D E V E L O P M E N T, T R A D E , A N D T H E W T O A Handbook Bernard Hoekman, Aaditya Mattoo, and Philip English, editors THE WORLD BANK Washington, D.C. © 2002 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 All rights reserved. Manufactured in the United States of America First printing June 2002 1 2 3 4 05 04 03 02 The findings, interpretations, and conclusions expressed here are those of the author(s) and do not necessar- ily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank cannot 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 on the part of the World Bank any judgment of the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is copyrighted. No part of this work may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or inclusion in any informa- tion storage and retrieval system, without the prior written permission of the World Bank. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint, 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, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pub- rights@worldbank.org. Cover and interior design by W. Drew Fasick, Serif Design Group, Washington, D.C. Cover photographs: background image, U.S. National Aeronatics and Space Administration. Inset images, clockwise from upper left: Edwin G. Huffman, the World Bank; Curt Carnemark, the World Bank; the World Bank; Edwin G. Huffman, the World Bank; PhotoDisc; Francis Dobbs, the World Bank. ISBN 0-8213-4997-X Library of Congress Cataloging-in-Publication Data has been applied for. Contents Foreword xi A Tribute to J. Michael Finger xiii Acknowledgments xix Contributors xxi Abbreviations xxiii Introduction xxvii Part I Trade Policy Reform in Context 1 1 Trade Policy Reform as Institutional Reform 3 Dani Rodrik 2 Economywide Dimensions of Trade Policy and Reform 11 Kym Anderson 3 Exchange Rate Overvaluation and Trade Protection 17 Howard J. Shatz and David G. Tarr 4 Fiscal Dimensions of Trade Liberalization 24 Liam Ebrill, Janet Stotsky, and Reint Gropp 5 Trade Policies for Poverty Alleviation 28 L. Alan Winters Part II The World Trade Organization 39 6 The WTO: Functions and Basic Principles 41 Bernard Hoekman 7 Reciprocity in the WTO 50 J. Michael Finger and L. Alan Winters 8 WTO Accession 61 Constantine Michalopoulos 9 Developing Countries and the WTO Dispute Settlement System 71 Valentina Delich 10 The Adequacy of WTO Dispute Settlement Remedies: A Developing Country Perspective 81 Robert E. Hudec Part III Selected Trade Policies Affecting Merchandise Trade 93 11 Market Access Issues and the WTO: An Overview 97 Sam Laird 12 Tariff Peaks and Preferences 105 Marcelo Olarreaga and Francis Ng iii D E V E L O P M E N T, T R A D E , A N D T H E W T O 13 Rules of Origin and Trade Preferences 114 Luis Jorge Garay S. and Rafael Cornejo 14 Nonpreferential Rules of Origin and the WTO Harmonization Program 122 Stefano Inama 15 Customs Valuation and Customs Reform 128 Vinod Rege 16 Trade Facilitation: Improving the Invisible Infrastructure 139 Brian Rankin Staples 17 Industrial Policy and Developing Countries 149 Mari Pangestu 18 Export Development Policies and Institutions 160 Philip English and Luc De Wulf 19 Trade-Related Investment Measures 171 Bijit Bora 20 Local Content Policies: Australia's Experience with Automobiles 179 Garry Pursell 21 Implementing the Agreement on Textiles and Clothing 186 Hanaa Kheir-El-Din 22 Safeguards: Making Sense of GATT/WTO Provisions Allowing for Import Restrictions 195 J. Michael Finger 23 Dealing with U.S. Trade Laws: Before, During, and After 206 Gary N. Horlick and Eleanor Shea Part IV Developing Countries and Negotiations on Trade in Services 213 24 Liberalization of Trade in Services in Developing Countries 221 James Hodge 25 The Magnitude of Flows of Global Trade in Services 235 A. Maurer and P. Chauvet 26 Quantifying Barriers to Trade in Services 247 Robert M. Stern 27 The GATS: Key Features and Sectors 259 Rudolf Adlung, Antonia Carzeniga, Bernard Hoekman, Masamichi Kono, Aaditya Mattoo, and Lee Tuthill 28 Negotiating Improved Market Access Commitments 280 Aaditya Mattoo 29 Domestic Regulations and Liberalization of Trade in Services 290 Carlo Gamberale and Aaditya Mattoo 30 Movement of Natural Persons and the GATS Major Trade Policy Impediments 304 Rupa Chanda 31 Electronic Commerce, the WTO, and Developing Countries 315 Catherine L. Mann 32 Completing the GATS Framework: Safeguards, Subsidies, and Government Procurement 326 Pierre Sauvé 33 Regional Liberalization of Trade in Services: Experiences in the Americas 336 Sherry M. Stephenson and Francisco Javier Prieto Part V Technology and Intellectual Property 347 34 International Technology Transfer and Economic Development 351 Kamal Saggi iv Contents 35 Implementing the TRIPS Agreement 359 Jayashree Watal 36 Benefiting from Intellectual Property Protection 369 Keith E. Maskus 37 Proprietary Protection of Genetic Resources and Traditional Knowledge 382 Arvind Subramanian 38 Cultural Industries and Intellectual Property Rights 390 Frank J. Penna and Coenraad J. Visser 39 Trademarks, Geographical Indications, and Developing Countries 403 Carsten Fink and Beata K. Smarzynska Part VI "Behind-the-Border" and Regulatory Issues 413 40 Multilateral Disciplines and Government Procurement 417 Simon J. Evenett 41 Standards, Regulation, and Trade: WTO Rules and Developing Country Concerns 428 John S. Wilson 42 Multilateral Disciplines and National Investment Policies 439 Bernard Hoekman and Kamal Saggi 43 Trade, Competition, and the WTO 447 Peter Holmes 44 Merger and Anticartel Policies in an Era of Integrating Markets 456 Simon J. Evenett 45 Trade and Labor: Text, Institutions, and Context 463 Simon Tay 46 Environmental Regulation and the WTO 472 Veena Jha Part VII The Trading System and Developing Countries 483 47 Developing Countries: Turning Participation into Influence 485 Diana Tussie and Miguel F. Lengyel 48 Implementation of WTO Commitments: The Development Challenge 493 J. Michael Finger and Philip Schuler 49 Special and Differential Treatment 504 T. Ademola Oyejide 50 Trade-Related Capacity Building for Enhanced African Participation in the Global Economy 509 David F. Luke Part VIII Trade Policy Questions and Guidelines 517 51 Rules of Thumb for Trade Policy 519 Constantine Michalopoulos, Maurice Schiff, and David G. Tarr 52 Arguments for and against Uniform Tariffs 526 David G. Tarr 53 Formula Approaches to Reciprocal Tariff Liberalization 535 Arvind Panagariya 54 Binding Tariffs: Why Do It? 540 Joseph F. Francois and Will Martin 55 Benefiting from Regional Integration 548 Bernard Hoekman and Maurice Schiff Appendixes 559 A Patterns of Trade and Protection: Selected Country Data 561 Francis Ng v D E V E L O P M E N T, T R A D E , A N D T H E W T O B Trade Indicators and Indices 585 Francis Ng C A Dataset on Trade and Production, 1976­99 589 Alessandro Nicita and Marcelo Olarreaga Glossary of Trade-Related Terms 593 Bibliography 605 Index 629 CD-ROMs 1 A Dataset on Trade and Production, 1976­99 Alessandro Nicita and Marcelo Olarreaga 2 Applied Trade Policy for Developing Countries: Outline,Content,and Readings for a Short Course Jaime de Melo and Marc Bacchetta Tables 3.1 Comparing "Great Depressions": Cameroon, Côte d'Ivoire, and the United States 21 4.1 Revenue Impact of Trade Liberalization 25 6.1 From GATT to WTO: Major Events 46 7.1 Control of Free-Riding in GATT Negotiations: U.S. Experience, 1947­67 53 7.2 U.S. Imports Covered by Kennedy Round Tariff Concessions (Reductions plus Bindings) as a Share of Total U.S. Imports from the Country Group 54 7.3 Tariff Concessions Received and Given at the Uruguay Round 57 8.1 Accessions to the World Trade Organization as of May 1, 2001 64 9.1 Number of Dispute Settlement Cases, 1995 through September 2000 76 11.1 Post­Uruguay Round Applied and Bound Rates of Industrial and Developing Economies by Major Product Group 98 11.2 Tariff Escalation on Products Imported by Industrial Economies from Developing Economies 99 11.3 Post­Uruguay Round Import-Weighted Applied and Bound Tariff Rates 99 11.4 Patterns of Protection in Manufacturing, 1995 102 12.1 Tariff Peaks and Imports, Quad Economies, 1999 108 12.2 Tariff Peaks and Preferential Duty Rates, Quad Economies, 1999 110 12.3 Effects of Granting Duty- and Quota-Free Access to Quad Markets to LDC Exporters 111 17.1 Evolution of Industrial Policies in East Asia, 1950s­1990s 153 17.2 Policies and Measures for Promoting Exports in Asia 155 21.1 Results of Integration Programs for ATC Stages 1 and 2 189 21.2 Number of Quotas Eliminated by Integration in ATC Stages 1 and 2 190 21.3 Restrained Trade Freed of Quotas, 1995­97 190 21.4 Expanded Market Access Attributable to Increases in Quotas (Stages 1 and 2 Combined) 190 22.1 Frequency of Use of GATT Provisions That Allow Trade Restrictions, 1948­94 196 IV.1 Summary of Selected GATS Negotiations and Domestic Policy Issues: Current Status and Desirable Outcomes 218 25.1 Inadequacies of Statistical Domains with Regard to Modes of Supply 237 25.2 Average Annual Growth of Services and Merchandise Trade, 1990­99 238 25.3 Trade in Commercial Services by Region, 1990­99 239 25.4 Share of Selected Economies in World Trade in Services, 1995 and 1998 239 25.5 Exports of Services of Selected Economies by Selected Partners, 1995 and 1998 240 25.6 Evolution of Transport, Travel, and Other Commercial Services, 1980­99 241 25.7 World Trade in Other Commercial Services by Category, 1996­98 241 vi Contents 25.8 Trade in Services by Mode of Supply, 1998 242 25.9 Shares in Inward and Outward FDI Stocks, Selected Country Groups, 1990 and 1999 243 25.10 Total U.S. Services Trade, 1994, 1996, and 1997 243 26.1 Constructed Ad Valorem Tariff Equivalent "Guesstimates" by One-Digit ISIC Services Sector, Selected Countries 250 26.2 FDI Restrictiveness Indices, Selected APEC Economies and Selected Services Sectors 250 26.3 Estimated Tariff Equivalents in Traded Services: Gravity Model­Based Regression Method 252 26.4 Average Gross Operating Margins of Firms Listed on National Stock Exchanges, 1994­96, by Economy or Region 253 26.5 Alternative Approaches to Modeling the Impact of Barriers to Trade and Investment 254 26.6 Welfare Effects of 33 Percent Reduction in Barriers to Trade in Agriculture, Manufactures, and Services 257 27.1 Main Provisions of the GATS 260 27.2 Format and Example of a Schedule of GATS Commitments 261 27.3 Number of Committed Services Sectors by Member, July 2000 263 27.4 Market Access Commitments on Insurance (Life and Nonlife) under the GATS 268 27.5 Market Access Commitments on Banking (Acceptance of Deposits and Lending of All Types) under the GATS 270 27.6 Current Status of GATS Commitments on Voice Telephone Services 273 28.1 Types of Market Access Restrictions on Commercial Presence in Services Sectors, All WTO Members, 2000 281 28.2 Developing Country Precommitments to Liberalize in Basic Telecommunications Negotiations, 1998 284 30.1 Commitments by Sector and Mode of Supply (Professional Services) 308 33.1 Specific Services Sectors Highlighted in the Subregional Integration Agreements of the Americas 342 V.1 Summary of Selected TRIPS Issues: Current Status and Possible Approaches from a Development Perspective 349 40.1 Performance Indicators for Government Procurement in Korea, 1993­98 424 41.1 Total Number of Countries That Have Established SPS Enquiry Points 435 44.1 Economies with Firms Convicted of Price Fixing by the United States or the European Commission during the 1990s 458 44.2 Cross-Border Mergers and Acquisitions in Latin America and Asia Pacific, 1991­98 459 48.1 Costs of World Bank Customs Reform Projects, Selected Countries 496 48.2 Costs of SPS-Related World Bank Projects 497 48.3 World Bank Projects Related to Intellectual Property Rights 499 54.1 Industrial Applied and Bound Tariff Rates 542 54.2 Unbound Tariff Lines by GATT Multilateral Tariff Negotiation (MTN) Category 543 A.1 Trends in Average Tariff Rates for Developing and Industrial Countries, 1980­99 562 A.2 Average Tariff Rates by Sector and FDI Inflows for All Countries in Recent Years 568 A.3 Tariff Escalation in Developing and Industrial Countries, 1994­2000 571 A.4 Pre­ and Post­Uruguay Round Nontariff Barriers for All Goods, Developing and Industrial Countries, 1984­93 573 A.5 Nontariff Barriers for All Products in Developing Countries, 1989­98 576 A.6 Tariff Peaks and Preference Margins, Canada, 1999 578 A.7 Tariff Peaks and Preference Margins, European Union, 1999 580 A.8 Tariff Peaks and Preference Margins, Japan, 1999 581 vii D E V E L O P M E N T, T R A D E , A N D T H E W T O A.9 Tariff Peaks and Preference Margins, United States, 1999 582 A.10 Comparison of MFN Applied Tariffs of Labor-Intensive Products, Selected Countries, 1997­99 583 Figures 11.1 Implications of a 40 Percent Reduction in Agricultural Trade Barriers 100 11.2 Share of Manufactures in Developing Country Merchandise Exports, Actual and Projected, 1965­2005 101 12.1 Projected Changes in LDC Export Revenues as a Result of Duty-Free Access, by Product and Importer 112 22.1 Renegotiations, Emergency Actions, and Voluntary Export Restraints (VERs), 1948­93 198 22.2 Renegotiations, Emergency Actions, Antidumping Actions, and Voluntary Export Restraints (VERs), 1948­93 200 27.1 Number of WTO Members Committing to at Least Part of an Aggregate Services Sector, July 2000 264 27.2 Market Access Commitments by WTO Members on Movement of Natural Persons (Services Mode 4) 265 29.1 Approaches toward Domestic Regulation at the Multilateral and National Levels 291 33.1 Matrix of Possible Elements in Services Trade Agreements 343 39.1 Global Demand for Trademarks and Patents 408 39.2 Nationality of Trademark and Patent Title Holders by Country Income Group, 1997 409 41.1 Number of Notifications under the SPS Agreement by Region, 1995­99 435 54.1 Implications of a Tariff Binding for the Applied Rate of Protection 545 54.2 Welfare Implications of Tariff Bindings for a Small Country 546 55.A1 A RIA Involving Small Countries 557 Boxes 4.1 Export Taxes 27 5.1 Key Questions for Determining the Poverty Impact of a Trade Reform 29 6.1 Transparency: Notification and Surveillance 45 9.1 Special and Differential Treatment and the DSU: Some Examples 73 9.2 The Advisory Centre on WTO Law 75 10.1 The Bananas Case 87 12.1 The European Union's "Everything but Arms" Initiative 106 13.1 The U.S. Trade and Development Act of 2000 115 15.1 The WTO Agreement on Preshipment Inspection 132 15.2 The International Convention on the Simplification and Harmonization of Customs Procedures: 1999 Revision of the Kyoto Convention 135 16.1 Increasing the Effectiveness of Preshipment Inspection Services 143 16.2 Improving Export Clearance in Jamaica 145 17.1 Nominal and Effective Rates of Protection 151 17.2 Instruments of Industrial and Export Promotion Policies Employed in the Republic of Korea and Japan 152 18.1 Matching Grant Schemes to Promote Exports 162 18.2 Duty Drawback for Direct Exporters: The Chilean Case 164 18.3 Subsidies, the WTO, and Developing Countries 168 21.1 The Cost of Trade Protection of Textiles and Clothing in the European Union 193 22.1 The Basics of WTO Antidumping Rules 201 22.2 The Flawed Economics of Basing Decisions on an Injury Investigation 203 24.1 Realizing the Gains from Financial Liberalization 223 viii Contents 24.2 Contribution of Inefficient Internal Transport Systems to the Concentration of China's Export Industries in Coastal Regions 225 24.3 Lessons from Reforming Argentina's Ports 227 24.4 Welfare Gains from Services Liberalization: The Case of Tunisia 229 24.5 Achieving Effective Competition in Telecommunications: The Asian Experience 233 25.1 Services Statistics and the GATS 236 25.2 Improving Data on Trade in Services 244 27.1 Financial Services: Market Access Commitments of Developing and Transition Economies 267 28.1 Services Exports by Developing Countries: Potential Gains and Current Barriers 286 29.1 Whose Regulations, and for What Purpose? Sticky Issues in Electronic Commerce 292 29.2 Assessing the WTO Telecommunications Reference Paper 293 29.3 Rules for Domestic Regulations: The Experience with Accountancy 295 29.4 Challenges in Implementing Procompetitive Regulation 297 29.5 Financial Sector Liberalization: The Need for Policy Coherence 299 29.6 Financial Liberalization and Access to Credit by the Poor 301 29.7 Pursuing Social Objectives in Competitive Telecommunications Markets 302 30.1 Immigration Patterns in OECD Countries 309 31.1 The Economics and Law of Duty-Free Electronic Commerce 319 32.1 ASEAN Proposal for a GATS Emergency Safeguard Mechanism 330 35.1 An Overview of Intellectual Property Rights 360 35.2 Combating Disease Worldwide: Fostering the Required R&D 366 36.1 Pharmaceutical Policies and the TRIPS Agreement 374 36.2 Differential Pricing 378 36.3 Health Care Systems and Essential Medicines 380 40.1 Accession to the GPA: Identifying Costs and Benefits 425 41.1 SPS Measures: The Kenya Fish Exports Case 430 41.2 Standards-Related Problems Encountered by Indian Exporters 433 41.3 Moving Forward: A Proposal 438 42.1 WTO Rules on FDI: A Positive View 441 43.1 Why Competition Policy Should Not Be Seen through a Market Access Lens 450 43.2 Free Trade and Competition Policy 452 45.1 What Are Core Labor Standards? 464 45.2 Trade and Labor Standards: Three Debates 466 46.1 Basic Principles of Environmental Regulation 474 47.1 Café au Lait Diplomacy 489 48.1 Costs of Implementing the WTO Agreements: The Jamaican Experience 502 51.1 Export Diversification in Africa: Elements of a "Business Plan" 523 55.1 Selected Major Regional Integration Agreements and Dates of Formation 549 ix blank Foreword T oday, trade policy is at the fore- markets should be more open to the goods of poor front of the development agenda, countries, and that developing countries should and it is a critical element of any strategy to fight open their markets as well as address a range of poverty. This renewed interest in trade liberaliza- institutional issues. tion does not come from dogma but instead is based The advance at Doha presents a unique opportuni- on a careful assessment of development experience ty for development, but it will require substantive over the last 50 years. participation from all countries to succeed. In partic- Developing countries that increased their integra- ular, each participating developing country will need tion into the world economy over the past two a thorough understanding of how trade liberaliza- decades achieved higher growth in incomes, longer tion can contribute to its national objectives of eco- life expectancy, and better schooling. These coun- nomic growth and poverty reduction. Such strategic tries, home to some 3 billion people, enjoyed an understanding will have to be supported by both the average 5 percent growth rate in income per capita trade negotiators and by civil society; at times, the in the 1990s compared to 2 percent in rich countries. medium-term goal of poverty reduction requires A common thread exists among these developing governments to challenge the interests of some par- countries that have been successful at generating ticular industries for short-term protection. greater growth and at lifting people out of poverty. In addition, many countries will have to break new They opened up their economies as part of a broad- ground. Today's trade issues go beyond the tradi- er development strategy that builds on two pillars: tional mechanisms of tariffs and quotas and include improving the investment climate for the private "behind-the-border" issues, such as the role of infra- sector to generate jobs and empowering poor peo- structure and governance in supporting a well-func- ple, so they can participate in growth. tioning trading economy. Many poor countries have This approach to development, with trade liberal- yet to create intellectual property regimes that make ization as one mechanism of improving the invest- traditional knowledge or cultural products into ment climate for private entrepreneurs, has gained negotiable and defensible assets; to identify options wide support among developing and industrial to upgrade and enforce national product, health, countries. All WTO member countries, including and safety standards; or to strengthen institutions those in the developing world, have reduced their for prudential and pro-competitive regulation of trade tariffs since the Uruguay Round. In its Every- services. Developing countries will have to acquire thing but Arms agreement, the European Union has quickly the needed expertise on these complex unilaterally lowered its trade barriers to the least issues, so they can negotiate more effectively and developed countries. The United States adopted the ensure that agreements serve their objective of African Growth and Opportunities Act. And in poverty reduction. November 2001, the members of the World Trade This Handbook is part of our efforts to prepare Organization launched a "Development Agenda" in developing countries to negotiate trade agreements. Doha. In doing so, they acknowledged that to make It is the product of a joint capacity-building effort progress in the fight against poverty, rich country involving a number of research institutes around xi D E V E L O P M E N T, T R A D E , A N D T H E W T O the globe. It aims to provide a summary of the eco- strong interest in, real-world trade policy making, nomics of sound trade policy and to be a guide to rather than the theory of international trade. Such many of the behind-the-border regulatory issues practitioners will be in ministries of trade, industry, that confront countries in the contexts of both and finance; parliaments; private sector associations domestic reform and international negotiations. such as chambers of commerce; consumer organi- Views and approaches to many of the issues that are zations, and policy institutes. dealt with in this volume differ substantially, and The diversity and pragmatism of the views repre- these differences are reflected here. All are motivat- sented contribute to the richness of this Handbook ed by the question of how the global trade architec- and make it a very worthwhile resource for all trade ture might be made more supportive of practitioners. It will help us "seize the moment" and development, and the question of how developing fulfill Doha's promise to focus on the need for trade countries can use international negotiations and to bring about greater growth and poverty reduction. cooperation as an instrument to further domestic reform and access to export markets. NICHOLAS STERN The Handbook is intended to be a source of SENIOR VICE PRESIDENT AND information and guidance for all practitioners, CHIEF ECONOMIST defined as those with either a responsibility for, or a THE WORLD BANK xii A Tribute to J. Michael Finger F ifteen years after the publication friends. This Handbook is dedicated to him both as of the World Bank's first handbook a practical tribute to his work and influence and in on trade policy and multilateral negotiations (Fin- the belief that his clear-sighted approach to trade ger and Olechowski 1987), the development dimen- policy will motivate researchers, analysts, and com- sions of trade policy and trade negotiations often mentators who have never had the opportunity to seem to be neglected. This is especially the case as meet him. regards trade agreements, where negotiations are Finger has noted that "trade theory is about iden- frequently driven by interest groups in high-income tifying whose hand is in whose pocket. Trade policy countries and where outcomes can have significant is about who should take it out" (Finger 1981). Both costs for developing countries, both in a monetary are important. Good policymaking requires a solid or resource-use sense and--since these obligations grounding in fact and analysis--an understanding may deflect attention and resources away from of the processes that are taking place--and a frank other, more important, tasks--in opportunity costs. recognition that, at least for international trade, The focus of much of the advice and assistance that there will be winners and losers from virtually any is offered to developing country policymakers cen- policy decision. Trade policy advice needs to pro- ters on enhancing their understanding of the rules vide this grounding, but it must also understand of the international trading game, as opposed to and internalize how the potential conflicts between determining what type of trade policy makes the winners and losers are played out in actual decision- most sense from a development perspective. The making institutions. Merely wringing one's hands latter is crucial, as only on that basis is a national and bemoaning the fact that policy advice is ignored "bottom-up" approach to the design of multilateral is not satisfactory; one needs to see why and ask rules possible. As noted by J. Michael Finger (1991a, how institutions can be designed to produce better 1991c), when it comes to the relationship between policy outcomes. the multilateral trading system and development, Five components of good trade policymaking can there is a widespread tendency to "think about be distinguished: economic analysis, information GATT only in the GATT way." Finger made this and data, political economy, operationalization of observation before the creation of the WTO and, policy advice, and a contestable market for policy characteristically, well before it became convention- research. Each is discussed below. al wisdom in the development community. Subse- quent experience has reinforced his insight.1 Economic Analysis This Handbook continues a series that Finger launched in 1987. Mike Finger, who retired from the Economic analysis is perhaps the most obvious and World Bank in 2001, has been a source of inspira- most easily provided input into the policy cocktail. tion, a guide, and a mentor to several generations of Many academic economists are active in this area, trade policy analysts. Many of the contributors to and theory is cheap. Getting concepts clear and this volume have been inspired by his writings, and showing how one thing implies another, and under a significant number have also been colleagues and what conditions, are necessary first steps toward any xiii D E V E L O P M E N T, T R A D E , A N D T H E W T O reliable policymaking. Finger's early work on the lizing support for disciplining the use of NTBs in usefulness of the concept of effective protection is the Uruguay Round. an example (Finger 1969). Such analysis might be Finger also produced the most useful summaries purely positive--describing relationships between of tariff concessions in the Uruguay Round (Finger, economic phenomena--or it might have normative Ingco, and Reincke 1996). They were useful partly dimensions in which conclusions about welfare are because he chose to measure concessions in a more inferred. Either way, it needs to be clear and logical- informative way than did national statisticians and ly watertight. the WTO. He generated the statistic, much cited in But although logical precision may be necessary, the mid-1990s, that trade reforms under IMF and it is far from sufficient for defining good policy World Bank programs liberalized more developing analysis. The other essentials--which are, regret- country trade than had the Uruguay Round. And tably, much scarcer--are that the analysis deal with recently he showed how developing countries a real problem that someone is interested in and received fewer Uruguay Round concessions (in a that it be couched in operational terms.2 Policy mercantilist sense) than they gave (Finger and research must ultimately refer to phenomena that Schuknecht 2001; Finger, Reincke, and Castro forth- can be observed--especially in terms of defining the coming). In each case, the secret was to base detailed circumstances in which the research can be applied work on the data on a well-specified question posed and the levels at which to set policy instruments-- in terms of clear operational concepts. This is, in and it must refer to policy levers that could, in prin- fact, not "mere" description but a sophisticated ciple, be manipulated. Unfortunately, all this makes marriage of theory and data that relies on simple such analysis less attractive to the editors of aca- statistics rather than high-technology econometric demic journals. Finger's work, and his publishing techniques.4 record, suggest, however, that policy analysis can be A further important dimension of good policy every bit as exciting intellectually as "pure" theory. research is to devise ways of helping policy analysts Some examples are early pathbreaking work assess- throughout the world to repeat and extend state of ing the relative impact of transport costs and tariffs the art analysis. Many theoretical analyses are easily (Finger and Yeats 1976) and assessments of the replicated, but once one begins to work with data, effects of the offshore assembly provisions in U.S. there can be high costs to replication even if there customs law (Finger 1976b). are no unforeseen difficulties to overcome. The first requirement for facilitating replication is to make data available. Great strides have been taken recent- Information and Data ly in this direction, but there are still many hurdles Great analysis will occasionally throw up universal to be overcome. For example, it is unfortunate that truths that do not require detailed empirical analy- members of the WTO continue to resist making the sis in order to be applied. More commonly, howev- Uruguay Round bound tariff rates available in a er, policy analysis is empty without a supporting convenient form for analysis.5 information base. The most obvious need in inter- The second component also focuses on conven- national trade is for data on international trade ience. Where complex operations or calculations are flows and policies, but information on the laws and required, it is essential to make tools and routines institutions that govern behavior is also essential. available to external researchers and so make it "Merely" collecting and presenting information can cheap and easy for them to carry out their own be spectacularly illuminating. For example, in the analysis. Although we tend to equate data with mid-1980s Finger inspired and managed the World numbers, it is equally important for policy analysts Bank's empirical work on nontariff barriers to have accurate representations of institutional and (NTBs); Nogués, Olechowski, and Winters (1986) is legal processes. Some of these may be summarized an example.3 At that time theory about the effects of numerically, as in, for example, Finger's compilation NTBs was not lacking, but there was almost no con- of tariff concessions granted and received in the cept of how pervasive these barriers were. Present- Uruguay Round (Finger, Ingco, and Reincke 1996). ing the data (and being clear about the adequacy or In other cases it is a matter of observing processes inadequacy of the measurements) took the debate and finding ways of distilling their essence into a much further and had a significant effect in mobi- few simple statements. The pioneering work by Fin- xiv A Tribute to J. Michael Finger ger, Hall, and Nelson (1982) on administrative pro- into an unbalanced game with a strongly biased tection (on which more below) is based on a careful outcome. The classic analysis by Finger, Hall, and institutional study of U.S. antidumping, counter- Nelson of antidumping actions makes clear how the vailing, and escape clause protection. technical process favors business interests and dis- Today, as we grapple with the subtleties and com- enfranchises consumers and users--in fact, the plexities of trade and development policies and the authors argue that this is precisely the political role of the WTO in supporting development, work function of the process. Only by covertly loading on information and data is prominent. For exam- the scales can a low-level (ostensibly nonpolitical) ple, many of the chapters on services in this Hand- process hope to resolve distributional struggles; if book deal with how to measure the progress of the loading is obvious, the matter tends to become liberalization in these areas and examine how gov- political and more difficult. Moreover, once estab- ernments have applied the agreements domestically. lished, such unbalanced processes tend to perpetu- ate themselves because the interest groups whose interests are served (including the people who man- Recognition of Political Economy age the processes) ensure that this happens. Finger, If entrepreneurs and workers respond to price sig- who has contributed massively and seminally to our nals and incentives in determining their behavior, it understanding of antidumping and safeguards over is hardly surprising that they respond to other eco- two decades, discusses the subject in Chapter 22 of nomic incentives, such as the opportunity to create this volume. economic rents by intervening in markets. Nor is it A second area in which political economy has surprising that politicians and bureaucrats also been and continues to be crucial is in the analysis of respond to incentives of various kinds. Political the GATT and the WTO. Finger (1974, 1976a) economy may be broadly thought of as exploring demonstrated the centrality of reciprocity to GATT the role of political activities and forces in shaping tariff negotiations: despite the requirement that all economic behavior. Policy analysis must take it into tariff cuts be extended to all GATT partners, nego- account not only in predicting the outcomes of par- tiators chose commodities in such a way that much ticular policy changes but also, and more particular- of the benefit of a cut went to the country request- ly, in designing institutions and policy regimes. ing it (the principal supplier), and nearly all of the Political economy and trade policy are closely benefit went to countries taking active part in the linked because every trade intervention potentially negotiations. In passing, Finger also showed that the creates conflict between winners and losers--pro- developing countries which made significant con- ducers and consumers, agriculture and industry, cessions in the multilateral negotiations received far skilled and unskilled labor. Because all political sys- more concessions on their exports than did passive tems find it difficult to deal with redistribution observers. explicitly, trade policy presents both problems and Finger (1979) was being relatively unfashionable opportunities. Supporting a trade liberalization in observing that the critical political balance of the while providing explicit redistribution to prevent GATT was internal to countries, as exporters seek- anyone from losing is a big political challenge, ing market access abroad pressed import-compet- whereas using trade intervention as a form of covert ing sectors to concede market access at home. When redistribution that is buried beneath a pile of arcane U.S. legislation provided an alternative route for technical detail often looks like an attractive way of exporters to open foreign markets, in the form of protecting or rewarding specific interests. Section 301, which authorized unilateral trade sanc- Opacity can make trade policy very attractive. tions against trade-restricting partners, the whole Finger (1981) and Finger, Hall, and Nelson (1982) dynamic of U.S. support for the multilateral system showed that the mechanisms involved can be changed (Finger 1991b). Finger was prominent extremely subtle. Making protection subject to legal among those who had argued that for developing requirements and establishing expert investigative countries the principal requirement was not to authorities to establish the "facts" of the case all open up others' markets, but to open up their own serve to remove the process from public scrutiny. By (Finger and Kreinin 1976) and that in this respect making procedures complex and expensive, one can the GATT approach (and GATT-think) based solely turn an apparently objective pseudo-legal process on reciprocity was not very useful. Developing xv D E V E L O P M E N T, T R A D E , A N D T H E W T O countries must address their own needs directly via forces and shining a light on them (Jagdish Bhag- their own policy and should not view the multilat- wati's "Dracula principle"); redressing the balance eral system as providing a shortcut to good trade of forces in trade debates to promote consumer policy or good trade outcomes; it just does not do interests; and making transparent the winners and that (Finger and Winters 1998). losers from any action (or inaction); see, for exam- These political-economy problems figure promi- ple, Finger (1982, 1986). It also illustrates the dan- nently in this Handbook, for it is only by recogniz- gers of complexity, suggesting a second aspect of ing their force that they can be overcome. Thus, for operationalization: the use of rules of thumb in pol- example, Chapter 52, by Tarr, stresses the political- icymaking. Among the rules of thumb advocated in economy advantages of uniform tariffs, which are this Handbook by some authors are the use of uni- much more robust to lobbying than tailor-made form tariffs as a robust antidote to sectoral special tariffs. Rodrik, in Chapter 1, emphasizes the impor- pleading and rent-seeking, and promotion of effec- tance of developing local institutions as local solu- tive competition as the single most important tions to local problems rather than adopting objective in services markets. uniform institutions imposed by the international community, and this analysis is consistent with Fin- A Contestable Market for Policy Research ger's point about the costs of certain Uruguay Round institutions (Finger and Schuler 2000). In An important dimension of such beneficial compe- Chapter 22, Finger explores the political economy tition relates to policy analysis. The social function of safeguard provisions, and in Chapter 7 Finger of such research is to improve policy outcomes by and Winters explore what reciprocity means in the basing them on the best possible understanding of current broad agenda of the WTO. (It has no "exter- the effects of policy. De facto, its political function is nal" definition; it is whatever deal the parties are to smooth the path of decisionmaking by ensuring willing to agree on.) that relatively minor issues do not destroy social consensus and impose huge costs in the form of strife. This second function is not unimportant (as Operationalization of Policy Advice Rodrik notes in Chapter 1), but it is often at vari- Policy analysis is ultimately sterile if it does not ance with the first. The tension between the two change behavior. How to present and package the roles of policy research is felt most immediately in findings of analysis in ways that both strike chords official policy research centers. If analysts there stick with decisionmakers and are (relatively) easy to to the objective side of their brief, they are ignored, apply is critical. Again, Mike Finger leads the way. abused for being irrelevant or obstructionist, and The clarity and directness of his writing is a model often, as happened to Finger's unit in the U.S. Trea- for all researchers. And it is substantially achievable sury, closed down. If they stress the political aspects, by them too, for while it certainly requires talent, it they discredit themselves and, ultimately, their insti- mostly relies on thinking hard--and with brutal tutions as purveyors of information; indeed, they objectivity--and on working hard (spending time may even discredit analysis itself. And by giving a refining one's prose). Finger also has a talent for the politically convenient compromise a gloss of spuri- memorable phrase or metaphor: "Antidumping is ous intellectual respectability, analysts can sow the ordinary protection with a great public relations seeds of further problems by establishing the wrong program" (Finger 1993); "Where the WTO got it basis for thinking about future decisions. The falla- wrong, it was perhaps because the World Bank did cy that trade liberalization creates jobs (perpetrated, not get it at all" (Finger and Nogués 2002, on the for example, in the debate on the North American inappropriateness of certain Uruguay Round out- Free Trade Agreement), and its refutation by experi- comes for development);"Half of domestic interests ence, have made rational trade policy more difficult have no chance to score" (on antidumping, in to achieve. The fallacy that reductions of tariffs on a Chapter 22 below, with a picture of a soccer field of developing country's exports are more important which only one end has a goal). than reductions of tariffs on its imports has led to The discussion of political economy in the pre- the waste of huge resources on instruments such as ceding section covered some aspects of operational- trade preferences, the Generalized System of Prefer- izing trade policy advice: recognizing reactionary ences (GSP), and the New International Economic xvi A Tribute to J. Michael Finger Order and, ultimately, to the false notion that the Notes GATT/WTO process and good trade policy are 1 Similar arguments are made in Finger and Kreinin (1976) and coterminous (see Finger 1975, 2001; Finger and Finger (1982). Kreinin 1976). 2 This is not to decry basic economic science but merely to place How can this tension be resolved? In his valedic- it outside the box of policy research. tory speech to the U.S. Treasury (Finger 1981), Mike Finger observed that "political responsibility is the 3 The actual collection was mostly (and continues to be) done ultimate intellectual vasectomy." What is the by the UNCTAD, but the presentation and use of the data for policy analysis was pursued more vigorously by the World answer? It is to ensure that the market for policy Bank. research is open and contestable. Governments and international organizations require research arms, 4 Another example was Finger and DeRosa (1980), which showed with the simplest of tools that the IMF's Commodity but it is vital that others, outside government, are Compensatory Fund might not have the desired effect of stabi- also able to participate fully in this market. Govern- lizing developing countries. ments thus have to make data and information eas- 5 A CD-ROM that replicates the country schedules is available, ily available publicly, accept criticism, and be but it is not an electronic file of data. prepared to justify distributive judgments and deci- sions. Finger has shown by example what type of analy- sis is necessary for better policy choices and out- comes. We hope that this Handbook, and the kind of collaborative, research capacity­building effort on which it draws, will help stimulate others to emulate the "Finger approach" to policy research and analysis. BERNARD HOEKMAN L. ALAN WINTERS xvii blank Acknowledgments T he chapters included in this Hand- (UNCTAD); Diana Tussie, LATN; Ademola Oyejide, book are a product of a collaborative University of Ibadan and AERC; Rajesh Chadha, research and capacity-building project that involved National Council of Applied Economic Research scholars in developing countries, international (NCAER), New Delhi; Hana' Al-Sagban, ERF; Thier- experts, and World Bank trade economists and that ry Noyelle, United Nations Department of Economic was designed to assess the costs and benefits of fur- and Social Affairs (UNDESA); Dominique Njinkeu, ther multilateral rule-making and liberalization. In AERC; Mari Pangestu, Center for Strategic and addition to researchers based in national research International Studies, Jakarta; Alan Winters, Sussex organizations, the project draws on the work of a University; and Stephen Yeo, Centre for Economic number of research networks, including the Latin Policy Research (CEPR), London. We are grateful as American Trade Network (LATN); the Economic well to Constantijn Claessens, Antonio Estache, Research Forum for the Arab Countries, Iran, and Carsten Fink, Faezeh Foroutan, John Hegarty, Turkey (ERF); the African Economic Research Con- Charles Kenny, Dorsati Madani, Amrita Narlikar, sortium (AERC); the Coordinated African Program Christina Neagu, Claudia Orozco, Randeep Rathin- of Assistance on Services (CAPAS); and the Trade dran, and Gomi Senadhira, who contributed part or Policy Forum of the Pacific Economic Cooperation all of the material for boxes on specific topics. Council (PECC). The views expressed are entirely Many of the papers that were prepared under the those of the contributors, and do not necessarily auspices of the research capacity­building project reflect the views of the World Bank Group, the insti- of which this Handbook is a part have been pub- tutions that the authors are affiliated with, or the lished in journals and conference volumes. Papers countries they represent. written by members of the project and dealing The project was supported by a grant from the with many of the subjects covered in this Hand- U.K. Department for International Development book can be found in the April 2000 issue of The (DFID) and by the World Bank Institute, the World World Economy; the May 2001 issue of The Review Bank Research Support Budget, the government of of International Economics; Olarreaga and Rocha the Netherlands, and the Société Générale de Sur- (2000); Hoekman and Martin (2001); Stern veillance (SGS), Switzerland. We are particularly (2001); Martin and Pangestu (forthcoming); grateful to Susan Prowse and Charlotte Seymour- Maskus and Wilson (2001); and Mattoo and Stern Smith for their encouragement and guidance. (forthcoming). The editors are indebted to Maria Kasilag, Rebecca All of the members of the World Bank's trade Martin, Ana Rivas, Rob Simms, and Lili Tabada for team have been active contributors to the project, in excellent administrative support throughout the particular Will Martin, who has played a key role in project. We thank the following people for their managing the project and provided leadership in active engagement and participation in the delivery working with partners in Asia. Carsten Fink, and of various parts of the project: Kym Anderson, Ade- Keith Maskus have been unstinting in their willing- laide University; Bijit Bora, at the time with the Unit- ness to contribute and comment on materials. Fran- ed Nations Conference on Trade and Development cis Ng played an essential role in providing data to xix D E V E L O P M E N T, T R A D E , A N D T H E W T O participants. The Trade and Production Database that helped improve the final product. We thank CD-ROM included with the Handbook is the result those who acted as readers and discussants of chap- of the painstaking work of Marcelo Olarreaga and ters and the handbook as a whole, including Claude Alessandro Nicita. Miroslava Zervoudakis and Barfield, Paul Collier, Richard Eglin, Alan Gelb, Joe Faezeh Foroutan were key contributors to the pro- Francois, Jacob Kol, Patrick Messerlin, Douglas Nel- ject's Website, , which son, David Palmeter, Garry Pursell, Jayanta Roy, has become the major dissemination vehicle for the Alan Winters, Luc de Wulf, Jamel Zarrouk, and output generated by team members. Ambassador B. K. Zutshi. The chapters included in the Handbook benefited Finally, we acknowledge the considerable time from comments and feedback obtained from par- and effort that has been devoted to the preparation ticipants in workshops, conferences, and seminars, of the second CD-ROM included with this Hand- who are too numerous to be mentioned by name book, "Applied Trade Policy for Developing Coun- here. We are very grateful to the people and institu- tries: Outline, Content, and Readings for a Short tions that hosted and helped arrange these meet- Course." Our thanks go to the principal authors, ings, in particular Richard Eglin, Sam Laird, and Jaime de Melo, Centre d'Etudes et de Recherches sur Peter Tulloch of the WTO secretariat, who organ- le Développement International (CERDI), and ized a major conference in September 1999 and a Marc Bacchetta, WTO, and to their organizations, as day-long review seminar of the first draft of the well as to Olivier Jammes (CERDI) for CD-ROM Handbook in May 2001. Members of the WTO del- development. egations in Geneva provided invaluable suggestions xx Contributors Rudolf Adlung World Trade Organization Kym Anderson Adelaide University, Australia, and CEPR Marc Bacchetta World Trade Organization Bijit Bora World Trade Organization Antonia Carzeniga World Trade Organization Rupa Chanda Indian Institute of Management Philippe Chauvet World Trade Organization Rafael Cornejo Inter-American Development Bank Valentina Delich FLACSO/Argentina and LATN Jaime de Melo University of Geneva, CERDI, and CEPR Luc De Wulf Independent Consultant Liam Ebrill International Monetary Fund Philip English World Bank Simon J. Evenett World Trade Institute, Berne, and CEPR J. Michael Finger American Enterprise Institute, Washington,D.C. Carsten Fink World Bank Joseph F. Francois Erasmus University, Rotterdam, and CEPR Carlo Gamberale World Trade Organization Luis Jorge Garay S. Inter-American Development Bank Reint Gropp International Monetary Fund James Hodge University of Cape Town Bernard Hoekman World Bank and CEPR Peter Holmes University of Sussex, U.K. Gary N. Horlick O'Melveny & Myers, Washington, D.C. Robert E. Hudec Tufts University, Boston. Stefano Inama UNCTAD Veena Jha UNCTAD Hanaa Kheir-El-Din Cairo University Masamichi Kono Financial Services Authority (Japan) Sam Laird UNCTAD and University of Nottingham, U.K. Miguel F. Lengyel FLACSO/Argentina and LATN David F. Luke Organization for African Unity Catherine L. Mann Institute for International Economics, Washington, D.C. Will Martin World Bank Keith E. Maskus University of Colorado Aaditya Mattoo World Bank Andreas Maurer World Trade Organization xxi D E V E L O P M E N T, T R A D E , A N D T H E W T O Constantine Michalopoulos Independent consultant Francis Ng World Bank Alessandro Nicita University of Geneva Marcelo Olarreaga World Bank and CEPR T. Ademola Oyejide University of Ibadan, AERC, and the Development Policy Center, Ibadan, Nigeria Arvind Panagariya University of Maryland Mari Pangestu Centre for Strategic and International Studies, Jakarta Frank J. Penna The Policy Sciences Center, Inc., New Haven, Conn. Francisco Javier Prieto Organization of American States Garry Pursell Independent consultant Vinod Rege International trade consultant Dani Rodrik Harvard University Kamal Saggi Southern Methodist University, Dallas, Tex. Pierre Sauvé OECD Maurice Schiff World Bank Philip Schuler University of Maryland Howard J. Shatz Harvard University Eleanor Shea O'Melveny & Myers, Washington, D.C. Beata K. Smarzynska World Bank and CEPR Brian Rankin Staples Trade Facilitation Services, Ottawa Sherry M. Stephenson Organization of American States Robert M. Stern University of Michigan Janet Stotsky International Monetary Fund Arvind Subramanian International Monetary Fund David G. Tarr World Bank Simon Tay Member of Parliament, Singapore Diana Tussie FLACSO/Argentina and LATN Lee Tuthill World Trade Organization Coenraad J. Visser University of South Africa Jayashree Watal World Trade Organization John S. Wilson World Bank L. Alan Winters University of Sussex, U.K. and CEPR xxii Abbreviations ACIS Advance Cargo Information System ACP African, Caribbean, and Pacific (Cotonou Convention, formerly Lomé) ACWL Advisory Centre on WTO Law AD antidumping AERC African Economic Research Consortium AGOA African Growth and Opportunity Act (U.S.) AMS aggregate measure of support APEC Asia-Pacific Economic Cooperation ASEAN Association of Southeast Asian Nations ASYCUDA Automated System for Customs Data and Management (UNCTAD) ATC Agreement on Textiles and Clothing (WTO) ATPA Andean Trade Preferences Act BDV Brussels Definition of Value BIT Bilateral Investment Treaty BTN Brussels Tariff Nomenclature CAP Common Agricultural Policy (EU) CBD Convention on Biological Diversity CCC Customs Cooperation Council (now the WCO) CCCN Customs Cooperation Council Nomenclature CEFACT Center for Facilitation of Procedures and Practices for Administration, Commerce, and Transportation (UN) CGE computable general equilibrium (model) c.i.f. cost, insurance, and freight CRM customs reform and modernization CRTA Committee on Regional Trade Agreements (WTO) CTE Committee on Trade and Environment (WTO) CTH change in tariff heading CVD countervailing duty DSB Dispute Settlement Body (WTO) DSP dispute settlement procedures (WTO) DSU Dispute Settlement Understanding (WTO) EBA Everything but Arms (EU initiative for LDCs) EC European Community EDI electronic data interchange EDIFACT Electronic Data Interchange for Administration, Commerce, and Transport (UN) EEC European Economic Community EFTA European Free Trade Association xxiii D E V E L O P M E N T, T R A D E , A N D T H E W T O EPZ export-processing zone ERP effective rate of protection EU European Union FAO Food and Agriculture Organization of the United Nations FDI foreign direct investment f.o.b. free on board FSC foreign sales corporation FTA free trade area FTAA Free Trade Area of the Americas GATS General Agreement on Trade in Services (WTO) GATT General Agreement on Tariffs and Trade (WTO) GDP gross domestic product GMO genetically modified organism GNP gross national product GPA Agreement on Government Procurement (WTO) GSP Generalized System of Preferences GTAP Global Trade Analysis Project HCC Heads of Customs Conference (NAFTA) HS Harmonized Commodity Description and Coding System ICC International Chamber of Commerce ICTSD International Centre for Trade and Sustainable Development IDB Integrated Data Base (WTO) IECC International Express Carriers Conference IF Integrated Framework for Technical-Related Assistance, Including Human and Institutional Capacity Building to Support Least-Developed Countries in Their Trade and Trade-Related Activities IFIA International Federation of Inspection Agencies ILO International Labour Office IMF International Monetary Fund IPRs intellectual property rights ISIC International Standard Industrial Classification ISO International Organization for Standardization ITC International Trade Centre (UNCTAD and WTO) ITC International Trade Commission (U.S.) ITCB International Textiles and Clothing Bureau ITO International Trade Organization LATN Latin American Trade Network LDC least-developed country (UN classification) MAI Multilateral Agreement on Investment MEA Multilateral Environmental Agreement MENA Middle East and North Africa MERCOSUR Common Market of the South MFA Multifibre Arrangement MFN most-favored-nation MRA mutual recognition agreement MTA multilateral trade agreement MTN multilateral trade negotiation NAFTA North American Free Trade Agreement NATO North Atlantic Treaty Organization NGO nongovernmental organization NRP nominal rate of protection xxiv Abbreviations NTB nontariff barrier NTM nontariff measure OAU Organization of African Unity OECD Organisation for Economic Co-operation and Development OMA orderly marketing arrangement PPM production and processing method PSI preshipment inspection PTA preferential trading agreements Quad Canada, European Union, Japan, and the United States QR quantitative restriction R&D research and development RCA revealed comparative advantage RIA regional integration agreement ROO rules of origin SADC Southern African Development Community SCM subsidies and countervailing measures S&D special and differential (treatment) SDR Special Drawing Right SGS Société Générale de Surveillance SITC Standard International Trade Classification SPS sanitary and phytosanitary STE state trading enterprise TABD Transatlantic Business Dialogue TBT technical barriers to trade TMB Textiles Monitoring Body (WTO) TPO trade promotion organization TPRB Trade Policies Review Body (WTO) TPRM Trade Policies Review Mechanism (WTO) TRIM trade-related investment measure TRIPS Trade-Related Aspects of Intellectual Property Rights (WTO agreement) TRQ tariff rate quota UNCITRAL United Nations Committee on International Trade Law UNCTAD United Nations Conference on Trade and Development UNDP United Nations Development Programme USTR U.S. Trade Representative VER voluntary export restraint WCO World Customs Organization WHO World Health Organization WIPO World Intellectual Property Organization WITS World Integrated Trade Solution (World Bank) WTO World Trade Organization xxv blank Introduction M any countries have been less than institutions, trade policies, narrowly defined, are successful in integrating into the still important in today's international economic world economy and benefiting from trade reform landscape. Barriers to exports of some products in programs. The reasons are multifaceted and com- which developing countries have a comparative prise a mix of domestic and international factors. advantage remain high--tariffs on some agricultur- Barriers to trade and investment remain high in al products are over 100 percent. Agricultural subsi- many nations, with policy regimes implying signifi- dies in OECD countries exceeded US$300 billion in cant anti-export bias. Numerous countries have 2000, contributing to global price instability and been affected by civil strife and war. And in spite of impeding the ability of developing countries to the trade preferences granted by member countries compete on export markets. of the OECD, industrial country tariff structures are Trade between developing countries began to still characterized by escalating tariffs, with high grow rapidly in the 1990s, increasing the signifi- tariff peaks for agricultural products and for labor- cance of their own trade barriers for export interests intensive products such as clothing. in these countries. Antidumping actions are no There is general agreement that many complementa- longer limited to OECD economies but have come ry policies and institutions are needed to support trade to be used intensively by a number of developing policy reforms in order to create an enabling environ- countries. Barriers to trade in services are many ment for supply-side responses that generate employ- times those that apply to trade in merchandise, ment and economic growth. As Dani Rodrik argues in especially where movement of the service provider Chapter 1 of this Handbook,if trade policy reform is to is necessary. In many cases these barriers and detri- be successful, it must be embedded in and supported mental policies can be removed only through inter- by an effective institutional setting,and it must be com- national negotiations. plemented by other reforms. A large and complex International trade agreements, in particular the "behind-the-border" agenda has to be addressed if WTO agreement, have become the focal point for trade reform is to have its intended effect. Much many discussions on trade and investment policy. depends on complementary policies that define the As a result, policymakers and citizens of developing business environment--on policies regarding invest- countries are confronted with demands that a num- ment in human capital (education),infrastructure,and ber of trade policy­related issues be addressed in the quality of public and private sector governance. the context of multilateral or regional negotiations. The Handbook focuses on a number of the elements of This offers opportunities to pursue what are regard- that agenda, as well as on more"traditional"trade poli- ed as desirable domestic reforms, but it also poses cy issues such as the design of the tariff regime. risks associated with agreements or rules that may not be supportive of development prospects. The traditional mechanism driving trade agree- Trends in the Multilateral Trading System ments has been the reciprocal exchange of commit- Although the challenges confronting developing ments to reduce trade barriers. This mechanism countries primarily concern domestic policies and results in greater welfare improvements than can be xxvii D E V E L O P M E N T, T R A D E , A N D T H E W T O obtained through unilateral reform, as it generates WTO auspices to address developing country con- liberalization both at home and abroad and makes cerns. Similar calls were put forward in the prepara- politically feasible domestic trade reforms that oth- tions for the 2001 ministerial meeting in Doha. erwise might be blocked by powerful vested inter- The Doha Development Agenda that emerged ests. International cooperation can also be a useful from the meeting clearly reflects the increased device for pursuing domestic reforms that are indi- prominence of development concerns in WTO rectly linked to trade. As tariff barriers have fallen deliberations--in turn, the result of increased par- and quantitative restrictions have disappeared, the ticipation by developing countries in the trading focus of trade agreements has increasingly shifted system. All that was done, however, was to define an toward regulatory regimes that can have an impact agenda. Achievement of a prodevelopment outcome on trade and investment. remains a major challenge. Resistance to liberaliza- Multilateral negotiations on nonborder policies, tion of "hard-core" sectors such as agriculture and administrative procedures, and domestic legal textiles that are of key interest to developing coun- regimes have proved much more complex than talks tries is very strong; conversely, many low-income on traditional market access. Because it is more dif- countries are unwilling to extend the reach of the ficult to trade"concessions,"the focus tends to be on WTO to cover new issues. Implicitly, if not explicit- the identification of specific rules that should be ly, much of the discussion and debate at Doha con- adopted. Given the disparities in economic power cerned defining the limits of the WTO. Developing and resources among countries, the outcome often countries played a central role in this debate, with reflects the status quo in high-income countries. many resisting the further expansion of the WTO These may be fully consistent with the development into the territory of domestic regulation. priorities of low-income countries, but there is no The Doha Ministerial Declaration launches nego- presumption that this will be the case. tiations on market access for manufactures, dispute Developing country misgivings regarding the settlement, WTO rules, disciplines on regional inte- rule-making dimensions of the WTO became gration, environment, and intellectual property increasingly prominent in the 1990s. These con- rights (geographical indications). These talks will cerns centered on the costs required for implement- complement ongoing negotiations on agriculture ing some WTO agreements, the lack of adequate and services, as mandated by the Uruguay Round financial assistance, and the failure of high-income agreements. Negotiations are to be concluded by countries to grant "special and differential" treat- 2005. At the next WTO ministerial meeting, in 2003, ment to developing countries. (Most of the provi- negotiations will be launched on four "Singapore sions in the WTO agreements calling for such issues"--competition, investment, trade facilita- treatment are "best endeavor" commitments that tion, and transparency in government procure- are not binding on high-income countries.) A more ment--if agreement on modalities can be obtained fundamental concern was that the rules of the game by explicit consensus at that time. were not always compatible with national efforts to Whether the end result will be prodevelopment reduce poverty and increase economic growth. will depend to an important degree on the extent to For the rules to make sense for all members, which developing and industrial country trade bar- stakeholders in developing countries must partici- riers are lowered, and on the rules that emerge. A pate in the domestic policy formation process, be key determinant of the outcome of the negotiations able to inform national representatives of their will be effective and proactive developing country views, and hold their representatives accountable participation. This, in turn, requires a good under- for outcomes. If WTO agreements were unambigu- standing of where national interests lie and a good ously seen by constituencies in developing countries understanding of the substantive issues, not just by as being conducive to (or consistent with) the government officials but also by the private sector attainment of development objectives, these agree- and civil society. There is clearly a need to strength- ments could play a much more beneficial and effec- en capacity to undertake analysis and to identify tive role. In the run-up to the 1999 WTO ministerial national reform priorities, market access con- meeting in Seattle, a number of prominent straints, and the potential merits and implications observers and policymakers called for the launch of of multilateral disciplines. This Handbook is a "Development Round" of negotiations under intended as a contribution to that effort--as a use- xxviii Introduction ful resource for analysts and stakeholders engaged views do not necessarily reflect those of the institu- in the design of trade-related policies. tions with which they are affiliated. Not everyone will necessarily agree with all the policy recommendations made by the authors. After Objectives of the Handbook all, as we noted above, on a number of issues there is A major challenge confronting developing countries no "one size fits all" answer, and this is especially is to use international negotiations and cooperation true of regulatory policies. What matters most is to as instruments for improving their terms of trade ask the right questions and to determine the status and their access to export markets and as mecha- quo in a given area. It is important to obtain as nisms for adopting and implementing domestic pol- much information as possible regarding alternative icy reforms that will raise living standards and policy options, to understand what type of analysis reduce poverty. The design of trade policy reform is is needed to provide policy guidance, and to have a a complex matter that extends far beyond tariffs and good understanding of the prevailing multilateral quotas applied at the border. It must be comple- rules of the game. mented by policies designed to ensure that enter- Although much of what is contained in the Hand- prises can compete on world markets. There is no book is motivated by the fact that the issues are on "one size fits all" package of policy reform, and no the agenda of international negotiations, the magic bullet. Approaches will and must differ across emphasis of many contributors is on economic and countries, reflecting different circumstances, endow- development dimensions. The institutions and poli- ments, legal systems, and cultures. cies that are important for development and eco- One goal of this Handbook is to provide informa- nomic growth extend far beyond the subject areas tion on the implications of--and options offered that the WTO deals with or can deal with. Although by--international trade agreements, especially the the WTO can be useful in helping countries address WTO, for developing countries that seek to use specific bottlenecks and constraints that impede trade as a vehicle for development. Contributors trade, most of the trade policy agenda is domestic. It were asked to write relatively short chapters on a is therefore vital that policymakers and civil society variety of trade policy­related topics that are have a good understanding of what their national important from a development perspective and that priorities are and what makes for good policy, are subject to or affected by multilateral rules, or informed by the experiences of other countries, in may become so. The chapters assess the economics order to determine what types of multilateral coop- of the issues, survey what cross-country experience eration can help countries benefit from trade inte- suggests are good practices, and consider the pros gration. and cons of the possibilities for using international Relatively little emphasis is given in the Hand- cooperation as an instrument for improving both book to an enumeration of WTO disciplines. There domestic policy and access to export markets. are many readily available resources that can pro- Although there is an emphasis on the WTO, many vide the interested reader with such information, of the issues addressed also arise in the context of starting with the WTO Website, . regional integration agreements. The CD-ROM provided with this Handbook, Notwithstanding its length, this Handbook can "Applied Trade Policy for Developing Countries," only partially address the many policy issues that contains all of the major agreements and many arise in the course of efforts to integrate into the other WTO documents. Therefore, only key aspects world economy. The focus is on trade policy, broad- of WTO rules are discussed. Relatively more atten- ly defined to cover both traditional instruments of tion is given to the General Agreement on Trade in commercial policy--tariffs, customs administra- Services (GATS) and the Trade-Related Aspects of tion, and so on--and "new" issues such as services, Intellectual Property Rights (TRIPS) agreement intellectual property, and the behind-the-border than to General Agreement on Tariffs and Trade regulatory agenda that has implications for market (GATT) disciplines, since a wealth of analysis and access conditions. The approach is one of multiple information exists on "traditional" trade policy voices; the contributors include many authors who instruments. The chapters in the Handbook dealing have no connection to the World Bank. In all cases, with merchandise trade issues focus primarily on contributors wrote in a personal capacity, and their those subjects that are of greatest interest to devel- xxix D E V E L O P M E N T, T R A D E , A N D T H E W T O oping countries--tariff peaks, preferences, rules of Each of the eight parts begins with a short intro- origin, customs clearance and trade facilitation, duction that is intended as a reader's guide to the local content and other industrial policy measures, issues and to further reading. Annotated references, and export promotion. drawn in part from Hoekman and Kostecki (2001), One topic that is of major importance to develop- are listed in each introduction for those who are ing countries--agriculture and agricultural trade interested in pursuing in-depth discussion and policies--is not addressed in any depth in this analysis. To facilitate consultation of the citations in Handbook because it is the subject of a companion individual chapters, we have compiled an integrated volume (Ingco and Nash forthcoming). The same is bibliography, found at the end of the Handbook. true for another major issue area: trade and poverty. The boxes included in the chapters illustrate spe- Winters, in Chapter 5 of this volume, summarizes cific points or describe specific cases. Many of these the main messages and conclusions that have boxes were prepared by the editors of the Hand- emerged from the literature on trade and poverty. book, drawing on papers prepared for this project Those seeking a more in-depth treatment should and on the literature. The chapter authors are not consult McCulloch, Winters, and Cirera (2001), responsible for the content of those boxes written which reviews the literature and good practices con- by the volume editors or by other contributors. cerning the design of trade policy reforms from a The appendix includes a glossary and a set of poverty alleviation perspective. tables that present data on trade barriers for a large sample of countries. The glossary provides a listing of major WTO articles and provisions for the con- Structure of the Handbook venience of readers who are not familiar with the This volume has eight parts. The chapters in Part I WTO, as well as succinct descriptions and defini- place trade policy reform in a development context tions of key trade-related institutions and policies. and discuss key dimensions of reform. Part II deals A CD-ROM containing data on tariffs, trade, and with the main aspects of the WTO. Parts III, IV, and production is packaged with the Handbook. The V cover the areas that are the subject of WTO rules: database is described briefly in the appendixes and trade in goods, trade in services, and the protection is more fully documented on the CD-ROM. We of intellectual property. Part VI contains discussions have also included in the appendixes and on the of a number of regulatory issues; many of these CD-ROM a short guide to the most commonly used have not been subject to multilateral rules but are indicators and indices that can be applied to the now being introduced into the WTO agenda due to data. More detailed datasets and analytical tools strong interest on the part of some high-income that can be used for negotiations are being devel- countries and nongovernmental organizations. Spe- oped in cooperation with UNCTAD; this set of tools cific process-related concerns of developing coun- will be released in mid-2002. tries are the subject of Part VII; these include A second CD-ROM contains the teaching modules participation in the WTO, capacity building, and developed by Jaime de Melo and Marc Bacchetta implementation of WTO agreements. Finally, the during their many years of conducting an intensive chapters in Part VIII summarize a number of rules two-week course for government officials, cospon- of thumb for good trade policy and review sored by the World Bank Institute and the WTO. The approaches to using the WTO (and regional agree- CD-ROM also includes an extensive set of readings ments) as instruments for promoting development. and, as noted above, official WTO documents. xxx I TRADE POLICY REFORM IN CONTEXT I t is useful to make a distinction between appropriate macroeconomic incentive environment reduction in border barriers to trade and and embedded in a comprehensive development and "behind-the-border" complementary poli- poverty reduction strategy. Complementary institution- cies that are critical in supporting trade policy reforms. al reform efforts and improvements in the legal and The first set of policies focuses on creating incentives regulatory environment that increase investor confi- for efficient growth by reducing the average level and dence are vital if trade liberalization is to serve as an the dispersion of border protection, eliminating non- engine of growth. Key elements of the associated tariff barriers (NTBs), and strengthening the public behind-the-border trade agenda include efficient regu- institutions needed to ensure that goods cross fron- latory regimes, institutions that support the participa- tiers with low transactions costs (i.e., efficient customs tion of national firms in international markets, and regimes that minimize red tape). The second set has measures to enhance the competitiveness of these firms to do with regulatory standards and policies to ensure by providing access to crucial services inputs. Dani that supply responses to liberalization are efficient, Rodrik, in Chapter 1, examines some of these issues. equitable, and enduring. Important issues here The trade agenda has become increasingly com- include liberalization of trade in services sectors to plex in the past decade, and where it starts and enhance competitiveness, policies to promote access ends is not clear. What is clear is that the standard to information and technology, and strengthening of approach found in most textbooks--which focuses institutions in order to benefit from participation in on policy instruments that are applied at the border regional and multilateral trade arrangements. and that affect the domestic prices of goods or The chapters in this part focus on aspects of the com- export prices (tariffs, quotas, export subsidies, and plementary agenda. A strategy for sustained trade taxes)--is too narrow. In practice, the trade agenda expansion and growth must be framed within an spans all policies that have the effect of discriminat- 1 T R A D E P O L I C Y R E F O R M I N C O N T E X T ing against foreign providers (suppliers) of goods, A key message that emerges from the chapters in services, and production factors (knowledge, labor, this part is the need for analysis that focuses not and capital), and it takes in the functioning of insti- only on trade policy narrowly defined but also on tutions that affect the investment climate in a coun- the complementary reforms and institutions that are try. The recognition that trade policy has a much required if trade reforms are to benefit society. Such wider ambit than border policies implies that gov- analysis should include a diagnosis of the current sit- ernments and civil society must have a broad focus uation, benchmarking in relation to good practice and must consider the interrelationships between and competitors, determination of the incentive different policy areas and the operation and effec- and redistributional implications of status quo poli- tiveness of existing institutions. Kym Anderson, in cies and possible changes, and identification of the Chapter 2, emphasizes the need for an economy- complementary actions that are needed to make wide perspective on trade policy reform. trade reform an effective component of a poverty- Key complementary factors that often determine reducing growth strategy. the success of trade policy reform are the real exchange rate and the ability of the government to Further Reading maintain revenue collection objectives. As Howard J. Schatz and David G. Tarr document in Chapter 3, Jeffrey Sachs and Andrew Warner, "Economic Reform although countries may maintain different types of and the Process of Global Integration," Brookings exchange rate regimes, allowing the real exchange Papers on Economic Activity, 1 (1995): 1­118, is a rate to appreciate significantly over time has often widely read and influential empirical study that finds led to the failure of trade reforms. Chapter 4, by an unambiguous positive relationship between open- Liam Ebril, Janet Stotsky, and Reint Gropp, examines ness and economic performance. Dani Rodrik, Has the fiscal implications of trade liberalization. Tariff Globalization Gone Too Far? (Washington, D.C.: Insti- revenue remains important for many low-income tute for International Economics, 1997), provides a countries. In pursuing further tariff reform, efforts skeptical view of the benefits of globalization for have to be made to develop alternative domestic growth and welfare in the absence of the institutions tax bases and to ensure that reliance on tariff rev- and policies needed to manage downside risks. An enues does not needlessly distort resource allocation accessible account of the effects of the inward-look- incentives. Cross-country experience suggests that ing, import-substituting development strategies pop- policy reforms can be designed so as to maintain or ular in the 1960s and 1970s, as well as the shift increase revenue collection. toward more outward-looking policies in the 1980s, Although the available research indicates that trade is given by Jagdish Bhagwati in Protectionism (Cam- liberalization reduces poverty overall, segments of the bridge, Mass.: MIT Press, 1988). Edward Buffie, Trade poor may be hurt by it, and in Chapter 5, L. Alan Win- Policy in Developing Countries (Cambridge, U.K.: Cam- ters looks at the interactions between trade reform bridge University Press, 2001), analyzes trade policy and poverty alleviation. Reform programs supporting in an integrated framework that allows for economic liberalization must be complemented by efforts to dynamics and incorporates the structural features of strengthen social safety nets. Since some of the poor developing countries. Neil McCulloch, L. Alan Win- are likely to be so destitute that any decrease in ters, and Xavier Cirera, Trade Liberalization and Pover- incomes will impose extreme hardship, it is important ty: A Handbook (London: Centre for Economic Policy to identify which of them may be adversely affected Research, 2001), provides a comprehensive treat- by reforms and to determine the most appropriate set ment of the links between trade and poverty in the of policies to complement trade reform. context of the WTO. 2 1 D A N I R O D R I K Trade Policy made and implemented, estab- lishes new constraints and oppor- tunities for economic policy more Reform as broadly, creates a new set of stake- holders while disenfranchising the Institutional previous ones, and gives rise to a new philosophy (alongside a new Reform rhetoric) on what development policy is all about. Hence, trade reform ends up being much more than a change in relative prices: it results in institutional reform of a major kind. E In the language of economics, conomists are trained to think institutional reform changes not only policy param- about trade policy reform in terms eters but also behavioral relationships. Correspond- of changes in the levels of tariffs and quantitative ingly, the resource-allocation and dynamic restrictions (QRs) and the shifts in relative prices consequences of trade reform become harder to dis- brought about by these alterations. They use eco- cern using the type of analysis that is the applied nomic models, supplemented by quantitative esti- economists' stock in trade. Household behavior and mates of elasticities, to analyze the implications of investment decisions get altered in ways that are dif- changes in tariffs and QRs for production, con- ficult to track in the absence of knowledge about the sumption, and trade. By tweaking their models suf- "deep parameters" of the economy. When the ficiently, they can predict the likely impacts on reform is well designed and consistent with the employment, poverty and distribution, macroeco- institutional needs of the economy, it can spur nomic balances, and the government budget. If they unexpected levels of entrepreneurial dynamism and are ambitious (reckless?), they will also pass judg- economic growth. When it is not, it can result in a ment on dynamic efficiency, technological progress, stagnation that will appear surprising. and long-run economic growth. Viewing trade reform as institutional reform Policymakers often have a different perspective on helps clarify the criteria by which trade reform trade reform. For them, the actual changes in tariff should be evaluated. My main argument in this schedules are typically only a small part of the process. chapter is that the relevant criterion is neither open- What is at stake is a deeper transformation of the pat- ness to trade nor consistency with existing WTO terns of behavior within the public sector, and of the rules.1 The yardstick that matters is the degree to government's relationship with the private sector and which trade reform contributes to the construction the rest of the world. The reform goes beyond particu- of a high-quality institutional environment at home. lar levels of tariffs and QRs: it sets new rules and My working hypothesis, supported by empirical expectations regarding how these policy choices are evidence to which I will refer below, is that a high- 3 T R A D E P O L I C Y R E F O R M I N C O N T E X T quality institutional environment has greater eco- oriented reforms in Latin America and the growing nomic payoffs than a liberal trade regime or adher- realization that these reforms have paid too little ence to WTO rules. attention to mechanisms of social insurance and to In practice, there may be some important safety nets. The third and most recent is the Asian spillovers among these objectives. To cite an impor- financial crisis, which has shown that allowing tant illustration, a free trade regime is likely to financial liberalization to run ahead of financial reduce the corruption and rent-seeking associated regulation is an invitation to disaster. A number of with trade interventions. Similarly, tariff bindings recent empirical studies have highlighted the under the WTO may generate greater predictability importance of high-quality institutions in shaping in incentives and solidify property rights--two economic performance (see, especially, Kaufmann, important attributes of a high-quality institutional Kraay, and Zoido-Lobatón 1999; Acemoglu, John- framework. But while free trade and WTO rules can son, and Robinson 2000). contribute to the emergence of high-quality institu- Following Lin and Nugent (1995: 2306­07), it is tions, these are not one and the same. Institutional useful to think of institutions broadly as "a set of development takes time and often requires humanly devised behavioral rules that govern and unorthodox and divergent choices. Some of the shape the interactions of human beings, in part by most spectacular cases of development in the post- helping them to form expectations of what other war period have been the product of gradualist, people will do." All well-functioning market two-track modes of institutional reform (Rodrik economies are "embedded" in a set of nonmarket 2000b). The type of investments in institution- institutions, without which markets cannot per- building required for full adherence to WTO agree- form adequately. I will highlight below five types of ments on, say, customs valuation or intellectual market-supporting institutions in particular: prop- property rights (IPRs) may not be the first order of erty rights, regulatory institutions, institutions for business for low-income countries with more macroeconomic stabilization, institutions for social urgent needs (Finger and Schuler 2000). Since insurance, and institutions of conflict management. human resources, administrative capacity, and I emphasize as well the variety of institutional political capital are scarce, especially in developing setups that is compatible with superior economic countries, policymakers need to have a good sense performance. of the priorities. An implication of this line of reasoning is that we Property Rights should think of the trade regime and WTO rules as being at the service of developing countries' institu- As North and Thomas (1973) and North and Wein- tional needs, not vice versa. Governments that gast (1989), among many others, have argued, the understand this are the ones that are likely to make establishment of secure and stable property rights the most of trade reform. has been a key element in the rise of the West and the onset of modern economic growth. It stands to reason that an entrepreneur would not have the Institutional Prerequisites for Development incentive to accumulate and innovate unless s/he Price reforms--in external trade, in product and had adequate control over the return to the assets labor markets, in finance, and in taxation--were the that are thereby produced or improved. Note that rallying cry of the reformers of the 1980s, along the key word is "control," rather than "ownership." with macroeconomic stability and privatization. By Formal property rights do not count for much if the 1990s, it had become clear that incentives would they do not confer control rights. By the same not work, or would generate perverse results, in the token, sufficiently strong control rights may do the absence of adequate institutions. Three sets of dis- trick even in the absence of formal property rights. parate developments have conspired to put institu- In Russia today, shareholders have property rights tions squarely on the agenda of reformers. One of but often lack effective control over enterprises, these was the dismal failure in Russia of price whereas in China's township and village enterprises reform and privatization in the absence of a sup- control rights have spurred entrepreneurial activity portive legal, regulatory, and political apparatus. A even in the absence of clearly defined property second is the lingering dissatisfaction with market- rights. 4 Trade Policy Reform as Institutional Reform As these instances illustrate, establishing "proper- States has both the world's freest markets and the ty rights" is rarely a matter of just passing a piece of world's toughest antitrust enforcement. The lesson legislation. Legislation in itself is neither necessary that market freedom requires regulatory vigilance nor sufficient for the provision of secure control has been driven home recently by experience in East rights. In practice, control rights are upheld by a Asia. In the Republic of Korea and in Thailand, as in combination of legislation, private enforcement, so many other developing countries, financial liber- and custom and tradition. They may be distributed alization and capital account opening led to finan- more narrowly or more diffusely than property cial crisis precisely because of inadequate prudential rights. Moreover, property rights are rarely absolute, regulation and supervision. even when set formally in the law. Each society In developing countries, where market failures are decides for itself the scope of allowable property pervasive, regulatory institutions may need to rights and the acceptable restrictions on their exer- extend beyond the standard list covering antitrust, cise. Intellectual property rights are protected assid- financial supervision, securities regulation, and the uously in the United States and most advanced like. Recent models of coordination failure and cap- societies but not in many developing countries. By ital market imperfections make it clear that strategic contrast, zoning and environmental legislation government interventions may often be required to restricts the ability of households and enterprises in escape low-level traps and elicit desirable private rich countries to do as they please with their "prop- investment responses.2 The experience of Korea and erty" to a much greater extent than is the case in Taiwan (China) in the 1960s and 1970s can be inter- developing countries. All societies recognize that preted in that light. The extensive subsidization and private property rights can be curbed if doing so government-led coordination of private investment serves a greater public purpose. It is the definition in these two economies played a crucial role in set- of what constitutes a "greater public purpose" that ting the stage for self-sustaining growth. It is clear varies. that many other countries have tried and failed to replicate these institutional arrangements. And even Korea may have taken a good thing too far by main- Regulatory Institutions taining the cozy institutional linkages between the Markets fail when participants engage in fraudulent government and chaebols well into the 1990s, at or anticompetitive behavior. They fail when trans- which point these ties may have become dysfunc- actions costs prevent the internalizing of technolog- tional. Once again, the lesson is that desirable insti- ical and other nonpecuniary externalities. And they tutional arrangements vary, and that they vary not fail when incomplete information results in moral only across countries but also within countries over hazard and adverse selection. Economists recognize time. these failures and have developed the analytical tools required to think systematically about their Institutions for Macroeconomic Stabilization consequences and the possible remedies. Theories of the second best, imperfect competition, agency, Markets are not necessarily self-stabilizing. Keynes and mechanism design, to name but a few, offer an and his followers worried about shortfalls in aggre- almost embarrassing choice of regulatory instru- gate demand and the resulting unemployment. ments for countering market failures. Theories of More recent views of macroeconomic instability political economy and public choice offer cautions stress the inherent instability of financial markets against unqualified reliance on these instruments. and its transmission to the real economy. All In practice, every successful market economy is advanced economies have come to acquire fiscal overseen by a panoply of regulatory institutions that and monetary institutions that perform stabilizing regulate conduct in goods, services, labor, asset, and functions, having learned the hard way about the financial markets. A few acronyms from the United consequences of not having them. Probably most States will suffice to give a sense of the range of important among these institutions is a lender of institutions involved: FTC, FDIC, FCC, FAA, last resort--typically, the central bank--that guards OSHA, SEC, EPA, and so on. In fact, the freer are the against self-fulfilling banking crises. markets, the greater is the burden on regulatory There is a strong current within macroeconomics institutions. It is not a coincidence that the United thought that disputes the possibility or effectiveness 5 T R A D E P O L I C Y R E F O R M I N C O N T E X T of stabilizing the macroeconomy through monetary represented well by the Japanese case, is one in and fiscal policies. There is also a sense in policy cir- which social insurance is provided through a com- cles, particularly in Latin America, that fiscal and bination of enterprise practices (such as lifetime monetary institutions, as currently configured, have employment and enterprise-provided social bene- added to macroeconomic instability, rather than fits), sheltered and regulated sectors (mom-and- reduced it, by following procyclical rather than anti- pop stores), and an incremental approach to cyclical policies. These developments have spurred liberalization and external opening. the trend toward central bank independence and Social insurance legitimizes a market economy have helped open a new debate on designing more because it renders it compatible with social stability robust fiscal institutions. Some countries (Argenti- and social cohesion. But the existing welfare states na being the most significant example) have given in Western Europe and the United States engender a up on a domestic lender of last resort altogether by number of economic and social costs--mounting replacing their central bank with a currency board. fiscal outlays, an "entitlement" culture, long-term The debate over currency boards and dollarization unemployment--that have become increasingly illustrates the obvious, but occasionally neglected, apparent. Partly because of this experience, devel- fact that the institutions needed by a country are oping countries, such as the countries in Latin not independent of that country's history. America that adopted the market-oriented model following the debt crisis of the 1980s, have not paid sufficient attention to creating institutions of social Institutions for Social Insurance insurance. The upshot has been economic insecuri- One of the liberating effects of a dynamic market ty and a backlash against the reforms. How these economy is that it frees individuals from their tradi- countries will maintain social cohesion in the face tional entanglements--the kin group, the church, of large inequalities and volatile outcomes, both of the village hierarchy. The flip side is that it uproots which are being aggravated by the growing reliance them from traditional support systems and risk- on market forces, is an important question that has sharing institutions. Gift exchanges, the fiesta, and no obvious answer. kinship ties--to cite just a few of the social arrange- ments for equalizing the distribution of resources in Institutions of Conflict Management traditional societies--lose many of their social insurance functions. And as markets spread, the tra- Societies differ in their cleavages. Some are made up ditional ways of managing the risks that have to be of an ethnically and linguistically homogenous insured against become much less effective. A mod- population marked by a relatively egalitarian distri- ern market economy is one where idiosyncratic bution of resources. Others are characterized by (individual-specific) risk to incomes and employ- deep cleavages along ethnic or income lines. These ment is pervasive. divisions often hamper social cooperation and The huge expansion of publicly provided social engender social conflict. Economists have used insurance programs during the 20th century is one models of social conflict to shed light on questions of the most remarkable features of the evolution of such as: Why do governments delay stabilizations advanced market economies. In the United States it when delay imposes costs on all groups? Why do was the trauma of the Great Depression that paved countries rich in natural resources often do worse the way for major institutional innovations in this than countries that are resource-poor? Why do area: social security, unemployment compensation, external shocks often lead to protracted economic public works, public ownership, deposit insurance, crises that are out of proportion to the direct costs and legislation favoring unions. In Europe the roots of the shocks themselves? of the welfare state reached in some cases to the tail Healthy societies have a range of institutions that end of the 19th century. But the striking expansion make such colossal coordination failures less likely. of social insurance programs, particularly in the The rule of law, a high-quality judiciary, representa- smaller economies most open to foreign trade, was tive political institutions, free elections, independent a post­World War II phenomenon. Social insurance trade unions, social partnerships, institutionalized need not always take the form of transfer programs representation of minority groups, and social insur- paid out of fiscal resources. The East Asian model, ance are examples of such institutions. What makes 6 Trade Policy Reform as Institutional Reform these arrangements function as institutions of con- tries that face the prospect of adopting the EU's flict management is that they entail a double "com- Common Agricultural Policy or its antidumping mitment technology": they warn the potential regime. It all depends on the circumstances and on "winners" from social conflict that their gains will be how national governments are able to use such cir- limited and assure the "losers" that they will not be cumstances. expropriated. They tend to increase the incentives One way that governments can use institutional for social groups to cooperate by reducing the payoff arbitrage to good effect is to enhance the credibility to socially uncooperative strategies. of domestic institutions. For example, the new disci- plines imposed on developing country governments by the WTO--in the areas of tariff bindings, quanti- Trade Policy and Institutional Reform tative restrictions, services, subsidies, trade-related What is the link between trade policy reform and investment measures (TRIMs), and intellectual these institutions? Trade reform often entails the property--can be viewed as helping these govern- importation of institutions from abroad. Sometimes ments overcome traditional weaknesses in their style this is the outcome of deliberate policy actions to of governance. These disciplines impose a certain "harmonize" a country's economic and social insti- degree of predictability, transparency, rule-bound tutions with those of its trading partners. Member- behavior, and nondiscrimination in areas of policy ship in the WTO, for example, requires the adoption often subject to discretion and rent-seeking. In the of a certain set of institutional norms: nondiscrimi- same vein, perhaps the greatest contribution of the nation in trade and industrial policies, transparency North American Free Trade Agreement (NAFTA) to in the publication of trade rules, WTO-consistent the Mexican economy was the element of irre- patent and copyright protection, and so on. Similar- versibility and "cementing" that the agreement has ly, membership in the European Union (EU) contributed to Mexico's economic reforms. In requires the adoption of wide-ranging legal and Europe the accession of Greece, Portugal, and Spain bureaucratic requirements set down in Brussels. to the EU has made return to military dictatorship in At other times, institutional arbitrage is the result those countries virtually unthinkable. of the working out of market forces. Mobility of Imported institutions, however, can also turn out employers around the world, for example, makes it to be ill suited or counterproductive. Many of the harder to tax corporations and tilts national regimes labor standards that some labor groups in the North toward the taxation of nontraded goods and factors, would like developing countries to adopt--such as such as labor. Financial integration raises the premi- higher minimum wages or restrictions on some um for macroeconomic stability and makes central kinds of child labor--may fit in this category. The bank independence look more desirable. Finally, new patent restrictions called for by the Trade-Relat- openness can change national institutions by alter- ed Aspects of Intellectual Property Rights (TRIPS) ing the preferences that underlie them. Civil liber- agreement of the WTO are at best a mixed blessing ties and political freedoms are among the most for countries such as India that have so far benefited important imported concepts in the developing from cheap pharmaceuticals. A similar argument world; the demands for democracy to which these can be made about pressures for tightening environ- ideas give rise are a direct product of openness in mental standards in developing countries. this broad sense. Successful institutional reforms typically combine Arbitrage in markets for goods and capital, in the imported blueprints with local flavor. A good exam- absence of second-best complications, is associated ple of this in the area of trade comes from Mauri- with normatively desirable outcomes; it increases tius, where superior economic performance has efficiency. One cannot make the same presumption been built on a peculiar mix of orthodox and het- where arbitrage in institutions is concerned. There erodox strategies. This economy's success derives in are no theorems stating that institutional conver- large part from an export-processing zone (EPZ), gence, harmonization, or "deep integration" which operates under free trade principles. The EPZ through trade is inherently desirable. While many of has enabled a boom in exports of garments to Euro- the examples cited above involve outcomes that are pean markets and an accompanying investment desirable (greater democracy, for instance), this is boom at home. Yet the island's economy has com- not true of all possible outcomes. Think of the coun- bined the EPZ with a domestic sector that was high- 7 T R A D E P O L I C Y R E F O R M I N C O N T E X T ly protected until the mid-1980s. The origins of this tary and phytosanitary measures (SPS), and intel- essentially dual-track strategy (not unlike that fol- lectual property rights (IPRs)--a sum equal to a lowed in China) lay in the social and political make- year's development budget for many of the least- up of the island and in the decision by policymakers developed countries. Would this be money well not to disrupt a fragile ethnic situation through an spent? Finger argues that for the vast majority of across-the-board liberalization that would have dis- developing countries, the answer is no. Although advantaged established import-substituting groups. these countries would benefit from the strengthen- The EPZ scheme, in fact, provided a neat way ing of their institutions in the relevant areas, the around the political difficulties. The creation of the reality is that "WTO obligations reflect little aware- EPZ generated new opportunities in trade and ness of development problems.""Other alternatives, employment without removing protection from the e.g., basic education for women and girls, would import-substituting groups or from the male work- have much more attractive rate-of-return numbers" ers who dominated the established industries. The (Finger 1999). It is a safe bet that any new trade segmentation of labor markets early on between round will shorten the leash on developing coun- male and female workers, with women predomi- tries further, even if pressure in the controversial nantly employed in the EPZ, was crucial, as it pre- areas of environment and labor can be fended off. vented the expansion of the EPZ from driving wages Integration into the world economy has other, up in the rest of the economy and hurting import- more subtle institutional requirements, as well. substituting industries. New profit opportunities Openness implies heightened exposure to external were created at the margin while leaving old oppor- risk and, consequently, greater demand for social tunities undisturbed. insurance. Greater provision of social insurance One can cite other instances of heterodox trade seems to be a key factor behind the empirical regu- reforms that proved successful because they suited larity that governments tend to be bigger in existing political and institutional realities. Korea's economies where trade makes up a higher share of outward orientation during the 1960s, for example, GDP (Rodrik 1998). More broadly, openness was achieved not through import liberalization (of increases the premium on institutions of conflict which there was little), but through export subsi- management (Rodrik 1999). dization (of which there was a lot). This type of It is often overlooked that the most successful reform is now prohibited under existing WTO rules "globalizers" of an earlier era--the East Asian on subsidies. Similarly, China's two-track reform "tigers"--had to abide by few international con- strategy in agriculture, industry, and trade, which straints and had to pay few of the costs of integra- maintained nonmarket institutional forms while tion during their formative growth experience in aligning incentives correctly at the margin, has been the 1960s and 1970s. Global trade rules essentially wildly successful. These are cases in which imagina- gave them a free ride, and capital mobility was hard- tive experimentation with institutional reform has ly an issue. This is why these countries can hardly be had, in all likelihood, greater payoffs than the considered poster children for today's globalization. wholesale transplantation of institutions from Korea, Taiwan (China), and the other East Asian advanced industrial countries would have had.3 economies had the freedom to do their own thing, and they used it abundantly. As noted above, they combined their reliance on trade with unorthodox Integration into the World Economy as a policies--export subsidies, domestic content Model of Institutional Reform requirements, import-export linkages, patent and WTO membership entails institutional reforms that copyright infringements, restrictions on capital are not only demanding, but also of a particular flows (including on foreign direct investment), kind. One can question, as Michael Finger has elo- directed credit, and so on--that are either preclud- quently done, the fit between these reforms and the ed by today's rules or greatly frowned on. The envi- needs of developing countries, particularly of the ronment for today's globalizers is quite different. least developed among them. Finger has calculated None of the institutional reforms needed for that it would cost a typical developing country insertion in the world economy is bad in and of US$150 million to implement requirements under itself, and in fact, many of them can be indepen- the WTO agreements on customs valuation, sani- dently desirable, as I argued above. Some can also 8 Trade Policy Reform as Institutional Reform have unintended benefits. For example, a govern- phenomena (overvalued currencies or macro insta- ment that is forced to protect the rights of foreign bility) or geographic determinants (such as location investors perhaps becomes more inclined to protect in the tropical zone) to trade policies proper. Once the basic human rights of its own citizens, too. This simple corrections are made for such problems, one was a potent argument in U.S. debates, prior to rarely finds a statistically significant relationship China's accession to the WTO, about China's per- between the level of tariff and nontariff barriers and manent normalized trade relations (PNTR) status. economic growth across countries. But one has to recognize that a strategy of institu- There are, in fact, reasons to be skeptical about tional reform based on global integration is a strate- the existence of a general, unambiguous relation- gy of trickle-down institutional reform. The ship between trade openness and growth. The rela- reforms may or may not trickle down, and even tionship is likely to be a contingent one, dependent when they do, they will rarely constitute the most on a host of country and external characteristics. effective way of targeting the desired ends, whether The fact that practically all of today's advanced those ends are legal reform, improved observance of countries embarked on their growth behind tariff human rights, or reduced corruption. Institutional barriers and reduced protection only subsequently change is costly and requires the expenditure of surely offers a clue of sorts. Moreover, the modern scarce human resources, administrative capabilities, theory of endogenous growth yields an ambiguous and political capital. The priorities implied by glob- answer to the question of whether trade liberaliza- al insertion will not always coincide with the priori- tion promotes growth. The answer varies depending ties of a more fully developmental agenda. on whether the forces of comparative advantage push the economy's resources in the direction of activities that generate long-run growth (via exter- Can We Rely on a Growth Payoff from nalities in research and development, expansion of Openness? product variety, upgrading of product quality, and Global integration carries opportunity costs because so on) or divert them from such activities. Finally, as of the institutional consequences that such a strategy I have stressed throughout, the institutional setting entails. These costs have to be traded off against the in which trade policy operates is more important expected benefits. All economists know that gains for economic performance than the levels at which from trade exist, but the standard gains from trade specific trade barriers are set. tend to be small. The tendency in policy discussions No country has developed successfully by turning has been to go considerably beyond the standard case its back on international trade and long-term capi- for trade and to claim that open trade policies pro- tal flows. Very few countries have grown over long duce significant boosts in economic growth rates. periods of time without experiencing an increase in This claim is apparently supported by a large cross- the share of foreign trade in their national product. national empirical literature. Recently, Francisco In practice, the most compelling mechanism that Rodríguez and I reviewed the extensive literature on links trade with growth in developing countries is the relationship between trade policy and growth that imported capital goods are likely to be signifi- (Rodríguez and Rodrik 2001) and reached the con- cantly cheaper than those manufactured at home. clusion that there is a significant gap between the Policies that restrict imports of capital equipment, message that the consumers of this literature have raise the price of capital goods at home, and thereby derived and the "facts" that the literature has actually reduce real investment levels have to be viewed as demonstrated. The gap emerges from a number of undesirable prima facie. Exports, in turn, are factors. In many cases the indicators of "openness" important, since that is what one purchases import- used by researchers are problematic as measures of ed capital equipment with. trade barriers or are highly correlated with other But it is equally true that no country has devel- sources of poor economic performance. In other oped simply by opening itself up to foreign trade cases the empirical strategies used to ascertain the and investment. The trick in the successful cases has link between trade policy and growth have serious been to combine the opportunities offered by world shortcomings, the removal of which results in signif- markets with a domestic investment and institu- icantly weaker findings.4 One common problem has tion-building strategy to stimulate the animal spir- been the misattribution of either macroeconomic its of domestic entrepreneurs. Almost all of the 9 T R A D E P O L I C Y R E F O R M I N C O N T E X T outstanding cases--East Asia, China, India since the along these lines. I have argued here that the first early 1980s--involve partial and gradual opening question policymakers contemplating trade reform up to imports and foreign investment. should ask is not whether the reform will result in The appropriate conclusion to draw from the evi- higher volumes of trade, render their trade regime dence is not that trade protection should, as a rule, more liberal, or increase market access abroad but be preferred to trade liberalization. There is no evi- whether it will improve the quality of institutions at dence from the past 50 years that trade protection is home. The results of trade negotiations--whether systematically associated with higher growth. The bilateral, regional, or multilateral--should be point is simply that the benefits of trade openness judged by the same yardstick. should not be oversold. When other worthwhile policy objectives are competing for scarce adminis- Notes trative resources and political capital, deep trade liberalization often does not deserve the high prior- This chapter draws heavily on several earlier papers, in particular ity it typically receives in development strategies. Rodrik 1999, 2000a, and 2000b. This is a lesson that is of particular importance to 1 It should go without saying that openness to trade and adher- countries, such as those in Africa, that are in the ence to WTO rules are not the same thing. A country can fol- early stages of reform. low free trade policies without being a member of the WTO, and many WTO rules are at variance with free trade (as in the cases of antidumping, safeguards, and regional agreements). Conclusion 2 See Hoff and Stiglitz (2000) for a useful survey and discussion. A high-quality policy environment is one that sends 3 See Kapur and Webb (2000) and Pistor (2000) for useful dis- clear signals to producers and investors, precludes cussions of the limitations of importing legal and institutional rent-seeking, does not waste economic resources, is forms from abroad. consistent with the administrative capabilities of the 4 Our detailed analysis covers the five papers that are probably government, and maintains social peace. Trade pol- the best known in the field: Dollar (1992); Sachs and Warner icy reform contributes to economic development (1995); Ben-David (1993); Edwards (1998); and Frankel and insofar as it helps build high-quality institutions Romer (1999). 10 2 K Y M A N D E R S O N Economywide taking an economywide perspec- tive when considering the effects of actual policies at home or Dimensions of abroad or of potential policy reforms. Given the significance of Trade Policy and agriculture in low-income coun- tries, the chapter focuses primari- Reform ly on the possible direct and indirect effects of policies on this sector, emphasizing the need to consider the impact of input as well as output price distortions on producer incentives. E very country has an interest in Direct Effects of Policies: A Single-Sector trade policy reform. This is true Perspective even for the most open of economies because, although that government may not be distorting Historically, the governments of poor agrarian incentives, government policies of many other economies have taxed farmers in one way or anoth- countries are distorting the prices received by the er (Krueger, Schiff, and Valdés 1988). Sometimes it open economy's exporters in international markets. has been an in-kind tax, such as a proportion of Moreover, it is relative prices that matter: the incen- grain output. In other settings, where a cash crop tives facing producers or consumers of a particular was being exported, producers often have been product can be distorted not only by policies direct- required to sell to a statutory marketing authority ly affecting the price of that product but also, and that paid them only a fraction of the export price. sometimes even more strongly, by policies affecting Either way, farmers receive less than the free-market the prices of products that are substitutes or com- price for their produce. Except in the unlikely event plements in production or consumption. Govern- that all of those taxes come back to farmers in the ment intervention in currency markets also can form of government goods and services they other- have nontrivial distortionary effects on incentives. wise would have purchased with that taxed income, Farmers, for example, may receive the international the incentive to produce and market farm products price for their produce and yet be harmed by having is reduced. to convert from foreign to domestic currency at an Governments of such agrarian economies typical- artificially low exchange rate. ly return little of the proceeds of those taxes to farm This chapter explores not just the direct but also the families, especially at early stages of the country's various indirect ways in which trade and trade-related development. Rather, the taxes tend to be used to policies affect the welfare of people in developing develop urban infrastructure, pay officials relatively countries. Its purpose is to identify the importance of high wages, subsidize food consumption, and so on. 11 T R A D E P O L I C Y R E F O R M I N C O N T E X T Until recently, it was widely believed that taxing pri- import-protection policy encourages the allocation mary producers for such purposes would not reduce of too many resources to agriculture's import-com- output significantly because farm families were poor peting industries, and it also harms consumers of or had no alternative uses for their time, land, and those importables via higher food prices. other resources. Empirical studies during the past What implications does the Krueger, Schiff, and half-century, however, have shown that farmers in Valdés study have for reforming agricultural policy even the poorest settings are quite price-responsive in the average developing country? Reducing export (Askari and Cummings 1977). When the proceeds restrictions would cause the domestic price of from growing a marketable product are reduced, exported farm products to rise by up to one-eighth, farm households divert at least some of their helping producers of those exportables but hurting resources to producing other products or to leisure domestic buyers of the products (who may be down- pursuits. Only the very poorest subsistence farmers stream processors). That reform might also encour- might be enticed by such taxes to work harder, but age producers of import-competing farm products even that response may be welfare reducing in that to switch their production to exportables that now they then have less recreational time and are likely to carry higher prices. If the country's food import live less healthy and shorter lives. restrictions also were reduced, producers of those It matters that farm household resources are importables would see their output price decline diverted from producing the taxed good because and would consider switching to other farm prod- such diversion means that society's resources are ucts. This would reinforce the encouragement of not being used where they are most productive. A exportable production in agriculture, insofar as the farmer discouraged from specializing in growing a resources used in the two different farm subsectors cash crop, for example, has less to spend on other are substitutable. Both types of reform improve the products and therefore is less able to encourage oth- efficiency of resource use in the sector by encourag- ers to specialize in doing what they do best also. ing greater exploitation of the country's agricultural Likewise, it matters if farmers have to pay more for comparative advantage. That is, reforms that boost inputs purchased from nonfarm sectors (for exam- the relative profitability of the industries previously ple, because of import taxes on these goods): they discouraged by the government's trade-restrictive then buy less of those inputs in relation to other policies tend to be welfare enhancing. inputs than is optimal. This direct effect is, however, of much less economic importance than the indi- Indirect Effects of Policies: An Intersectoral rect effect of such industrial or service sector pro- Perspective tection policies, as is made clear in the next section. Of course, not all agricultural producers in devel- The above lesson applies not just within the agricul- oping countries face artificially depressed prices for tural sector but also to interactions between it and their products. Indeed import-competing produc- other sectors. That is, farmers also can be discour- ers of some key food items enjoy protection from aged, albeit indirectly, by nonagricultural policy import competition that raises the domestic price of interventions. One source of such discouragement their produce above free-market levels. An empiri- comes from import protection to producers of non- cal study of 18 developing countries between the farm products. In an economy producing just two mid-1970s and the mid-1980s contrasted the treat- sets of goods, importables and exportables, a tax on ment of major farm export products with that of imports is equivalent to a tax of equal size on key imported foods (Krueger, Schiff, and Valdés exports whenever the two sets of goods use com- 1988). The authors found that domestic prices of mon resources such as labor and capital. Both taxes the imported foods averaged about 20 percent raise the prices of importables relative to exporta- above prices at the countries' borders, whereas bles, and by the same amount, and it is that price domestic prices for key agricultural export items ratio which determines the allocation of resources were, on average, 11 percent below international between the two sectors (Lerner 1936). levels. Both types of distortion are harmful to More generally, when domestic prices of some national economic welfare: whereas depressed industrial or services sector products are raised arti- export prices result in too few resources being ficially by restrictions on their importation or by devoted to the production of those exportables, an other price-support measures, resources are drawn 12 Economywide Dimensions of Trade Policy and Reform to those import-competing sectors at the expense of toral resource use remain and could worsen when other industries in the primary sectors, including the average level of agricultural assistance is raised. export industries (Clements and Sjaastad 1984). Second, the economy is invariably made up of more Historically, industrial tariffs have been a major than just those two sectors, so similar levels of assis- source of indirect discrimination against agricul- tance would have to be provided to fishing, mining, ture, but a wide range of other distortionary mea- and other sectors to ensure an overall improvement sures is found in service industries as well. in the efficiency of national resource use. Third, sup- The importance of this cause of inefficient pose the farm assistance were to be provided via, say, resource allocation cannot be stressed enough, as it input subsidies for fertilizer and water, as in fact has crucial implications for reform. Two examples often happens, even in poor countries. It turns out will illustrate the point. Taking, again, the average that agricultural support via input policies would be country in the Krueger, Schiff, and Valdés study, less efficient and possibly even counterproductive suppose that the agricultural sector is a net exporter because it would encourage the use of only a subset (which means the country is a net importer of non- of inputs rather than all farm inputs (Warr 1978). farm products) and that the food-importing sub- Most damning of all, manufacturers would perceive sector is almost as large as the agricultural their situation as deteriorating if support for pri- export­focused subsector. Within agriculture, the mary production were increased and, if no change in restrictions that reduce the domestic price of farm the political-economy forces had been at work, they exportables by 11 percent and raise the price of food would presumably demand a return to the status importables by 20 percent would boost the overall quo ante, perhaps through another hike in industri- average price of farm products, but by less than 10 al tariffs. Clearly, tariff compensation to farmers is a percent. Taking a single-sector perspective as in the far riskier reform strategy for improving the use of previous section might lead one to believe that national resources than the first-best strategy of eliminating those agricultural policies and thereby reducing industrial tariffs. reducing farm prices on average would be welfare improving. The Additional Indirect Effect of Distorting As it turns out, however, such a conclusion would Exchange Rates follow only if there were no distortions in the rest of the economy. If the manufacturers in this economy The Krueger, Schiff, and Valdés (1988) study also were to enjoy an average nominal rate of protection examined the extent to which unsustainable current from import competition of, say, 25 percent (for account deficits, overvalued official exchange rates, example, as a result of a uniform 25 percent tariff), and the like artificially inflate the value of a nation's then, prior to reform and notwithstanding the posi- currency from the viewpoint of farmers. Such poli- tive direct assistance to farmers, there would already cies encourage the production (and discourage the be too many resources in industrial relative to agri- domestic consumption) of nontradables relative to cultural pursuits. In that case reducing support for tradables and thereby represent another source of farming would be likely to exacerbate the inefficient inefficiency in national resource use and another resource allocation rather than improve it. To disincentive to farm. ensure a welfare-improving policy reform in this Empirically, for the 18 countries studied by case, it would be necessary to first lower the degree Krueger, Schiff, and Valdés, these macroeconomic of assistance to manufacturers and then, when the policies proved less of a disincentive to agricultural level of assistance to industry equaled that to farm- producers than did industrial protectionism. ers, phase down both simultaneously.1 Nonetheless, they added to farmers' difficulties. If it is too difficult politically to lower tariff protec- Together, the indirect negative impact of industrial tion to manufacturers, might a similar national wel- and macroeconomic policies on farmers' incentives fare improvement be achievable by raising the level was two-and-a-half times as large as the direct neg- of assistance to agriculture? In theory, maybe, but in ative effects of agricultural export policies in the practice such a tariff-compensation strategy would decade 1974­84, equivalent to depressing the price be unwise, on a number of grounds. First, if rates of of farm exportables by 38 percent, compared with assistance to different industries within each of the just 11 percent by direct measures. This indirect dis- two sectors are not equal, inefficiencies in intrasec- incentive also applied to import-competing farm- 13 T R A D E P O L I C Y R E F O R M I N C O N T E X T ers. In this sample of developing countries, the latter used intensively in the industry in which it has a enjoyed direct nominal protection of 20 percent in technological advantage. that decade, so even the most favored farmers in For nationalistic and cultural reasons, permanent those countries were being disadvantaged by the immigration of labor has not been made easy in dominance of the adverse indirect effects of non- recent decades, but numerous countries have toler- agricultural policies on agricultural incentives. ated temporary movements of labor, bringing mutu- What would be the economywide implications of al gains to the countries involved. Much more reducing import tariffs in the above case? Reducing important in the past two decades, however, has the food import restrictions alone would probably been the growth in movement of capital across boost production of exported farm goods, which national borders. Foreign direct investment can would improve resource allocation within the farm bring with it not just financial capital but also mana- sector. But it would also free mobile resources that gerial and marketing skills, technological knowl- could then move to nonfarm activities, which, on edge, and intellectual property--forms of capital average, are more protected than farming. Hence, that foreign firms might not be willing to see export- whether the overall efficiency of national resource ed if they were unable to retain control over them. use would rise or fall is an empirical question if only Developing countries seeking to exploit fully their a subset of import restrictions and exchange rate comparative advantages therefore need to relax their distortions is to be removed. Only if the most pro- restrictions on foreign investment inflows. By the tected industries were to be liberalized first would same logic they also need to allow foreign invest- resources necessarily move to less protected indus- ment outflows so that domestic owners of capital tries and sectors and thereby guarantee an improve- also can earn the highest rewards possible. ment in the efficiency of the use of these resources in producing tradables. Even then, there is the possibil- The Dynamic Consequences of ity that those mobile resources would move into the Trade Reform production of more nontradables if the currency remained overvalued. This is the reason for the value Freeing up trade in goods, services, currencies, and of comprehensive reform that simultaneously frees capital not only improves the efficiency of national trade in goods, services, and currencies. resource use and consumer welfare at a point in time but also contributes to economic growth. The mech- anisms by which openness contributes to growth are What about Markets for Factors of gradually becoming better understood, thanks to the Production? pioneering work of such theorists as Grossman and An economywide perspective on trade reform would Helpman (1991) and Rivera-Batiz and Romer be incomplete unless it also extended to restrictions (1991). In a helpful survey of the subsequent litera- on factor flows. Theorists in the 1950s pointed to the ture, Taylor (1999) identifies several channels possibility that trade in goods could be a complete through which openness to trade can affect an econ- substitute for trade in productive factors in terms of omy's growth rate. They include the scale of the mar- both the volume of product trade and the welfare ket when knowledge is embodied in the products gains from trade (Mundell 1957). That theoretical traded, the effect of knowledge spillovers, and the possibility holds only under fairly restrictive condi- degree to which redundant creation of knowledge is tions, however. More recently, attention has been avoided through openness. More important from a drawn to the possibility that trade in some produc- policymaker's viewpoint, the available empirical evi- tive factors complements rather than substitutes for dence strongly supports the view that open trade in products (Markusen 1983). That can hap- economies grow faster (Edwards 1993; USITC 1997). pen when other productive factors are sector-specif- ic and goods trade is thus insufficient to equalize What if Trade Reform Harms the factor prices across countries. In that case trade in Environment? internationally mobile factors can generate further welfare gains from trade. It can also happen when Ideally, in adopting an economywide perspective, all there are differences in technologies across coun- significant influences of trade reform on human tries; then each country should import the factor welfare should be considered. That could include a 14 Economywide Dimensions of Trade Policy and Reform whole range of so-called noneconomic policy the Krueger, Schiff, and Valdés (1988) study shows, objectives, as well as standard economic effects such the indirect effect of nonagricultural and macroeco- as those on the natural environment, poverty, nomic policies on farmers' welfare can be several unemployment, food security, and distribution of times as large as the direct influence on incentives of income and wealth across regions and households. agricultural polices affecting export-oriented farm- Space is not available to discuss each of these here, ers. This is also true within a sector, and even more but excellent treatments are available in such books so to the extent that productive factors are more as Corden (1997). The main conclusion to be drawn readily substitutable within than between sectors. from that literature is that whatever the domestic In lobbying for trade reform, care is needed to policy objectives one has in mind, trade policy ensure that trade liberalization is not accompanied instruments are virtually never first-best ways of or followed by "re-instrumentation" of support. achieving those objectives. There are numerous ways to support producers This conclusion does not mean that trade reform other than through trade policy, and many of them can be undertaken without regard for society's are even more inefficient than trade measures. It other objectives. Welfare improvement via trade lib- would be counterproductive to lobby for the eralization cannot be guaranteed if optimal domes- removal of a trade restriction if it led to such an tic policies are not in place. There is no better inferior replacement. illustration of this than with respect to the natural A major aspect of exporters' lobbying activities environment. Reducing restrictions on exports of often involves encouraging the removal of impedi- logs, for example, in the absence of any other forest ments to market access abroad. Here again, an econ- resource policies is likely to lead to excessive defor- omywide perspective is needed (as is vigilance in estation. Another example is the reduction in Mon- preventing re-instrumentation). Consider, for golia's export tax on cashmere, which encouraged example, the interests of developing countries with the excessive grazing of common pastures. In these a strong comparative advantage in agriculture. They and in many other such cases overexploitation was would be likely to benefit directly from reduction in the result of property rights being poorly defined or agricultural protectionism in advanced industrial poorly policed. Clearly, better resource and environ- countries, but they could also benefit, albeit indi- mental policies are required before optimal social rectly, from a reduction in manufacturing protec- welfare can be achieved. tion in those same countries. The most obvious Note, however, that those resource and environ- example is a reduction in the very high barriers to mental policies are warranted, regardless of the imports of textiles, clothing, and footwear. Greater degree of openness of the economy. All that trade global production and trade in those products reform requires in addition is that the levels of envi- would result from reduced protection, with the out- ronmental policy intervention be adjusted when put expansion concentrated in newly industrializ- trade is liberalized to ensure that any additional ing countries. A direct consequence would be an environmental damage which accompanies opening expanded demand for cotton, wool, and leather up is matched in value terms with the marginal inputs--but that is only part of the impact on gains from trade expansion. Of course, trade reform agrarian developing countries. Probably more need not cause additional environmental damage; important is that such reform would speed the at least equally possible is the prospect that the industrialization of the more densely populated changes in production and consumption that developing countries, which would attract resources accompany trade liberalization will actually reduce away from their farm sectors. An indirect conse- pollution or resource depletion (Anderson 1997). quence, therefore, would be increased demand for food imports by those newly industrializing coun- tries. This suggests there is scope for agrarian and Implications for Reform-Minded Producers newly industrializing developing countries to act and Trade Policymakers collectively in pushing hard for greater market A clear implication of this economywide perspec- access for farm and textile products in advanced tive for producers seeking to influence government economies. In return, developing countries would policy is that their focus should not be confined to be expected to provide more access to their markets measures directly affecting their own industries. As for the goods and services exported by advanced 15 T R A D E P O L I C Y R E F O R M I N C O N T E X T economies--another dimension of the intersectoral connectedness of the global economy.2 Notes 1 In practice, a greater degree of refinement is possible, taking into account not only input price distortions (to get a measure of effective assistance to value added rather than just the nom- inal boost to the price of output) but also the degree of inter- sectoral substitutability or complementarity in production and consumption. See Corden (1971); Vousden (1990, ch. 9). 2 For recent empirical studies on the economywide effects of global trade reform and their implications for farm production and trade, see, for example, Hertel and others (forthcoming); Anderson, Hoekman, and Strutt (2001). 16 3 H O W A R D J . S H AT Z D A V I D G . TA R R Exchange Rate Although as a group, develop- ing countries progressively liber- alized their trade regimes during Overvaluation the 1980s and 1990s, some gov- ernments continue to take and Trade actions to defend their exchange rates that are counter to their Protection long-run trade liberalization efforts. One classic pattern is to attempt to defend an overvalued exchange rate through protec- tionist trade policies.3 Experi- ence shows that protection in A defense of an overvalued lthough both fixed and flexible exchange rate will significantly retard the country's exchange rate systems (and their medium-to-long-run growth prospects. In fact, an variants) have their advantages and disadvantages, overvalued exchange rate is often the root cause of more than half the countries in the world maintain protection, and the country will be unable to return fixed or managed exchange rates.1 While we do not to the more liberal trade policies that allow growth discuss the relative merits of these exchange rate without exchange rate adjustment. systems in this chapter, we note that as a practical Moreover, a devaluation of the nominal matter, exchange rate management in many coun- exchange rate appears to be a necessary condition tries in the world has resulted in overvaluation of for achieving a large depreciation of the real the real exchange rate, in some cases leading to gross exchange rate, as virtually all real devaluations distortions.2 (For further discussion of the links (above 25­35 percent) have been accompanied by between trade and macroeconomic management, nominal devaluations (Ghei and Hinkle 1999). see the CD-ROM, "Applied Trade Policy," that Sustained efforts to use downward adjustment of accompanies this Handbook.) wages and prices as a means of restoring a compet- Since governments are frequently confronted itive real exchange rate have frequently led to with the problems of external shocks and external severe recessions or depressions. trade deficits in the context of a fixed exchange rate Worldwide experience has shown that defending regime, a concise survey of worldwide experience the exchange rate has no medium-run benefits, with the effects of overvalued exchange rates in since falling reserves will eventually force devalua- terms understandable to policymakers should be tion. It is better that the devaluation be accom- useful. This chapter presents theory, cross-country plished without further debilitating losses in econometric evidence, and important case studies reserves and lost productivity due to import con- of the effects of overvalued exchange rates. trols. Experience with devaluations shows that after 17 T R A D E P O L I C Y R E F O R M I N C O N T E X T the devaluation, the exchange rate will reach a new result, less foreign exchange is available for need- equilibrium and that the equilibrium is strongly ed imports. influenced by the policies of the central bank and · Foreign exchange may be rationed and allocated the government. inefficiently by the government. · Efforts to defend an overvalued exchange rate through very tight monetary policy can plunge The Problems of an Overvalued the country into severe recession. Exchange Rate Countries that attempt to maintain overvalued The Need to Restore Internal Balance exchange rates significantly impede their growth in the medium to long term. Theory, cross-country When a country experiences a deficit in its trade bal- statistical studies, and case histories all reinforce the ance, it is not in "external" balance. It follows from a basic findings that exchange rate overvaluation can national income accounting identity that a trade reduce economic efficiency, misallocate resources, deficit means the country is spending more than its increase capital flight, and, most perniciously, lead income. That is, the trade deficit allows the country to exchange and trade controls. to consume or spend beyond its income (or beyond the value of what it is producing). When a country's expenditure does not equal its income, it is not in The Theory "internal" balance. These external and internal Theory suggests that there are many channels imbalances can severely impede country economic through which an overvalued exchange rate hurts performance, and it is these imbalances that coun- the economy and growth: tries suffering from external shocks often face. Although a nominal devaluation is designed to · It discriminates against exports. Since a signifi- correct the problem of external balance, it will also cant portion of the costs of production is paid in be important to ensure internal balance; otherwise, domestic currency, the overvalued exchange rate the trade deficit may not be corrected by the nomi- results in a reduction of exporters' incentives and nal devaluation. For many developing countries the ability to compete in foreign markets. This chokes trade deficit reflects the government's fiscal deficit, foreign exchange receipts and damages a coun- which is often financed by monetary expansion. try's ability to purchase the imports needed for The monetary expansion in turn leads to inflation. economic activity. In this environment the impact on the real exchange · Import-competing industries are faced with rate of a nominal devaluation is likely to be eroded increased pressure from foreign companies, by inflation, since high inflation tends to appreciate resulting in calls for protection against imports the real exchange rate, making elimination of the from industrial and agricultural lobbies. The trade deficit problematical. political pressures for protection eventually prove In general, monetary or fiscal policies will have to to be overwhelming, and governments yield to be combined with exchange rate policies to achieve lobbying and impose higher tariffs on imports. both internal and external balance simultaneously. This closes the economy to international compe- This is a special case of a more general principle of tition and reduces access to needed imported economics: multiple policy targets typically require inputs and technology. As a result, growth falls. multiple policy instruments. In this chapter, howev- Devaluation serves the dual purpose of uniformly er, we focus on the experience of countries that have protecting import-competing industries and limited the use of exchange rate adjustment as an increasing incentives for exporters. economic policy instrument. · Productivity advances are less rapid because the export sectors and the import-competing sectors, Problems with "Automatic" Adjustment where productivity advances are often fastest, are Mechanisms disadvantaged by an overvalued exchange rate (Cottani, Cavallo, and Khan 1990). Unless the central bank takes offsetting action, a · Overvaluation induces capital flight among trade deficit will result in a decline in the domestic domestic citizens anticipating a devaluation. As a money supply. Thus, one response to an overvalued 18 Exchange Rate Overvaluation and Trade Protection exchange rate is to hold the nominal exchange rate with similar results, Ghura and Grennes (1993) ana- fixed and assume that domestic prices and wages lyzed the relationship between the real exchange will fall and so help bring tradable goods prices back rate and macroeconomic performance in 33 Sub- to internationally competitive levels. This is the Saharan African countries between 1972 and 1987. "specie flow mechanism" described by David Hume They found that misalignment, or overvaluation, in the 18th century. The problem with this strategy was associated with lower levels of growth of real is that in most modern economies, prices and wages GDP per capita, lower levels of exports and imports, tend to be sufficiently inflexible downward that sus- lower levels of investment, and lower levels of sav- tained and substantial periods of unemployment ings, even when they corrected for other causes. must be endured if the strategy is to have a chance of succeeding. Most countries are unwilling to Case Studies of the Effects of Overvaluation endure these high costs. (See Sachs and Larraín 1999 for a further discussion.) For example, as is The economic histories of developing countries that described below, Chile endured a deep recession in followed a classic import-substituting industrializa- 1982­83 before it devalued in 1984, and the fran- tion strategy after World War II provide good illus- cophone African countries in the CFA zone experi- trations of the negative effects of an overvalued enced disastrous consequences from overvaluation; exchange rate combined with trade controls. Latin in some, the economic contractions were compara- America, more than any other region, followed this ble to the Great Depression in the United States. strategy, but it was not alone. We select illustrative The CFA zone experience also casts doubt on the episodes from Argentina, Chile, Uruguay, Turkey, claim that countries should avoid devaluation in and the CFA zone of Africa. order to retain international investors. The zone certainly had stable prices and exchange rates, but Argentina, Chile, and Uruguay its failure to solve the problems brought on by the overvalued real exchange rate substantially Argentina, Chile, and Uruguay all followed import- decreased its attractiveness to foreign investors. substituting industrialization policies that led to a Capital flight increased in anticipation of an even- bias against exports, extremely uneven rates of trade tual devaluation (Clément and others 1996). protection across sectors, and controlled financial systems. They also experienced recurrent balance of payments crises and slow growth (Corbo, de Melo, Cross-Country Economic Performance and Tybout 1986). By the early 1970s, all three had Cottani, Cavallo, and Khan (1990) investigated the accelerating inflation, bottlenecks in production, effects of real exchange rate misalignment and vari- slow export growth, and balance of payments diffi- ability on the economic performance of 24 develop- culties (Corbo and de Melo 1987). In response, they ing countries between 1960 and 1983. They found went through two phases of stabilization and that exchange rate misalignment was strongly relat- reform, one in the mid-1970s and the other during ed to low growth of per capita GDP. Misalignment 1979­82. The second phase is most relevant for was also related to low productivity (capital did not evaluating the effects of an overvalued exchange go to the companies or sectors that could make the rate and import controls on economic perform- best use of it), slow export growth, and slow agricul- ance. tural growth. In the second phase all three countries used a A study of growth in 12 countries between 1965 nominal exchange rate anchor to halt inflation. The and 1985 (Edwards 1989) reinforced these exchange rate appreciated, and when it became findings.4 The greater the misalignment, the lower apparent that the nominal rate could not be sus- the growth during the period. Furthermore, tained, capital flight resulted. In Uruguay and exchange controls and trade impediments, proxied Argentina, where there were no capital controls, by the black-market exchange rate premium, were major capital outflows occurred. In Chile, where negatively related to growth. there were capital controls, people engaged in capi- There is strong evidence that overvaluation of real tal flight by buying imported consumer durables. exchange rates was greatly implicated in Africa's This capital flight occurred in all three countries poor economic performance. Among other studies well before the onset of the debt crisis in 1982. 19 T R A D E P O L I C Y R E F O R M I N C O N T E X T Other problems resulted. Profitability fell in the and a staged lowering of uniform tariffs, from 35 tradable goods sectors. In Argentina, which percent in 1984 to 11 percent by 1991. An important remained quite restrictive to imports throughout, feature of the new nominal exchange rate system the gross margins of exporting businesses were hurt was a crawling band, which policymakers intended much more than those of import-competing busi- to use to maintain the international competitiveness nesses. In Uruguay the rate of growth of nontradi- of Chilean exports (Dornbusch and Edwards 1994). tional exports fell sharply between 1979 and 1981. In fact, although they used the nominal rate as the In Chile the leading growth sectors during the peri- policy variable, they focused on the real exchange od were construction, internal trade, and financial rate, adjusting the nominal rate for the differential services--all nontradables--even though reforms between domestic and foreign inflation. Taking an during the 1975­79 period had reduced the bias index of 100 as the value of the real rate in 1977, the against exports significantly by June 1979. real exchange rate appreciated to 84.5 in 1981, fell to 118.2 in 1984, and then, following the introduction of the new policy, depreciated to 145.2 in 1985. It Chile: The Aftermath continued depreciating, to 180.1 in 1990 (Corbo Chile is now well known for its economic success. and Fischer 1994). In 1998 the Chilean legislature Since 1984, it has had an average annual rate of approved a further lowering of the uniform tariff to growth of real GDP of more than 7 percent. Its poli- 6 percent, in stages, and in late 1999 Chile aban- cies following the 1982­83 crises are instructive. doned the exchange rate band system for a float. Chile experienced high rates of growth in the late The improved incentives to exporters from the 1970s, following a deep contraction in 1974­75. The reduction in the import tariff and the devaluation growth surge was the result of a number of deregula- led to an expansion of nontraditional exports (by 10 tion and reform measures, including institution of a percent a year from 1985 to 1995) and to efficient uniform 10 percent tariff on all goods except auto- import substitution. Macroeconomic stabilization, mobiles. Nevertheless, inflation persisted, hurting tax reform, and cuts in government spending com- the reforms, and in 1979 Chile fought back by set- bined to promote savings and investment. And pri- ting a fixed exchange rate as a nominal anchor. Com- vatization of state-owned firms, rehabilitation of bined with other policies, this at first led to large the financial sector through recapitalization, and external borrowings, most of which were at variable strengthened bank regulation spurred private busi- interest rates. In the early 1980s the external financ- ness activity. ing dried up as confidence in the sustainability of the exchange rate ebbed. Making matters worse, Chile Turkey experienced a deterioration in the terms of trade. Then foreign interest rates rose, further hurting the Three episodes from the post­World War II history of Chilean financial and business sectors. In 1982­83 Turkey, recounted in Krueger (1995), provide another Chile experienced its worst depression since the illustration of the problems created by an overvalued 1930s, as real GDP fell 15 percent. exchange rate combined with import restrictions. During and immediately after the recession, Chile Like the Latin American countries, Turkey followed experimented with a number of policies, including an import-substituting industrialization growth strat- an increase in tariff rates to switch domestic spend- egy. Starting in 1953, export growth ceased for a num- ing to domestic products. In June 1982 the govern- ber of reasons, and inflation accelerated. The ment abandoned the fixed exchange rate, combination of inflation and a fixed nominal eliminated compulsory wage indexation, and initi- exchange rate meant a strengthened real exchange ated a series of nominal devaluations. For a short rate and a bias against exports. Foreign exchange time, Chile allowed the exchange rate to float became scarce, and the country started import licens- (Corbo and Fischer 1994). Then, however, it fol- ing in 1954. By 1957, export earnings were falling, and lowed an erratic policy, implementing five different imports were severely restricted, damaging domestic exchange rate regimes (Labán and Larraín 1995). economic activity. In 1958, Turkey could not finance In 1985 the government embarked on the strategy imports, and it appeared that the country would not it maintains to this day: an export-oriented struc- even be able to obtain gasoline for trucks to move that tural adjustment. This included steady devaluations year's harvest to ports. 20 Exchange Rate Overvaluation and Trade Protection In response, Turkey adopted an IMF stabilization countries, average growth was 2.8 percent. In addi- plan that featured devaluation, import liberaliza- tion, CFA countries achieved an annual average tion, and fiscal and monetary restraint. Real GDP, export growth rate of 7 percent. which had been declining, started growing immedi- In the mid-1980s the economic performance of the ately in response to the availability of imports. Infla- CFA zone countries began to deteriorate, for two rea- tion fell, and export earnings began to rise again. In sons: the appreciation of the French franc, and a the 1960s Turkey was among the most rapidly grow- series of primary-commodity price shocks (Azam ing developing countries. and Devarajan 1997). Devarajan (1997), in a study of In the late 1960s, Turkey's exchange rate again 12 CFA countries, found an average overvaluation of became overvalued as a result of moderate inflation 31 percent in 1993 on the eve of the devaluation, with throughout the decade (5 to 10 percent annually) Cameroon's real exchange rate the most overvalued and a fixed nominal exchange rate. The high (78 percent) and Chad's real rate the only underval- demand for imports, together with the bias against ued one.5 Eight of the 12 had overvaluations of 20 exports, caused foreign exchange to become scarce. percent or more. Making matters worse, other The resulting problems in obtaining imports led to African countries were devaluing during the 1980s, a slowdown in both production and real invest- contributing to the overvaluation of the real rates of ment. The country responded in 1970 with a nomi- the CFA zone countries compared with those of their nal devaluation, and the result was extremely rapid export competitors. Elbadawi and Majd (1996) export growth. Turkey then experienced rapid eco- showed statistically that CFA membership and, by nomic growth through 1975. implication, the high level of the real exchange rate The third episode occurred in the late 1970s. were partly to blame for the poor economic perform- Large fiscal deficits, failure to adjust the internal ance of the CFA countries in the late 1980s. price of oil following the 1973 oil shock, and an Because of the overvaluations and mounting overvalued exchange rate, made worse by extremely structural problems, such as rigidly high wages, eco- high inflation, spurred this new crisis. Once again, nomic performance started to deteriorate. The zone the country ended up with severely constrained saw no economic growth between 1986 and 1994, a imports, falling real output, and falling income. period when other Sub-Saharan African countries were growing at 2.5 percent a year (Clément 1994). In fact, some of the countries suffered an output The CFA Zone Countries contraction comparable to that of the Great The currency of the countries of the CFA zone of Depression in the United States (Table 3.1). Africa was fixed precisely to the French franc and is A number of other ill effects stemmed from the now fixed to the euro. Until the second half of the period of overvaluation in the CFA zone. Several 1980s, these countries experienced stable and posi- countries suffered large increases in poverty tive economic performance (Elbadawi and Majd (Devarajan and Hinkle 1994). For example, in Côte 1996). For example, their average annual real GDP d'Ivoire the incidence of poverty doubled between growth rate between 1973 and 1981 was 5.7 percent, 1985 and 1992, from 30 to 60 percent. Devarajan whereas for 18 non-CFA Sub-Saharan African and Hinkle also note that banking systems in a Table 3.1 Comparing "Great Depressions": Cameroon, Côte d'Ivoire, and the United States (percentage decline in per capita GDP) Measure of output decline Cameroon Côte d'Ivoire United States Purchasing power paritya 31.4 29.1 -- Purchasing power parity with terms of trade adjustmenta 38.5 34.5 -- Market pricesb 41.5 18.8 30.9 -- Not available. a. Authors' calculations for 1986­92 from the Penn World Table Mark 5.6, described in Summers and Heston (1991) and available on the Website . Data for after 1992 are unavailable. b. Authors' calculations from the peak to the trough of the depression period (1986­94 for Cameroon and Côte d'Ivoire; 1929­33 for the United States). Data are from World Bank (1999) and U.S. Bureau of the Census (1975). 21 T R A D E P O L I C Y R E F O R M I N C O N T E X T number of countries became insolvent or illiquid as countries often apply high or prohibitive trade pro- a result of private sector inability to repay debts, tection on selected products or vis-à-vis selected government and public enterprise arrears, and capi- countries. Even given a limited objective of reducing tal flight. Export earnings collapsed in response to the demand for foreign exchange, an increase in the adverse terms of trade shocks and the overvalu- imports will occur through informal channels, ation of the real exchange rate. The contractionary depending on how porous the borders are. With macroeconomic policies adopted by most CFA diverse protection, while some sectors will be pro- countries reduced import levels, and inflation tected, the burden of the costs of adjustment to the remained low, but budgetary and external deficits overvalued exchange rate will be borne by the rose. The fixed nominal rate and various policy- unprotected sectors, by those sectors that are more induced rigidities in domestic prices, particularly in susceptible to informal or illegal imports, and by the wages and nontradable goods prices, meant that export sectors. Countries typically eventually deval- adjustment had to come through reduced employ- ue, but it is better that the devaluation be accom- ment, output, and growth.6 plished without debilitating losses in reserves and Constrained by their fixed exchange rates, at least lost productivity due to import controls. two of the CFA zone countries tried to carry out As the experience cited here shows, governments "mock devaluations," with subsidies to exports and must avoid policies that contribute to an overvalued increases in import tariff rates. In Côte d'Ivoire the exchange rate. Although we do not advocate any scheme collapsed after a short trial because of particular type of exchange rate regime in this chap- administrative difficulties, inability to give the ter, we emphasize that whatever regime is employed, export subsidy plan a sufficient budget, and lack of policies should be aimed at maintaining a competi- support by the government. In Senegal administra- tive real exchange rate. tion of the plan proved difficult, and the scheme encouraged overinvoicing by exporters and smug- Notes gling and underinvoicing by importers. The plan also proved costly to the budget, as tariffs were The authors thank Arup Banerji, Julian Berengaut, Dominique already high and the increases could not generate Desruelle, Lawrence Hinkle, Fred King, Kiyoshi Kodera, Albert Mar- tinez, Will Martin, Francis Ng, Paul Ross, Maurice Schiff, and semi- much more revenue. nar participants at the World Bank for helpful comments on earlier Finally, on January 12, 1994, the countries held a drafts of this paper. "maxi-devaluation," changing the rate to the French franc from 50:1 to 100:1.7 The CFA devaluation had 1 As of the beginning of 1999, the IMF (1999: app. I) reported arrangements for 185 countries. The exchange rate regimes excellent intermediate-term effects on growth. For can be categorized as pegged (84 countries), floating (75 the 12 CFA countries in Devarajan's sample, World countries), and limited flexibility (26 countries). Of the 84 Bank data showed that real GDP growth between countries with pegged exchange rates, 37 have no separate 1990 and 1993 averaged almost minus 0.3 percent legal tender, 8 use a currency board arrangement, 24 peg to annually, weighted by GDP (World Bank 1999). another currency, and 15 peg to a composite of currencies. Of those using a floating rate, 27 maintain a managed float and From 1994 to 1997, however, growth in the sample 48 an independent float. countries averaged 5.1 percent annually, according to the same data source.8 Cameroon, the largest 2 See Global Currency Report (1999). Of 160 countries listed, 38 had black market premiums of more than 10 percent at the country in the CFA zone, grew at an annual rate of end of 1998. Of the 38, 19 had premiums of more than 25 minus 3.4 percent in the first period but by 4.5 per- percent, 13 had premiums of more than 50 percent, and 10 cent in the second period (World Bank 1999). (Afghanistan, Algeria, Angola, Iraq, the Democratic Republic of Devarajan (1997) found that a year after the devalu- Korea, Liberia, Libya, Myanmar, São Tomé and Principe, and ation, the average undervaluation for the group was Somalia) had premiums of more than 100 percent. The black- 2 percent, but with significant variance. market exchange rate is likely to be overly depreciated in rela- tion to an equilibrium long-run real exchange rate, since an actual real depreciation would increase the supply of and Conclusion reduce the demand for foreign exchange. See Ghei and Kamin (1999) for a detailed explanation and econometric evidence. Worldwide experience has shown that defending 3 Ghei and Pritchett (1999) call this the "import compression syn- the exchange rate has no medium-run benefits. In a drome." Since devaluations (which reduce imports) are often classic pattern, once reserves are drawn down, accompanied by reductions of trade barriers (which increase 22 Exchange Rate Overvaluation and Trade Protection imports), econometric evidence on the import-reducing impact of devaluation has been weak. Ghei and Pritchett argue that devaluations significantly reduce imports if there is proper adjustment for the simultaneous reduction of trade protection. 4 The 12 countries studied were Brazil, Colombia, El Salvador, Greece, India, Israel, Malaysia, the Philippines, South Africa, Sri Lanka, Thailand, and Yugoslavia. 5 The countries in the study were Benin, Burkina Faso, Cameroon, Central African Republic, Chad, Congo, Côte d'Ivoire, Gabon, Mali, Niger, Senegal, and Togo. 6 For example, both Senegal and Côte d'Ivoire had rigid labor laws that kept wages high throughout the predevaluation peri- od (Foroutan,1997). Clément (1994) noted that throughout the CFA zone, rising wage costs contributed to substantial drops in public enterprise profitability, expanding the public sector financing requirement. Extensive controls over both producer prices and retail prices, particularly nontradable goods prices, added to the price rigidities in many countries. 7 The Western and Central African monetary unions (comprising Benin, Burkina Faso, Cameroon, Central African Republic, Chad, Congo, Côte d'Ivoire, Equatorial Guinea, Gabon, Mali, Niger, Senegal, and Togo) changed their rates from 50 CFA francs: 1 French franc to 100 CFA francs: 1 French franc. At the same time, Comoros changed its rate from 50 Comoros francs: 1 French franc to 75 Comoros francs: 1 French franc. 8 The unweighted averages are 0.1 percent for 1990­93 and 4.7 percent for 1994­95. 23 4 L I A M E B R I L L J A N E T S T O T S K Y R E I N T G R O P P Fiscal Dimensions the reforms undertaken and the initial circumstances of the par- ticular country. of Trade Table 4.1 shows a taxonomy of trade reform measures and their Liberalization expected effect on the govern- ment's fiscal position. As noted above, the impact can be positive, negative,or neutral,depending on the nature of the restrictions and the characteristics of the particu- lar country. For the most part, however, the reforms will general- D ly enhance revenue collections or espite substantial trade liberaliza- will have an ambiguous effect.The sequencing of trade tion over the past decade, many liberalization in programs supported by the IMF and developing countries continue to have restrictive the World Bank normally gives the highest priority to trade regimes characterized by high tariffs and per- the removal of nontariff barriers, which tend to be the vasive nontariff barriers (NTBs). Given the now most distortionary, followed by measures to rational- well-established nexus between open trade regimes ize the tariff structure. Below, we take a brief look at and improved export and growth performance, fur- some of these measures and their implications for the ther trade liberalization to promote sustainable country's fiscal position. growth and integration into the global trade system Nontariff barriers encompass a whole range of remains essential. Because trade liberalization has practices, including quotas, bans, export and import implications for fiscal revenues, and because many licensing, and state trading monopolies. In addition low-income countries continue to rely to a signifi- to the economic efficiency arguments, the initial focus cant extent on trade taxes as a source of revenue, on removing NTBs has the advantage of also increas- attention must center on the fiscal dimension of ing fiscal revenues. Quotas and bans provide no rev- trade reform in designing a strategy for trade liberal- enue to the budget and offer ample opportunities for ization. rent-seeking behavior or for smuggling. Accordingly, the conversion of quotas into equivalent tariffs (nor- mally accompanied by a scheduled timetable for fur- The Revenue Impact of Trade Liberalization ther reduction in tariff rates) or the removal of bans Trade liberalization has often been delayed by con- will, other things being the same, have an immediate cerns that it will have a negative impact on fiscal positive effect on fiscal revenues as rents are trans- revenues and contribute to macroeconomic insta- ferred to the government in the form of trade tax rev- bility. In fact, the impact of trade liberalization on enues. For these reasons, removal of NTBs should be revenues is generally ambiguous and depends on addressed early in the reform process. 24 Fiscal Dimensions of Trade Liberalization Table 4.1 Revenue Impact of Trade Liberalization Trade reform Expected revenue impact Replace nontariff barriers with tariffs Positive Eliminate tariff exemptions Positive Eliminate trade-related subsidies Positive Reduce tariff dispersion Ambiguous/positive Eliminate state trading monopolies Ambiguous/positive Reduce high average tariffs Ambiguous Lower maximum tariff Ambiguous Reduce moderate or low average tariffs Negative Eliminate export taxes Ambiguous/negative Sources: IMF and World Bank staff estimates. Eliminating tariff exemptions (excluding export structure, or one with few tariff bands, will mini- duty drawback schemes) and trade-related subsi- mize tax evasion and ease the task of customs dies should have a direct positive effect on the gov- administrators by reducing opportunities for mis- ernment's fiscal position. Moreover, not only does classification and valuation mistakes. the existence of tariff exemptions, especially discre- If a country has already implemented substantial tionary exemptions, provide an incentive for trade reforms, at some point further reductions in importers to seek additional exemptions, but their rates (other things being the same) will result in lower proliferation also increases the incentive for classify- revenue collection, at least in the short run. Given the ing taxable products as exempt, which has a nega- longer-term growth benefits of trade reform, howev- tive impact on revenues. Thus, in addition to their er, the appropriate response would be to offset any direct positive fiscal effect, eliminating discre- potential revenue loss by using other, less distorting, tionary exemptions and other complexities can broader-based taxes (for example, a value-added tax), contribute to improved governance. applied equally to both domestically and foreign-pro- The fiscal impact of reducing tariffs depends on duced goods. The distortion to the economy from their initial levels and coverage and on the extent to taxing both imports and domestic substitutes at which they are reduced. In principle, given an equivalent tax rates is generally less than that of taxing unchanged level of imports, lowering tariffs will imports alone, and taxing both yields larger revenues. reduce trade taxes. Since, however, the lower rates For developing countries in which trade taxes are are also likely to increase the demand for imports, an important source of revenue, a further reduction the net impact on revenues will depend on the price in the average tariff could be perceived to have a elasticity of import demand. The higher the elastic- negative effect on revenues and to inhibit the pace ity is, the more likely it is that a reduction in tariffs of further reforms. In these cases mobilizing alter- will have a net positive impact on fiscal revenues. native sources of revenue and diversifying tax In countries with prohibitively high tariffs, there sources away from trade taxes is critical but is likely is a strong incentive for tax evasion, either through to be a long-term process requiring an early start misclassification or by smuggling and avoiding pay- toward a broader-based tax. Since such a process ing the tax altogether. Therefore, lowering such tar- takes time to prepare and implement, technical iffs is likely to generate higher revenues because it assistance from the IMF should be sought at an reduces the cost of compliance and increases the early stage of the liberalization process and should volume of recorded trade subject to taxation as also be used to support the trade reform measures smuggling activities subside. More generally, a by improving customs administration. But even in reduction in tariff dispersion will tend to bolster countries that are highly dependent on trade taxes, revenues by reducing incentives for tax evasion. there is no reason to delay implementing trade The reform of the trade regime in the direction of reform measures that have a positive or neutral a more uniform tariff structure could increase fiscal impact on revenues. In fact, heavy reliance on such revenues as a result of increased transparency and taxes strengthens the case for proceeding more rap- simplification of tax administration. A uniform idly with the revenue-increasing elements of trade 25 T R A D E P O L I C Y R E F O R M I N C O N T E X T reform, especially the tariffication of NTBs and the dence suggests that liberalization of quantitative curtailment of exemptions. restrictions tended to bolster revenues and that tar- Significant progress has been made in reducing iff reforms did not result in revenue losses. the role of export taxes, in part because their elimi- · Fiscal considerations were the main factors cited nation is now generally accepted as a way of as limiting the extent of targeted trade reforms. enhancing growth prospects and strengthening a Greater trade liberalization could have been tar- country's external position. The effect on revenues geted and achieved if more attention had been from lowering export taxes will depend on how given to supportive fiscal policies and to revenue- much and how rapidly the reduction expands total neutral trade measures. trade and reduces illegal activities such as smug- · The effects of trade reform on revenues also gling. Since export taxes are often claimed to be depend significantly on the accompanying substitutes for some form of income tax on hard-to- macroeconomic policies and, in particular, on an tax sectors such as agriculture, their reduction or appropriate exchange rate policy. elimination is likely to be accepted by the country authorities if it is implemented as one element of an Lessons for the Design of Trade Policy overall tax reform package for broadening the tax base. Elimination of export taxes will also have a The discussion in this chapter suggests that there is positive impact on the producers of the commodi- scope for so tailoring the pattern of trade liberaliza- ties that are affected. In the case of agricultural tion as to avoid adverse consequences for revenues. products, these producers may be among the poorer Accordingly, adjustment programs supported by the segments of society (Box 4.1). IMF and the World Bank should, at the start of the Finally, in many programs supported by the IMF reform process, focus on broad-based trade liberal- or the World Bank, substantial trade liberalization ization measures, with a front-loading of those ele- has been accompanied by a devaluation of the ments that are likely to have a positive impact on exchange rate in order to, among other things, pro- revenues. The discussion also underscores the vide incentives for exporters so that they can take importance of sound macroeconomic policies and, advantage of the more liberal trade regime. In gen- in particular, the need for an appropriate exchange eral, the effect of devaluation on trade taxes is rate and for efforts to broaden the domestic tax base. ambiguous (Tanzi 1989) and will depend on the Nevertheless, the problems posed by trade liberal- price elasticity of import demand; if import ization in cases where it is perceived to have an ini- demand is inelastic, the devaluation will result in a tial negative effect on revenue (especially in higher value of imports in local currency terms and countries that rely significantly on trade taxes) will increase revenues at any given level of tariffs. should not be minimized. Even if alternative rev- enue and expenditure measures are readily avail- able, there are likely to be political and economic Case Studies and Other Empirical Evidence challenges. In addition, the consideration of fiscal Several studies (for example, Ebrill, Stotsky, and alternatives will take place in the context of pro- Gropp 1999; Sharer and others 1998) have exam- gram design, which usually involves fiscal pressures ined the actual impact on fiscal revenues following in many areas of both revenue and expenditures, the implementation of trade reforms. These studies including outlays in support of structural reform strengthen the observation noted above that the other than trade policy. sequencing of trade reforms can be done in a way Since sustained trade liberalization could eventu- that minimizes its adverse effect on revenues. Some ally lead to a reduction in the share of trade tax rev- of the main conclusions from these studies follow. enues in total receipts, maintaining revenue performance will require compensating domestic · For countries that initially started with highly tax reforms. Given the long gestation period of tax restrictive trade regimes, trade reforms were imple- policy and administrative reforms, it is critical that mented with a view toward protecting budgetary the reform of domestic taxes be considered at the revenues, and for the most part countries were able very outset of the trade reform exercise and that to achieve significant liberalization without com- technical assistance be sought at an early stage of promising their fiscal objectives. Empirical evi- the liberalization process. 26 Fiscal Dimensions of Trade Liberalization B O X 4 . 1 E X P O R T TA X E S Developing countries often impose export taxes on GI's interventions in the domestic market had a simi- primary commodity exports. Export taxation is one lar depressing effect on producer prices, which fluc- policy instrument that is not subject to WTO disci- tuated around US$5 per kilogram during the 1980s. plines, reflecting WTO members' focus on import Who benefited from the bourbon vanilla cartel policies. This makes it particularly important to and CAVAGI's domestic policies? Indonesian pro- determine the economic effects of such policies. In ducers were clearly the winners. The losers were some cases taxes are imposed in lieu of royalties for Madagascar's producers--mainly smallholders, the extraction of minerals; in others they are used numbering about 60,000, with an average pro- to provide protection to industries that process pri- duction of 130 kilograms and an average income mary commodities. In the latter case they can have of US$650 per plantation. adverse impacts on the poor that need to be care- A recent study provides estimates of the produc- fully monitored and analyzed. The export taxes er prices that would have prevailed in Madagascar mean that primary producers and farmers receive a had the marketing board been abolished. These price below that prevailing in world markets for are close to US$26, well above the US$5 price their commodities. Elimination of the tax will raise fixed by CAVAGI. Taking into account the increase their incomes but may bankrupt established pro- in production that such a change in prices would cessing facilities that are viable only if they pay have generated, laissez-faire policies would have lower-than-world prices for their inputs. Such increased the vanilla producer surplus eightfold. plants may employ poor urban labor, giving rise to Perhaps surprisingly, given the market power that a policy conundrum. In such cases a careful analysis Madagascar had in international markets, free of the appropriate trade regime for poverty allevia- trade (no intervention) would have also increased tion and the provision of safety nets is needed. Madagascar's welfare, by 0.5 percent of GDP--the Sometimes export taxes are used in an attempt to outcome of a major gain equivalent to 2.2 percent exercise market power, and in such cases the policy of GDP for producers, partially offset by a 1.7 per- can have a very adverse effect on the poor. An exam- cent of GDP loss to the marketing board. ple is Madagascar's marketing board for vanilla. An alternative to free trade would have been for In 1960 Madagascar, the world's lowest-cost CAVAGI to eliminate its interventions in the domestic producer of high-quality bourbon vanilla, account- market but to continue to exploit its market power in ed for 60 percent of world exports of natural vanil- international markets through an export tax. Esti- la. From its dominant position, Madagascar mates suggest that the optimal export tax would organized a bourbon vanilla cartel, with Comoros have been close to US$25 per kilogram instead of and Reunion, which set high export prices. Mada- the US$61 implicit tax that CAVAGI was imposing on gascar restricted supply by regulating its domestic producers. This would still have resulted in a dou- market through a marketing board (CAVAGI) that bling of the vanilla producers' surplus and, when fixed low producer prices and required licenses for combined with the tax revenues, would have gener- growing, preparing, and exporting vanilla. ated a welfare gain close to 1 percent of GDP. If this strategy were to be assessed by the effect it A likely explanation as to why these alternative poli- had on export prices of vanilla from Madagascar, it cies were not pursued is that the marketing board's was a clear success. The export price of vanilla revenue would have declined under both scenarios. increased from US$10 per kilogram in the late 1960s This suggests that Madagascar's marketing board to more than US$65 in the early 1990s. However, pricing policies had objectives other than welfare Madagascar's share of world markets declined to 30 maximization and that the heavy implicit taxation of percent as Indonesia, which was outside the cartel, small producers generated an important income took advantage of high world prices to develop its redistribution from the rural poor to the urban elite. export capacity. The entry of Indonesia into world markets left the total value of Madagascar's exports Source: Prepared by the volume editors, based on de Melo, constant throughout the 1970s and 1980s. CAVA- Olarreaga, and Takacs (2000). 27 5 L . A L A N W I N T E R S Trade Policies ed with an eye to its direct poverty consequences. It should, rather, be set on a sound basis for Poverty overall, with recognition that some modification may be Alleviation inevitable for political and other reasons. The primary way to deal with poverty is through general antipoverty policies. Trade Reform and Poverty International trade scholars have long understood that although H owever it is defined, poverty is not for small countries, trade interventions are generally a direct result of international inefficient and wasteful, their inefficiency is usually trade. Rather, poverty reflects low earn- dominated quantitatively by their redistributive ing power, few assets, poor access to communal effects. That is, the net losses from intervention will resources, poor health and education, powerless- generally represent large positive effects for some ness, and vulnerability. It does not matter what people and households and large negative effects for causes these features so long as they exist, nor what others. Correspondingly, although removing inter- relieves them if they can be relieved. Trade policy ventions will generally be income enhancing overall, matters only to the extent that (a) it affects the it is likely to generate both winners and losers.1 For direct determinants of poverty and (b) relative to example, liberalizing an import sector typically the whole range of other possible policies, it offers redistributes real income from producers to con- an efficient policy lever for poverty alleviation sumers as prices fall, and between different factors of (more poverty bang for a buck of forgone opportu- production in such a way that some gain while oth- nities). ers lose more than average. Trade liberalization may have adverse conse- The important positive issues are empirical: does quences for some--including some poor people-- trade liberalization generally create poverty, and that should be avoided or ameliorated to the under what circumstances might it do so in specific greatest extent possible. My fundamental belief, cases? In Winters (2000a), I develop a detailed tax- however, is that trade liberalization aids growth, onomy linking trade shocks to household and indi- which, in turn, aids poverty alleviation. I also vidual poverty and extract 11 key questions that believe that a widespread reform will contain help to answer the latter question (see Box 5.1). enough positive elements so that, in general, only a Finding an instance in which a trade reform causes few people will end up as net losers. Trade policy poverty may not constitute a refutation of an intel- should therefore generally not be closely manipulat- lectually interesting hypothesis, but it does pose a 28 Trade Policies for Poverty Alleviation B O X 5 . 1 K E Y Q U E S T I O N S F O R D E T E R M I N I N G T H E P O V E R T Y I M PA C T O F A T R A D E R E F O R M Will the effects of changed border prices be passed dicting either the price effects or the factor inten- through to the rest of the economy? Trade policy sities of affected sectors can be complex, as was and shocks operate primarily through prices. If seen with the Latin American reforms of the price changes are not transmitted--for example, 1980s and 1990s. In addition, if factor supplies because governments continue to fix the internal show some elasticity, part of a trade shock will prices of goods that they have ostensibly liberal- show up as changes in employment rather than ized internationally--the most direct effects on in factor prices. At the limit, a factor with a per- poverty, whether positive or negative, will be nul- fectly elastic supply will experience only employ- lified. ment effects. This is most pertinent for labor Is reform likely to give poor consumers access to markets. If the prevailing wage is determined by new goods? Perhaps the most direct effect of subsistence levels, switching people from one trade reform on poverty is through the prices of activity to another has no perceptible effect on goods and services in which poor households poverty. If, however, the trade-affected sector have large net positions. The biggest price shocks pays higher wages (because, say, it has an institu- occur when either the initial or final price is finite tionally enforced minimum wage), increases in and the other is infinite (that is, when there is no activity will tend to reduce poverty, and declines market). A shock that completely undermines an will tend to increase it. The formal-informal divide important market--say, for a cash crop or a form is important in this respect. In all this, it is impor- of labor--is likely to have major implications for tant to remember the difference between the poverty. Similarly, making new opportunities, functional and the personal distribution of goods, or services available to the poor can great- income. Falling wages for unskilled labor gener- ly enhance welfare. ate poverty only to the extent that the poor Is reform likely to affect different household mem- depend disproportionately on such wages. bers differently? Within a household, claims on Will the reform actually affect government rev- particular goods and endowments of particular enue strongly? One's immediate reaction is that assets (labor) are typically unevenly distributed. It cutting tariffs will reduce government revenue. is possible that poverty impacts will be concen- Although at the limit this is clearly true (zero tar- trated on particular members--usually females iffs yield zero revenue), many trade reforms actu- and children, who may lose personally even ally have small or even positive revenue effects, when the household gains in aggregate. especially if they convert nontariff barriers into Will spillovers be concentrated on areas and activ- tariffs, remove exemptions, and get tariff rates ities that are relevant to the poor? The sectors of an down to levels that significantly reduce smug- economy are interlinked, and, if substitutability is gling. Even where revenue falls, it is not inevitable high, a shock will be readily transmitted from one that expenditure on the poor will decline. That, sector to another. Frequently the diffusion will be ultimately, is a policy decision. so broad that it has little effect on any particular Will reform lead to discontinuous switches in activ- locality or sector, but sometimes--for example, ities? If so, will the new activities be riskier than the where trade in services is very local--the trans- old ones? If a trade liberalization allows people to mission is narrow but deep. Then it is necessary combine "national" and "international" activities, to ask whether the second-round effects have it is most likely to reduce risk: foreign markets are serious poverty implications. Agricultural stimuli likely to be less variable than domestic ones, and can confer strong propoor benefits on local even if they are not, risk spreading is likely to economies via benign spillovers. decrease overall risk. If, however, trade reform What factors are used intensively in the most leads to more or less complete changes in activi- affected sectors? Changes in the prices of goods ties, there is a possibility that risk will increase, if affect wages according to factor intensities. Pre- the new activity is riskier than the old one. (continued) 29 T R A D E P O L I C Y R E F O R M I N C O N T E X T B O X 5 . 1 ( C O N T I N U E D ) Does the reform depend on or affect the ability of Thus, if a reform implies very large shocks for par- poor people to take risks? The very poor cannot ticular localities, mitigation through phasing, or, bear risk easily. Because the consequences of better, through compensatory and complemen- even small negative shocks are so serious for the tary policies, could be called for. There is a trade- poor, they may be unwilling to take opportunities off, however, because, typically, larger shocks will that increase their average income if the chance reflect bigger shortfalls between current and of losses also increases. This might leave them potential performance and hence larger long-run with only the negative elements of a reform pack- gains from reform. age. Similarly, if a reform makes it more difficult Will transitional unemployment be concentrated for the poor to continue their traditional risk-cop- on the poor? The nonpoor typically have assets ing strategies, it may increase their vulnerability that carry them through periods of adjustment. to poverty even if it raises mean incomes. The situation might be unfortunate for them, but If the reform is broad and systemic, will any it is not poverty strictly defined. The poor have growth it stimulates be particularly unequalizing? few assets, so even relatively short periods of Economic growth is the key to sustained poverty transition could induce a descent deep into reduction. Only if it is very unequalizing will it poverty. If the transition impinges on the poor, increase absolute poverty. there is a strong case for using some of the long- Will the reform imply major shocks for particular run benefits of reform to ease their adjustment localities? Large shocks can create qualitatively strains. different responses from smaller ones; for exam- ple, markets can seize up or disappear altogether. Source: Winters (2000a). real policy question: should we automatically con- theoretical. For this and other reasons, losers will demn a trade reform because it means that one usually be better able to articulate their interests poor person loses or one person is pushed into than gainers, and so the volume of opinion is not a poverty? I believe very strongly that we should not. sufficient indicator of the relative strengths of the Rather, the identification of hardship arising from a pluses and minuses of a policy change. This is par- generally desirable policy reform should stimulate ticularly true given that the poor are generally much the search for complementary policies to minimize less able to advertise and defend their interests than the adverse consequences and reduce the hurt that are wealthier groups. they cause. Rejecting any reform that adversely In what follows, I explore three responses to the affects any poor person is a recipe for long-run stag- possibility that trade reform can create poverty: nation and for an ultimate increase in poverty. Even manipulating trade policy itself, compensating the the requirement that no household fall temporarily losers or the poor, and pursuing complementary into poverty is likely to be extremely restrictive in policies to try to ensure that as few people as possi- poor countries. The more utilitarian view that the ble are net losers. number of households (or persons) in poverty should not increase is more appropriate, although Can Trade Policy Be Managed to Alleviate even then, consideration of the depth of poverty is Poverty? required. All judgments ultimately must be quantitative, One natural response to the possibility that trade not just qualitative. In practical circumstances, it is liberalization could exacerbate poverty in certain easier to identify losers from trade policy than sections of a society is to "manage" liberalization in potential gainers. Losers are identifiable, concrete, a way that eliminates or at least reduces the prob- and personified (see Krueger 1990), whereas the lems. At the conceptual level, this is just common gains are diffuse and appear merely prospective and sense: poverty alleviation is arguably our highest 30 Trade Policies for Poverty Alleviation priority, whereas trade policy is just a means to an growth as a cause of growth) or include much more end. It makes sense to marshal all the tools we have than just trade policy--and that their econometrics toward achieving our principal goals, and indeed, it are weak. But Rodriguez and Rodrik do not argue would be perverse to do anything different. that trade liberalization is harmful to growth, nor But on a practical level, the question is how to use do they deal with other evidence for a beneficial trade policy to achieve poverty objectives. First, relationship, such as the manifest failure of closed there is the possibility that we do actually have goals economies and the findings of a number of case other than poverty alleviation--for example, as studies (see, for example, Srinivasan and Bhagwati regards average incomes, security, foreign policy, or 1999). Thus while Rodriguez and Rodrik should environmental sustainability--and these would certainly inspire greater modesty in policy advice need to be factored in. Second, even leaving aside and renewed research efforts, they have not (yet) additional objectives, we need to decide which reversed the presumption that openness is likely to measure of poverty we are aiming at: there are boost long-run growth. choices even among income- or consumption- The difficulty of establishing an empirical link based measures, let alone among the various con- between liberal trade and growth arises at least part- cepts and dimensions that characterize modern ly from two difficulties, both of which should views of poverty. Third, there may be questions inform our policy attitude. The first is the difficulty about trading poverty in one region against that in of measuring trade stances once one comes inside another, and there will certainly be, fourth, tradeoffs the boundary of near autarchy: tariffs need to be between poverty today and poverty tomorrow. aggregated, quantitative restrictions assessed and Fifth, what else figures in the policy packages among then aggregated, and the degrees of credibility, vul- which we are deciding? Are other policy instru- nerability to lobbying, and enforcement measured ments frozen at current levels, so that the question (see Winters 2000b). This suggests that while one is only one of how trade reform impinges directly should staunchly recommend openness, one needs on the real incomes of the poor? Or can we presume to be cautious about declaring particular regimes that other policies will be optimized, so that, say, open or not. Which was the more open in 1997, boosting incomes in the top decile at the expense of Brazil, or Chile? Both had average most-favored- higher prices for the poor is acceptable because it nation (MFN) tariffs of around 11­12 percent, but will permit a redistribution via the tax-benefit sys- in Chile there appeared to be little discretion and tem that more than offsets the initial growth in the little sensitivity to industrial lobbying, whereas in income gap? These questions illustrate that saying Brazil political pressures could be observed almost "manage trade policy" is not helpful until one spec- every day. ifies how to manage it. The second difficulty is that, although liberal trade policies are likely to be beneficial under any circumstances (because they enlarge the set of Don't Do It opportunities), a quasi-permanent effect on growth One response to the fear that a trade liberalization almost certainly requires combination with other will cause poverty is, "don't do it." But even if the good policies as well. The latter point is made direct effect of a reform might be to worsen poverty repeatedly by the IMF and the World Bank in their overall, this is not generally a satisfactory response. policy advice.2 Krueger (1990) has argued that Although it has proved difficult to isolate the effects openness is likely to be correlated with better policy of trade liberalization on economic growth empiri- in a number of dimensions, and supporting evi- cally, the predominant view is that it has an impor- dence for this assertion might be detected in Ades tant role. The well-publicized cross-country studies and Di Tella (1997, 1999), on corruption, and in that supported this view in the 1990s (for example, Romer (1993), on inflation. Thus, openness brings Dollar 1992; Sachs and Warner 1995; Edwards advantages not only on its own but also as part of a 1998) have recently received rough treatment from constellation of policies designed to ensure efficien- Rodriguez and Rodrik (2001). The latter argue, with cy and competition in markets, and transparency some justification, that these studies' measures of and predictability in policymaking. openness are flawed--in particular, because they The second part of the openness-poverty link either are endogenous (at least as much due to concerns the connection from growth to poverty. 31 T R A D E P O L I C Y R E F O R M I N C O N T E X T Growth needs to be strongly biased against the poor a series of exceptions is not strong. One needs very before it is likely to worsen poverty absolutely. (The compelling evidence of the efficacy of such inter- effect on inequality is a different story but, in my ventions, and such evidence is, on the whole, miss- view, a distracting one.) There are examples of such ing. Simply appealing to the experience of East Asia a negative relationship (see, for example, White and is not persuasive; it is not beyond dispute that these Anderson 2000), but they are equally balanced by countries' trade interventions were important or cases in which growth disproportionately favors the beneficial (Lee 1996 suggests the opposite for poor. Thus, Dollar and Kraay's (2001) finding that, Korea), and it is far from certain that other coun- on average, growth is good for the poor does seem tries have the policymaking institutions to be able to be robust (as it was in the earlier work of Gallup, to replicate East Asian policy stances effectively. Radelet, and Warner 1998), and so does their con- In addition to efficiency considerations, we must clusion that growth driven by trade liberalization is recognize that trade liberalization is a political act no different in that respect. A challenge to the latter and that governments must generate sufficient view can be found in the early work of Lundberg political support to sustain the reform. Even the and Squire (2000), but on further investigation they most rigorous reforms need tempering for political too concluded that trade liberalization benefited the reasons; see, for example, Edwards and Lederman poor, albeit by less than it helped better-off house- (1998) on Chile, where certain agricultural goods holds. were granted special protection in the form of price To conclude, I would argue that although there bands. Recognizing the need for such compromise remain a number of pressing research questions in is not the same as recommending it, however, and it the area, a liberal trade regime almost certainly is important to remember that the poor are even assists poverty alleviation in the long run. Thus, lib- weaker in political markets than in economic ones. eralization should have a place in the armory of a Rarely will protecting the poor and reaping signifi- poverty-conscious government. This does not cant political support for a reform coincide. Gov- imply a call for the immediate dismantling of all ernments are well advised to do everything possible trade restrictions, and it certainly does not imply to avoid using the instruments of trade policy for that opening the border is all that is needed, but it political purposes. One of the most powerful tools does, I believe, mandate a serious and credible com- for avoiding political pressures is uniformity-- mitment to openness in the foreseeable future. explicitly treating all commodities equally. Such a motivation was clearly articulated in Chile as it entered its big reform in the mid-1970s (Edwards Don't Do It All and Lederman 1998). A second response is,"don't do it all: while everyone There may be a stronger case for exceptions to lib- is in favor of liberalization in general, certain sectors eralization for the sake of direct poverty alleviation: or products should be exempt." In fact, all countries the outcome is objectively measurable and can have such exceptions (agriculture in Europe and arguably be isolated politically from general inter- clothing in the United States, for example), but that vention. Thus, if particular products can be clearly does not necessarily make them good economics. identified with the poor as either consumption or Considering overall economic performance, there production goods, it may be justified to postpone undoubtedly are cases in which an isolated inter- their liberalization significantly. There are, however, vention in trade would be beneficial to immediate some important caveats to such a recommendation. economic welfare--where externalities, informa- First, the calculation needs to be rigorous in defin- tion failures, or just random shocks can be usefully ing "the poor" whose interests are being protected. overcome by a well-judged intervention. But given (Ravallion and van de Walle 1991 show how in the difficulty of identifying these cases, of prevent- Indonesia the poor and the very poor had conflict- ing their capture by interest groups, and of avoiding ing interests in rice liberalization.) Second, the giving a systemic signal that lobbying for interven- products do need to be tightly linked to the poor in tion pays, it is not clear that it will be beneficial order that the distributional gains of protecting overall to pursue them. Thus, although one does not them are not offset by efficiency losses elsewhere in need to progress all the way to free trade to reap the the economy. The goods concerned need to be of benefits of liberalism, the general case for planning great significance to the poor--almost always, a 32 Trade Policies for Poverty Alleviation foodstuff on the consumption side, and frequently with long adjustment periods and postponing liber- an agricultural good on the production side--and alization because "the time is not ripe." The key is of little interest to other sections of society. On the credibility that reform will actually occur. Adjust- latter dimension, for example, it was reported in the ment costs may be lower if adjustment can be 1970s that in addition to any effects on the poor that spread somewhat through time, but they are proba- Egypt's bread subsidies may have had, they also bly increased if adjustment is resisted in the hope made it worthwhile for pig farmers to feed their that the threat of liberalization will go away. Several stock on fresh bread. trade reforms have been accelerated once they have Third, close monitoring is required to ensure that been launched; examples include implementation the desired effects do actually emerge. An important of free trade in the European Economic Communi- goal to keep in mind is that the poor continue to ty, of the Kennedy Round tariff cuts, and of the tar- have access to the effects of the policy and do not get iff cuts planned in the Association of Southeast hustled out of the way by more articulate and pow- Asian Nations (ASEAN) Free Trade Agreement. erful middle-class interests. In general, the poor fare Usually the acceleration happens at the behest of the very badly in discretionary allocation mechanisms private sector, presumably because, once it is precisely because they are marginalized politically accepted that reform will occur, business is keen to and socially, as well as economically. Thus, for adjust rapidly. example, export restrictions to keep down the price Thus, undertaking a major trade liberalization in of a local staple will probably not benefit the poor if phases is probably desirable, just as the Uruguay low prices mean that nonprice rationing is required. Round, for example, permitted long adjustment Fourth, a long-term plan is needed to help reduce periods. The phasing should, however, not merely the dependence of the poor on the policy interven- entail postponing the largest adjustments longest; tion. Otherwise, the intervention just amounts to it should pay attention to the different adjustment stopping the clock, which offers little prospect of needs of different sectors and to the interactions long-run development. between different parts of the package. For exam- Overall, "don't do it all" is not a suitable policy ple, if the inputs and outputs of a particular sector recommendation. The politics, especially protecting are liberalized at very different rates, the sector the interests of the poor, will be easier if the govern- could face either negative or excessively positive ment can explicitly reject special pleading on the incentives for production during the transition.3 grounds that everyone is receiving equal treatment. Whatever the transition period, credible commit- De facto, there may be--there may have to be-- ment to the final goal is important, for without it, some slippage in such an attempt, but it seems to neither current nor potential future activities will me undesirable to go into the process expecting or look desirable, and there will be a diversion of recommending slippage. The only exception I effort into lobbying. would make would be for temporary exemptions for goods or services that can be clearly and closely Compensatory Policies for Developing linked to the poor. A high burden of proof should Countries be placed on candidates for such exemptions to prove their efficacy. If trade liberalization causes poverty among certain sections of society, the next question is whether soci- ety can offset the effect directly. Despite the theoreti- Don't Do It Now cal attractions of lump-sum budgetary transfers for "Don't do it now" is a more useful response than economists, governments are not generally attracted the others in some circumstances. For example, to them because of their cost, their transparency trade reform in the midst of recession seems likely (and the transparency of their abuse), and the to give rise to more, and more durable, transitional appearance that they do little to cure "the problems" unemployment than reform in a boom. Again, that individuals face. Rather, assistance is usually where investment is necessary to allow the produc- offered, if at all, in terms such as retraining, reloca- tion of export-quality goods, it may be desirable to tion assistance, and temporary income support. In allow time for that to occur. There is, however, a fact, while these approaches probably do have a con- world of difference between committing to policies tribution to make, even they face severe difficulties. 33 T R A D E P O L I C Y R E F O R M I N C O N T E X T Official retraining has mixed success under any cir- thresholds are set low enough; and, since relieving cumstances, and, what is worse, it is difficult to sepa- poverty is more or less universally recognized as a rate those cases where trade is to blame from those responsibility of the state, there is little argument where it is not. Unless one is willing to underwrite about the legitimacy of such interventions. almost any adjustment, identification of cases is a Targeting is a major problem for safety nets, not major difficulty. Making a general commitment, only technically but also because the middle classes however, is not attractive because of the potentially are often better able to access them than the poor. huge cost and because doing so shifts private risk to Sustainability is another difficulty; a major trade the public sector, with all the attendant moral hazard shock could put severe financial pressure on a problems. It is not the role of the state, nor is it feasi- scheme just at a time when it is most needed. Raval- ble, to absorb every negative shock that might afflict lion (1999) offers some useful thoughts on setting individuals. Yet it is difficult to make a moral case as up safety nets. Workfare is a good start, provided to why trade shocks warrant adjustment assistance that the wage is low enough, that there is little or no while other shocks do not.4 administrative discretion in its application, and that A further complication arises in giving compen- the tasks set are seen to be of communal interest. In sation in a way that encourages rather than discour- fact, Ravallion suggests that local communities ages adjustment. European agricultural policy is select the projects to be undertaken under workfare essentially designed to protect farmers from the and that better-off communities should be asked to consequences of declining comparative advantage, cofinance the projects. Workfare has to be supple- yet it has the effect of rewarding current, not for- mented, however, by schemes to provide food to mer, farmers. Compensation may be decoupled people such as the elderly and infirm who cannot from current output but not from farming as an work and to children (through, for example, food- activity. for-education schemes). These supplementary In cases where liberalization leads to the loss of schemes may be tripped on and off according to jobs, government can insist on, and perhaps help need, but they should have a permanent infrastruc- finance, redundancy payments. These payments can ture and sensitive and quick triggers. Expenditure help some people avoid poverty, if they use their on safety nets is almost by definition countercycli- money productively, but they are not guaranteed to cal, and so a firm commitment by government is do so. (See Winters 2000a on the "new poor" in required to ensure that the money does not dry up Zimbabwe.)5 Moreover, redundancy payments typi- in times of greatest need. cally reward past service, not current need, and so Examples of useful safety nets can be found in they are not particularly well targeted for poverty Bangladesh. According to the Consumer Unity & alleviation purposes. Trust Society, General compensatory policies, including safety nets, are designed to alleviate poverty from any It is generally recognized that programs such as source directly. They replace the problem of identi- Food for Education (FFE), Vulnerable Group fying the shock with the task of identifying the poor. Development (VGD), Test-Relief, and Food for Ideally, countries should already have such pro- Work positively induce alleviation of poverty. grams in place. Indeed, a major part of the effect of For example, during the unprecedented floods these programs arises from their mere existence of 1998, about 4.5 million VGD cards were dis- rather than their use: they facilitate adjustment by tributed in Bangladesh, which provided crucial assuring the poor that there is a minimum (albeit a help at a critical time. The FFE program has barely acceptable one) below which they will not be helped increase school attendance of poor chil- allowed to fall. Such schemes, if trade-adjusting dren by 21%. (CUTS 1999: 110) countries do already have them, offer the advan- tages over tailor-made schemes of automaticity, The safety nets in Zambia and Zimbabwe, by con- immediacy, and a degree of "road-testing," and they trast, are currently regarded as too poorly run and also avoid the problems associated with targeted underfunded to be able to offer serious assistance to trade adjustment assistance. If they are sensibly losers from trade liberalization. constructed, they need not entail huge expenditure; Safety nets are not the only answer to the threat of there is little chance of moral hazard problems if the increasing poverty from trade liberalization, but 34 Trade Policies for Poverty Alleviation they are an important part of the response. They relatively underdeveloped in most of the smallhold- can generally be targeted better than other policies, er areas, increasing numbers of resettled and com- and they are not very distortionary of market forces. munal households are now becoming involved as If countries do not have safety nets already, they producers of the main crops. This has primarily should consider setting them up as part of the con- been the result of "outgrower" schemes and of text for a trade liberalization that may create short- sourcing or subcontracting by large-scale commer- term poverty. The safety nets should not, however, cial farms. The Horticultural Promotion Council be trade shock­specific. (HPC) estimates that around 3,000 small-scale farmers are now growing for export on a contract basis, accounting for approximately 10 percent of Complementary Policies for Zimbabwe's exports.6 In January 1999 the HPC Better-Functioning Markets established the Small-Scale Linkage Programme, A critical issue concerning the poverty impacts of designed to provide communal and resettled farm- trade liberalization, especially for surprises connect- ers with the knowledge and skills to produce high- ed with it, is the functioning of markets. Trade liber- value, off-season export crops. alization must be accompanied by monitoring to determine whether any markets are failing. Policies Credit Markets designed to ensure that markets continue to func- tion or to develop, where required, will have high Development economics affords many examples of payoffs for both aggregate income and poverty alle- how missing credit markets have prevented devel- viation. Some important circumstances are dis- opment, and the same phenomenon is visible in cussed next. responses to trade liberalization. Thus, for example, achieving minimum consignment size might entail hiring draft power or seasonal labor, but this is not Infrastructure Support possible without credit. Similarly, establishing Potential opportunities for poor producers to bene- informal businesses in areas such as trading may fit from a more open trading regime have been lost require more capital than the poor can raise. These because critical infrastructure was either absent or cases in which the poor are not able to respond to had deteriorated. In both Zimbabwe and Zambia incentives as strongly as the less poor replicate the remote farmers found their opportunities con- results of López, Nash, and Stanton (1995) in their strained by inability to reach major market centers. panel study of Mexican agriculture. In the same way, many of the benefits from relaxed retailing regulations and from availability of new or Labor Mobility cheaper goods have been confined to urban and periurban areas. The secret of spreading the benefits of increasing labor demand widely is labor mobility. If markets are segmented for cultural or geographic reasons, Market Institutions breaking down these barriers through information The poor frequently seem unable to attain the eco- and facilitating physical mobility will have an equal- nomic mass required for the establishment of mar- izing effect. kets that, once in place, may be viable. Policy should aim at the creation of the market as an institution, Establishing Businesses not at the ongoing subsidization of market activity. One aspect of facilitating the participation of the If the regulations for establishing new businesses are poor in markets may be to find means to allow them cumbersome, if the businesses' ability to obtain to combine very small consignments of inputs or inputs (especially utilities) is weak, or if regulations outputs into reasonably sized bundles and so reduce on expansion and on labor recruitment and separa- transactions cost sufficiently to make dealing with tion are restrictive, this could curtail the willingness poor producers worthwhile. Horticulture in Zim- of entrepreneurs to start or expand operations. A babwe offers an illustration of a successful policy of success story of business deregulation is the growth this kind (Winters 2000a). Although horticulture is of maize hammer milling in Zimbabwe. Following 35 T R A D E P O L I C Y R E F O R M I N C O N T E X T domestic deregulation, 3,500 new hammer mills This is not, however, a license to postpone the opened, mainly in rural areas, and the share of ham- design, announcement, and locking in of the reform mer millers in total maize milling has increased to itself. Any of these delays--for example, announc- almost 80 percent.7 These mills are mechanically ing that liberalization is necessary but that its form simple and robust (being based on swinging or will be worked out once certain other reforms have rotating hammers in a grinding chamber) and can been implemented--would seem likely to result in be used by unskilled labor. They provide quality the worst of all worlds. It would create uncertainty maize meal products to nearby customers in poor and incentives to lobby government and, indeed, communities, saving them significant transport would look to many commentators like a de facto costs. In 1995 hammer mills were estimated to reluctance to liberalize trade. In particular, in the employ 7,512 permanent workers (751 in urban absence of a clear and monitorable plan for specific areas); when casual workers and rural activities are pieces of infrastructure, a general wish to wait until included, the sector employs some 13,000 workers. the roads or ports are "ready" is just a recipe for About 18 percent of the employees in urban ham- indefinite postponement. A credible plan for liber- mer mills are female, as are 8 percent of the employ- alizing the borders--albeit one with significant ees in rural areas.8 transition periods--will be an important stimulus to reforming these other areas in ways that will typ- ically have other benefits as well. Prerequisites or Concomitants? It is also well to record that there are disadvan- In many cases actual policy debate appears to hinge tages as well as advantages to phased adjustment. on whether complementary policies of the sort just Populations can certainly suffer from reform fatigue described should be prerequisites for a trade liberal- and would actually be more comfortable with a def- ization."Everyone accepts that trade liberalization is inite, even if ambitious, reform plan than with one desirable in the long run," the argument goes, "but that drifts into the indefinite future. Phased adjust- various supporting policies must be in place before ment implies a longer time spent out of equilibri- it is attempted." Here, even more than in the matters um, and in most discussions it is not proved that the discussed above, we have no formal analysis to fall integral of shallow adjustment costs over a long back on. There is a literature on sequencing reform period is smaller than that of deep costs over a within the trade sector and between trade and capi- shorter period.9 Moreover, delay postpones the ben- tal accounts, but there are no convincing empirical efits of full reform. Finally there are likely to be generalizations about sequencing in the sense dis- aggregate gains from trade reform even in the cussed here. Moreover, the question is only partly absence of complementary policies. A trade reform economic; part of it is political and concerns increases opportunities for desirable exchange, and whether a reform postponed is a reform preempted. these will exist even with poor infrastructure, and I argued above that there may well be a case for even though there would have been more opportu- phasing in a reform over a long period provided nities had the infrastructure been better. And this that the final destination is clear (and not likely to applies to the poor as much as to other people. It is be contested) and that the transition is well possible that in the absence of complementary poli- designed and does not amount just to postponing cies, the poor will suffer (say, because of rising all effective change until the last moment. Given prices) whereas with such policies they will gain that a well-conceived and well-executed reform because they will receive income gains to offset the generates a potentially infinite-lived stream of bene- price rises. But there is no general theorem to this fits, whether this occurs over three years or, say, nine effect; the case remains to be made. years is not that important. The same logic applies to delays required to put complementary policies in Conclusion place (or, indeed, to compensatory mechanisms if that is the route chosen). Thus, for example, there Trade reform almost invariably brings with it two may be a case for delaying the implementation of a changes that help in the battle against poverty: it liberalization while legislation on business forma- induces efficiency in the use and allocation of tion or labor market operation is put in place and resources (the economist's beloved static gains), and plans for protecting market institutions are laid. it fosters long-run growth. It also entails temporary 36 Trade Policies for Poverty Alleviation adjustment costs that reduce incomes immediately, Notes although these costs are almost always outweighed Preparation of this chapter was supported by Globkom, the Parlia- by the long-term benefits. Finally it has a host of mentary Commission on Swedish Policy for Global Development, direct and indirect effects on poverty that could go to which I am most grateful. I also thank Costas Michalopoulos for either way, depending on consumption and produc- comments on the outline and Rosie Bellinger for logistic help. tion patterns and on the nature of reform. 1 I say "generally" because second-best considerations or market The general presumption is that reform will help failures could reverse the result. alleviate poverty, but the direct and indirect effects 2 Mosley (2000) argues that the attempts of the IMF and the just mentioned make it likely that some will lose World Bank to prove this proposition have not been very suc- from liberalization--especially one that is narrowly cessful. (His attempts to prove the contrary are similarly focused sectorally--and it is certainly possible that flawed, however.) some of the poor will suffer. Still, others will gain, 3 Technically, the effective rate of protection (ERP) could and these will quite possibly include others among become very distorted. In Zambia, Oxfam and the Institute for the poor. Thus, tradeoffs are necessary. There is no Development Studies (IDS) found decreases in the ERP for alternative to case-by-case analysis if policymakers maize, as fertilizer prices (which were said to account for 76 wish to predict and preempt adverse effects, even percent of the cost of production) increased by more than out- put prices. This accounts for the loss of output there. though prediction is very difficult. One must be alive to the possibility that "predict and preempt" 4 See Decker and Corson (1995) on the U.S. Trade Adjustment policies will be captured by powerful interest Assistance Program, which doubles the length of unemploy- groups. Given these groups' strong interests in trade ment insurance coverage, from 26 to 52 weeks, for workers certified as displaced by trade liberalization. After serious abuse policy and the apparent ease with which trade poli- in its early years, when it was merely a transfer (over 70 per- cy can be captured (because its domestic costs are cent of claimants went back to work for the employer from usually hidden and the issues can be so easily pre- whom they were said to have been displaced), a training ele- sented in terms of standing up to foreigners), there ment was added. This had the effect of screening out are grave dangers in setting out to manipulate trade claimants who did not want or need training, but it apparently did nothing to increase the earning power of recipients. policy directly to avoid adverse poverty impacts. Only in the most obvious cases are the dangers like- 5 The "new poor" are retired public sector officials who have not ly to be worth incurring. managed to invest their redundancy payments sufficiently pro- ductively to maintain themselves above poverty levels. My general prescription, then, is for a rigorously liberal trade policy (even though it is recognized 6 These small-scale "outgrowers" tend to supply the four main that some slippage may occur for political reasons). packinghouses in Zimbabwe, which are the large-scale pro- General compensatory policies should then be used ducers that seek to add volume and diversify risk. to cure immediate hardship, and complementary 7 The 1995­96 Zimbabwe National Hammer Miller Status Study, policies should be pursued to enlarge long-term funded by the U.S. Agency for International Development. gains. Assessing likely impacts in the design of poli- 8 Ibid. cy reforms is of great importance. The set of ques- 9 This is not to deny the possibility--as, for example, if a major tions posed in Box 5.1 can help policymakers in the shock creates hysteresis in labor markets--but it needs to be design and implementation of reforms. proved. 37 II THE WORLD TRADE ORGANIZATION T he WTO was created in 1995 as one of discussions of the "engine" of the WTO--the princi- the outcomes of the Uruguay Round of ple of reciprocity (Chapter 7, by J. Michael Finger multilateral trade talks. The Uruguay and L. Alan Winters); the accession process (Chapter Round, which concluded in 1994 after eight years of 8, by Constantine Michalopoulos); and the dispute complex and sometimes contentious negotiations, settlement mechanism (Chapters 9, by Valentina was a landmark in the history of the trading system. Delich, and 10, by Robert E. Hudec). The last is the Agriculture and textiles and clothing became subject aspect of the WTO that attracts most attention. The to stronger multilateral disciplines, and the trading WTO is unique among international organizations system was extended to include intellectual property in that it has a well-functioning, binding dispute set- and trade in services. The WTO establishes the rules tlement mechanism. This is of great importance to of the trade policy game for its members, which developing countries, which generally will not be increasingly include developing countries. (Member- able to induce compliance with negotiated rules in ship at the time of writing stood at 144, but more bilateral disputes with large industrial economies. In than 50 developing countries have yet to join the practice, because countries value the trading sys- WTO.) A good understanding of how the WTO tem, the large and powerful tend to abide by the works and what it does is a necessary condition for rulings of dispute settlement panels, providing an maximizing the benefits of membership. incentive for developing countries to ensure that The chapters in this part discuss some of the they are able to use the system. major features of the WTO that are relevant to The ability to use the WTO system is a function of developing countries. A brief summary of the basic many factors. Among the necessary conditions are rules and the institutional mechanisms of the WTO that countries participate in the negotiations on the (Chapter 6, by Bernard Hoekman) is followed by rules of the game and that they use the WTO in a 39 T H E W O R L D T R A D E O R G A N I Z AT I O N proactive manner. Much of this Handbook is aimed Kluwer Law International, 1999). An early study of at helping countries do so. the GATT system that remains well worth reading is The chapters in Part II are not intended to provide Gerard Curzon, Multilateral Trade Diplomacy (Lon- comprehensive coverage of the WTO. Those seek- don: Michael Joseph, 1965). Robert Hudec's seminal ing an in-depth treatment of the WTO, its negotiat- Developing Countries in the GATT Legal System (Lon- ing history, and its dispute settlement case law are don: Trade Policy Research Centre, 1987) is an indis- referred to the sources listed below. pensable source for those seeking to understand the approach taken toward development issues in the WTO. An informative history of the Uruguay Round Further Reading and Sources of negotiations is presented in John Croome, Reshaping Information the Trading System (Deventer: Kluwer, 1999). The WTO Website, , provides direct Michael Trebilcock and Robert Howse, in The Regula- access to most of the documents submitted to the tion of International Trade (London: Routledge, institution, as well as to reports and case law. The 1998), provide a comprehensive treatment of WTO International Center for Trade and Sustainable Devel- rules, as well as a comparison between WTO disci- opment has links to all the major nongovernmental plines and those that apply in the European Union organizations on its Website, , and and the North American Free Trade Agreement it publishes an informative newsletter, Bridges, that (NAFTA). A recent analysis and description of the monitors WTO issues from a development perspec- economics and politics of the world trading system tive. WTO dispute settlement procedures are dis- can be found in Bernard Hoekman and Michel cussed in detail by David Palmeter and Petros C. Kostecki, The Political Economy of the World Trading Mavroidis in Dispute Settlement in the World Trade System: The WTO and Beyond, 2d ed. (New York: Organization: Practice and Procedure (The Hague: Oxford University Press, 2001). 40 6 B E R N A R D H O E K M A N The WTO: Since 1947, the GATT has been the major focal point for industrial country governments Functions and seeking to lower trade barriers. Although the GATT was initially Basic Principles largely limited to a tariff agree- ment, over time, as average tariff levels fell, it increasingly came to concentrate on nontariff trade policies and domestic policies having an impact on trade. (See the Glossary to this volume for a list of trade-related policies used T by countries.) Its success was he WTO, established in 1995, adminis- reflected in a steady expansion in the number of ters the trade agreements negotiated contracting parties. By the end of the Uruguay by its members, in particular the General Agree- Round (1994), 128 countries had joined the GATT. ment on Tariffs and Trade (GATT), the General Since the entry into force of the WTO, membership Agreement on Trade in Services (GATS), and the has grown to 144, as of the end of 2001. Trade-Related Aspects of Intellectual Property The WTO differs in a number of important Rights (TRIPS) agreement. (These and other major respects from the GATT. The GATT was a rather WTO agreements are contained in the CD-ROM flexible institution; bargaining and deal-making lay "Applied Trade Policy," which is included with this at its core, with significant opportunities for coun- Handbook.) The WTO builds on the organizational tries to "opt out" of specific disciplines. In contrast, structure that had developed under GATT auspices WTO rules apply to all members, who are subject to as of the early 1990s. binding dispute settlement procedures. This is The origins of the GATT were in the abortive attractive to groups seeking to introduce multilater- negotiations to create an International Trade Orga- al disciplines on a variety of subjects, ranging from nization (ITO) following World War II. Negotiations the environment and labor standards to competi- on the charter of such an organization were con- tion and investment policies to animal rights. But it cluded successfully in Havana in 1948, but the talks is a source of concern to groups that perceive the did not lead to the establishment of the ITO because (proposed) multilateral rules to be inappropriate or the U.S. Congress was expected to refuse to ratify the worry that the adoption of specific rules may affect agreement. Meanwhile, the GATT was negotiated in detrimentally the ability of governments to regulate 1947 by 23 countries--12 industrial and 11 develop- domestic activities and deal with market failures. ing--before the ITO negotiations were concluded.1 The main function of the WTO is as a forum for As the ITO never came into being, the GATT was the international cooperation on trade-related poli- only concrete result of the negotiations. cies--the creation of codes of conduct for member 41 T H E W O R L D T R A D E O R G A N I Z AT I O N governments. These codes emerge from the small number of contracting parties to the GATT exchange of trade policy commitments in periodic (only 23 countries), the benchmark for MFN is the negotiations. The WTO can be seen as a market in best treatment offered to any country, including the sense that countries come together to exchange countries that are not members of the GATT. market access commitments on a reciprocal basis. It National treatment requires that foreign goods, is, in fact, a barter market. In contrast to the markets once they have satisfied whatever border measures one finds in city squares, countries do not have are applied, be treated no less favorably, in terms of access to a medium of exchange: they do not have internal (indirect) taxation than like or directly money with which to buy, and against which to sell, competitive domestically produced goods (Art. III, trade policies. Instead they have to exchange apples GATT). That is, goods of foreign origin circulating for oranges: for example, tariff reductions on iron in the country must be subject to taxes, charges, and for foreign market access commitments regarding regulations that are "no less favorable" than those cloth. This makes the trade policy market less effi- that apply to similar goods of domestic origin. cient than one in which money can be used, and it is The MFN rule applies unconditionally. Although one of the reasons that WTO negotiations can be a exceptions are made for the formation of free trade tortuous process. One result of the market exchange areas or customs unions and for preferential treat- is the development of codes of conduct. The WTO ment of developing countries, MFN is a basic pillar contains a set of specific legal obligations regulating of the WTO. One reason for this is economic: if pol- trade policies of member states, and these are icy does not discriminate between foreign suppliers, embodied in the GATT, the GATS, and the TRIPS importers and consumers will have an incentive to agreement. use the lowest-cost foreign supplier. MFN also pro- vides smaller countries with a guarantee that larger countries will not exploit their market power by Basic Principles raising tariffs against them in periods when times The WTO establishes a framework for trade poli- are bad and domestic industries are clamoring for cies; it does not define or specify outcomes. That is, protection or, alternatively, give specific countries it is concerned with setting the rules of the trade preferential treatment for foreign policy reasons. policy game, not with the results of the game. Five MFN helps enforce multilateral rules by raising principles are of particular importance in under- the costs to a country of defecting from the trade standing both the pre-1994 GATT and the WTO: regime to which it committed itself in an earlier nondiscrimination, reciprocity, enforceable com- multilateral trade negotiation. If the country desires mitments, transparency, and safety valves. to raise trade barriers, it must apply the changed regime to all WTO members. This increases the political cost of backsliding on trade policy because Nondiscrimination importers will object. Finally, MFN reduces negoti- Nondiscrimination has two major components: the ating costs: once a negotiation has been concluded most-favored-nation (MFN) rule, and the national with a country, the results extend to all. Other coun- treatment principle. Both are embedded in the main tries do not need to negotiate to obtain similar WTO rules on goods, services, and intellectual treatment; instead, negotiations can be limited to property, but their precise scope and nature differ principal suppliers. across these three areas. This is especially true of the National treatment ensures that liberalization national treatment principle, which is a specific, not commitments are not offset through the imposition a general commitment when it comes to services. of domestic taxes and similar measures. The The MFN rule requires that a product made in requirement that foreign products be treated no less one member country be treated no less favorably favorably than competing domestically produced than a "like" (very similar) good that originates in products gives foreign suppliers greater certainty any other country. Thus, if the best treatment grant- regarding the regulatory environment in which they ed a trading partner supplying a specific product is must operate. The national treatment principle has a 5 percent tariff, this rate must be applied immedi- often been invoked in dispute settlement cases ately and unconditionally to imports of this good brought to the GATT. It is a very wide-ranging rule: originating in all WTO members. In view of the the obligation applies whether or not a specific tar- 42 The WTO: Functions and Basic Principles iff commitment was made, and it covers taxes and establish "ceiling bindings": the member concerned other policies, which must be applied in a nondis- cannot raise tariffs above bound levels without criminatory fashion to like domestic and foreign negotiating compensation with the principal sup- products. It is also irrelevant whether a policy hurts pliers of the products concerned. The MFN rule an exporter. What matters is the existence of dis- then ensures that such compensation--usually, crimination, not its effects. reductions in other tariffs--extends to all WTO members, raising the cost of reneging. Once tariff commitments are bound, it is impor- Reciprocity tant that there be no resort to other, nontariff, Reciprocity is a fundamental element of the negoti- measures that have the effect of nullifying or ating process. It reflects both a desire to limit the impairing the value of the tariff concession. A num- scope for free-riding that may arise because of the ber of GATT articles attempt to ensure that this MFN rule and a desire to obtain "payment" for does not occur. They include Article VII (customs trade liberalization in the form of better access to valuation), Article XI, which prohibits quantitative foreign markets. As discussed by Finger and Winters restrictions on imports and exports, and the Agree- in Chapter 7 of this volume, a rationale for reciproc- ment on Subsidies and Countervailing Measures, ity can be found in the political-economy literature. which outlaws export subsidies for manufactures The costs of liberalization generally are concentrat- and allows for the countervailing of production ed in specific industries, which often will be well subsidies on imports that materially injure domes- organized and opposed to reductions in protection. tic competitors (see Chapter 17, by Pangestu, in this Benefits, although in the aggregate usually greater volume). than costs, accrue to a much larger set of agents, If a country perceives that actions taken by anoth- who thus do not have a great individual incentive to er government have the effect of nullifying or organize themselves politically. In such a setting, impairing negotiated market access commitments being able to point to reciprocal, sector-specific or the disciplines of the WTO, it may bring this situ- export gains may help to sell the liberalization polit- ation to the attention of the government involved ically. Obtaining a reduction in foreign import bar- and ask that the policy be changed. If satisfaction is riers as a quid pro quo for a reduction in domestic not obtained, the complaining country may invoke trade restrictions gives specific export-oriented WTO dispute settlement procedures, which involve domestic interests that will gain from liberalization the establishment of panels of impartial experts an incentive to support it in domestic political mar- charged with determining whether a contested kets. A related point is that for a nation to negotiate, measure violates the WTO. Because the WTO is an it is necessary that the gain from doing so be greater intergovernmental agreement, private parties do than the gain available from unilateral liberaliza- not have legal standing before the WTO's dispute tion. Reciprocal concessions ensure that such gains settlement body; only governments have the right to will materialize. bring cases. The existence of dispute settlement pro- cedures precludes the use of unilateral retaliation. For small countries, in particular, recourse to a mul- Binding and Enforceable Commitments tilateral body is vital, as unilateral actions would be Liberalization commitments and agreements to ineffective and thus would not be credible. More abide by certain rules of the game have little value if generally, small countries have a great stake in a they cannot be enforced. The nondiscrimination rule-based international system, which reduces the principle, embodied in Articles I (on MFN) and III likelihood of being confronted with bilateral pres- (on national treatment) of the GATT, is important sure from large trading powers to change policies in ensuring that market access commitments are that are not to their liking. implemented and maintained. Other GATT articles play a supporting role, including Article II (on Transparency schedules of concessions). The tariff commitments made by WTO members in a multilateral trade Enforcement of commitments requires access to negotiation and on accession are enumerated in information on the trade regimes that are main- schedules (lists) of concessions. These schedules tained by members. The agreements administered 43 T H E W O R L D T R A D E O R G A N I Z AT I O N by the WTO therefore incorporate mechanisms mitments on trade policies that are subject to bind- designed to facilitate communication between ing dispute settlement, can also have this effect. WTO members on issues. Numerous specialized committees, working parties, working groups, and Safety Valves councils meet regularly in Geneva. These interac- tions allow for the exchange of information and A final principle embodied in the WTO is that, in views and permit potential conflicts to be defused specific circumstances, governments should be able efficiently. to restrict trade. There are three types of provisions Transparency is a basic pillar of the WTO, and it in this connection: (a) articles allowing for the use of is a legal obligation, embedded in Article X of the trade measures to attain noneconomic objectives; (b) GATT and Article III of the GATS. WTO members articles aimed at ensuring "fair competition"; and (c) are required to publish their trade regulations, to provisions permitting intervention in trade for eco- establish and maintain institutions allowing for the nomic reasons. Category (a) includes provisions review of administrative decisions affecting trade, allowing for policies to protect public health or to respond to requests for information by other national security and to protect industries that are members, and to notify changes in trade policies to seriously injured by competition from imports. The the WTO. These internal transparency require- underlying idea in the latter case is that governments ments are supplemented by multilateral surveil- should have the right to step in when competition lance of trade policies by WTO members, becomes so vigorous as to injure domestic competi- facilitated by periodic country-specific reports tors. Although it is not explicitly mentioned in the (trade policy reviews) that are prepared by the sec- relevant WTO agreement, the underlying rationale retariat and discussed by the WTO General Coun- for intervention is that such competition causes cil. (The Trade Policy Review Mechanism is political and social problems associated with the described in Box 6.1.) The external surveillance need for the industry to adjust to changed circum- also fosters transparency, both for citizens of the stances. Measures in category (b) include the right to countries concerned and for trading partners. It impose countervailing duties on imports that have reduces the scope for countries to circumvent their been subsidized and antidumping duties on imports obligations, thereby reducing uncertainty regard- that have been dumped (sold at a price below that ing the prevailing policy stance. charged in the home market). Finally, under category Transparency has a number of important bene- (c) there are provisions allowing actions to be taken fits. It reduces the pressure on the dispute settle- in case of serious balance of payments difficulties or ment system, as measures can be discussed in the if a government desires to support an infant industry. appropriate WTO body. Frequently, such discus- sions can address perceptions by a member that a From GATT to WTO specific policy violates the WTO; many potential disputes are defused in informal meetings in Gene- Over the more than four decades of its existence, the va. Transparency is also vital for ensuring "owner- GATT system expanded to include many more ship" of the WTO as an institution--if citizens do countries. It evolved into a de facto world trade not know what the organization does, its legitimacy organization, but one that was increasingly frag- will be eroded. The trade policy reviews are a mented as "side agreements" or codes were negoti- unique source of information that can be used by ated among subsets of countries. Its fairly complex civil society to assess the implications of the overall and carefully crafted basic legal text was extended or trade policies that are pursued by their govern- modified by numerous supplementary provisions, ments. From an economic perspective, transparency special arrangements, interpretations, waivers, can also help reduce uncertainty related to trade reports by dispute settlement panels, and council policy. Such uncertainty is associated with lower decisions. Some of the major milestones are sum- investment and growth rates and with a shift in marized in Table 6.1. resources toward nontradables (Francois 1997). The GATT's early years were dominated by acces- Mechanisms to improve transparency can help sion negotiations and by a review session in the lower perceptions of risk by reducing uncertainty. mid-1950s that led to modifications to the treaty. WTO membership itself, with the associated com- Starting in the mid-1960s, recurring rounds of mul- 44 The WTO: Functions and Basic Principles B O X 6 . 1 : T R A N S PA R E N C Y: N O T I F I C AT I O N A N D S U R V E I L L A N C E Transparency at both the multilateral (WTO) level enhances communication, thereby strengthening and the national level is essential to ensure owner- the multilateral trading system. Country-specific ship of commitments, reduce uncertainty, and reviews are conducted on a rotational basis, and enforce agreements. Efforts to increase the trans- the frequency of review is a function of a mem- parency of members' trade policies take up a good ber's share in world trade. The four largest play- portion of WTO resources. The WTO requires that ers--the European Union, the United States, all trade laws and regulations be published. Article Japan, and Canada--are subject to review by the X of the GATT, Article III of the GATS, and Article 63 WTO General Council every two years. In princi- of the TRIPS agreement all require that relevant ple, the next 16 largest traders are subject to laws, regulations, judicial decisions, and administra- reviews every four years, and the remaining tive rulings be made public. More than 200 notifi- members are reviewed every six years. A longer cation requirements are embodied in the various periodicity may be established for least-devel- WTO agreements and mandated by ministerial and oped countries. The trade policy review (TPR) for council decisions. The WTO also has important sur- a country is based on a report prepared by the veillance activities, since it has a mandate to period- government concerned and on a report by the ically review the trade policy and foreign trade WTO Trade Policies Review Division. TPRs are regimes of members. The WTO's Trade Policy supplemented by an annual report by the Direc- Review Mechanism (TPRM), established during the tor-General of the WTO that provides an Uruguay Round, builds on a 1979 Understanding overview of developments in the international on Notification, Consultation, Dispute Settlement, trading environment. and Surveillance under which contracting parties By subjecting the trade policies of the largest agreed to conduct a regular and systematic review industrial country markets to regular public peer of developments in the trading system. The objec- review, the TPRM shifts the balance of power in tive of the TPRM is to examine the impact of mem- the WTO ever so slightly in favor of the develop- bers' trade policies and practices on the trading ing countries (Francois 2001). Equally important, system and to contribute to improved adherence the TPRM provides domestic interest groups with to WTO rules through greater transparency. The the information necessary to determine the costs legal compatibility of any particular measure with and benefits of national trade policies. The WTO disciplines is not examined, this being left for reports are not analytical in the sense of deter- members to ascertain. mining the economic effects of various national The TPRM was originally motivated in part by policies--the size of the implied transfers and the concerns stemming from the fact that the only beneficiaries and losers under the prevailing poli- available review of global trade policies at the cies. This task is left to national stakeholders time was produced by the United States (Keesing (think tanks and policy institutes). 1998). The TPRM is an important element of the WTO because it fosters transparency and Sources: Hoekman and Kostecki (2001); Francois (2001). tilateral trade negotiations gradually expanded the There are many similarities between the GATT and scope of the GATT to take in a larger number of the WTO, but the basic principles remain the same. nontariff policies. Until the Uruguay Round, how- The WTO continues to operate by consensus and to ever, no progress was made on agriculture or on tex- be member driven. There were, however, a number of tiles and clothing. The deal that finally allowed these major changes. Most obviously, the coverage of the sectors to be subjected to multilateral disciplines WTO is much wider. A change of great importance is included the establishment of rules for trade in that in contrast to the GATT, the WTO agreement is a services and enforcement of intellectual property "single undertaking"--all its provisions apply to all rights (IPRs), as well as the creation of the WTO. members. Under the GATT there was great flexibility 45 T H E W O R L D T R A D E O R G A N I Z AT I O N for countries to "opt out" of new disciplines, and in intended to strengthen the political guidance of the practice many developing countries did not sign spe- WTO and enhance the prominence and credibility cific agreements on issues such as customs valuation of its rules in domestic political arenas. Article II of or subsidies. This is no longer the case, implying that the Marrakech Agreement that established the WTO the WTO is much more important for developing charges the organization with providing a common countries than the GATT was. Also important were institutional framework for the conduct of trade changes in the area of dispute settlement, which relations among its members in matters to which became much more "automatic" with the adoption agreements and associated legal obligations apply. of a "negative consensus" rule. (All members must Four annexes to the WTO define the substantive oppose the findings in a dispute settlement to block rights and obligations of members. Annex 1 has adoption of reports.) Finally, the secretariat acquired three parts: Annex 1A, Multilateral Agreements on much greater transparency and surveillance func- Trade in Goods, which contains the GATT 1994 (the tions through the creation of the Trade Policy Review GATT 1947 as amended by a large number of Mechanism. understandings and supplementary agreements negotiated in the Uruguay Round); Annex 1B, which contains the GATS; and Annex 1C, the TRIPS Scope, Functions, and Structure of agreement. Annex 2 contains the Understanding on the WTO Rules and Procedures Governing the Settlement of The WTO is headed by a ministerial conference of all Disputes (DSU)--the WTO's common dispute set- members that meets at least once every two years. By tlement mechanism. Annex 3 contains the Trade contrast, under the GATT a decade could pass Policy Review Mechanism (TPRM), an instrument between ministerial meetings. The more frequent for surveillance of members' trade policies. Finally, participation by trade ministers under the WTO was Annex 4, Plurilateral Trade Agreements, consists of Table 6.1 From GATT to WTO: Major Events Date Event 1947 The GATT is drawn up to record the results of tariff negotiations among 23 countries. The agreement enters into force on January 1, 1948. 1948 The GATT provisionally enters into force. Delegations from 56 countries meet in Havana, Cuba, to consider the final draft of the International Trade Organization (ITO) agreement; in March 1948, 53 countries sign the Havana Charter establishing an ITO. 1950 China withdraws from the GATT. The U.S. administration abandons efforts to seek con- gressional ratification of the ITO. 1955 A review session modifies numerous provisions of the GATT. The United States is granted a waiver from GATT disciplines for certain agricultural policies. Japan accedes to the GATT. 1965 Part IV (on trade and development) is added to the GATT, establishing new guidelines for trade policies of and toward developing countries. A Committee on Trade and Develop- ment is created to monitor implementation. 1974 The Agreement Regarding International Trade in Textiles, better known as the Multifibre Arrangement (MFA), enters into force. The MFA restricts export growth in clothing and textiles to 6 percent per year. It is renegotiated in 1977 and 1982 and extended in 1986, 1991, and 1992. 1986 The Uruguay Round is launched in Punta del Este, Uruguay. 1994 In Marrakech, on April 15, ministers sign the final act establishing the WTO and embody- ing the results of the Uruguay Round. 1995 The WTO enters into force on January 1. 1999 Ministerial meeting in Seattle fails to launch a new round. 2001 A new round of trade talks (the Doha Development Agenda) is agreed on in Doha, Qatar. Source: Hoekman and Kostecki (2001). 46 The WTO: Functions and Basic Principles Tokyo Round codes that were not multilateralized ment and between trade and competition policy, in the Uruguay Round and that therefore bind only and with the issue of transparency in government their signatories. Together, Annexes 1 through 3 procurement. Specific committees address matters embody the multilateral trade agreements. Article II relating to the GATS or the TRIPS agreement. All of the WTO specifies that all the agreements con- WTO members may participate in all councils, tained in these three annexes are an integral part of committees, and other bodies, with the exceptions the WTO agreement and are binding on all mem- of the Appellate Body, dispute settlement panels, the bers. All of these instruments are discussed further Textiles Monitoring Body, and committees dealing in this chapter or in other chapters of this volume, with plurilateral agreements. The WTO is charged with facilitating the imple- About 40 councils, committees, subcommittees, mentation and operation of the multilateral trade bodies, and standing groups or working parties agreements, providing a forum for negotiations, functioned under WTO auspices in 2000, more than administering the dispute settlement mechanism, twice the number under the GATT. Such bodies are exercising multilateral surveillance of trade policies, open to all WTO members, but generally only the and cooperating with the World Bank and the IMF more important trading nations (less than half of to achieve greater coherence in global economic the membership) regularly send representatives to policymaking (Art. III WTO). Between meetings of most meetings. The degree of participation reflects the ministerial conference, which is responsible for a mix of national interests and resource constraints. carrying out the functions of the WTO, the organi- The least-developed countries, in particular, tend zation is managed by the General Council, at the not to be represented at these meetings; often, they level of diplomats. The General Council meets do not have delegations based in Geneva. All of about 12 times a year. On average, about 70 percent these fora, plus working parties on accession (aver- of all WTO members take part in its meetings, at aging close to 30 in the late 1990s), dispute settle- which members are usually represented by delega- ment panels, meetings of regional groups, meetings tions based in Geneva. The General Council turns of heads of delegations, and numerous ad hoc and itself, as needed, into a body that adjudicates trade informal groups add up to 1,200 events a year at or disputes (the Dispute Settlement Body, or DSB) or near WTO headquarters in Geneva. Most WTO that reviews members' trade policies (the Trade Pol- business is conducted in English, but many official icy Review Body, or TPRB). WTO meetings require French and Spanish inter- Three subsidiary councils, on goods, on services, pretation. and on intellectual property rights, operate under The main actors in the day-to-day activities are the general guidance of the General Council. Sepa- officials affiliated with the delegations of members. rate committees deal with the interests of develop- The WTO--like the 1947 GATT--is therefore ing countries (Committee on Trade and something of a network organization (Blackhurst Development); surveillance of trade restriction 1998). The WTO secretariat is the hub of a very actions taken for balance of payment purposes; sur- large and dispersed network comprising official veillance of regional trade agreements; trade-envi- representatives of members based in Geneva, civil ronment linkages; and WTO finances and servants based in capitals, and national business and administration. Additional committees or working nongovernmental groups that seek to have their parties deal with matters covered by the GATT, the governments push for their interests at the multilat- GATS, or the TRIPS agreement. There are commit- eral level. The operation of the WTO depends on tees, functioning under the auspices of the Council the collective input of thousands of civil servants on Trade in Goods, on subsidies, antidumping and and government officials who deal with trade issues countervailing measures, technical barriers to trade in each member country. (product standards), import licensing, customs val- Initiatives to launch multilateral trade negotia- uation, market access, agriculture, sanitary and phy- tions and to settle disputes--the two highest-profile tosanitary measures, trade-related investment activities of the WTO--are the sole responsibility of measures, rules of origin, and safeguards. In addi- WTO members themselves, not the secretariat. The tion, working groups have been established to deal member-driven nature of the organization puts a with notifications, with state-trading enterprises, considerable strain on the national delegations of with the relationships between trade and invest- members. Many countries have no more than one 47 T H E W O R L D T R A D E O R G A N I Z AT I O N or two persons dealing with WTO matters; a large and mechanisms have therefore been developed minority has no delegations in Geneva at all. over the years to reduce the number of members that are active participants in WTO deliberations. The first and most important device is to involve Decisionmaking only "principals," at least initially. To some extent Most decisionmaking in the WTO follows GATT this is a natural process; a country that has no agri- practices and is based on consultation and consen- cultural sector is unlikely to be interested in discus- sus. The consensus practice is of value to smaller sions centering on the reduction of agricultural countries, as it enhances their negotiating leverage trade barriers. In general the "Quad" economies-- in the informal consultations and bargaining that Canada, the European Union, Japan, and the United precede decisionmaking, especially if they are able States--are part of any group that forms to discuss to form coalitions. Although recourse to voting may any topic. They are supplemented by countries that be had if a consensus cannot be reached, in practice have a principal supplying interest in a product and voting occurs only very rarely. If a vote is needed, it by the major (potential) importers whose policies is based on the principle of "one member, one are the subject of interest. Finally, a number of vote." Unanimity is required for amendments relat- countries that have established a reputation as ing to general principles such as MFN or national spokespersons tend to be involved in most major treatment. Interpretation of the provisions of the meetings. Historically, such countries have included WTO agreements and decisions on waivers of a Egypt, India, and Yugoslavia. member's obligations require approval by a three- During the Tokyo and Uruguay Rounds, con- quarters majority vote. A two-thirds majority vote is tentious issues as to which deals had to be struck sufficient for amendments relating to issues other were often thrashed out in the "green room," a con- than the general principles mentioned above. ference room adjacent to the Director-General's Where not otherwise specified, and where consen- offices. Green-room meetings were part of a consul- sus cannot be reached, a simple majority vote is, in tative process through which the major countries principle, sufficient. In practice, voting does not and a representative set of developing countries--a occur. Indeed, in 1995 WTO members decided not total of 20 or so delegations--tried to hammer out to apply provisions allowing for a vote in the case of the outlines of acceptable proposals or negotiating accessions and requests for waivers but to continue agendas. Such meetings generally involved the to proceed on the basis of consensus (WT/L/93). active participation and input of the Director-Gen- Legislative amendments are also likely to be quite eral. The convention now is to call such meetings rare, as, in practice, changes to the various agree- green-room gatherings, no matter where they are ments occur as part of broader multilateral rounds. held. The green-room process became a contentious issue during the Seattle ministerial meeting; many developing countries that were excluded from criti- Management of the Secretariat and cal green-room meetings, where attempts were Daily Operations being made to negotiate compromise texts of a draft Unlike the World Bank and the IMF, the WTO does agenda for a new multilateral trade negotiation, felt not have an executive body or a board comprising a that they were not being kept informed of develop- subset of members some of whom represent a num- ments and were not being granted the opportunity ber of countries. Such executive boards facilitate to defend their views. Proposals have been made decisionmaking by concentrating discussions with- periodically to formalize the green-room process by in a smaller but representative group of members. creating an executive committee to manage the The closest the GATT ever came to such a forum WTO agenda, based on shares in world trade was the Consultative Group of Eighteen (CG18), (Schott and Buurman 1994). To date, no progress in established in 1975. It ceased meeting in 1985 and this direction has proved possible in the WTO. never substituted for the GATT Council of Repre- sentatives (Blackhurst 1998). Conclusion As of January 1, 2002, the WTO had a member- ship of 144. Achieving consensus among such a The Uruguay Round and the establishment of the large number of members is not a simple matter, WTO changed the character of the trading system. 48 The WTO: Functions and Basic Principles The GATT was very much a market access­oriented customs valuation, a good case can be made that institution: its function was to harness the dynamics implementation should be linked to national capac- of reciprocity for the global good. Negotiators could ity and international assistance (Hoekman 2002). be left to follow mercantilist logic, and the end A lesson from post­Uruguay Round experience result would be beneficial to all contracting parties. and thinking is that trade policy should be made This dynamic worked less well for developing coun- more central to the development process and devel- tries, where the burden of liberalization rested opment strategies. This needs to be done at both the much more heavily on the shoulders of govern- national and international levels. At the national ments. Even if they wanted to, their scope to use the level it is necessary in order to ensure that govern- GATT was often limited because exporters had ments have a basis on which to resist efforts to fewer incentives and were less powerful than in negotiate agreements in an area. Governments must industrial countries. The reciprocal, negotiation- be able to identify what types of rules will promote driven dynamic also worked much less well for development and what types would lead to an inap- issues that were "lumpy" and where the terms of the propriate use of scarce resources. At the interna- debate revolved around what rules to adopt, not tional level such a change is necessary in order to around how much of a marginal change was appro- enhance the communication between trade and priate. Once discussions center on rules, especially development assistance bodies in member coun- on disciplines for domestic policy and regulations, tries. One reason for the implementation assistance it is more difficult to define intraissue compromises problems that were encountered in the late 1990s that make economic sense. Cross-issue linkage was that the best-endeavors commitments on assis- becomes necessary. Disengagement was not an tance that were made by industrial country trade option during the Uruguay Round (because of the negotiators were not "owned" by counterpart agen- "single undertaking"), so the task was to come up cies in their governments that controlled develop- with a balanced package that ensured gains for all ment assistance money. Progress on both fronts players. One can argue whether the package that would do much to ensure that future negotiations emerged from the round was a balanced one; views do not give rise to problems of the type that were on this point differ widely. created in the Uruguay Round. Whatever the conclusion, it is clear that the approach taken toward ensuring and supporting Notes implementation of WTO agreements by developing This chapter draws on Hoekman and Kostecki (2001). countries was not an effective one. Limiting recog- nition of this problem to the setting of uniform 1 The founding parties to the GATT (giving the names used at transition periods was clearly inadequate. The case the time) were Australia, Belgium, Brazil, Burma, Canada, Cey- lon, Chile, China, Cuba, Czechoslovakia, France, India, for uniform application of agreements that involve Lebanon, Luxembourg, the Netherlands, New Zealand, Nor- reducing trade barriers--tariffs and nontariff barri- way, Pakistan, Southern Rhodesia, Syria, South Africa, the Unit- ers--is very strong. But in other areas requiring ed Kingdom, and the United States. Subsequently, China, minimum levels of institutional capacity, such as Lebanon, and Syria withdrew. 49 7 J . M I C H A E L F I N G E R L . A L A N W I N T E R S Reciprocity in eralization by developing coun- tries, and an "apples versus oranges" problem that arises the WTO because the WTO spans both border trade restrictions (tar- iffs, quotas, and the like) and within-border regulatory struc- Reciprocity: Mutual or correspondent concession of advantages or tures such as standards and privileges, as forming a basis for the commercial relations intellectual property. Failure to between two countries. recognize the apples versus --The Oxford English Dictionary1 oranges problem, we argue, has led to a troublesome Uruguay Round outcome. R eciprocity has been a motivating Reciprocity in GATT Rules principle of the GATT/WTO system. Although the economics of import restrictions recog- The thrust of the GATT/WTO system is that agree- nizes that the losses from a country's own restric- ment defines reciprocity (or balance), not the other tions exceed domestic gains, the politics has not way around. An agreed outcome from a negotiating found a way to enfranchise the domestic interests round, the system presumes, is an outcome that each that bear these domestic losses--users and con- member considers advantageous, by whatever stan- sumers of imports. When trade policy involves an dard the member chooses to apply. Beyond that, vari- exchange of domestic restrictions for foreign restric- ous provisions for adjustment, such as renegotiation tions, this amplifies the voice of export interests. The and safeguard actions,attempt to maintain the balance success of the GATT/WTO system manifests the that the agreement has established. The same holds for ingenuity of reciprocally agreed liberalization as a dispute settlement. In this section we look at how reci- means of transferring political power over domestic procity enters into each of these parts of the system. import restrictions to export interests, and it also manifests the power of these interests. Negotiations In this chapter we explore the role of reciprocity in GATT/WTO negotiations and in the processes of Reciprocity serves to motivate negotiations. Partici- making adjustments and settling disputes under or pants and commentors use reciprocity, or its func- within an agreement. We look at the role of reci- tional equivalent, "balance," as a standard against procity in past agreements, and we present evi- which to evaluate an outcome. The rules, however, dence suggesting that reciprocity is not the only do not define that standard; determining the stan- force that shapes the outcome of a negotiation. We dard is part of the evaluation itself.2 then turn to two issues that relate to reciprocity: The GATT, and the Marrakech Agreement that "credit" in reciprocal negotiations for unilateral lib- established the WTO, refer in their preambles to 50 Reciprocity in the WTO "entering into reciprocal and mutually advanta- my question of what it means in negotiating prac- geous arrangements directed to the substantial tice--what countries have interpreted as equivalent reduction of tariffs and other barriers to trade." concessions, and what they have not. We take up GATT Article XXVIII bis, the article that provides this topic below. for negotiations to be held, refers also to "negotia- tions on a reciprocal and mutually advantageous Treatment of Developing Countries in Negotia- basis." Neither the GATT nor the WTO provides tions. Part IV of the GATT provides elaborate com- further specification of what is "reciprocal" or of mitments to the developing countries. For what is "mutually advantageous." The logic of the example, Article XXXVI.8 states, "The developed GATT/WTO is that in the negotiations each mem- contracting parties do not expect reciprocity for ber is sovereign to determine for itself whether a commitments made by them in trade negotiations proposed agreement is to its advantage--to decide to reduce or remove tariff and other barriers to the the criteria by which to identify the pluses and trade of less-developed contracting parties." The minuses, and to apply those criteria by whatever commitments of Part IV, however, are not legally formula the member considers appropriate. The binding. Exhortations such as the one quoted are GATT's tradition of decision by consensus rein- qualified by other phrases: for example,"The devel- forces the idea that an agreement is an outcome that oped countries shall to the fullest extent possible-- each member considers to be to its benefit. If any that is, except where compelling reasons, including one member does not find the outcome advanta- legal reasons, make it impossible . . ."(Art. geous, the proposed agreement does not go into XXXVII.1), and "The adoption of measures to give effect. effect to these principles and objectives shall be a matter of conscious and purposeful effort on the GATT Elaboration on Reciprocity in Negotiations. part of the contracting parties both individually An early (1955) GATT working party, in response to and jointly" (Art. XXXVI.9). The operational a proposal to establish rules for how concessions meaning of such phrases is to make clear that they should be measured, concluded that "governments are not legal commitments. The commitment is to participating in negotiations should retain com- a nonmeasurable "conscious and purposeful plete freedom to adopt any method they might feel effort," not to a measurable result. most appropriate for estimating the value of duty Although such statements do not express legal reductions and bindings . . ." The working party obligations, they do have behind them the weight of went on to note that "there was nothing in the moral suasion; they are intended to influence behav- Agreement, . . . to prevent governments from adopt- ior without going so far as to regulate it. This moral ing any formula they might choose, and therefore suasion has not delivered much. For example, the considered that there was no need for the Contract- Trade-Related Aspects of Intellectual Property ing parties to make any recommendation in this Rights (TRIPS) agreement, the customs valuation matter" (GATT 1994a: 912­13). Similarly, Arthur agreement, the Sanitary and Phytosanitary (SPS) Dunkel, Director-General of the GATT from 1980 agreement, and several other Uruguay Round agree- to 1992, observed, "Reciprocity cannot be deter- ments suggested that industrial country members mined exactly; it can only be agreed upon" (GATT furnish technical assistance to developing country Press Release 1312, March 5, 1982). members that request it. This provision, however, is Since the GATT and the Marrakech Agreement not a binding commitment; the developing coun- are silent on how a member might measure the tries undertook to implement bound commitments advantage it draws from the agreement, they say in exchange for unbound commitments for assis- nothing about how much one country should gain tance. Although developing countries pressed hard from the negotiations relative to another. The word at the WTO for delivery on such promises, bilateral- "balance" does not appear in the GATT/WTO text ly or through an increased WTO technical budget, on negotiations. An agreement (the outcome of a the high-income countries have done little. The negotiating round) defines balance, not the other stalemate has prompted Rubens Ricupero (2000) to way around. Although the GATT/WTO rules make suggest that in the future, negotiations on topics no demands as to what reciprocity means in a nego- that will involve expensive implementation be tiation, there remains the practical political-econo- accompanied by an "implementation audit" that 51 T H E W O R L D T R A D E O R G A N I Z AT I O N will identify concretely what developing countries involved some degree of objectivity. In the end, will have to do and what it will cost. Short of a however, appropriate compensation or retaliation is bound commitment from the high-income coun- what the parties agree on, not what an objective and tries to meet such costs, statements about imple- exogenous standard dictates. mentation assistance should be omitted. There should be no more instances of creating the rhetoric Safeguard Actions (only) of reciprocity by exchanging bound commit- ments for unbound promises. GATT Article XIX, the "escape clause," or "safe- guard" article, includes a similar provision. (The article, roughly speaking, allows a country to Renegotiations restrict imports that cause injury to domestic pro- Political reality will require from time to time that ducers.) Implicitly, the article calls on the country adjustments be made in the outcome of an agree- that takes safeguard action to provide compensa- ment, and domestic politics will demand that some tion. Explicitly it provides that exporting countries "concession" be withdrawn.3 The GATT article on may retaliate if satisfactory compensation is not renegotiations states, "In such negotiations and offered: "If agreement among the interested con- agreement, which may include provision for com- tracting parties with respect to the action is not pensatory adjustment with respect to other prod- reached, . . . the affected contracting parties shall ucts, the contracting parties concerned shall then be free . . . to suspend . . . the application to the endeavor to maintain a general level of reciprocal trade of the contracting party taking such action . . . and mutually advantageous concessions not less of substantially equivalent concessions or other favorable to trade than that provided for in this obligations under this agreement the suspension of Agreement prior to such negotiations" (Art. XXVI- which the Contracting Parties do not disapprove II. 2). If the importing country wanting to raise a . . ." (GATT Art. XIX.3[a]).4 tariff above a previously bound level does not reach The Uruguay Round Agreement on Safeguards agreement with supplying countries on appropriate mentions compensation explicitly, and, in its Article compensation, the exporting countries will, in time, 8.1, it includes the exhortation in GATT Article be entitled to retaliate-- "to withdraw . . . substan- XXVIII (renegotiations) "to maintain a substantial- tially equivalent concessions" (Art. XXVIII.3a, 3b, ly equivalent level of concessions and other obliga- 4d, 5). Adjustments of concessions should maintain tions." Retaliation, as in the GATT safeguard article, the balance that the previous agreement established. is the suspension of the application of "substantially In practice, many renegotiations have eventually equivalent concessions or other obligations . . ." been taken up as part of a next round of negotia- (Art. 8.2). In practice, the determination of what is tions, and it is not possible to identify in these "substantially equivalent" has been determined instances whether the compensation that the parties strictly by negotiation among the interested parties. agreed to was appropriate. In other instances, deter- The GATT Contracting Parties have never disap- mining what is a "substantially equivalent conces- proved of a countermeasure to an action under sion" has centered on finding an equivalent amount Article XIX (GATT 1994a: 490). of trade and an equivalent change in the degree of protection. One of the less complex--but still not Dispute Settlement simple--parts of the process has been to agree on an appropriate base period in which to measure the Although compensation and retaliation are part of amounts of trade involved. Other parts were more the vocabulary of GATT/WTO dispute settlement, difficult; for example, often what was at issue was the process primarily has to do with maintaining not simple tariffs but more complicated tariff quo- behavior within an agreement rather than adjusting tas. Many renegotiations have stemmed from cre- what was agreed.5 The GATT text that is relevant to ation of a customs union, and here the task is to compensation and retaliation reads: "If the Con- compensate for discrimination, not merely for tracting Parties consider that the circumstances are change in a tariff rate. serious enough to justify such action, they may Maintaining balance--determining what is an authorize a contracting party or parties to suspend equivalent adjustment to an agreed outcome--has the application to any other contracting party or 52 Reciprocity in the WTO parties of such concessions or other obligations Control of Free-Riding under this Agreement as they determine to be appro- priate in the circumstances" (GATT Art. XIX.2; The initial GATT rounds consisted of bilateral emphasis added). negotiations on modification of most-favored- Robert Hudec (1978) has explained that the nego- nation (MFN) schedules, conducted among a limit- tiators of the proposed International Trade Organi- ed number of countries. (For example, in the 1947 zation (ITO) and the GATT were ambiguous as to round the United States negotiated with 16 coun- whether dispute settlement concerned compensa- tries that supplied about two-thirds of U.S. tion or compliance. In the 48 years of the GATT, imports.) Efforts were made in these negotiations to there was only one instance in which the dispute limit concessions to products imported in large part settlement mechanism authorized retaliation. In the from other participants. Table 7.1, a tabulation of WTO's first six years, the WTO dispute settlement U.S. experience in early rounds, shows, for example, mechanism has twice authorized retaliation: in the that at the Dillon Round 96 percent of U.S. tariff Bananas case and the Beef Hormone case. As of late cuts--all made on an MFN basis--were on imports 2001, both were still in political dispute. from countries that made concessions in return. At that time, 66 percent of U.S. imports came from these countries. The difference between the 96 per- Reciprocity and Other Influences on an cent and the 66 percent reflects the emphasis on Agreement limiting concessions to products imported almost One can find, in the results of negotiations, clear entirely from countries that reciprocated. Attention evidence of the influence of reciprocity. One can to internalizing the concessions (that is, to limiting also find evidence that more is at play than reaching free-riding) led, however, to low coverage of the tar- a mercantilist balance of concessions received ver- iff cuts--for the United States, the reductions cov- sus concessions given. In this section we review ered 15 percent of dutiable imports in the 1956 other influences that are likely to shape a negotia- round and 20 percent in the 1960­61 round. tion. We also present scattered evidence of the influ- At the Kennedy Round, bilateral bargaining over ence of reciprocity and of other factors. tariff cuts was replaced by formula cuts. The shift to Table 7.1 Control of Free-Riding in GATT Negotiations: U.S. Experience, 1947­67 Kennedy Round, 1964­67 Geneva Annecy Torquay Geneva Dillon Major All Round, Round, Round, Round, Round, partici- partici- 1947 1949 1951 1956 1960­61 pantsa pants Percentage of dutiable imports from all countries subject to cuts 35 37 26 15 20 -- 44 Percentage of dutiable imports coming from participants 65 6 34 67 66 68 72 Percentage of dutiable imports subject to cuts coming from participants 84 39 64 89 96 81 91 --Not available. a. Austria, Canada, Denmark, European Economic Community, Finland, Japan, Sweden, Switzerland, and the United Kingdom. Source: Finger (1979): 424­25. 53 T H E W O R L D T R A D E O R G A N I Z AT I O N a formula approach did lead to broader reduc- although public works and other benefits have also tions--U.S. cuts applied to 44 percent of imports. been used.6 Exclusion of free riders took the form of negotiation Another way in which governments attempt to over "exclusion lists" and, as the figures show, did minimize the problem of compensating losers is by limit spillover to free riders to only 9 percent of con- taking advantage of the large volume of intraindus- cession imports. try trade that characterizes the modern trading sys- tem. To the extent that concessions given by an industry can be offset by concessions received on Fewer Concessions Given, Fewer Received the products exported by the same industry, the Table 7.2 provides another indication that to receive government need not develop interindustry mecha- concessions, a country has to give concessions. The nisms for balancing losers against winners. Gilbert message is obvious: the lower the degree of partici- Winham (1986: 65) notes that from the Kennedy pation in the negotiations, the lower the share of Round forward, there has evolved a tendency to exports affected by the concessions of other partici- look for such "self-balancing sectors." In regional pants. agreements such balancing can be increased through the use of rules of origin. An example is the way in which the North American Free Trade Agree- Domestic Reciprocity ment (NAFTA) and other regional agreements in At the same time, there is more to the liberalization which the United States participates condition process than concessions given over the interna- access to the U.S. market for textile products on the tional table versus concessions received. The bar- use of U.S.-made fibers or fabrics. gaining process ties access to foreign markets to the granting of access to the domestic market and Noneconomic Objectives thereby mobilizes export interests to favor import liberalization. But the domestic politics of setting War, according to Clausewitz, is the pursuit of gains for export industries against losses for diplomacy through other means. Often, so is trade import-competing industries is not frictionless. For policy. Freedom of international commerce was the a government motivated toward trade liberaliza- third of U.S. President Woodrow Wilson's Fourteen tion, the tough tradeoffs are not between it and for- Points. To Cordell Hull, secretary of state for Presi- eign governments but between domestic winners dent Franklin D. Roosevelt, the link was straightfor- and losers. ward: "Unhampered trade dovetailed with peace; Overcoming such frictions has been, in practice, high tariffs, trade barriers and unfair economic partly a matter of power--in the simple case in competition with war" (Hull 1948: 81). After World which negotiating authority must be specifically War II, leadership in Europe and in the United granted, using export industries to win more con- States saw economic union in Europe and the con- gressional votes than the opposition can rally. It has struction of an open global trading system more as also been partly a matter of compensation. Adjust- strategic objectives than as economic ones. A gov- ment assistance is the straightforward example, ernment that can mobilize noneconomic motives Table 7.2 U.S. Imports Covered by Kennedy Round Tariff Concessions (Reductions plus Bindings) as a Share of Total U.S. Imports from the Country Group (percent) Country group Share (percent) Major participants 70 Other industrial country participants 49 Active developing country participants 33 Other developing countries 5 Note: Data are for 1994 imports. Source: Finger (1979): 435. 54 Reciprocity in the WTO into significant support for trade liberalization will and intellectual property rights unless high-income be in a position to play a hegemonic role (as the countries negotiated on agriculture and on textiles United States did in early GATT rounds) and to and clothing. make larger concessions than it receives in exchange. Equal Sacrifice, Softly Applied When, however, the time came for totting up, the Locking In equal sacrifice concept was the one that delegations Developing countries that have unilaterally liberal- used.7 Particularly in the last month of the negotia- ized sometimes view binding such liberalization tions (the mopping-up phase) the negotiators internationally as a defense against the risk of back- devoted significant attention to ensuring that each sliding should political authority shift or popular participating country had made an appropriate support wane. Such an objective, like a noneconom- contribution to the tariff reduction exercise. Dele- ic one, can motivate a government to accept what a gations widely but informally accepted that the tar- calculation based strictly on considerations of mer- gets were average reductions of one-third for cantilist market access would view as a bad bargain. industrial countries and one-fourth for developing countries.8 The negotiating guidelines lacked preci- sion; for example, was the one-third cut to be a Individual Sacrifice for the Common Good weighted or an unweighted average? Over all prod- Many of the participants in the initial ITO and ucts, or only over dutiable items? GATT negotiations viewed their task as the con- Tied up with achievement of these targets was the struction of a system from which all countries would question of how countries would receive "credit" for derive significant noneconomic benefits (perhaps unilateral tariff reductions and for extensions of economic gains as well, but the emphasis was usual- bindings that did not imply tariff cuts. The agricul- ly on noneconomic considerations). This view of tural negotiations established formal negotiating reciprocity differs from the mercantilist bargaining guidelines, not only on the amounts by which model in that the benefits a participant gets are not import restrictions and other agricultural supports unequivocally identified with the particular market were to be reduced by each country but also for how access concessions that the country receives; the link agricultural nontariff barriers were to be converted between contribution and benefit is amorphous, to tariffs. resulting from the collective nature of the system These percentages were negotiating guidelines, rather than from any particular element of it. not bound commitments. Even in agriculture, Robert E. Hudec, in his 1987 book Developing where the negotiating guidelines were circulated as Countries in the GATT Legal System, builds on the a GATT document (GATT 1993b), legal commit- common-good view to provide a convincing inter- ments were the rates notified on each country's pretation of how GATT members came to accept schedule. The GATT/WTO members seem to have "special and differential treatment" as the appropri- policed these guidelines rather softly. Interviews ate attitude toward developing countries. In con- with more than a dozen delegations after the round structing any system from the contributions of its found none that had attempted to calculate the members, Hudec notes, it is difficult to ask the less depth of cut by each country, or even for major well off to contribute proportionally with the bet- trading partners. Likewise, no delegation had tabu- ter-off members. lated concessions received--that is, the coverage of its exports by concessions scheduled by other coun- tries. A number of developing country delegations The Uruguay Round Tariff Reductions pointed to the agriculture and the textiles and cloth- Scorecard: What We Learn from It ing agreements as evidence that they had paid atten- Reciprocity--"get what you pay for," or, more tion to what they would receive, but in agriculture, aggressively, "pay for what you get!"--was clearly too, although the guidelines were more precise, the motivating principle of the Uruguay Round either the numbers were checked only casually or negotiations. For example, developing countries disregard of the guidelines was widely accepted. would not negotiate in "new areas" such as services After-the-fact examination has turned up a lot of 55 T H E W O R L D T R A D E O R G A N I Z AT I O N "dirty tariffication"--tariff rates considerably in ever established for what any country should "give" excess of those that the guideline formulas generate or could expect to "receive." It is hence impossible to (Hathaway and Ingco 1996). measure the extent to which credit for unilateral lib- The interviews revealed that as the negotiations eralization was given. were being completed, selling the agreement at On the question of how to take into account home--that is, gaining approval--was an impor- bindings that did not imply tariff cuts, such as ceil- tant consideration. The issue was not the overall ing bindings, not even an unofficial approach balance of concessions; it was to make sure that evolved. Toward the end of 1990 the Mexican dele- powerful domestic constituencies were accommo- gation circulated a nonpaper arguing that credit dated. The focus was on the effect on the big trees, should be given for expansion of the scope of bind- not the forest. ings, but it did not offer a method for measuring the "tariff cut equivalent."11 Later, the chair of the GATT Market Access Group provided guidelines for Concessions Given versus Concessions Received: such measurement, including a matrix of suggested Great Differences among Countries equivalents between depth of tariff cut and scope of "Concessions given" is a familiar concept. Its com- expansion of bindings. The view of the negotiators plement, "concessions received," refers to the con- with whom we spoke was that there never emerged cessions made by trading partners that apply to a even notional agreement on how to convert exten- given country's exports.9 In Table 7.3, columns (1) sion of bindings into a tariff cut equivalent. through (3) show the depth of tariff reductions, and Giving credit for unilateral liberalization by columns (4) through (6) attempt to introduce the developing countries is part of the standard scope as well as the depth of tariff changes. pro­developing country list of what a negotiation It is hard to find equal sacrifice in this subpart of the should do. But the fact that what is an "appropriate" Uruguay Round outcome. If all the reductions were outcome is defined by the agreement and not by an equal, column (2) would show identical numbers for exogenous standard means that calls for credit for each country, as would column (5). It is also hard to unilateral liberalization are an exercise in moral find mercantilist balance. If such balance had been suasion, not an application of economic or account- achieved--that is, if each country received conces- ing science. Finger, Reincke, and Castro (2002: table sions roughly in line with the concessions it gave--all 2) found that the suasion did have an effect. Bind- the numbers in column (6) would be zero, which is ings of unilateral tariff cuts (but not unbound uni- clearly not the case. The summary statistic in the last lateral cuts) do seem to have been counted toward row reports that, on average, a country's imbalance developing countries' fulfillment of their "obliga- (positive or negative) was over half as large as the tion" to reduce tariffs by one-fourth. value of the concessions the country received.10 In sum, calls for credit for unilateral liberaliza- tion--where that liberalization has then been bound under the GATT/WTO--have been effective. Credit for Unilateral Liberalization Calling for a "credit rule," however, reveals a misun- Many developing countries undertook unilateral derstanding of how the GATT/WTO works. liberalization in the 1980s and 1990s. Credit in the reciprocal negotiations for this liberalization was Apples versus Oranges part of the informal guidelines for meeting the equal sacrifice criterion. From what delegations told The "grand bargain," as Sylvia Ostry (2000) has us in interviews, informal practice was more or less labeled it, that was struck at the Uruguay Round was to calculate tariff cuts from the rate prevailing in that the developing countries would take on signifi- 1988 to the rate bound at the Uruguay Round. cant commitments in "new areas" such as intellectu- Developing countries were given credit for unilater- al property and services, where industrial country al liberalization by allowing them to count from the enterprises saw opportunities for expanding inter- rates applied earlier, at the beginning of the 1980s. national sales. The industrial countries, in In any case, no tabulation of country-by-country exchange, would open up in areas of particular tariff cuts was made, either by the GATT secretariat export interest to developing countries: agriculture, or by individual countries, and no formal target was and textiles and clothing. 56 Reciprocity in the WTO Table 7.3 Tariff Concessions Received and Given at the Uruguay Round Mercantilist balance Percentage tariff reductiona (percentage point dollars)b (1) ­ (2) as (4) ­ (5) as percentage Concessions Concessions percentage Received Given of (1) received given of (4) Economy (1) (2) (3) (4) (5) (6) High-income Australia 0.76 3.35 ­341 21,032 88,162 ­319 Austria 2.64 3.74 ­42 74,602 108,820 ­46 Canada 0.22 0.89 ­305 5,291 26,205 ­395 European Union 1.94 2.19 ­13 578,816 627,939 ­8 Finland 3.47 2.52 27 63,924 44,021 31 Hong Kong (China) 2.36 0.00 100 60,258 0 100 Iceland 1.59 0.20 87 2,151 299 86 Japan 2.06 1.06 49 481,006 143,142 70 New Zealand 0.84 0.83 1 5,126 4,155 19 Norway 1.15 2.17 ­89 24,250 44,263 ­83 Singapore 1.96 0.85 57 50,294 32,741 35 Switzerland 2.15 0.89 59 100,659 46,829 53 United States 1.21 1.07 12 214,791 283,580 ­32 Transition Czech and Slovak Union 2.06 1.05 49 9,773 7,312 25 Hungary 1.82 1.69 7 7,755 13,727 ­77 Poland 1.36 1.26 7 8,609 7,112 17 Developing Argentina 0.98 0.00 100 6,331 0 100 Brazil 1.37 0.00 100 38,037 98 100 Chile 0.50 0.00 100 3,291 0 100 Colombia 1.25 0.02 98 6,323 81 99 India 1.22 6.16 ­405 14,380 67,172 ­367 Indonesia 0.87 0.25 71 16,222 3,355 79 Korea, Rep. of 1.87 5.99 ­220 100,809 262,918 ­161 Malaysia 1.46 1.97 ­35 36,108 28,966 20 Mexico 0.16 0.00 100 960 3 100 Peru 0.57 0.03 95 1,586 58 96 Philippines 2.43 1.29 47 19,748 12,847 35 Sri Lanka 1.36 0.01 99 1,595 33 98 Thailand 1.33 5.93 ­346 20,564 95,953 ­367 Tunisia 1.42 0.02 99 2,506 72 97 Turkey 1.72 3.00 ­74 12,557 32,661 ­160 Uruguay 0.52 0.00 100 772 6 99 Venezuela, R. B. de 0.21 0.13 38 2,051 806 61 Summary Sum of absolute differences/ Sum of absolute differences/ statistics sum of received = 86 percent sum of received = 58 percent a. Weighted average of change measured as dT/(1 + Tavg) * 100, where Tavg is the average of the before- and after-change rates, calculated across all tariff lines, including those on which there was no reduction. Why this formula? Whereas cutting by half a tariff of 2 percent saves the importer only 1 cent, cutting by half a tariff of 50 percent saves the importer 25 cents. As a part of what the importer pays, the tariff reduction relates to the tariff charge plus the price received by the seller--to Ps(1 + T) rather than simply to T. Finger, Ingco, and Reincke (1996) provide a more detailed explanation. b. Tariff cut as measured in column (1) or (2) multiplied by the value (in millions of dollars) of the imports or exports to which the import- ing country applies MFN tariff rates. A percentage point dollar is a 1 percent tariff change on 1 dollar of exports or imports. 57 T H E W O R L D T R A D E O R G A N I Z AT I O N What the North gave in this exchange was tradi- The problems that developing countries face in tional market access, reduction of import restric- the new areas largely have to do with project design tions, and in agriculture, reduction of export and cost-benefit analysis--with development eco- subsidies and production subsidies. What the South nomics, not market access. The World Bank and the gave in the new areas was different. WTO obliga- GATT/WTO are different institutions that work in tions on services, on intellectual property rights, different ways. These differences are not arbitrary; and on standards basically have to do with the they reflect what the international community structure of the domestic economy. The industrial knows about how to deal with trade issues versus countries that wanted these areas to be covered by how to deal with development issues (see Finger the WTO rationalized their inclusion by reference and Nogués 2001). to their "trade-related" attributes (although the actual motivation was the trade interests of their The Uruguay Round Scorecard enterprises). Whatever the fig leaf, regulation here is, figuratively speaking, nine-tenths concerned with The outcome of the Uruguay Round was a good one the domestic economy and one-tenth about trade. for the North. Not only did the industrial countries The two sides of the grand bargain have funda- gain from the concessions they received; the eco- mentally different economics. In real economics, nomics of the concessions they gave was also posi- giving away an import restriction is not a cost; it is tive, through the opening up of their own something that enhances the national economic agriculture and textiles sectors. And for the South? interest. GATT bargaining is a response to the diffi- On the gain dimension--market access--develop- cult politics of liberalization, not to the good sense ing countries did not achieve a mercantilist surplus of its economics. The economics of new area (Table 7.4). Their tariff reductions covered as large a responsibilities are different in two respects (see also share of their imports as did those of the industrial Chapter 48, by Finger and Schuler): countries, and their tariff cuts, measured by how these reductions will affect importers' costs, were · Implementing such responsibilities will cost deeper than those of the industrial countries. This is money--for example, for laboratories to develop true even when we take into account the tariff and enforce standards.12 equivalent of the Multifibre Arrangement (MFA) · The result can be a substantial net cost rather quotas that the industrial countries have committed than a benefit. For many developing countries the themselves to remove. economics of TRIPS is the same as the economics For developing countries as well as industrial to oil importers of oil price increases. Just the countries, concessions made in the grand bargain patent changes required by TRIPS will cost some make for difficult domestic politics. For developing countries more than they gain from the whole of countries, these concessions will also mean real eco- the market access liberalization package (see Fin- nomic costs. The scorecard on the Uruguay Round ger 2001). grand bargain? The South's concessions in the new Table 7.4 Uruguay Round Tariff Concessions, All Merchandise Industrial Developing economies economies Percentage Depth of Percentage Depth of of imports cuta of imports cuta Includes tariffication and bound reductions on agricultural products 30 1.0 29 2.3 Includes the above plus the tariff equivalent of elimination of the Multifibre Arrangement 30 1.6 29 2.3 a. Depth of cut, dT/(1 + T), is a weighted average across all products, including those on which no reduction was made. Source: Finger and Schuknecht (2001): table T-1, based on Finger, Ingco, and Reincke (1996). 58 Reciprocity in the WTO areas are, as mercantilism, unrequited--and as real 5 The WTO Dispute Settlement Understanding (Art. 3.7) establish- economics, they are costly. es explicit priorities among different outcomes: "The aim of the dispute settlement mechanism is to obtain a positive solution to a dispute; A solution mutually acceptable to the parties to a dis- Conclusions pute and consistent with the covered agreements is clearly to be preferred; Withdrawal of the measures concerned if they are Reciprocity in negotiations is a motivation and an found to be inconsistent with . . . the covered agreements; objective, not a criterion. Within an agreement, rec- Compensation . . . only if the immediate withdrawal of the pro- iprocity--better known in this context by its other vision is impractical . . . a temporary measure pending the with- drawal of the measure . . . inconsistent with the covered name, balance--comes closer to having an opera- agreements; and, the last resort . . . suspending . . . concessions tional meaning. Still, it is to a large extent some- or other obligations . . ." thing that can be agreed on but not measured. 6 Zeiler (1992) provides examples of the trades U.S. President "Credit" is moral suasion--a useful rallying cry for John F. Kennedy made to win congressional approval of the driving a better deal for the South. It is, however, authority to negotiate in what came to be called the Kennedy futile, and a basic misunderstanding of the Round. Providing quota protection for the textile industry was GATT/WTO, to think that credit can be converted one; another was an extensive waterways project for the state into the "shall" language of obligation. of Oklahoma. Mistaking clean clothes for dirty was acceptable 7 Provided that each country's trade is balanced (that is, exports when that was what both sides brought to the table. equal imports) and that there are uniform cuts with complete Each took home what its politics saw as the other's coverage (of goods and countries), "equal sacrifice" comes to dirty laundry. In economics, each was doubly better the same arithmetic as "get what you pay for." off; trading market access "concessions" was good 8 These guidelines were cited by many of the Geneva delegations economics for the giver as well as for the receiver. that were interviewed by Finger and colleagues as part of the Bringing in the new areas changed things. What the research published as Finger, Reincke, and Castro (2002). As to the origin of the figures, at the July 7­9, 1993, G-7 summit in developing countries are now asked to put on the Tokyo, the "Quad" (Canada, EU, Japan, and the United States) table can have domestic economic costs as well as trade ministers announced a substantial market access agree- domestic political costs. Reciprocal bargaining over ment, as well as their goals for what they hoped to achieve the political dimension may not be enough. overall: reductions to zero for selected products or harmoniza- Progress in these areas may require management of tion at low levels; a 50 percent cut in tariffs 15 percent and above; and, for other tariffs, a negotiated reduction of at least their economic dimensions, as well. one-third. The one-third reduction for industrial countries may thus have come from this agreement; we have not identified Notes the origin of the one-fourth target for developing countries. 9 The formula dT/(1 + T), not dT/T, was used to measure tariff 1 According to the Oxford English Dictionary, the first recorded change. Concessions given for a country are the familiar sum use of "reciprocity" in this sense can be found in the Prelimi- for the country of all its MFN tariff cuts across all tariff lines, nary Articles for the Peace between the United States and weighted by imports. For concessions received, if Dij is the Great Britain, in 1782. reduction of the MFN tariff rate of country i on tariff line (that 2 Finger, Hall, and Nelson (1982) sort decision processes into is, "product") j, and Wijk is the share or weight (by value) of country k's total exports of product j to country i, then the "political" versus "technical" ones. At the technical end of the "reduction received" (column 1 of Table 7.3) by country k is spectrum, the criteria are given, and the decision turns on whether or not the criteria are met. Examples are an the sum, across countries and across products, of Dij multiplied antidumping determination or a jury's decision in a court trial. by Wijk. To calculate percentage point dollars of concessions At the other end of the spectrum, a "political" decision received, for Wijk in that formula we substitute Vijk, the value of k's exports of product i to country j. The countries in the table involves debate over what the criteria are, as well as about are those for which data were available from the WTO Inte- whether the criteria are met. A legislative decision on tax grated Database. See Finger, Ingco, and Reincke (1996) for a reform might be an example. In this framework, reciprocity in description of the database and a more detailed explanation of a negotiation is a political concept. What it means operational- the calculations. ly is not specified by the rules of the negotiation. 10 For some of these countries, counting only the concessions 3 Finger (1998) discusses in more detail GATT/WTO safeguards made at the Uruguay Round leaves out the unilateral liberaliza- and other pressure valve provisions. tion they implemented in the 1980s. If, however, we drop from 4 There are time limits as well as notification and consultation the table all the Latin American countries plus Sri Lanka and requirements. Tunisia, the average (absolute) imbalance is still 56 percent. 59 T H E W O R L D T R A D E O R G A N I Z AT I O N 11 In GATT/WTO usage, a nonpaper is a way of circulating an idea for discussion without proposing that the idea be adopt- ed; it is a means of promoting preliminary discussion. The nonpaper was cosponsored by 19 other developing countries. 12 Although considerable amounts of money will flow in different directions as a result of a tariff reduction and the political fall- out may be severe, implementing the reduction costs nothing. The signature of an executive or ratification by a legislature does it. 60 8 C O N S TA N T I N E M I C H A L O P O U L O S WTO Accession tries applying to join the WTO, many developing countries and economies face very similar chal- lenges in establishing the institu- tions needed to implement WTO commitments. Perhaps the most important of these challenges is the need to introduce laws and institutions for the operation of private enterprises and markets free from government controls-- other than those explicitly pro- vided under WTO regulations-- A regarding, for example, standards, ccession to the WTO is a complex, sanitary and phytosanitary (SPS) provisions, intellec- difficult, and lengthy process. In tual property rights, and state-trading practices. May 2001 it was a process being faced by 28 coun- Equally important to a country's economy is the tries, 9 of them transition economies and about half introduction of greater stability in commercial poli- of the remainder, least-developed countries (LDCs). cy, which is a consequence of adherence to WTO This chapter analyzes the WTO accession process rules and legally binding agreements. Stability is and identifies the main issues and challenges faced important both to domestic producers and to by acceding countries. exporters from other countries wishing to access these economies' markets.Adherence to WTO provi- sions--for example, by binding tariffs and by speci- Benefits of Membership fying conditions for foreign direct investment (FDI) There are three main benefits of WTO membership: in the services agreement--would improve the effi- (a) strengthening of domestic policies and institu- ciency and productivity of acceding countries. tions for the conduct of international trade in both WTO membership also offers the opportunity for goods and services, which is required before acces- new members to lock in existing, relatively liberal sion into the WTO can be accomplished; (b) trade regimes. Although the trade regimes in acced- improvements in the ease and security of market ing economies vary considerably, many have estab- access to major export markets; and (c) access to a lished regimes with relatively low tariffs and no dispute settlement mechanism for trade issues. significant formal nontariff barriers. For these countries, membership provides the opportunity to lock in these regimes by assuming legally binding Policies and Institutions obligations regarding tariff levels. This not only per- Although there are significant differences in the insti- mits them to enjoy the benefits of liberal trade but tutional and policy environment of the various coun- also gives them a first line of defense against the 61 T H E W O R L D T R A D E O R G A N I Z AT I O N domestic protectionist pressures that are present in cies and institutions that have a bearing on the con- all market economies. duct of international trade; (b) the members' fact- finding phase; and (c) the negotiation phase. The last two phases, while conceptually separate, tend to Market Access overlap in practice. Throughout, the applicant is Two main dimensions of market access are of faced with meeting WTO requirements and provi- importance to acceding economies. The first is the sions, as well as demands by existing members. With extension of permanent and unconditional most- very few exceptions, negotiation is in one direction favored-nation (MFN) status, which comes with only: the applicant is asked to demonstrate how it WTO membership. At present, economies that are intends to meet the existing WTO provisions--it not members of the WTO have been granted MFN cannot change them. Existing members can ask the treatment voluntarily by major trading partners, applicant to reduce the level of protection in its but there is nothing to guarantee that they will con- markets, but the reverse does not usually occur. tinue to be accorded such treatment. For example, in the United States extension of MFN to Russia and The Formalities several other economies in transition is contingent on the economies' adherence to the provisions of After a country sends a letter to the Director-Gener- the Jackson-Vanik amendment to the 1974 Trade al of the WTO expressing its desire to accede to the Act regarding freedom of emigration.1 The second organization, the request is considered by the WTO point is the substantial evidence that the incidence General Council, which consists of representatives of antidumping actions (both investigations and of all members and which meets frequently during definitive measures) is much higher against non- the course of the year. The General Council routine- WTO members than against members. ly decides to set up a working party, with appropri- ate terms of reference, to consider the accession application, and it nominates a chairman of the Dispute Settlement working party.2 Membership in the working party is Access to an impartial and binding dispute settle- open to all members of the WTO. In the case of ment mechanism, the decisions of which have a sig- applications by large countries such as China or nificant chance of being enforced, is an important Russia, many countries participate; in the case of potential benefit for the acceding economies, many smaller countries, the working party is usually made of which are small and heavily dependent on inter- up only of the "Quad" (Canada, the European national trade. The WTO's dispute settlement Union, Japan, and the United States) plus a number mechanism has proved successful in providing of other members, including neighboring countries opportunities for members to obtain satisfaction that are significant trading partners of the appli- regarding grievances stemming from practices of cant. The formalities phase can be quite short--no other members that cause trade injury. Although more than a few months. developing countries face some problems in access- ing this mechanism, membership provides an The Memorandum opportunity that, with proper assistance, can be beneficial to new members, especially in their rela- The preparation of the memorandum on the for- tionships with large trading partners. eign trade regime by the applicant explaining its policies and institutions can be a demanding task because of the range of issues that the memoran- The Accession Process dum has to address and the degree of detail The process of accession to the WTO is demanding required. The issues include much more than sim- and lengthy. It can be divided into an introductory ply trade in goods and services, although describing phase of formalities and three substantive phases. the trade regime for services, which encompasses The three substantive phases are (a) the applicant's the financial sector, insurance, telecommunications, preparation of a memorandum on the foreign trade professional services, and the like, is a large task in regime (hereafter referred to as the "memoran- itself. Relevant subjects also include various aspects dum"), which describes in detail the country's poli- of foreign exchange management and controls, 62 WTO Accession investment and competition policy, protection of ent rules and criteria, as opposed to administra- intellectual and other property rights, and enter- tive discretion. A key issue for enterprises that are prise privatization. The preparation of the memo- expected to remain state owned is whether they randum is solely the responsibility of the applicant, operate under market conditions or enjoy special and so is any delay in its preparation. monopoly rights and privileges. Even if the original memorandum is prepared · Some issues relate to the jurisdiction and capacity quickly, if it is incomplete in its details or if the leg- of national agencies to implement policies on islation and practices described are inconsistent which commitments are being made. The funda- with WTO provisions, the subsequent question- mental concern is one of governance: do the and-answer period can be protracted. At times, agencies have the authority and capacity to imple- members have asked the WTO secretariat to review ment the commitments that they are making in draft memoranda before their circulation to prevent the context of WTO accession regarding the laws incomplete documentation from being disseminat- and regulations that affect the conduct of interna- ed. The secretariat, however, assumes no responsi- tional trade? A related concern has to do with the bility regarding the contents of the memorandum. role and jurisdiction of local authorities and whether they have the right and opportunity to nullify commitments made by the national Questions and Answers authorities in the context of accession negotia- Once the memorandum has been circulated to tions. WTO members, the accession process enters the second stage, in which members ask questions and Negotiations obtain clarifications on the applicant's policies and institutions. This typically takes several months. (In At some point during the question-and-answer the case of Russia, it took more than a year.) The phase--after most, but frequently not all, the points working party usually does not meet until the mem- raised by working party members have been orandum and the initial questions and answers have answered--the applicant is requested to submit its been distributed. so-called initial schedule of offers in goods and The purpose of the detailed review that takes services. This consists of (a) the detailed schedule of place during this phase and that may involve several tariffs the applicant proposes to impose on goods working party meetings is to make sure that the leg- and the level at which the tariffs are "bound," and islation and institutions of the applicant are in con- (b) the commitments it makes (and the limitations formity with WTO provisions. The applicant is it sets) on providing access to its market for ser- requested to submit for the consideration of the vices.3 In addition, the applicant is requested to working party members relevant legislation on a make commitments regarding the level of support it variety of issues covered by the WTO. Delays during plans to provide to its agriculture in relation to a this phase are frequent; if a member feels that the base reference period (usually three representative answers submitted to a question or the actions years before the application for accession), as well as taken to remedy an inconsistency are inadequate, it other aspects of its support for agricultural trade, simply resubmits the question for the next round. such as export subsidies. Although the issues raised in each accession Once these offers are tabled, the accession process working party vary somewhat depending on the enters its final phase, which involves specific bilater- country, some common themes emerge in the dis- al negotiations between the applicant and each cussions of accession, especially, but not exclusively, WTO member that wishes to hold such talks in the case of countries in transition. regarding the tariff level or the degree of openness of the services sector proposed by the prospective · Within the context of laws and the operations of member. The actual timing of the original offers government institutions, two broad issues typi- varies considerably, and sometimes they are tabled cally receive special attention: the degree of priva- very early in the question-and-answer phase, as tization in the economy, and the extent to which happened, for example, in the case of Georgia. government agencies involved in the regulation of Often, bilateral negotiations take place in parallel economic activity do so on the basis of transpar- with formal meetings of the working party that 63 T H E W O R L D T R A D E O R G A N I Z AT I O N continue to deal with questions and answers regard- General Council (usually a formality), the country ing the foreign trade regime. The negotiations phase is invited to sign a protocol of accession. can also be lengthy, depending on the degree of openness the applicant proposes and the demands Progress in Accession for market access made by members. When these negotiations are in the process of As of May 1, 2001, 28 working parties had been for- being finalized and the applicant has provided mally established to consider the accession applica- assurances that the legislation and institutions that tions of prospective WTO members (Table 8.1). Most would permit compliance with WTO provisions are of the working parties were established some time in place, a draft report on accession, including the ago, with the oldest, those for Algeria and China, dat- schedule of agreed commitments on goods and ing from 1987. With few exceptions (that of Algeria services, is prepared by the secretariat for considera- being one) most of the working parties are active. In tion by the working party. After approval by the practice, accession has taken, on average, a little more working party, the report is forwarded to the Gener- than five years, from the establishment of the work- al Council. Following a favorable decision by the ing party to entry into force of WTO membership. Table 8.1 Accessions to the World Trade Organization as of May 1, 2001 Establishment of Tariff Services Draft working Economy working party Memorandum offers offers party report Algeria 06/87 07/96 -- -- -- Andorra 10/97 02/99 09/99 09/99 -- Armenia 12/93 04/95 01/99 10/98, 07/99 08/99 Azerbaijan 07/97 04/99 -- -- -- Belarus 10/93 01/96 03/98 05/99 -- Bosnia and Herzegovina 4/00 -- -- -- -- Bhutan 10/99 -- -- -- -- Cambodia 12/94 06/99 -- -- -- Cape Verde 07/00 -- -- -- -- China 03/87 02/87, 09/93 04/94 09/94, 11/97 12/94, 05/97 Kazakhstan 02/96 09/96 06/97 09/97 -- Lao PDR 02/98 -- -- -- -- Lebanon 04/99 06/01 -- -- -- Macedonia, FYR 12/94 04/99 -- -- -- Nepal 06/89 02/90, 09/98 -- -- -- Russian Federation 06/93 03/94 02/98 10/99 -- Samoa 07/98 -- -- -- Saudi Arabia 07/93 07/94 09/97, 06/99 09/97, 06/99 -- Seychelles 07/95 08/96 06/97 05/97 -- Sudan 10/94 01/99 -- -- -- Taiwan (China) 09/92 10/92 02/96, 08/99 09/94, 08/99 08/98 Tonga 11/95 05/98 -- -- Ukraine 12/93 07/94 05/96 02/98, 06/98 -- Uzbekistan 12/94 09/98 -- -- -- Vanuatu 07/95 11/95 11/97, 05/98 11/97, 11/99 11/99 Vietnam 01/95 09/96 -- -- -- Yemen 07/00 -- -- -- -- Yugoslavia, FR 01/01 -- -- -- -- -- Not yet done. Source: World Trade Organization. 64 WTO Accession A number of economies are at an early stage in result in long delays while governments ascertain the negotiation process. They include Azerbaijan, consistency between existing legislation and regula- Bosnia and Herzegovina, Cape Verde, FYR Macedo- tions and WTO requirements and while they design nia, and Yemen. Several others, such as Armenia, and put in place the amendments or new legislation Belarus, Kazakhstan, Russia, and Ukraine, have been or regulations needed. This is compounded by the involved in the accession process for a long time reality that legislative processes are themselves but, for various reasons, are not yet close to com- lengthy. The WTO secretariat can be of assistance pleting it. The process was finally completed for only in a very limited way in the accession process, China and Taiwan (China) by the end of 2001. as the WTO budget allocates very few resources to accession of new members. The five staff members in the WTO Accession Division are thinly stretched Why Does It Take So Long? in servicing even the procedural needs and paper- To understand why accession to the WTO is such a work generated by more than 25 active accession long process, it is necessary to look first at the vari- working parties. ous phases of the accession process and the reasons why delays may occur. Difficulties in the Negotiations Phase The negotiations phase can be and frequently has Weak Follow-Up been the most time-consuming phase of accession. In several cases a government has taken the initial Negotiations partly have to do with whether the step to apply for accession and have a working party acceding member's policies and institutions are set up, but then it fails to follow up the accession consistent with various aspects of the WTO agree- process by preparing a memorandum on its policies ments and partly with the specific tariff bindings or taking subsequent steps, or it does so after a long and commitments in agriculture and services. interval. Working parties for Uzbekistan and Sudan Delays can occur on both sides. The acceding gov- were set up in 1994, but the memoranda of foreign ernment may be unwilling to make needed liberal- trade policy were only submitted in September ization commitments--for example, it may not 1998, for Uzbekistan, and January 1999, for Sudan. offer to liberalize nontariff barriers, or it may pro- pose binding tariffs at levels much higher than existing ones. Members, for their part, may not be Political Issues satisfied with the level of liberalization proposed or In a few cases political issues between an applicant may be unwilling to accept delays in bringing the and one or more influential WTO members have laws and institutions of the applicant into conform- introduced delays. This happened in the past to ity with WTO provisions. Sometimes, as in the cases some extent with the accessions of China (and of of Albania, Croatia, Estonia, and Latvia, delays have Taiwan, China, which is linked to it) and of FYR stemmed not from the WTO accession process as Macedonia. such but from disagreements between the European Union (EU) and the United States over the commit- ments of acceding countries in the WTO (for exam- Inherently Time-Consuming Processes ple, in audio-visual services) and the possible future Even if the above problems did not exist, accession association of these countries with the EU. is inherently a time-consuming process. The prepa- ration of the memorandum presents serious diffi- Strategy and Tactics of the Applicant Country culties for governments that typically do not have sufficient human or material resources to address Within the rules and disciplines of the WTO, each the issues that have to be discussed in detail. Most country has considerable scope as to how restrictive countries have had to seek assistance from outside or liberal its trade regime will be. There are no spe- experts funded by bilateral aid agencies, from the cific rules as to the maximum level at which a coun- WTO itself, and from the World Bank. try has to bind its tariffs, how many services it will The question-and-answer process is also time liberalize, whether to establish antidumping legisla- consuming; applicants' institutional weaknesses tion, or how fast to liberalize its agricultural trade. 65 T H E W O R L D T R A D E O R G A N I Z AT I O N Countries thus have a strategic choice to make dur- duce uncertainty about trade policy among the ing the negotiations phase: how liberal their trade country's trading partners. regime will be, consistent with overall WTO disci- Several transition countries that have recently plines. become WTO members, such as Albania, Estonia, A strategy that some countries have pursued in Georgia, the Kyrgyz Republic, Latvia, and Mongolia, their accession negotiations is to try to liberalize as pursued a different strategy. In most respects their little as is necessary to ensure accession. Since such governments adopted a liberal trade strategy as part applicants cannot negotiate significant improve- of the process of accession.4 This entails (a) binding ments in their access to other markets, they try to tariffs at the usually low currently prevailing levels maintain significant levels of protection to use as or agreeing to reduce and bind tariffs at low levels as bargaining chips for obtaining improved access in part of the accession negotiations; (b) agreeing to a future negotiating rounds. Some of the countries liberal trade regime in agriculture and services; and that are using this strategy, such as China and Rus- (c) at an early date after accession, participating in sia, also feel that significant levels of protection are such agreements as the government procurement necessary during a transition period when ineffi- code, which increases competition and transparen- cient state-owned enterprises are being restructured cy in the operation of their markets. (see Gabunia 1998). These countries have typically The fundamental benefits of such a strategy are presented initial offers that propose to bind tariffs at economic: these countries reap the benefits of liber- rates much higher than those currently applied. al trade and investment. But the strategy has a num- Similar issues arise in services. Many transition ber of other advantages as well: it tends to facilitate and developing economies feel that their services negotiations for accession; it provides governments sectors are underdeveloped and would like to limit with political cover against domestic protectionist the commitments they make to open these sectors interests that may otherwise succeed in subverting to foreign competition. This is especially an issue in an existing liberal trade regime; and the legally such areas as financial services and telecommunica- binding WTO commitments lock in reforms by tions, in which countries frequently face requests making it more difficult for future governments to from WTO members to establish liberal policies reverse the liberalization. Increased protection to regarding commercial presence. Such policies "safeguard" against serious injury to domestic would permit foreign services suppliers to establish industry is permitted under WTO rules, but it is subsidiaries or joint ventures based on the principle based on a detailed and transparent investigation to of national treatment, which prohibits discrimina- demonstrate injury, which is then notified to the tion against foreign services providers and thus has WTO and subjected to the scrutiny of other mem- a direct bearing on foreign direct investment. bers. This is far more difficult than for a powerful There are significant dangers to a "minimum lib- domestic industry to simply seek government sup- eralization" accession strategy. Individual countries, port for raising tariffs beyond the applied level but especially small developing economies, have little below the higher bound level, which a government leverage in market access negotiations, and so the can do almost without any constraint. The point potential benefits they may be able to obtain about the WTO is not that it prohibits protection through such a strategy may be very small. At the but, rather, that it permits it only according to cer- same time, maintaining protection through rela- tain rules; obeying these rules makes protection tively high tariffs and protected agriculture and more transparent as well as more difficult to initiate services sectors imposes costs on the applicants' and expand. own economies: they forgo the benefits of a more The Chinese accession, concluded in 2001, has liberal trade regime, which, in the first instance, combined elements of both strategies and has raised accrue to the country itself. If countries bind tariffs a number of additional issues. First, China has used at levels higher than those applied and assume few the process of WTO accession to stimulate and commitments regarding agriculture and services make irreversible substantial trade liberalization (both of which are possible under WTO rules), they and more broadly based reforms. Second, China, in are subject to another risk: they create an opening many ways an economy in transition, considers for domestic interests to exert political pressure for itself a developing country and has been seeking to additional protection in the future, and they pro- obtain transition periods and other special and dif- 66 WTO Accession ferential treatment that WTO agreements extend to WTO agreements. When such weaknesses are developing countries. The latter includes nonreci- brought out in negotiations, members suggest that procity, preferential market access, and different the applicant seek technical assistance, available commitments and time limits in the implementa- from a variety of bilateral and multilateral donors, tion of the provisions of various aspects of the and that it present a detailed plan regarding the par- agreements, ranging from agriculture to subsidies ticular aspects of the relevant WTO provisions in and trade-related intellectual property rights which weaknesses exist and how and within what (TRIPS). China, because of its position as a large time period it proposes to remedy them. market, has also bargained on certain aspects of UNCTAD, the World Bank, the European Union, market access, such as on textiles and on issues Switzerland, and the United States, as well the related to its designation as a nonmarket economy. WTO, have programs that provide technical assis- tance on various aspects of the accession process, especially in the preparation of the initial country Attitudes and Policies of WTO Members memorandum. Anecdotal evidence about these The demands made on newly acceding countries are programs suggests a somewhat uneven perfor- greater than WTO disciplines on existing members. mance. Most countries report very helpful contri- Based on recent accession experience, the areas dis- butions by foreign consultants and advisers in the cussed below are ones in which members typically preparation of the memorandum. In some cases, request that acceding countries make more far- however, it appears that advice provided by outside reaching commitments than those made by many experts has actually slowed the accession process existing members at similar levels of development. because the consultants suggested, and the country agreed to, a "bargaining" strategy of tariff binding at Tariffs. Acceding countries are requested to bind all high levels and limited offers on services. In addi- tariffs, whereas many developing countries contin- tion, there have been problems of coordination ue to have a large portion of their tariff schedule among the various donors, as well as between the outside agriculture unbound. Ceiling bindings have bilateral aid agencies providing the assistance and been accepted, but there is pressure to bind close to their colleagues in the trade ministries who negoti- applied rates. ate the accession. Agriculture. In addition to binding the tariff sched- Plurilateral Agreements. There is pressure for ule, commitments are expected on aggregate mea- countries to begin examining the provisions of the sures of support (AMS), export subsidies, and the plurilateral agreements (for example, on govern- like. Since many acceding countries did not provide ment procurement and civil aviation) at the time of substantial support to agriculture but, rather, penal- accession and to commit to a timetable for complet- ized it, the requests they face for reductions in AMS ing negotiations soon after accession. may not be warranted, and in any case meaningful calculation of commitments in this area is subject to "Market Economy" Issues. Although there is no serious statistical difficulties. explicit requirement in the WTO agreements that a member have a market economy, a requirement Rules and Disciplines. Acceding countries are typi- that acceding countries have, fundamentally, such cally requested to meet all commitments at entry an economy is being pushed de facto by existing with regard to, for example, TRIPS, customs valua- members as part of their leverage in the accession tion, standards, and SPS regulations, without time process.5 The pressures have been felt by all acced- limits such as those available to existing members at ing countries, including China, where explicit similar levels of development, and regardless of understandings were reached with regard to the whether institutional weaknesses make it difficult existence of state trading in specific sectors. At the for them to fulfill such commitments. Such weak- same time the Quad countries have been unwilling nesses relate broadly to the operations of a market to modify their own antidumping procedures economy; it takes time to establish the institutional regarding the designation as "nonmarket infrastructure that would enable the applicants to economies" of transition countries that have discharge their responsibilities properly under the become WTO members. Under this designation, 67 T H E W O R L D T R A D E O R G A N I Z AT I O N different, less transparent, and potentially discrimi- It is politically difficult to adopt a liberal trade natory practices can be applied in the determina- strategy at accession, especially when major trading tion of whether dumping has occurred and, in the partners, which are WTO members, take advantage case of EU safeguard actions, against imports from of opportunities that are perfectly legal under the a number of these countries, including all the mem- WTO to limit market access--for example, by bers of the Commonwealth of Independent States maintaining high levels of protection in agriculture. (CIS) and China (Michalopoulos and Winters Even recognizing the political difficulties involved, a 1997). For this reason, the nonmarket economies strong argument can nonetheless be made that if label has been a major cause of trade friction developing countries and transition economies cur- between many transition economies, on the one rently applying for WTO accession adopt a liberal hand, and the United States and the European trade strategy at entry, they will maximize the bene- Union, on the other. fits and opportunities for integration in the interna- Legal justification for using such procedures can tional community that WTO membership offers be found in GATT provisions that permit different and will accede more quickly, as well. treatment "in the case of imports from a country It is fair to ask whether countries should not which has complete or substantially complete maintain some flexibility in their initial offers, as monopoly of its trade and where all domestic prices they are bound to face demands to liberalize by are fixed by the State" (Palmeter 1998: 116). These existing members almost irrespective of the level of practices were perhaps fully justified when practical- protection they initially propose. Although there is ly all trade was controlled by state trading enter- merit in this point, it probably should not be prises or ministries and prices were fixed by the pushed too far. Experience in recent accession nego- state. Countries in transition, however, have made tiations suggests that countries which make initial great progress in introducing market forces in recent offers to bind their tariffs at levels significantly dif- years. It would be difficult to argue that, say, China ferent from the applied level encounter serious diffi- or Russia has at present "a substantially complete culties in accession--even though the practice is monopoly on trade" or that all domestic prices in widespread among existing developing country these countries are fixed by the state. Thus, continu- members, many of which have not bound large por- ation of the traditional EU and United States tions of their tariff schedule. When such an initial antidumping practices no longer appears justified in offer is put on the table (as has happened with sev- the new setting (Michalopoulos and Winters 1997).6 eral countries of the former Soviet Union, as well as Because the GATT antidumping provisions accept other applicants for accession), working party national legislation and practices as decisive, the odd members basically refuse to consider it or to enter situation can arise in which countries become new negotiations on that basis. They simply ask the WTO members but are still designated as nonmar- country to submit a revised offer with bound rates ket economies for antidumping purposes. closer to the applied ones before serious negotia- tions take place. China's accession is unique, for both political and Lessons of Experience and Issues for the economic reasons, and lessons from it have to be Future drawn with extreme care. Undoubtedly, China has The first important lesson of experience is that each used WTO accession to promote and lock in wide- accession case involves a different negotiation, with ranging reforms. China differs from most recently different dynamics. This makes it difficult to gener- acceding countries in that it has been able to negoti- alize. Nevertheless, the cases of a number of small ate a number of transition periods--for example, countries that have recently concluded the accession for eliminating quantitative restrictions, licensing, process suggest that the smaller the country and the and state trading--as well as the maintenance of more liberal its regime, the faster the accession tariff quotas in agriculture. It probably has much process. There are two reasons for this: smaller more bargaining power than all the recently acced- countries realize that the costs of protection are ing countries taken together. It is a moot point high for them, and the small size of their economies whether the time limits and extensions obtained by poses fewer market access issues for major WTO China (which were much less than it requested) are members. compatible with its economic interests or whether a 68 WTO Accession faster liberalization of its trade regime would have complex, and challenging for all countries, especial- been more conducive to the country's longer-term ly the LDCs. The process is inherently time consum- development. To WTO members negotiating ing, but there are a number of steps that acceding China's accession, it is almost irrelevant what Chi- countries and WTO members could take which nese protection does to China's economy; they are would facilitate and expedite accession. concerned about the impact of such protection on Governments seeking accession need, first, to their exports to the Chinese market and its large establish a central coordination point to provide potential. direction and manage the multiplicity of legislative At the same time, China has had to accept limita- and regulatory changes in their foreign trade regime tions on its market access that other developing that are necessary for accession. Second, they need countries have not. For example, it has agreed to be to adopt liberal trade policies, which will both con- subject to product-specific selective safeguards; it tribute to their effective integration into the inter- accepted three more years of restrictions in the national economy and facilitate WTO entry. Third, implementation of the Agreement on Textiles and governments need to focus on and identify those Clothing (ATC), from which it had been completely areas of the WTO agreements in which weaknesses excluded; and it agreed to be designated a nonmar- in their institutional infrastructure require that they ket economy for 15 years. In this context, the more delay implementation of WTO provisions. They exceptions and the longer transitions a large coun- should actively solicit technical assistance, prepare a try such as China seeks to obtain, the more WTO realistic plan for implementing remedial actions, members will strive to maintain provisions that and seek agreement to obtain suitable delays in the inhibit full access of the acceding country's prod- implementation of the agreements as part of the ucts in their markets. accession process (despite the apparent reluctance It can be argued that WTO members' insistence of members to agree to such extensions so far). on a liberal commercial policy at entry is likely to WTO members can also take steps to help expe- serve the acceding countries' long-term develop- dite the accession process. It is in their interest that ment interests, as well as WTO members' commer- the organization achieve universal membership cial objectives. But insistence on adherence to all sooner rather than later, as current members would WTO commitments at entry and without transition benefit if all countries adhered to the rules and pro- periods in areas (such as customs valuation, TRIPS, visions of the WTO. In this regard, members should standards, and SPS) where there are obvious institu- attempt to ensure that accession is not delayed on tional weaknesses in LDCs and transition account of high-income countries' own disagree- economies raises serious problems. That is, acced- ments or disputes. ing countries, motivated by their strong desire for WTO members also need to consider the institu- membership, may agree to obligations that they tional weaknesses of acceding governments and to cannot implement, leaving them open to subse- moderate their demands by agreeing to suitable, quent complaints. Alternatively, providing generous time-bound extensions in meeting WTO obliga- transition periods at a time when the transition tions. This should not mean lowering the require- periods for other countries that are already mem- ments but, rather, allowing more time to meet bers are expiring would create inequities between them. If such extensions are not provided, either the existing and new members. The solution to this negotiations become stalled or the acceding country problem may be the substantial extension of some ends up accepting obligations that it cannot imple- of these transition periods both for existing low- ment. In particular there is merit to extending and income members of the WTO and for acceding standardizing transition periods for acceding coun- LDCs and transition economies; after all, these tries in the areas of standards, TRIPS, and custom transition periods were arbitrarily set in the first valuation, where countries invariably face serious place. constraints in meeting requirements at accession. Industrial countries should continue to provide assistance to developing countries and countries in Conclusion transition that are not members in order to For a variety of reasons, the process of WTO acces- strengthen their institutional capacities so that they sion has been and is likely to continue to be lengthy, are better able to meet the requirements for WTO 69 T H E W O R L D T R A D E O R G A N I Z AT I O N accession. Such assistance needs to be better coordi- ance and receive "permanent" conditional MFN treatment. nated. The Integrated Program of Trade-Related When, however, Georgia, the Kyrgyz Republic, and Mongolia became WTO members, the United States exercised its right of Technical Assistance to the Least-Developed Coun- nonapplication under WTO Article XIII; that is, it did not pro- tries has the potential to benefit a number of acced- vide these countries with unconditional MFN status and thus, ing countries. Consistent with preserving the WTO de facto, had not accepted their accession. Subsequently, leg- as a member-driven institution, industrial country islation was enacted permitting the United States to notify the members should also consider substantially increas- WTO that it has accepted these countries' membership. In the ing the resources available to the WTO secretariat case of China, the United States had to address this important issue through amended legislation before membership negoti- for assisting acceding governments in the prepara- ations were concluded. tion of the original memorandum and in the design of legislation and regulations that would enable the 2 Usually the chairman is an ambassador, a permanent represen- tative to the WTO. Countries often request and obtain observ- applicants to meet WTO obligations. Channeling er status at the WTO to familiarize themselves with the more resources through the WTO would permit the institution before they make a formal request for accession. secretariat to play a more active role in coordinating 3 Services commitments are typically more general and open- assistance efforts in support of accession and would ended than commitments in the sphere of goods. For a discus- give greater assurance that the outside experts who sion, see Part IV of this volume. assist governments with the preparation of the 4 The trade regimes in Croatia and Jordan, which also acceded needed documentation and the modification of leg- recently, were somewhat less liberal. islation and regulations do so in ways that more effectively meet WTO requirements. A reasonable 5 GATT Article XVII calls for notification of enterprises engaging objective would be to cut the processing time of in state-trading practices. The article, however, was never intended to address problems that come up when the bulk of accessions to no more than two years, a time frame external trade is controlled by the state. Indeed the GATT that is feasible if the above steps are taken. If that accommodated several countries, such as Romania and schedule were attained by all acceding countries, the Czechoslovakia, that at the time had centrally planned WTO would be able to achieve universal member- economies. ship in the next five years--a worthwhile objective 6 In 1997 the European Commission announced proposals for for the international community. liberalization of EU policy on this issue vis-à-vis Russia and China, which would terminate their designation as nonmarket Notes economies at the country level and would permit determina- tions to be made case by case, taking into account the market 1 Belarus is subject to annual waivers (as was China, until recent- conditions prevailing in each commodity in which dumping ly); the other countries have been found to be in full compli- had been alleged (Croft 1997). This is similar to U.S. practice. 70 9 VA L E N T I N A D E L I C H Developing AB report. In these instances, only "negative" consensus can stop the process; that is, all Countries and members must agree not to pro- ceed or not to adopt panel and the WTO Dispute AB recommendations or rul- ings. This reversal of the consen- Settlement System sus rule led to a radical change in the dynamics of dispute set- tlement, making it more auto- matic and less dependent on the power of the countries involved in a dispute. Since there is an W extensive literature comparing ith the creation of the WTO, devel- the GATT and WTO systems, we confine ourselves oping and industrial countries to briefly summarizing the salient features of the became subject to the same set of rules and to simi- DSU before turning to developing country experi- lar commitments. A new Dispute Settlement ence and concerns.1 Understanding (DSU) was negotiated to enforce multilateral disciplines. The DSU is widely regarded The WTO Dispute Settlement System as one of the positive outcomes of the Uruguay Round, marking a move toward a more "automatic" The Dispute Settlement Body (DSB), which com- and rule-oriented system (Jackson 1997). This prises all WTO members, has the authority to estab- chapter evaluates the functioning of the DSU from a lish panels, adopt panel and AB reports, maintain developing country perspective. surveillance of the implementation of rulings and Although the cornerstone of the WTO dispute recommendations, and authorize suspension of settlement mechanism remains Articles XXII and concessions and other obligations under WTO XXIII of the GATT, the DSU brought about a sub- agreements (Art. 2 DSU). A member must first stantial change in the workings of the system. A request bilateral consultations if it considers that a major improvement was to remove the consensus benefit accruing to it directly or indirectly under the requirement at key stages of the process. The DSU WTO agreements is being nullified or impaired states that "where the rules and procedures of this (Art. 4 DSU). If consultations fail to settle the dis- Understanding provide for the Dispute Settlement pute, the complaining party may request the estab- Body to take a decision, it shall do so by consensus," lishment of a panel, which must be created unless but this general rule does not apply to the establish- the DSB decides by consensus not to do so (Art. 6 ment of a panel of experts, the adoption of its DSU). report, or, if the report was subject to an appeal A panel is generally composed of three panelists. before the Appellate Body (AB), the adoption of the Its deliberations are confidential, and the opinions 71 T H E W O R L D T R A D E O R G A N I Z AT I O N expressed in the panel report by individual panelists may seek to suspend concessions under another are anonymous (Arts. 6, 14 DSU). Nationals of covered agreement (Art. 22 DSU). countries that are parties to the dispute may not Suspension of concessions (retaliatory action) is serve on a panel unless the parties to the dispute the last recourse for countries in enforcing compli- agree otherwise. Panels must conduct examinations ance with DSB recommendations and rulings. Of within six months (Art. 12 DSU). Within 60 days of course, effective retaliatory measures are available to the date of circulation of a panel report to WTO countries with economic power. As noted by Hoek- members, the report must be adopted at a DSB man and Mavroidis (2000: 531): meeting unless a party to the dispute formally noti- fies the DSB of its decision to appeal or the DSB Those WTO members that can afford to either decides by consensus not to adopt the report (Art. take countermeasures or to incur the costs of 16 DSU). action being taken against them are in a better The Appellate Body, a standing tribunal created in position. When acting as complainants they the Uruguay Round, considers any appeals. The tri- will use threat and/or imposition of counter- bunal consists of seven members, of whom three measures in order to induce compliance; when serve on any given case. The members are appointed acting as defendants, they will have at least the for four years and may not be connected with any luxury of weighting the pros and cons between government. Appeals are limited to issues of law changing the domestic policies at stake (in covered in the panel report and to legal interpreta- order to avoid imposition of countermeasures) tions developed by the panel. AB proceedings are or simply keeping the domestic policies at stake not to exceed 60 days and are confidential. The intact (and see countermeasures imposed reports are drafted in the absence of the parties to against them). the dispute, and the opinions expressed in them are anonymous (Art. 17 DSU). When a panel or the AB A key principle, nevertheless, is that multilateral concludes that a measure is inconsistent with a cov- authorization of retaliatory action is required. In ered agreement, it must recommend that the mem- this regard there is a "sequencing problem." Article ber concerned bring its measures into conformity 22.6 prescribes that retaliation must be authorized with the WTO agreement (Art. 19 DSU). within 30 days of the time a country is supposed to Article 21.5 of the DSU states that "where there is comply with a WTO ruling. This deadline, however, a disagreement as to the existence or consistency does not allow enough time for completion of a with a covered agreement of measures taken to compliance review under Article 21.5.2 Valles and comply with the recommendations and rulings, McGivern (2000) conclude that three different such dispute shall be decided through recourse to precedents have been established for determining these dispute settlement procedures, including the consistency of implementing measures and the wherever possible resort to the original panel." In suspension of concessions: the Bananas model, in turn, Article 22 of the DSU foresees that "if no satis- which the arbitrators first determined the WTO- factory compensation has been agreed, (. . .) any consistency of the implementing measures before party having invoked the dispute settlement proce- assessing the level of the suspension of concessions; dures may request authorization from the DSB to the Salmon model, in which the parties provided for suspend the application to the member concerned "sequencing" on an ad hoc basis; and the Subsidies of concessions or other obligations under the cov- and Countervailing Measures (SCM) model, in ered agreements." The complaining party should which the parties used a provision of the SCM first seek to suspend concessions or other obliga- agreement to extend Article 22 retaliation deadlines. tions with respect to the same sector as that in In each case, the complaining party requested the which the panel or the Appellate Body has found a establishment of an Article 21.5 panel on the basis violation or a nullification or impairment of bene- of a bilateral agreement to extend the Article 22 fits. Then, if the party considers that such action is deadlines until after completion of the Article 21.5 not practicable or effective, it may seek to suspend review. concessions or other obligations in other sectors A number of provisions in the DSU relate to under the same agreement. Finally, if the circum- developing countries. Article 4.10 calls for members stances are serious enough, the complaining party to give special attention to the particular problems 72 Developing Countries and the WTO Dispute Settlement System and interests of developing countries in consulta- Developing Countries' Experience with the tions, and Article 12.10 allows for the extension of WTO Dispute Settlement System the consultation period in cases of measures taken by developing countries if the parties agree. Article Most of the clauses in the DSU regarding develop- 8.19 provides that a developing country involved in ing countries have proved to be more declarative a case can request that the panel include at least one than operative. For instance, the concept, in Article person from a developing country, and Article 12.11 4.10, of giving "special attention" to the particular states that in such cases the panel report is to indi- problems and interests of developing countries cate how account was taken of relevant provisions during consultations has no operative content and concerning differential and more favorable treat- has not been developed in panel or AB reports. ment for developing countries that are embodied in Although in one case this article was mentioned in the WTO agreements referred to in the dispute. a DSB meeting to support a developing country If a case is brought by a developing country, the position, there was no substantive discussion of the DSB, in considering what appropriate action might "special attention" concept. A similar problem be taken, is to take into account not only the trade arose with special and differential treatment (S&D) coverage of the measures complained of but also clauses in such agreements as that on antidumping their impact on the economy of the country con- (see Box 9.1). Although some panels have dealt cerned (Art. 21.8 DSU). Article 27.2 provides for with S&D clauses, given that these clauses have neutral legal advice (technical assistance) to be fur- been invoked in fewer than 10 cases involving nished by the WTO secretariat to developing country developing countries, it seems that they are not members. Finally, Article 24.1 calls for due restraint very relevant for these countries, either in defend- in invoking the DSU against least-developed coun- ing or in claiming their rights. 3 (See Chapter 49, by tries (LDCs), in asking for compensation, or in seek- Oyejide, in this volume; see also Whalley 1999.) ing authorization to suspend the application of DSU provisions related to the surveillance of concessions or other obligations to these countries. implementation of DSB recommendations and rul- B O X 9 . 1 S P E C I A L A N D D I F F E R E N T I A L T R E AT M E N T A N D T H E D S U : S O M E E X A M P L E S The Scallops Case cles 3(7), 4(2) and 4(5) of the DSU. This In 1995 Chile and Peru requested that a panel be request had been disregarded by the Com- established on the trade description of scallops munities thus discriminating against and drawn up by the European Communities (EC). impairing Chile's interests in deviation from The EC asked that the item be removed from the the provisions of Article 4(10) of the DSU DSB agenda, arguing that the time periods for which stated that Members "should give spe- consultations and for the inclusion of those items cial attention to the particular problems and on the agenda stipulated in the DSU had not interests of developing country Members." This been respected. According to Chile, however, was a discrimination against Chile which was not being granted the same treatment as the Communities did not take into account Canada and was not in conformity with the that consultations with Chile had already obligations of WTO members towards a been initiated when the Communities had developing country. agreed to Chile's participation in consulta- tions held with Canada on the same subject The Bed Linen Case. . . . [or that] [w]hen Canada-EC consultations In European Communities: Antidumping Duties on had been completed, Chile had requested Imports of Cotton-Type Bed Linen, India contended further consultations to resolve this matter in that the EC had not taken into account the spe- accordance with the letter and spirit of Arti- cial situation of India as a developing country. (continued) 73 T H E W O R L D T R A D E O R G A N I Z AT I O N B O X 9 . 1 ( C O N T I N U E D ) India asserted that the EC had not acted consis- provides procedural safeguards, and thus, tently with Article 15 of the Agreement on does not require any particular substantive Antidumping, which recognizes that "special outcome, or any specific accommodations regard must be given by developed country to be made on the basis of developing members to the special situation of developing country status. In the United States' view, country members when considering the applica- [Article 15] does not impose anything tion of anti-dumping measures" and calls for other than a procedural obligation to exploring constructive remedies before applying "explore" possibilities of constructive anti-dumping duties in instances where they remedies. The word "explore" cannot fair- would affect the essential interest of developing ly be read to imply an obligation to reach a country members. India asserted that the EC did particular substantive outcome; it merely not explore any such possibilities prior to the requires consideration of these possibilities. imposition of antidumping duties and did not react to detailed arguments from Indian The panel's view was that exporters pertaining to Article 15: "[D]espite repeated and detailed arguments by the Indian the imposition of a lesser duty or a price parties stressing the importance of the bed linen undertaking would constitute constructive and textile industries to India's economy, the EC remedies but we come to no conclusions failed to even mention India's status as a develop- as to what other actions might in addition ing country, let alone consider or comment on be considered to constitute constructive possibilities of constructive remedies." India sug- remedies as none have been proposed to gested that "such remedies may consist of, us. . . . In our view, Art. 15 imposes no obli- among others, the non-imposition of anti dump- gation to actually provide or accept any ing duties or undertakings." India rejected the constructive remedy that may be identified notion that any procedural mechanism, such as and/or offered. It does, however, impose simplified questionnaires or extensions of time, an obligation to actively consider, with an satisfied the requirements of Article 15. The EC open mind, the possibility of such a reme- agreed in principle and accepted that undertak- dy prior to imposition of an anti-dumping ings could be a remedy, but it argued that Indian measure that would affect the essential exporters did not offer undertakings within the interests of a developing country. time limits set by the EC regulation. The United States, a third party in this dispute, Source: WTO, WT/DSB/M/7 (scallops); WTO, argued that Article 15 WT/DS/141 (India). ings are too weak to imply any difference between transparent liaison between the state and industry the possibilities open to industrial and developing in order to obtain up-to-date information on trade countries. Article 21.7, however, mandates that problems in which developing countries have a when a matter is raised by a developing country, stake. Developing countries lack the high-level the DSB is to consider what further action might be expertise and resources to devote to such activities. appropriate to the circumstances. To date, this pro- International financing for training public officials, vision has not been used by a developing country, screening industrial countries' trade policies, and perhaps because a precondition is that the country building a network with other developing countries devote resources to analyzing and following cases. with the aim of jointly presenting cases could help This involves checking arguments, issues, and pos- address some of these problems. sibilities and comparing experiences and results; The technical assistance called for in Article 27.2 exploring new legal as well economic arguments; is provided by only a few consultants and is inade- and, domestically, building up an efficient and quate, given the large number of cases. In addition, 74 Developing Countries and the WTO Dispute Settlement System since the WTO secretariat must be impartial, its lat- ity. These proposals are supported by many devel- itude for helping developing countries with legal oping countries in particular, as there is a common strategic issues is limited. In this context, the Advi- concern regarding the costs associated with submit- sory Centre on WTO Law (described in Box 9.2) ting, pursuing, and defending cases and the scarcity could play an important role in helping developing of human resources for dealing with increasingly country governments present and pursue cases. complex issues. Venezuela has noted the need to increase the num- Finally, provisions related to least-developed ber of legal assistants to the secretariat to help countries have not been invoked at all because no developing countries and has called for the creation least-developed country has been involved in a dis- of a trust fund to establish alliances with private law pute, as a complainant or as a respondent. firms to augment developing countries' legal capac- B O X 9 . 2 T H E A D V I S O R Y C E N T R E O N W T O L AW Claudia Orozco ratification process, and the Advisory Centre On the side of the Third WTO Ministerial Confer- became operational in October 2001. ence, held in Seattle in 1999, ministers from 29 The Advisory Centre provides legal advice on WTO member countries signed an agreement WTO law to developing countries and to establishing the Advisory Centre on WTO Law economies in transition. This legal advice might (ACWL). The establishment of the ACWL is a con- take the form of advisory opinions on particular crete action toward addressing the needs of questions of law, analysis of situations involving developing countries for advice and training on trade concerns, or legal advice provided WTO law. The contractual nature of the WTO throughout a dispute settlement proceeding. In requires that members have a full understanding recognition of the differences among developing of the content and scope of their rights and obli- countries, the extent of the support to be provid- gations and that they are able to access the dis- ed will depend on the needs and requirements pute settlement mechanism. Otherwise, the of each member in each case. Examples might ever-growing complexity and breadth of the sys- include outlining the legal questions of a case, tem, coupled with the relative scarcity of special- drafting submissions, and commenting on drafts ized human resources in developing countries prepared by government officials. In addition, and the costs of specialized external legal coun- the ACWL holds regular in-house seminars on sel, would marginalize many members. jurisprudence for Geneva-based officials and To help address these needs, a legal aid facility regional yearly seminars for officials based in was proposed, with two goals: (a) training gov- capitals. Finally, and most important, the ACWL ernment officials in WTO jurisprudence, and (b) offers on-the-job training for government offi- providing specialized legal advice on WTO law, to cials in charge of a particular case and intern- include support throughout legal proceedings. ships for government lawyers responsible for The response was the establishment of the Advi- WTO issues. sory Centre as a small, independent international organization based in Geneva and open to all Note: Claudia Orozco was minister counselor at the WTO members. By March 31, 2000, the final mission of Colombia to the WTO between August 1994 date for becoming a founding member, the and July 2000. She served as legal counselor and was a treaty had been signed by 9 industrial and 22 panelist in several cases. On February 1998 she submit- developing countries. The 38 least-developed ted the project proposal for the Advisory Centre on countries that are members of the WTO are prior- WTO Law to the Netherlands, the United Kingdom, and ity beneficiaries of the ACWL's services. The Norway and led consultations with an informal group agreement entered into force in July 2001, after of WTO members. The result was a proposal that was the requisite number of countries completed the offered to all WTO members. 75 T H E W O R L D T R A D E O R G A N I Z AT I O N Developing Country Participation in matters ranging from patent protection under the the DSU [Trade-Related Aspects of Intellectual Property Rights, or TRIPS] Agreement, to balance-of-pay- As of September 2000, 207 complaints had been ments restrictions, safeguard measures, and the tax- notified to the WTO (Table 9.1). Of these, 16 were ation of alcoholic beverages" (Lacarte-Muro and active, 40 had concluded with the adoption of either Gappah 2000). The exception is intellectual proper- an AB or a panel report, 34 had been settled bilater- ty rights (IPRs), with many complaints alleging vio- ally or were inactive, and 12 were being implement- lation of the TRIPS agreement by both developing ed (WTO 2000).4 Industrial countries brought the and industrial countries. As of 1999, the number of most cases, and their share of total complaints (74 TRIPS cases had already reached 16, equivalent to percent) was greater than their share of world 10 percent of all filings under the DSU. Eleven of exports. Among the different categories of cases, these filings were brought by the United States those brought by industrial countries against devel- (Geuze and Wager 1999). As regards developing oping countries appear to have increased the most countries, Correa (2001) notes that "although the between the GATT period and the WTO era, from adoption by another Member of unilateral trade 10 to 31 percent. Over 40 percent of industrial sanctions would be incompatible with the multilat- country cases were against developing countries-- eral rules, developing countries have continued to higher than the developing countries' 27 percent be under unilateral demands by some developed share of industrial countries' exports in 1998. The countries, notably the United States in the area of proportion of cases by developing countries against IPRs, in some cases aiming at ensuring protection of industrial countries was also higher than might such rights beyond the minimum standards set have been expected (66 percent of all developing forth by the Agreement" (Correa 2001: 22). country complaints) and was higher than the share Although developing countries are likely targets of industrial countries in developing country for intellectual property rights cases, IPRs may also exports, 57 percent (Weston and Delich 2000). become their most effective means of exerting pres- Latin America and Asia are the developing coun- sure and eventually retaliating. Ecuador's threat to try regions most involved in the dispute settlement suspend its TRIPS concessions in the Bananas case process. To date, African countries have not initiat- (see Chapter 10, by Hudec, in this volume) and the ed or been respondents in any case, although sever- strategy used by Brazil with regard to public health al, including Nigeria and Zimbabwe, have made and patents would appear to be the first steps in this presentations as third parties. direction (Dyer 2001; see also Chapter 36, by No single theme dominates the substance of the Maskus, in this volume).5 Subramanian and Watal cases involving developing countries. "As respon- (2000) have proposed that "developing countries dents, developing countries have been involved in convert their TRIPS obligations into instruments of Table 9.1 Number of Dispute Settlement Cases, 1995 through September 2000 Complaint by Share of Industrial Developing Total total cases countries countries complaints (percent) Complaint against: Industrial countries 89 35 124 60 Developing countries 65 18 83 40 Total 154 53 207 100 Share of total cases (percent) 74 26 Memorandum: Share of cases under GATT (percent) 84 16 Note: Based on number of cases brought by each country. The European Union and its member countries are counted jointly. Source: Weston and Delich (2000); WTO (2000b); IMF, Direction of Trade Statistics, various issues. 76 Developing Countries and the WTO Dispute Settlement System multilateral enforcement embodying the retaliation The compliance panel would comprise the mem- possibilities in domestic legislation" (p. 415). bers of the original panel, if its report had not been According to these authors, "domestic legislation appealed, or the members of the Appellate Body implementing the TRIPS agreement must clearly that considered the appeal if the report of the origi- specify that the country's executive reserves the nal panel had been appealed. The compliance panel right to revoke or dilute these rights in the event would be required to circulate its report within 90 that partner countries are found to be in non-com- days of the date of its establishment, after which any pliance with commitments that affect the country's party to the compliance panel proceeding would be interest" (p. 411). In addition, they hold that "if permitted to request a meeting of the DSB to adopt designed with care, retaliation in TRIPS can be fea- the report within a period of 10 days. The report sible, effective, and legal. Further, it has one really would be subject to the negative consensus rule: it attractive attribute that distinguishes it from con- would be automatically adopted unless the DSB ventional trade retaliation in the area of goods: decided by consensus not to adopt. retaliation in TRIPS can be genuinely welfare Compliance panel reports would not be subject to enhancing in a way that conventional retaliation--a appeal. If the compliance panel found that the case of shooting oneself in the foot to shoot at the member concerned had failed to bring its measures other person's foot--is not" (p. 405). into compliance within the reasonable period of time determined by the original panel, the com- plaining party could request authorization from the Proposals for Reforming the Dispute DSB to suspend the application of concessions to Settlement System the member concerned or to suspend other obliga- A number of proposals have been made by develop- tions under the covered agreements. ing countries and by scholars to improve the func- The joint proposal also modifies Article 22.2 to tioning of the dispute settlement system. This entitle the complaining party to request authoriza- section briefly summarizes these suggestions. tion to suspend concessions if a compliance panel report pursuant to Article 21 bis finds that the member concerned has failed to bring its measures On Implementation of Recommendations and into compliance with the ruling of the DSB. If the Rulings and Suspension of Concessions member concerned objects to the level of suspen- Three proposals have been made: amend the system sion proposed, the proposal states that "the matter to resolve procedural problems such as the sequenc- shall be referred to arbitration. The arbitration shall ing issue described earlier; allow financial compen- be completed and the decision of the arbitrator sation for developing countries; and turn retaliation shall be circulated to Members within 45 days after into a collective action. the referral of the matter. The complaining party A large number of WTO members have made a shall not suspend concessions or other obligations joint proposal that Article 21.2 of the DSU be during the course of the arbitration." reformed to address the sequencing problem.6 The In regard to financial compensation, Pakistan has proposal foresees the creation of Article 21 bis, enti- commented that "[i]t would be useful to clarify that tled "Determination of Compliance," that would the term `compensation' used in Article 22 includes establish the following procedures. A complaining grant of financial compensation to the complaining party may request the establishment of a compli- party by the country which has been found to be in ance panel (a) any time after the member concerned violation of the rules. Panels should be authorized states that it does not need further time for compli- to recommend payment of such financial compen- ance; (b) any time after the member concerned has sation in disputes between developed and develop- submitted a notification that it has complied with ing countries where they find that as a result of the recommendations or rulings of the DSB; or (c) WTO inconsistent measures taken by developed 10 days before the date of expiration for the "rea- countries, the developing country has lost its trade sonable period of time" to comply. While consulta- in the affected product" (WT/GC/W/162). It is not tions between the member concerned and the the first time that a developing country has called complaining party are desirable, they are not for the inclusion of financial compensation in the required prior to a request for a compliance panel. dispute settlement system: a similar proposal was 77 T H E W O R L D T R A D E O R G A N I Z AT I O N made during the GATT era (see Chapter 10, by porary developments." In particular, Pakistan pro- Hudec, in this volume).7 posed that such clarification make clear that panels Proposals have also been made to make violation or the Appellate Body were not permitted to take of WTO rules a collective problem and, accordingly, into account "unsolicited information" including to require collective retaliatory actions. Pauwelyn "amicus curiae briefs from private parties" (2000: 6), for instance, argues that (WT/GC/W/162). In November 2000, at a special WTO General with the advent of the WTO--its legal refine- Council meeting, developing countries called for ment, quasi-judicial dispute settlement system, the Appellate Body to exercise extreme caution in and, in particular, major expansion into new inviting amicus curiae briefs from nongovernmental fields that directly affect individuals--it may be organizations (NGOs). (The context was the devel- time to move away from the idea of the oping countries' reaction to the AB ruling on the GATT/WTO only as a package of bilateral bal- Asbestos case.) Developing countries sought to limit ances between governments. Has the time not the Appellate Body's "interpretation powers" and to come to introduce the WTO as a truly multilat- prevent NGOs from participating in the dispute set- eral construct providing legal rules as public tlement system. Brazil, Egypt, India, Pakistan, goods that merit collective enforcement for the Uruguay, and the Association of Southeast Asian good of governments and economic operators? Nations (ASEAN) countries argued that a decision . . . [T]he enforcement of WTO rules can and to admit amicus curiae briefs was a substantive and should be seen as a collective rather than a not a procedural one and was therefore something mainly bilateral exercise. for WTO members to decide. Moreover, "develop- ing countries took the view that non-governmental In addition, Pauwelyn has proposed that "coupled organizations are not accountable to sovereign par- with countermeasures, a broad scheme of compen- liaments and have no contractual rights and obliga- sation--additional market access offered by the los- tions in the WTO. The AB had let itself be unduly ing party to WTO members--would provide influenced by the campaign of NGOs of major trad- genuine leverage to induce compliance, a move ben- ing entities. In effect, NGOs were being accorded eficial to all WTO members, and not just `compensa- privileges greater than those enjoyed by WTO tion' to the one or few that brought the case" (p. 9). Members."9 Finally, Pakistan has presented a proposal that As for the composition of the Appellate Body, Article 22.3 be amended to eliminate the possibility India has proposed that, to promote an atmosphere of cross-retaliation by industrial against developing conducive to impartial and independent function- countries. This would impede, for instance, retalia- ing of the Appellate Body, all future appointments tion against trade in goods if a developing country of AB members should be for a nonrenewable fixed has been found to be in violation of the TRIPS term of five or six years, to ensure that members agreement (WT/GC/W/162, p. 3). have no incentive to seek support for their reap- pointment (WT/DSB/W/117). On the Appellate Body On Time Lines The role of the Appellate Body--in particular, the extent to which it has gone beyond its mandate and The joint proposal mentioned earlier would shorten undertaken to "make rules" through interpretation the consultation period from 60 to 30 days; the peri- of WTO agreements--has been severely questioned od could be extended by up to 30 additional days if by developing countries.8 Pakistan has called for an one or more of the parties in the dispute were a interpretation of "the relevant provisions in the developing country and the parties agreed. In addi- DSU to make it clear that the responsibility for clar- tion, in the Working Procedures, the proposal ifying or modifying the provisions of the WTO would reduce the time for receipt of the complain- Agreements clearly rests with the WTO member ing party's first written submissions to three­four countries and that it would not be appropriate for weeks (currently, it is three­six weeks), while the Appellate Body to usurp these functions under increasing the time for the party complained against the guise of interpreting law on the basis of contem- to respond to four­five weeks instead of two­three 78 Developing Countries and the WTO Dispute Settlement System weeks, as at present. Since the proposal unifies the provisions relating to enforcement of S&D language reports (there would be a single report, including in WTO agreements are ineffective, developing the descriptive sections and the panel's findings and countries do not enjoy a "neutral" playing field. conclusions), it eliminates the period during which Although the DSU is not biased against any party in parties submit their comments on the descriptive a dispute, developing countries are less well report and, consequently, the possibility that at the equipped to participate in the process: they have request of a party the panel would hold a further fewer people with the appropriate training, they are meeting on the issues identified in the written com- less experienced, and they can bring fewer financial ments. After making other adjustments on time resources to bear. Therefore, although the DSU is an lines, the proposal states that "the total reduction of asset, developing countries must work to obtain time is up to approximately 47 days, and the time international financing for training and capacity frames in Article 20 (the reference to 9 months and building and for the establishment of a joint mech- to 12 months), and the periods in Art. 21.4 (the ref- anism among developing countries to screen indus- erence to 15 months and to 18 months) shall be trial country trade policies of interest to them--not reduced by one month" (WT/MIN[99]8, p. 7). only to reduce the costs of the screening but also to coordinate the submission of joint cases. In addi- tion, developing countries could use cases in which On Third Parties they are involved as a way to identify gaps in WTO In relation to third parties, the joint proposal retains agreements that need to be addressed through the obligation contained in Article 10 that a copy of negotiations. all documentation submitted in a case be given to Reform of the dispute settlement system does not third parties. It allows exclusion, however, of certain appear to be a priority on the negotiating agenda of factual confidential information (designated as such developing countries. Their efforts are mainly by the disputing party) and sets a period of 15 days directed toward defending their interests as best for the party to provide a nonconfidential summary they can in current cases, bridging the gap with that can be disclosed to the public of the informa- industrial countries in terms of legal expertise, and tion contained in the confidential submission. establishing more effective enforcement and retalia- tory devices. A Special Prosecutor, "Light" Procedures, and Customs Unions Notes Hoekman and Mavroidis (2000) have proposed a 1 See for example, Komuro (1995); Lafer (1996); Jackson kind of "special prosecutor," able to act on an ex (1997); Montaña Mora (1997). officio basis, to detect illegalities. They also suggest 2 On the sequencing problem, see the Salmon dispute between "light" procedures for cases involving less than Austria and Canada; the U.S.-Australian dispute over leather US$1 million of exports; in such cases a single pan- subsidies; and the Bananas case. See also O'Connor and elist would be asked to address the dispute within Vergano (2000); Rhodes (2000); Valles and McGivern (2000). three months. Turkey has proposed amending Arti- 3 Cases in which S&D clauses were invoked by a party in a dis- cle 10 of the DSU to grant all parties to customs pute were European Communities: Antidumping Duties on unions the right to participate in panel and AB pro- Imports of Cotton-Type Bed Linen, complaint by India; Korea: ceedings in disputes concerning measures intro- Measures Affecting Imports of Fresh, Chilled, and Frozen Beef, complaint by the United States; India: Quantitative Restrictions duced pursuant to a common trade policy of the on Imports of Agricultural, Textile and Industrial Products, com- union (WT/MIN[99]/15). plaint by the United States; Brazil: Export Financing Programme for Aircraft, complaint by Canada; Canada: Measures Affecting the Export of Civilian Aircraft, complaint by Brazil; and Indone- Conclusion sia: Certain Measures Affecting the Automobile Industry, com- The Dispute Settlement Understanding brought plaints by the United States, the European Communities, and Japan. There were also cases in which S&D clauses were about a positive and beneficial change for develop- invoked by a third party or some kind of statement was made ing countries. Weaker states have a better chance to about a developing country's preferential treatment on the defend their interests in a rule-oriented than in a basis of its status as a developing country. Examples include power-oriented system. However, since the DSU Guatemala: Antidumping Investigation Regarding Imports of Port- 79 T H E W O R L D T R A D E O R G A N I Z AT I O N land Cement from Mexico, complaint by Mexico; European Com- 6 See "Proposed Amendment of the DSU," WT/MIN(99)8, sub- munities: Measures Affecting the Prohibition of Asbestos and mitted by the government of Japan on behalf of cosponsors Asbestos Products, complaint by Canada; and Mexico: Canada, Costa Rica, the Czech Republic, Ecuador, the Euro- Antidumping Investigation of High-Fructose Corn Syrup, com- pean Communities (and its member states), Hungary, Japan, plaint by the United States. These cases may be found on the Korea, New Zealand, Norway, Peru, Slovenia, Switzerland, WTO Website, . Thailand, and Venezuela. 4 There was one active case on implementation of WTO rulings, 7 For a detailed discussion on the Uruguay-Brazil plan to reform as well as six adopted AB and panel reports on implementation the dispute settlement system, including financial compensa- of WTO rulings (Art. 21.5 DSU), one active arbitration on the tion, see Dam (1970): 368­73. level of suspension of concessions (Arts. 22.6­7), and four authorizations of suspension of concessions (under Art. 22.7 8 Industrial countries have also questioned panel rulings on the DSU and Art. 4.10 of the Subsidies Agreement). same grounds. For instance, when the DSB, in a case involving subsidies to an automotive leather manufacturer, adopted a 5 Brazil linked developing countries' right to access affordable panel ruling that required, for the first time, that a private medicines to patents. First, Brazil put forward a very broad and company repay in full an illegal export subsidy, the countries ambitious plan to fight HIV in its territory and pressed corpora- involved--the United States and Australia--commented that tions to reduce drug prices. Simultaneously, Brazil obtained a "the ruling should not set a precedent for future disputes"; the declaration at the World Health Organization on the virtues of European Union said that the implications needed more dis- its HIV program. Finally, at the TRIPS Council, Brazil submitted cussion; and the United States stated that "the payback reme- a document highlighting the need to interpret the TRIPS dy went beyond that sought by the US." "Canada and Brazil agreement in a way that did not impede countries' ability to expressed their serious concerns about the decision and Japan implement health policies. Brazil received rapid and wide- and Malaysia voiced misgivings" (Financial Times, weekend, spread support from developing countries (and from public February 12­13, 2000). opinion as well). As a result of this strategy, the United States withdrew the panel against Brazil on intellectual property 9 The question of amicus curiae briefs is part of a broader debate rights, and at the Doha ministerial meeting a separate declara- on the governance of the trading system; see Chapter 47, by tion was made asserting that the TRIPS agreement does not Tussie and Lengyel, in this volume. See also "Developing Coun- and should not prevent members from taking measures to tries Make Their Mark on WTO Appellate Body Controversy," protect public health. World Trade Agenda, no. 00/22 (December 4, 2000), p. 11. 80 10 R O B E R T E . H U D E C The Adequacy WTO system allow larger coun- tries to exert significantly stronger enforcement pressures of WTO Dispute against developing countries than developing countries can Settlement exert in the reverse situation. The shortcomings of the WTO legal Remedies system in this regard thus raise a legitimate issue for developing country governments when they must decide whether to employ A Developing Country Perspective the dispute settlement procedure against larger countries. In addi- A tion, these shortcomings raise ccording to conventional wisdom, the question of whether it would be worthwhile for it is a waste of time and money for developing countries to expend negotiating capital developing countries to invoke the WTO's dispute in an effort to remedy these shortcomings, and, if so, settlement procedure against industrial countries. what particular reforms should be sought. Even if, the argument runs, a developing country In this chapter, I examine the facts behind the con- obtains a clear legal ruling that an industrial coun- ventional wisdom more carefully than is usually try has violated its legal obligations, the developing done. My purpose is not to prove that the conven- country has no effective way to enforce the ruling. tional wisdom is entirely wrong. Rather, I hope to The only enforcement sanction provided by the show that the issue here, as in most issues, is not quite WTO dispute settlement procedure is trade retalia- the open-and-shut proposition that is usually tion--the imposition of discriminatory trade sanc- advanced. I think it is important that officials respon- tions by the complaining country against the trade sible for deciding these issues have an accurate of the defendant country. And trade retaliation by understanding of what is and is not wrong with the smaller developing countries, it is argued, simply remedies offered by the existing WTO system, and of does not inflict any significant harm on larger how well or how poorly the system works in practice. industrial countries. In the end, the argument con- My purpose is to outline at least some of the infor- cludes, retaliation will harm the developing country mation needed to arrive at such an understanding. imposing it far more than it will harm the industri- al country it is supposed to punish. Enforcement under the GATT Dispute The conventional wisdom has a great deal of truth Settlement System to it. The "law" of the WTO does not, in fact, give weaker countries the same protection that well- The weaknesses of the GATT dispute settlement sys- developed domestic legal systems usually afford tem, which operated from 1948 to 1994, have been their weaker citizens. The remedies provided by the described so often that they need little elaboration. 81 T H E W O R L D T R A D E O R G A N I Z AT I O N The entire system was based on consensus decision- response to legal rulings. In the 1980s, when gov- making, which meant that the consent of the defen- ernments began to use the dispute settlement sys- dant was required before the procedure--creating a tem to deal with more politically controversial panel, defining its terms of reference, appointing its matters, the success rate dropped to about 81 per- members, adopting its ruling, and authorizing retali- cent--not up to the standards of most domestic ation--could move forward at all. The central reform legal systems, but still a very impressive perfor- made by the WTO Dispute Settlement Understand- mance for an international legal regime, especially ing (DSU) was to make the procedure go forward in the politically sensitive area of trade policy automatically on the request of the complainant, (Hudec 1993: 285­94). with or without the consent of the defendant. Although complaints by developing countries did In addition to its central weakness, the formal not achieve the same level of success as those remedies provided by the GATT legal system when brought by larger countries, the results were still legal violation was found were also rather limited. A favorable in a significant percentage of the cases. ruling of violation entitled the complaining govern- Over the GATT's entire history, 28 complaints were ment to a rather general "recommendation" calling brought by developing countries. Of these, 17 ended on the defendant government to comply with its in legal rulings, 11 of which were rulings of legal obligations. The recommendation was directed only violation, and 10 of the 11 (91 percent) had a suc- toward future conduct, with no compensation for cessful outcome. Of the 22 complaints known to be harm done while the violation was in force. There based on a valid legal claim, satisfaction was was no time limit on the order to comply, and the achieved in 18 of the cases (82 percent). Even in the process of seeking compliance could drag on for more contentious cases of the 1980s, legally valid years. The complaining government could at some complaints by developing countries achieved a 73 point request authorization to retaliate by imposing percent success rate (Hudec 1993: 315­26). approximately equal trade barriers in return, but The paradoxical contrast between the voluntary the request could be vetoed by the defendant. In procedures and weak remedies of the GATT dispute modern GATT practice, only two requests for retal- settlement system, on the one hand, and its rather iation authority were made, both against the United strong record of success, on the other, contains a les- States, and both were vetoed.1 son. It teaches that the enforcement of international Despite the defendant's ability to block the proce- legal obligations cannot be explained by superficial dure, the GATT disputes procedure produced a con- analysis of dispute settlement procedures and reme- siderable number of dispute settlement complaints dies. Enforcement requires that governments be during its almost 50-year history. My own study of persuaded to reverse decisions they have taken in GATT cases from 1948 to the end of 1989 counted violation of the agreement. Governments are not 207 cases filed during that period, of which 88 pro- private litigants. They are complex institutions that duced legal rulings; of the 88, 68 were rulings of vio- make decisions in their own peculiar, often irra- lation. In the decade of the 1980s, when the GATT tional, manner, which we call "politics." Even small- system had matured, there were 115 complaints er governments are strong enough to be able to yielding 47 legal rulings, of which 40 were rulings of resist coercive forces that would move private liti- legal violation (Hudec 1993: 277­78). Provisional gants. Governments, however, usually have a data from a continuation of that study list 71 more longer-term interest in the efficacy of the legal rela- complaints in the final five years of GATT opera- tionships they have established with other govern- tions (1990 to 1994), with 22 legal rulings, 20 of ments, and so they are more inclined to act in ways which were rulings of legal violation.2 designed to preserve those relationships. Ultimately, Notwithstanding the defendant's ability to block the compliance decisions of governments are deter- adoption of adverse rulings, the great majority of mined more by calculated self-interest than by the violation rulings were in fact adopted. More- force. over, the bulk of these violation rulings, including In my view, government compliance with legal many of those not adopted, did produce a satisfac- rulings is usually the product of at least three inter- tory correction of the practice at issue. In its first related factors that influence the way in which gov- three decades the GATT system achieved almost a ernments make trade policy decisions. First, some 100 percent success rate in producing a satisfactory parts of the defendant government's decisionmak- 82 The Adequacy of WTO Dispute Settlement Remedies: A Developing Country Perspective ing apparatus usually want the conduct called for by effective, in many cases it may well be more effective GATT legal obligations to be pursued for its own than the other practical alternatives. Policy deci- sake, simply because it is good policy. Such officials, sions that focus only on the availability of retalia- and the private interest groups that share this view, tion thus run the risk of ignoring the other, quite constitute an existing political force within the valuable, gains that can be achieved from a legal rul- defendant government, and the effect of GATT legal ing alone. rulings is to give them greater influence in the national decisionmaking process. Second, many The WTO Reforms officials and private interest groups within the national government's decisionmaking process per- In the Uruguay Round, WTO member governments ceive a value in the legal system itself, believing that agreed to establish a more rigorous dispute settle- both they and their country will gain more over the ment system. As noted above, they began by making long term from an effective legal system than they the disputes procedure move forward automatically. will gain from noncompliance in this or that indi- The automaticity of the procedure makes it more dif- vidual case. Although these actors may not want to ficult for larger countries to bully smaller countries tie their country's hands through rigid commit- into giving up their legal complaints. If developing ments to a particular legal system--or may not have countries want to have a legal ruling, it will now enough political support to go that far--they will require less diplomatic confrontation to get one. nonetheless argue strongly against noncompliance The remedies granted for enforcing a binding in individual cases that would damage respect for ruling were also strengthened. In addition to mak- the system. Finally, one should not underestimate ing retaliation more readily available, the Uruguay the influence of active pressure by other govern- Round reforms adopted a number of reforms ments. If a majority of member governments intended to strengthen the effect of the ruling itself. believes in the value of an existing legal system, The primary remedy set forth in the DSU is still the those governments will have the same incentive to legally binding "recommendation" ordering the discourage noncompliance with legal rulings and defendant to bring its conduct into compliance. will express their views in the form of collective Although the recommendation is still only future- condemnation of noncompliance. directed (it provides no remedy for the harm done But even if all three factors existed in cases where by the violation so far), some steps were taken to no legal rulings were provided, legal rulings sharpen make that future-directed order more effective.3 the focus on the issue of compliance, and the nor- Panels were given explicit power to make nonbind- mative force of such rulings increases the power of ing suggestions for how compliance can be those participants who favor compliance. This is so achieved--a power that, if used, could sharpen the whether or not the ruling is enforced by coercive focus of compliance pressures. Quite a bit more sanctions. was added to the procedure for following up a rec- As noted above, the GATT dispute settlement ommendation after it has been issued. There is now procedure almost never employed retaliation as an a procedure for establishing a time limit for com- enforcement device. The fact that the GATT pliance, which, so far, has ranged from 6 to 15 nonetheless produced a large number of successful months. During that period, the illegal measure is legal rulings indicates, therefore, that the internal under periodic review, and it remains under peri- government forces just described frequently did odic review, without further action by the com- play a significant role in bringing about successful plainant, as long as noncompliance lasts. For the outcomes. This is not to say, of course, that enforce- government that does not, or cannot, retaliate, ment would not have been even more effective if these changes make it easier for the complainant to more retaliation had been employed. Other things focus and maintain community pressures for com- being equal, one would expect a better chance of pliance. compliance with a retaliation tool than without it. Retaliation is still the final remedy for eventual The key point, however, is that a legal ruling with- noncompliance. In contrast to the GATT disputes out retaliation can still be an effective policy tool for procedure, under which retaliation was a vague and a developing country seeking to reverse a legal vio- seldom-used remedy, the new WTO procedure lation by a larger country. Although not invariably appears to make retaliation the central objective of 83 T H E W O R L D T R A D E O R G A N I Z AT I O N the remedy structure. The defendant no longer has utility of the dispute settlement procedure for a the power to veto retaliation requests, making it cer- developing country complainant. tain that retaliation will be authorized whenever noncompliance is established. The use of retaliation Proposals for Additional Reforms to Cure is subject to time-limited procedures to resolve dis- the Imbalance putes over whether the defendant has in fact failed to comply with the ruling, and over the amount and This section contains background information nature of the retaliation, but these procedures only about some of the reforms that have been advanced delay the remedy slightly. in the past to cure the perceived imbalance in GATT In the light of past experience, the WTO's greater and WTO dispute settlement procedures. emphasis on retaliation as an enforcement tool would appear to be somewhat misguided. The Compensation for Harm Done emphasis on retaliation seems to have been the result of an effort to accommodate pressures for For most of the GATT's history, the prevailing view stronger enforcement. The U.S. negotiators had to has been that the only remedy for violation of a persuade the U.S. Congress that the WTO dispute legal obligation is a forward-looking order directing settlement procedure had strong enforcement pow- the defendant to comply in the future. Except for a ers. To do this, the negotiators had to satisfy the leg- string of antidumping and countervailing duty islators' rather simplistic view that enforcement can (AD/CVD) cases in the late 1980s and early 1990s, be achieved by retaliation. As is usually the case defendant governments have not been required to when negotiators try to convince domestic legisla- compensate for harm done before the illegal mea- tures to support a new trade agreement, the sure was brought into conformity. Uruguay Round negotiators probably promised The most important challenge to the exclusively more enforcement power from retaliation than forward-looking view of GATT remedies was a 1965 retaliation can deliver. Threats of retaliation can be effort by GATT developing countries to add mone- useful, but they can also become counterproductive tary compensation to the list of dispute settlement if used too forcefully or too often. Governments remedies. This occurred at a time when developing must remember that enforcement is a more com- country GATT members were using the threat of plex process than mere retaliation, involving the abandoning the GATT for UNCTAD to ask for a bet- generation of the political forces needed to bring ter deal from GATT. The developing country caucus about the desired compliance decision. made several proposals to improve the operation of The new emphasis on retaliation probably makes the dispute settlement procedure. One proposal the WTO dispute settlement system even more one- resulted in the adoption of a special accelerated pro- sided than before. Retaliation by larger countries cedure for complaints by developing countries, tends to be most effective when used against smaller which is still in force today.4 Two other proposals countries, and so, by making access to retaliation concerned the improvement of remedies: a proposal more available, these new reforms give larger coun- for monetary damages to be paid to developing tries a still greater advantage over smaller countries countries injured by GATT-illegal trade restrictions, that cannot effectively retaliate. (Needless to say, the and a proposal for collective retaliation.5 increased one-sidedness would be viewed as an The theory behind the developing country pro- advantage by industrial country governments.) posal for monetary compensation was that GATT- Once again, however, it must be stressed that the illegal trade restrictions caused serious harm to the greater one-sidedness of the procedure does not fragile economies of developing countries and that mean that legal complaints by developing coun- this harm would be multiplied by its retarding effect tries--that is, legal complaints without the retalia- on the development process. In these circum- tion option--cannot be a useful and effective policy stances, the developing countries argued, forward- tool. To the contrary, the smaller reforms made in looking remedies were not enough to remedy the the Uruguay Round do make legal complaints with- harm already done. Instead, they proposed, devel- out retaliation quite a bit more effective than they oping countries should be entitled to collect were before. One-sidedness is a problem, but it is a retroactive damages in the form of money awards. separate problem that has nothing to do with the The money would compensate the government's 84 The Adequacy of WTO Dispute Settlement Remedies: A Developing Country Perspective economic development program, rather than pri- validity of the proceedings under review. Finally, it vate interests, which would have removed many of must be recognized that the disreputable character the problems in calculating the harm. It was under- of AD/CVD measures makes them a natural target stood that the obligation to extend monetary com- for aggressive regulation. pensation would remain in force until the measure At first, it seemed that GATT governments them- was corrected. selves agreed that a refund remedy should be avail- Developing countries strenuously advocated the able in such cases. The first such ruling, in a 1985 money damages proposal through a long series of case brought by Finland against New Zealand, was committee meetings in 1965. Industrial countries adopted by the GATT Council, and New Zealand opposed the proposal with equal conviction, assert- did in fact issue a refund.8 Afterward, six subse- ing that money damages were simply outside the quent GATT panel decisions ordered refunds, but in realm of the possible. In effect, they were saying, the each case the result was inconclusive. Two of these GATT was never meant to be taken that seriously. panel reports were adopted, but only after the issues The proposal was not adopted. had become moot, and over the express reservation The GATT practice of denying compensation for of the defendant as to the panel's rulings on past wrongs clearly reflects a view of GATT law as refunds.9 The other four rulings were blocked having a lower status than domestic law. Under the entirely by the defendant government, with at least domestic law of most GATT members, taxes or part of the objection to adoption being the remedy other charges imposed in violation of national law order.10 The principal opponent of such refund are a legal nullity, and government authorities are orders was the United States, joined later by the required to refund any monies so collected. The European Community (EC).11 trade laws of such countries similarly provide for Although governments renegotiated many provi- refund of tariffs collected in violation of national sions of the GATT Antidumping Code and the tariff law. To my knowledge, however, few if any GATT Subsidies Code in the Uruguay Round, the countries authorize such refunds when tariffs or negotiations yielded no answer to the impasse over other charges are found GATT-illegal.6 By limiting refund orders. The United States then cast its posi- GATT law in this fashion, governments are saying tion in cement when the U.S. Congress adopted a that they do not want GATT legal obligations to statutory provision, in the 1994 legislation imple- have such direct legal effect. Ultimately, it is a state- menting the Uruguay Round agreements, that ment that governments do not want (or do not have AD/CVD or safeguards duties already paid in "liq- sufficient political support) to make trade agree- uidated" entries would not be refunded, although ments that binding. the GATT-illegal duties could be revoked for all The one exception to the GATT's consistent prac- "unliquidated" entries.12 The issue of AD/CVD tice of issuing only forward-looking remedies was a refund orders has come up only once so far under series of GATT panel decisions between 1985 and the new WTO dispute settlement procedure, in the 1995, all involving antidumping and countervailing Guatemala Cement case.13 The panel in that case did duties, in which panels ordered refunds of duties not rule on the issue, however, and its entire opin- imposed in violation of GATT rules. It is not clear ion was set aside by the Appellate Body on other why these panels singled out antidumping and grounds. countervailing duties for more demanding remedies In conclusion, it bears repeating that the call for than those normally employed against other GATT- refund of GATT-illegal antidumping and counter- illegal charges on imports. One GATT panel vailing duties is an exception to the perfectly consis- referred to provisions in the 1979 Antidumping tent GATT practice of denying refunds of Code requiring government to refund overcharges. GATT-illegal tariffs and all other kinds of GATT- This obligation applies only to overcharges as illegal charges. One evident reason for the absence defined by national antidumping laws, and national of a refund remedy has been that many govern- governments do seem to comply with it when they ments have lacked domestic legal authority to find that, under national law, overcharges have been refund taxes or charges in such cases. Given the made.7 Furthermore, AD/CVDs rest entirely on usual controls on contingent government expendi- specific proceedings against specific firms