures whose effects are known. However, it should be thought of as only the
first stage of a process of regulatory improvement covering not just the stock of
existing measures, but also the flow of new ones, lest streamlining efforts would
have to start all over again as poorly designed new measures keep on appearing.
Figure 11.1 A flowchart of the NTM toolkit
Reduction of the knowledge gap
Governments Private
sector
Research
community
Policy
dialogue
Regional
dialogue
Review of existing
NTMs
Better procedures
for new ones (RIA)
Lower barriers to regional and
South-South trade
Improved regulatory
environment for reinforced
competitiveness
NTM
REVIEW
TOOLKIT
DATA
COLLECTION
This approach relies on a more sustainable institutional set-up that would ensure
continuity in the process of improving the trade competitiveness of firms as
the business environment evolves over time and the stock of regulations keeps
growing. It should be based on the following three pillars:
• A body dedicated to public-private dialogue (eg an ‘NTM committee’)
serving as an entry point for the private sector to flag problems and
contribute to the solution;
• A technical team dedicated to carry out substantial analysis (eg
a permanent secretariat for the NTM committee) with analytical
capabilities akin to those of a productivity or competition commission,
to lead the dialogue into policy action;
• A broad participation drawing into the review process outside expertise
and collaboration from line ministries involved in the issuance and
enforcement of NTMs, to ensure ownership.

Regulatory reforms are intensive in political capital, and the political economy
of reform is usually complex.1
Depending on the strength of the political
commitment, reform process is more or less successful in streamlining NTMs
(Figure 11.2). A necessary condition to undertake the process is to ensure
political support and commitment of the government and all relevant authorities
that are involved in the process. This means that when designing an institutional
arrangement to conduct the review of existing regulations, the process has
better chances to succeed when (i) driven by a high level of the administration,
which would have the appropriate mandate (supported by the law, decree or any
other decision); (ii) it involves all the concerned agencies; (iii) it ensures the
participation of the highest officials responsible for the issuing and administration
of the measures, including the agency’s staff; and (iv) last but not least, it has the
necessary technical and financial resources to conduct its mandate. This means
that the officials involved in the regulatory process must benefit from training,
improvement of their skills as well as tools to perform their tasks.
Figure 11.2 Political commitment and NTMs reforms
Relatively
inocuous
Severe
Relatively
inocuous
Severe
Full review with
institutional
setup
Systematic, light
review
Guillotine
approach
Korea, Mexico
Ad -hoc--start
with low -
hanging fruit
Indonesia, Nigeria
Many NTMsFew NTMs
Mauritius
Strong
Weak
Prevalence/severity of NTMs
Politicalcommitment
Source: Author’s elaboration.
First, strong political support is essential to achieve results and effectiveness,
more than any formal basis such as law, decree, directive or resolutions – to
set up any institutional set-up to lead the review process (OECD, 2009).
Depending on each country’s legal environment, the level of the regulations
(laws, decrees, administrative decisions, etc), the level of the authority involved
(national, subnational, or municipal), or the entity responsible for issuing the
regulation, the reform process may be more or less complex and the institutional
arrangements proposed may be more or less appropriate. The proposal outlined
1 See Tools for Effective Regulatory Reform Tools and Approaches to Review Existing Regulations,
IFC Business Regulation Group and OECD Regulatory Impact Analysis: A Tool for Policy
Coherence, 2009. When engaged in wide regulatory reform, other factors such as importance of
external pressure, taking advantage of the context in which reforms take place (for instance economic
crises) are examined in Akinci and Ladergaard (2009).

could be implemented at different levels of government, but flexibility should
be considered when assessing the final approach in a specific country context.
Thesecondconditionofsuccessisthechoiceoftheagency/ministryresponsible
for the conduct of the review process. International experience with streamlining
of business regulations to improve domestic competitiveness indicates that the
reviewing agency/ministry should be provided with adequate independence
and authority to conduct the review from the agencies responsible for issuing
the regulations (Akinci and Ladergaard, 2009; Jacobs and Ladegaard, 2010).
Effective results are more limited when the review is conducted internally by a
ministry/agency. Regulators tend to keep most of their regulations. However, if
such an institutional design is not politically feasible, the reviewing unit should
be located in a politically and technically strong ministry that could provide
the necessary support to conduct the process and is technically capable of
conducting it. A key element of success is that the agency/ministry responsible
for the review must report and be accountable for the work undertaken. Another
option is to create an inter-ministerial committee responsible for conducting the
review as in the case of Japan where the Regulatory Reform Ministerial Council
(Chair by the Prime Minister and a Coordination Agency lead the process or
in the case of Mexico where the COFEMER, a public agency, oversees the
regulatory making process.
Third, the agency/ministry responsible for conducting the review process
must be endowed with the necessary resources to undertake a thorough review
of regulations (Jacobs and Ladegaard, 2010). Experience shows that lack
of resources can jeopardise the process because the entity is not capable of
undertaking or outsourcing complex regulatory issues. This means, among other
things, that the agency responsible for conducting the review must have the staff
and skills to conduct the review. For instance, if the review agency is endowed
with one or two staff it probably will not be able to conduct a thorough review
process.
Fourth, beyond existing measures, the responsibility for improving the overall
regulatory environment goes beyond a single unit and is a continuous process.All
relevant agencies must adopt best regulatory practices (IFC, 2010). The agency
responsible for the review should also have the following responsibilities:
• todefinethegeneralprinciplesofregulatoryreformbasedoninternational
best practice and to ensure that these are applied consistently across line
ministries and departments;
• to oversee the introduction of regulatory impact analysis as a key tool
across the government;
• to facilitate the intra-ministry coordination that is essential to address a
wide range of complex multidisciplinary regulatory issues;

• to ensure transparency and public dissemination of existing regulations,
their policy objectives, and provide technical background information
that support the rationale for the regulations; and
• to ensure that the agencies responsible for issuing regulations are an
integral part of the review process, the staff benefit from training, their
skills are continuously upgraded, and the review process is not against
them but with them.
When due to political factors the institutional set-up is not feasible or possible
in the short term, the review process could be developed exclusively within the
existing institutional framework. This would require inserting a review process
within the agencies responsible for issuing the regulations. For example,
Australian authorities mandate s review process within each regulatory agency,
such as the Department of Agriculture, Fisheries, and Forestry; Department of
Health and Ageing; the Treasury; and the Reserve Bank of Australia, among
others. In order to achieve the broader goal of regulatory reform, the resources
should focus on the strengthening of the regulatory agencies. More specifically
this will require introducing the regulatory review process as part of the regulatory
process, establishing a transparent and participatory process which will ensure
accountability, and improving staff skills and knowledge on regulatory matters.
The Indonesian government has just adopted a decree to institutionalise the
review process for the introduction of new NTMs (see Box 11.1 for more details).
While the toolkit focuses on the review of the existing NTMs, such institutional
set-ups may be appropriate for both the review of existing and new NTMs.
Finally, the single most important initiative that countries could pursue
to improve their regulatory environment is to enhance transparency. Strict
transparency requirements will provide important information regarding the
nature of existing regulations, the objectives they aim to achieve, and whether
they are based on international standards and recommendations, among other
aspects. Transparency is usually an international requirement embedded in
agreements such as the WTO as well as regional and bilateral trade agreements
which countries must comply with. Usually, the establishment of enquiry point
which must provide information regarding regulations, including proposed
regulations, is mandatory, but countries can adopt additional transparency rules
beyond what is required by international requirements. For instance, they could
ensure that proposed new regulations or changes to existing regulations are
publicised and ample opportunity for comments and suggestions for domestic and
international interested parties are provided. Enhancement of transparency has
been facilitated in recent years by information and communication technologies
which lower the costs of improving access to relevant information as well as
the costs of providing and managing information. Moreover, transparency is
both an important driver as well as a supporter of regulatory reform, increased
transparency reduces policy risks for market actors and of the potential capture

of regulators by interested parties (Akinci and Ladergaard, 2009; Jacobs and
Ladegaard, 2010).
Box 11.1 Institutional set-up to streamline new NTMs in Indonesia
Streamlining of NTMs in Indonesia has received a new impetus from the Ministry of
Trade in September 2011 with the passing of a decree to launch a pilot programme
establishing a review process of NTMs issued by this ministry, currently responsible
for 61% of all NTMs issued.2
The main thrust of the programme is to strip the role
for reviewing NTM from unit implementing NTM and to equip the unit responsible
for the former with an adequate capacity to conduct regulatory impact analysis. The
decree establishs a Non-Tariff Policy Team (KNT) within the Minsitry of Trade that
would essentially (i) coordinate with relevant agencies for input into formulation and
establishment of non-tariff policies; (ii) formulate and submit recommendations to the
Minister of Trade on non-tariff policies; (iii) monitor and evaluate the implementation of
non-tariff policies; (iv) introduce non-tariff policies to stakeholders; and (v) participate
in international trade negotiations in the framework of bilateral, regional and multilateral
cooperation with implications for Indonesia’s non-tariff policies. The Ministry has also
introduced a standard operating procedure for reviewing NTMs to conduct objective
and independent assessments within a specified time. The KNT Team will analyse the
eligibility of a proposal related to non-tariff policies; analyse the impacts of non-tariff
policies using appropriate analytical tools; verify consistency with other policies and
with the World Trade Organization (WTO) rules or other international agreements; and
hold a public consultation through meetings with stakeholders or field surveys. The KNT
Team should settle the problem within a maximum of 60 working days.
2 Decree 709/M-DAG/KEP/9/2011.
11.3 Improving NTMs in Mauritius
With the reduction in tariffs, non-tariff barriers (NTBs) have emerged as a clear
obstacle dampening the competitiveness of the domestic economy. Compliance
with foreign NTBs was also harming Mauritius’s competitiveness and integration
with the world economy by holding back its export performance (see Box 11.2
for examples). When the benefits that would come from a systematic process
for the review of regulations and their implementation became apparent, the

governmentdecidedtoaddressit.FollowingthebudgetofJune20083
theMinistry
of the Economy was given the task of reviewing all permits and licences. The
Permit Review Committee (PRC) was established to review the need for import/
export permits where they exist and also to submit proposals for the simplified
processes and procedures for the issue of such permits and clearances.
The PRC conducted 13 meetings and consultations to carry out its mandate.
First it identified a number of tariff lines requiring permits/clearances from
various ministries and organisations and requested each ministry to submit the
necessary justifications for maintaining permits under their purview. The PRC
also requested all players to submit the detailed procedures, fees payable and
processing time for the issuing of permits as well as the relevant legislation
related thereto. As a result of this exercise, the PRC identified about 19 types
of permits affecting trade, issued mainly by six ministries, affecting almost a
quarter of the tariff lines (2,610 out of 6,298 tariff lines), and for reasons of
consumer protection, safety, phytosanitary, compliance to norms or standards or
compliance to international agreements. The need to have import/export permits
was found unnecessary or unwarranted in 72 instances.
The PRC also noted that some ministries or their departments experience
capacity constraints in terms of human resource and/or logistics, which greatly
hampers the delivery of timely services, thus unduly increasing the processing
time. It was therefore recommended that apart from streamlining the procedures
and processes involved, permit issuing bodies must ensure that the relevant unit
or section are adequately staffed and equipped. There should also be a general
shift in mindset with focus on a high standard of service delivery and only
essential control exercised.
In addition to this important identification step, the World Bank has
recommended the government to take this effort to the next step in order to
further streamline the regulations by:
1. Setting-up a permanent regulatory review committee. The committee
would have the responsibility to (i) define the general principles of
regulatory reform based on international best practice and to ensure
that these are applied consistently across government, (ii) review all
new and important existing regulations, (iii) oversee the introduction
of regulatory impact analysis as a key tool across the government and
(iv) facilitate the intra-ministry coordination that is essential to address
3 The 2008–9 Budget Speech said ‘We also want to significantly simplify the processes for exports and
imports where bureaucracy can be unnecessarily exasperating in many cases. Our aim is to reduce
the number of permits relating to imports and exports to the essential minimum, by 1st July 2009.
Provisions will therefore be made under the Customs Act to suspend as from 1 July 2009 all permits
relating to imports and exports, except those that are considered essential. All permit authorities
will have until end December 2008 to submit to a Committee chaired by the Ministry of Business,
Enterprise Cooperatives, any justifications for maintaining the permits they issue. Furthermore, the
Committee will recommend measures to lower compliances costs.’

a wide range of the regulatory constraints, including duplication of
requirements, that currently undermine competitiveness in Mauritius.
2. Introducing an appeal mechanism to allow affected stakeholders the
opportunity to contest the decisions of civil servants that they feel are
incorrect, unfair or arbitrary.4
3. Assessing capacity gaps that undermine effectiveness and efficiency
of implementation. There are clearly critical gaps in the standards and
conformity assessment infrastructure that need to be identified and
prioritised in terms of the needs of business and the effectiveness of
regulations.
11.4 A broader business regulatory review
The streamlining of permits and licences quickly showed its limitations when it
came to trade regulations.A diagnostic of the trade regulations revealed a number
of cases that deserved a deeper regulatory analysis (Box 11.2) and weaknesses
in the regulatory system in Mauritius that were undermining competitiveness,
including the following:
• Lack of systematic approach to assessment of regulations and their
implementation. There is no clear procedure by which actual and
proposed regulations are examined and analysed in terms of whether
the design is (i) consistent with the underlying public policy objective;
(ii) with the capacity in government ministries and agencies and in
the private sector to effectively implement them; and (iii) in terms of
impacts on competitiveness, growth, employment and poverty.
• Lackoftransparencyindesignandapplicationofregulations. Regulations
are not designed on the basis of consultation with interested stakeholders,
even with different parts of the government.
• Resources are poorly utilised due to lack of risk-based approach in
many parts of the government. With the exception of customs, most
government ministries and agencies take a very rigid approach to the
implementation of regulations and do not apply risk-based approaches
that target resources, such as those for inspection, at the most risky
transactions.Hence,forexample,firmsandindividualsthathaveinvested
in processes and procedures that reduce risks face similar compliance
costs to those whose products or activities pose a greater risk. These
high compliance costs and lack of flexibility inhibits international
4 One possibility that is used in other countries is to establish an Ombudsman Office. Best practices
have been defined by the International Ombudsman Association at http://www.ombudsassociation.
org/standards.

competitiveness relative to countries where the regulatory system and
its application encourages rather than stifles innovation.
• No recourse to dispute procedures. Individuals and firms have no formal
mechanism by which to challenge the decisions of civil servants as
they implement regulations. In practice, large firms and well connected
individuals are able to exert influence and obtain a review of decisions
that they deem to be incorrect or inappropriate. Rather, it is small and
new firms and ordinary individuals who are unable to dispute decisions.
• Duplication of requirements across government ministries and agencies.
In many cases different parts of the government are making similar
requirements, such as permits or licences or the testing of products,
to achieve the same objective. For example, the export of fish requires
permits from the Ministry of Fisheries and the Ministry of Agriculture
to achieve essentially the same purpose of monitoring and collecting
information – information that is recorded by Customs and is readily
available from them. This also reflects the lack of coordination and
cooperation across ministries and agencies. While different groups of
the government have responsibility for achieving the same public policy
objective, the efficiency in achieving this objective is often undermined
by lack of coordination and cooperation.
• Lack of competitiveness perspective in many ministries and agencies. The
culmination of the weaknesses discussed above is that in designing and
implementing regulations few ministries or agencies consider the impact
of their actions on the competitiveness of the country. An approach,
similar to that of trade facilitation in customs, should spread throughout
the government leading to a move from a rigid, non-contestable mindset
in which ministries and agencies seek to shelter their requirements and
approaches to one in which civil servants are continually exploring ways
in which to better design and implement regulations through openness
and cooperation.
Recognising the need to carry out rigorous regulatory review for NTMs, the
government of Mauritius established a public-private Review Committee for
NTBs chaired by the Ministry of Foreign Trade in April 2009 to (i) define the
general principles of regulatory reform across government; (ii) review all new
and important existing regulations; (iii) oversee the introduction of regulatory
impact analysis as a key tool across the government; (iv) facilitate the intra-
ministry coordination, and (v) encourage and assist the roll out of IT solutions
for trade facilitation across ministries and agencies.
While this initiative has generated a productive dialogue among various
agencies and the private sector and helped identify various cases requiring
government attention, it was unable to address the issues because of lack of

human and technical capabilities dedicated to this process. Given government
commitment to remove non-tariff barriers, it further strengthened the business
regulatory review mechanism in early 2012 by establishing a joint public-private
business facilitation task force to coordinate and strengthen the review process
for business regulations and procedures, including a committee dedicated to the
review of NTMs. This new task force will bring together the programs of four
existing subcommittees dealing with business regulations, including the NTB
Review Committee. It also put in place a secretariat, with administrative and
technicalstaffandadedicatedbudget,tomakesureregulatoryimpactassessments
would be conducted when needed. In addition, with the assistance of the World
Bank, two pilot regulatory impact assessments were conducted on the two cases
presented in Box 11.2, and helped interested parties, the government and the
private sector, identify the exact nature of the problem and think about measures
to address it.
References
Cadot, O, Malouche, M and Saez, S (2012), Streamlining Non-Tariff Measures: A
Toolkit for Policy Makers, World Bank Group
IFC (2010), Better Regulation for Growth: Governance Frameworks and Tools
for Effective Regulatory Reform—Tools and Approaches to Review Existing
Regulations, International Finance Corporation/World Bank, Washington, DC.
Jacobs, S and P Ladegaard (2010), Regulatory Governance in Developing Countries,
International Finance Corporation/World Bank, Washington, DC.
OECD (2009), Regulatory Impact Analysis: A Tool for Policy Coherence, Paris:
OECD.

Box 11.2 Two regulations hurting the competitiveness of Mauritian firms
Regulation designed without reference to impact and capacity to implement
Problem: The government has defined the public policy objective of shielding citizens
from toxic chemicals. The relevant regulation, the Dangerous Drug Act, seeks to achieve
this through a ban on the importation of a defined list of toxic chemicals. While it is
straightforward to prevent the importation of these materials in bulk form there is lack
of capacity to inspect and test imports of final products containing the toxic ingredients,
such as paints.
Impact: While Mauritian firms can no longer use the toxic chemicals, products containing
toxic chemicals are still being sold on the domestic market. According to the Mauritius
Paint Association the production cost of paint produced in Mauritius has increased since
the non-toxic ingredients that must now be used are more expensive. Thus Mauritian
producers of paint find it more difficult to compete in the domestic market with imported
paints that contain the toxic materials. They also are at a competitive disadvantage in
export markets that have not banned such ingredients.
Application of plant health regulations creates obstacles to flower exporters
Problem: Mauritius was a global market leader in traditional (red) anthurium flowers.
International competitors have developed new species with various colours as a marketing
tool. Local producers would like to respond by growing a wider range of varieties.
There is a risk of bacterial infection from imported plants that could spread to domestic
production. In response the government has banned the import of adult plants regardless
of the risk that the particular imports may contain the bacteria. There is a lack of capacity
to test plants for the presence of the bacteria and the government currently does not
accept certificates confirming that plants are disease free from foreign laboratories. This
forces exporters to import only baby plants (which do not contain the bacteria) which
take up to two years to grow and be sold on the market.
Impact: The import ban of adult plants prevents exporters from following market trends.
The industry association claims that Mauritius has been marginalised as a single variety
exporter.

12
Addressing Non-Tariff Barriers on
Regional Trade in Southern Africa1
Ian Gillson and Nick Charalambides
12.1 Introduction
A recent and important trend in global trade has been the proliferation of
regional trade agreements (RTAs) and Southern Africa is no exception.
Regional integration efforts in Southern Africa, such as the Common Market for
Eastern and Southern Africa (COMESA), the Southern African Development
Community (SADC) and the Southern African Customs Union (SACU), have
all sought to liberalise trade between countries so as to increase bilateral trade
flows, diversify exports by overcoming the limits of small markets, and deepen
specialisation through achieving economies of scale. Yet despite these efforts,
regional trade outcomes in Southern Africa have been limited due to barriers,
particularly non-tariff barriers (NTBs), that persist on regional trade.
This chapter summarises a number of new studies that identify the most
restrictive NTBs in Southern Africa. In doing so it illustrates some of the costs
associated with them using information gathered from some of the largest firms
engaged in cross-border trade. For example, Shoprite (a South African retailer)
reports that it costs US$500 for each day one of its trucks is delayed at a land
border in the region. These costs serve to thicken regional borders and prevent
the development of regional production chains. While there have been some
success stories in the region with countries attempting to tackle NTBs, more
needs to be done. For example, Mauritius has established a joint Public-Private
Standing Committee that is undertaking a review of all NTBs there. Nevertheless
in Mauritius, as well as elsewhere, knowledge-sharing of good practice for
regulatory reform remains lacking. This chapter, therefore, aims to be a first step
in filling this gap by providing some practical policy recommendations on how
countries, either unilaterally or through a regional process where there might
1 This work was part funded by the Multi-Donor Trust Fund for Trade and Development supported
by the governments of Finland, Sweden, Norway, the United Kingdom, and the Bank Netherlands
Partnership Program (BNPP). The views expressed in this chapter reflect solely those of the authors
and not necessarily the views of the funders, the World Bank Group or its Executive Directors.

be gains from cooperation, can remove existing NTBs as well as discipline the
development of new ones.
12.2 Non-tariff barriers in Southern Africa are critically hindering
regional trade
Harnessing regional integration more effectively would help all countries lower
their cost base and enhance global competitiveness. For the smaller Southern
African countries, regional integration also offers the prospect of improved
access to neighbouring markets as well as the potential to attract greater
SADC-orientated foreign direct investment (FDI). In some of these countries
(eg Lesotho) greater exploitation of the regional market is critical to reduce
reliance on the export of a single product to a single market (eg clothing to the
USA under the African Growth and Opportunity Act). For the larger countries,
especially South Africa, regional integration offers opportunities to enhance the
sustainability of existing exports (eg light manufacturing) on world markets by
lowering costs through specialisation within the context of regional value chains.
However, while Southern African countries have been able to successfully
eliminate a significant proportion of their tariffs on intra-regional trade,
2
corresponding increases in regional trade have only been relatively minor. For
example, while their trade with the world tripled in value between 2000 and 2008
(from US$50 billion to US$153 billion) the share of regional trade has remained
steady at close to 10% over the last decade despite Southern African countries
growing faster than the world average for most of this period. In contrast, the
most successful RTAs in Asia and Latin America (eg Association of Southeast
Asian Nations, Mercado Común del Sur) have reached and maintained much
higher degrees of regional trade (typically over 20% of their total trade), often
through intensified intra-industry linkages. Furthermore, traditional exports of
agricultural raw materials and minerals continue to dominate regional trade in
Southern Africa. Cases of diversification into higher value-added manufacturing
exports to the region remain limited (eg Mauritian clothing to South Africa) and
strong trade imbalances persist between South Africa and the smaller countries.
Regional production chains for exports to the world market remain virtually
non-existent.
A lesson from successful regional integration experiences elsewhere in the
world is that tackling tariff barriers is necessary but not sufficient to enhance
trade. Countries must also aim to facilitate trade by addressing NTBs and,
2 For example, SADC has been trading on preferential terms since 2000 and, based on the
implementation of tariff phase down commitments under the Trade Protocol, formally launched a
free trade area (FTA) inAugust 2008. Under this, 85% of intra-SADC merchandise trade is now duty-
free with most of the remaining 15% comprising sensitive products scheduled to be liberalised by
2012 (2015 for Mozambique). A subset of five SADC members have already established a customs
union under SACU which is 98% duty-free. COMESA has had an FTA since 2000 and formally
launched a customs union in June 2009. Trade between FTA and non-FTA countries is conducted on
reciprocal terms under a preferential trade agreement.

Non-Tariff Barriers on Regional Trade in Southern Africa 227
here, Southern Africa has made notably less progress than with its tariffs. For
example, despite the SADC Trade Protocol making specific reference on the
need to eliminate NTBs and urging countries to refrain from imposing new
ones, moving from identifying NTBs to their reduction and removal has proven
to be challenging. Consequently, borders remain thick as major obstacles to
trade remain which undermine the predictability of the trade regime and reduce
investment in the region.
NTBs reported by firms in SADC countries under the Monitoring Mechanism
(explained in greater detail later) affect products which jointly account for
US$3.3 billion, or one-fifth, of regional trade (see Table 12.1 for a summary
and the Appendix for the detailed mapping). In other words, even those NTBs
that have been reported in SADC (and others may have yet to be notified) affect
products in which there is already significant regional trade. This is also a least
cost estimate of the impact of NTBs on trade in the region since some barriers
are so restrictive that preferential trade is effectively prohibited (eg wheat flour)
and, of course, there are others which affect all trade and not just individual
products (eg customs delays, transport costs) which are not captured here. So
NTBs are widespread in their effect on regional trade, even more so than these
figures suggest.
The mapping also shows that some sectors are affected more than others by
these barriers and individual sectors are often affected by more than one type
of barrier. For example, regional trade in wheat is affected by import bans,
import quotas, import levies, single marketing channels and rules of origin
(for flour), although obviously not all countries impose all barriers. NTBs also
disproportionately affect regional trade in agricultural commodities, particularly
sugar, maize, wheat, meat products (including poultry) and dairy products.
Regional trade in manufactures is mostly affected by restrictive rules of origin
(ROOs) as well as product standards.
The impact of these barriers on firms is pervasive. In a recent firm survey,
which included five SADC countries, roughly 80% of respondents indicated that
they faced some form of trade barrier within the region (RTFP, 2009). Over half
of the respondents indicated that the cost of these was equivalent to 5% of the
c.i.f. value of their imports/exports. A further 24% of respondents indicated a
5–15% attribution to trade barriers; and 23% faced increased trade costs of over
15%. There is also evidence to suggest that barriers in one form or another exist
in all countries throughout the region. In an inventory of NTBs in SADC (RTFP,
2007), all countries were found to maintain at least ‘moderate’ barriers.

Table 12.1 NTBs that have been notified to SADC affect at least one-fifth of
regional trade
Barrier Examples of products affected
Volume of intra-SADC
trade potentially affected
(% of total)
Import bans, quotas and
levies
Wheat, beer, poultry, flour, meat,
maize, UHT milk, cement, sugar,
eggs, pasta, sorghum, pork, fruit and
vegetables
6.1
Preferences denied Salt, fishmeal, pasta 0.4
Import permits and levies
UHT milk, bread, eggs, sugar, fruit
and vegetables, livestock, liquor,
cooking oils, maize, oysters
5.4
Single marketing channels
Wheat, meat, dairy, maize, tea and
tobacco, sugar
5.3
Rules of origin
Textiles and clothing, semi-trailers;
palm oil; soap; cake decorations; rice;
curry powder; wheat flour
3.0
Export taxes
Dried beans, live animals, hides,
skins, sugar, tobacco, maize, meat,
wood, coffee
4.8
Standards/SPS/TBT
Milk, meat, canned tuna, beer, honey,
maize bran, cotton cake, poultry,
batteries, sugar, coffee, ostriches
2.5
Customs-related
Wine, electronic equipment,
copper concentrate, salt, cosmetics,
medicines
5.2
Source: Authors’ calculations based on NTBs reported to the SADC-EAC-COMESA Non-Tariff Barrier
Monitoring Mechanism.
Global evaluations of NTBs indicate that they are in most cases more restrictive
than tariffs. While quantitative assessment of their impact is challenging, in great
part due to large gaps and errors that persist in the data on NTBs, there is a
growing body of work that attempts to estimate their tariff equivalence. A recent
survey of this analytical work (Carrere and De Melo, 2009a, b) indicates that
on average the tariff equivalent of NTBs is 40%, which for most products is
higher than the MFN tariff applied by most countries.Assuming 40% ad valorem
equivalence on the SADC NTBs cited above, would imply a crude cost estimate
of US$1.3 billion per year attributable to those barriers that have been notified –
roughly equivalent to the GDP of Lesotho.
Consequently, NTBs significantly increase costs for consumers as well as
for those firms that source intermediate inputs from the region. For example,
Woolworths reports that prices in its franchise outlets in non-SACU SADC
countries are 1.8 times higher than those within SACU because of higher
expenditures associated with sending goods to these markets as well as the
higher costs of doing business in them.

12.3 What are the main types of NTBs and how much do they cost?
There are opportunities for Southern African firms to trade across regional
borders that currently remain unexploited due to policy constraints that serve to
raise trade costs. Five main types of NTB can be broadly identified as follows.
12.3.1 Inefficiencies in transport, customs and logistics raise costs
In order for RTAs to be effective, it is critical that regional trade be able to move
without hindrance. Many Southern African countries are landlocked, making
road and rail networks very important in linking these countries to the regional
market as well as to the rest of the world via the main ports in South Africa,
Mozambique, Angola and Namibia. However, high transactions costs are being
incurred from inadequate transport infrastructure, inefficiencies in customs
procedures (including delays at road checks, borders and ports) as well as poor
quality and costly logistics due to weak competition among service providers.
They serve to raise logistics costs; require suppliers to keep higher levels of
inventories; result in a higher percentage of goods not reaching final markets;
raise the rate of spoiled agricultural goods; and, ultimately stunt the development
of new exported products. They also restrict the potential of countries to scale
up existing production for global markets by first exploiting regional sources
of comparative advantage. Shoprite reports that for each day one of its trucks is
delayed at a land border costs $500. Port congestion is also an important source
of delay, particularly at Durban – the busiest container port in Africa. The Citrus
Growers’ Association in South Africa has estimated that delays at the Port of
Durban cost its growers US$10.5 million per season (on approximately US$400
million of exports), based on an average delay per load of 12 hours for each
of the 20,000 citrus laden trucks that enter the port during peak season. One
measure being used to reduce these costs is to increasingly use the Maputo port
for citrus exports, which can offer a saving of up to US$0.50 per carton loaded
from the Maputo Cold Store compared with the cost in Durban (Cargo Info
Africa, 2010).
Another source of delay within the region concerns work permit regimes for
foreign truck drivers. In South Africa, visitor visas were once accepted for this
purpose but foreign drivers will soon be required to obtain work permits that
necessitate trucking companies proving that the skills being sought outside of
South Africa are not available locally and involves vacant posts being advertised.
There are between 1,600 and 2,000 foreign drivers in South Africa who will
require these permits, affecting 6,000–8,000 deliveries per month. While
ostensibly designed to protect employment opportunities, the new approach does
not take into account prospects for South African drivers operating elsewhere in
the regional market. In particular, it risks SouthAfrica’s neighbours reciprocating
with similar measures that will force South African drivers working in these
countries to also apply for work permits. For example, Angola has already
signalled its intention to put in place a similar requirement for South African

drivers crossing its border. Such restrictions could significantly impede the
movement of trucks in and out of countries and make trade even more difficult
for regional exporters than it is now.
12.3.2 Cumbersome fiscal arrangements widen borders
FiscalbordersbetweenSouthernAfricancountriesareunnecessarilycomplicated,
inefficient and contribute to higher trade costs. The three main reasons SACU
retains internal border posts, even though it is a customs union, are to capture
data on intra-SACU trade for revenue sharing purposes; administer NTBs such
as infant industry protection; and, because domestic sales taxes have not yet been
harmonised, requiring refunds and payments. The costs and delays associated
with these procedures reduce trade flows between Southern African countries.
Those costs attributable to the differences in VAT alone have been estimated to
be up to 2% of the value of each transaction on intra-SACU trade (Jitsing and
Stern, 2008).
12.3.3 Restrictive rules of origin limit preferential trade
Onerous local content requirements in ROOs (see Box 12.1), particularly
in labour intensive sectors (eg clothing) that use capital intensive inputs not
produced competitively in the region (eg fabrics), and high compliance costs with
administering certificates of origin reduce the utilisation of tariff preferences
offered by RTAs and therefore the incentive for Southern African firms to trade
regionally.
Box 12.1 Rules of origin in Southern African RTAs
ROOs are particularly restrictive in SADC as they are more product- and process-specific
than the simple value-added criterion that can be used under the COMESA FTA, for
example.
SADC ROOs have also continued to evolve over the past few years. The treatment of
products in some sectors has remained subject to review and negotiation. One of the
most contentious issues has been the differentiated treatment afforded to SACU clothing
imports from four of the Least Developed Countries in SADC. Mozambique, Malawi,
Tanzania and Zambia (MMTZ) have in the past received a derogation to rules of origin
(single transformation) under which certain lines of clothing exports from these countries
to SACU would qualify for preferences under SADC even if they were cut and sewn
(locally) from third country fabric. This derogation from the main rules has, however,
been subject to various restrictions, including quantitative restrictions and administrative
arrangements to ensure compliance as well as time limits. It was last granted at the end
of March 2007 and was set to run through 31 December 2009. Since then, however, the
derogation has not been renewed.

A recent example of the costs associated with meeting ROOs involves SACU
moving to more restrictive rules (double transformation) on selected clothing
imports from Malawi, Mozambique, Tanzania and Zambia following the
expiration of the MMTZ-SACU Market Access Arrangement at the beginning
of 2010. This has resulted in some clothing producers in these countries (eg
Bidserv in Malawi) being no longer able to compete in the regional market.
It has also further distorted investment decisions as some of these firms have
relocated to Botswana, Lesotho, Namibia and Swaziland (the BLNS countries)
as a result of the change to avoid the loss of preferences in supplying the
South African clothing market. For other products where ROOs have been so
contentious (eg wheat flour) or simply not agreed (eg electrical products for
which rules were only finalised in April 2010), preferential trade within the
region has been effectively prohibited (Naumann, 2008). Further costs arise
from the administrative requirements for certificates of origin that can account
for nearly half the value of the duty preference. For example, Shoprite spends
US$5.8 million per year in dealing with red tape (eg filing certificates; obtaining
import permits) to secure US$13.6 million in duty savings under SADC – see
Box 12.2.
COMESA ROOs have also caused problems. While COMESA has a 35% value-added
rule, not all countries have adopted this. Egypt unilaterally imposes a 45% local content
requirement. Until recently, Zambia, Uganda and Malawi did the same. Rules in two
sectors have also proven to be particularly contentious under COMESA. For wheat flour,
the 35% value added rule has generated difficulties for exporters in Egypt and Mauritius
that do not produce wheat grain, but import the raw material from the world market. In
periods of high wheat prices, such as those experienced in 2007, this meant that these
countries were unable to meet the value added requirement. With palm oil, there have
been disputes over ROOs (eg Zambia-Kenya) because of difficulties assigning value
added. These have arisen because a number of products can be produced from the raw
material such as cooking oil, soap and margarine.
Box 12.2 Obstacles facing the utilisation of tariff preferences in SADC – the
case of Shoprite
Shoprite is a South African-based retail and fast food company that operates over 1,200
stores under various names (eg Shoprite, OK Furniture, Checkers, Hungry Lion, Usave,
House Home) in 17 African countries as well as in India. The company accounts for
30% of South African retail trade.
In 2009, over 15% of Shoprite’s revenues came from stores based outside of South
Africa. In securing these a key challenge has been administering compliance with the
rules of origin to qualify for SADC preferences on consignments sent to its stores outside
of SACU. In particular:

Woolworths does not use SADC preferences at all in sending regionally produced
consignments of food and clothing to its franchise stores in non-SACU SADC
markets. Instead it simply pays full MFN tariffs because it currently deems the
process of administering ROO documentation to be too costly. Estimates indicate
that in 2009 it could have benefited from duty savings of US$0.6 million on
US$3.2 million of exports to Mozambique, Tanzania and Zambia – a cost saving
on imports for franchise holders in these markets of up to 19% and a strong
incentive to source more of its products regionally (Charalambides, 2010).
12.3.4 Poorly designed technical regulations and standards limit consumer
choice and hamper trade
Standards regimes in Southern Africa are characterised by an over-reliance on
mandatory inspections and certifications; unique national (rather than regional or
international) standards and testing; overlapping responsibilities for regulation;
and occasional heavy government involvement in all dimensions of the
standards system. These factors create unnecessary barriers to trade, especially
when technical regulations and standards are applied in a discriminatory fashion
against imports.
International good practice is to use technical regulations only to ensure
core public policy objectives such as maintaining safety. Voluntary standards
should be used in all other cases, including indicating quality attributes. But in
several Southern African countries, scarce public resources are being wasted on
developing and enforcing technical regulations that go well beyond issues of
purely public interest. One example is imported shoes in Mauritius where the
Chamber of Commerce has proposed the development of a regulation to govern
their quality to prevent the entry of low-cost Chinese sandals that are perceived to
have a tendency to wear more quickly than domestically produced ones. However,
these are often the only shoes that the poorest people in Mauritius can afford to
buy. One way to overcome such problems would be to use regulatory impact
assessment (RIA) more systematically. However, in most Southern African
countries there are often no such procedures by which technical regulations are
assessed in terms of their consistency with public policy objectives; whether
countries have the capacity to implement them; or their impact on trade and
competitiveness. In particular, no ‘Office of Regulatory Reform’ exists in any
Southern African country to review the justification for both new and existing
• The value of SADC preferences to Shoprite was US$13.6 million in 2009 on US$550
million of exports to the region (implying an average margin of preference of 2.4%).
• The cost of proving eligibility for preferences on this trade was US$5.8 million in
the same year. These costs comprised: 40% for staff to maintain customs data for
shipments; 40% on in-house clearing and forwarding; and, 20% on the maintenance
of a library to demonstrate compliance with rules of origin for suppliers.
Source: Charalambides (2010).

technical regulations. This absence of RIA causes problems and raises costs. For
instance, the environmental levy on plastic bags in South Africa was introduced
to reduce problems associated with litter, but the technical regulation governing
it also affects unrelated issues such as the minimum thickness of the plastic to be
used as well as the size of the text that must be printed on the bags.
Regional efforts to harmonise standards are another way to ensure that national
differences between them do not restrict trade. While these are under way (eg
SADCSTAN),3
application remains lacking. Only Namibia and Swaziland have
adopted all 78 (to date) of the harmonised standards for the region. Participation
in regional standards development has also been uneven. For example, 60–70%
of regional standards developed have been to South African standards although
increasingly other countries are becoming more active in making proposals such
as Mauritius (for agricultural products), Botswana (for construction services)
and Zimbabwe. A key challenge, therefore, is participation of the region. There
are different levels of standardisation competence, skills and knowledge between
countries. The national standards bodies in a few SADC countries (eg Tanzania,
Malawi and Zambia) are also involved with efforts to harmonise standards
in other regional groups, namely COMESA and EAC, which has resulted in
different harmonised standards being agreed between the RTAs. It will be
important for the various regional groups in Southern Africa to coordinate on
standards issues so that there is a consistent approach. Another challenge is
language. Portuguese, French and English are used in the region so if a regional
standard is developed in, say, English then Mozambique will be required
to translate this into Portuguese and then back into English again before the
harmonised standard is agreed. This is a costly process and can cause disputes.
For example, a standard on salt developed by Mozambique was translated into
English by South Africa but when translated back to Portuguese was objected
to by Mozambique on the grounds of it not being the same as the original.
Some regional standards have also been developed without any real sense of
prioritisation and so are unlikely to bring significant increases in regional trade
(eg for frozen peas and dried apricots).
12.3.5 Other NTB restrict opportunities for regional sourcing
A plethora of other barriers such as trade permits, export taxes, import licences
and bans also persist. Shoprite spends $20,000 a week on permits to distribute
meat, milk and vegetables to its stores in Zambia alone. For all countries it
operates in, approximately 100 (single entry) import permits are applied for
every week; this can increase to up to 300 per week in peak periods.As a result of
these and other documentary requirements (eg rules of origin) there can be up to
1,600 documents accompanying each truck that Shoprite sends with a load that
3 All the national standards bodies in SADC participate in the SADC Cooperation in Standardization
(SADCSTAN). Within it a country can propose a regional standard for a good or service that if
accepted by other Member States is adopted as a harmonised SADC standard.

crosses a SADC border. Lack of coordination across government ministries and
regulatory authorities also causes significant delays, particularly in authorising
trade for new products. Another South African retailer took three years to get
permission to export processed beef and pork from South Africa to Zambia.
In SACU, national protection for infant industries is often used to justify import
bans. Namibia has used the provision to protect a pasta manufacturer and broilers
and maintains protection on UHT milk. Botswana has recently limited imports
of specific varieties of tomatoes and UHT milk. Seasonal import restrictions on
maize wheat and flour also ensure that domestic production is consumed first.
For example, Swaziland’s imports of wheat flour were effectively banned for
six months of 2009 since no import permits were issued for half of that year.
Again, the costs associated with these types of NTBs are high and do not appear
to be based on developing globally competitive firms nor creating significant
numbers of sustainable jobs. For example, Erasmus and Flatters (2003) find that
the infant industry protection for Namibian pasta created less than 20 jobs in
the domestic firm that the measure was designed to protect. Its main impact
was to limit consumer choice, since quality imported pasta made using durum
wheat became almost unavailable. Flatters (2010) estimates that protection for
Botswana UHT milk costs consumers there US$16 million per year – equivalent
to US$160,000 per job created where the annual wage of a factory worker is less
than US$1,500. While South Africa is not allowed to protect its industries from
regional competition under the SACU Agreement, it has managed to shield its
mature sectors through imposing anti-dumping duties (most recently on imports
of chicken).
Export restrictions also impose costs and inhibit the development of regional
supply chains. A case in point is small stock exports from Namibia. Since 2004,
the Government has established a scheme ostensibly designed to encourage local
slaughtering. Initially under this scheme quantitative restrictions were imposed
on live sheep exports together with a discretionary permit system. These have
since been replaced by a flexible levy. Under the former arrangement, the
quantity of live sheep allowed to be exported from Namibia was set as a function
of the number of sheep slaughtered domestically, initially set at 1:1 but later on
restricted to 1:6. Under the latest arrangement, quotas for live sheep exports have
now been abolished and replaced with a flexible levy of between 15 and 30%.
The new arrangement is proving to be more restrictive than the old quota system
and has virtually closed the border for the export of any live sheep from Namibia
to South Africa. The impact of these restrictions on the live sheep industry in
both Namibia and SouthAfrica has been highly adverse, particularly on abattoirs
situated in the Northern and Western Cape. Between July 2004 and May 2008,
Namibian exports of live sheep to South Africa decreased by 84%. Total sheep
production in Namibia also fell from one million in 2008 to 800,000 in 2009 as
traditional sheep farmers have switched to alternative activities such as cattle
and game farming. There have also been cases of livestock smuggling from
Namibia to avoid the tax. In SouthAfrica, 975 full-time jobs are at risk due to the
scheme, especially in the bigger abattoirs that focus on slaughtering Namibian

sheep during the low season to better utilise their slaughter capacity (Talijaard
et al, 2009). In addition, Namibian sheep farmers have become almost entirely
dependent on the four Namibian export abattoirs where they were previously able
to sell more sheep to the South African market to receive a higher price (PWC,
2007). Over 90% of the sheep carcasses slaughtered in Namibia are exported
to South Africa, so the small stock export abattoirs in Namibia have essentially
been granted an oligopsony where they do not have to perform efficiently nor
pay a competitive price for their inputs. The same abattoirs were supposed to do
value addition to the primary product for the benefit of domestic industry but
over the past six years have largely failed to do so apart from slaughtering and
skinning sheep locally.
12.4 What is being done at the regional level to combat NTBs?
The implication of the various types of NTBs described above persisting on
Southern African regional trade is that they impose unnecessary costs for
producers that limit trade and raise prices for consumers. Many of these barriers
are simply wasteful and do not serve any real purpose. Import bans and delays
create uncertainty over market access and limit investment. Thick borders and
fragmented regional markets limit possibilities for the development of regional
production chains in which countries can exploit their comparative advantage
in specific tasks and intra-industry trade. Finally, the heavy bureaucratic burden
imposed on all regional trade flows ties up regulatory and customs resources,
limiting their attention on achieving the most pressing public policy objectives
such as promoting effective border management.
Nevertheless, RTAs in SouthernAfrica have made some progress in addressing
NTBs but much more needs to be done.
A key consideration has been those conditions under which non-tariff
measures are needed to achieve legitimate public policy objectives such as health
and safety, versus those in which they are not such that they qualify as non-
tariff barriers and are simply protectionist either by design or intent and impose
unjustified burdens on trade. At the multilateral level, some core principles
have been specified in making this distinction. These are based on ensuring
transparency, non-discrimination, adopting international best practice wherever
possible (eg using harmonised standards for goods and services where these have
been developed) and proportionality to risk. However, NTBs may also include
measures that are applied but not legislated for, such as the very real possibility
in some African countries that a customs official asks to see an import permit
when one is not legally required. NTBs may also include those regulations that
while seeking to achieve legitimate objectives nevertheless impose unnecessary
burdens on trade through inappropriate application. In other words, the barrier
can relate to the administration of the measure as well as the measure itself.
Article 6 of the SADC Trade Protocol calls for the elimination of all existing
forms of NTBs and for Member States to refrain from imposing new ones. While
implementing this Article remains a major challenge, SADC Ministers of Trade

have identified ten categories of NTBs for ‘immediate’ action.4
In some of these
areas there has been progress, but in most barriers still remain. For example, the
2007 Audit on the implementation of the SADC Trade Protocol found that all
Member States were implementing many of the trade facilitation instruments
that had been rolled out by SADC. However, those governing the creation of
a regional transit system, for example, have yet to be implemented (see Table
12.2).
At the country level there have also been some significant improvements. For
example, Mauritius has made strong progress in the automation and streamlining
of its customs clearance and port management processes.
Nevertheless, success in removing NTBs in other areas has been lacking.
Instead, efforts have tended to focus on improving the monitoring and reporting
of NTBs rather than elimination. Monitoring has taken two main forms:
i. Audits on the implementation of the SADC Protocol on Trade have been
undertaken every year since 2007. Their main focus has been on progress
in removing tariffs facing regional trade, as per countries’commitments,
but they also review some NTBs, in particular those relating to trade
facilitation and rules of origin.
ii. A SADC Trade Monitoring and Compliance Mechanism (TMCM) was
established in mid-2008. This has two distinct elements: an online NTB
Monitoring Mechanism (NTBMM) which records reported NTBs by
firms; and the elimination and reduction of barriers (both tariffs and
NTBs) following bilateral negotiation or outcomes from the various
Dispute Settlement Mechanisms.
The publication of NTBs under the auspices of the NTBMM is a major step
forward. However, while the NTBMM is now well established (see Box 12.3),
there are problems with it including misidentification of some of the barriers
reported and, most importantly, very slow progress in resolving the barriers once
they have been notified.
4 These being: (i) cumbersome customs documentation and procedures; (ii) cumbersome import
and export licensing/permits; (iii) import and export quotas; (iv) unnecessary import bans and
prohibitions; (v) import charges not falling within the definition of import duties; (vi) restrictive
single channel marketing; (vii) prohibitive transit charges; (viii) complicated visa requirements; (ix)
pre-shipment inspection; and (x) national food security restrictions.

Table 12.2 Some progress has been made in implementing trade facilitation
instruments in Southern Africa
Instrument Bot Les Mal Mau Moz Nam RSA Swz Tan Zam Zim
WTO Valuation
Agreement
Y Y Y Y Y Y Y Y Y Y Y
HS Coding
System a)
Schedule of
Concessions b)
Migration to
2007
Y
ongoing
Y Y
Y
ongoing
Y Y Y Y N Y Y
SADC
Certificate of
Origin
Regulations on
SADC ROOs
ROOs manual
for customs
Y N Y Y Y Y Y Y Y Y Y
ROOs manual
for traders
N N N N N N N N N N N
SADC SAD SAD 500
SAD
500
N N N
SAD
500
SAD
500
SAD
500
N N N
Voucher for
correction of
SAD
Y Y N N N Y Y Y N N N
Guidelines for
completion of
SADC customs
documentation
Y N N N N N N N N N N
SADC Transit
Regulations
SADC Transit
Documentation
SADC Transit
Customs Bond
Guarantee
SADC Integrity
Plan to fight
corruption
Y Y Y Y Y N Y Y Y Y Y
MOU for
SADC customs
administrations
Y Y Y N Y N Y Y Y Y Y
Conformity
assessment
Source: USAID (2009).

An example of the former is that the NTBMM once reported South Africa as
maintaining a single marketing channel for maize, whereas this system has long
been abolished since the 1990s. Another example of misidentification concerns
Angola, where weak ‘postal services’ were cited as an NTB.
Regarding the latter, there is presently no mechanism for ensuring that
countries follow a process of either justifying their NTBs or agreeing to remove
or reform them once a barrier has been notified. The absence of a clearly defined
enforcement mechanism with strict time limits for action and sanctions for
non-compliance means each country is effectively responsible for voluntarily
removing or reforming their NTBs. This ‘moral suasion’ approach to removing
NTBs in Southern Africa has largely failed to yield results, in contrast to more
legally binding mechanisms used elsewhere (eg in the European Union) with
sanctions. Consequently, within the SADC Region, of the eight complaints
Box 12.3 The Tripartite NTB Monitoring Mechanism
The NTBMM is shared between SADC, COMESA and EAC and is a web-based ‘post
box’ where the private sector can report complaints against NTBs to regional trade in
Southern and Eastern Africa. So far under this 335 complaints of NTBs have been made
against barriers originating in 20 countries. The greatest number of complaints have been
made by Namibia (66), followed by South Africa (46), Zimbabwe (39) and Malawi (30).
The three most cited countries for imposing NTBs are South Africa (40 cases), Namibia
(36) and Malawi (33).
An assessment of the types of barriers cited in the NTBMM is set out in Table 12.3. Trade-
related administrative barriers are reported most frequently by firms as an impediment to
regional trade, followed by import licensing.
Table 12.3 NTBs cited in the NTB Monitoring Mechanism
Barrier No. of complaints
Trade-related administrative NTBs 74
Export and import licences 39
Transit issues 36
Technical barriers to trade 32
SPS measures 28
Rules of origin 26
Clearance procedures 24
Quotas 19
Payments 21
Customs documentation 17
Pre-shipment inspection 8
Customs valuation 6
Immigration requirements for cross-border traders 4
Safeguards 1
Source: Charalambides (2010).

raised to date against South Africa (forty if complaints from EAC and COMESA
countries are included), only four have been resolved. For complaints against
other SADC countries there has only been one resolution of an NTB reported
via the NTBMM.
12.5 Priorities for NTB removal in the Southern African market and
implementing reform
The commitment of Southern African countries to remove NTBs has, to date,
focused on identifying and monitoring them. Raising awareness and improving
transparency are necessary steps but it is becoming increasingly apparent that
they are not sufficient due to the lack of progress in removing these barriers that,
as we have shown, remain extremely costly. It is therefore important to consider
how these remaining NTBs should be dealt with and which among them should
be prioritised for reform.
First, one of the biggest issues for regional trade integration in SouthernAfrica,
especially for manufactures and agro-processed products is undoubtedly ROOs.
The issue has gained particular prominence in light of the planned (almost
Africa-wide) Tripartite FTA where one set of rules for all countries will have
to be agreed. Harmonisation of the different rules among the regional groups
will not be possible because process requirements, employed for example under
SADC, cannot be easily harmonised with the value addition criteria under, for
example, COMESA. So a new set of ROOs will need to be agreed, either based
on one of the existing arrangements or completely redesigned. Characteristics
of ROOs that would encourage the development of new export industries would
include:
• Providing exporters with a choice as to which rule (defined simply and
transparently) they apply, for example either a change in tariff heading
test (ideally at a disaggregated product level) or a reasonable value-
added rule (20%).
• Eliminating process-specific ROOs which set out how a product is to be
made for originating status to be conferred.
• Removing the requirements for certificates of origin for products with
nuisance tariffs, for example those with preference margins below three
percentage points.
• Enforcing these simplified rules more consistently and effectively at
customs to mitigate any concerns over leakage or trade deflection.
• Greater use of risk assessment, especially for large, trusted regional
traderswhoshouldnotrequireacertificateoforiginforeachconsignment
but, instead, should be able to submit these electronically per batch.

Secondly, resolving the other types of NTBs, both existing and curtailing the
development of new ones, is also vital as these are critically restricting trade
in the region particularly for primary agricultural commodities. Among these
the most serious barriers are import bans, quotas, permits and licensing that are
often implemented by countries with little or no consultation with their trading
partners.
Thirdly, while tariffs have been largely reduced across the region, NTBs arise
in those sectors where tariff peaks persist. One advantage with addressing any
remaining tariffs is that tariff reform can often be dealt with by a ‘stroke of
the pen’ approach, as opposed to some of the other barriers where reform will
be complex, perhaps more costly and certainly more involved. High tariffs are
especially restrictive because concerns of leakage from third countries can
create the need for NTBs (eg ROOs) as well as affecting trade in all sectors
as border checks are intensified to check for transshipments of these products.
Lower, more uniform external tariffs would significantly reduce the need for
many of the barriers which persist on regional trade in Southern Africa as would
the development of policies that directly address the difficulties that protected
sectors may be facing such as assisting labour in these industries to retrain in
tasks where employment opportunities are much better.
Fourthly,reducingbureaucraticrequirements,streamliningbordermanagement
procedures and implementing trade facilitation measures, including one-stop
border posts (OSBPs), have significant potential to lower border crossing times
and reduce transport costs, at least along the main corridors in Southern Africa.
There is also increasing political willingness among governments for this type
of reform to go ahead sooner rather than later. For example, the South African
Government has recently identified OSBPs as one focus area it wishes to develop
for regional integration in the next 12 months. Namibia has already put forward
five proposals for OSBPs that link it to South Africa and one with Botswana.
However, revenue concerns among the smaller SACU countries risk impeding
reform. Overcoming this challenge will require the development of better ways
to capture trade flows across SACU borders than those currently employed as
well as an open discussion about alternatives to the current revenue sharing
arrangement that might be more effective and sustainable.
In undertaking these reforms, how can existing structures for NTB removal in
Southern Africa be strengthened?
Most NTBs can be addressed nationally. Few reforms need wait for regional
agreement and much can be done both unilaterally and bilaterally to increase
regional trade in both goods and services. Most of what is needed to improve
any country’s international competitiveness and business climate can therefore
be done at home. In particular, unilateral action focused on the needs of firms
and services providers would do much to take countries further along the road to
deeper regional integration by lowering the costs they face as they trade across
borders. For example, using regulatory impact assessment more systematically
would ensure that each country’s regulations are consistent with the public policy
objectives they are designed to meet; are minimally trade-distorting; and that

those countries have the capacity to implement them. Unilateral action also does
not require all countries to move at the same pace. Countries that wish to reform
deeper or sooner can do so and in the process provide valuable experience to
other countries on what works and what does not, including through the use of
regional knowledge platforms to share good practice in regulatory reform for
goods and services.
Nevertheless, regional interventions can also be important through the
realisation of scale economies in the development of more efficient and effective
regulatory frameworks.There are clear benefits from closer regional coordination
in regulation and standards, particularly where national agencies face technical
skill shortages or capacity constraints. For the smaller SouthernAfrican countries
it may be better to seek closer collaboration within SADC or COMESA by
relying on fewer regulatory agencies and accredited regional providers of
testing, inspection and certification for both goods and services instead of many
national ones that each require recurrent budgets and often have trouble with
staffing. Continuing to develop appropriate and prioritised regional standards
(eg through SADCSTAN) would also reduce the costs of market participants
operating across borders and reduce the scope of capture by national private
sector interests. Unlike other regions in Africa, Southern Africa has the distinct
advantage that world class testing bodies and accreditation services are already
available in South Africa. Regional integration could leverage the capacity of
these institutions to support exports and capacity in the less advanced countries
as well as eliminating the need for double testing or certification of regional trade.
Alternatively, countries could consider steps towards implementing a regional
framework for mutual recognition so that conformity assessment procedures,
for example, are recognised in other countries or qualifications and licences for
service providers issued by one country are accepted by all. Otherwise, suppliers
are obliged to perform additional or repeated tests of their goods and services in
regional markets or to invite foreign inspectors, thereby increasing trade costs.
The most significant regional initiative to remove NTBs in Southern Africa is
the NTBMM. However, as discussed earlier the main achievements to date of
this have been monitoring and reporting of barriers rather than actually removing
them. In order to make the NTBMM more effective the following steps should
be taken:
• Proper identification of the barriers, whether they really exist and
whether the measures being notified are unnecessarily trade restricting
given the public policy objectives they are trying to achieve.
• Prioritising NTBs for removal, in particular by linking those that have
been notified in the NTBMM to core barriers identified by SADC for
‘immediate removal’ (eg unnecessary import bans and prohibitions) as
well as those types identified in this paper (eg ROOs).

• Developing a pilot between a few countries (eg South Africa-Mauritius)
to ensure the mechanism is well adapted to the needs of its users, as well
as providing valuable demonstration effects to the rest of the region on
the benefits of removing NTBs.
Finally, any mechanism aimed at removing NTBs will be of little use if
new barriers can be freely erected to take their place. New NTBs should be
disciplined through the greater use of regulatory impact assessment. In addition
there should be stricter notification requirements such that information on any
planned measure is provided to regional trade partners in good time before it
is implemented. Another significant problem is that many NTBs in the region
are not legislated for, so the existing monitoring mechanism must also be
strengthened to deal with new barriers of this type that cannot be picked up
through regulatory impact assessments.
References
Cargo Info Africa (2010), Durban Efficiency Breakdown Forces Switch to Maputo:
Creating Competitive Cost Chains.
Carrere, C and De Melo, J (2009a), ‘Notes on Detecting the Effects of Non Tariff
Measures’, CERDI Working Document E 2009.32.
Carrere, C and De Melo, J (2009b), ‘Non Tariff Measures: What Do We Know?
What Should Be Done?’, CERDI Working Document E 2009.33.
Charalambides, N (2010), Addressing NTBs in Regional Goods Trade in Southern
African countries, Imani Development.
Erasmus, H and Flatters, F (2003), ‘Rent-seeking in SADC Trade Liberalization:
Rules of Origin and Other Barriers to Trade in Wheat Products’, Research Report
for World Bank Netherlands Partnership Program (BNPP).
Flatters, F (2010), ‘Implementing the SADC FTA: Where Are We? What Next?’,
Technical Report for USAID.
Jitsing, A and Stern, M (2008), ‘VAT Practices within SACU and Possibilities for
Harmonisation’, Southern African Regional Integration Project, World Bank.
Naumann, E (2008), Intra-SADC and SADC-EU Rules of Origin – Reflections on
Recent Developments and Prospects for Change, TRALAC.
PWC (2007), Evaluation of the Implementation of the Small Stock Marketing
Scheme in Relation to the Namibian Government’s Value Addition Goals and
Objectives, Windhoek.
RTFP (Regional Trade Facilitation Programme) (2007), ‘Inventory of Regional Non-
tariff Barriers: Synthesis Report’, RTFP, Pretoria.
RTFP (Regional Trade Facilitation Programme) (2009), ‘Non-tariff Barrier Impact
Study for COMESA Region’, RTFP, Pretoria.
Talijaard, P, Alemu, Z, Jooste, A, Jordaan, H and Botha, L (2009), The Impact of
Namibian Small Stock Marketing Scheme on South Africa, National Agricultural
Marketing Council, South Africa.

USAID (United States Agency for International Development) (2009), ‘Technical
Report: 2009 Audit on the Implementation of the SADC Protocol on Trade’,
AECOM International Development, Gaborone.

Appendix: Mapping of NTBs reported by SADC countries to products and
potential trade affected
NTB Products affected
Value of intra-SADC trade,
2008 US$ million
(% of total non-fuel intra-SADC trade)
Import bans Wheat 83.71
Beer 46.81
Poultry 43.79
Wheat flour 15.40
Meat products 69.22
Maize 228.45
Milk (sterilised, UHT) 36.59
Total 523.97 (3.2%)
Import quotas Maize 228.45
Wheat 83.71
Maize meal 24.24
Flour 16.46
Cement 143.71
Sugar 93.86
Salt 46.43
Poultry 43.79
Eggs 9.26
Tobacco 150.90
Fruit and vegetables 172.34
Total 1013.153 (6.1%)
Import levies Milk (sterilised, UHT) 36.59
Pasta 15.34
Sorghum 19.62
Wheat 83.71
Chicken and eggs 16.54
Dairy 144.41
Pork and poultry 54.87
Beer 46.81
Total 417.89 (2.5%)
Preferences denied Salt 46.43
Fishmeal 0.02
Pasta 15.34
Total 61.79 (0.4%)
Import permits and licensing Milk (sterilised, UHT) 36.59
Bread 35.10
Eggs 9.26
Sugar 93.86
Fruit and vegetables 172.34
Livestock 26.58
Liquor 35.06
Cooking oils 255.60

2008 US$ million
Maize 228.45
Oysters 0.01
Total 892.85 (5.4%)
Single marketing channels Wheat 83.71
Meat 69.22
Dairy 144.41
Maize 228.45
Tea and tobacco 183.81
Sugar 170.92
Total 880.52 (5.3%)
Rules of origin Textiles and clothing 306.60
Semi-trailers 37.47
Palm oil 28.42
Soap 46.96
Cake decorations 0.04
Rice 53.65
Curry powder 0.01
Wheat flour 15.40
Total 488.55 (3.0%)
Export taxes Dried beans 15.28
Live animals, hides and
skins
26.58
Sugar 93.86
Tobacco 150.90
Maize 228.45
Meat 69.22
Wood 194.90
Coffee 12.73
Total 791.92 (4.8%)
Standards/SPS/TBT Milk 36.59
Meat 69.22
Canned tuna 1.46
Beer 46.81
Honey 0.05
Maize bran 3.91
Cotton cake 71.62
Poultry 43.79
Batteries 21.27
Sugar 93.86
Coffee 12.73
Ostriches 3.85
Total 405.16 ((2.5%)
Customs related Wine 39.44
Electronic equipment 2.4

2008 US$ million
Copper concentrate 491.40
Salt 46.43
Cosmetics 41.08
Medicines 232.86
Total 853.61 (5.2%)
Note: Volumes of trade do not reflect those actually affected by NTBs since not all countries in SADC
impose all these NTBs. It should simply be interpreted as an indication of how important to regional
trade the products affected by these NTBs are.
Sources: Authors’ calculations based on complaints made by SADC countries to NTB Monitoring
Mechanism and UN Comtrade.

13
Streamlining NTMs: How Mexico Did It
Ali Haddou
13.1 Introduction
Regulatory reform has been a fundamental element of Mexico’s transition from
a closed to an open market-based economy. The process started in the mid-
1980s, when general frustration with macroeconomic instability and years of
stagnant growth and inflation in the wake of the 1982 debt crisis led to a change
in economic policy, based on three interdependent pillars: trade liberalisation,
privatisation and regulatory reform. The change entailed a complete recasting of
Mexico’s system of regulatory governance.
The progressive trade liberalisation began with a significant unilateral
reduction of tariffs and import licences, then continued with Mexico’s accession
to the GATT in 1986 and the negotiation of free trade agreements (FTAs) with
the United States and Canada (1994), the European Union (2000), and Japan
(2005), among others. The process went hand in hand with the establishment
of new regulatory institutions, tools and processes in order to make the most of
opportunities created by the FTAs, and the promotion of the structural reforms
– in recently privatised network industries and other areas – needed to boost
the productivity and competitiveness of the economy as a whole (Salas, 2005;
Córdova and Haddou, 2008).
Regulatory reform and trade liberalisation play complementary roles in
that they enhance competition to give consumers and businesses competitive
and non-discriminatory access to a wider selection of inputs and final goods
in domestic markets; they facilitate international trade to enhance foreign
market access for Mexican goods and the entry of capital goods, key inputs
and investment on more favourable terms, which expands productive capacity,
generates employment and favours the diffusion of new technologies; and they
contribute to the development of efficient regulation that reduces the costs of
market entry and operation in general.
But each FTA brings with it new rules and administrative processes designed
to ensure that goods benefitting from preferential treatment satisfy geographical
origin requirements. The complexity and asymmetry of the overall tariff structure
that results can produce unwanted consequences, such as:

• Inefficient trade diversion. The preferential access for FTA partner
goods and services may result in domestic consumers and producers
paying more for goods and services than they would otherwise, as a
result of the higher tariffs applicable to non-FTA products.
• Transshipment and fraud to bypass rules of origin requirements. This
risk of rules of origin bypass and the multiplicity of procedures to verify
the origin of goods forces customs authorities to invest significant
resources in the classification and inspection of goods that might be
better applied to other uses. The increased transaction costs raise the
prices of all goods, not just those from non-FTA partners.
• Distortion of incentives in production chains. Final products often wind
up receiving different tariff rate ‘protection’ than the one resulting from
the summation of tariffs on inputs along the production chain. When the
equivalent tariff rate on inputs is greater than the nominal rate paid for
the final product, the incentives to produce goods locally are reduced
and the results are precisely the contrary of those intended.
When a significant proportion of trade is with preferential trading partners – well
over 80% in the case of Mexico – the rationality of maintaining such a complex
system of trade regulation is naturally put into question (Zabludovsky, 2005).
The uniform application of regulatory reform principles (transparency, public
consultation, non-discrimination and cost-benefit analysis) to trade-related
measures can help.
This paper presents an overview of the parallel chronological evolution
of trade and regulatory reform policy in Mexico over the last 20 years, with
particular emphasis on the institutional and procedural aspects of the review of
trade-related measures (NTMs) and the role of regulatory impact analysis (RIA)
in this process.
13.2 Unilateral liberalisation and the road to NAFTA: 1989–1994
After Mexico had slashed its tariffs unilaterally in the mid-1980s and joined
GATT in 1986, it soon became clear to the incoming administration in 1988 that to
integrate Mexico fully into the world economy and to effectively tap into foreign
savings, it would need to develop its export markets, accelerate economic reform
and form part of a regional trading bloc. Amid preparations for the negotiation
of a trade agreement with the United States – and subsequently Canada – an
Economic Deregulation Unit (Unidad de Desregulación Económica, or UDE)
was created within the trade ministry in 1989 to review the national economic
regulatory framework in order to promote competition, economic development
and job creation.
Theinitialroundofunilateraltariffreductionsinthemid-1980squicklyrevealed
the inefficiencies of state-owned enterprises that dominated the economy, poor

Streamlining NTMs: How Mexico Did It 249
infrastructure and high entry barriers that resulted in an underperforming private
sector.1
The UDE’s response to this was a reform strategy that would make
Mexico’s regulatory framework compatible with a modern free-market economy,
and reduce operating costs across the board to promote competitiveness. The
reforms were quickly implemented. Between 1989 and 1992, for example,
airlines and telephony were privatised, road haulage was deregulated and private
investment – both foreign and domestic – was allowed in road construction,
ports and electricity generation.
The UDE’s actions implied reducing the discretion and power of many
regulatory authorities, and ran counter to a Mexican bureaucratic culture still
grappling with its vestiges of corruption and opaqueness. The only way to
overcome resistance forcefully was to secure committed political support at
the highest level of government. The shift towards actively seeking to improve
the business environment and the rule of law required active support from the
President to discipline the executive branch and Congress (where the incumbent
party still held a strong majority) and allow a small group of economists and
lawyers to design and push for the implementation of major reforms. The
reforms were carried out in an opportunistic manner, establishing a clear priority
for projects with high economic returns and low political costs (Salas, 2004).
On the trade front, while the private sector may have expressed some
reservations regarding the scope and speed of tariff reductions, and the even
greater changes that would result from NAFTA, top business representatives were
given the chance to participate closely in the negotiations. A consultative council
composed of major business organisations with an interest in foreign trade called
COECE (Coordinadora de Organismos Empresariales de Comercio Exterior)
was set up, and given the opportunity to comment on the nature of commitments
before they were agreed on by the negotiating teams (Ortiz Mena, 2006).2
In the
end, the prospect of having the government return the operation of the economy
to private hands was a big enough incentive to participate constructively in the
negotiation process (Elizondo Mayer-Serra, 2001).
The negotiators also needed to dispel fears – expressed on both sides of the
border – of potential opacity and arbitrariness in regulatory rules affecting
foreign trade. COECE was concerned about the openness of US government
procurement procedures, protectionism in sensitive goods, and agricultural
subsidies. Mexico, with its long history of poorly designed and inconsistently
applied regulations, needed to show that its investment environment would be
stable and predictable, and that NTMs would not be used as a way to get around
the treaty’s obligations in sensitive sectors. The UDE stepped in dynamically in
1991–92 to design and promote four modern laws that have been instrumental
1 Average MFN tariffs were lowered from over 25% in 1982 to approximately 10% in 1987 (source:
UNCTAD-TRAINS Database).
2 During the NAFTA negotiations, COECE set up 19 different working groups that frequently met
with the ministry of trade’s negotiation team, and COECE was always physically present during the
negotiations themselves to allow for real time consultation on delicate issues. This created a very
strong sense of buy-in and cemented unwavering business sector support for the process.

in the process of guaranteeing effective market access, transparency and legal
certainty: the industrial property law, the foreign trade law (specifying, among
others, general tariff, NTM, licensing, quota, safeguards and antidumping
regulations), the economic competition law (which introduced modern antitrust
legislation for the first time in Mexico), and the standards and metrology law
(which regulates the establishment of technical rules and standards).
While UDE was set up as a specialised task force to dismantle regulatory
bottlenecks to make the most of a more open trade regime, the foreign trade
law (Ley de Comercio Exterior, or LCE) established a foreign trade commission
(Comisión de Comercio Exterior, or Cocex) to review and to prepare opinions
on all tariffs, safeguards, antidumping duties, rules of origin requirements, and
technical standards relevant to foreign trade. The Cocex, which still exists today,
is a commission without a permanent structure, comprised of representatives
from ministries that most often generate NTMs, the foreign affairs ministry, the
Bank of Mexico, and the Federal Competition Commission. Cocex was only
given the power to issue non-binding opinions (that weren’t made public at the
time), but it was significant in that it was the first intragovernmental process
designed specifically to promote uniformity in domestic trade-related measures
and lessen the probability of regulatory capture. Thus, the tandem of UDE and
Cocex would tackle existing regulatory bottlenecks and new NTM proposals to
avoid entry barriers and unnecessary restrictions on trade.
The standards and metrology law (Ley Federal sobre Metrología y
Normalización, or LFMN) created a more open process for the development of
new mandatory technical rules (Normas Oficiales Mexicanas, or NOMs). The
law required the publication of proposed technical rules, and the presentation of
cost-benefit analyses to consultative committees composed of public and private
sector representatives for review and approval. This was the first time that
notice and comment, as well as cost benefit analysis was formally introduced
into the Mexican legal system. It completely eliminated the possibility of
changing technical rules from one day to the next without giving affected parties
an opportunity to comment on the nature and effects of the rules. The process
has been improved over the years, but the NOM process was the seed of the
regulatory impact assessment process that is in place today.
The NAFTA text eventually incorporated disciplines relating to non-tariff
barriers, competition policy, intellectual property protection, foreign investment,
and protection for aggrieved investors or exporters via novel dispute settlement
procedures. These issues were not typical of trade negotiations at the time, and
show to what extent the issues were important to Canada, the US and Mexico
(and for future multilateral trade negotiations as well). The NAFTA, more than
any other trade agreement, generated effective market access by stimulating
mechanisms to promote efficient market structure, and implementing open
processes for the development of rules that could affect any of the parties.
Mexico gained the added benefit of anchoring its broad economic modernisation
programme to a trilateral treaty that would make backtracking very difficult
(Tornell and Esquivel, 1995; Kose et al, 2004; Córdova et al, 2009).

13.3 Economic crisis and the systematisation of regulatory reform:
1994–2000
The euphoria generated by NAFTA was short-lived. In 1995, Mexico suffered
arguably the worst economic crisis of its modern history. The private sector
clamoured for government action that would allow for a rapid recovery and a
level playing field for companies facing increased competition in the domestic
market as a result of NAFTA entry into force in 1994. With financing completely
dried up and domestic consumption anaemic, the government offered a new
regulatory reform programme as one of the cornerstones of its industrial policy.
In the face of strict fiscal restrictions the new government could hardly accede
to the private sector’s calls for subsidies, and NAFTA eliminated the possibility
of increased tariff protection, so regulatory reform emerged as an effective – and
fiscally inexpensive – business facilitation measure that would complement the
devaluation of the peso in helping Mexican companies become internationally
competitive.3
President Zedillo signed the Agreement for the Deregulation of Economic
Activity (Acuerdo para la Desregulación de la Actividad Económica, or ADAE)
in 1995 to implement the new regulatory reform strategy. Its three main elements
were the obligation to catalogue and review all formalities that applied to
business; the requirement that all regulatory proposals and their corresponding
justifications be sent to the UDE for review; and the creation of an Economic
Deregulation Council (Consejo para la Desregulación Económica, or CDE) to
assist in the identification and review of relevant regulatory reform measures,
to supervise implementation by the ministries, and to hold the UDE to account
with respect to its mandate.4
The ADAE completely changed the nature of the
regulatory reform programme from one of opportunistic ad hoc deregulation to
a systematic process of regulatory improvement.
First, it called for a full review of existing formalities prior to including them
in a central registry of all business formalities, with a guillotine rule, by which
any formality not contained in the registry was deemed automatically void was
also applied. This applied not only to formalities relating to opening a business,
but to any formality that could be required of business by any branch of the
federal government, from patent application forms to the process for requesting
the start of an antidumping investigation.
Second, all regulatory proposals having business impacts – however small –
now had to be sent to the UDE for review, along with a technical justification
3 In contrast, Mexico raised tariffs by 100% after the 1982 financial crisis, and American exports to
Mexico fell by half and did not recover for seven years. In 1995, Mexico continued to implement its
NAFTA obligations even as it raised tariffs on imports from other countries. As a result, American
exports recovered in 18 months and were up nearly 37% by the end of 1996 relative to pre-NAFTA
levels, even though Mexican consumption was down 3.3% (source: Office of the United States Trade
Representative).
4 The CDE was presided by the minister of commerce and membership included the ministers of
finance and labour, the comptroller general, and representatives from business, academia and
organised labour.

of the measure. In addition, an amendment to the administrative procedures law
(Ley Federal de Procedimiento Administrativo, or LFPA) was passed in 1996
to formally require regulatory impact analyses (RIAs) in these cases. Although
there was no obligation to make the proposals public, the systematic review
by the UDE completely changed the dynamics of federal regulatory decision
making. This meant that most NTMs were now subject to full UDE analysis and
review before being submitted to Cocex for its opinion. The UDE’s reviews were
naturally more thorough than Cocex’s opinions, because it was staffed by experts
completely independent of the ministries proposing the rules. Nevertheless, this
process did not apply for fiscal regulation, an important loophole that still exists
today and that exempts tariff decisions, safeguards, antidumping duties and
customs procedures from the general regulatory reform programme.
Third, the ADAE and the CDE significantly strengthened the UDE’s position
vis-à-vis the rest of the administration, especially with regard to its role in
reviewing proposed regulations. The RIA requirement shifted the burden of
proof to the ministries proposing regulation, which forced them to make their
motivations more transparent. RIAs had to include the general public objectives
of their proposals, an explanation of how the alternative proposed would be the
least restrictive option, an estimation of costs and benefits, and a description
of the resources that would be necessary for effective implementation. The
UDE would then review this information and present recommendations to the
ministry. If the recommendations were not accepted, the issue would be raised
to the CDE to pressure the ministry into compliance.5
The fact that the UDE
itself had to report to the CDE on the effectiveness of the review process at least
quarterly meant that there was also quite a bit of pressure on it to deliver results.
Alongside the implementation of the RIA process, the UDE remained very
active in pushing for major reforms of its own initiative. This included the
streamlining of judicial procedures to reduce the number and the duration
of commercial trials in the Federal District, the elimination of regulatory
restrictions on setting up loan guarantees to ease access to credit, amending the
foreign investment law to allow greater participation of neutral foreign capital in
restricted sectors, replacing the traditional defined benefit pension scheme with
the creation of a system of individual retirement accounts, facilitating mining
concession schemes, and the adoption of a new insolvency law. In 1996, the
UDE was successful in pushing for the elimination of mandatory affiliation
to business associations, and for the elimination of a wide range of sanitary
licences except in areas where the cost of imposing such licences was justified
by the risks to public health. In 1997, the UDE sponsored amendments to the
LFMN that allowed companies to demonstrate compliance with technology-
specific technical rules via alternative means (regulating the ends rather than the
means), instituted a fast-track procedure for the elimination of rules that do not
5 A key element of this process was that council members were all ministers or their equivalents, and
it is obviously very costly for a minister in terms of time and reputation to appear recurrently before
the CDE in order to defend questionable regulatory practices.

impose stricter or technical specifications, and mandated a five-year review and
sunset clause for all technical rules.6
The UDE used the sense of urgency generated by the economic crisis to its
advantage, pushing difficult reforms that might otherwise have failed to generate
enough political support. Mexico’s response to the crisis was swift and effective,
driven in large part by its increased export capacity and the support given by
the US Treasury quickly putting together a package of over $50 billion in loan
guarantees, neither of which would have been possible without NAFTA (Rubin,
2003).7
Nevertheless, there was a strong public backlash as a result of the crisis, fuelled
by criticisms of lack of transparency and crony capitalism in the privatisation of
state enterprises and other aspects of economic policy. Support for structural
reform began to wane significantly in 1997, the year in which the PRI (Partido
Revolucionario Institucional) lost its majority in Congress for the first time in
the 20th century.
Again, the UDE found an opportunity to capitalise on an apparently adverse
environment, by shifting the focus away from the deregulation and reform
of specific economic sectors to a regulatory agenda of transparency, public
consultation and technical analysis for the federal government as a whole.
The UDE also realised that the conditions that allowed it to have the success
it did during the 1990s, particularly the nearly unconditional support from the
President and the trade minister, would probably not last forever. It became
keenly aware of the fragility of the regulatory reform process that it had pushed
for over a decade. There were a few instances of flagrant non-compliance with
the regulatory review process when political pressures overwhelmed the UDE’s
opposition, and there was mounting and recurrent pressure to tone down or not
publicise its opinions, or to expedite the review process. An example of the high
risks involved came with the change in legal status of the Federal District, where
an incoming administration in 1998 essentially scrapped the reforms that the
UDE had promoted in the deregulation of road freight and parking rules.
With the support of the CDE, the UDE designed reforms to the LFPA (2000)
to fully institutionalise the regulatory reform programme and set its legal basis
in law. The reforms created the Federal Regulatory Improvement Commission
(Comisión Federal de Mejora Regualtoria, or Cofemer), giving it full technical
autonomy to design and implement regulatory reform policy in Mexico. The
reforms gave Cofemer full review jurisdiction over NOMs, government social
security regulation, and also public procurement, that had been important
loopholes in the past, although fiscal regulations affecting foreign trade were
still outside its jurisdiction and could only be reviewed by Cocex. In line with
6 NOMs are automatically repealed if the five-year review is not carried out.
7 Following Mexico’s 1982 financial crisis, Mexican output drifted down for nearly two years before
rising again and did not recover to pre-crisis levels for five years. Although Mexican economic output
dropped more quickly in 1995, it also rebounded more quickly, reaching pre-crisis peaks by the end
of 1996. Similarly, following the 1982 crisis, it took Mexico seven years to return to international
capital markets, while in 1995, it took seven months.

the UDE’s practice, Cofemer’s powers included the review of regulations, the
diagnosis of their application, and the ability to propose reforms in specific areas
or economic sectors directly to the President. The programme consisted in four
main actions: the deregulation of formalities, the review of regulatory proposals
and their RIAs, the drafting of studies and reform proposals upon its own
initiative, and the support for regulatory reform efforts at the state and municipal
level. The reforms maintained a Federal Council for Regulatory Improvement
(Consejo Federal para la Mejora Regulatoria, or CFMR), similar to the CDE,
but with the addition of consumer-oriented bodies such as the consumer
protection agency (Procuraduría Federal del Consumidor, or Profeco) and the
Federal Competition Commission (Comisión Federal de Competencia, or CFC).
RIA was further strengthened by the 2000 reforms, by setting the regulatory
review process in law for all regulations, including full transparency of proposals
and RIAs and a minimum period of public consultation of 30 days. As of the
date of those amendments, no regulation could be published in the official
government gazette without having passed through the open regulatory reform
process, with severe penalties for non-compliance.8 The amendments were
unanimously approved by Congress, in spite of some acrimony between political
parties, showing clearly that there was still appetite for ‘horizontal’ reforms that
enhanced predictability and transparency, and that kept the executive branch
regulatory machinery in check. This profoundly changed the way of regulating
in Mexico.
8 Up to dismissal and a one-year suspension from public service.

Box 13.1 The Mexican Regulatory Impact Assessment process
Institutional design
Cofemer’s RIA programme was consolidated and detailed in reforms to theAdministrative
Procedures Act in 2000, which were greatly influenced by the UDE’s five years of
experience in the review of regulations and international best practices. The main aspects
of the programme were:
• The Presidential appointment of Cofemer’s head and the granting of technical
autonomy to the institution.
• The creation of the Federal Council for Regulatory Improvement as a means of
ensuring Cofemer’s accountability and giving the business, labour and academic
sectors an important role in the direction of the reform programme’s work programme.
• Requiring each ministry and regulatory agency to name a vice-minister in charge of
coordinating in-house regulatory reform efforts, and of submitting two-year work
programmes to Cofemer for review and public comment.
• Implementing a detailed RIA review process and mandatory minimum 30-day period
of public comment for all regulations. Cofemer can question and require more
detailed analysis of RIAs within ten working days of submission, and has 30 working
days to issue its opinion on the proposal itself. Ministries and agencies must publicly
respond to all Cofemer comments and suggestions.
Guidance
To enhance transparency and to provide guidance on how to conduct RIAs, Cofemer
developed an electronic platform that allowed for the centralisation of communication
between Cofemer and the vice-ministry in charge of regulatory reform in each institution,
easy access to technical reference materials, and prompt publication of relevant
consultation documents. The online RIA template includes sections on:
• The description of the regulatory action, identifying the nature of the problem and
risks that each proposal seeks to address.
• Legal analysis, including a clear review of powers, and compatibility with the existing
regulatory framework.
• The alternatives considered (both regulatory and non-regulatory), and an explanation
of how the proposal is considered to be the least intrusive option to attain the stated
goals.
• The implementation strategy and the consideration of resources necessary to ensure
proper compliance and supervision.
• The quantifiable and non-quantifiable costs and benefits of the proposal. In addition
to requiring explicit consideration of effects on market competition, domestic and
international trade, small and medium size enterprises, and consumer access to
goods and services, all relevant capital, operation and transaction costs, and effects
on health, safety and the environment must be itemised.

Training
The quality and usefulness of RIA depends crucially on the capacities of the staff entrusted
with it, on the quality of Cofemer’s review, and on its integration into the decision-making
process of each ministry or regulatory agency. Cofemer has invested heavily in training
programmes for officials involved in the process. From 2000 to 2006, for example, more
than 50 training sessions were set up for well over 2,000 government staff. These included
guidance on the content and timing of RIA and the development of case studies and
hypothetical examples. While the hiring of external consultants was necessary for some
very high impact regulatory decisions, for the most part Cofemer encourages analysis to
be carried out in-house in parallel to the development of the regulatory proposal itself.
This allows for key RIA decision-making criteria to be considered and internalised when
it is easiest to incorporate them into the regulatory process, rather than at the end as a
mere formality that is undertaken as a mere ex post justification of decisions already
taken. The fact that Cofemer used to evaluate and publish the relative quality of different
ministries’ RIAs was also useful in pushing for greater quality in RIA development.
Cofemer’s review
Cofemer’s review of RIA is a crucial part of this process. It has ten working days to
comment on and to request more detailed or precise analysis of specific issues.
Regulators must then incorporate recommendations or respond to comments before
Cofemer can issue an opinion on the regulatory proposal itself within 30 working days
of the submission of the final version of the RIA. It is at this stage that Cofemer can first
and most effectively challenge key aspects of a regulatory proposal. If a questionable
regulatory strategy is insufficiently justified, Cofemer can ask for more detailed
analysis of the motivations for pursuing it and of the potential effects it might have.
In fact, Cofemer often sits informally with the proponents of the regulation to discuss
alternative regulatory scenarios and their likely outcomes in order to eventually arrive at
a more efficient and precompetitive regulation solution to the problems identified. In this
sense, the technical quality of the analysis of potential effects is not as important as the
identification of the key decision points. Prospective analysis of potential effects is an
inherently difficult task and it is often much more productive to focus on key regulatory
actions and to evaluate if they conform to basic regulatory quality criteria (such as
adoption of least restrictive regulatory options and international best practice, avoidance
of unnecessary barriers to entry, minimisation of paperwork burden and supervision and
inspection costs, etc). When analysing RIA, the identification of flawed reasoning and
exaggeration of potential benefits of poor regulatory strategies is one of the most effective
ways of exposing weaknesses in regulatory design, and of keeping agencies from issuing
ineffective regulations that run contrary to the public interest.

13.4 The growing regulatory complexity of foreign trade: 2000–2005
By the turn of the century, it was clear that the Mexican economy was facing
some very serious structural problems.9
There seemed to be a general consensus
regarding the need for important reforms in energy, telecommunications,
labour and fiscal policy, but the political climate made it difficult to make
major changes to sector-specific regulatory policy.10
In spite of the changes
in the political landscape, Cofemer maintained and even increased its public
visibility during this period. Horizontal reforms for increasing the transparency
and accountability of government remained feasible. In response to this trend,
Cofemer drafted another groundbreaking legislation, the transparency and
freedom of information law that was approved by Congress in 2002.
Cofemer also prepared a set of regulatory reform proposals in natural gas, road
haulage,11 water and environmental regulation that did not fare as well in the
political process. Cofemer was also one of the first institutions to point out the
need for a detailed review of customs regulations and procedures to better meet
the needs of exporting and import firms, despite the fact that it does not have
jurisdiction over tariff and customs matters, only over technical standards and
administrative rules pertaining to access to certain foreign trade programmes.
Among some of the issues Cofemer pointed to was the increase in border
crossing times and bureaucratic formalities that were beginning to seriously
burden Mexico’s trade competitiveness.12 For example, the growth of customs
brokers, which hold a monopoly on the processing of foreign trade transactions,
did not keep pace with the growth in trade volume in the 1990s, and has remained
stable for the last decade, and in 2000 the trade ministry was handling foreign
trade regulations and formalities with the same budget it had ten years earlier.
9 The success of the structural reform programme and growing international confidence in Mexico
clearly show in the foreign investment data that jump from just over $4 billion in 1993 to a peak of
nearly $30 billion in 2001. But the data also show a prolonged period of stagnation in FDI inflows
since 2000. The data also show us a clear trend of falling market share in the North American market
and a general decline in GDP growth rates beginning in 2001. Cuevas et al (2002) link the growth
slowdown in Mexico in part to the halt of economic reforms.
10 Eventually limited energy and fiscal reforms were approved in 2008.
11 For an overview of some of the issues discussed in the road haulage sector, see Box 13.2.
12 Soloaga et al (2006) estimated that unilateral trade facilitation in Mexico could lead to increases of
20% for exports and 11% for imports.
Box 13.2 Review of courier services rules
Ever since the enactment of the Law on Road, Bridges and Road Transport in 1993, the
regulation of road haulage and courier services has been a hotly debated topic, mainly due
to the fact that road haulage is an activity reserved for Mexican nationals, while courier
services are completely open to foreign investment. The law called for the development
of implementing rules for courier services within 6 months of the law’s entry into force.
It was not until 2011, however, that these implementing rules were finally issued, with
Cofemer playing an important role in the process.

The law rather vaguely defined courier services as those approved by the Ministry
of Communications and Transport (SCT) according to the criteria specified in the
implementing rules. As for the packages themselves, the only characteristics specified
were that they were to be properly wrapped and labelled (envueltos y rotulados) and
accompanied by a shipping guide.
Road haulage companies are a rather combative interest group and they expressed their
concerns regarding ‘unfair competition’ from courier services infringing on their turf.
Tensions in the sector were exacerbated by the fact that NAFTA called for the complete
opening of NorthAmerican road haulage services in 1994, but was postponed for political
reasons until 2011. Road haulage companies therefore argued that courier companies
were competing unfairly in road haulage (since foreign companies were prohibited from
providing the service) and that they were not being granted access to the US market as a
result of NAFTA obligations not being honoured.13
Through the Cofemer review process, road haulage companies argued for limits on the
gross weight of trucks for courier services and for maximum size and weight limits on
the packages they could transport. Courier companies argued that the limits on truck and
package dimensions would unnecessarily limit their competitiveness, and that the limits
would serve no safety purpose, which is what SCT should focus on. The RIA prepared by
SCT distilled the arguments in support of the proposed implementing rules, and revealed
certain bias towards creating entry barriers that favoured domestic companies. The initial
proposal was that gross truck weight for courier services should not exceed 9 tons and
that packages should not exceed 50 kg, and included an implied authorisation of routes.
The courier companies argued that there should be no specific regulation for truck size
or weight and that package limits should be set at no lower than 70 kg. Bus companies
would be allowed to offer courier services, while road haulage companies would not.
The RIA presented only general information on the size and composition of the cargo
and courier fleets, and on the number of people working in the sector and so forth, but
relatively little on the effects that the rules would have on them. Only some direct costs
were quantified, and benefits were limited to stating that the measure would attract foreign
investment and increase legal certainty. The justification for truck size limits was to enable
authorities to visually distinguish courier from cargo trucks. Package size and weight
limits were based on the argument that heavy packages were a health risk for courier
employees. Cofemer’s opinion made reference to the Federal Competition Commission’s
(CFC) vehement criticism of the unnecessary barriers to entry contained in the proposal,
and in response SCT raised the truck gross weight limit to 11 tons (respecting the size and
weight of courier trucks currently in use), and eliminated package restrictions altogether.
Finally a compromise was reached in that bus companies and cargo companies would
be allowed to provide courier services, all route authorisations would be eliminated,
maximum truck weight was raised to 11 tons and package weights were limited to 31.5
kg. Though the outcome was not as procompetitive as Cofemer and the CFC may have
desired, the adamancy with which the Cofemer opposed the original proposal helped
13 Fox et al (2003) estimate welfare losses of $3.2 billion derived from the US–Mexico trucking
dispute.

While COECE continued supporting free trade, the rival Mexican Foreign Trade
Council (Consejo Mexicano de Comercio Exterior, or COMCE) openly called for
a moratorium on new FTA negotiations during these years, which the government
formally established in late 2003, as the FTA negotiations with Japan were
coming to a close. COMCE also spearheaded the calls for an increased focus
on ‘national competitiveness’, rather than foreign trade, which eventually led to
the creation of the Mexican Institute for Competitiveness (Instituto Mexicano
para la Competitividad, or IMCO) in 2003.14
The government again called for
the politically expedient deregulation of formalities, and Cofemer responded by
completing the RFTS and identifying a list of over 60 high-impact formalities
for immediate improvement. The Office of the President and the ministry of
public administration (Secretaría de la Función Pública, or SFP) assisted in
the implementation of this project by extending it to customs regulation and
setting up working groups with business representatives to agree on a set of
presidential commitments of formalities improvement. Some relevant measures
were adopted through this process, most notably the elimination of business
associations’ approval of the inscription in the national importers registry.15
The negotiations excluded the CFMR, however, which never really recovered
the influence it once had in the regulatory improvement programme.16
Since
then, regulatory improvement has been led in large part by the SFP. Given that
the minister of public administration operates at the leisure of the president,
it is much more exposed to political pressures than Cofemer. This does not
mean that it does poor work, but it is not subject to the same mandate, nor the
accountability mechanisms, as the CFMR. The reduction and elimination of
information requests and the increased use of information technology that SFP
focuses on are important from an administrative development standpoint, but
they are not the essence of building an effective regulatory governance system.17
The relevance of promoting good regulatory governance rather than
administrative simplification is clear. The design of policies, institutions,
processes tools and evaluation of regulatory outcomes is beyond the legal
14 IMCO eventually published a very influential 2004 report on competitiveness that includes an in-
depth analysis of foreign trade regulation.
15 Until then, incumbent business associations could oppose the inclusion of new competing companies
in the national import registry, without which it was impossible to import goods into Mexico.
16 The meetings of the CMFR turned into public presentations of Cofemer’s activities, without much
scrutiny or analysis of individual ministries’ activities. It has not held a single meeting during 2006–
2010.
17 For instance, SFP has traditionally been worried about the establishment of national preferences in
public procurement in order to ‘promote national interests’.
avert a more anticompetitive regulation from being issued. Though costs and benefits
were never properly quantified, RIA played a key role in making the bias inherent in
the proposal clearly visible, and it raised the public cost to SCT for not conforming to
Cofemer’s regulatory principles.

scope and abilities of the SFP, but they are the essence of Cofemer’s mandate.
In addition, given that one of the main obstacles to the creation of efficient
regulatory systems has been the problem of regulatory capture by political or
business interests, it is at the very least questionable to put the process in the
hands of a government ministry instead of an autonomous institution with the
necessary accountability mechanisms to avoid the problem of capture.
Without strong support of the CFMR – or the president – however, Cofemer
lacksthepowerandindependencetoconvinceorcompelagenciesintocompliance
with the regulatory reform programme. To put it bluntly, if the head of Cofemer
is too aggressive in pushing issues considered damaging to the interests of the
political party in power, he can quickly be replaced by someone more malleable.
Since 2005, therefore, its role has been largely relegated to promoting formalities
deregulation, the diffusion of regulatory reform programmes at the state and local
level, and ensuring opportunities for public comment on regulatory proposals.
These programmes are relevant, but Cofemer is no longer the epicenter for
structural change that it once was.
Despite these obstacles, Cofemer devoted itself from the outset to consolidating
the regulatory reform, including the continued development of RIA (see Box
13.1), and the launching of a state-of-the-art electronic submission platform
that not only set a common template for the presentation of data and results,
but also enormously facilitated public consultation and comment on regulatory
proposals.18
This activity, although more relevant because it addresses the
underlying regulatory problems that generate poorly designed formalities, did not
draw as much attention.And yet, it is probably the most enduring action Cofemer
has implemented in the last decade, along with the freedom of information law.
Indeed, it is through the review of the RIAs for the rules for the administration
of preferential import tariffs programmes and of agricultural import quotas that
the issues of regulatory complexity and potential discrimination first became of
interest to Cofemer.
Eventually, several relevant institutions prepared detailed reviews and analysis
of customs and foreign trade regulatory policy, including IMCO (2005), the CFC
(2008), and the vice-ministry for standards and foreign trade services. The latter
commissioned a detailed study on the impact of foreign trade regulation on
competitiveness in 2004, which would later be influential in the policy shift that
took place after 2006 (Bracho et al, 2005).
If there was any doubt in 2000 that Mexico’s asymmetric tariff structure,
combined with the administration of selective tariff preferences, countervailing
duties, rules of origin requirements and border inspection of NOMs were
creating high costs for importers and users of foreign inputs, by 2006 those
were completely dispelled. Clearly, the benefits of preferential access to the US
market were being eroded by the gradual lowering of MFN tariffs and diverse
18 Before the electronic RIA platform, only the titles of regulatory proposals were published once a
month in the official gazette. Interested parties would then have to go to Cofemer to ask for a copy,
which left little opportunity to comment effectively and within the review times stipulated by the
LFPA.

arrangements for preferential access to the US market. Regulatory and logistical
problems were damaging the other main contributor to Mexico’s comparative
advantage, its geographical proximity to the US. China, generally seen as a
direct competitor of Mexico’s in international markets, had begun its vertiginous
export growth, and Mexico was going to have to face the effects of its dwindling
advantage as a location with intermediate wages and technological sophistication
for certain key industries. China replacing Mexico as the number two exporter to
the US confirmed many of the fears.19
Cofemer’s growing concern was that foreign trade regulation and processing
was beginning to have a large impact on Mexico’s competitiveness. The diversity
of regulatory requirements to obtain the preferential FTA benefits under the
agreements were becoming increasingly complex (rules of origin accreditation
was often, and still is, named as one of importers’ major obstacles). Mexico also
needed to improve its trade processing capacity, both regulatory and logistical,
to take full advantage of its proximity to the US market, particularly in light of
the emergence of East Asia as an important provider of industrial inputs.
By the year 2000, 90% of Mexico’s trade was being conducted with its FTA
partners, and roughly 80% with the US alone. However, Mexico has been known
to selectively raise tariffs to non-FTA countries in response to political pressures,
and its MFN tariff levels have been quite high. In fact, Mexico actually increased
MFN tariffs quite significantly as recently as 1999.
High average MFN tariffs and very low preferential FTA tariffs made Mexico’s
tariff structure very complex. This became a problem for many important
manufacturing industries that depended on low cost inputs and machinery from
non-FTA countries. In response, the government designed programmes like
the Programas Sectoriales (Prosec) in 2001, which essentially granted North
American tariff preferences for non-originating inputs used in the production of
exports. Prosecs were created for industries such as chemicals, footwear, steel,
autos, textiles and electronics, which benefitted from a 10 to 15% tariff reduction
on imports under the programme. Programmes like these amount to an export
subsidy and were scrutinised by Cofemer because of their discriminatory nature.
Mexico also became a frequent user of antidumping measures, applying more
than 90 countervailing duties, which affected nearly 1,400 tariff lines in 2002.
The majority of these duties exceeded 200% and many were well over 500%.20
With such a huge disparity in duties, the incentives to evade them and to transship
goods through the US or incur in other forms of contraband were significant. It
was estimated that the amount of contraband goods sold in street markets ranged
from 50 to 90% for clothing, music, software, cigars and footwear (Bracho et
al, 2005).
Finally there is the case of NOMs. These technical rules are generally applied
in a non-discriminatory fashion throughout Mexico. However, because of the
purported difficulty of identifying someone responsible for compliance in the
19 These concerns led Mexico to be the last country in accepting China’s membership in the WTO.
20 Source: WTO Trade Policy Reviews of Mexico.

case of imported goods, many NOMs are verified at the border. This creates two
significant problems. First, it implies a complicated coordination issue between
customs officials, responsible ministries, and private conformity assessment
bodies, which can generate costs and delays. Second, these procedures have
been used as a way of containing imports in defence of certain products and
industries. The proportion of tariff lines subject to border inspection of NOM
standards is as high as 70 to 90% in the case of food, textiles and footwear. The
inherently discriminatory nature of border conformity assessment procedures
was also a common theme of Cofemer’s opinions (see Box 13.3).21
21 Econometric studies have documented the increase of production costs due to standards and how
they may constitute technical barriers to trade (Maskus et al, 2004). Recently, the Mexican Federal
Competition Commission has also actively promoted the competition screening of NOMs (Conrath,
2009) to help avoid such problems.
Box 13.3 Review of food labelling standards
The 1996 mandatory technical standard on food labelling in Mexico was a contentious
trade-related measure that had been the subject of much debate for the better part of
10 years. The standard mandates the labelling of nutritional content and consumer
information for most packaged foods in Mexico, but did not until recently conform to the
international standard set by Codex.
Foreign imports were subject to ex ante and sometimes discretional conformity
assessment procedures at the border, where their products were sometimes immobilised
for containing additional information or health claims not covered by the standard.
Domestic interests were mostly content with the fact the standard acted as a trade barrier
that obstructed access to foreign goods. The government was dissatisfied with the fact
that the regulation did not appear to be efficiently achieving its stated health goals. The
complaints and defences of the standard were, for the better part of a decade, generally
unorganised and led to sterile discussions that did not produce any concrete outcomes.
Finally, in 2008 a number of factors came into play to push for the review of the standard
that eventually led to its modification and harmonisation with international standards. The
first has to do with a Cofemer sponsored reform to the Federal Standards and Metrology
Law in 1997 that required a five-year review and automatic sunsetting clause of technical
standards, and a default requirement of harmonisation with international standards. Upon
the date of the second five-year review date in 2008, importers made it clear that the
standard was obsolete and out of line with international standards. The second major
factor was that the government was pushing for a major campaign against obesity and
wanted to ensure that the standard was generating useful consumer health information in
line with the government policy.
A review and RIA of the standard was initiated as a combined effort of the Economics
Ministry and Federal Commission for the Protection Against Sanitary Risks (Cofepris).
The first advantage of the RIA was that it clearly stated which issues were to be reviewed
(consumer identification of health risks, clarification of information and health claims to

The problem of the illegality spawned by this complex system is compounded by
the fact that it has led to a number of measures designed to combat the problems,
which paradoxically make it even more complex, creating not only the possibility
of corruption, but also creating new incentives to evade regulation. Outright
pressure to protect certain industries was also a clear problem, as evidenced by
importers summarily eliminated from the import registry and the increase in
customs embargos by over 600% from 2000 to 2004 (Bracho et al, 2005).
All of the aforementioned measures and others such as the general and
sectorial registries of importers, exclusive custom entry points for certain goods,
and estimated prices, can be attributed to one of three causes: (1) the complexity
be included in labels, harmonisation with international standards, compliance costs and
conformity assessment procedures), which helped enormously in ordering the discussion.
During the time where the regulatory process was opaque, the government had little
incentive to confront protectionist interests and the inefficient aspects of the standard
were never really publicly debated. The mandatory public five-year review allowed
parties interested in modifying the standard to be heard and to push for harmonisation.
Once the review process was opened, everyone had an equal opportunity to publicly
present their views.
The RIA also shifted the debate towards a more technical discussion. The government
was forced to present credible studies and data regarding the incidence and effects of
health claims, and the importance of reliable nutritional information. It also began to
present trade facilitation arguments as an additional benefit to consumers, thanks to the
involvement of the trade areas of the Economy Ministry. The business sector opposed
to harmonisation had to present reasonable arguments against changing the standards,
and could no longer rely on directly lobbying the government on protectionist grounds.
It focused therefore on the potential compliance costs of the regulatory change derived
from the large inventories that would be ‘caught’in non-compliance. Importers presented
information regarding the costs due to uncertainty of the rules themselves and regarding
the discriminatory nature of at-the-border inspections, rather than regular ‘retail shelf
inspections’ like those domestic manufacturers were subject to.
Although the quality of the data presented was certainly not immune from criticism of
bias, and one cannot say that the estimation of compliance costs was entirely satisfactory,
the mere presentation of legitimate issues rather than backdoor lobbying was enormously
helpful in shaping the debate. Because the default position required by law is one of
harmonisation with international standards, the Codex standard was used as a model, and
it proved impossible for opposing parties to discredit it or allege that domestic conditions
in Mexico were different. The principles of harmonisation and non-discrimination of
foreign products were agreed upon, and all that was left was to attend to the stranded
inventories argument. This issue was resolved by incorporating a gradual implementation
of the new obligations over one full year. This was sufficient to overcome business and
conformity assessment bodies’ opposition to the changes and the new standard was
published in February 2010.

of foreign trade regulation; (2) the need to combat fraud and enforce the law; or
(3) pressure to protect certain sectors.
Unfortunately, the result of all this has been the erosion of two of the main
factors that contribute to Mexico’s comparative advantage: the preferential
access to the US market because of NAFTA, and the geographic proximity
to the US. According to one study, an average US to Mexico border crossing
could take two to five days and cost between 1.5 and 6% of the merchandise
in question, while in the other direction it would only take between 1.5 and 13
hours and cost only 0.5 to 2.6% of the merchandise value (Hummels, 2001).
Other studies confirm a clear dampening effect of NTMs on the welfare effect of
tariff reductions (Anderson and van Wincoop, 2001), and a ‘border effect’nearly
an order of magnitude greater in US–Mexico prices versus US–Canadian prices
(Rogers and Smith, 2001).
13.5 Unilateral trade opening and application of regulatory reform
principles to trade related measures: 2006–2011
The dwindling preferential access to the US market, the rise of China and other
nations as exporting powers and direct competitors of Mexico in world markets
pointed to the need for moving to higher value-added products and exploiting the
geographical proximity to the US as effective policy options.An effective way of
contributing to this is by reducing the complexity of foreign trade regulation by
applying simple regulatory reform principles to promote entrepreneurship and
allowing greater industry flexibility, and business competitiveness.22
The first major steps were taken in 2006. The trade ministry reduced MFN
rates of over 6,000 tariff lines of inputs and raw materials (especially in textiles,
chemicals and ore) relevant for 18 sectors in which tariff heterogeneity was high
and where the cumulative tariff rates on inputs was higher than for the finished
goods, which obviously reduced incentives to produce in Mexico. The move
taken at the end of the Fox administration was a major shift in policy and did not
generate as much opposition as might have been expected.
The process continued under the Calderón administration. In 2008, Mexico
pushed the process of unilateral market opening further, based on a general
reduction of MFN tariff levels, the simplification of exceptions and customs
procedures, and the strengthening of Cocex’s institutional design.
The economy ministry reduced industrial MFN tariff levels from 10.4% in
2008 to 5.3% in 2010 and they are scheduled to be reduced further, to 4.3% by
2013. The number of duty-free tariff lines was also increased from 20% in 2008
to 63% in 2010, meaning that over 90% of imports now enter Mexico duty-
free. To address the complicated administrative structure of granting duty-free
access to inputs and machinery for certain industries, the preferential treatment
granted under programmes such as Prosec are being extended to all, eliminating
22 For a description of how Asian competition has spurred the impetus for reform in Mexico see Ibarra-
Yúnez (2004).

discriminatory effects. Prosec preferences applied to 4,800 tariff lines will be
completely eliminated by 2013 (down to 1,300 already in 2010). Quotas are
being changed to tariffs in the case of textiles, and auctioned for sugar and toys
(another frequent Cofemer recommendation). NOM conformity assessment
procedures – which regulate over 3,700 internationally traded products and
services – are being reviewed, and now nearly 60% of the value of Mexican
imports enter free of non-tariff regulation (Martínez Trigueros, 2010a).
Mexico also formally recognised the equivalence between Canadian, US and
Mexican safety standards for household appliances and business electronic
equipment. The decree accepts certification of compliance with North American
standards for commercialisation of products in Mexico. Certification bodies
and laboratories in Mexico were opposed to the measure, but in the end, the
major business associations agreed to it, since it would reduce unnecessary
certification procedures. The government did not consult formally with the
private sector through the usual Cofemer procedure, however. The LFPA allows
for an exemption of RIA for rule changes that are judged to reduce compliance
costs, and of public consultation procedures in cases where the publicity is
deemed to compromise the effects of the regulation. The fear was that the issue
would stir up controversy or court actions that would eventually halt the process.
Experience has shown, however, that it is usually better to treat all arguments
on their merits and give assurances that all public concerns have been properly
addressed. Nevertheless, the measure is a positive one.
The same measure was taken for medical devices in October 2010 by health
officials (Comisión Federal para la Protección contra Riesgos Sanitarios,
or Cofepris) to expedite access to safe and effective medical devices that
comply with North American standards, an issue had been raised in Cofemer
recommendations since the year 2000.
These decisions represent an important policy shift in practice, because
traditionally the Mexican government had sought to promote harmonisation
(convergence of standards) or mutual recognition of conformity assessment
procedures, rather than outright recognition of foreign certification. Laboratories
and conformity assessment bodies were obviously against the measure, because
equivalence eliminates the need for Mexican certification for imports, but the
gains from avoiding unnecessary delays and testing costs which negatively
affected the availability of goods and prices for consumers sufficed to overcome
domestic resistance.
On the customs administration side, the work with business associations has led
to the elimination of the sectorial import registries, the burdensome requirement
of having to identify and label individual units of goods at the border, and the
posting of bonds in the case of goods that were subject to estimated pricing
rules. A one-stop shop project for customs and foreign trade regulation – similar
to Cofemer’s system for rapid business registration in design – is also being

implemented.23
All ministries’ regulations that need to be enforced by customs
authorities will be included in a single portal for foreign trade formalities, and
direct access will be granted to the private sector (not only customs brokers, but
also trade advisors and shipping companies).
Cocex will be strengthened by making its comments open and public (including
resolutions, opinions and reports), and requiring its opinion before any non-tariff
measure can be published in the official gazette and effectively applied.
Rules of origin requirements are now more uniform. Negotiations for the
accumulation of rules of origin in textiles for Mexico and Central American
nations, as well as with the Dominican Republic and the US, have been concluded
to support the competitiveness of the textile industry at the regional level and of
Mexican exports in particular. The agreements make it easier for producers to
draw on the participating countries for their supplies at preferential tariff levels,
thereby allowing the finished products to acquire originating status.
Finally, Mexico and China entered into an agreement in 2008 that obligates
Mexico to repeal a large number of antidumping duties in areas such as textiles,
apparel, footwear, toys, bicycles, tools, chemical products, valves, and locks, by
2011. To minimise opposition to the agreement, reconversion funds have been
approved by Congress.
13.6 Conclusions
Trade policy is a central element of regulatory governance. Greater openness to
foreign trade enhances competition, expands productive capacity, and benefits
consumers in domestic markets. It also generates better regulation by providing
the incentives to reduce transaction costs, improve transparency, predictability
and legal certainty. In the vast majority of cases, tariff cuts are regulatory
improvements in and of themselves because they reduce barriers to entry and
competition, and improve the diversity, quality and prices of goods offered.
Trade facilitation should be the sole objective of trade policy, and the creation
of barriers, either in the way of tariffs or NTMs, should occur only on the basis
that the benefits of doing so would outweigh the costs to society as a whole. To
achieve this, trade policy should also be subject to regulatory reform principles
and review in general.
It is important to be clear about the objectives of trade policy and their relative
prioritisation. This is a primary responsibility of the trade ministry, naturally,
but the coordination with authorities responsible for regulatory NTMs, and of
course with customs administration is essential to clarify the overall direction of
government policy. Specifically, Mexico’s experience shows that the following
measures could lead to more efficient regulation of trade:
23 The Sistema de Apertura Rápida de Empresas (SARE) created by Cofemer, which led to an
improvement in the IFC/World Bank Doing Business rankings for Mexico (World Bank, 2009).

• Review of tariff structure. Reducing fiscal burden, eliminating
contradictions and reducing tariff dispersion should all be central
goals. In the case of Mexico, this meant switching from a bilateral
trade negotiation strategy based on reciprocal concessions, back to a
unilateral trade liberalisation strategy. Ideally, tariff reduction schedules
in different FTAs should try to be accelerated and synchronised. The
process should be accompanied by a parallel reduction and eventual
elimination of preference programmes such as Mexico’s Prosecs.
• Establishment of full regulatory reform disciplines for all trade-
related measures. Mexico’s experience illustrates the importance of
including trade issues under the fold of a government-wide regulatory
improvement programme. The fact that tariffs and customs regulations
were not reviewed by Cofemer because of its lack of jurisdiction over
fiscal issues, probably meant that they were more susceptible to capture
than other forms of regulation, which may be part of the reason for which
it took the government the better part of a decade to attack the issue of
the complexity of trade regulation in a coordinated way. Transparency,
public comment and technical justification are needed in the case of
trade-related measures just as for any other sort of regulation. Cocex
was an early regulatory reform-type body for tariff and NTM regulation,
and is useful as an intragovernmental discussion forum, but it lacked the
transparency, staff and power to contest actions proposed by powerful
ministries. The recent measures making its opinions public are a step in
the right direction, but it is still far from regulatory best practice.
• Continuous review of the impact of regulation on foreign trade and
competitiveness. Regulatory impact analysis of NTMs and regulatory
proposals in general should be applied both ex ante and ex post to ensure
proper design, to correct assessment of benefits and cost (effects not only
on government revenue, but also on the volume of trade, consumers,
business competitiveness, employment, competition and social
objectives) and to quickly identify ways of improving the regulatory
environment. RIA in Mexico explicitly requires consideration of trade
effects by all regulation submitted to Cofemer, but ex post review is
still rare. Mexico could have benefited from a much earlier response to
accumulating NTM barriers to trade.
• Coordination and evaluation of trade policy and its regulation. A central
body should be entrusted with the overall design and implementation of
policy, and the coordination of all of the different agencies involved.
The goal should be to reach consensus with regard to the objectives and
use of instruments of trade policy, and to periodically and transparently
evaluate its effectiveness with respect to pre-established parameters.
This would help in conferring a clear common vision to trade policy and

to reduce the occurrence of ad hoc regulatory reactions to protectionist
pressures and special interests. The government should also engage
business, the legislative branch and consumer groups to clearly
communicate the contributions of regulatory reform of trade measures
to competitiveness and social welfare.
References
Anderson, J and vanWincoop, E (2001), ‘Borders,Trade andWelfare’, paper prepared
for Brookings Trade Forum 2001 on ‘Globalization: Issues and Implications’.
Anson, J, Cadot, O, de Melo, J, Estevadeordal,A, Suwa-Eisenmann,A and Bolormaa,
T (2003), ‘Rules of Origin in North-South Preferential Trading Arrangements with
an Application to NAFTA’, CEPR Discussion Paper 4166.
Aspe Armella , P (1993), El camino mexicano de la transformación económica,
México, Fondo de Cultura Económica.
Bazdresch , C, Bucay, N, Loeaza, S and Lusting, N (eds) (1992), México. Auge,
crisis y ajuste. 3 vols, México, Fondo de Cultura Económica.
Blanco Mendoza, H (1994), Las negociaciones comerciales de México con el
mundo, México, Fondo de Cultura Económica.
Bracho, M E, Haddou-Ruiz, A, Hernández, F, López-Ayllón, S, Montañés, J and
Rodán Xopa, J (2005), El impacto de la regulación del comercio exterior en la
competitividad y el bienestar económico. Report prepared for the Secretaría de
Economía, México, DF.
CFC (2008), Opinión con el fin de promover la aplicación de los principios
de competencia y libre concurrencia en el diseño y aplicación de políticas y
regulaciones del comercio exterior de mercancías, May. Available at: http://
resoluciones.cfc.gob.mx/Docs/COMENTARIOS/opinioncomercioexterior.pdf
Comisión Federal de Mejora Regulatoria. Informes de labores de la Comisión
Federal de mejora Regulatoria. México, DF. 2000-2009.
Conrath, C (2009), OECD Competition Assessment Toolkit: Competition Screening
of Normas. Available at: http://www.oecd.org/dataoecd/32/11/45049530.pdf
Córdova, C and Haddou-Ruiz, A (2008), Regulatory Transformation in Mexico,
1988–2000: Case Studies on Reform Implementation Experience, Investment
Climate Advisory Services, The World Bank Group.
Córdova, C, Kim, J S, Kim, T Y, Yang, J, Jacobs, S, Haddou-Ruiz, A, Varley, C,
Deighton-Smith, R and Carbone, L (2009), Lessons for Reformers: How to
Launch, Implement, and Sustain Regulatory Reform. An Analysis of Six Case
Studies in Developing and High-Income Countries, Investment ClimateAdvisory
Services, The World Bank Group.
Cuevas,A, Messmacher, M and Werner,A (2002), ‘Macroeconomic Synchronization
between Mexico and its NAFTA Partners’, Mimeo, Banco de México.
Elizondo Mayer-Serra, C (2001), La importancia de las reglas: Gobierno y
empresario después de la nacionalización bancaria, México, DF, Fondo de
Cultura Económica.

Espinosa Velasco, J E and Serra Puche, J (2004), ‘Diez Años del Tratado de Libre
Comercio deAmérica del Norte’, in P GarcíaAlba, L Gutiérrez and G Torres (eds),
El Nuevo Milenio Mexicano. I. México en el mundo, México, UAM-Editorial
EON.
Fox, A, François, J and Lodoño-Kent, P (2003), ‘Measuring Border Crossing
Costs and their Impact on Trade Flows: The United States-Mexican Trucking
Case’, GTAP Resource 1282, Center for Global Trade Analysis, Department of
Agricultural Economics, Purdue University, April.
Hummels, D (2001), ‘Time as a Trade Barrier’, GTAP Working Papers 1152,
Center for Global Trade Analysis, Department of Agricultural Economics, Purdue
University.
Ibarra-Yúnez, A (2004), ‘NAFTA as a Vehicle for Regulatory and Institutional
Convergence in the North American Region’, Institute for Research on Public
Policy Working Paper No. 2004-09g.
IMCO (2005), Situación de la competitividad de México 2004: hacia un
pacto de competitividad, Monterrey, NL, Mexico, Instituto Mexicano para la
Competitividad.
Kose, M, G Meredith and C M Towe (2004), ‘How Has NAFTAAffected the Mexican
Economy? Review and Evidence,” IMF Working Papers 04/59, International
Monetary Fund.
Lederman , D, Maloney, W and Serven, L (2003), Lessons from NAFTA for Latin
America and the Caribbean Countries: A Summary of Research Findings,
Washington DC, The World Bank.
López Ayllón, S and Haddou-Ruiz, A (2007), ‘Rendición de cuentas y diseño
institucional de los órganos reguladores en México’, Política y Gestión Pública
XVI(1), pp 101–46.
Martínez Trigueros, L (2010a), Facilitación Comercial. Presentation at the OECD
Forum on Regualtory Reform, Mexico City, 12 January.
Martínez Trigueros, L (2010b), ‘Promoviendo la Competitividad de Latinoamérica
y el Caribe’, Mimeo.
Maskus, K, Otsuki, T and Wilson, J (2004), ‘The Costs of Complying with Foreign
University of Colorado at Boulder: Institute of Behavioral Science Working Paper
PEC2004-0004.
OECD (1999). OECD Reviews of Regulatory Reform: Regulatory Reform in
Mexico, Paris, OECD.
OECD (2004). OECD Reviews of Regulatory Reform: Mexico 2004. Progress in
Implementing Regulatory Reform, Paris, OECD.
OECD (2009). Indicators of Regulatory Management Systems. Regulatory Policy
Committee 2009 Report, Paris, OECD.
OECD (2009). Overcoming Border Bottlenecks: The Costs and Benefits of Trade
Facilitation, OECD Trade Policy Studies, Paris, OECD. 241 pp.
Ortiz Mena, A (2006), ‘The Domestic Determinants of Mexico’s Trade Strategy’,
Working Paper No 138. Centro de Investigación y Docencia Económicas: Mexico,
DF.

Presidencia de la República (1995). Acuerdo para la Desregulación de la Actividad
Empresarial, Diario Oficial de la Federación, 24 November.
Rogers, J and Smith, H (2001), ‘Border Effects within the NAFTA Countries’, Board
of Governors of the Federal Reserve System International Finance Discussion
Paper No 698.
Rubin, R (2003), In an Uncertain World: Tough Choices from Wall Street to
Washington, New York: Random House.
Salas, F (2004), ‘Mexican Deregulation: Smart Tape on Red Tape’, Background
paper 31435 prepared for the World Development Report, World Bank.
Salas, F (2005), Regulatory Reform. Institution Building: Lessons from Mexico.
Public Policy for the Private Sector Note 282. The Washington, DC, World Bank
Group, January.
Secretaría de Comunicaciones y Transportes (2011), Reglamento de Paquetería y
Mensajería, Diario Oficial de la Federación, 6 April.
Secretaría de Economía (2010), Acuerdo por el que se aceptan como equivalentes
a la NOM-016-SCFI-1993 Aparatos electrónicos-aparatos electrónicos de uso
en oficina y alimentados por diferentes fuentes de energía eléctrica-requisitos de
seguridad y métodos de prueba y sus resultados de evaluación de la conformidad,
los reglamentos técnicos y procedimientos para la evaluación de la conformidad
de los Estados Unidos de América y de Canadá, Diario Oficial de la Federación,
17 August.
Secretaría de Economía and Secretaría de Salud (2010), Norma Oficial Mexicana
NOM-051-SCFI/SSA1-2010, Especificaciones generales de etiquetado para
alimentos y bebidas no alcohólicas preenvasados-Información comercial y
sanitaria, Diario Oficial de la Federación, 5 April.
Soloaga, I,Wilson, J and Mejía,A (2006), ‘Moving Forward Faster:Trade Facilitation
Reform and Mexican Competitiveness’, World Bank Policy Research Working
Paper 3953, June.
Tornell, A and Esquivel, G (1995), ‘The Political Economy of Mexico’s Entry to
NAFTA’, NBER Working Paper 5322.
UNCTAD, World Investment Reports 2000–2008, Geneva, United Nations
Conference on Trade and Development.
Urquiza , R, Wilson, J S, Mann, C L and Otsuki, T (2004), Assessing the Potential
Benefit of Trade Facilitation: A Global Perspective, World Bank Policy Research
Working Paper 3224.
World Bank (2009), Doing Business 2010: Comparing Regulation in 183
Economies, Washington DC, World Bank.
WTO (2008), Trade Policy Review: Mexico, Switzerland,World Trade Organization.
WTO (2010), World Tariff Profiles 2006–2010, Switzerland, World Trade
Organization.
Zabludovsky, J (2005), ‘El TLCAN y la política de comercio exterior en México:
Una agenda inconclusa’, Información Comercial Española, ICE: Revista de
Economía No 821, March–April.

14
Regulatory Impact and Cost-Benefit
Analyses – Tools to Evaluate
Regulatory Quality and Improve
Transparency among Countries
Sílvia Helena G de Miranda
14.1 Outline
Regulatory impact analysis (RIA) is a tool that has been adopted by several
countries as part of a policy to improve their regulatory quality, providing more
legitimacy to regulations and reducing conflicts between government agencies
and private sectors. When a country implements an ex-ante or ex-post evaluation
of public policies this gives more room to involve private agents in participating
in the regulatory process, making them aware of it since the conception of a
new policy or programme proposed by public agencies until the monitoring of
its consequences. In RIA’s scope, one of the main analytical techniques used is
cost-benefit analysis (CBA).
In Brazil, some governmental agencies are currently implementing RIA,
aiming at having a common and quite harmonised methodology. For achieving
such a goal, those agencies have been providing capacity building programmes
to their staff. The Casa Civil, which is a branch of the Republic Presidency,
with a ministry status, and which aims at assisting and advising directly and
immediately the President, is in charge of a programme, the PRO-REG
(Programa de Fortalecimento da Capacidade Institucional para a Gestão em
Regulação1
), which is focused on promoting and implementing RIA in Brazilian
regulatory agencies, and is also supported by the International Development
Bank (IDB). Several agencies are taking part in this initiative, although there
are other individual programmes taking place in Brazil, also having RIA
implementation as a target. That is the case of INMETRO’s (the National Institute
of Metrology, Standardization and Industry Quality) programme to apply RIA
to conformity assessment programmes. PRO-REG and the Analysis of Technical
1 In English, literally, Programme to Strengthen the Institutional Capacity for Regulatory Management.
More information can be found at http://www.regulacao.gov.br/.

and Economic Viability conducted by the Directory of Quality from INMETRO
are two examples to be highlighted in this new regulatory environment that has
taken place in Brazil recently.
As far as the sanitary policy is concerned, a Brazilian institution, which
is focused on promoting research, the CNPq – Conselho Nacional para o
Desenvolvimento Científico e Tecnológico2
– sponsored a large project regarding
agriculture and cattle health and safety, particularly focused on evaluating the
current Brazilian system of Sanitary Defence, which involves universities,
agencies and private companies in a broad discussion to build up a diagnosis
of major problems and priorities to be worked on, and to identify opportunities
to improve the sanitary and phytosanitary safety services. Likewise, this project
has also aimed at constituting a network of experts in such a way as to enhance
the dissemination of technical and scientific information and promote a broader
discussion of themes of interest in this field. This effort to build up a network
of experts and officials and other stakeholders resulted in the creation of the
RIT-DA (Rede de Inovação Tecnológica – Defesa Agropecuária3
). One of the
studies supported by this project deals with proposing a methodology to evaluate
the relation between costs and benefits of sanitary programmes and policy in
agriculture and livestock sectors, conducted by governmental agencies and
departments, especially those related to prevention, control and eradication of
exotic pests. This is one of the cases that will be presented in the following
pages.
Those cases mentioned above indicate that the Brazilian government, either
through direct administration or through research and development institutions,
is sensitive to improving transparence in its regulatory actions and is searching for
analytical tools to give more support to policymakers and to establish priorities
in terms of regulation and budget allocation. Moreover, such developments
in terms of RIA (both applying to ex-post or ex-ante regulations) consist of
important sources of data and information for other sectors of the economy also,
not necessarily directly involved in the sanitary and phytosanitary issues. These
points will be detailed later.
As a country or agency implements RIA, besides providing more transparency
for government decisions, an additional benefit may derive from having the very
detailed technical and legal dossiers available for other uses, and containing
important information regarding specific programmes, regulations or rules
analysed. RIA’s documentation usually comprises not only the quantitative
results of cost-benefit analysis and cost-effectiveness analysis, but also qualitative
information about environmental and social impacts expected from the object
analysed. Additionally, these reports usually include a legal and technical
2 ‘The National Council for Scientific and Technological Development (CNPq) is an agency linked
to the Ministry of Science and Technology (MCT), dedicated to the promotion of scientific and
technological research and to the formation of human resources for research in the country’(available
at: http://cnpq.br/english/cnpq/index.htm).
3 Literally, Network for Technological Innovation in Agriculture and Livestock Safety.

Regulatory Impact and Cost-Benefit Analyses 273
background section that should be part of the report, as it consists of important
information to decide whether or not to implement a new programme or policy.
As these dossiers contain, generally, this detailed information, one could
also use them to give technical and legal arguments to those regulations that
are being notified to the World Trade Organisation (WTO), particularly those
related to non-tariff measures, such as sanitary and technical measures that
are, respectively, negotiated in the TBT (Technical Barriers to Trade) and SPS
(Sanitary and Phytosanitary) Agreements.
If a regulation to be issued by government is expected to generate significant
trade impacts, RIA’s report is supposed to contain some results from the
assessment of this potential effect. Therefore, the dossier may supply the
notification process with information necessary to justify the measure that is
being notified, as well as to describe its impacts to different stakeholders. This
procedure provides more transparency also for notifications to WTO agreements
and so those gains relating to transparence improvements are extended from the
domestic to the international scope.
However, this tool of regulatory quality is well spread in developed countries
but it has just emerged in developing countries, where some additional difficulties
are faced by policymakers and agencies. Therefore, it is important to enlighten
the discussion about RIA implementation in developing countries. Several
factors can postpone or jeopardise the use and success of RIA in countries that
do not count on large budgets and institutional infra-structure to deal with this
new tool of regulatory management. Some of these factors are pointed out below,
being illustrated by the experience of recent studies conducted in Brazil.
14.2 Application of RIA and the OECD model
Several papers published by the OECD explain the concept of RIA, the main
steps to develop it and some examples of its application (OECD, 2004, 2008,
n.d.). Methodological aspects are also explored by some studies and it is very
clear that even in the European Union, countries have implemented RIAs
using different approaches and selecting different criteria to ranking their
priorities. There are countries applying simple questionnaires and using them
as the analytical background for conducting their regulatory impact analysis,
and likewise there are countries applying comprehensive analysis such as cost-
benefit techniques and valuation of intangible effects, most of them related to
social and environmental costs and benefits, which demand a lot of skilled people
and a proper infrastructure to be developed. So, there are a number of different
approaches and results to present and the United Kingdom example is probably
the most advanced in terms of scope and use of different methodologies, besides
being very pragmatic in terms of application.
It is interesting to mention that more than quantitative results, those countries
more advanced in implementing the RIA and making it useful to policymakers
have been very critical about the importance of qualitative discussion whenever
the impacts can be identified, but not properly estimated. Moreover, a recent

development of the RIA basic model shows that aspects related to improving
transparence and the inclusion of sustainability are becoming more and more
relevant to evaluate the public policy role. OECD (n.d.) presents a model called
RIA-S, which has basically the same steps of RIA, but emphasises how the
sustainability aspects could be enhanced in this tool application.
When governments or agencies decide to establish RIA procedures ex-ante,
which means that it is conducted before a certain technical regulation, standard,
conformity assessment or a programme is issued, the policy’s transparency and
legitimacy are improved when it is finally adopted. Consequently, decisions
made according to these studies of RIA contribute to establish priorities in terms
of public expenditures, in such a way as to reduce the number and intensity of
disputes and social controversies. Moreover, usually the RIA procedures include
consulting processes to those sectors and actors affected by the regulation, in
order to collect data or even to ask them to help identify potential impacts or
regulations and identify potential conflicts caused by an eventual regulation.
When the RIA tool is applied to evaluate a regulation that has already been
implemented (ex-post approach), it makes it possible to analyse its effects ex-
post like a surveillance tool, for example, it is possible to verify if the expected
and estimated impacts have occurred or if they were under- or overestimated.
Additionally, government has an opportunity to change or redirect or correct the
path of that regulation and its implementation. It is a way to improve regulatory
quality and public policy efficiency.
According to OECD (2004), the checklist to guide a study of RIA should
comprise the following basic questions:
1. Is the problem correctly defined?
2. Is government action justified?
3. Is regulation the best form of government action?
4. Is there a legal basis for regulation?
5. What is the appropriate level (or levels) of government for this action?
6. Do the benefits of regulation justify the costs?
7. Is the distribution of effects across society transparent?
8. Is the regulation clear, consistent, comprehensible and accessible to
users?
9. Have all interested parties had the opportunity to present their views?
10. How will compliance be achieved?

In order to evaluate the effects of a certain policy and of different scenarios or
alternativestoit,quantitativetoolsareproposedtoevaluatetangibleandintangible
costs and benefits, previously identified for each alternative scenario. It is still
complicated to quantify environmental and social impacts due to regulations
imposed by agencies; however, according to experts, some policymakers would
rather have good qualitative information than doubtful quantitative results.
14.3 Regulatory quality and WTO agreements
The Technical Committee of the Technical Barriers to Trade Agreement has
been stimulating discussion about regulatory impact analysis, although the
first advances were made by OECD countries, as a general effort to build up a
framework for RIA and spread it to a number of countries, even though they do
not achieve the same level of implementation and a large heterogeneity of scope
and methods.
Another illustration of the approach of evaluating impacts related to
international agreements is the case of Pest Risk Analysis (PRA), which has
a legal provision in the SPS Agreement and is regulated internationally by the
International Standards for Phytosanitary Measures (ISPM) no. 11. This measure
establishes the procedures for PRA, and clearly recommends that countries apply
economic, social and environmental impact assessment besides the biological
and agronomical aspects of PRA. In fact, CBA was used in two emblematic cases
involving the North American trade of fruits with its partners. One case relates
to the dispute between Mexico and the United States on avocados (Peterson
and Orden, 2006) and the second one refers to the apple trade flows with Japan
(Krissoff and Calvin, 1998). In both studies, the CBA was an auxiliary method to
estimate the impacts of government decisions on imposing restrictions to trade.
So, in terms of relating the evaluation of public policy and the international
agreements, considering trade aspects, there is room to work on the RIA and on
the particular tool of CBA in order to provide more information and transparency
to commerce and to countries’ regulatory policy.
14.4 Brazilian experience in RIA
Two initiatives held in Brazil illustrate the adoption of RIA in a developing
country and provide empirical elements to work on improving the use of tools
to examine the impacts of policies in developing countries compared to its use
in developed countries. One is the implementation of RIA by INMETRO, which
basically relies on the use of CBA techniques to quantify economic impacts of
new programmes of conformity assessment; and the second is the use of CBA
to evaluate federal programmes for prevention, control and eradication of pests
and that was proposed as the analytical framework in the scope of a big research
and technological project sponsored by CNPq and led by the Federal University
of Viçosa, in which the Center for Advanced Studies on Applied Economics
(CEPEA) was invited to develop this specific topic.

The first initiative relates to INMETRO’s efforts in quantifying impacts of its
own regulatory role in terms of technical measures and conformity assessment
programmes. INMETRO belongs to the Ministry of Development, Industry and
Foreign Trade and is the enquiry point of the TBT Agreement in Brazil. This
interface with the international negotiations on technical issues has certainly
contributed to make INMETRO a pioneer in terms of creating and implementing
regulatory impact assessments inside its activities, in Brazil.
Its Directory of Quality faces a large number of demands to establish rules,
technical regulation and new conformity assessment programmes, both emerging
from government and from private sectors. At the same time, INMETRO has
been following the most recent international developments in this area, and
consequently there was a natural motivation to implement RIA as a procedure to
evaluate the conformity assessment programmes.
This is important to harmonise procedures in Brazil with other trade partners,
even though some of these procedures are not yet compulsory in terms of
international relations.
Although the Directory of Quality had already developed and adopted a
procedure to have reports checking the technical viability of new conformity
assessment programmes, INMETRO’s staff has noticed that a world trend
towards having impact analysis would sooner or later challenge its regulatory
activities. So, in a pro-active action, INMETRO has invested initially in
consulting international experts and evolved to develop its own RIA model,
similar to what other OECD countries had already done. This process started
consolidating in 2008.
In order to adjust the RIA model proposed by the OECD to Brazilian
agencies and specifically to be applied in evaluating programmes of conformity
assessment, in 2009 INMETRO established a partnership with CEPEA, which
is a research group that belongs to the University of São Paulo (USP). The
main goal was to have a guide to apply the RIA to conformity assessment
programmes, but that additionally, comprised also procedures to support the
process of ranking regulatory priorities and considered the particular features of
Brazilian institutional structure and capacity building.As INMETRO already had
a very good system to evaluate technical viability of technical regulations and
conformity assessment programmes, and was also investing in hiring economists
to compose its team, the proposal to work in a partnership with CEPEA aimed
at improving in particular the knowledge and practical application of economic
tools to quantify and analyse impacts, as well as to better examine also social
and environmental effects of regulations.
The RIA system proposed by CEPEA was called RIA-S – Regulatory Impact
Assessment with Sustainability – which was based upon the core model of RIA
fromtheOECDbutaddedsomenewelementsfromthesustainabilityassessments
proposed also by the OECD in more recent studies. The model proposed by
CEPEA comprised six basic steps: triage (to help ranking priorities using statistic
tools), scope delimitation (definition of problem, alternative scenarios, potential
impacts, agents affected – step of collecting secondary data and determining the

need of surveys to collect other data), collecting primary data (implementing the
collection of primary data), measurement of impacts (application of methods
chosen to quantify impacts or to describe them qualitatively), identification of
synergies and trade-offs (analysis of results, considering the political aspects and
infra-structure available to implement the scenarios), communication of results
(elaboration of final dossier).
The main contributions were to make INMETRO’s staff advance in terms
of evaluation of economic impacts of its regulatory action and to introduce
analysis of social and environmental aspects as well. Despite the fact that most
of the social and environmental impacts cannot be quantified yet, the proposed
methodology considers the use of qualitative analysis to handle them, taking
those effects also into consideration to have a final diagnostic.
The RIA method adjusted to INMETRO needs was proposed at the end of
2009 through a Guide of RIA-S (Miranda et al, 2009), and later, in May 2010,
this Guide was used as a reference to update an internal instruction issued by
INMETRO (NIT-DIPAC – 014), defining not only risk analysis requirements but
also economic impact analysis as part of the viability and impact study required
to justify the proposition of new conformity assessment programmes.
INMETRO has been applying RIAs in an ex-ante approach for conformity
assessment programmes and this experience shows that it is necessary to have
a very good delimitation of the target of new regulation, as well as an efficient
identification of the potential impacts, in the sense that those that should be
considered are supposed to be the most effective and relevant impacts. So, the
implementation of RIA has started, step by step, and is still in progress. This
gradual implementation has been important to identify bottlenecks and propose
solutions to deal with them. For example, re-evaluating the number of people
and their qualification to be part of the technical staff in order to apply RIA and
amplify the scope of impacts considered.
The process of developing the Guide of RIA-S to INMETRO provided an
opportunity to have a better understanding about the difficulties of implementing
this kind of tool to examine public policy quality in developing countries.
Some points that became a constant concern during the development of the
methodology can be mentioned. First of all, it is important to consider the staff
available to conduct RIAs and the specific skills required to do it, particularly
to allow dealing with quantifying economic, social and environmental impacts.
Although the technical staff of public agencies are, in general, highly qualified
in Brazil, the number of staff and the specific knowledge required to work
with RIAs studies might become a caveat. One example of this difficulty is the
possibility to apply different econometric and statistics methods to valuation of
impacts and that require very specific knowledge, like the Willingness to Pay,
used to estimate demand of goods or services when there is not a formal market
organised.
Moreover, staff size and time to conduct the studies have a negative correlation,
in general. If deadlines are short to provide studies of RIA by agencies, it is
supposed to have enough people involved in the work to deal with that or,

otherwise, it is important to delimitate the scope of the analysis to match it with
the available resources to conduct the studies. If one agency intends to apply
sophisticated methods to quantification and collect data to supply such studies
it will be certainly necessary to have higher investments in collecting data and
in having skilled people to work with them using more complex economic and
statistical methods.
A second important point to make about the requirements to succeed in
establishing regulatory impact assessments is taking into consideration the
administrative structure of the agency, in terms of hierarchy, tasks sharing and
competences. Very often, in developing countries, there are many departments
or divisions committed with different tasks, but they can be related in terms of
building the RIA studies. It is important to have all these people and their leaders
following a mutual goal and guarantee that excessive paperwork and political
aspects do not jeopardise their communication.
A third and very restrictive factor to implement RIA in developing countries
refers to the scarcity of databases to provide enough information to estimate
impacts. There is a direct correlation between the quality of analysis and the
availability of data and good quality information. For example, if it is identified
that a certain regulation will have benefits in reducing the occurrence of health
problems, like cases of contamination by agricultural chemicals, which is a
social and economic effect, it is necessary to have initial information about the
number of occurrences of this problem in order to have an accurate estimate of
the expected benefit. However, only a few countries can currently count on such
kinds of statistics. One strategy is to adopt parameters and other data available in
foreign countries, usually developed ones, although very often this information
does not reflect the reality of developing or less developed countries.
And finally, another important factor to be considered and related to all the
above-mentioned points, is financial support. It is well known that improvements
in staff, databases and infrastructure are dependent on the availability of financial
resources. In developing countries there are usually restrictions on budget, as
governments face major problems to focus on, such as growth and employment
promotion, reduction of poorness and improvement of income distribution.
Regulatory quality and other public management issues are not usually ranked
as priorities. So, competition for financial resources might compromise the
development and implementation of new tools regarding RIA, even though these
tools could provide gains in optimising the allocation of governmental funds.
In terms of political aspects, there are two points to emphasise. First of all,
it is essential that agents involved in RIA are convinced about the importance
and benefits of this tool and motivated to implement it. This is so, because the
process of implementing RIA or any kind of impact assessment will require
the determination of the whole chain of empowered people involved in it. As it
will also be necessary to have more and better communication between different
governmental agencies, a political desire to have these evaluations should prevail
from the highest positions to the lowest in the ranking of officials in order to

provide the support, technical and political, necessary to accomplish the RIA
goals.
A second political aspect to be considered is that the model of OECD for
RIA, and also the one proposed to INMETRO, counts on a wide participation
of stakeholders through the whole process of developing the RIA studies.
RIA allows for a consultancy process at every juncture, from the definition of
regulations that should be first submitted to a RIA up to the validation of the
process at the end of the studies, when the results are brought to the awareness
of society to be discussed and validated, before the regulation is issued by the
government. The greatest advantage of conducting a transparent process of RIA
is that when the regulation is finally enforced the main sectors of society affected
are supposed to have already participated and be conscious about it. They have
had opportunity to propose, provide information, criticise and get prepared to
face the new regulation, anticipating its effects.
The second initiative that is highlighted here relates to a study developed
to apply the CBA to evaluate impacts of public policy in Plant and Animal
Health, focusing on programmes of prevention, control and eradication of
pests. This study was conducted by CEPEA (Miranda et al, 2010), sponsored
by a CNPq project entitled ‘Inovação Tecnológica para Defesa Agropecuária’
(Technological Innovation for Agriculture and Livestock Safety). This CNPq
project aims to have a diagnostic of the agricultural safety service in Brazil
and opportunities to improve it and one of the goals proposed was to develop
and discuss a methodological approach that could be replicated to analyse the
effectiveness and the net benefits of sanitary and phytosanitary policy in Brazil.
In this sense, having the information on the benefit-cost relations provided by
public policies might become a strategic tool to determine priorities for budget
allocation. Moreover, society has a poor understanding about the importance of
controlling diseases and pests’entrance and dissemination in the territory, which
makes it even more difficult to mobilise the general population, consumers and
producers, to cooperate with the sanitary policy. This is so probably because
although costs are usually identified and quantified by those agents that incur
them, the benefits (economic, social and environmental) of safety policy are
less transparent, diffuse and hardly quantified. Therefore, it is more difficult to
justify public expenses in this area.
So this study is included in the scope of CNPq’s project to provide more
information, technical and science based, about these costs and benefits
regarding sanitary and phytosanitary policies and to establish a method that
could be replicated easily by agents and experts to examine different cases, and
eventually be part of the Pest Risk Analysis, which is conducted by the Ministry
of Agriculture in Brazil. The CBA method was chosen as a basic framework and
three study cases were proposed in order to illustrate its use. These study cases
were presented in the II Conference of Plant and Animal Safety, in May 2010,
with an audience of officials, sanitary agents, private companies, veterinarians,
agronomists, experts in different fields related to agriculture, cattle and safety

issues and several points were made to improve the methodology, the scope of
impacts analysed and the specific themes of study cases.
It is interesting to note that the application of this methodological framework
is an ongoing work, as some challenges are still to be overcome, most of them
related to the evaluation of social and environmental aspects of pests’ entrance
and of sanitary programmes. Likewise, there are also some methodological and
database challenges regarding the quantification of effects over public finance,
either in identifying the effective expenditures to support these programmes or
the impacts that diseases might cause to tax revenues.
Three study cases were chosen: the case of a fruit-fly (Bactrocera carambolae),
which is present in Brazil but restricted to one northern state, and that is being
controlled by a federal government programme to avoid dissemination to other
states; Avian Influenza, which has never occurred in Brazil and is being targeted
by the Ministry of Agriculture (MAPA) programme to prevent entrance; and
the Greening, which is a disease that spread very fast through São Paulo State –
the major citrus producer in Brazil, and the government, federal and state, used
to have an eradication programme, implemented through a partnership with a
private organisation (Fundecitrus) and cancelled in December 2009, particularly
affecting the inspections and eradication procedures, raising even more concerns
about the risk of dissemination of the disease.
Basically, the approach applied to evaluate the impacts of the programmes
mentioned above was the CBA method, which was adjusted to deal with sanitary
and phytosanitary issues. Some of the main points to make, which differ from the
usual approach of CBA applied to investment projects and other types of objects,
are: (i) it is recommendable to have a biological model that shows how the pest
is expected to behave in the country, its dissemination pattern, its severity in the
territory, what are the conditions that favour or restrict its expansion, and the
damages that it causes in such conditions, among others; (ii) the definition of a
time path that will be proper to conduct the CBA analysis; (iii) the discount rate
that should be used in the CBA in order to have results that reflect the fact that
sanitary policy also has social goals and not only that private costs and benefits
should be accounted for.
Basically, the first step in applying the CBA to analyse the MAPA programmes
was to collect the information about the pests and their behaviours; secondly,
different and possible scenarios were drawn, and basically two were taken – one
being ‘what if’ the disease or pest spreads in the territory and there is no public
programme to control it; alternatively, ‘what if’ there is a programme preventing
the entrance or spread of the pest? In the analysis, the agents and impacts of the
pest entrance and dissemination were identified in each scenario proposed. The
next step was to build the matrices to calculate the pattern of dissemination and
to estimate impacts along the time-path of projection, impacts on production,
commercialisation, exports or imports, jobs, small companies’ competitiveness,
government expenditures and revenues and so on.
The impacts were identified and valued directly whenever this was possible,
by having market prices and other information available. For example, the

effects of having the fruit-fly expanding to regions that export fruits to Europe
and to the United States would certainly result in reducing the export income.
So, considering the size of imports in countries that do not allow importing from
territories infected by the pest, the volume of exports lost was multiplied by an
average price. So, the loss effect was evaluated. The same rationale was applied
to estimate losses in production, because of lower productivity or reduction of
plants and for jobs.
However, other impacts had to be only qualified as the environmental impact
that the growing number of pesticides application in citrus production is causing
to other agriculture activities, like production of honey, by killing bees that
pollinate the orchards. Or, the social losses that will arise from having a reduction
of poultry exports in regions that depend basically on this activity and that will
become impoverished by an eventual Avian Influenza outbreak.
The losses estimated for each case, in the scenario that considered the presence
of public policy to control or eradicate the pests were taken as the avoided losses
and so, indirectly, they represent the gains or benefits of maintaining MAPA’s
programmes. These gains were compared to costs of those programmes, which
were assumed to be only the direct expenditures of federal government to
conduct them. Unfortunately, due to database restrictions, it was not possible
to include also the expenses made by government with the general services of
inspections and laboratories that are used by the programmes analysed but are
also benefitting other governmental programmes. The comparison between
avoided losses and costs of programmes resulted in net benefits, which were all
monetised, and allowed calculating benefit-cost relations for each case.
Despite the difficulty of estimating all the potential impacts and the
simplifications in scenarios and assumptions, all the benefit-cost relations
calculated were found positive, showing that expenditures of government in
those programmes are benefiting society and that it is worth keeping these
programmes. Miranda et al (2010) calculated these relations of benefit-cost,
which ranged from a ratio of 26.4 to 92.4 for the Bactrocera case (considering
mangoes, citrus fruits and guavas); from 477.4 to 636.3 for the Avian Influenza
programme; and 72 to 85.8 in the case of Greening. Of course, these numbers
vary according to the time-path analysed, the assumptions on dissemination
patterns and on impacts, discount rates and many other variables.
A difference between this application and the one of INMETRO, is that in the
former the CBA was conducted as an ex-post evaluation, while INMETRO’s
RIA was first designed to be implemented as an ex-ante tool.4
Nevertheless,
it is interesting to comment that also for the sanitary policy there are several
reasons and opportunities to use the CBA as an ex-ante approach. It is known
that many pests are present in countries that share borders with Brazil and the
future MAPA programmes of sanitary defence could use previous information
4 It is important to emphasise and make clear that INMETRO has already implemented impact
evaluation of conformity assessment programmes, while the sanitary case is only illustrating a
research initiative to propose a model that could be eventually implemented by governmental
agencies.

about potential impacts in order to guide its decisions in terms of investments in
new programmes.
Brazilian private sectors could also benefit from having these kinds of results
available, to help in making decisions on technology and research investments
(in the chemicals sector, for example), or on capacity building of human capital
(agricultural economists, for example). Strategic plans in private sectors, as well
as public planning for governments, could take into consideration the expected
challenges identified by studies applying the ex-ante approach.
References
Calvin, L and Krissoff, B (1998), ‘Technical Barriers to Trade: A Case Study of
Phytosanitary Barriers and U.S.-Japanese Apple Trade’, Journal of Agricultural
and Resource Economics, 23(2), 351–66.
Miranda, S H G, Bartholomeu, D B and Lima, L M (2009), ‘Guia para Avaliação de
Impacto Regulatório com Sustentabilidade (AIR-S)’, Research report.
Miranda, S H G, Nascimento, A M and Ximenes, V P (2010), ‘Uma aplicação da
Análise Benefício-Custo para políticas de defesa sanitária: alguns estudos de caso
para o Brasil’, Research report.
Organisation for Economic Cooperation and Development (OECD) (2004),
‘Regulatory Impact Analysis Inventory’, Notes by the Secretariat, Public
Governance and Territorial Development Directorate – Public Governance
Committee, Paris, 29th Session do Comitê, 15–16 April. Available at: http://www.
oecd.org/dataoecd/22/9/35258430.pdf.
Organisation for Economic Cooperation and Development (OECD) (2008),
‘Measuring Regulatory Quality’, Policy Brief, April. Available at: http://www.
oecd.org/dataoecd/38/13/40395187.pdf.
Organisation for Economic Cooperation and Development (OECD) (n.d.), Guide to
Sustainability Assessments.
Peterson, E and Orden, D (2006), ‘Linking Risk and Economic Assessments in
the Analysis of Plant Pest Regulations: The Case of U.S. Imports of Mexican
Avocados’, Contractor and Cooperator Report No 25, October.

15
How Can Risk Management Help
Enforce Technical Measures?
Christopher Grigoriou
15.1 Introduction
Two major trends have shaped international trade in recent years: deeper market
integration on the one hand, and lower tariffs on the other. At the same time, the
use of non-tariff and behind-the-border measures (NTM and BTB) has become
commonplace, reflecting the need for regulations that seek to protect consumer
health and safety. In 2006, UNCTAD created ‘The Group of Eminent Persons
on Non-Tariff Barriers’ in order to discuss, define, classify, collect and quantify
non-tariff barriers (Multi-Agency Support Team, 2009). An NTM classification
has since been established, adding a new category specifically for technical
measures which aim to address the sanitary, phytosanitary, and technical barriers
to trade. Sanitary and phytosanitary measures (SPS) have been devised primarily
with the agricultural and food markets in mind, and as such, received a great deal
of attention during the Uruguay Round of agricultural trade talks.1
Measures on
technical barriers to trade (TBT) cover other standards, technical regulations, as
well as compliance assessment procedures for all products.2
While these measures are welcome in terms of consumer health and safety,
they are often seen as substitutes for classic protectionist measures (tariffs), and
even as a potential impediment to trade because they could complicate border
clearance procedures. Efficient checks of shipments’ SPS and TBT compliance
is a costly undertaking. It requires not only the involvement of numerous skilled
personnel, but also coordinated action from the various government agencies,
not to mention the sharing of information and practices between public and
private operators. Developing countries have rarely adopted such an approach,
leading to the use of inappropriate processes and the misallocation of compliance
management resources.As a result, clearance times increase and/or public health
and safety is comprised. Agencies in charge of monitoring compliance are
1 The GeneralAgreement on Tariffs and Trade (GATT) leaves their regulation to sovereign governments
as long as they are neither discriminatory nor connected to internal taxes or regulations.
2 Thirty-seven per cent of the TBT regulations reported to the WTO in 1997 had human health as their
objective (see SPS agreement training module chapter 9, available at: http://www.wto.org/english/
tratop_e/sps_e/sps_agreement_cbt_e/c9s3p1_e.htm).

increasingly faced with a trade-off between trade facilitation and public health
and safety concerns, rather than the revenue maximisation trade-off typically
faced by customs agencies. The inspection of most (if not all) SPS- and TBT-
related transactions prolong clearance times,3
yet the failure to identify fraudulent
transactions can have dramatic consequences for public health and safety.
The purpose of this paper is to demonstrate how risk management can help
SPS and TBT administrations handle this trade-off more effectively by focusing
attention and resources on the riskiest transactions. Thanks to optimised and
targeted checks, inspection opportunity costs will fall and the efficiency of
inspection personnel will rise. Here, risk management can rely on one of three
principles.Thefirstistherandomrule,wherebyagivenpercentageoftransactions
are inspected on a random basis. The second is the selectivity rule, whereby
transactions are chosen for inspection according to predetermined criteria or
modalities. The third is a risk-based analysis, whereby transactions are selected
for inspection according to their risk profile. The latter is established from an
assessment of the probability that a false declaration has been made, derived
from historical data pertaining to the transaction. Of the three approaches, only
the risk-based analysis fulfils what will be called hereafter ‘the golden rules of
risk management’. According to these rules, the procedure must be: (i) objective
(non-arbitrary) and standardised; (ii) dynamic and evolutive; (iii) not decodable
by economic agents; and (iv) implemented through a computerised process in
accordance with the revised Kyoto Convention.
This paper develops and details a risk-based econometric analysis that should
ensure the systematic targeting and risk-profiling of transactions.An econometric
method will serve to identify those elements of the declaration that are key to
assessing the risk of non-compliance. We then demonstrate empirically that
fewer inspections lead to better detection rates. Using hypothetical data we show
that limiting the inspection rate to under 30% will lead to the identification of
90% of fraudulent declarations, while an inspection rate of 50% for lower-risk
transactions will lead to the detection of only less than 2% of infringements.
Section 15.2 discusses how risk management can enhance and modernise
SPS and TBT administrations and help them manage the trade-off between
trade facilitation and TBT-compliance inspections more effectively. Section
15.3 gives reasons why risk management solution must be performed through
risk-based assessments, according to the golden rules of risk management.
Section 15.4 shows that econometrics is the appropriate basis for a dynamic and
objective risk-based approach to compliance inspections. Section 15.5 explains
how the risk-based approach optimises processes by demonstrating that fewer
inspections mean a better detection rate. The conclusions are detailed in Section
15.6.
3 Djankov et al (2006) show that a one-day delay in shipment leads to a trade reduction of at least 1%.
The rate can rise to 7% if the exports are agricultural products.

How Can Risk Management Help Enforce Technical Measures? 285
15.2 Risk-based analysis as a modernisation tool for SPS and TBT
administrations
15.2.1 The need for modernisation
AnincreasinglycomplexenvironmentmeansthatSPSandTBTagenciesareoften
powerless to enforce and check the compliance of declarations and shipments
with adequate efficiency. Added to this is the fact that they tend to lack modern
IT technologies, not to mention poor coordination with high-tech customs’
agencies. Widdowson and Holloway (2011) point out common weaknesses
in the compliance inspection procedures of SPS agencies. Besides inadequate
information technology resources, systematic data gathering and feedback from
inspections are rare. This hampers the establishment of risk profiles, which in
turn leads to higher-than-necessary inspection rates. These shortcomings also
favour recurrent bias in interventions aimed at generating revenue from fees and
informal payments arising from corruption and rent-seeking.
Given that SPS agencies are much smaller than customs’ agencies and do not
have the same level of modern technology at their disposal, particularly an IT
infrastructure, there is an expectation that customs’ agencies could help build
the capacities of SPS agencies. For example, much of the data required for the
risk-based analysis of technical measures are already integrated in the customs’
management system. Moreover, cooperation would be in the interests of both
the private and public sector, as it would reduce clearance times and increase the
effectiveness of inspections. To date, however, cooperation between SPS services
and customs has tended to be the exception rather than the rule.Van der Meer and
Ignacio (2011) report that capacity-building projects between customs and SPS
agencies are virtually non-existent and that their relationship, in certain cases,
can border on the adversarial.Administrative processes, such as declarations, the
payment of duties, data gathering, quarantine agencies and border police, tend to
be duplicated and performed in isolation. In addition, Widdowson and Holloway
(2011) point out that the risk management process is generally not an integral
part of the management framework of the given organisation and, as such, is not
systematically implemented and incorporated within border management.
15.2.2 The need for a risk-based approach
When checking compliance with technical measures, quarantine officers focus
first on the examination and tracking of the necessary paperwork from which
they extract all the relevant statistics and information.Afterwards, they may carry
out an inspection or tests on either the entire transaction or a sample thereof.
Finally, if an infringement is found, the transaction is forwarded for destruction,
quarantine or treatment. In all other cases, it is sent on for release or clearance.
Many countries do not use the results of a risk-based analysis to decide on which
channel the transaction should be directed towards. This means that they carry
out more inspections and laboratory tests than required. Furthermore, they tend,

in the interests of revenue maximisation, to bias inspections in favour of formal
enterprises rather than high-risk producers. This is obviously a lose-lose system
as neither trade facilitation nor public health and safety benefit. Increases in
the incidence of unnecessary inspections have resulted in higher costs for the
private sector. Moreover, it is not only impossible but also counter-productive
for the agency to inspect and check every transaction. With limited resources,
the opportunity costs of inspecting both low- and high-risk importers are
considerable. Concentrating time and resources on compliant importers limits
the resources available for the inspection of high-risk importers. The adoption
of a risk-based approach would reverse this situation by effectively tackling the
trade-off problem associated with the enforcement of technical measures. Some
customs’ and tax agencies have successfully adopted such practices, generating
benefits for both the public and private sector.
15.2.3 More to risk management than risk assessment
World Customs Organisation (WCO) guidelines4
recommend that the risk
management process is implemented in five stages: (1) context analysis; (2)
risk identification; (3) risk analysis; (4) risk assessment and prioritisation; and
finally (5) treatment and evaluation as described in Figure 15.1 from the Risk
Management Guide (WCO, 2003).
Figure 15.1 Original WCO Risk Management Framework
Address risks
Establish the context
IdenƟfy risks
Analyse risks
Assess and prioriƟse risks
Monitor and review
compliance
measurement
Source: World Customs Organization (2003).
Each stage provides insights that are used to inform the subsequent stage. It may
take some time before the process is implemented effectively due to the gradual
improvements that must be made in order to establish and update the fraud
database. Risk-based processes determine what action (such as sample tests and
full inspections) should be taken on each transaction according to its risk profile.
4 WCO guidelines to its Revised Kyoto Convention (International Convention on the Simplification
and Harmonization of Customs Procedures), available at: http://www.wcoomd.org/Kyoto_New/
Content/content.html.

This is established from an assessment of the probability of non-compliance,
which is derived from historical data pertaining to the transaction, that is, whether
or not elements of the declaration have been linked to previous cases of fraud. By
ensuring that the correct priority is given to the correct action by singling out the
riskiest transactions, risk-based processes lead to a more efficient allocation of
resources. Risk management is therefore not only a selection tool and router but
also greatly helps to increase efficiency by optimising resource allocation and
modernising the administrative structure through the use of modern technologies
and infrastructure. Moreover, risk management is per se an incentive to comply
with regulations. In the context of the principal-agent model with asymmetric
information, Alm et al (1993a, 1993b) show how risk-profiled importers have an
incentive to adopt compliant strategic behaviours. Risk management naturally
influences importers’ behaviour and advances compliance through inspections,
detection and sanctions (penalties or destruction of the merchandise). This
incentive will be further reinforced because the targeting strategy will concentrate
on high-risk declarations. Also, risk-management techniques generate business
intelligence as well as a reporting basis for the administration to assess how
efficient its compliance inspections are. Finally, given the lack of private-public
cooperation, inspections provide a good opportunity for the administration to
remind importers of their legal obligations and therefore may also serve as an
educational tool.
15.3 Risk-based approaches to risk profiling
We call the golden rules of risk management the requirements that a proper
profiling method must fulfil. First, the basic rules used to profile the risk of a
transaction must rely on an objective and standardised method. The aim here is
to avoid arbitrary decisions and exclusive reliance on human intervention as well
as to eliminate possible collusion or corruption opportunities. Second, given that
the world trade structure is constantly shifting, the risk management system has
to be dynamic and evolutive. As previously stated, the implementation of a risk
management system will imply changes in strategic behaviour as well as new
incentives that need to be considered. Third, the rules underlying specific risk
profiles should not be decodable by the importers so as to prevent them by-passing
the rules. Finally, a risk management system should be implemented through a
computerised process, in accordance with both the revised Kyoto Convention
and the recommendations of international institutions on modernising processes
through non-intrusive and standardised methods.
Depending on their level of modernisation, customs’ agencies can choose
from the selectivity rule, the random rule or risk-based selection when deciding
on which transactions are to be inspected. Indeed, the risk-based approach is
part of the modernisation programme for transition and developing countries,
devised by international institutions (see among others Widdowson, 2005). The
selectivity rule refers to a set of specific criteria on which the future route of the
declaration is determined. These draw primarily on the inspectors’knowledge of

importer behaviour and background. For instance, there will be a special focus
on all imports from countries A and B, or imports with specific HS codes will
be uniformly considered as high risk.5
While this technique might appear fit for
purpose, given that it requires only a limited amount of data and relies on local
knowledge, it does have major shortcomings: (i) it increases the risk of collusion
or corruption between the inspector and importer since the former manages the
information and therefore the selectivity criteria may be chosen arbitrarily; (ii)
the criteria to be used to route the declarations to the high-risk channel can
be decoded by non-compliant actors; (iii) it is not dynamic because the fraud
patterns will change as soon as non-compliant importers adapt their behaviour to
the monitoring strategy and because the selectivity rule is, by definition, static.
These drawbacks mean that this approach is of limited interest. However, it
could be used as a complement to the risk-based approach due to the fact that it
takes into account intelligence on a specific cargo or declaration which has been
gathered from cooperation with other services, for example.
As the name suggests, the random rule means that transactions are randomly
selected for inspection. This method does not require a sophisticated IT
infrastructure and ensures that importers are treated equally. Given that the
inspectionlikelihoodisthesameforallimporters,risksofcorruptionandarbitrary
selection are significantly reduced (providing that random inspections actually
are random). In addition, it serves as an incentive for compliant importers to
remain so, because they are just as likely to be subject to an inspection. This last
point is in contrast to the selectivity rule, which would never choose a ‘known’
compliant importer for inspection. Finally, the random selection method is
capable of detecting new types of infringements. However, it also has major
limitations. First, it is not dynamic, by definition. Its randomness means that
there is no opportunity to use information gleaned from previous infringements
and compliance patterns. Moreover, it generates substantial opportunity costs
because the high-risk declarations will have the same likelihood of being selected
for inspection as low risks. Ultimately, this leads to an inefficient allocation of
resources. These shortcomings, therefore, limit the interest of the random rule.
However, it could be used to complement a risk-based approach due to the fact
that it checks the consistency of the low-profile declarations, a common concern
for all risk management systems.
In recent years, modern administrations have increasingly implemented risk-
based selection methods to profile, target and inspect non-compliant declarations.
The thinking behind this practice is to attribute scores to each declaration with
a view to ensuring that the lion’s share of inspection resources are concentrated
on declarations that present with the highest risk of non-compliance (see
Geourjon and Laporte, 2005). The scores depend on the declarations’ attributes
and the infringement patterns observed over past transactions. This type of risk
profiling is dynamic because it is continuously updated to take account of new
5 This method is implemented via Asycuda, the custom management system developed by the United
Nations.

information and changing patterns. Moreover, if the scoring method is based on
a scientifically objective approach, it is by definition non-arbitrary and cannot
be decoded by importers. Finally, risk-based selection can, and should, be
combined with the selectivity approach to allow the system to gather intelligence
on specific cargos or declarations. The same applies to the random rule, as this
solution tracks low-risk profiles and thus sends a clear message to compliant
importers. The next section describes how econometrics can be used to run this
type of risk profiling.
15.4 Adaptable and objective risk-based methods from econometrics
Econometrics has many advantages for risk-based approaches, particularly as
regards technical measures and public health and safety issues, where the risk
profiles of declarations need to be ranked accordingly. Regression analysis first
filters the elements of the declaration (eg country of origin and the HS code),
or criteria as they will be called here, that are of relevance for the risk-profiling
process. It then weighs the scored criteria to arrive at the total score, which
will determine what risk profile will be attributed to the transaction: the higher
the score, the riskier the profile. Moreover, econometrics allows for several
approaches depending on data quality and the type of fraud. These approaches
come with a very large subset of methods, from linear to non-linear estimates
depending on the type of non-compliant behaviour that is being targeted or
profiled, including time or individual specific effects through panel modelling.
The observed (measure of) non-compliance is the dependent variable which
represents the behaviour we aim to model. This variable can be either categorical
or continuous. The explanatory variables are the scored criteria. Regression
econometrics defines which of these are likely to affect the dependent variable.
For example, SPS infringements may be influenced by the importer, the country
of origin, the HS code or the broker, but not by the packaging or the mode of
transport.
Apart from technology issues, the input for any risk management application
is the fraud database. Of course, because the purpose is to model non-compliant
behaviour, this must be observed. The database therefore should gather data
on the transactions, such as the importer code, HS code, country of origin
given in the declaration, as well as on inspection details like the release or
destruction of merchandise, or a penalty. This database is built up incrementally
and continuously updated to reflect possible changes in the behaviour to be
modelled. If no such database exists, the first step is to prepare and integrate one
in the administration system. Scores are then computed for each element of the
declaration accordingly with the priorities of the SPS and TBT administration.
These scores reflect the risk profiles resulting from the compliance history of
these elements. For instance, the score for each country of origin will be higher
if it has been frequently linked to non-compliant declarations. As the fraud
database is continuously updated through the inclusion of new declarations, so
too are the scores given to each element, thus ensuring a dynamic and permanent

adjustment of the process to any changes in fraud patterns. The next step is to
identify the scored criteria which can be used as effective predictors of fraud,
that is, the scored criteria that are significantly correlated to the fraud measure.
The regression analysis has two aims. It first defines which scored criteria are
effective predictors of fraud, that is, which criteria have a coefficient that is
significantly different from zero. These can then be used as risk profilers when
a new transaction comes up for assessment. Second, the regression analysis
provides the coefficient estimates that will be used to weight the scored criteria
in order to compute the total score. Once the regression has been run, each
new declaration will be risk-profiled from the scores of its modalities weighted
through the estimated coefficients to provide the total score for the declaration,
which determines the risk profile it will be given. It should be noted that the
use, albeit partial, of the selectivity and random rules will continue, the first in
order to draw on specific intelligence arising from inter-agency cooperation, the
second to provide compliant importers with incentives to remain so.
Once the predictions have been made, econometrics can still bring considerable
added value to the system because it makes it possible to ‘go beyond the black
box’, providing evidence on the relationships (causality) between the observed
infringement and each criterion as well as on the contribution of each criterion to
theoverallriskprofile.Moreover,italsogathersscoredcriteriaaswellasstatistical
results and predictions, allowing for the detailed analysis required for reporting
purposes or for evaluating system performance. Finally, the performance and
consistency of the system can be assessed by means of contingency tables. This
reporting step is vital because it is a powerful demonstration of the need and
added value of a risk (-based) management approach by validating the accuracy
of the model.
Table 15.1 matches the predicted and actual status of the declarations
(compliant versus non-compliant) and presents type 1 and 2 errors as well as
the percentage of correct predictions, which itself is a measure of the model
accuracy rate. Prediction efficiency measures the percentage of non-compliant
declarations that the model predicted correctly. The strike rate measures the
percentage of non-compliant declarations that will be detected when non-
compliant predictions are controlled. Consequently, a trade-off results between
efficiency and strike rates: prediction efficiency must be reduced to improve the
strike rate and vice versa (see Gupta and Nagadevara, 2007). Indeed, improving
the strike rate implies lowering the number of inspections in order to reduce the
number of false positives (erroneous non-compliant predictions, ie inspection of
compliant transactions), which may in turn lead to a rise in the number of false
positives (erroneous compliant predictions, ie no inspection of non-compliant
transactions), thus lowering the prediction efficiency rate.

Table 15.1 Contingency table for the predicted and actual status of the
declarations
Predicted
Compliance Infringement
Real
Compliance True negative False positive
Infringement False negative True positive
15.5 Application
This section aims to illustrate the added value of the risk-based method relying
on econometrics to target non-compliant declarations and to optimise inspections
through a set of fake data generated for the purpose of this exercise.
15.5.1 Data
This application is run over the following hypothetical data which are based on
100,000 observations. The assumptions underlying this database are that 24.8%
of declarations are non-compliant and are consequently considered fraudulent.
The following elements of the declarations are considered as potential criteria
to target the declarations to be inspected. (i) The internationally standardised
system of names and numbers for classifying traded products, (HS codes): 7,406
kinds of goods (ii) the trade partners: 130 countries of origin; (iii) the importers:
8,364 TIN numbers; (iv) the brokers: 1,606 brokers; (v) the final destination:
are the goods intended for consumption, for trans-shipment or for a warehouse?
(vi) the entry site: 14 different sites are considered as potential entrance points
in the importing country; (vii) the packaging: are the goods from the declaration
packed in cartons, traditional packaging or something else? The description and
number of associated modalities for each criterion are reported in Table 15.2.
Table 15.2 Criteria used to predict fraud – 100,000 observations
Criteria No. of modality Criteria description
HS code 7,406 Item HS classification
Country of origin 130 Country of origin
Importer 8,364 Importer name
Final destination 4 Consumption/trans-shipment/warehouse
Entry port 14 Entry port identifier
Broker 1,606 Broker name
Packaging 3 Cartons (CT)/packages (PK)/others
Figure 15.2 describes the main features of these elements from the declarations;
no more than 40% of the HS codes and country of origin were compliant
(Figure 15.2a and b). Almost 50% of importers were linked to at least one
infringement (Figure 15.2c), while more than 80% of brokers were linked to
at least one fraudulent declaration (Figure 15.2d). Regarding the destination of

the goods, approximately a quarter of goods intended for consumption (Figure
15.2e, column C) were linked to at least one infringement, which is somewhat
higher than the share observed for goods in transit (T) or for warehouse (W).
Regarding packaging, goods packed in cartons (CT) were more often associated
with fraudulent transactions than other packaging schemes. Finally, of the 14
sites defined as entry points, two had never been linked to any kind of fraudulent
declaration, nine have had less than 10% of fraudulent declarations and, in three
sites 10% of imports were non-compliant.
Figure 15.2 Declaration elements and infringements
0%
10%
20%
30%
40%
50%
0 1 2 and
more
0%
10%
20%
30%
40%
50%
0 1 2 and
more
0%
10%
20%
30%
40%
50%
60%
0 1 2 and
more
0%
20%
40%
60%
80%
0 1 2 and
more
(a) Distribution of the HS codes
per number of infringments
(b) Distribution of countries of origin
per number of infringments
(c) Importer distribution per
number of infringements
(d) Broker distribution per
number of infringements
0%
20%
40%
60%
80%
100%
C T W
Fraudulent Compliant
0%
20%
40%
60%
80%
100%
CT PK Others
Fraudulent Compliant
0
2
4
6
8
10
0% 0-10% 10%
(e) Infringement distribution
per destination
(f) Infringement distribution
per type of packaging
(f) Port distribution per %
of fraudulent declarations
15.5.2 The limited dependent variable model
The logit and probit estimators are specifically designed for binary dependant
variable models (see Laporte, 2011). Because a regression with a binary
dependent variable Y models the probability that Y = 1, it makes sense to adopt
a non-linear formulation that forces the predicted values to be between 0 and 1.

Given that cumulative probability distribution functions produce probabilities
between 0 and 1, they are used in probit and logit regressions, whereby probit
regressions, which will be used here, use the standard normal cumulative
distribution function. This implies the use of the maximum likelihood method
to estimate the weighting of the explanatory variables in order to maximise the
chances of correctly predicting the dependent variable.6
The binary variable identifies two different cases, the probability of which
we want to assess. The added value of this method is that it covers the entire
consistent set of information related to the targeted event and to associate it with
an assessed probability. For our purposes, we differentiate between fraudulent
and compliant declarations with a view to predicting the probability that a new
declaration will be fraudulent. In other words, the probit regression model is:
P(Y = 1/X1
, X2
, …, Xn
) = ϕ(β0
+ β1
.X1
+ β2
.X2
+ … + βn
.Xn
)
P(Y = 1) is the probability that the declaration is fraudulent. ϕ is the cumulative
standard normal distribution function. X1i
, X2i
, …, Xni
are the scored modalities
of the associated criterion, accordingly with Section 15.4. Here, we have X1i
, X2i
,
X3i
, X4i
, X5i
, X6i
, X7i
, which are the scored HS code, country of origin, importer,
final destination, entry point, packaging and broker of the declaration (item i),
in that order. Finally, β1
, β2
… βn
are the parameters we wish to estimate. These
parameters reflect the impact of any change in X (the scored modalities) on the
probability of non-compliance.
15.5.3 The regression
Table 15.3 reports the econometric regression.
Table 15.3 Fraud measure on the scored criteria
Dependent variable Infringement
Importer 0.67*
Country of origin 3.23*
HS code 1.01*
Final destination –2.21
Entry port 0.76*
Broker 0.16*
Packaging 0.53
Constant –2.26*
Obs 90,760
Estimator Probit
Log likelihood 16,004
Pseudo-R2 0.59
Note: * Refers to a coefficient significant at the 10% level from robust standard errors
6 Instead of minimising the errors as in the linear case.

This regression provides four insights for the risk-based assessment of
declarations. The first two results from the regressions table (Table 15.3 – fraud
measure on the scored criteria). (i) It indicates which criteria are significant
fraud predictors, in other words the criteria that should be applied in order to
select which goods will be inspected. For these particular criteria, the likelihood
of fraud will rise, the higher the scored modalities are. It appears from Table
15.3 that only the HS code, country of origin, importer, entry port and broker
are good predictors of fraud. (ii) It provides the weights to be used to compute
the probability of fraud (pr[Yi
= 1]) associated with a specific declaration (βi
)
through the normal cumulative distribution function ϕ. (iii) The third element
provided by the econometric regression is the contingency table to figure out the
consistency of the modelling (Table 15.4). Matching the predicted fraudulent/
compliant cases, in other words the predicted 0 and 1 to the actual fraudulent/
compliant declaration, we can compute the following contingency table, which
provides evidence on the quality of the predictions.
Table 15.4 Contingency table
‘Real’ fraud
Predicted fraud
Total
0 1
0 70% 5% 75%
1 4% 21% 25%
Total 74% 26% 100%
It is possible to compute from this table several parameters that characterise the
quality of the predictions. We can also conclude from the accuracy rate that the
model performs well, given that the success rate as regards the predictions is
91%. Moreover, by reporting the number of true positives (21%) as a share of the
total number of fraudulent declarations (25%), we obtain a measure of prediction
efficiency (84%). Finally, the strike rate (81%) is defined as the number of true
positives (21%) in all cases predicted to be fraudulent (26%).
Ordering the declarations by decile of level of assessed risk demonstrates that
focusing on the riskiest transactions is the best way to optimise the inspection
process. The data are first ordered by risk of fraud (assessed probability) and
then classified by decile, with the first decile corresponding to the riskiest 10%
of transactions, and so on. For each decile, the corresponding number of actual
fraudulent transactions is computed. It is then cumulated to derive the percentage
of fraudulent transaction that would have been detected had we used this model
to direct and select them for inspection. Figure 15.3 shows that inspecting only
the first decile (10% of the riskiest transactions) would result in the detection of
almost 40% of infringements, while inspecting 20% of the riskiest would result
in the detection of almost three-quarters of infringements, and inspecting 30%
would lead to the identification of over 90% of the non-compliant declarations.
On the other hand, inspecting 50% of declarations that were predicted to have
the lowest level of risk would only detect less than 2% of additional cases of
fraud.

Figure 15.3 Distribution of actual infringements by deciles of predicted level of
risk
0.00%
20.00%
40.00%
60.00%
80.00%
100.00%
10% 20% 30% 40% 50%
These findings are all the more promising when we compare them to the findings
from other methods. Indeed, while we have seen that inspecting under 30% of
declarations would lead to the detection of more than 90% of infringements, a
random rule would require the inspection of 90% of declarations to achieve the
same rate of success.7
Moreover, the econometrics method is by and large more
convincing than the selectivity rule due to the fact that it is continuously updated
and adjusted to changing patterns and trends. As new infringements are detected,
the scored modalities change, as does the output of the econometric regression
(coefficients etc.).
15.6 Conclusion
This paper has shown how risk management can help SPS and TBT agencies to
handle increasingly complex regulations more effectively by focusing attention
and resources on the riskiest transactions. Risk-based analysis is the most
suitable method as it fulfils the golden rules of risk management, which require
the methodology to be objective (non-arbitrary) and standardised, dynamic and
evolutive, not decodable by the economic agents, and implemented through
a computerised process in accordance with the revised Kyoto Convention.
Moreover, econometrics is essential in this risk-profiling process as it identifies
the elements of the declaration that should receive the most attention and enables
the assessment of new declarations. We then demonstrate empirically that fewer
inspections can mean better detection rates. Using hypothetical data, we show
that limiting the number of inspections to under 30% of transactions will lead
to a 90% detection rate, while inspecting 50% of lower-risk transactions would
only detect less than 2% of infringements.
7 Given the probability of detecting fraudulent transactions with a random inspection (24.8%) and the
90% detection target rate (which corresponds to 22,390 declarations), one would need to inspect
90,000 of the 100,000 transactions.

References
Alm, J, Bahl, R and Murray, M N (1993a), ‘Audit Selection and Income Tax,
Underreporting in the Tax Compliance Game’, Journal of Development
Economics, 42(1), pp 1–33.
Alm, J, Cronshaw, M B and McKee, M (1993b), ‘Tax Compliance with Endogenous
Audit Selection Rules’, Kyklos, 46, pp 27–45.
Djankov, S, Freud, C and Pham, C S (2006), ‘Trading on Time’, World Bank Policy
Research, Working Paper No 3909.
Geourjon, A-M and Laporte, B (2005), ‘Risk Management for Targeting Customs
Controls in Developing Countries: A Risky Venture for Revenue Performance?’,
Public Administration and Development, 25, pp 105–13.
Gupta, M and Nagadevara, V (2007), ‘Audit Selection Strategy for Improving Tax
Compliance – Application of Data Mining Techniques’, in A Agarwal and V
Ramana (eds.), Foundations of E-government, ICEG.
Laporte, B (2011), ‘Risk Management Systems: Using Data Mining in Developing
Countries’ Customs Administrations’, World Customs Journal, 5(1), pp 17–27.
Multi-Agency Support Team (2009), ‘Report to the Group of Eminent Persons on
Non-Tariff Barriers’ (draft), September.
Van der Meer, K and Ignacio, L (2011), Sanitary and Phytosanitary Measures and
Border Management’, in G McLinden, E Fanta, D Widdowson and T Doyle (eds),
Border Management Modernization, Washington DC, The World Bank.
Widdowson, D (2005), ‘Managing Risks in the Customs Context’, in Customs
Modernization Handbook, Washington DC, The World Bank.
Widdowson, D and Holloway, S (2011), ‘Core Border Management Disciplines:
Risk Based Compliance Management’, in G McLinden, E Fanta, D Widdowson
and T Doyle (eds), Border Management Modernization, Washington DC, The
World Bank.
World Customs Organization (2003), Risk Management Guide, Brussels, WCO.

Non-Tariff Measures - A Fresh Look at Trade Policy's New Frontier

More Related Content

What's hot

Viewers also liked

Similar to Non-Tariff Measures - A Fresh Look at Trade Policy's New Frontier

More from Ira Tobing

Recently uploaded

Non-Tariff Measures - A Fresh Look at Trade Policy's New Frontier