GLOBAL AGRICULTURAL Tradeand Developing Countries Editors M. Ataman Aksoy · John C. Beghin THE WORLD BANK GLOBAL AGRICULTURAL TRADE AND DEVELOPING COUNTRIES GLOBAL AGRICULTURAL TRADE AND DEVELOPING COUNTRIES Editors M. Ataman Aksoy and John C. Beghin THE WORLD BANK Washington, D.C. © 2005 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 Telephone 202-473-1000 Internet www.worldbank.org E-mail feedback@worldbank.org All rights reserved. 1 2 3 4 07 06 05 04 The findings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of the World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, 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 pubrights@worldbank.org. Library of Congress Cataloging-in-Publication Data Global agricultural trade and developing countries / editor M. Ataman Aksoy, John C. Beghin. p. cm. ­ (Trade and development) Includes bibliographical references and index. ISBN 0-8213-5863-4 1. Produce trade--Developing countries. 2. Produce trade--Government policy--Developing countries. 3. International economic relations. I. Aksoy, M. Ataman, 1945- II. Beghin, John C. (John Christopher), 1954- III. Trade and development series HD9018.D44G565 2004 382'.41'091724--dc22 2004058811 Contents Acknowledgments xiii Contributors xv Acronyms and Abbreviations xvii 1. INTRODUCTION AND OVERVIEW 1 M. Ataman Aksoy and John C. Beghin Part I. Global Protection and Trade in Agriculture 2. THE EVOLUTION OF AGRICULTURAL TRADE FLOWS 17 M. Ataman Aksoy 3. GLOBAL AGRICULTURAL TRADE POLICIES 37 M. Ataman Aksoy 4. THE IMPACT OF AGRICULTURAL TRADE PREFERENCES, WITH PARTICULAR ATTENTION TO THE LEAST-DEVELOPED COUNTRIES 55 Paul Brenton and Takako Ikezuki 5. EXPERIENCE WITH DECOUPLING AGRICULTURAL SUPPORT 75 John Baffes and Harry de Gorter 6. AGRO­FOOD EXPORTS FROM DEVELOPING COUNTRIES: THE CHALLENGES POSED BY STANDARDS 91 Steven M. Jaffee and Spencer Henson 7. GLOBAL AGRICULTURAL REFORM: WHAT IS AT STAKE 115 Dominique van der Mensbrugghe and John C. Beghin Part II. The Commodity Studies 8. SUGAR POLICIES: AN OPPORTUNITY FOR CHANGE 141 Donald O. Mitchell 9. DAIRY: ASSESSING WORLD MARKETS AND POLICY REFORMS: IMPLICATIONS FOR DEVELOPING COUNTRIES 161 Tom Cox and Yong Zhu 10. RICE: GLOBAL TRADE, PROTECTIONIST POLICIES, AND THE IMPACT OF TRADE LIBERALIZATION 177 Eric J. Wailes v vi Contents 11. WHEAT: THE GLOBAL MARKET, POLICIES, AND PRIORITIES 195 Donald O. Mitchell and Myles Mielke 12. GROUNDNUT POLICIES, GLOBAL TRADE DYNAMICS, AND THE IMPACT OF TRADE LIBERALIZATION 215 Ndiame Diop, John C. Beghin, and Mirvat Sewadeh 13. FRUITS AND VEGETABLES: GLOBAL TRADE AND COMPETITION IN FRESH AND PROCESSED PRODUCT MARKETS 237 Ndiame Diop and Steven M. Jaffee 14. COTTON: MARKET SETTING, TRADE POLICIES, AND ISSUES 259 John Baffes 15. SEAFOOD: TRADE LIBERALIZATION AND IMPACTS ON SUSTAINABILITY 275 Cathy A. Roheim 16. COFFEE: MARKET SETTING AND POLICIES 297 John Baffes, Bryan Lewin, and Panos Varangis INDEX 311 ANNEX CD-ROM Baris Sivri List of Boxes 2.1 Role of Demand and Changes in Market Share 24 4.1 The EU System of Entry Prices: The Example of Tomatoes 62 10.1 Definitions of Rice Trade Flows in This Study 178 13.1 The European Union's Entry Price Scheme: Hindering Cost-Based Competition in the EU Market 247 13.2 The U.S.-Brazilian Trade Dispute over Orange Juice 249 13.3 Peruvian Asparagus Exports--A Standard Success Story? 254 15.1 Impacts of Trade Liberalization in Uganda's Fishing Industry 282 15.2 Foreign Fishing Access Agreements Involving Mauritania 284 16.1 Coffee Supply Controls in the Twentieth Century 303 List of Figures 2.1 Ratio of Farm Household Income to Nonfarm Household Income for Selected Developing Countries, Various Years 19 2.2 Ratio of Farm Household Income to All Household Income for Selected High-Income Countries 19 2.3 Ratio of Farm Income to Total Income of Farm Households, Selected Countries and Years 21 3.1 Market Price Support and Average Tariffs for Selected OECD Countries 40 3.2 Nominal Rates of Agricultural Support in OECD Countries 1965­2002 41 3.3 Rates of Agricultural Support in OECD Countries and Real U.S. Agricultural Price Index 42 3.4 Average Most-Favored-Nation Applied Tariffs for Agricultural and Manufacturing Products in Developing Countries, 1990­2000 43 Contents vii 3.5 Non-Ad-Valorem Tariff Lines as a Share of Total 45 3.6 Share of Output under Tariff Rate Quotas 50 4.1 The Value of Preferences Requested under GSP and AGOA Programs of the United States, as a Share of Agricultural Export to the United States 67 4.2 The Value of Preferences Requested under Cotonou and GSP Programs of the EU, as a Share of Agricultural Export to the EU 68 4.3 The Value of Preferences for LDCs under the GSP Program of Japan, as a Share of Agricultural Export to Japan 69 7.1 Output Structure in Base Year, 1997 122 7.2 Change in Rural Value Added from Baseline in 2015 124 7.3 Percentage Change in Rural Value Added from Baseline in 2015 125 7.4 Welfare Impacts of Productivity Changes 130 7.5 Sugar Output in Europe 132 7.6 Real Income and Trade Elasticities 135 7.7 Exports and Trade Elasticities 135 8.1 World Sugar Exports and Net Imports of Selected Countries (million of tons) 142 8.2 U.S. Sugar and HFCS Consumption 143 8.3 Sugar Prices, 1970­2003 152 8.4 Japanese Sugar Trends, 1970­2000 154 8.5 U.S. Sugar Loan Rates, U.S. Prices, and World Prices, 1980­2002 157 10.1 World Rice Trade and Share of Total Use 179 11.1 Wheat Yields, U.S. and India, 1900­2000 197 11.2 Per Capita Food Consumption of Wheat 198 11.3 Wheat Ending-Stocks vs. Prices 199 11.4 U.S. Wheat Price 199 11.5 U.S. Wheat Food Aid vs. Prices 203 11.6 Global Wheat Imports 207 11.7 Wheat Net Imports, Average for 1990­2000 207 11.8 Wheat Producer Prices in 2001 for Selected Countries 210 11.9 Emerging Wheat Net Exports of Emerging Exporters in the FSU 210 12.1 Global Groundnut Consumption, Exports, and Market Shares 219 12.2 Unit Price of Raw Edible Groundnuts Produced in The Gambia, Senegal, and South Africa 221 12.3 Rotterdam Prices of Groundnuts, 1970­2000 221 12.4 U.S. Domestic Groundnut Prices, 1993­2003 225 13.1 Production of Fruit and Vegetables by Region 238 13.2 Annual Growth Rates of World Imports of Selected Fruits 239 13.3 Annual World Import Growth Rates of Selected Vegetables 240 13.4 Annual Growth Rates of World Import of Major Processed Fruits and Vegetables 240 13.5 World Unit Values for Fresh Fruits, Fresh Vegetables, and Prepared Vegetables 241 13.6 Import Penetration Ratios in U.S. Fruit and Vegetable Markets, 1970­2001 243 13.7 The Hierarchy of Preferences in the European Fruits and Vegetables Market 250 13.8 Value of Fruit and Vegetable Preference for Major ACP Exporters, 2002 251 13.9 Value of Fruit and Vegetable Preference under AGOA 252 13.10 EU­Third Country Imports and the Share of Key Exporters 253 14.1 Cotton's Share in Total Fiber Consumption and Polyester to Cotton Price Ratio, 1960­2002 260 15.1 Fish Catches by Leading Countries, 1991­2000 276 15.2 World Aquaculture Production of Shrimp, by Volume, 2000 277 15.3 World Aquaculture Production by Value, 1991­2000 278 viii Contents 15.4 Food Fish Exports by Top Countries, 2000 279 15.5 World Food Fish Exports by Value of Major Commodity Group, 2000 280 15.6 Tariff Structure by Level of Processing, (1998­2001) 285 16.1 (a) Nominal Coffee Prices, 1990­2003 301 (b) Real Coffee Prices, 1960­2003 301 (c) Nominal Price Indexes for Coffee and Other Commodities, 1990­2002 302 (d) Real Price Indexes for Coffee and Other Commodities, 1960­2003 302 List of Tables 2.1 Distribution of Poor People in Developing Countries, 1999 18 2.2 Rural Population and Poverty for a Sample of 52 Developing Countries 18 2.3 Structure of Rural Household Incomes, Selected Developing Countries 20 2.4 Urban and Rural Income Inequality, Selected Countries and Years 21 2.5 Average Annual Real Export Growth Rates, 1980s and 1990s 22 2.6 Shares of Developing and Industrial Countries in World Exports, 1980­81 to 2000­01 23 2.7 Changes in Agriculture Price Indices, 1980s and 1990s 23 2.8 Average Annual Agricultural Output Growth Rates, 1980s and 1990s 23 2.9 Global Agricultural Trade Flows 25 2.10 Agricultural Flows (excluding Intra-EU and Intra-NAFTA Trade), 1980­81 to 2000­01 26 2.11 Agricultural Trade Flows of Developing Countries, by Groups, 1980­81 to 2000­01 27 2.12 Annual Import Growth Rates for Four Classifications of Agricultural Products, 1980s and 1990s 29 2.13 The Structure of Agricultural Exports, 1980­81 to 2000­01 30 2.14 Share of Agricultural Final Products in Exports, 1980­81 to 2000­01 31 2.15 Export Shares by Level of Processing, 1980/81 to 2000/01 32 2.16 Export Shares by Product and Region, 1980­81 to 2000­01 33 3.1 Percentage of Farm Gate Prices Attributable to Border Protection and Direct Subsidies by Country and Group, 1986­2002, Evaluated at World Prices 41 3.2 Agricultural Protection Rates in Selected Developing Countries 43 3.3 Agricultural Support in OECD Countries, 2002­02 44 3.4 Average Ad Valorem and Specific Duty Rates 46 3.5 Proportion of Non-Ad-Valorem Tariff Lines by Degree of Processing 46 3.6 Average Agricultural Tariffs, Selected Country Groups and Years 47 3.7 Tariff Peaks and Variance in Selected Countries 48 3.8 Tariff Rate Escalation in Agriculture, Selected Country Groups and Years 48 3.9 Tariff Escalation in Selected Agricultural Product Groups 49 3.10 Tariffs in the European Union and the United States Before and After Average Reduction from Applied Tariffs under the Harbinson Proposal 51 3.11 Tariffs in Selected Developing Countries Before and After Average Reductions from Bound Rates 51 4.1 Average Unweighted Tariffs on Agricultural Products in the United States, 2003 59 4.2 Average Unweighted MFN Tariffs on Agricultural Products Covered by GSP and AGOA in the United States, 2003 59 4.3 Number of Agricultural Tariff Lines Liberalized under GSP and AGOA Programs in the United States, 2003 60 4.4 Average Unweighted Tariffs on Agricultural Products in the European Union, 2002 61 Contents ix 4.5 Average Unweighted MFN Tariffs on Agricultural Products Covered by GSP and Cotonou Agreement, 2002 63 4.6 Number of Tariffs Lines Liberalized under EU Preferences for ACP Countries, 2002 63 4.7 Average Unweighted Tariffs on Agricultural Products in Japan, 2002 64 4.8 Average Unweighted MFN Tariffs on Agricultural Products Covered by GSP in Japan, 2002 64 4.9 Tariffs Lines Liberalized under Japan's GSP Preferences, 2002 64 4.10 Exports to the United States under AGOA and by other LDCs under the GSP, 2002 65 4.11 Exports to the European Union from ACP Beneficiaries, 2002 66 4.12 Exports to Japan from LDCs in 2002 66 5.1 Chronology of Broader Decoupling and Recoupling Episodes, 1985­2004 79 5.2 Composition of Agricultural Support in the United States, 1986­88 to 1999­2001 80 5.3 Composition of Agricultural Support in the European Union, 1986­88 to 1999­2001 82 5.4 Composition of Agricultural Support in Mexico, 1986­88 to 1999­2001 84 5.5 Composition of Agricultural Support in Turkey, 1986­88 to 1999­2001 85 6.1 Costs of Compliance with Export Food Safety Requirements in the Shrimp Processing Industry in Bangladesh and Nicaragua 98 6.2 Estimated Value of World Agricultural and Food Trade Directly Affected by Import Border Rejections Based on Technical Standards, 2000­01 102 6.3 Number of Counter-Notifications to the Sanitary and Phytosanitary Committee Relating to Reported Measures, 1995­2002 104 7.1 Trends in Agriculture, 2000­15 117 7.2 Trends in Processed Foods, 2000­15 117 7.3 Real Income Gains and Losses from Global Merchandise Trade Reform: Change from 2015 Baseline 119 7.4 Agricultural Output Gains and Losses from Global Merchandise Trade Reform: Change from 2015 Baseline 121 7.5 Real Income Gains from Agricultural and Food Trade Reform: Change from 2015 Baseline 121 7.6 Impact of Global Agricultural and Food Reform on Agricultural and Food Trade: Change from 2015 Baseline 123 7.7 Impact of Global Agriculture and Food Reform on Agricultural Employment and Wages: Change from 2015 Baseline 126 7.8 Impact of Global Agricultural and Food Trade Reform on Agricultural Capital: Change from 2015 Baseline 127 7.9 Impact of Global Agriculture and Food Reform on Agricultural Land: Change from 2015 Baseline 128 7.10 Net Trade Impacts Assuming Lower Agricultural Productivity in Developing Countries 129 7.11 Impacts on Output Assuming Lower Agricultural Productivity for Developing Countries 129 7.12 Baseline Trends in Agriculture with Higher Agricultural Productivity in Middle-Income Countries 131 7.13 Baseline Trends in Food Processing with Higher Agricultural Productivity in Middle-Income Countries 131 7.14 Impact of Lower Land Supply Elasticities in Rest of South Asia and Sub-Saharan Africa 133 8.1 Average Costs of Producing Cane Sugar, Beet Sugar, and High-Fructose Corn Syrup by Categories of Producers, and Actual Sugar Prices, 1994­1999 144 8.2 Raw Sugar Produced Annually per Sugar Industry Employee, Selected Developing Countries 145 8.3 Major Sugar Producers, Net Exporters and Net Importers, 1999­2001 Average 146 x Contents 8.4 Government Support to Sugar Producers, 1999­2001 150 9.1 Dairy Import Quotas for the Developed Countries under the URAA (1000 tons) 166 9.2 Tariff Reductions for Dairy Products under the URAA, by Region (US$/ton for specific duties, percent for ad valorem tariffs) 167 9.3 Full Liberalization of Trade and Domestic Support: Changes from 2005 Baseline 172 10.1 Share of Calories from Rice by Region and Income Level, 2000 178 10.2 Leading Rice-Producing, -Consuming, -Exporting, and -Importing Countries 180 10.3 Net Rice Trade, 1982­2002 181 10.4 Schedule of Tariffs, Tariff Rate Quotas, and Quotas in Rice, 2002­03 Levels 185 10.5 Simulation Results for Rice Trade Liberalization Using RICEFLOW, 2000 187 11.1 Wheat Production, Trade, and Growth Rates 1989­91 to 1999­2001, by Region 196 11.2 Global Wheat and Wheat Products Exports, Selected Periods 197 11.3 Tariffs and Tariff Rate Quotas, by Country 200 11.4 Average Tariff Rates and Imports for Wheat and Wheat Products 202 11.5 Support to OECD Wheat Producers, 1999­2001 204 11.6 Major Wheat Exporters' Shares of Global Wheat Net Exports 205 11.7 Producer Support Estimates, 1986­88 and 2000­02 205 11.8 Percentage Change of Wheat Production, Area Harvested, Yields, and Net Exports of Major Exporters from 1990­95 to 1996­2001 206 12.1 Production, Use, and Export of Groundnuts, Average 1996­2001 216 12.2 Costs of and Revenues from Groundnuts in China and the United States 217 12.3 Value of Net Exports, by Groundnut Product, 1996­2000 218 12.4 Share of Groundnut Products in Total Merchandise Exports 220 12.5 U.S. Producer Support Prices for Groundnuts, 1993­94 to 1998­99 222 12.6 U.S. Aggregate Support to Groundnuts, 1986­88 to 2000­01 223 12.7 U.S. Edible Groundnut Tariff Rate Quota Allocation, 1995­2008 223 12.8 U.S. Over-Quota Tariffs, 1994­2008 224 12.9 U.S. Imports of Edible Groundnuts 224 12.10 Groundnut Trade Policy Distortions in Argentina, China, and India 225 12.11 Tariffs on Groundnut Products in The Gambia, Malawi, Nigeria, Senegal, and South Africa 227 12.12 Average Tariffs on Edible Unprocessed and Processed Groundnuts 228 12.13 Welfare Effects of Policy Scenarios, 1999­2001 Average 229 12.14 Impact of Different Liberalization Scenarios on Groundnut Trade and Prices 230 13.1 World Fruit and Vegetable Imports, 1980­2001 239 13.2 Average Annual Growth Rates in World Import Volumes, 1980­2001 241 13.3 Import and Export of Fruits and Vegetables by Region or Country 242 13.4 Concentration of Fresh Fruit and Vegetable Exports among Developing Countries, 2001 244 13.5 Concentration of Processed Fruit and Vegetable Exports among Developing Countries 245 13.6 Export Subsidy Expenditures for Horticultural Products 245 13.7 Applied MFN Tariffs for Fresh Fruit and Vegetables in the Quad Countries, 1999 and 2001 246 13.8 Percentage of Tariff Lines on Fresh Fruits and Vegetables in Selected OECD Countries by Tariff Levels 247 13.9 Percentage of Tariff Lines on Processed Fruits and Vegetables in Selected OECD Countries by Tariff Levels 248 13.10 Percentage of Tariff Lines at Different Levels in Selected Developing Countries, 2001 250 14.1 Cotton's Importance to Developing and Transition Economies, 1989­99 Average 262 14.2 Direct Government Assistance to Cotton Producers, 1997­98 to 2002­03 265 14.3 Government Assistance to U.S. Cotton Producers, 1995­96 to 2002­03 265 Contents xi 14.4 Estimated Effect of Removal of Distortions (Percentage Changer over Baseline) 269 15.1 Trade-Weighted Tariff Averages for Developing Countries' Fish Product Exports to OECD Countries, by Processing State 285 15.2 Simulated Changes in the Real Value of World Exports in 1995 Prices (percent) 289 15.3 Simulated Benefits from Tariff Reductions, by Country 289 15.4 Effects of Relaxing Trade Barriers 291 15.5 Effects of a Rise in the Price of Cultivated Fish on Aquaculture Output and Fisheries Catch If Feed Is Held Constant 292 15.6 Effects on Price and Quantities of Market Liberalization: Relaxing Border Measures in the Importing Country 292 16.1 The Changing Structure of the Coffee Market 298 16.2 Coffee Production, Selected Years 299 16.3 Coffee's Importance to Developing Countries, 1997­2000 Averages 304 16.4 Per Capita Coffee Consumption of Major Consumers, 1993­99 308 ACKNOWLEDGMENTS This book is a joint effort by the Prospects Group in Development Department that reviewed the manu- DEC and the Trade Group in PREM. Support has script and helped to improve it. been given by the trade group in DECRG and We also benefited from presentations and feed- through the Knowledge for Change (KCP) trust back at the 2003 World Bank ABCDE Conference funds. Supporting donors for KCP include Canada, in Paris; to the board of Executive Directors of the Finland, Norway, Sweden, Switzerland, the United World Bank; at the World Bank's international Kingdom, and the European Commission. trade workshop, the WTO, UNCTAD, FAO, the The completion of this book would not have 2003 World Outlook Conference at the OECD in been possible without the help of numerous col- Paris, the 2003 American Agricultural Economics leagues inside and outside of the World Bank. Col- Meetings in Montreal, the European Commission, leagues in the Development Prospects Group and the French Ministry of Agriculture; and at the Uni- throughout the Development Economics Vice Pres- versity of California at Berkeley. Outside the Bank, idency and the World Bank's operational units pro- we would like to particularly thank Bruce Babcock, vided critical help and feedback. Support by the Pierre Bascou, Jean Christophe Bureau, Tassos former and current Chief Economists, Nicholas Haniotis, Chad Hart, David Roland-Holst, Daniel Stern and Francois Bourguignon was instrumental. Sumner, Peter Timmer, and Pat Westhoff for discus- Bernard Hoekman supported this project in all its sions and comments that helped to shape our views. stages, and without his support this book would Finally, we would like to thank Baris Sivri who not have happened. We are particularly grateful for carried out most of the data work for the book, to the ideas and insights of Uri Dadush, Hans Meta de Coquereaumont and Steven Kennedy for Timmer, Richard Newfarmer, Will Martin, Yvonne editing the manuscript and making it readable, to Tsikata, John Redwood, Kutlu Somel, Tercan Awatif Abuzeid and Cathy Rollins for preparing Baysan, and especially to John Nash, Kevin Cleaver, the manuscript in record time, and to Santiago Sushma Ganguly, Cornelis Van Der Meer, and Pombo-Bejarano and Mary Fisk for managing the their colleagues in the Agricultural and Rural publishing process. xiii CONTRIBUTORS M. Ataman Aksoy has recently retired from the Myles Mielke is Senior Commodity Specialist in Prospects Group of Development Economics at Basic Foodstuffs Service at Commodities and the World Bank and is now a Consultant at the Trade Division of the Food and Agriculture World Bank in Washington, D.C. Organization of the United Nations in Rome, John Baffes is Senior Economist in the Prospects Italy. Group of Development Economics at the World Donald O. Mitchell is Lead Economist in the Bank in Washington, D.C. Prospects Group of Development Economics at John C. Beghin is Professor, and Martin Cole the World Bank in Washington, D.C. Endowed Chair for the Department of Econom- Cathy A. Roheim is Professor of Economics in the ics and Center for Agricultural and Rural Devel- Department of Environmental and Natural opment at Iowa State University in Ames. Resource Economics at the University of Rhode Paul Brenton is Senior Economist in the Trade Island. Department of the World Bank in Washington, Mirvat Sewadeh is Consultant in the Trade D.C. Department of the World Bank in Washington, Tom Cox is Professor in the Agricultural and D.C. Applied Economics Department at the Univer- Baris Sivri is Consultant in the Development sity of Wisconsin in Madison. Economics Group of the World Bank in Ndiame Diop is Economist in the Trade Depart- Washington, D.C. ment of the World Bank in Washington, D.C. Dominique van der Mensbrugghe is Lead Econo- Harry de Gorter is Associate Professor of Agricul- mist in the Prospects Group of Development tural Economics at Cornell University in Ithaca, Economics at the World Bank in Washington, New York. D.C. Spencer Henson is Professor of Economics in the Panos Varangis is Lead Economist in the Agricul- Department of Agriculture, Economics, and ture and Rural Development Department of the Business at the University of Guelph in Canada. World Bank in Washington, D.C. Takako Ikezuki is Junior Professional Associate Eric J. Wailes is the L. C. Carter Professor in the Economist in the Trade Department of the Department of Agricultural Economics and World Bank in Washington, D.C. Agribusiness at University of Arkansas in Steven M. Jaffee is Senior Economist in the Trade Fayetteville. Department of the World Bank in Washington, Yong Zhu is a Research Associate in the Agricul- D.C. tural and Applied Economics Department at Bryan Lewin is Consultant in the Agriculture and University of Wisconsin in Madison. Rural Development Department of the World Bank in Washington, D.C. xv Acronyms and Abbreviations ABARE Australian Bureau of Agricultural FSU Former Soviet Union and Research Economics FTA free trade agreement ACP Africa, Caribbean, and Pacific GAP good agriculture practice ACPC Association of Coffee-Producing GATT General Agreement on Tariffs Countries and Trade AFIPEK Kenya Fish Processors and GBC Guatemalan growers association Exporters Association GDP gross domestic product AGOA African Growth and GSP generalized system of preferences Opportunity Act GTAP Global Trade Analysis Project AGRM Arkansas Global Rice Model HACCP hazard analysis and critical AMAD Agricultural Market Access control point Database HFCS high-fructose corn syrup APEC Asia Pacific Economic HS harmonized system Cooperation IADB Inter-American Development APTA Andean Trade Preference Act Bank BAAC Bank for Agriculture and ICAC International Cotton Advisory Agricultural Cooperatives Committee BULOG Nacional Logistics Agency ICO International Coffee Organization (Indonesia) LDC least-developed countries CAP Common Agricultural MAFF Ministry of Agriculture, Forestry, Policy (EU) and Fisheries (Japan) CBERA Caribbean Basin Economic MFN most-favored nation Recovery Act MPS market price support CIMMYT Internacional Maite and Wheat NAFTA North American Free Trade Improvement Center Agreement CMO Common Market Organization ODA official development assistance Conasupo Compania Nacional de OECD Organisation for Economic Subsistencias Populares (Mexico) Co-operation and Development CTE Committee on Trade and the OPEC Organization of Petroleum Environment (WTO) Exporting Countries EBA Everything But Arms (EU) PIPAA Integrated Program for EEZ exclusive economic zones Agricultural and Environmental EPA economic partnership agreements Protection EU European Union PROCAMPO Programa de Apoyos Directos al EVSL early, voluntary sector- Campo liberalization PROMPEX Peruvian Commission for Export FAIR Federal Agricultural Improvement Promotion and Reform Act (U.S.) PS&D Production supply and FAPRI Food and Agricultural Policy distribution Research Institute PSE producer subsidy equivalents FAO Food and Agriculture SACU Southern Africa Customs Union Organization of the United SAM social accounting matrix Nations SMP skim milk powder xvii xviii Acronyms and Abbreviations TCK tilletia controversa kuhn fungus USFDA United States Food and Drug TRQ tariff rate quota Administration UKP UzKhlopkoprom/ USGAO United States General Accounting UzPakhtasanoitish Office UNCTAD United Nations Conference on USDA United States Department of Trade and Development Agriculture UNEP United Nations Environment USITC United States International Trade Programme Commission URAA Uruguay Round Agreement on UW-WDM University of Wisconsin-Madison Agriculture World Dairy Model USAID United States Agency for VAT value-added tax International Development WTO World Trade Organization 1 INTRODUCTION AND OVERVIEW M. Ataman Aksoy and John C. Beghin In recent years, agricultural protection and its those leaving the farm, growth and modernization impact on developing countries have attracted of agriculture create jobs in agricultural processing growing attention. While manufacturing protec- and marketing, as well as the expansion of other tion has declined worldwide following substantial nonfarm jobs. reforms of trade policies, especially in developing Although most successful developing countries countries, most industrial and many developing have not relied on agriculture for export expansion countries still protect agriculture at high levels. and growth, growth in agriculture has a dispropor- Agricultural protection continues to be among the tionate effect on poverty because more than half of most contentious issues in global trade negotia- the populations in developing countries reside in tions, with high protection in industrial countries rural areas and poverty is much higher in rural being the main cause of the breakdown of the areas than in urban areas. Some 57 percent of the Cancún Ministerial Meetings in 2003. developing world's rural population lives in lower- middle-income countries, and 15 percent lives in the least-developed countries. Even though histo- Why Highlight Agriculture? rical trends show that agriculture's importance What happens in the global agricultural market is diminishes over time and the share of population important for developing countries beyond the in rural areas declines, there will still be more poor price changes triggered by global reforms.For coun- people in rural areas than in cities for at least a tries with a small urban population, increasing agri- generation. cultural exports can accelerate growth more than expanding domestic market demand can. Although Why This Book? food production for home consumption and sale in domestic markets accounts for most agricultural This book explores the outstanding issues in global production in the developing world, agricultural agricultural trade policy and evolving world exports and domestic food production are closely production and trade patterns. Its coverage of agri- related. Export growth contributes significantly to cultural trade issues ranges from the details of the growth of agriculture overall by generating cash cross-cutting policy issues to the highly distorted income for modernizing farming practices. For agricultural trade regimes of industrial countries 1 2 Global Agricultural Trade and Developing Countries and detailed studies of agricultural commodities general issues of competition, entry, and exit, which of economic importance to many developing coun- are major issues for products with distorted poli- tries. The book brings together the background cies, are equally important for the less-protected issues and findings to guide researchers and policy- traditional export products such as coffee, tea, and makers in their global negotiations and domestic cocoa. Exporters of such products still face long- policies on agriculture. The book also explores the term price declines, price volatility, and other key questions for global agricultural policies, both problems usually associated with products with dis- the impacts of current trade regimes and the impli- torted policy regimes. Seafood also faces fewer trade cations of reform. It complements the recent agri- distortions but is included as representative of the cultural trade handbook that focuses primarily on problems facing new, expanding sectors in the pres- the agricultural issues within the context of the ence of domestic subsidies in industrial countries. World Trade Organization (WTO) negotiations The commodity studies analyze the current (Ingco and Nash 2004). trade regimes in key producing and consuming The first part of the book replies to the broad, countries, document the magnitude of distortions cross-cutting questions raised by researchers and in these markets, and assess the distributional policymakers about agricultural trade regimes and impacts (across countries and across groups of con- trade performance. What has happened to the sumers, taxpayers, and producers within countries) structure of agricultural trade over the last two of trade and domestic policy reforms in developing decades? What is the level of protection across and industrial countries. These assessments are commodities and countries? Do tariff preferences based on rigorous quantitative analyses of various make a big difference in the levels of protection fac- reform scenarios and disaggregated partial equilib- ing developing-country agricultural products? Has rium models. The impacts of current agricultural the move toward decoupling agricultural support trade policies and of policy reforms vary substan- from production reduced the effects of agricultural tially across commodities, and different reforms support? Do stricter food safety standards consti- result in very different gainers and losers. tute a new barrier to market access by developing countries? How big are the potential gains from Some Key Findings global liberalization, and how sensitive are esti- mates to various assumptions? While these topics Despite the diversity of the cross-cutting analyses have been analyzed before, much of the work here and commodity studies, it is possible to draw some relies on new information. The answers to these general conclusions. First, these commodity mar- questions give a clearer picture of global agricul- kets exhibit a complex political economy, both tural policies and reforms. domestically and internationally. The arcane nature However, broad answers to these questions typi- of many policy interventions in these commodity cally do not convince the critics and, more impor- markets and the many heterogeneous interests exac- tant, provide little implementable guidance on erbate this complexity. Identifying superior policy specific policy issues. Micro details and partial options is not difficult, but the feasibility of reform equilibrium analyses at country and commodity depends on the power of vested interests and the levels are necessary to ensure that these broad ability of governments to identify tradeoffs and pos- results are credible and specific enough to be a basis sible linkages that will allow them to pursue multi- for policies. The second part of the book comple- ple goals (food security, income transfers, expansion ments the broad answers with detailed studies of of domestic value added) more efficiently. commodities that are of considerable economic Second, a narrow sectoral or product approach importance to many developing countries and that is unlikely to be fruitful in WTO negotiations. The are representative of the export bundle of develop- commodity studies illustrate why. They also illus- ing countries. The commodities selected are sugar, trate that potential tradeoffs exist even within agri- dairy, rice, wheat, groundnuts, fruits and vegeta- culture, as interests differ across commodities. bles, cotton, seafood, and coffee. Most of the prod- Third, and perhaps most important, the studies ucts selected have highly distorted policy regimes reveal the importance of microanalysis for identify- in industrial and some developing countries. The ing both the key policy instruments that distort Introduction and Overview 3 competition and the likely winners and losers from petitive producers from exit decisions, making global reforms (producers, consumers, taxpayers decoupling of these policies a moot point. If U.S. within and across countries). Knowing who is likely cotton subsidies were abolished, revenues for cot- to gain or lose from reform is critical for sequenc- ton farmers in West and Central Africa would ing reforms and putting in place complementary increase by some $250 million. Total official devel- policies, including assistance to reduce the cost of opment assistance (ODA) to the region in 1999 was adjustment in noncompetitive sectors. $1.9 billion, 15­25 percent of which typically goes Fourth, the studies identify trade distortions to agricultural assistance, not all of it directly reach- (border protection) and domestic subsidies as ing producers. One can see the incompatibility major factors affecting world markets and thus between ODA and farm policy in donor countries developing-country consumers and producers. A that subsidize their rich farmers. common theme is that border protection is more Fifth, a development strategy based on agricul- distorting in most markets, with the notable excep- tural commodity exports is likely to be impoverish- tions of cotton and seafood (corroborating the find- ing in the current agricultural policy environment ings of Hoekman, Ng, and Olarreaga 2002). Both in which policymakers in many countries have domestic subsidies and border protection con- mercantilist and protectionist reflexes that, when tribute to making commodity markets artificially aggregated, compromise world trade in agricultural thin, with small trade volumes and a small number and food products. The emergence of competitive of agents, in turn leading to high variability in prices producers in developing countries does not lead to and trade flows. Large trade distortions impede a rationalization of production among noncom- trade flows, depress world prices, and discourage petitive producers as it would in a liberalized market entry or delay exit by noncompetitive pro- market. Instead, noncompetitive producers remain ducers. Border barriers are high in most of the com- in business, buffered by extensive protection and modity markets studied (the exceptions are cotton, support. coffee, and seafood), including industrial countries and many developing countries. For example, the Potential Winners and Losers global trade-weighted average tariff for all types from Trade Liberalization of rice is 43 percent and reaches 217 percent for Japonica rice. Many Asian countries remain bas- Agricultural trade liberalization would create win- tions of protectionism in their agricultural and food ners and losers. The studies conclude that reform markets. would reduce rural poverty in developing eco- Subsidies have similar effects, depressing world nomies, both because in the aggregate they have a prices and inhibiting entry by inducing procyclical strong comparative advantage in agriculture and surplus production by noncompetitive (often large) because the agricultural sector is important for producers. In dairy and sugar markets, the effects of income generation in these countries. export subsidies have been smaller than those of Resource reallocation within agriculture would tariffs and tariff rate quota schemes, partly because be substantial. For example, production of ground- of the export subsidy disciplines introduced in the nut products in India would likely contract as would Uruguay Round Agreement on Agriculture. Many vegetable oil production in China, but dairy produc- domestic subsidies in Organisation for Economic tion and exports would expand in India, and rice Co-operation and Development (OECD) coun- production and exports would expand in China. tries, such as cotton subsidies in the United States, Liberalization of value-added activities is crucial for are countercyclical. expanding employment and income opportunities Domestic support and protection policies have beyond the farm gate. Such findings illustrate the substantial negative effects on producers in devel- importance of a multicommodity approach to oping countries, because of the sheer size of the reform, as gains and losses will differ by market. subsidies relative to the size of the market. Cotton They also illustrate the importance of social safety subsidies in the United States and European Union nets and other complementary policies. (EU), for example, reached $4.4 billion in a $20 bil- Consumers in highly protected markets will lion market. Such large subsidies shield noncom- benefit greatly from trade liberalization as domestic 4 Global Agricultural Trade and Developing Countries (tariff-inclusive) prices fall and product choice traditional agricultural commodity markets (sugar, expands. Consumers in poor, net-food-importing cotton, dairy, groundnuts, rice, and to a lesser countries could face higher prices if these markets extent, wheat) if trade and domestic distortions were not protected before liberalization, because of were removed, prospects of continuing high prices higher import unit costs. In practice, however, such are limited because of the nature of these markets concerns have often been exaggerated. For example, (a large number of low-cost competitors and dairy consumption in the Middle East and North inelastic demand). The bulk-commodity route to Africa would be little affected by trade liberaliza- export expansion requires low-cost conditions and tion because, while world prices would rise, high achievement of economies of scale. These markets import tariffs would be removed, so that the net face a long-term decline in prices as economies of impact on dairy consumer prices would be negligi- scale and competitive pressures yield lower costs ble. Consumer prices would rise for rice, however, and margins. Domestic farm subsidies in industrial since the removal of low tariffs would not offset the countries have exacerbated this low-price tendency increase in border prices. by fostering production beyond what free markets Other winners and losers would also emerge. would demand, with dramatic immiserizing conse- Multilateral trade liberalization erodes the benefits quences in some cases, such as cotton. from preferential bilateral trade agreements and Better opportunities exist in new markets such pits low-cost producers in some developing coun- as horticulture and seafood and in more differenti- tries (such as sugar producers in Brazil and ated products (niche coffee markets, confectionary Thailand) against less efficient producers in the peanuts). The high-quality differentiated-product least-developed countries who are currently helped alternative requires quality upgrades and the neces- by preferential access. The actual gains from such sary infrastructure and institutions to certify prod- preferences, however, have been smaller than ucts. These new markets imply increased costs to expected because of efficiency differences. meet quality standards and higher rewards. Pro- How these reforms occur will have important ducers have to be able to demonstrate quality, an consequences for developing countries. The best institutional challenge in many countries. This sec- approach is coordinated global liberalization of ond strategy can be successful only when supply policies. This approach would yield the largest price constraints are alleviated. Trade barriers also exist increases to offset some of the lost rents. For exam- in these new markets, especially with higher safety ple, world sugar price increases alone would offset standards. However, while the findings show that about half the lost quota rents, or about $0.45 bil- food safety standards are becoming more stringent, lion, for countries with preferential access. The the view that standards are simply new barriers to analysis shows that losses in rents would be much trade has been somewhat oversold. less than is commonly expected, because high pro- duction costs eat up much of the potential benefit What the Book Covers from preferential access to the high-price markets. Part 1 contains six chapters on cross-cutting issues, Moreover, the cost to the European Union and the and Part 2 includes nine commodity studies. While United States of each $1 in preferential access is the chapters in Part 1 are sequenced to provide a estimated at more than $5, a very inefficient way to detailed picture of cross-cutting issues in global provide development assistance. Global liberaliza- agricultural trade, they can be read individually as tion of primary commodity markets should be self-contained pieces. The accompanying CD-ROM accompanied by further effective opening of value- contains detailed supplementary tables and annexes. added markets, along with some targeted assistance to overcome supply constraints. Supply constraints Changes in Agricultural Trade Flows are particularly acute in Africa and some Latin American countries but are not insurmountable, as Chapter 2, "The Evolution of Agricultural Trade success stories in horticultural and seafood markets Flows," by Ataman Aksoy, gives a bird's-eye view of in Kenya show. the changes in global agricultural trade flows since Although the commodity case studies provide the early 1980s and contrasts these with the pro- evidence that higher market prices would prevail in gressive global integration of manufacturing.World Introduction and Overview 5 trade in agriculture, broadly defined throughout exports of temperate-climate products (grains, the book to include seafood, processed foods, and meats, dairy products, edible oils and seeds, and some agro-processing such as wine and tobacco animal feed) have surpassed exports of traditional products, was $467 billion in 2001­01, up from tropical products (coffee, tea, cocoa, textile fibers, $243 billion in 1980­81. During the 1980s real sugar, and nuts and spices). More important, manufacturing and agriculture exports expanded at exports of fruits and vegetables are now greater similar rates of 5.7 and 4.9 percent a year. However, than total exports of traditional products. Seafood during the 1990s real agricultural export growth exports are larger still, with a growing portion of decelerated to 3.7 percent a year, falling well behind exports coming from aquaculture. the 6.7 percent annual growth in manufacturing. Developing countries increased their share in State of Agricultural Protection manufacturing exports during the 1990s but saw little expansion in agricultural exports, barely main- Chapter 3, "Global Agricultural Trade Policies," by taining their share of around 36 percent after losing Ataman Aksoy, summarizes the state of agricultural market shares during the 1980s. All of their gains in protection, using data on domestic support policies agriculture during the 1990s came from expansion from the OECD and tariff data from the WTO for a of their exports to other developing countries. More large set of developing and industrial countries. than 48 percent of world agricultural trade is still The analysis of experience with the new rules on accounted for by trade between industrial coun- market access, export subsidies, and domestic sup- tries--about the same share as in 1980­81. port indicates that the effects of implementation of This stability of trade shares comes as a surprise, the Uruguay Round Agreement on Agriculture since it was during the 1990s that Uruguay Round have been modest. Within OECD countries, pro- commitments in agriculture began to be imple- ducer support in agriculture was about $230 billion mented and rapid trade reforms were introduced in in 2000­02, or almost 46 percent of production developing countries. More than a third of world value (evaluated at world prices), down from agricultural exports are traded within EU member approximately 63 percent in 1986­88, but still very nations and among the three signatories of the high. Of producer support, 63 percent came North American Free Trade Agreement (NAFTA). through higher prices associated with border pro- Low-income countries' agricultural trade tection (so-called Market Price Support or MPS) surpluses against both middle-income and indus- and 37 percent from direct subsidies. trial countries has increased. Low-income develop- While protection remained high in industrial ing countries now export more to middle-income countries, many developing countries have signifi- countries than they do to the European Union, their cantly liberalized their agricultural sectors since the largest export market in the early 1980s. The agri- early 1980s. Average agricultural tariffs, the main cultural trade surpluses of middle-income countries source of protection in developing countries, have diminished. Among industrial countries, Japan declined from 30 percent to 18 percent during the has the largest agricultural trade deficit (almost 1990s. In addition, these countries eliminated $50 billion in 2000­01); the European Union, once import restrictions, devalued exchange rates, aban- the largest net buyer of agricultural commodities, doned multiple exchange rate systems that penal- has seen its deficits decline; and NAFTA's trade sur- ized agriculture, and eliminated almost all export plus has shrunk considerably. Developing-country taxes. As overall taxation of agriculture declined regions, after losing market shares during the 1980s, in developing countries, reactive protection in regained most of them by the end of 1990s. The response to industrial-country support to agricul- only exception is Sub-Saharan Africa, which lost tural producers increased, especially in food prod- market shares during the 1980s and did not regain ucts. All these measures increased incentives for them during the 1990s. agricultural production in many developing coun- The structure of world trade has changed, espe- tries. However, without compensating reductions cially for developing countries. Nontraditional in protection in industrial and some middle- products, especially seafood and fruits and vegeta- income countries, the result was overproduction bles, now constitute almost half their exports. Also, (beyond competitive and undistorted market 6 Global Agricultural Trade and Developing Countries levels) and price declines for many commodities, countries, so tariff preferences are irrelevant. reducing opportunities for competitive developing Although duties on other primary agricultural countries to expand exports and rural incomes. products and processed products are often very The structure of agricultural tariffs is compli- high, few of these products receive preferences. cated and nontransparent. More than 40 percent of Nevertheless, for a small number of products the agricultural tariff lines in the European Union substantial preferences are available for certain and the United States contain specific duties, which countries, usually within strict quantitative limits. make it difficult to calculate average tariffs, obscure Countries that produce sugar and tobacco, for true levels of protection, and penalize developing example, have received large transfers as a result of countries that supply cheaper products. Specific these preferences. duties, which are rare in manufacturing, are also Comparison of different preference schemes is used to hide high rates of protection in agriculture. difficult because the schemes differ substantially. The ad valorem equivalents of specific duties, when They differ in the group of eligible countries, the they can be measured, are much higher than the products covered, the size of the preferences average ad valorem duties. Also, a much higher pro- granted, and administrative requirements, espe- portion of tariff lines in final products than in raw cially rules of origin. These differences are a major and intermediate products have specific rates. Low- weakness of the current system of preferences. Dif- income countries have more transparent tariff ferences between preference schemes constrain the regimes and tend to use ad valorem tariffs. ability of developing-country suppliers to develop Average agricultural tariffs in industrial coun- global market strategies. tries, when they can be measured, are some two to In general, preferences are unilateral concessions four times higher than manufacturing tariffs. by industrial countries. The agreements require Developing-country exports confront tariff peaks renewal, and specific products can be withdrawn at as high as 500 percent in some industrial countries. short notice. This uncertainty has impeded new High variance and high peaks make it difficult to investment. The most highly protected products, measure the real impact of protection on key prod- which would have the highest potential margins of ucts, whose high tariff rates are buried in lower preference, are often excluded or preferences are average tariffs. This is why the OECD measure of small. Rules of origin for processed products often protection, market price support, which compares constrain the ability of countries to expand into local and international prices, shows much higher these products. rates of protection than do average tariffs. Tariffs The value of preferences is largest in the EU mar- also increase by the degree of product processing, ket, driven mainly by the very high EU prices for creating an escalating tariff structure that impedes sugar. For some countries, such as Mauritius, prefer- access to processed food markets. In addition, ences seem to explain at least part of the relatively almost 30 percent of domestic production in OECD strong economic performance and economic diver- countries is protected by tariff rate quotas. sification. For the majority of low-income coun- tries, however, EU, Japanese, and U.S. preferences have had little impact and have done little to stimu- Trade Preferences late the export of a broader range of products. Industrial countries have established tariff prefer- ence schemes to create market access opportunities Decoupling Agricultural Support for developing countries, especially for low-income countries. In chapter 4,"The Impact of Agricultural One key challenge is to lower the effect of domestic Trade Preferences on Low-Income Countries," Paul subsidies on world production and prices. Although Brenton and Takako Ikezuki examine the impacts official export subsidies may be small and shrink- of these preferences. For most developing coun- ing, implicit export subsidies created by domestic tries, preferences have provided limited gains at support are increasing, lending unfair advantage to best. Many agricultural products exported from producers in industrial countries. More generally, developing countries, especially traditional tropical there is a move toward supporting agriculture products, are subject to zero duties in industrial through direct subsidies rather than through border Introduction and Overview 7 barriers. Some domestic support to agriculture has duction has continued. One-time buyouts have had moved away from being directly linked to produc- greater success in eliminating very inefficient tion to being partially decoupled, with payments arrangements, but their range is limited. More atten- made based on historical production levels and tion should be given to constraints on input use, other mechanisms. Decoupling should reduce the government credibility, other support programs, output effects of support and thus increase world and time limits. Unless these aspects are addressed, prices for the exports of developing countries. The decoupled support is likely to have the same kinds of move to decoupled agricultural support policies is undesirable effects as other subsidy programs. Pay- therefore a step in the right direction. ments should be time limited, provided only to help How much has the world actually moved to producers adjust. The European Union and Turkey decoupled payments? What has been the net effect have no time limit. The United States had (at least on resource use, efficiency, and trade distortions? In implicitly) a time limit in the 1996 Farm Bill but chapter 5, "Experience with Decoupling Agricul- violated it three years later. Mexico has a time limit tural Support," John Baffes and Harry de Gorter and has complied with it so far. evaluate the impact of decoupling measures in The coexistence of coupled and decoupled pro- industrial and developing countries. From 1986­88 grams means that incentives to overproduce to 2000­02, domestic subsidies paid to farmers in remain. In the four decoupling cases examined, all OECD countries increased 60 percent. Output and either left some coupled support programs in place input subsidies ("large"impact programs) increased or added new ones. Eligibility rules need to be fixed moderately compared with the substantial increases and clearly defined. Updating the bases for pay- in payments linked to land area or number of ani- ment of subsidies and adding crops results in a gov- mals, decoupled historical entitlements, or input ernment credibility problem and reduces the effect use and overall farm income ("smaller" impact pro- of the decoupling programs. grams). Payments based on area planted and num- ber of animals have increased most, followed by his- Food Product and Safety Standards torical entitlements. The United States took the first step toward With the decline in traditional barriers to trade, decoupling in the 1985 Farm Bill, which shifted the attention has focused on the potential role of stan- base of support from current yields to historical dards as technical barriers to trade. Zero-duty yields. In the 1996 Farm Bill the United States access means little if countries cannot meet prod- replaced deficiency payments with decoupled sup- uct standards. Chapter 6, "Agro-Food Exports from port. The European Union partially replaced inter- Developing Countries: The Challenges of Stan- vention prices with decoupled payments following dards," by Steven M. Jaffee and Spencer Henson, the Common Agricultural Policy reform of 1992. provides an overview of the impact of food safety Mexico replaced price supports with decoupled and agricultural health standards on developing payments in 1994 with the introduction of the country agro-food exports. Standards have become National Program for Direct Assistance to Rural an increasingly important influence on the interna- Areas (Programa de Apoyos Directos al Campo tional competitiveness of developing countries, [PROCAMPO]). More recently, Turkey replaced especially in the context of high-value agricultural some price supports and input subsidies with and food products. Some well-established sectors decoupled payments. In addition to broad decou- that are highly export dependent have been hurt by pling attempts, there have been numerous one-time new and stricter standards. In several cases, devel- buyouts, including New Zealand's exit grant in oping countries have faced restrictions because of 1984, the buyout of Canada's grain transportation their inability to meet food safety or agricultural subsidy in 1995, and the buyout of the U.S. peanut health requirements. At the same time, other devel- marketing quota under the 2002 Farm Bill. oping countries have gained access to high-value Experience designing and implementing these markets in industrial countries despite these programs has been mixed. Although decoupling has stricter standards. led to a reallocation of resources in agriculture, its The evidence in this chapter suggests a less pes- effects have been modest. In many cases, overpro- simistic picture for developing countries than that 8 Global Agricultural Trade and Developing Countries commonly presented, which sees standards as bar- competitiveness. Failing this, countries face the riers to developing-country trade. Rising standards need for potentially large-scale investments over accentuate underlying supply chain strengths and long periods of time to remedy violations of stan- weaknesses and thus have different effects on the dards as they arise. In all of this, the public and pri- competitive position of different countries. In this vate sectors need to work together to identify the perspective, food safety measures must be viewed most efficient and effective ways to develop capac- within the context of more general capacity ity. Food safety and agricultural health controls constraints. must be seen as a collaborative effort in a system Much of the impetus for stricter food safety and that is only as strong as its weakest link. agricultural health standards is coming from con- sumer and commercial interests, magnified by Welfare Gains from Global Agricultural Reform advances in technology and new security concerns. Thus prospects are slim for slowing this movement Given the magnitude of the distortions in agricul- or allowing poorer countries to meet lower stan- tural sectors in all countries, an obvious question dards. Developing countries need to find ways to concerns the net impact of status quo policies and develop and improve food safety and agricul- of global reform. Models of global trade and tural health management systems to meet these domestic policy reforms often yield very large standards. welfare gains for both industrial and developing A crucial need is for management capacity, not countries. Critics argue that many of the assump- only to comply with the different requirements in tions of these studies are exaggerated and that different markets but also to demonstrate that their results should be treated with caution. In compliance has been achieved. While many coun- chapter 7, "Global Agricultural Reform: What Is at tries have struggled to meet ever-stricter standards, Stake?" Dominique van der Mensbrugghe and John even some very poor countries have managed to C. Beghin look beyond the estimates of aggregate implement the necessary capacity, especially where welfare gains to structural changes that would the private sector is well organized and the public emerge from multilateral trade liberalization in sector supports the efforts of exporters. Many poor agricultural and food markets, including cross- countries have successfully entered the demanding regional patterns of output and trade. They address seafood and fresh fruit and vegetable markets. Most some of the common criticisms of these aggregate violations reported at border controls involve fail- models and explore the implications for welfare, ures to meet simple hygiene standards. trade, output, and value added of changing key There is no single model for all countries striv- modeling assumptions. The real gains often ing to meet the challenges posed by standards. amount to 1 percent or less of base income, Institutional frameworks are required, however, to whereas the structural changes (resource realloca- overcome the problems associated with being poor tion) can be greater than 50 percent. The chapter or small. These can include outgrower1 programs decomposes the impacts of partial reforms both for smallholder farmers, systems of training and regionally and across instruments to determine the oversight for small and medium-size enterprises share of the global gains that comes from reform in established through associations and other groups, industrial countries and the share from reform in and twinning and regional networking for small developing countries. It also examines the extent to countries. Such efforts undoubtedly need to be which border protection and various forms of improved and refined, but they offer useful guid- domestic support drive global gains. ance on effective ways to proceed. The second part of the chapter addresses some of The chapter clearly demonstrates the need for the issues raised by critics of trade reform--notably, developing countries to be proactive when facing that the estimated gains for developing countries are new food safety and agricultural health standards. too optimistic and that the transitional costs for By thinking strategically, countries can program industrial-country farmers are high and too often capacity enhancement into wider and longer-term ignored.The analysis looks at three assumptions that efforts to enhance domestic food safety and agri- could influence the level of gains: the consequences cultural health management systems and export of lowering agricultural productivity growth in Introduction and Overview 9 developing countries, the impact of constraining support to their producers. Thus 80 percent of output supply response in low-income countries, world production and 60 percent of world trade and the assumptions on the magnitude of trade take place at prices much higher than world prices. elasticities. The chapter also examines the impact of There are pressures on the European Union and lowering the rate of exit of industrial-country farm- the United States to reform their sugar markets ers, including adjustments to transition. because of internal market changes and interna- The results are broadly robust to the range of tional commitments already made under NAFTA, sensitivity analyses undertaken, but trade elastici- the EU Everything but Arms Program, and the ties are the most important. Assuming low produc- Uruguay Round Agreement on Agriculture. Their tivity gains in agriculture in developing countries protectionism is unravelling, another case of leads to a reversal in the estimated impact of global border opening forcing domestic policy discipline. liberalization for industrial countries, with an Needed reforms could be carried out in conjunc- increase in the net food trade surplus. If productiv- tion with scheduled reviews of the EU Common ity grows slowly in developing countries, they Agricultural Policy in 2006 and expiration of the become much larger importers of food and agricul- U.S. Farm Bill in 2007, which could provide a target tural products, and trade reform accentuates this period for getting reforms agreed on and in place. tendency. Low-income developing countries expe- Japan remains a bastion of protectionism, with tar- rience an increase in net food trade surplus that is iffs, price surcharges, and trade management by much smaller than under the higher productivity state agencies. assumption. Thus different assumptions about Preferential and regional agreements often bar productivity could lead to different conclusions low-cost producers from entering the internal mar- about the direction of food self-sufficiency in the kets covered by the agreements. Quota allocations aftermath of reform. Supply constraints do not are concentrated in a few, often high-cost countries, qualitatively affect the estimated impact of trade which are generally not the poorest. For example, reform on agricultural output, although estimated Mauritius has 38 percent of EU quotas. Thailand, a changes tend to be smaller. Higher trade elasticities very low-cost producer, is limited to a 15,000 ton dampen the adverse terms-of-trade shocks from quota in the United States, whereas the Philippines reforms, leading to larger income gains and higher has a quota 10 times larger that often goes unfilled. variations at the country level. Multilateral negotiations provide an opportu- nity to rationalize the proliferation of preferential agreements, by phasing in multilateral liberaliza- Commodity Studies tion and allowing markets to allocate access on a Nine chapters analyze the impact on global markets competitive basis. Reforms would result in a con- of policies for selected commodity groups. The traction of output in both industrial countries and commodity groups were selected to provide a beet-producing developing countries. World prices broad range of policy environments, to deal with would rise by about 40 percent. The big gainers different groups of countries, and to show the would be producers in Thailand, Latin America, diversity of gainers and losers. and southern Africa among developing countries and Australia among industrial countries. Con- Sugar Chapter 8,"Sugar:Opportunity for Change," sumers would gain in almost every country, since by Donald O. Mitchell, looks at the sugar market, even competitive producers cover their export one of the most distorted markets in the world. The losses with higher-price domestic sales. The losses European Union, United States, and Japan together to quota holders, many of them very high-cost pro- protect sugar at some $6.4 billion a year, about the ducers, would be much smaller because of the value of total developing-country exports. On aver- world price increases. age, domestic producers in these countries receive more than triple the world price for their output. Dairy In chapter 9, "Dairy: World Markets and Among middle-income countries, Mexico, Poland, the Implications of Policy Reform for Developing Turkey, and almost all beet-producing, northern Countries," Tom Cox and Yong Zhu analyze the developing countries also provide significant dairy market, which is the most distorted of all the 10 Global Agricultural Trade and Developing Countries markets examined in this volume. The sector is driven by demographics rather than by income distorted by a complex system of domestic and inter- growth. Prospects for growth in trade therefore rely national trade barriers, including surplus disposal in on policy reforms. the Quad countries (Canada, Japan, the European Tariff and related border protection are very Union, and the United States) and the Republic of high, averaging about 40 percent globally and rising Korea. OECD support totaled $41 billion in 2002, to 200 percent in some markets. Support in OECD and tariff rates are above 30 percent worldwide. The countries is almost $25 billion. Support in Japan, Quad countries and Australia and New Zealand expressed in ad valorem form, is a staggering dominate the export market. Although Australia and 700 percent of world prices. Tariff escalation is New Zealand are competitive exporters, with few systematically practiced (from paddy to milled rice) distortions, dairy interest groups in the Quad coun- in many countries. In the European Union the tries are strongly entrenched. Prospects for policy tariff on milled rice (80 percent) is prohibitive, reforms appear dim, especially in the European except for small preferential import quotas granted Union and Japan. Domestic price discrimination to a few countries. Tariff escalation is also prevalent schemes in the European Union, the United States, in Central and South America. Mexico has a 10 per- and Canada rely on the ability to close borders, sug- cent tariff on paddy rice and a 20 percent tariff gesting that the emphasis in the Doha Round negoti- on brown and milled rice. This pattern of protec- ations should be on commitments to lower border tion depresses world prices for milled high-quality protection. long grain rice relative to prices for brown and Despite high distortion levels, the global dairy rough rice, creating economic hardship for millers market is dynamic, with much growth potential. of high-quality long grain rice in exporting coun- Dairy consumption in Asia has been expanding dra- tries such as Thailand, the United States, and matically with income growth, urbanization, and Vietnam. the westernization of diets. Innovations in food pro- Net rice consumers would be negatively affected cessing also contribute to the sector's dynamism, by trade liberalization if the new consumer price with new value-added opportunities such as dry rises with reform. Prices would rise wherever cur- whey and lactose, for which trade barriers are low. rent ad valorem tariffs are lower than the potential Innovations have also expanded trade opportunities world price increase following liberalization, such for traditional milk products such as milk powder as in the Middle East. and butter-oil, which are transformed into final products after importation to circumvent protec- Wheat In chapter 11, "Wheat: The Global Mar- tion on finished products. Concentration and verti- ket, Policies, and Priorities," Donald O. Mitchell cal integration in industrial countries are also and Myles Mielke analyze the world wheat market, important sources of economies in procurement, which has become less distorted since 1990. A processing, and logistics and lead to high levels of number of countries have undertaken reforms uni- foreign direct investment. Global reforms could laterally or as a consequence of commitments raise prices by 20­40 percent and lead to production under the Uruguay Round. The European Union declines in the Quad countries and increases in and the United States have ended their export sub- Australia, New Zealand, Latin America, and India. sidies, but other surplus-disposal programs, such as nonemergency food aid and export credits, are still Rice In chapter 10, "Rice: Global Trade, Protec- in place. Most importing countries have reduced tionist Policies, and the Impact of Trade Liber- their tariffs on wheat or allowed duty-free imports alization," Eric J. Wailes analyzes rice, the most from regional trading partners and thus benefit important food grain in the world. On average, con- from low world market prices. A few importers, sumers in low-income food-deficit countries get such as Japan, continue with high levels of protec- 28 percent of their calories from rice. Production tion that raise internal prices to more than five and consumption are concentrated in Asia (China, times world market levels. India, and Indonesia). The rice market is a mature While wheat trade has become less distorted, market, with static demand in industrial countries tariff escalation is high. Tariffs on flour are well and growing demand in developing economies above those on wheat, and tariffs on bakery and Introduction and Overview 11 pasta products are even higher. Consequently, trade The policy dimension of international ground- in wheat products is confined largely to free-trade nut markets is a challenge largely for developing areas such as the European Union and NAFTA. countries. India and, to a lesser extent, China are A major concern for wheat-importing countries large, protected groundnut markets, and low-cost is the lack of assured access to wheat markets in producers in Argentina and Sub-Saharan Africa are periods of high prices. In the 1970s the United States potential gainers from global reforms. The United imposed an export embargo on wheat, to protect States, which once strongly supported the peanut domestic consumers from high world prices. In sector, eliminated major distortions with a one- 1995 the European Union imposed an export tax on time buyout in 2002, but a now-redundant tariff of wheat for a similar reason. Such actions increase 160 percent remains. Liberalization would make international price volatility and reinforce the desire India and China net importers of some peanut for self-sufficiency in importing countries. Import- products. With trade liberalization, the bulk of ing countries need to pressure exporting countries world welfare gains would occur with groundnuts for assured market access as part of the Doha Round rather than with derivative products, although lib- of multilateral trade negotiations. eralization of the value-added markets (groundnut OECD countries still provide substantial sup- oil and meal) would lead to larger welfare gains port to wheat producers, but the production effects and higher rural incomes for African countries have been partially offset by land set-aside pro- ($72 million in aggregate welfare and $124 million grams and by the way support is provided. Global in farm profits). Consumers in OECD countries liberalization is expected to raise world wheat would pay higher prices for these products, but prices by a relatively small amount (5­10 percent) there would be little effect on poverty. Consumers because of large surplus capacity in major in India and southern China, who pay for heavy exporters. This capacity could return to production and inefficient government intervention in the sec- following policy reforms, preventing prices from tor, would be better off. rising significantly. Big gainers would be Argentina, The major challenge in successful negotiations Kazakhstan, and Ukraine, with some output reduc- to open groundnut product markets is to overcome tion by the United States and the European Union. entrenched interests in India and China. Except for Further reforms of the global wheat market should the United States, industrial countries have limited focus on ensuring access to wheat exports during interests at stake in these markets and should not price spikes, reducing producer support in OECD be an impediment to reform. Moreover, U.S. pro- countries, reducing protection in the few remaining ducers would benefit from the higher world prices highly protected markets, and reducing tariff esca- that would prevail under free trade, helping to off- lation on wheat products. set reductions in U.S. tariffs. Groundnuts In chapter 12,"Groundnuts: Policies, Fruits and vegetables In chapter 13, "Fruits and Global Trade Dynamics, and the Impact of Trade Vegetables: Global Trade and Competition in Fresh Liberalization," Ndiame Diop, John C. Beghin, and and Processed Products," Ndiame Diop and Steven Mirvat Sewadeh analyze groundnuts, an important M. Jaffe look at another dynamic product group, product for many low-income producers and con- which now constitutes almost 21 percent of devel- sumers. There are two main groundnut markets, oping-country exports. World imports of fruits and one for edible groundnuts (confectionary, processed vegetables grew 2­3 percent a year during the 1990s, butter and paste) and one for crushed groundnuts a slowdown over the 1980s. Low population and (oil and cakes) used in livestock feed. The peanut oil income growth in the European Union, where prod- market is declining because of the availability of uct markets were already mature and saturated, had lower-priced vegetable oils, but the confectionary much to do with the slowdown.Adverse price move- nuts market is expanding. African producers have ments for fresh and processed products from the considerable potential in this sector, but supply mid-1990s onward also contributed to the decelera- volatility, inefficient processing, and uneven quality tion. Trade growth remained robust among NAFTA are challenges to their becoming dependable countries, for exports to high-income Asian coun- exporters of confectionary products. tries and for trade between developing countries. 12 Global Agricultural Trade and Developing Countries Although many developing-country suppliers especially following the technological improve- have entered this market, relatively few countries ments of the early 1970s that brought prices down have achieved significant success on a sustained to those for cotton. Since 1975 polyester and cotton basis. This is a highly competitive and rapidly have traded at roughly the same price levels. Cot- changing industry, with multiple influences on ton's share of total fiber consumption has dropped competitiveness. from 68 percent in 1960 to 40 percent in 2001­02. Unlike the case in many other agricultural sec- Cotton demand has grown at the same rate as pop- tors, production and export subsidies are not per- ulation growth during the last 40 years. vasive in horticulture. Border controls are the main The major challenge for cotton is to cut back instrument of protection. The United States, the support policies, particularly in the United States, European Union, and Japan use a range of complex which subsidized cotton at a cost of $3.7 billion in tools, including highly dispersed ad valorem tariffs, 2001­02, and the European Union (Greece and specific duties, seasonal tariffs, tariff escalation, and Spain), which provided subsidies of almost $1 bil- preferential access with tariff rate quotas. Many lion. These are extremely high subsidies in a market industrial countries have set up complex systems of in which production was valued at $20 billion in preferential access to provide a few privileged trade 2001­02. At this level of support, U.S. and EU cot- partners with favorable entry without undermining ton producers receive prices that are 87 percent and protection of domestic producers. The product 160 percent, respectively, above world prices. China coverage of preferential access schemes is wide, but has also supported its cotton sector. Many cotton- entry is often limited by quotas for "sensitive prod- producing developing countries have reacted to low ucts." Tariff escalation is widespread, although its world prices by introducing offsetting support. extent varies significantly across countries. Support in Brazil, Egypt, India, Mexico, and Turkey Further tariff liberalization would be needed to totaled $0.6 billion in 2001­02. reduce tariff peaks,especially in the European Union Cotton support policies reduce world prices by and the Republic of Korea. Changes in domestic some 10­15 percent, cutting the incomes of poor support will not affect the sector significantly farmers in West Africa and Central and South Asia. because most countries have low levels of direct gov- Cotton has important implications for poverty ernment intervention. Reductions in tariffs and reduction in these countries as it is one of the most other import restrictions are thus critical for deter- important sources of cash in these economies. If mining the impact of trade agreements and policies support were removed completely, Africa would on world horticultural trade. Still, as experience sug- increase production by 6 percent and Uzbekistan gests,the main beneficiaries of such reforms will be a by 4 percent, while the United States would reduce limited number of middle-income countries that production by 7 percent and the European Union have developed strong production, post-harvest by 10 percent. processing, logistical marketing, and sanitary and phytosanitary management systems and that con- Seafood In chapter 15, "Seafood: Trade Liberal- tinue to attract new investment. With few excep- ization and Impacts on Sustainability," Cathy A. tions, low-income countries still face substantial Roheim looks beyond global trade policies to supply-side challenges in taking advantage of exist- examine the complementary issues of management ing and future international market opportunities. and sustainability. Seafood is one of the most traded food commodities in the world. Developing Cotton In chapter 14, "Cotton: Market Setting countries account for more than 50 percent of the and Policies," John Baffes explores cotton, a market global fish product trade by value. This trade now with minimal border restrictions but considerable constitutes 20 percent of their agricultural and food domestic support. Cotton production is an impor- processing exports, more than tropical beverages tant source of rural income and exports in Africa (coffee, cocoa, and tea), nuts and spices, cotton, and and Central Asia. In 1998­99, cotton accounted sugar and confectionary combined. Aquaculture for more than 30 percent of merchandise exports has expanded to 30 percent of world seafood in Benin, Burkina Faso, Chad, Mali, Togo, and production. The most valuable component of the Uzbekistan, and 15 percent in Tajikistan. Cotton seafood trade is shrimp, with total world trade of faces intense competition from synthetic fibers, more than $10 billion in 2000. Introduction and Overview 13 Capture fisheries still supply the majority of fish try. Except in Brazil, Colombia, Ethiopia, and Mex- production, but 60 percent of the world's fisheries ico, little coffee is consumed in developing coun- are either overused or fully used. Even with the tries. Efforts to expand coffee consumption in establishment of the 200-mile exclusive economic developing countries are likely to come at the zones in 1977, which brought a third of the world's expense of tea, a commodity produced by the same oceans under the jurisdiction of coastal states, most countries that produce coffee. fisheries management plans have not achieved their Although a few large producers produce most of stated goal of maintaining sustainable fisheries. the coffee, several small countries depend heavily Most seafood product trade flows from develop- on coffee. In Burundi, Ethiopia, and Rwanda, coffee ing countries to industrial countries. In several accounts for more than half of total merchandize developing countries, fish products are a primary exports. The coffee market had supply controls in source of export earnings. Trade barriers may have place longer than any other important commodity. significant potential for harm for these countries. In addition to stabilizing (and perhaps raising) Among trade barriers, tariffs are low compared prices in the short term, these agreements brought with the effects of sanitary and phytosanitary new entrants into the coffee market. With the measures and, increasingly, countervailing and exception of Colombia, Ethiopia, and, to a lesser antidumping measures. Many industrial countries degree, Côte d'Ivoire, Kenya, and Tanzania, the heavily subsidize their fishing sector, including marketing regimes in coffee-producing countries buying access to the waters of developing nations. are liberal. Some 6­8 percent of coffee output is These subsidies and other fishing arrangements traded outside of traditional marketing channels, as mean that industrial countries capture a significant organic, fair-trade, gourmet specialty, and eco- portion of fishing value added. Many developing friendly products. These new markets provide countries do not have management policies or lack higher prices to producers. the resources to enforce them, with the result that During the 1990s, Brazil expanded its coffee capture fisheries are being depleted. Increased production to areas less subject to frost, reducing aquaculture production in developing countries, weather-induced supply disruptions. Vietnam particularly of shrimp, has had adverse environ- emerged as the dominant supplier of robusta mental impacts along coastal areas. coffee, currently producing as much coffee as The effects of trade liberalization will differ by Colombia. New technologies on the demand side country, depending on domestic policies for fish- have enabled roasters to be more flexible in switch- eries and aquaculture. If trade liberalization in fish ing quickly among coffee types, implying that products leads to higher prices for exporters, fish premiums for certain types of coffee cannot be catches may decline as already overstressed resources retained for long. Thus the so-called coffee crisis is are pushed past sustainable levels. This in turn will more a case of new entry, faster technological lead to a decline in food security and, ultimately, to change, and so far, little exit. unsustainable international seafood markets. Note Coffee In chapter 16,"Coffee: Market Setting and 1. Outgrower refers to farmers producing for a larger proces- Policies," John Baffes, Bryan Lewin, and Panos sor under some contractual arrangement and technical advice or Varangis look at a traditional tropical product, one oversight. that does not have major trade distortions. Tariffs are low, and there is only slight tariff escalation on References processed coffee. Yet despite this, coffee prices have been highly volatile. This volatility reflects mainly Hoekman, Bernard, Francis Ng, and Marcelo Olarreaga. 2002. "Reducing Agricultural Tariffs versus Domestic Support: weather-related conditions (and to a lesser extent What's More Important for Developing Countries?" World currency fluctuations) in Brazil. Bank Policy Research Working Paper 2918. Washington, Coffee consumption has been stagnant (com- D.C. mon among primary commodities), in part Ingco, Merlinda, and John D. Nash, eds. 2004. Agriculture and the WTO: Creating a Trading System for Development. because of competition from the soft drink indus- Washington, D.C.: World Bank. Part I Global Protection and Tr ade in Agriculture 2 The EVOLUTION OF AGRICULTURAL TRADE FLOWS M. Ataman Aksoy Despite tremendous change in the past 20 years in the least-developed countries (table 2.1). Although global specialization and trade in manufacturing, most of the world's poor countries are in Sub- remarkably little structural change has occurred in Saharan Africa, the region accounts for only about global agricultural trade flows. This chapter exam- 12 percent of the developing world's rural popula- ines the growth and structure of agricultural trade tion. Asia accounts for 65 percent. since the 1980s, looking at the performance of Although the share of the population in rural industrial and developing countries and of specific areas is declining, more poor people will live in commodity groups. To place arguments about agri- rural areas than in cities in developing countries for cultural policies in perspective, it also presents at least a generation. With urbanization, the rural basic statistics on rural income and poverty. share of poor households will decline, but based on current trends that share will not fall below 50 per- cent before 2035 (Ravallion 2001). Agriculture and Rural Income The share of agriculture in global trade has been Poverty shrinking, as has its share in global gross domestic product. Most successful developing countries have By the international $1-a-day poverty line, most of not relied on agriculture for their exports. Yet for the world's poor live in China,India,and"other low- most developing countries, growth in agriculture income" countries (see table 2.1). Least-developed has a disproportionate effect on poverty because countries constitute 15 percent of the world's popu- more than half of the people in developing coun- lation but almost 24 percent of the world's poor. tries reside in rural areas.1 Some 57 percent of the National poverty data, which disaggregate informa- developing world's rural population lives in lower- tion by rural and urban households but are not middle-income countries, and 15 percent lives in available for all countries, yield similar results. They Research support for this chapter was supplied by Baris Sivri, Tarek Souweid, Konstantin Senyut, and Zeynep Ersel. The author would like to thank John Beghin, Donald Mitchell, John Baffes, Steve Jaffee, Ndiame Diop, Sushma Ganguly, Harry De Gorter, and anonymous reviewers for their comments. Some of the preliminary findings of this chapter have been incorporated into Beghin and Aksoy (2003) and World Bank (2003). 17 18 Global Agricultural Trade and Developing Countries TABLE 2.1 Distribution of Poor People in Developing Countries, 1999 Poverty Headcount (under $1/day) Percentage of Population 2001 Developing Number of (millions) World's Rural Rate Poor People Country Category National Rural Urban Population (percent) (millions) Least-developed countries 596 443 153 74 15 49 292 Other low-income countries 839 501 338 60 17 26 218 excluding India Middle-income countries 1,435 478 957 33 16 8 114 excluding China China 1,272 805 467 63 27 18 226 India 1,032 745 288 72 25 35 358 Total 5,175 2,972 2,203 57 100 23 1,209 Source: World Bank data. TABLE 2.2 Rural Population and Poverty for a Sample of 52 Developing Countries (percent) Sample Countries All Developing Countries Share of Rural Share of Poor Share of Rural Income Group Dwellers in Rural Areas Dwellers Upper-middle-income countries 19 37 22 Lower-middle-income countries 64 72 61 Low-income countries 65 74 60 Least-developed countries 76 82 68 All developing countries 63 73 56 Note: Sample consists of 52 countries for which separate rural and urban income data are available. Source: World Bank data. show that four countries--Bangladesh, China, oping countries, rural households have lower aver- India, and Indonesia--account for 75 percent of the age incomes than nonrural households (figure 2.1). world's rural poor. It is in Asia, therefore, that rural The ratio of rural incomes to nonrural incomes income growth will have the greatest impact on ranges from 40 to 75 percent, a relationship that poverty. remains consistent across groups of developing In the 52 countries for which separate rural and countries. The same relationship holds for the urban income data are available, 63 percent of the middle-income OECD (Organisation for Eco- population lives in rural areas, slightly more than nomic Co-operation and Development) countries, the 56 percent for developing countries as a whole such as Greece, the Republic of Korea, and Turkey.2 (table 2.2). Some 73 percent of poor people live Farm household incomes are around 75­80 percent in rural areas and the incidence of poverty is higher of nonfarm incomes. in rural areas in all groups of developing coun- The opposite is true in many high-income tries, whatever their income level. In the least- OECD countries. Average farm household incomes developed countries, 82 percent of the poor live in are higher than average household incomes rural areas. (figure 2.2). Average farm household incomes are On average, farmers are poorer than nonfarmers almost 275 percent of average household incomes in developing countries but are better off than non- in the Netherlands, 175 percent in Denmark, farmers in industrial countries. In almost all devel- 160 percent in France, and 110 percent in the The Evolution of Agricultural Trade Flows 19 FIGURE 2.1 Ratio of Farm Household Income to Nonfarm Household Income for Selected Developing Countries, Various Years Pakistan 1990­91 Colombia 1988 India 1997 Sri Lanka 1990­91 Honduras 1993 Côte d'Ivoire 1988 Indonesia 1990 Malaysia 1990 Bangladesh 1996 Philippines 1991 Tanzania 1995 China 1995 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Source: Eastwood and Lipton 2000. FIGURE 2.2 Ratio of Farm Household Income to All Household Income for Selected High-Income Countries Netherlands, 1997 Denmark, 1999 France, 1995 Finland, 1999 Belgium, 1999 United States, 2000 Japan, 2000 Australia, 1998­99 Canada, 1999 Sweden, 1997 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5 2.75 Note: The ratio is for farm household income to all households except in Japan, where it is farm household income to workers' household income. Source: OECD 2002 and 2003. 20 Global Agricultural Trade and Developing Countries United States and Japan. In most other high- and social transfers from health, pension, unem- income countries, average farm incomes are either ployment, and child-allowance schemes). While equal to or very slightly lower than the average ratios of farm to nonfarm income are higher for household income (OECD 2002). some European countries, definitional differences make reliable comparisons across countries very difficult (OECD 2002). Structure of Income Sources In addition to these differences in relative rural and Income Distribution nonrural income levels between developing and industrial countries, the two groups of countries It is often argued that income distribution in rural have different structures of income sources. Most areas of developing countries is highly unequal and rural households in poor countries are dependent that the gains from global reforms could accrue pri- on agriculture. Rural households in Ethiopia, marily to the well-to-do rather than to the rural Malawi, and Vietnam, for example, derive about poor. Gini coefficients for a group of developing three-quarters of their income from agricultural and industrial countries indicate that despite claims activities, mainly subsistence farming (table 2.3). to the contrary, income distribution in most devel- Wages are the second-largest income source, with oping countries is more equitable in rural house- some of the wage income originating in agricul- holds than in nonrural households (table 2.4). This ture. For example, in Malawi, where 8 percent of is true for both low- and middle-income countries. total income is from wages, 3 percentage points of The opposite is true in industrial countries. that income is from agriculture. In Mexico, where In industrial countries the largest farm 40 percent of total income is from wages and only operations, generally the most profitable and 26 percent is directly from agriculture, 24 percent- wealthiest, receive most of the benefits of support age points of wage income is from agriculture, systems. Subsidy programs are not intended to keep bringing agriculture's contribution to almost small, struggling family farms in business but to 50 percent. provide large rents to large-scale farmers. Current As countries develop, the share of nonfarm production-based policies, by increasing land prices, income in rural households increases, so that agri- also encourage the creation of larger farms and the cultural price and output variations have a smaller elimination of small family farms. The unintended direct impact on rural households (figure 2.3).3 In spillover effects of these policies on other countries most industrial countries, the share of farm income and on global markets are large and negative. in total household income declines even further, as Agricultural protection in rich countries would other sources of income gain a larger share (salaries appear to worsen global income distribution. Farm- and wages from other activities; investment income; ers in industrial countries earn more on average TABLE 2.3 Structure of Rural Household Incomes, Selected Developing Countries (percent) Ethiopia Malawi Vietnam Pakistan Mexico Type of Income 2000 1997 1993 1989 2000 Total agricultural income 77 76 63 45 26 Agricultural cash income 18 16 -- -- 22 Subsistence farming 59 60 -- -- 4 Transfers 16 7 1 9 23 Wages 3 8 21 31 40 Other 4 9 15 15 11 Total 100 100 100 100 100 -- Not available. Source: World Bank household data. The Evolution of Agricultural Trade Flows 21 FIGURE 2.3 Ratio of Farm Income to Total Income of Farm Households, Selected Countries and Years 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 States, ­2001 Japan, Canada, ­1999 ­2000 Finland, ­1999 ­1999 Ireland, 1999 Norway, Denmark, Kingdom, ­1998 ­1999 Australia, ­1999 United1999 1997 1998 1998 1997 1994 1996 1997 1997 United 1996 Note: Data are averages of three most recent years available. Source: OECD 2002. TABLE 2.4 Urban and Rural Income Inequality, Selected Countries and Years Low-Income Rural Urban Middle-Income Rural Urban High-Income Rural Urban Countries Gini Gini Countries Gini Gini Countries Gini Gini Bangladesh 1996 0.26 0.36 Mexico 1996 0.34 0.40 Australia 1994/95 0.36 0.31 Pakistan 1991 0.41 0.39 Turkey 1994 0.46 0.58 Canada 1994 0.30 0.29 Sri Lanka 1990/91 0.28 0.35 Colombia 1988 0.47 0.49 Denmark 1992 0.32 0.23 Indonesia 1990 0.26 0.35 Costa Rica 1984 0.41 0.48 Finland 1995 0.26 0.22 Lesotho 1993 0.55 0.58 Peru 1994 0.37 0.35 France 1994 0.29 0.29 Madagascar 1980 0.44 0.49 India 1997 0.30 0.36 Ireland 1987 0.37 0.32 Tanzania 1993 0.35 0.42 China 1995 0.34 0.28 Italy 1995 0.43 0.34 Uganda 1992 0.35 0.44 Rep. of Korea 1987 0.12 0.42 Netherlands 1994 0.31 0.26 Malawi 1997­98 0.33 0.52 Thailand 1986 0.45 0.46 Norway 1995 0.20 0.24 Nigeria 1996 0.42 0.50 Malaysia 1987 0.42 0.43 Spain 1990 0.28 0.31 Burkina Faso 1995 0.40 0.45 Philippines 1991 0.39 0.47 United States 1994 0.37 0.37 Source: OECD 1999 for the high-income countries; National Statistics Office for Malawi; World Bank data for Nigeria; Ozmucur and Silber 2000 for Turkey; and Eastwood and Lipton 2000 for the remaining countries. than the national income average, and most farm in developing countries, will thus improve global aid goes to the largest and wealthiest farmers. At the income distribution while reducing global poverty. other end of the global income spectrum, more poor people in developing countries tend to live in Broad Trends in Agricultural Trade rural areas. Agricultural support in industrial coun- tries tends to depress world prices and demand for The last two decades have been a period of very the agricultural products of developing countries rapid export growth from developing countries, and to lower rural incomes. Global trade reforms, to aided by the growth of the world economy and the the extent that they transfer resources from well-to- lowering of trade barriers, as well as by increasing do farmers in industrial countries to poorer farmers supply capabilities in developing countries. The 22 Global Agricultural Trade and Developing Countries resulting increased import and export shares in increased both to other developing countries and total output have been a key source of growth in to industrial countries, while agricultural export many developing countries. This growth has been growth rates increased to other developing coun- fastest in manufacturing, where global levels of tries but decreased to industrial countries. protection have been reduced significantly. Growth These differential growth rates are reflected in has been slower in agriculture, where significant the shares of exports in world trade in developing protection still remains.4 countries (table 2.6). Their share in manufacturing While the 1990s were a period of rapid trade exports rose dramatically, from 19 percent in reform in developing countries and of implemen- 1980­81 to 33 percent in 2000­01, with higher tation of Uruguay Round commitments, the exports to both developing countries and industrial Uruguay Round seems not to have yielded any countries. In agricultural trade, developing coun- meaningful reduction in protection in industrial tries lost market shares during the 1980s and barely countries (see chapter 3). Protection in OECD recovered during the 1990s to their 1980­81 level of countries increased during the 1960s and 1970s, around 36 percent. All of this gain in the 1990s reaching its peak in the late 1980s. There is little came from expansion of exports to other develop- evidence that protection decreased significantly in ing countries. Despite these changes in the shares, the 1990s. In many cases, protection might even nearly half of world agricultural trade takes place have increased in the 1990s through "dirty tariffica- among industrial countries. tion" (Nogues 2002; Ingco 1997). The deceleration in growth of world agricultural trade reflects the decline in real import growth rates of industrial countries, from 4.8 percent a year Growth in Agricultural Trade in the 1980s to 2.3 percent in the 1990s.6 Over that World agricultural trade in 2000­01 was $467 bil- same period, real import growth rates for develop- lion, up from $243 billion in 1980­81.5 Real manu- ing countries accelerated from 4 percent to 6.1 per- facturing and agriculture trade expanded at similar cent a year. rates during the 1980s (5.7 and 4.9 percent a year), Two explanations have been proposed for the but real manufacturing export growth accelerated decline in import growth in industrial countries: a to 6.7 percent a year during the 1990s, while agri- lower elasticity of demand for agricultural products cultural export growth decelerated to 3.4 percent in industrial countries and the decline in commod- (table 2.5). ity prices in the 1990s. Gross domestic product The picture is similar for developing countries. (GDP) growth slowed from 3.0 percent a year in Their manufacturing export growth accelerated the 1980s to 2.3 percent a year during the 1990s in and agricultural export growth stagnated during industrial countries, while rising from 3.1 percent the 1990s. Manufacturing export growth rates to 3.7 percent in developing countries. Unless there TABLE 2.5 Average Annual Real Export Growth Rates, 1980s and 1990s (percent) Developing Countries Developing to Developing to World Total Developing Countries Industrial Countries 1980­81/ 1990­91/ 1980­81/ 1990­91/ 1980­81/ 1990­91/ 1980­81/ 1990­91/ Sector 1990­91 2000­01 1990­91 2000­01 1990­91 2000­01 1990­91 2000­01 Agriculture 4.9 3.4 5.3 5.3 4.2 7.2 5.9 4.4 Manufacturing 5.7 6.7 7.4 10.9 7.1 12.1 7.5 10.3 Note: Manufacturing imports are adjusted by the manufactures' unit value. World agricultural trade is adjusted by commodity price index with world trade weights, and developing country exports are adjusted by the same index with developing country trade weights. Source: COMTRADE. The Evolution of Agricultural Trade Flows 23 TABLE 2.6 Shares of Developing and Industrial Countries in World Exports, 1980­81 to 2000­01 (percent) Developing Countries Industrial Countries Sector by Destination 1980­81 1990­91 2000­01 1980­81 1990­91 2000­01 Agriculture Total 37.8 33.0 36.1 62.2 67.0 63.9 To developing countries 13.4 10.5 13.7 18.9 14.5 15.6 To industrial countries 24.3 22.4 22.4 43.4 52.5 48.3 Manufacturing Total 19.3 22.7 33.4 80.7 77.3 66.6 To developing countries -6.6 7.5 12.3 21.7 15.2 19.0 To industrial countries 12.7 15.2 21.1 59.0 62.1 47.6 Source: COMTRADE. TABLE 2.7 Changes in Agriculture Price Indices, 1980s and 1990s (percent) Item 1980­81/1990­91 1990­91/2000­01 U.S. farm products (producer price index) 4.7 -6.8 Raw commodities (world trade weights) -8.3 -6.6 Raw commodities (developing countries' weights) -22.7 -15.2 Source: World Bank. was a significant change in income elasticities growth rates for industrial countries and no change between 1980s and 1990s, however, these changes for developing countries (table 2.8). Thus the decel- in GDP growth rates are not large enough to eration in import growth rates is not reflected in a explain the declines in real import growth rates. deceleration in supply, and a significant component Faster liberalization in developing countries can of demand is met by domestic supply. explain some of the increases in their faster import growth rates. However, experience in the last two Agricultural Trade Shares decades also shows that the correlation between demand and trade growth is not very high over the The evidence that the agricultural trade shares medium and short run, when changes in trade of developing countries have not increased is regimes and competitiveness will have bigger impacts (box 2.1 shows the relationship between TABLE 2.8 Average Annual Agricultural demand and import growth for selected products Output Growth Rates, 1980s in industrial countries). and 1990s As for the decline in commodity prices, these (percent) were greater during the 1980s than the 1990s and so could not have been the cause of the decline in Industrial Developing Period Countries Countries growth rates (table 2.7). In the absence of specialization, slowing demand 1980­81/1990­91 0.88 3.67 growth will lead to slowing import growth if output 1990­91/2000­01 1.13 3.68 growth does not also slow. Agricultural production indexes show a slight acceleration of production Source: FAO Agriculture Production Index. 24 Global Agricultural Trade and Developing Countries BOX 2.1 Role of Demand and Changes in Market Share Low income elasticities for agricultural products, then changes in trade regimes can explain a sig- especially in industrial countries, are identified as nificant part of trade growth. Since the mid- the primary reason for the slowdown in global 1970s merchandise trade has expanded much agricultural trade growth. These low income faster than demand, showing the importance of elasticities are contrasted with higher income production restructuring. Unfortunately, the sys- elasticities for manufactured products. While temic information that is necessary to decom- trade and demand growth are highly correlated pose the determinants of export growth exists in the long run, it is not clear whether they are in only for manufacturing. The information for agri- the medium run. Variables such as level and culture is very limited. changes in protection and the degree of com- When manufacturing (including food process- parative advantage play an important role. ing) import growth to industrial countries If world trade expands primarily because of (Canada, Germany, Japan, and the United States) increases in demand, then slower agricultural is decomposed between demand and market trade can be explained by lower income elastici- share changes, demand growth accounted for ties and lower income growth in industrial coun- 32 percent of import growth and changes in tries. But if the primary cause of trade expansion market share for 68 percent. For imports from over the medium run is restructuring of produc- developing countries, growth contributed only tion and changes in both imports and exports, 21 percent while changes in market share con- without commensurate changes in total demand, tributed 79 percent (Aksoy, Ersel, Sivri 2003). Demand and Import Growth in Selected Industrial Countries, 1991­99 (percent) 1991 Market Shares Import Import Growth Industrial Country Growth from from Developing Developing Sectors Demand Growtha the World Countries World Countries Food processing 15.82 26.65 14.46 6.41 2.42 Garments 14.35 57.29 73.08 43.19 33.80 Glass products 13.06 63.54 71.99 14.24 4.30 a. Includes Canada, Germany, Japan, and the United States. Source: Aksoy, Ersel, Sivri 2003. The examples of food processing, garments, ent not only for imports from developing coun- and glass products illustrate the lack of a strong tries but for imports from the rest of the world as relationship between import and demand well. Depending on policy regimes and changes growth. The three subsectors have similar in policy regimes, trade growth rates can be very demand growth rates but very different import different from growth rates in demand. growth rates. The import growth rates are differ- consistent with other partial findings. Within a nar- A trade flow matrix for the years 1980­81, rower definition of agriculture, and focusing 1990­91, and 2000­01 shows the details of nominal mostly on key commodities, OECD (2001) data agricultural trade flows among different groups of show that import shares of these agricultural com- countries (table 2.9). The European Union is the modities in key industrial countries have not largest trader, with exports of $181 billion and increased since 1986. For many agricultural com- imports of $197 billion. Developing countries as a modities, imports as a share of world consumption block are the second largest trader, with exports of stagnated. For some commodities, such as sugar $162 billion and imports of $128 billion. and wheat, there has been significant import sub- Trade among industrial countries dominates stitution since the 1960s and 1970s, when the global agricultural trade, most of it within the trade OECD countries greatly increased their protection. blocs such as the European Union and NAFTA The Evolution of Agricultural Trade Flows 25 TABLE 2.9 Global Agricultural Trade Flows (US$ billion) Exporters Low- Middle- Other Income Income Developing Industrial Total Importers Countries Countriesa Countries EU-15 Japan NAFTA Countries Imports Low- 1980­81 0.86 2.16 3.03 2.19 0.20 1.42 0.63 7.47 income 1990­91 0.81 2.52 3.33 1.17 0.06 1.22 0.73 6.52 countries 2000­01 1.50 4.48 5.98 2.01 0.06 1.99 1.78 11.82 Middle- 1980­81 3.05 25.73 28.78 14.55 1.02 20.03 6.51 70.88 income 1990­91 4.05 29.72 33.77 17.41 1.32 19.30 7.18 78.99 countriesa 2000­01 9.20 48.44 57.64 22.85 1.74 23.42 10.71 116.36 Developing 1980­81 3.91 27.89 31.80 16.74 1.21 21.45 7.14 78.34 countries 1990­91 4.85 32.25 37.10 18.59 1.39 20.52 7.92 85.51 2000­01 10.70 52.92 63.63 24.86 1.80 25.41 12.49 128.18 EU-15 1980­81 7.20 22.89 30.09 53.82 0.24 15.44 5.55 105.15 1990­91 7.66 33.76 41.42 116.81 0.28 9.99 9.42 177.92 2000­01 9.65 37.81 47.46 131.33 0.15 9.57 9.38 197.89 Japan 1980­81 1.13 6.64 7.77 1.22 -- 9.20 2.56 20.74 1990­91 1.85 14.61 16.47 3.78 -- 14.65 4.32 39.23 2000­01 2.52 19.21 21.73 4.83 -- 17.61 5.11 49.28 NAFTA 1980­81 2.62 11.67 14.30 4.42 0.37 8.86 2.78 30.73 1990­91 2.06 15.02 17.08 7.96 0.42 15.52 3.54 44.53 2000­01 3.72 21.95 25.67 12.60 0.54 34.80 4.77 78.38 Other 1980­81 0.47 1.68 2.14 3.79 0.06 1.53 0.62 8.15 industrial 1990­91 0.40 2.31 2.71 7.01 0.07 1.66 1.09 12.54 countries 2000­01 0.54 3.24 3.79 7.22 0.08 2.15 1.70 14.94 Total 1980­81 15.33 70.77 86.10 79.99 1.89 56.48 18.64 243.10 exports 1990­91 16.81 97.95 114.77 154.16 2.15 62.35 26.29 359.72 2000­01 27.14 135.13 162.27 180.84 2.57 89.55 33.45 468.67 -- Not available. Note: All data are import-based and all the exports and imports are evaluated at c.i.f. (cost, insurance, and freight) prices. a. Includes China and India. Source: COMTRADE. (North American Free Trade Agreement). This Trade among developing countries is also intrabloc trade accounts for more than a third of increasing, with more than 50 percent of their agri- global agricultural trade. In 2000­01 industrial cultural imports coming from other developing country agricultural exports to other industrial countries. Only 39 percent of their agricultural countries totaled $226 billion. Of that, $131 billion exports are to other developing countries, however, was intra-EU trade (almost 58 percent) and $35 showing the continuing importance of industrial- billion was intra-NAFTA trade. Agricultural trade country markets for their exports. Other develop- among industrial countries excluding intra-EU and ing countries accounted for 39 percent of exports intra-NAFTA was only $60 billion. from low-income countries and 51 percent of Agricultural trade within trade blocs as a share imports in 2000­01, increases from 26 percent and of total trade is not only high, but it has increased 41 percent, respectively, in 1980­81. Shares for during the last 20 years. Intra-EU agricultural middle-income countries were similar, with other imports increased from 51 percent of total agricul- developing countries accounting for 39 percent of tural imports in 1980­81 to 66 percent in 2000­01; their exports and 50 percent of their imports in intra-NAFTA imports rose from 29 percent to 44 2000­01. Developing countries have become major percent. This increase shows how removing tariff players in the world agricultural trade, especially if barriers can stimulate trade. intra-EU and intra-NAFTA trade is excluded. 26 Global Agricultural Trade and Developing Countries TABLE 2.10 Agricultural Trade Flows (excluding Intra-EU and Intra-NAFTA Trade), 1980­81 to 2000­01 (US$ billion) Country Group and Period Exports Imports Net Imports Low-income developing countries 1980­81 15.33 7.47 -7.86 1990­91 16.81 6.52 -10.30 2000­01 27.14 11.82 -15.32 Middle-income developing countriesa 1980­81 70.77 70.88 0.11 1990­91 97.95 78.99 -18.96 2000­01 135.13 116.36 -18.77 EU-15 1980­81 26.17 51.32 25.16 1990­91 37.34 61.10 23.76 2000­01 49.51 66.56 17.06 NAFTA 1980­81 47.62 21.86 -25.75 1990­91 46.83 29.01 -17.82 2000­01 54.75 43.57 -11.17 Japan 1980­81 1.89 20.74 20.74 1990­91 2.15 39.23 39.23 2000­01 2.57 49.28 49.28 Other industrial countries 1980­81 18.64 8.15 -10.49 1990­91 26.29 12.54 -13.76 2000­01 33.45 14.94 -18.51 a. India and China are included under the middle-income developing countries. Source: COMTRADE and computations by the author. Since 1980­81 the biggest change in net agricul- The combined trade surplus of developing tural trade flows has been the relative decline in EU countries increased to $34 billion in 2000­01, from imports from the rest of the world and the increase $8 billion in 1980­81. They have a trade surplus in its export share (table 2.10). In 1980­81 the with all groups of countries except Australia and European Union was the largest importer in the New Zealand. world, accounting for 32 percent of world imports. By 2000­01 its import share had dropped to 23 per- Distribution of the Trade Expansion cent and its export share had increased to 16 per- cent (from 13 percent). Its trade deficit declined as A contentious issue in the literature has been the well, from $25 billion in 1980­81 to $19 billion in trade performance of low-income countries. Many 2000­01. The opposite has happened in NAFTA, analysts have argued that the low-income countries whose trade surplus has decreased. Japan has been have not benefited from the expansion in global the world's largest net importer of agricultural trade. This is only partially true in agriculture. Low- products since 1990­91, and Australia and New income countries' share of world exports fell from Zealand combined have surpassed NAFTA as net 6.3 percent in 1980­81 to 4.3 percent in 1990­01 exporters. and barely recovered to 5.8 percent in 2000­01. The Evolution of Agricultural Trade Flows 27 However, if intra-EU and intra-NAFTA trade are countries. In 2000­01 low-income countries excluded, their share increases from 8.5 percent in exported more to other developing countries than 1980­81 to 8.9 percent in 2000­01. As measured by to the European Union, while in 1980­81 they export and import performance, the 1980s were a exported only half as much. Some analysts have period of decline for low-income countries, while argued that it is primarily small low-income coun- the 1990s were a period of major expansion. tries that have performed poorly, but the results do Their overall trade surpluses, however, have not change if the low-income countries are divided risen throughout the period, from $7.8 billion in into small and large countries. Trade expanded 1980­81 to $15 billion in 2000­01. Low-income for both groups during the 1990s, and both have developing countries have a trade surplus with increased their trade surpluses in agriculture industrial countries and with middle-income (table 2.11). Smaller low-income countries did per- developing countries, and both of these surpluses form much worse than large low-income countries have increased since 1980. Their exports have during the 1980s, however, when their exports and increased as well, primarily to other developing imports declined. TABLE 2.11 Agricultural Trade Flows of Developing Countries, by Groups, 1980­81 to 2000­01 (US$ billion) Group and Period Exports Imports Net Imports Low-income, small 1980­81 10.63 3.26 -7.37 1990­91 10.06 2.39 -7.67 2000­01 14.95 4.45 -10.5 Low-income, largea 1980­81 4.7 4.21 -0.49 1990­91 6.75 4.13 -2.62 2000­01 12.19 7.38 -4.81 Middle-income, large exportersb 1980­81 20.26 17.73 -2.53 1990­91 25.94 18.47 -7.47 2000­01 38.4 18.11 -20.29 Middle-income, Asian importersc 1980­81 5.28 12.62 7.34 1990­91 9.54 22.77 13.23 2000­01 7.22 28.49 21.27 China and India 1980­81 7.14 5.87 -1.27 1990­91 15.13 6.56 -8.57 2000­01 23.67 14.12 -9.55 Other middle-income 1980­81 38.09 34.65 -3.44 1990­91 47.34 31.19 -16.15 2000­01 65.85 55.64 -10.21 a. Bangladesh, Ethiopia, Indonesia, Nigeria, and Pakistan. b. Argentina, Brazil, and Thailand. c. Republic of Korea, Hong Kong (China), Singapore, Taiwan (China). Source: COMTRADE and World Bank calculations. 28 Global Agricultural Trade and Developing Countries Middle-income countries, however, performed grains, animal feed, and edible oil and oilseeds. A worse during the 1990s, becoming smaller net third category consists of dynamic nontraditional exporters, with a shrinking trade surplus with the products, such as seafood, fruits, vegetables, and cut rest of the world. There are large differences in agri- flowers, for which global protection rates are lower. cultural trade performance among the middle- A fourth group consists of other products,including income countries. Argentina, Brazil, and Thailand processed agricultural products such as tobacco and are becoming major exporters (see table 2.11). cigarettes, beverages, and other processed foods. These countries, which do not have highly distorted Import growth rates in industrial countries have agricultural trade regimes, are frequently cited as declined across all these agricultural product potential gainers from global liberalization. The groups (table 2.12). The decline does not originate upper-middle-income manufacturing exporters in with price declines, which were greater during the East Asia, another group of developing countries, 1980s than the 1990s, or with slower import growth are becoming major importers of agricultural com- of tropical products, whose share was only 16 per- modities, along with Japan. Of these, the Republic cent in 1990­91. Industrial countries' growth in of Korea and Taiwan (China) have distorted trade imports from both developing and other industrial regimes, while Hong Kong (China) and Singapore countries declined during the 1990s, while develop- have liberal trade regimes. With liberalization, ing countries' import growth rates accelerated in all China and India, with one-third of the world's pop- four product groups. Again, the differences in ulation, could emerge as major global exporters import growth rates of developing countries and importers. While they have trade surpluses, the between the 1980s and the 1990s are striking, sug- surpluses did not increase significantly during the gesting a significant role for the trade liberalization 1990s. The remaining middle-income countries of the late 1980s and 1990s (see chapter 3). experienced rapid trade growth during the 1990s, but their trade surpluses shrank considerably dur- Changes in Trade Structure ing this period. The significant trade liberalization among developing countries since the 1980s, espe- The structure of world trade in agriculture has cially among middle-income countries, could changed since the 1980s along with overall trade explain some of the expanding imports of these growth rates. Expanding groups include fruits and countries. vegetables, which now have the largest share of world exports at 19 percent; fish and seafood, at 12 percent; and alcoholic and nonalcoholic bever- Disaggregated Export and Import Performance ages, at almost 9 percent (table 2.13). While these To get an accurate sense of changes in trade, it is product groups tend to have high income elastici- important to measure the contributions of differ- ties, they also have low rates of protection in indus- ent product groups to those changes.7 Many ana- trial and large developing countries. lysts argue that the markets for traditional exports Product groups that show significant declines to industrial countries are static because of both are grains, from 17 percent to 10 percent; coffee, low income elasticities and product substitution. cocoa, and tea, from 8.5 percent to 5.4 percent; For example, coffee and tea have been partially sugar and confectionary products, from 6.4 percent displaced by soft drinks, cotton by synthetic fibers, to 3.1 percent; and textile fibers, from 5.9 percent to and sugar by high-fructose corn syrup (see com- 2.8 percent. These declines result from a combina- modity chapters). tion of price declines, low demand elasticities, and, To examine the detailed flows, agricultural prod- in the case of sugar and grains, expanded produc- ucts were separated into four groups. One group tion in industrial countries. consists mainly of developing-country tropical For developing countries the biggest decline in products, such as coffee, cocoa, tea, nuts, spices, tex- export shares has come in their traditional tropical tile fibers (mostly cotton), and sugar and confec- products, such as coffee and cocoa, while the biggest tionary products. A second is made up of highly gains have come in nontraditional exports, such as protected temperate zone products of industrial seafood and fruits and vegetables. For protected countries, such as meats, milk and milk products, products, such as grains, the increase in export The Evolution of Agricultural Trade Flows 29 TABLE 2.12 Annual Import Growth Rates for Four Classifications of Agricultural Products, 1980s and 1990s (percent) Developing Countries Industrial Countries 1980­81/ 1990­91/ 1980­81/ 1990­91/ Product Classification 1990­91 2000­01 1990­91 2000­01 Tropical products Coffee, cocoa, and tea, 1.9 5.1 -0.6 1.6 raw and processed Nuts and spices 1.4 4.7 5.0 3.8 Textile fibers 3.8 0.8 0.2 -5.9 Sugar and confectionary -5.7 3.7 0.4 0.2 Subtotal 0.3 2.9 0.2 0.1 Temperate products Meats, fresh and processed 2.2 2.9 6.1 1.2 Milk and milk products 1.9 3.0 6.3 1.8 Grains, raw and processed -1.3 1.6 0.4 1.8 Animal feed 5.3 5.9 3.8 1.2 Edible oil and oil seeds 2.0 6.8 1.3 1.0 Subtotal 0.7 3.5 3.6 1.4 Seafood, fruits, and vegetables Seafood, fresh and processed 8.8 7.7 10.4 3.3 Fruits and vegetables, fresh and processed 2.8 6.4 8.3 1.9 Subtotal 4.4 6.8 9.0 2.4 Other processed products Tobacco and cigarettes 8.5 4.1 6.6 3.3 Beverages, alcoholic 4.9 6.6 8.8 4.6 and nonalcoholic Other processed food 5.6 11.9 13.6 4.9 Other -2.0 2.6 0.2 0.6 Subtotal 3.9 6.0 7.4 4.0 Total 1.4 4.3 5.1 2.0 Source: COMTRADE. shares during the 1990s are due exclusively to other processed products, which together consti- expanding trade among developing countries; these tute almost 50 percent of the exports of developing products lost shares in industrial-country markets countries. Temperate zone products constitute and gained them in developing-country markets. another 28 percent, while the traditional products Market share gains for beverages come primarily that have received most of the attention in the liter- from expanding exports of wine and beer to both ature now constitute only 19 percent of the exports developing- and industrial-country markets. of developing countries. Attention also has to be Whatever the causes for these changes, analysis placed on further expanding trade within develop- of agricultural trade for developing countries now ing countries in temperate zone products such as needs to focus on the new commodities, such as milk, grains, and meats, whose trade within devel- seafood, fruits, vegetables, and cut flowers, and on oping countries has already increased significantly. 30 Global Agricultural Trade and Developing Countries TABLE 2.13 The Structure of Agricultural Exports, 1980­81 to 2000­01 (percent of total world trade) Developing-Country Industrial-Country Exports Exports World Exports 1980 1990 2000 1980 1990 2000 1980 1990 2000 Product Classification ­81 ­91 ­01 ­81 ­91 ­01 ­81 ­91 ­01 Tropical products Coffee, cocoa, and tea, 18.3 11.0 8.5 2.5 2.9 3.6 8.5 5.6 5.4 raw and processed Nuts and spices 2.4 2.7 2.8 0.7 0.7 0.8 1.3 1.3 1.5 Textile fibers 8.0 6.2 3.3 4.5 3.9 2.6 5.9 4.7 2.8 Sugar and confectionary 10.5 4.6 4.3 3.9 2.8 2.3 6.4 3.4 3.1 Subtotal 39.2 24.4 18.9 11.6 10.3 9.3 22.0 14.9 12.7 Temperate products Meats, fresh and processed 7.2 8.3 6.0 14.8 15.7 15.4 11.9 13.2 12.0 Milk and milk products 0.3 0.7 1.1 7.9 7.9 7.6 5.0 5.5 5.2 Grains, raw and processed 9.3 4.9 7.0 21.6 13.8 11.6 16.9 10.9 9.9 Animal feed 7.5 7.9 8.5 7.7 5.1 5.3 7.7 6.0 6.4 Edible oil and oil seeds 4.6 5.7 5.5 4.8 4.4 4.4 4.7 4.8 4.8 Subtotal 28.8 27.5 28.1 56.9 46.8 44.2 46.3 40.4 38.3 Seafood, fruits, and vegetables Seafood, fresh and processed 6.9 15.9 19.4 5.5 8.2 8.0 6.0 10.8 12.2 Fruits, vegetables, and cut 14.7 22.2 21.5 13.1 17.2 17.3 13.7 18.9 18.9 flowers Subtotal 21.6 38.2 41.0 18.7 25.5 25.4 19.8 29.7 31.0 Other processed products Tobacco and cigarettes 2.6 3.1 3.3 3.0 4.2 4.8 2.8 3.8 4.2 Beverages, alcoholic and 1.1 1.8 3.6 6.9 9.5 11.5 4.7 6.9 8.6 nonalcoholic Other products and 6.7 5.0 5.2 3.0 3.8 5.0 4.4 4.2 5.1 processed food Subtotal 10.4 9.9 12.1 12.8 17.5 21.2 11.9 15.0 17.9 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: COMTRADE. These developments show that many developing Degree of Processing countries can compete in the product categories Despite significant tariff escalation in processed historically dominated by industrial countries and products, trade has moved toward processed (final) that trade reforms in industrial sectors could lead agricultural products and away from raw material to a large expansion of exports from these develop- and intermediate products.8 In 1980­81 final prod- ing countries. ucts made up slightly more than a quarter of world Industrial-country export structures have also exports, and raw and intermediate products made changed, with a decline in exports of protected up two-thirds. By 2000­01 the share of final prod- products and expansion in exports of beverages and ucts had increased to 38 percent of total exports fruits and vegetables (including intra-EU trade). (table 2.14). The share of final products in exports Greater domestic production of sugar, grains, and increased for both developing and industrial coun- other protected products has made many industrial tries, but in 2000­01 final products still constituted countries more self-sufficient and reduced their only 10 percent of the exports from low-income exports to each other. The Evolution of Agricultural Trade Flows 31 TABLE 2.14 Share of Agricultural Final Products in Exports, 1980­81 to 2000­01 (percent) Developing Developing Low- Developing Middle- Industrial Years World Countries Income Countries Income Countriesa Countries 1980­81 27.3 15.5 6.6 17.4 33.8 1990­91 33.2 19.1 7 21.2 39.8 2000­01 38.3 24.8 10.4 27.8 45.6 a. Includes China and India. Source: COMTRADE. countries, compared with 46 percent from indus- Sub-Saharan Africa is the only region that has not trial countries. made up the market share losses of the 1980s. Tariff escalation has slowed the growth of trade Despite preferential access, Africa's export share in in final products. Shares of final products are much industrial-country markets has halved. The other higher within trading blocs, where there are no tar- regions made a comeback in the 1990s. iffs, than as shares of exports to the rest of the world. For example, in 2000­01 final products con- Conclusion stituted 49 percent of intra-EU exports but 39 per- cent of EU exports to the rest of the world. For The incidence of poverty is much higher in rural NAFTA final products constituted 38 percent of areas than in urban areas in developing countries, intra-NAFTA exports but 32 percent of NAFTA the average incomes are much lower, and even with exports to the rest of the world. For developing rapid urbanization, the rural share of the poor will countries, however, the share of final products in not fall below 50 percent before 2035. In industrial 2000­01 exports was the same (around 25 percent) countries average farm household incomes are for exports to developing countries and to indus- higher than average household incomes. The shares trial countries (table 2.15). of nonfarm income in total farm household More detailed disaggregation of export flows by incomes are much higher in industrial countries degree of processing does not yield much more than in developing countries, partially shielding information than the aggregate flows. The export farmers from price and supply shocks. Finally, the share of final products increased for tropical and distribution of income is more equitable in rural temperate product groups. For seafood and fruits areas in developing countries than in urban areas, and vegetables, the shares of final product stayed while the opposite is true for industrial countries. the same because of the higher value of fresh pro- Remarkably little structural change has occurred duce. In tropical products trade among industrial in global agricultural trade since the early 1980s, countries is now primarily trade in final products. unlike the significant changes in global specializa- tion and trade in manufacturing. Unlike the case with manufacturing, developing countries have not Export Shares by Product and Region been able to increase their export shares in agricul- Developing countries lost agricultural market ture. They have maintained their global trade shares during the 1980s, mainly because the shares by expanding exports to other developing increase in their shares of seafood and fruit and countries. Again unlike the case with manufactur- vegetable exports was not great enough to compen- ing and services, trade-to-output ratios in agricul- sate for the decline in tropical product exports ture have not increased. Import growth rates accel- (table 2.16). During the 1990s developing countries erated in developing countries and decelerated in increased their export shares for most product industrial countries during the 1990s. These results groups, while the loss of market share in tropical are consistent with significant trade liberalization products slowed. in manufacturing in both developing and industrial The geographical structure of developing- countries and reforms in agricultural trade regimes country exports has changed little since the 1980s. only in developing countries. Developing countries 32 Global Agricultural Trade and Developing Countries 8.49 3.33 1.821 5.852 2.491 8.343 3.812 8.12 1.933 3.49 4.421 7.911 1.636 8.373 2000­01 100.00 Exports 1.031 3.02 4.051 8.712 1.711 0.424 3.072 7.47 0.543 3.96 11.03 4.991 6.776 3.233 otalT 1990­91 100.00 4.10 5.08 5.22 6.69 17.10 21.20 34.88 11.44 46.32 15.50 20.58 11.91 72.69 27.31 1980­81 100.00 3.39 0.26 3.66 3.19 1.23 4.42 9.15 2.91 1.05 1.29 2.35 5.69 12.05 16.78 22.47 2000­01 to Countries 4.83 0.17 5.01 3.72 1.05 4.78 8.19 2.56 1.41 0.51 1.92 4.29 10.75 18.16 22.45 1990­91 Developing Industrial 8.86 1.09 9.95 4.67 0.86 5.53 5.21 1.39 6.60 1.89 0.36 2.25 3.70 20.63 24.33 1980­81 2.16 0.48 2.64 4.45 1.30 5.75 2.56 0.76 3.32 0.99 1.04 2.03 3.58 10.16 13.75 to 2000­01 correlated. countries 2.14 0.38 2.52 3.62 0.69 4.31 1.89 0.47 2.36 0.84 0.51 1.35 8.49 2.05 highly 10.54 1990­91 are Developing Developing 3.77 0.65 4.42 4.79 0.56 5.36 1.56 0.44 2.00 1.33 0.33 1.66 1.98 11.45 13.43 2000­01 1980­81 to movements 81­ 1.61 2.09 3.69 7.22 3.82 0.96 9.67 their 12.53 19.75 10.23 14.05 10.63 25.33 22.80 48.12 2000­01 to 1980 Countries because 2.74 1.91 4.64 7.90 4.05 1.32 8.16 9.49 14.64 22.54 11.81 15.86 30.51 22.01 52.52 1990­91 Industrial goods, Industrial Processing, 3.02 1.40 4.42 7.40 7.62 2.61 1.39 4.63 6.02 15.31 22.72 10.22 27.35 16.03 43.38 of 1980­81 intermediate Level 1.33 0.50 1.83 5.68 2.74 8.42 1.87 0.63 2.50 0.49 2.42 2.91 9.36 6.29 trade) 15.65 by 2000­01 and to raw world Countries e Shares of 1.31 0.56 1.87 6.73 2.07 8.80 1.18 0.39 1.58 0.39 1.85 2.24 9.61 4.87 14.48 both 1990­91 vegetables Industrial oducts cent and pr agricultur Developing Export (per 1.45 0.96 2.41 oducts 2.61 1.11 0.65 1.75 0.60 1.37 1.97 5.59 includes 1980­81 oducts pr 10.10 12.71 and 13.26 18.85 pr fruits, ocessed COMTRADE. pr 2.15 food "Raw" ce: Product opicalrT Raw Final otalT emperateT Raw Final otalT Seafood, Raw Final otalT Other Raw Final otalT otalT Raw Final otalT Note: Sour ABLET The Evolution of Agricultural Trade Flows 33 TABLE 2.16 Export Shares by Product and Region, 1980­81 to 2000­01 (percent of world trade) Exports to Exports to Developing Countries Industrial Countries Total Exports 1980 1990 2000 1980 1990 2000 1980 1990 2000 Item ­81 ­91 ­01 ­81 ­91 ­01 ­81 ­91 ­01 Tropical products Industrial countries 2.4 1.9 1.8 4.4 4.6 3.7 6.8 6.5 5.5 Developing countries 4.4 2.5 2.6 10.0 5.0 3.7 14.4 7.5 6.3 Americas 1.8 0.7 0.8 4.9 2.2 1.5 6.7 2.9 2.4 East Asia And Pacific 1.1 0.7 0.6 1.5 0.9 0.8 2.5 1.6 1.3 Europe and Central Asia 0.4 0.1 0.3 0.3 0.2 0.2 0.7 0.4 0.5 Middle East and North Africa 0.1 0.1 0.2 0.1 0.1 0.1 0.3 0.1 0.2 South Asia 0.5 0.5 0.3 0.4 0.2 0.2 0.9 0.8 0.5 Sub-Saharan Africa 0.5 0.3 0.5 2.8 1.4 0.9 3.2 1.7 1.4 Subtotal 6.8 4.4 4.5 14.4 9.6 7.4 21.2 14.0 11.8 Temperate products Industrial countries 12.7 8.8 8.4 22.7 22.5 19.8 35.4 31.3 28.2 Developing countries 5.4 4.3 5.8 5.5 4.8 4.4 10.9 9.1 10.2 Americas 2.2 1.3 2.7 2.1 2.3 2.2 4.3 3.7 4.9 East Asia And Pacific 1.7 1.9 1.7 2.0 1.4 1.2 3.7 3.2 3.0 Europe and Central Asia 0.7 0.5 0.5 0.7 0.7 0.6 1.3 1.2 1.2 Middle East and North Africa 0.1 0.2 0.2 0.1 0.1 0.1 0.2 0.3 0.3 South Asia 0.4 0.3 0.4 0.1 0.1 0.1 0.5 0.4 0.6 Sub-Saharan Africa 0.3 0.1 0.1 0.5 0.2 0.1 0.8 0.3 0.3 Subtotal 18.1 13.1 14.2 28.2 27.3 24.2 46.3 40.4 38.3 Seafood, fruits, and vegetables Industrial countries 1.8 1.6 2.5 10.2 15.9 14.0 12.0 17.4 16.5 Developing countries 2.0 2.4 3.3 6.6 10.7 12.1 8.6 13.1 15.4 Americas 0.4 0.4 0.8 2.2 3.8 4.3 2.6 4.1 5.1 East Asia And Pacific 0.8 1.3 1.4 2.3 3.8 4.3 3.0 5.1 5.7 Europe and Central Asia 0.2 0.2 0.4 0.7 1.2 1.3 0.9 1.4 1.7 Middle East and North Africa 0.3 0.2 0.3 0.4 0.6 0.5 0.7 0.9 0.8 South Asia 0.2 0.2 0.2 0.4 0.5 0.6 0.6 0.6 0.8 Sub-Saharan Africa 0.1 0.1 0.2 0.7 0.8 1.0 0.8 0.9 1.2 Subtotal 3.8 3.9 5.8 16.8 26.6 26.1 20.6 30.5 31.9 Other processed products Industrial countries 2.0 2.2 2.9 6.0 9.5 10.6 8.0 11.7 13.5 Developing countries 1.7 1.4 2.0 2.3 1.9 2.3 3.9 3.3 4.4 Americas 0.1 0.1 0.5 0.4 0.5 0.9 0.5 0.6 1.4 East Asia And Pacific 1.2 0.9 1.0 1.3 0.8 0.8 2.5 1.7 1.8 Europe and Central Asia 0.2 0.1 0.2 0.2 0.3 0.2 0.4 0.4 0.4 Middle East and North Africa 0.1 0.1 0.2 0.0 0.0 0.0 0.1 0.1 0.2 South Asia 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1 Sub-Saharan Africa 0.0 0.1 0.1 0.2 0.2 0.2 0.3 0.3 0.4 Subtotal 3.6 3.6 4.9 8.3 11.4 13.0 11.9 15.0 17.9 Total Industrial countries 18.9 14.5 15.7 43.4 52.5 48.1 62.2 67.0 63.8 Developing countries 13.4 10.5 13.7 24.3 22.4 22.5 37.8 33.0 36.2 Americas 4.6 2.6 4.8 9.5 8.8 9.0 14.1 11.4 13.8 East Asia And Pacific 4.7 4.8 4.8 7.1 6.9 7.1 11.8 11.7 11.9 Europe and Central Asia 1.4 0.9 1.4 2.0 2.4 2.4 3.4 3.3 3.7 Middle East and North Africa 0.7 0.6 0.9 0.6 0.8 0.7 1.3 1.4 1.5 South Asia 1.2 1.1 1.0 0.9 0.9 1.0 2.1 2.0 2.0 Sub-Saharan Africa 0.9 0.7 0.9 4.2 2.6 2.3 5.1 3.3 3.3 Total 32.3 25.0 29.4 67.7 75.0 70.6 100.0 100.0 100.0 Source: COMTRADE. 34 Global Agricultural Trade and Developing Countries lost export market shares during the 1980s, mainly ment of production to more competitive producers because of the collapse in the value of tropical and created much more static global trade flows. products, and made up the loss during the 1990s by increasing their shares of other commodities. Notes Trade among industrial countries still domi- 1. Global poverty rates estimated on a consistent interna- nates world agricultural trade flows, with much of tional poverty line of $1 a day are not disaggregated by rural and the trade taking place within trading blocs, such as urban populations. Such disaggregated data are available only for national poverty rates, which vary across countries, and the the European Union and NAFTA. Trade among country coverage of these surveys is limited. Data here are from developing countries has expanded, especially dur- 52 country household surveys conducted between 1990 and ing the 1990s, when most developing countries 2001. grew faster than they had in the past and liberalized 2. The information and data are not identical, however. There is a difference between rural households and farm house- their trade regimes. The middle-income developing holds. One is a locational definition, while the other is defined countries have now become the biggest single mar- by the sources of income. ket for the exports of low-income developing coun- 3. Of course, in most regions where agriculture is the pri- mary activity, income from nonfarm sources is also related to tries. Despite the belief of many to the contrary, agriculture. In regions where there are other nonfarm-related low-income countries have increased their trade activities, or other transfers, the relationship between off-farm surplus in agricultural commodities over the last income and farm income will not be so close. 4. Annex 2 in the attached CD-ROM has detailed product two decades, especially during the 1990s. coverage by degree of processing, description of the commodity Some change has taken place in the product mix groups, the concordance between nomenclatures, the country of global agricultural trade. The shares of nontradi- coverage, country income and geographic classifications, and tional products, such as seafood, fruits, and vegeta- detailed trade flows by more detailed commodity groups. 5. This study uses a broad definition of the agricultural sec- bles, have increased, and the shares of traditional tor that includes fisheries as well as raw agricultural commodi- tropical products have decreased. Seafood, fruits, ties and processed food products. This classification includes all and vegetables, and processed foods now constitute stages of processing and results in economically consistent data series. See the CD-ROM for the details of the coverage and defi- about 50 percent of the agricultural exports of nition of subgroups. Data for the European Union-15 have been developing countries. Temperate zone products, used for all periods. Mexico is included in NAFTA and not in such as grains, dairy, and meats, constitute another developing countries. For comparability over time, trade within the Commonwealth of Independent States is excluded from 28 percent. Traditional exports, such as tea, coffee, developing-country trade data for 1990­2001, as is trade within cocoa, sugar, cotton, nuts, and spices, now consti- the former Yugoslavia and within the Southern African Customs tute a very small share of exports. This suggests the Union. Data on imports are used in most cases, but export data need for more attention to global and country poli- are used for the following countries and years: United Arab Emirates 2000­01, Bulgaria 1980­81 and 1990­91, German cies for nontraditional product groups. Democratic Republic 1980­81, Iran 1980­81 and 1990­91, There is also a move toward greater trade in final Kuwait 2000­01, Lebanon 1980­81 and 1990­91, Libya products. However, most of this trade takes place 2000­01, Romania 1980­81, Sudan 1990­91, Soviet Union 1980­81, South Africa 1990­91, China 1980­81, and intra-EU within trade blocs, such as the European Union and flows for 2000­01. NAFTA, primarily because of steeply escalating tar- 6. The deceleration of the world trade growth rates was not iffs. Despite significant reforms, the European caused by price declines in the 1990s. In nominal terms, import growth declined from 5.1 percent a year in the 1980s to 2.1 per- Union has become more self-sufficient in agricul- cent in the 1990s in industrial countries, while rising from ture, and its net trade deficit has shrunk. During the 1.4 percent to 4.3 percent in developing countries. 1990s, Japan became the biggest net importer of 7. The price series are not consistent with the trade cate- agricultural commodities, followed by the Asian gories so the disaggregated flows discussed in this section are based on nominal trade data. Tigers: the Republic of Korea, Taiwan (China), Hong 8. To have consistent data going back to 1980, this analysis Kong (China), and Singapore. Sub-Saharan Africa is uses Standard International Trade Classification (SITC 1), which the only developing-country region that has not is not as precise as the Harmonized System in separating the products by degree of processing. Thus the results are not as pre- regained the market share lost during the 1980s. cise as they are under the Harmonized System classification. Although linking this lack of change to trade poli- cies is not straightforward, the next chapter shows References that agricultural trade policies tend to be much more Aksoy, M. Ataman, Z. Ersel, and B. Sivri. 2003."Demand Growth restrictive than manufacturing policies. This very versus Market Share Gains: Decomposing Export Growth in high protection in agriculture has slowed the move- the 1990s." World Bank, Washington, D.C. The Evolution of Agricultural Trade Flows 35 Beghin, John C., and M. Ataman Aksoy. 2003. "Agricultural tries." Directorate for Food, Agriculture, and Fisheries, Trade and the Doha Round: Preliminary Lessons From the Committee for Agriculture, Paris. Commodity Studies." Paper presented at the Annual World ------. 2001. The Uruguay Round Agreement on Agriculture: An Bank Conference on Development Economics Europe Con- Evaluation of Its Implementation in OECD Countries. Paris. ference, May 15­16, Paris. ------. 2002."Farm Household Income Issues in OECD Coun- Eastwood, R., and M. Lipton. 2000. "Rural-Urban Dimensions tries: A Synthesis Report." Directorate for Food, Agriculture, of Inequality Change." University of Sussex, Brighton, and Fisheries, Committee for Agriculture, Paris. United Kingdom. ------. 2003. Farm Household Income: Issues and Policy Ingco, M. 1997."Has Agricultural Trade Liberalization Improved Responses. Paris. Welfare in the Least-Developed Countries? Yes." Policy Ravallion, Martin. 2001. "On the Urbanization of Poverty." Research Working Paper 1748. World Bank, Washington, World Bank, Country Economics Department, Washington, D.C. D.C. Nogues, Julio J. 2002. "Comment to `Trade, Growth, and Silber, J., and S. Ozmucur. 2000. "Decomposition of Income Poverty--A Selective Survey' by Andrew Berg and Anne Inequality: Evidence from Turkey." Topics in Middle Eastern Krueger, and `Doha and the World Poverty Target' by L. Alan and North African Economics, Electronic Journal 2. Eco- Winters." Commentary presented at the Annual World Bank nomic Association and Loyola University, Chicago. Conference on Development Economics, May 10­11, World Bank. 2003. Global Economic Prospects 2004: Realizing the Brussels. Development Promise of the Doha Agenda. Washington, D.C. OECD (Organisation for Economic Co-operation and Develop- ment). 1999. "Low Incomes in Agriculture in OECD Coun- 3 GLOBAL AGRICULTURAL TRADE POLICIES M. Ataman Aksoy Agricultural protection continues to be the most supplying the world market with tropical com- contentious issue in global trade negotiations.1 modities that industrial countries could not easily The high protection in industrial countries was the produce. Some countries exported limited amounts main cause of the breakdown of the Cancún Minis- of products, such as sugar and beef, in which terial Meetings in 2003. Although protection for they competed with industrial countries under manufacturing products in both industrial and preferential-access programs. Many governments developing countries has declined significantly and levied export taxes on agricultural products to overall trade reforms have been adopted in devel- generate revenues while protecting manufacturing oping countries, agricultural protection in indus- through high tariffs and other import restrictions. trial countries has changed very little. These countries also used price controls, exchange Untilthe1990sindustrialcountriesgenerallypro- rate policies, and other restrictions to keep agricul- tected agriculture while developing countries gener- tural prices low for urban consumption.Thus,many ally taxed it (Krueger, Schiff, and Valdes 1992; World policy analysts focused more on the taxation of agri- Bank 1986). Industrial countries supported their culture and its negative effects on supply in develop- agricultural sectors through subsidies to producers, ing countries than on protection in industrial coun- high tariffs, and other nontariff measures such as tries.2 In industrial countries the higher returns import restrictions and quotas.While this protection created by protection led to capital-intensive and was acknowledged in the economic literature and in supposedly efficient agricultural sectors, creating global discussions, its implication for developing the impression that their higher yields reflected countries received much less attention. comparative advantage rather than public support. Until the late 1980s and 1990s many developing This pattern of incentives began to change with countries generated a large portion of their agricul- the reforms in developing countries. Over the last tural gross domestic product (GDP) in lower- two decades many developing countries have efficiency production for the domestic market, moved from taxing agriculture to protecting it. Research support for this chapter was supplied by Baris Sivri and Konstantin Senyut. The author would like to thank John Beghin, Donald Mitchell, John Baffes, Harry de Gorter, and the reviewers for their comments. Some of the findings of this chapter have been incorporated into Beghin and Aksoy (2003) and World Bank (2003). 37 38 Global Agricultural Trade and Developing Countries Most of this change has come not through increas- With the intention of aligning agricultural trade ing protection on agricultural products but rules with the rules applying to trade in other goods, through eliminating import restrictions and lower- negotiators agreed that all barriers to imports, other ing tariffs on manufactured products, devaluing than those in place for health and safety reasons, exchange rates, abandoning multiple exchange rate should be subject to tariffs only. Before agreeing on systems that penalized agriculture, and eliminating tariff reductions, countries had to convert all border export taxes (World Bank 2001; Jansen, Robinson, measures to their tariff equivalents--a process and Tarp 2002; Quiroz and Opazo 2000). called tariffication--by calculating the difference Meanwhile, reforms in most industrial countries between domestic and world market prices (the have been modest--despite the inclusion of agri- price-gap method). Once tariff equivalents were culture under the World Trade Organization established, reductions were applied to bound tar- (WTO) Uruguay Round of international trade iffs. Developed countries were to reduce tariffs by an negotiations. Increasing the incentives for agricul- average of 36 percent and a minimum of 15 percent tural production in many developing countries over 6 years. Developing countries had lower targets without lowering the incentives in industrial coun- of a 20 percent reduction and a minimum of 10 per- tries led to overproduction and price declines for cent over 10 years.3 For cases of very high tariffs or many commodities, reducing opportunities for import quotas that had allowed in some imports, many developing countries to expand exports and minimum and current market access opportunities rural incomes (see chapter 2). were also negotiated. Usually, a minimal tariff rate This chapter evaluates both the broad trends (called a tariff rate quota, or TRQ) was set for a lim- in agricultural protection and the structure of ited volume of imports. protection in key industrial and developing With the removal of nontariff measures, some countries. Specific issues, such as the impact of countries were concerned about not being able to preferences, decoupled support, and other forms prevent sudden surges in imports. To allay these of protection, are covered in the following chapters, concerns, negotiators agreed that a special agricul- as are the structure and levels of protection for tural safeguard could be applied to certain products. selected individual commodities. The URAA offered limited opportunities for undertaking minimum import commitments for certain products rather than adopting tariffs for Uruguay Round Agreement on them. This option was taken by Japan, the Republic Agriculture of Korea, and the Philippines for rice and by Israel Since the 1980s major reforms have been made in for certain sheep and dairy products. Japan and protection regimes around the world, both through Korea have now tariffed their rice imports. unilateral reform of tariffs and quantitative import Similar efforts were made to reduce the distort- restrictions and through undertakings within the ing effect of subsidies. Subsidies were classified by Uruguay Round of multilateral trade negotiations. degree of distortion: a Red Box for prohibited sub- Most developing countries have eliminated export sidies, an Amber Box for subsidies that had to be taxes; average tariffs have declined rapidly; and reduced, and a Green Box for nondistorting subsi- other import restrictions, such as foreign exchange dies. The negotiators decided to treat export subsi- allocations for import, have effectively disappeared dies separately, so the Red Box disappeared, and (World Bank 2001). the Amber Box became the core of the negotia- Industrial countries have also started to reduce tions. A new Blue Box was created to cover direct distortions in their agricultural trade policies. payments to producers under production-limiting Agricultural trade policies were brought into the programs that were considered to be less trade dis- global trade negotiations for the first time in the torting than pure market price supports (Ingco 1994 Uruguay Round Agreement on Agriculture and Nash 2004). (URAA). Before then, import barriers in agricul- ture were coupled with the widespread use of Amber Box production-related subsidies, such as price supports, which in some countries increased production above To measure domestic support and establish a basis the competitive market equilibrium level. for reductions, a total aggregate measure of support Global Agricultural Trade Policies 39 was created based on support to agriculture during comprehensive coverage is for OECD (Organisation the base period, 1986­88. The measure covered for Economic Co-operation and Development) market price support and production-related countries: all the industrial countries and a few subsidies to farmers. Each country agreed to reduce middle-income developing countries. The focus is its supports on the basis of this measure. Industrial on selected agricultural commodities that constitute countries committed to reduce support by 20 per- 60­70 percent of domestic agricultural output. cent by 2000, and developing countries committed Food processing and seafood are generally not to a 13.3 percent reduction by 2004. Countries with covered. no Amber Box supports agreed not to use supports Agricultural protection in OECD countries is over a de minimis level of 5 percent (10 percent for measured using three instruments. One is market developing countries) of the total value of agricul- price support, the difference between domestic and tural production. international prices caused by border barriers such as tariffs and quantitative restrictions. It measures Green Box the total impact of border barriers on the prices of domestic production and is equivalent to border To qualify as a Green Box measure, requiring no protection weighted by domestic production. Bor- reduction, a subsidy must have no or almost no der barriers are the major tool of protection and trade-distorting effect and must be provided account for about 70 percent of total protection in through publicly funded government programs. OECD countries. A second instrument is direct Despite these general requirements, the Green Box support, the direct production-related subsidies covers a wide range of programs. given to farmers. A third is the general support given to agriculture through research, training, Blue Box marketing support, and infrastructure. This instru- A special exemption from reduction commitments ment is not usually included in overall production covers payments made under production-limiting support estimates. In addition, many countries have programs, provided that the payments are based subsidies for consumers. These subsidies generally on fixed areas, crop yields, livestock numbers, or, if do not affect production and so are not included in the payments are variable, on 85 percent of the producer support estimates. base level of production. These payments replaced The second measure of support is the border traditional market support payments in the Euro- protection measured by average tariffs, a measure pean Union (EU) and elsewhere that had led to available for all countries. Both the market support overproduction or had become too expensive to price and the average tariff rate are used to com- maintain. pare protection across time and across countries. Both measures have limitations. Average tariffs Evolution of Agricultural Protection measure protection on all agricultural commodi- in Industrial and Developing ties, including products that are not produced Countries domestically, while the market price support meas- ures show only the protection rate for locally pro- Review of the experience with the new rules on duced commodities. In countries such as the market access, export subsidies, and domestic sup- United States that produce a large number of agri- port shows only modest effects. One reason is that cultural commodities and have a diversified agri- support levels were at historically high levels during cultural sector or in which the degree of protection the base period selected (1986­88). In some coun- on locally produced and imported commodities is tries, such as the United States, reforms undertaken similar, these two measures tend to be very close before the negotiations were adequate to achieve (figure 3.1). In countries such as Japan where local compliance with the new rules on reducing domes- production is highly specialized or locally pro- tic support (OECD 2001). duced commodities have different rates of protec- tion from imported commodities, the two meas- OECD Countries ures will differ much more. Average tariffs also fail Two different sets of data are available to estimate to give a clear picture of real protection for domestic the degree of protection in agriculture. The most producers when the variances in tariff rates are large 40 Global Agricultural Trade and Developing Countries FIGURE 3.1 Market Price Support and Average Tariffs for Selected OECD Countries (percent) Percent 140 120 100 80 60 40 20 0 EU US Japan Iceland Norway Canada Slovak MexicoHungary PolandTurkey Republic Market price support Average tariffs Note: Market price support figures are calculated using the 2000 and 2001 average except for the Slovak Republic, which uses just the 2000 average. Source: Organisation for Economic Co-operation and Development and World Trade Organization Integrated Database. and the peaks on key domestically produced com- of gross agricultural output at world prices to modities are very high. 45 percent in 2000­02 (table 3.1). The contribution Average tariffs underestimate the real degree of of border barriers to total protection came down protection given to local producers in industrial from 77 percent to about 63 percent. If the 1960s countries and overestimate protection in the and 1970s are used as the base, however, protection OECD developing countries (see figure 3.1). Thus has risen in most OECD countries. the low average tariffs in industrial countries, The overall protection rate, which declined rap- which are compared with higher average tariffs idly after 1986 to a low of 42 percent in 1995­97, in developing countries, are highly misleading. began to rise after 1997 as world agricultural prices Industrial countries protect commodities produced declined (figure 3.3). This recent increase is driven domestically much more than commodities that both by higher domestic prices compared with are not produced locally. Developing countries, in international prices and by increases in direct sup- contrast, seem to protect commodities that are not port. This overall cyclical movement is observed in produced locally more than commodities that are. most major countries and groups (European Most of the analysis of protection in OECD Union, Japan, and the United States). The counter- countries covers the post-1986 period because sys- cyclical movement of border protection indicates temic data have been collected since then. Other that the concept of full ad valorem tariffication is estimates, though not exactly comparable over not complete and that the instruments for increas- time, indicate that the 1986­88 baseline was a ing protection as global prices decline are still oper- period of peak protection levels in the OECD (fig- ative. Direct subsidies also increased as world prices ure 3.2) and that the significant increase in protec- declined because most direct subsidies are tied to tion took place during the 1960s and 1970s. the differences between a floor and a world price Since 1986­88, when data become more consis- and increase when world prices decline. tent, overall protection (total support) for agricul- The European Union and the United States tural producers in the OECD, including border marginally reduced their overall support during protection and direct subsidies, fell from 63 percent 1986­2002. In the European Union the prices Global Agricultural Trade Policies 41 FIGURE 3.2 Nominal Rates of Agricultural Support in OECD Countries 1965­2002 (percent of total value of production evaluated at world prices) Percent 70 60 50 40 30 20 10 0 1965­74 1979­81 1986­88 1995­97 2000­02 Source: OECD PSE database, except ABARE (1999) for 1965­74 and author's calculation for 2000­02. TABLE 3.1 Percentage of Farm Gate Prices Attributable to Border Protection and Direct Subsidies by Country and Group, 1986­2002, Evaluated at World Prices Market Price Support Total Producer (Border Protection) Direct Subsidies Support Estimate 1986 1995 2000 1986 1995 2000 1986 1995 2000 Country or Group ­88 ­97 ­02 ­88 ­97 ­02 ­88 ­97 ­02 OECD 48.2 28.2 28.1 4.3 13.3 16.7 62.5 41.5 44.9 European Union 65.3 28.3 30.3 10.5 20.4 23.1 75.8 48.8 53.4 United States 16.0 7.5 9.3 18.3 7.4 16.9 34.3 14.9 26.2 Japan 145.4 131.7 131.5 16.8 13.0 14.4 162.1 144.7 146.0 Eastern European 45.2 8.7 14.1 18.3 4.8 8.0 63.6 13.5 22.1 countriesa Australia and 4.2 2.8 0.3 6.4 3.9 3.2 10.6 6.8 3.6 New Zealand Canada 53.1 42.6 10.9 11.1 12.8 12.1 64.2 55.4 23.0 Other developing 31.4 38.1 44.2 6.4 8.0 8.4 37.8 46.1 52.6 OECDb countries a. Czech Republic, Hungary, Poland, and the Slovak Republic. b. Republic of Korea, Mexico, and Turkey. Source: OECD. PSE Database. received by farmers were 65 percent higher than or close to world levels. During the 1990s the international prices in 1986­88 and 30 percent European Union also lowered many domestic higher in 2000­02. Similarly, in the United States prices and moved to support farmers through domestic prices declined from 16 percent higher direct subsidies, some coupled and some partially than international prices to 9.3 percent higher. In decoupled. Thus, direct production-related pay- the United States the primary source of support is ments to farmers increased from 10.5 percent to direct subsidies to farmers. The level of subsidy 23 percent, partially compensating for the decline stayed around 17 percent, much higher than the in border barriers. So, while the type of support level of border barriers. The prices are set at world changed from border measure to different forms of 42 Global Agricultural Trade and Developing Countries FIGURE 3.3 Rates of Agricultural Support in OECD Countries and Real U.S. Agricultural Price Index Estimated protection Prices 80.0 140.00 70.0 120.00 60.0 100.00 50.0 80.00 40.0 60.00 30.0 40.00 20.0 10.0 20.00 0.0 0.00 19861987198819891990199119921993199419951996199719981999200020012002 Agricultural Support Real Prices Source: OECD for 1986­2001; author's calculation for 2002. Agricultural price data is from the U.S. Department of Agriculture and is deflated by the manufacturing unit value index. direct support, there was very little reduction in tion in most developing countries have been signif- overall protection (see chapter 5). icant. From the 1960s to the 1980s, despite high Among the middle-income countries of the tariffs on agricultural products, most developing OECD, the Eastern European countries had the countries had negative total protection rates on largest reductions in protection, from about 64 per- agriculture, a result of both direct protection, cent in 1986­88 to 18 percent in 2000­02. The including tariffs and taxes on agricultural products, Republic of Korea always had very high protection, and indirect protection caused by protection of and it has stayed high, with small variations. industry and exchange rate overvaluation (Schiff Mexico and Turkey, which started with low protec- and Valdes 1992; World Bank 1986). In a sample of tion, increased it over this period, mainly through 15 developing countries studied by Schiff and higher border protection. Valdes (1992), all but the 3 OECD middle-income These numbers support the hypothesis that the countries had negative direct protection rates and Uruguay Round did not have a significant impact negative total protection rates on agriculture. Of on the levels of agricultural support in OECD the 3 OECD middle-income countries, the total countries, especially the large industrial countries protection rate was marginally positive for the (Ingco 1997; Messerlin 2002; Nogues 2003; OECD Republic of Korea and Portugal (table 3.2). 2001). Thus, despite the implicit promise by indus- The average agricultural tariff in developing trial countries that agriculture would follow the countries declined from 30 percent in 1990 to path of manufacturing, with protection rates con- 18 percent in 2000, a significant drop (figure 3.4).4 tinuously declining--one of the reasons developing These reductions were complemented by elimina- countries embraced trade liberalization--this has tion of import licensing, most export taxes, and not happened. many quantitative restrictions (World Bank 2001). Overvaluation of exchange rates, the main source of the bias against agriculture, decreased or was Other Developing Countries eliminated during the 1990s in most developing In contrast to the modest changes in agricultural countries. On average, tariffs are now much higher protection in OECD countries, changes in protec- in agriculture than in manufacturing, a reversal of Global Agricultural Trade Policies 43 TABLE 3.2 Agricultural Protection Rates in Selected Developing Countries Tax Due to Group Direct Protection Industrial Protection Total Protection Developing countriesa -13.0 -27.8 -35.7 OECD middle-income countriesb 17.8 -28.4 -3.6 a. Argentina, Brazil, Chile, Colombia, Côte d'Ivoire, Dominican Republic, Egypt, Ghana, Malaysia, Morocco, Pakistan, Philippines, Sri Lanka, Thailand, and Zambia. b. Republic of Korea, Portugal, and Turkey. Source: Schiff and Valdes 1992, table 2-1. FIGURE 3.4 Average Most-Favored-Nation Applied Tariffs for Agricultural and Manufacturing Products in Developing Countries, 1990­2000 (percent) Average tariff (percent) 35 29.6 30 25.8 25 22.9 20 18.4 16.1 15 11.4 10 5 0 Agricultural products Manufacturing products 1990 1995 2000 Note: Tariff rates are simple averages of countries' unweighted tariffs. Source: TRAINS. the tendency during the 1980s of greater protection Current Structure of Agricultural for the industrial sector. Protection A study of 15 developing countries by Jansen, The overall support given to agricultural producers Robinson, and Tarp (2002) also concludes that the in OECD countries through higher domestic prices bias against agriculture had been largely elimi- and direct production-related subsidies was $228 nated. They find that by the end of the 1990s the billion during 2000­02 (table 3.3). About 63 per- economywide system of indirect taxes, including cent, or $143 billion, of this came from border bar- tariffs and export taxes, significantly discriminated riers and market price support, and 37 percent against agriculture in only one country, was largely from direct subsidies to farmers. The bulk of the neutral in five, provided a moderate subsidy to support went to temperate-climate products such agriculture in four, and strongly favored agriculture as milk, meats, grains, and sugar. in five. Quiroz and Opazo (2000), updating Schiff Aggregate support levels in OECD countries vary and Valdes (1992) for Latin America, also conclude significantly. Iceland, Norway, and Switzerland have that direct protection and protection due to higher very high levels of support, through both high border tariffs in manufacturing have fallen but that protection and high direct payments. At the other exchange rates appreciated, reversing some of the extreme, Australia and New Zealand have very low lower protection for exportable commodities. 44 Global Agricultural Trade and Developing Countries TABLE 3.3 Agricultural Support in OECD Countries, 2002­02 (billions of U.S. dollars) Eastern United European Emerging European Support States Union Japan Supportersa Countriesb Total OECD Who receives support Producers 46.97 92.19 47.50 30.49 4.41 227.54 General services 24.29 8.02 12.25 5.98 0.57 53.08 Consumers 22.24 3.64 0.42 0.97 0.06 34.26 Total 93.50 103.85 60.17 37.44 5.05 314.88 Products that receive support Milk 11.25 16.11 4.63 2.53 1.03 40.14 Beef and pork 1.99 25.05 3.50 2.63 0.73 36.65 Rice 0.92 0.25 16.47 7.21 na 25.00 Wheat 3.99 8.97 0.89 0.36 0.31 15.31 Corn 6.80 2.41 na 1.32 -0.10 10.64 Other 22.02 39.40 22.00 16.46 2.45 99.81 Source of producer support Border measuresc 16.63 52.24 42.80 25.60 2.81 142.66 Domestic measuresd 30.34 39.95 4.70 4.89 1.60 84.89 na ­ not applicable. a. Republic of Korea, Mexico, and Turkey. b. Czech Republic, Hungary, Poland, and Slovak Republic. c. Tariffs and tariff equivalents of other border measures. d. Direct payments to producers. Source: OECD 2003 and authors' calculations. support levels. Japan and the Republic of Korea have (Bangladesh, Guatemala, Indonesia, Kenya, Malawi, high support levels mainly through higher tariffs and Togo, Uganda, and Zimbabwe). The analysis focuses quantitative restrictions. In between are the European on tariffs because they are the only comparable Union toward the higher end and Canada toward the measure of protection and support across countries lower end. and because lower, more transparent tariff struc- This section evaluates tariff regimes for agricul- tures were a key objective of the Uruguay Round. tural products for 6 industrial and 24 developing countries within the context of the objectives of the Tariff Transparency Uruguay Round. The selection of countries was constrained by the lack of recent detailed tariff The objective of achieving greater transparency of schedules for most countries.5 protection levels through tariffication has not been The countries are placed in four groups for fully realized, especially in the key industrial coun- analysis: the Quad countries (Canada, the European tries and some middle-income countries. Many Union, Japan, and the United States); eight large tariffs are still specific, compound, or mixed, making middle-income countries with significant agricul- it almost impossible to estimate real protection lev- tural sectors (Brazil, China, India, the Republic of els, since these will change with the price of imports. Korea, Mexico, South Africa, Russian Federation, Protection rates rise as the world prices of products and Turkey); eight other middle-income countries, decline, increasing protection levels for lower-priced to ensure regional balance (Bulgaria, Costa Rica, products originating from developing countries.6 Hungary, Jordan, Malaysia, Morocco, Philippines, Transparency in agriculture is significantly and Romania); and eight low-income countries greater in developing countries than in industrial Global Agricultural Trade Policies 45 FIGURE 3.5 Non-Ad-Valorem Tariff Lines as a Share of Total (percent) Percentage of tariff lines 35 31.3 30 25 20 15 10 9.3 5.7 5 3.1 2.6 0.1 0.2 0.5 0 Quad Large middle- Other middle- Low-income countries income income countries countries countries Agriculture Manufacturing Note: Covers tariff lines with specific, compound, or mixed duties. Source: World Trade Organization Integrated Database (most-favored-nation applied duties). countries (figure 3.5). Of the 24 developing coun- oping countries that have specific duties, they are tries in the sample, only 4 have non-ad-valorem clustered within a few product groups. For exam- rates in more than 5 percent of tariff lines--Bulgaria ple, in Malaysia they are on tobacco and alcohol (13.5 percent), South Africa (25 percent), Russian products, in Mexico on chocolate and confec- Federation (31 percent),and Turkey (6 percent)--all tionary products, sugar, nuts, and spices. of them middle-income countries. Of the remaining Specific duties are found almost exclusively 20 countries, 4 have them in less than 5 percent of in agriculture. For example, in the United States, tariff lines, 5 in less than 1 percent; 11 have none. which has the highest percentage of non-ad- Within the Quad, Japan has specific, compound, or valorem duties in manufacturing, only 8 percent of mixed rates in 15 percent of its tariff lines, Canada tariff lines in manufacturing are non-ad-valorem, in 24 percent, the United States in 40 percent, the compared with 43 percent in agriculture. The European Union in 44 percent,and Norway,with the European Union has almost no non-ad-valorem highest share of any industrial country, in 54 per- duties in manufacturing, but 44 percent of its tariff cent. The European Union also has duties that vary lines in agriculture have non-ad-valorem rates. Thus according to the content of the products in 4 percent the use of specific duties is not a general administra- of its tariff lines,and the United States in 1 percent of tive arrangement but is limited to agriculture. its tariff lines. Thus, transparency of tariff rates is More detailed analysis of the incidence of spe- consistently weaker for industrial countries and a cific duties suggests that they are being used prima- few middle-income countries than for most devel- rily as an instrument of disguised protection. First, oping countries. the average ad valorem equivalents of specific The pattern of specific duties varies across coun- duties, where available, are much higher than the tries. In the United States almost all categories of average ad valorem rates, as shown for four coun- products have non-ad-valorem rates between 30 tries that reported the ad valorem equivalents of and 60 percent of tariff lines. In the European non-ad-valorem rates (table 3.4). This suggests that Union some product groups, such as milk, grains, reported average duties are seriously underesti- sugar, and beverages have non-ad-valorem duties mated for countries with a large proportion of in more than 90 percent of tariff lines. In the devel- non-ad-valorem duties. 46 Global Agricultural Trade and Developing Countries TABLE 3.4 Average Ad Valorem and Specific Duty Rates (percent) Country or Average ad Valorem Average ad Valorem Tariff Share of Non-ad- Group Tariff Equivalent of Specific Duties Valorem Lines Australia 1.2 5.0 0.9 United States 10.6 35.2 43.6 European Union 21.6 58.0 40.4 Jordan 8.1 11.7 0.8 Note: Average applied, out-of-quota, ad valorem, and ad valorem equivalents of non-ad-valorem tariffs for which equivalents are reported. Source: World Trade Organization Integrated Database (most-favored-nation applied duties). TABLE 3.5 Proportion of Non-Ad-Valorem Tariff Lines by Degree of Processing (percent) Country or Group Raw Intermediate Final Norway 41.39 58.84 68.53 European Union 22.05 45.27 57.54 United States 37.91 43.05 41.34 Canada 17.14 23.01 30.20 Russian Federation 11.79 9.74 53.06 Turkey 0 5.22 12.70 Note: Tariff Lines containing specific, compound, or mixed duties, as a percentage of all lines. Source: World Trade Organization Integrated Database (most-favored-nation applied duties). Second, the share of tariff lines with non-ad- The tariff data presented here, especially for valorem duties increases with the degree of pro- industrial and some middle-income countries, cessing and is highest in final products, which are seriously underestimates actual border protection generally classified under food-processing indus- for domestic producers. Specific duties are not tries. For example, in the European Union, the reflected in the averages, and they are generally share of non-ad-valorem tariff lines is 22 percent higher than ad valorem rates (see table 3.4). The for raw materials but 43 percent and 58 percent for reported ad valorem equivalents of specific duties intermediate and final products (table 3.5). In the for the European Union and the United States are Russian Federation the share of non-ad-valorem much higher than the ad valorem rates. Assuming duties in tariff lines is 12 percent for raw materials the same pattern for Canada and Japan, which have but 53 percent for final products. non-ad-valorem rates for 25 percent and 15 percent of their tariff lines, respectively, Quad average tariffs are being significantly underestimated. The Levels of Tariff Protection degree of bias is indicated by the third column in The conversion of nontariff barriers to tariffs under table 3.6 showing the proportion of tariff lines to the Uruguay Round Agreement on Agriculture was which the average tariffs apply. an important step forward, but in most industrial Except for Canada, which has a large proportion and developing countries average agricultural tar- of non-ad-valorem tariffs without equivalents, aver- iffs are much higher than average tariffs for non- age tariffs are much higher in agriculture than agricultural products and continue to restrict trade in manufacturing. The difference is especially (table 3.6). pronounced in the European Union, where the Global Agricultural Trade Policies 47 TABLE 3.6 Average Agricultural Tariffs, Selected Country Groups and Years (percent) Share of Lines Covered Country or Group Agriculture Manufacturing in Agriculture Quad countries 10.7 4.0 86.7 Canada (2001) 3.8 3.6 76.0 Japan (1999) 10.3 3.7 85.5 United States (2001) 9.5 4.6 99.4 European Union (1999) 19.0 4.2 85.9 Large middle-income countriesa 26.6 13.1 91.3 Other middle-income countriesb 35.4 12.7 97.7 Lower-income countriesc 16.6 13.2 99.8 Note: Most-favored-nation, applied ad valorem, out-of-quota duties. a. Brazil (2001), China (2001), India (2000), the Republic of Korea (2001), Mexico (2001), Russian Federation (2001), South Africa (2001), and Turkey (2001). b. Bulgaria (2001), Costa Rica (2001), Hungary (2001), Jordan (2000), Malaysia (2001), Morocco (1997), Philippines (2001), and Romania (1999). c. Bangladesh (1999), Guatemala (1999), Indonesia (1999), Kenya (2001), Malawi (2000), Togo (2001), Uganda (2001), and Zimbabwe (2001). Source: World Trade Organization Integrated Database (most-favored-nation applied duties). average tariff is 19 percent in agriculture and 4.2 per- tries (table 3.7). Furthermore, actual protection for cent in manufacturing. Among developing countries local producers is much higher than these average tar- the results are similar, with a few exceptions such as iffs in industrial countries and much lower than the Brazil and Indonesia, whose manufacturing tariffs average tariffs in selected developing countries, as are marginally higher (less than 1 percentage point). shown previously (see figure 3.1). Only in Malaysia are tariffs much higher in manufac- The difference between average rates and maxi- turing (9.7 percent) than in agriculture (2.8 percent). mum tariff rates and the relative domestic price dif- Developing countries in the sample have higher ferences for local production measured by market agricultural tariffs than industrial countries, with price support data from the OECD indicate that Morocco (64 percent), the Republic of Korea protection is very uneven, with domestic produc- (42.2 percent), and Turkey (49.5 percent) having tion being protected much more significantly. the highest average tariff rates, and Indonesia Japan, with an average tariff of 10 percent and a (8.5 percent) and Malaysia the lowest (2.8 percent). maximum ad valorem tariff of 50 percent, has esti- Again, it is important to recall that average tariffs in mated market price support of 130 percent. The dif- countries with a high share of non-ad-valorem ference can only be attributed to specific duties not rates in tariff lines are seriously underestimated; included in the data set. The situation is similar for examples are the Russian Federation (a non-ad- the European Union, with an average tariff of about valorem rate in tariff lines of 31 percent), South 19 percent and market price support of 30 percent. Africa (25 percent), Bulgaria (14 percent), and For both Japan and the European Union, tariffs Turkey (6 percent). for many locally produced items are very high. For In addition,average tariffs are not reflective of pro- example, in the European Union average tariffs are tection because the tariffs have wide dispersion and 34.6 percent for grains, 54.6 percent for milk and very high peaks. While tariffs on average are lower in milk products, and 32.5 percent for meats. industrial countries, significant tariff peaks indicate Another issue is the product coverage of the tariffs high rates of protection for specific products--almost presented here and included in the market price sup- 1,000 percent in the Republic of Korea, 506 percent in port measures used by the OECD. The tariffs the European Union, and 350 percent in the United reported here include seafood, tobacco and cigarettes, States.7 Many low-income countries have lower peaks wine, and tropical products, none of which is and variance than many of the middle-income coun- included in the market price support measures for 48 Global Agricultural Trade and Developing Countries TABLE 3.7 Tariff Peaks and Variance in Selected Countries (percent) Average Maximum Standard Share of Lines Country or Group Tariff Tariff Deviation Covered Canada 4.1 238.0 13.5 74.2 Japan 10.9 50.0 10.1 84.8 United States 9.9 350.0 26.5 99.5 European Union 19.0 506.3 27.3 85.9 Republic of Korea 39.9 917.0 107.9 97.9 Brazil 13.2 55.0 5.6 100.0 Costa Rica 14.2 154.0 18.0 100.0 Morocco 67.4 376.5 70.6 100.0 Indonesia 8.9 170.0 25.6 100.0 Malawi 16.5 25.0 8.5 100.0 Togo 15.6 20.0 6.1 99.9 Uganda 13.6 15.0 3.2 100.0 Note: Most-favored-nation, out-of-quota, applied tariffs. Source: World Trade Organization Integrated Database. TABLE 3.8 Tariff Rate Escalation in Agriculture, Selected Country Groups and Years (percent) Share of Lines Country or Group Raw Intermediate Final Average Covered Quad countries 6.1 9.3 14.8 10.7 86.7 Canada 1.4 3.4 6.5 3.8 76.0 Japan 4.2 10.2 15.9 10.3 85.5 United States 5.5 7.1 12.6 9.5 99.3 European Union 13.2 16.6 24.3 19.0 85.9 Large middle-income countriesa 21.9 23.3 34.4 26.6 91.3 Other middle-income countriesb 21.6 31.7 49.0 35.4 97.7 Lower-income countriesc 13.2 14.8 23.0 16.6 99.8 Note: Most-favored-nation applied, ad valorem, out-of-quota duties. a. Brazil (2001), China (2001), India (2000), the Republic of Korea (2001), Mexico (2001), Russian Federation (2001), South Africa (2001), and Turkey (2001). b. Bulgaria (2001), Costa Rica (2001), Hungary (2001), Jordan (2000), Malaysia (2001), Morocco (1997), Philippines (2001), and Romania (1999). c. Bangladesh (1999), Guatemala (1999), Indonesia (1999), Kenya (2001), Malawi (2000), Togo (2001), Uganda (2001), and Zimbabwe (2001). Source: World Trade Organization Integrated Database (most-favored-nation applied duties). the OECD countries. If seafood, beverages, tobacco, Tariff Escalation and noncompetitive tropical products are excluded, Protection escalates with the level of processing in the average tariff rises from 3.8 percent to 10.4 per- almost all countries and across all products cent in Canada and from 10.7 percent to 24.7 percent (table 3.8). Escalation slows diversification into in Japan (excluding specific tariffs). This supports the value added and processed products. The manufac- hypotheses that the low average tariffs are misleading turing component of agriculture and food process- and that protection is uneven and focused primarily ing have very high rates of protection. on selected domestically produced commodities. Global Agricultural Trade Policies 49 TABLE 3.9 Tariff Escalation in Selected Agricultural Product Groups (percent) Product European Union United States Japan Traditional tropical products Coffee Raw 7.3 0.1 6.0 Final 12.1 10.1 18.8 Cocoa Raw 0.5 0.0 0.0 Intermediate 9.7 0.2 7.0 Final 30.6 15.3 21.7 New expanding products Fruits Raw 9.2 4.6 8.7 Intermediate 13.3 5.5 13.2 Final 22.5 10.2 16.7 Vegetables Raw 9.9 4.4 5.0 Intermediate 18.5 4.4 10.6 Final 18.0 6.5 11.6 Seafood Raw 11.5 0.6 4.9 Intermediate 5.1 3.2 4.3 Final 16.2 3.5 9.1 Note: Most-favored-nation applied, ad valorem, out-of-quota duties. Source: World Trade Organization Integrated Database. Tariff escalation occurs in all types of products, Tariff Rate Quotas not just those produced in industrial countries. Data Tariff rate quotas, designed to ensure some degree of on products with low tariffs on raw commodities, market access despite protection, have resulted in both traditional products (coffee and cocoa) and more complex tariff regimes. While the number of new products (fruits and vegetables, seafood), show tariff lines under tariff rate quotas is small, these that tariff escalation is common to both (table 3.9). lines cover some of the main commodities produced Tariffs are extremely low on the raw stages of tra- in OECD countries. According to OECD data, ditional products, whereas the final stages and almost 28 percent of domestic agricultural produc- processed products have extremely high tariffs. Sim- tion is protected by tariff rate quotas. Rates range ilar tariff escalation is apparent in fruits and vegeta- from a high of 68 percent in Hungary to 38 percent bles, which are supposed to be less protected and in in the European Union and 26 percent in the United which developing country exports have expanded. States to 13 percent in Japan (figure 3.6). Australia In addition, these averages mask very high peaks and New Zealand have no tariff rate quotas. on individual products. In the United States maxi- mum tariffs are 136 percent on final fruit products and 186 percent on cocoa products. In the European Export Subsidies Union the maximum rates are 98 percent and 146 percent on processed fruits and vegetables and 63 per- Although lower tariffs and the move toward direct cent on cocoa products. And again, many of the final production subsidies are beginning to reduce the product tariffs are non-ad-valorem, meaning that the need for export subsidies in agriculture (they have averages underestimate the full extent of high tariffs. been illegal on nonagricultural products since 50 Global Agricultural Trade and Developing Countries FIGURE 3.6 Share of Output under Tariff Rate Quotas (percent) Percent 60 50 40 30 20 10 0 States Japan average Europe Zealand countries countries CommunityUnited New OECD Eastern industrial European developing Australia, Other Other Source: OECD, Agricultural Market Access Database (AMAD). 1955), export subsidies continue to distort world their costs of production) by 51 percent, leading to markets. The European Union accounts for almost increased production that depressed the world 90 percent of all OECD export subsidies. The price. U.S. export prices were 58 percent of the aver- Uruguay Round Agreement on Agriculture placed age costs of production for wheat, 67 percent for limits on export subsidies for individual commodi- corn, and 77 percent for rice (Watkins 2003). The ties but allowed some flexibility. With usage levels move toward replacing border barriers with direct low early in the implementation period, when world subsidies in industrial countries will increase the prices were high, several countries carried forward importance of these implicit export subsidies.8 unused export subsidy credits for later use. Circum- vention, through the subsidy elements of export Implications of Reform credits, export restrictions, and revenue-pooling arrangements in major products, is a concern. One trade reform proposal that would have cut Even if tariffs were eliminated altogether along agricultural tariffs substantially was put up by with the official export subsidies, current agricul- Stuart Harbinson, chairman of the agricultural tural production subsidies would keep the domestic negotiations in the Doha Round of the WTO trade and export price of many commodities lower than negotiations (DRIFE 2003). The proposal was their costs of production in industrial countries. By rejected by industrial-country trade ministers as lowering production costs, production subsidies too radical, however, and brought the Cancún favor industrial-country producers over developing- Ministerial Meetings to a close. The implications of country producers, who do not receive direct subsi- this proposal in terms of actual tariff outcomes is dies. Consider cotton subsidies in the European presented below as an illustration. Union and the United States. Tariffs are zero, and Harbinson proposed that industrial countries domestic prices are the same as world or export cut average agricultural tariffs 60 percent on bound prices (Baffes 2004; Watkins 2003).Yet in the United tariffs above 90 percent, 50 percent on bound tariffs States in 2001, production subsidies effectively between 15 and 90 percent, and 40 percent on increased the prices farmers received (or reduced bound tariffs below 15 percent.9 For developing Global Agricultural Trade Policies 51 TABLE 3.10 Tariffs in the European Union and the United States Before and After Average Reduction from Applied Tariffs under the Harbinson Proposal (percent) United States European Union Before Harbinson After Harbinson Before Harbinson After Harbinson Product Average Peak Average Peak Average Peak Average Peak Raw 5.5 350.0 2.7 140.0 13.2 131.8 6.9 52.7 Intermediate 7.1 159.3 3.8 63.8 16.6 284.8 8.3 113.9 Final 11.7 180.8 6.2 72.3 26.8 506.3 13.1 202.5 Overall 8.8 350.0 4.6 140.0 19.7 506.3 9.9 202.5 Note: The analysis excludes cigarettes and alcoholic drinks. Source: World Trade Organization Integrated Database. TABLE 3.11 Tariffs in Selected Developing Countries Before and After Average Reductions from Bound Rates (percent) Costa Rica India Jordan Korea, Rep. of Category Average Peak Average Peak Average Peak Average Peak Before Harbinson 49.0 245.0 115.3 300.0 21.5 180.0 50.8 917.0 After Harbinson 33.8 147.0 72.3 180.0 14.9 108.0 33.2 550.2 Current applied rates 13.1 154.0 36.7 115.0 18.5 120.0 42.7 917.0 Note: The analysis excludes cigarettes and alcoholic drinks. Source: World Trade Organization Integrated Database. countries and for products that are not considered of the reform process under an optimistic scenario strategic, average tariffs would be cut 40 percent for in which all tariffs are cut by the average rate from bound tariffs above 120 percent, 35 percent for tar- the applied rates (table 3.10).10 EU tariffs would iffs between 60 percent and 120 percent, 30 percent come down from 20 percent to about 10 percent, for tariffs between 20 percent and 60 percent, and while U.S. tariffs would drop from 9 percent to 25 percent for tariffs below 20 percent. These cuts below 5 percent. Even so, the average agricultural would be implemented over 5 years for industrial tariffs in both areas would remain significantly countries in equal installments and over 10 years higher than the average manufacturing tariffs of for developing countries (WTO 2003). 4.2 percent in the European Union and 4.6 percent While the proposed cuts look significant--some in the United States. Tariff peaks would remain groups have called them radical--their impact above 200 percent in the European Union and would not be as great as might appear. For develop- above 140 percent in the United States. ing countries the key issue is reductions from the For developing countries the optimistic scenario bound, not the applied, rates. Most developing lowers all the bound rates by the amount of the countries have bound their tariffs at relatively high average cut. Cuts from bound rates do not signifi- rates, but applied rates are much lower. If cuts are cantly lower protection in most developing coun- made to the bound rates, countries would get credit tries. At the end of 10 years the Harbinson reform for the unilateral reforms, but the reductions would would leave bound tariffs significantly above the not lead to significant actual reductions in tariffs. currently applied rates in Costa Rica and India and For the United States and the European Union, only marginally below the current applied rates in average effective tariffs would be halved by the end Jordan and the Republic of Korea (table 3.11). 52 Global Agricultural Trade and Developing Countries Because these results would hold for most develop- industrial countries in their structure of protection. ing countries, existing levels of protection in the More generally, as taxation of agriculture dimin- developing world would not be significantly ishes in developing countries, reactive protection in reduced under the Harbinson proposals or under response to industrial-country agricultural support any other proposals that start with bound rates. is increasing. Many developing countries have Thus even significant cuts in tariffs by industrial increased protection of domestic food products and developing countries will leave agricultural against cheaper, subsidized exports from industrial sectors with highly distorted tariff structures. In countries. addition to average cuts, however designed, there Although official export subsidies may be small has to be an agreement on tariff peaks, which and shrinking, implicit export subsidies resulting should be capped at reasonably low rates. from domestic support are increasing, lending unfair advantage to industrial-country producers. In the United States and the European Union, Conclusion domestic and export prices of cotton are the Within OECD countries, budgetary subsidies and same--but those prices are less than half the aver- subsidies from consumers (through high tariffs and age cost of production. Similar differences exist for quantitative restrictions on domestic production of many other products, a gap that will increase as selected commodities) to agricultural producers industrial countries move from protection through totaled about $228 billion in 2000­02, or 45 percent border barriers and high support prices to support of farm revenues. That was down from 62 percent through coupled or partially decoupled subsidies. in 1986­88 but is still very high. Some 63 per- Two other dynamics complicate protection. cent of this support was through the higher prices First, many agricultural policies are anticyclical, associated with border protection and 37 percent with protection increasing when agricultural prices through direct subsidies. In developing countries are low. Thus protection levels fell as commodity almost all support is generated through border prices increased in the early 1990s and then rose barriers. again as prices declined in the late 1990s. Second, Average agricultural tariffs in industrial coun- rapid and sustained technical progress in agricul- tries, when they can be measured, are two to four ture has lowered the costs of production and thus times higher than average manufacturing tariffs. lowered prices. Countries that have been able to Even at that, these averages seriously underestimate enjoy the benefits of technological change have the actual level of protection to local producers. managed to maintain their production and com- Almost 30 percent of domestic production in pete with subsidized production. OECD countries is protected by tariff rate quotas. Significant reforms are needed to make a dent in More than 40 percent of the tariff lines in the rural poverty in most developing countries (see European Union and the United States include spe- chapter 2). Given the magnitude of the distortions cific duties, which make it difficult to calculate in the agricultural sectors in all countries, the pro- average tariffs, obscure actual levels of protection, posals for reform have been quite modest. Yet even and penalize developing countries that supply the modest proposals have not been accepted by the cheaper products. Tariff peaks as high as 500 per- key industrial countries. cent confront imports from developing countries. A few simple issues stand out. Given the com- Tariffs also rise by degree of processing, creating a plexity of the protection regimes, all non-ad- highly escalating tariff structure that limits access valorem tariffs should be converted into ad to processed food markets. valorem tariffs. Variances in tariff rates are so high Developing countries, too, have maintained that the only way to reduce protection significantly high border protection and have higher average is through binding ad valorem, nonseasonal tariff agricultural tariffs than industrial countries. What caps that are gradually reduced to zero or to very is worse, many of the protectionist developing low levels. Otherwise, high tariffs on selected prod- countries are middle-income economies, where the ucts will continue under all modalities of reform. demand for agricultural products is growing rap- Finally, direct support programs have to be fully idly. These countries are beginning to resemble decoupled from production in industrial and Global Agricultural Trade Policies 53 middle-income countries (see chapter 5), and other Beghin, John C., and M. Ataman Aksoy. 2003. "Agricultural instruments have to be used to support the rural Trade and the Doha Round: Preliminary Lessons From the Commodity Studies." Paper presented at the Annual World sector in these countries. Bank Conference on Development Economics Europe Con- ference, May 15­16, Paris. DRIFE (Danish Research Institute of Food Economics). 2003. Notes "Note on the Harbinson Draft on Modalities in the WTO Agriculture Negotiations." Agricultural Policy Research 1. Annex 3 in the attached CD-ROM contains detailed tariff Division, Frederiksberg, Denmark. tables for 31 countries. Ingco, Merlindo. 1997. "Has Agricultural Trade Liberalization 2. For example, most of the policy work on agricultural poli- Improved Welfare in the Least-Developed Countries? cies in the World Bank in the 1970s and 1980s focused on supply Yes." Policy Research Working Paper 1748. World Bank, enhancement and the elimination of taxation of agriculture. Washington, D.C. 3. These were simple averages and were not weighted for the Ingco, Merlindo, and John D. Nash. 2004. Agriculture and volume of trade. Thus some countries made large reductions in the WTO: Creating a Trading System for Development. tariffs that were already low (from 2 percent to 1 percent, for Washington, D.C.: World Bank. example, for a 50 percent reduction), while making only the min- Jansen, H. Tarp, Sherman Robinson, and Finn Tarp. 2002. imum reduction in sensitive product groups with high tariffs. "General Equilibrium Measures of Agricultural Policy Bias 4. It has not been possible to generate consistent agricultural in Fifteen Developing Countries." Discussion Paper 105. manufacturing and agricultural tariffs for earlier years. International Food Policy Research Institute, Trade and 5. The annex in the attached CD-ROM presents the detailed Macroeconomics Division, Washington, D.C. structure of tariffs for the individual countries and the year for Krueger, Anne, M. Schiff, and A. Valdes. 1992. The Political which the tariff information applies for each country. The years Economy of Agricultural Pricing Policy. A World Bank Com- are also presented in table 3.6. parative Study. Baltimore: Johns Hopkins University Press. 6. For example, EU duties on wine are 13 euros a hectoliter, Messerlin, Patrick A. 2002. "Agriculture in the Doha Round." or about $0.15 a bottle. For a $1 (c.i.f.) bottle of wine from Paper presented at the World Bank Roundtable on Policy developing countries such as Bulgaria and Moldova, that gives a Research in Preparation for the 5th WTO Ministerial Meet- high tariff rate of 15 percent. For a $10 dollar bottle of wine ing, May 20­21, Cairo. from California, the tariff rate would be just 1.5 percent, a very Nogues, Julio J. 2003. "Agricultural Exports in a Protectionist low one. World: Assessing Trade Strategies for MERCOSUR." Inter- 7. Peaks for the European Union and the United States are all American Development Bank, Trade and Integration Divi- specific tariffs, whereas the variance and peaks for Canada and sion, Washington, D.C. Japan probably do not reflect the real peaks because specific OECD (Organisation for Economic Co-operation and Develop- duties are excluded. ment). 2001. The Uruguay Round Agreement on Agriculture: 8. Elimination of the Peace Clause, which effectively prohib- An Evaluation of Its Implementation in OECD Countries. ited legal action against implicit export subsidies, could change Paris. the legality of having domestic costs much higher than export ------. 2003. Agricultural Policies in OECD Countries: Monitoring prices. Decoupling payments to producers from production lev- and Evaluation. Paris. els is another alternative that would allow income support to Quiroz, J., and L. Opazo. 2000."The Krueger-Schiff-Valdés Study farmers but eliminate its link with production decisions (see 10 Years Later: A Latin American Perspective." Economic chapter 5). Development and Cultural Change 49(1): 181­96. 9. These are average cuts, so actual cuts in each line could be Schiff, M., and A. Valdes. 1992. The Political Economy of Agricul- lower. tural Pricing Policy: A Synthesis of the Economics in Developing 10. The European Union and United States were selected Countries. Baltimore: Johns Hopkins University Press. because there are tariff equivalents for the specific duties. The Watkins, Kevin. 2003. "Northern Agricultural Policies and data for the European Union are for 1999, the last year for which World Poverty: Will the Doha `Development Round' Make a the tariff equivalents were available. The difference between Difference?" Oxfam. Paper presented at the Annual World bound and effective rates is very small in most industrial coun- Bank Conference on Development Economics Europe Con- tries and for ease of presentation, the reductions were taken ference, May 15­16, Paris. from the effective rates. World Bank. 1986. World Development Report 1986. Washington, D.C. References ------. 2001. Global Economic Prospects and the Developing Countries 2001. Washington, D.C. ABARE (Australian Bureau of Agricultural and Resource ------. 2003. Global Economic Prospects 2004: Realizing the Economics). 1999. "Reforming World Agricultural Trade Promise of the Doha Agenda. Washington, D.C. Policies." Research Report 99.12. Canberra. WTO (World Trade Organization). 2003. "Negotiations on Baffes, John. 2004. "Cotton: Market Setting and Policies." Chap- Agriculture First Draft of Modalities for the Further ter 16 in this volume. World Bank, Washington D.C. Commitments."WTO Document TN/AG/W/1/Rev.1. Geneva. 4 THE IMPACT OF AGRICULTURAL TRADE PREFERENCES, WITH PARTICULAR ATTENTION TO THE LEAST-DEVELOPED COUNTRIES Paul Brenton and Takako Ikezuki Improving the ability of the least-developed coun- of origin for many processed products, have tries (LDCs) to participate fully in world markets severely limited the role of trade preferences in can accelerate development and poverty reduction. encouraging agricultural diversification in develop- Their dependence on agriculture, together with the ing countries. Many countries remain dependent high duties levied on many agricultural imports by on the export of staple products, for which world industrial countries, suggests that preferences on prices have fluctuated wildly. agricultural products could help boost exports and While the United States, the European Union growth in developing countries. (EU), and Japan all offer preference schemes, com- In practice, however, preferences have had little paring them is difficult since each scheme differs in impact for most developing countries. First, many important respects: the group of eligible countries, agricultural products produced in developing the products covered, and the magnitude of the countries are subject to zero duties in industrial preference granted. Administrative requirements, countries, and therefore no trade preference can be especially rules of origin, also vary across schemes given. Usually these are tropical products that and across products. These differences are a major are not produced in industrial countries. Second, weakness of the current system of preferences. the primary agricultural products and processed This chapter reviews these schemes, concentrat- products with very high duties are typically ing on the preferences offered to least-developed excluded from preferences or the preference margin countries, and discusses some of the key problems is very small. For a small number of products, how- with preferences: ever, preference margins are substantial, although usually within strict quantitative limits and only for · They are unilateral concessions from industrial certain countries. Countries that have been granted countries that must be renewed, and specific preferential access for sugar and tobacco, for exam- products can be withdrawn at short notice, ple, have received large transfers because of prefer- creating too much uncertainty to stimulate new ences. These factors, together with restrictive rules investment. The authors are grateful for the comments of M. Ataman Aksoy and Harry de Gorter. 55 56 Global Agricultural Trade and Developing Countries · The most highly protected products, with the countries' products under the importing country's highest potential margins of preference, are most-favored-nation (MFN) tariffs. The principal often excluded or receive only small preference scheme governing such preferences is the General- margins. ized System of Preferences (GSP), which originated · Rules of origin for processed products constrain in the work of the United Nations Conference on the ability of countries to expand into these Trade and Development (UNCTAD) in the 1960s products. to introduce a harmonized preference scheme · Differences and inconsistencies between prefer- across donor countries (UNCTAD 2001). Because ence schemes prevent developing-country sup- preferences for particular countries are at odds pliers from developing global market strategies. with the fundamental nondiscrimination principle of the General Agreement on Tariffs and Trade If trade preferences are to assist developing (GATT) and the WTO, the Decision on Differential countries, and More Favorable Treatment, Reciprocity, and Fuller Participation of Developing Countries · The schemes should be made permanent and (called the Enabling Clause) was adopted under the comprehensive, with no product exclusions. GATT in 1979 to allow industrial countries to offer · They should be harmonized, preferably at the more favorable treatment to developing countries World Trade Organization (WTO), with com- on a nonreciprocal basis. mon and simple rules of origin. · The domestic investment environment in bene- Potential Benefits of Trade Preferences ficiary countries must be improved so that producers and investors can exploit the oppor- Tariffs introduce a wedge between the world price tunities that arise from trade preferences to of a product and the price in the domestic market. develop competitive businesses that will survive Tariff preferences give suppliers in beneficiary once those preferences are eroded. developing countries access to part or all of this · Developing countries need to diversify into a price premium that normally accrues to the broader range of exports and not become importing-country government as tariff revenue. dependent on the preferential access granted for The acquisition of these rents raises returns in the a narrow range of products. developing country and, depending on the nature · Beneficiaries should ensure preferences are inte- of competition in domestic product and factor grated as one element of a strategy for broad- markets, stimulates expansion of the activity based export expansion. concerned, with implications for wages and · Preferences for a small group of developing employment. countries should not act as a brake on the multi- The arguments underlying trade preferences are lateral liberalization of agricultural products that the small scale of industry and the low level of under the WTO. Many developing countries development in developing countries lead to high receive little or nothing from preferences but costs, which reduce the ability to compete in global would gain from a reduction of subsidies in rich markets, and to lack of diversification, which countries (which, for example, would benefit increases risks. Developing countries, especially cotton producers in Western Africa) and from least-developed countries, face much higher trade- multilateral tariff reductions in all countries. related costs than other countries in getting their Such liberalization can be achieved only through products into international markets. Some of these negotiations at the WTO. costs may reflect institutional problems within the countries themselves, such as inefficient practices and corruption, and these problems require a Trade Preferences in Principle and domestic policy response. But some costs also in Practice reflect weak transportation infrastructure in many Trade preferences allow products from developing countries and firms' lack of access to standard countries to enter industrial-country markets with trade-facilitating measures such as insurance and lower import duties than are applied to other trade finance. The Impact of Agricultural Trade Preferences, with Particular Attention to the Least-Developed Countries 57 Trade preferences can provide the premium over Most highly protected products are excluded the normal rate of return that is required to from preference schemes. When preferences are encourage investment in these economies. The granted on some products for which domestic increase in trade attributable to preferences leads to prices in industrial countries are much higher than more output and, if there are scale economies, to world prices, such as sugar in the European Union, lower costs, which stimulate further trade.1 It is traded quantities are limited to avoid undermining important, however, that the sectors that receive the distortionary policies that generate the large preferences and investment are those in which the divergence between domestic and world prices. country has a comparative advantage in the long Nevertheless, in these instances preferential access term and that investment not be based on a false can lead to substantial gains for preferred suppliers. comparative advantage based on the margin of How much of the available rents are actually preference. obtained by suppliers in developing countries depends on the nature of competition in the indus- try and the rules and regulations governing the Why Do Trade Preferences Fall Short granting of preferential access, among other fac- of Their Potential? tors. If there is little effective competition among Assessments of the impact of trade preferences sug- buyers, then exporters may be unable to acquire gest that they have not transformed the export and much of the price premium. Ozden and Olarreaga growth performance of most developing-country (2003) find that only a third of the available rents beneficiaries, although performance may have been for African exports of clothing to the United States worse without them and a few countries may have under the African Growth and Opportunity Act benefited substantially. Trade preferences have not (AGOA) actually accrue to exporters. Furthermore, enabled beneficiaries as a group to increase their satisfying the rules governing preferences raises market shares in the main preference-granting costs and reduces the extent to which the prefer- markets.2 Why? ences raise actual returns. The cost of satisfying the UNCTAD's objectives of harmonizing prefer- rules of origin in preferences schemes is a major ence regimes across countries and making prefer- reason for low rates of utilization (UNCTAD 2001; ences general and nondiscriminatory among devel- Brenton and Imagawa 2004).3 oping countries were never achieved. Industrial countries have often excluded the most heavily pro- Undesirable Effects of Preference Schemes tected products, many of which offer the greatest scope for gains by developing countries. The seg- Tariff preferences can lead to several adverse effects. mented markets for preferential-access goods make Negotiations under the Doha Round have shown the program a weak mechanism for integrating that preferences can be used to bolster external sup- developing countries into the world economy. port for highly protectionist policies in industrial Industrial countries that grant preferences uni- countries and to weaken proposals that would sub- laterally determine which countries and which stantially reduce such levels of protection. Prefer- products are included in their schemes and what ences can also create a degree of dependence that rules govern the provision of preferences--and constrains flexibility and diversification and results graduation from the program. Preference schemes in high-cost production of preferred products typically are not permanent programs but require (Topp 2001). And the beneficiaries of trade prefer- legislative renewal. And preference-granting coun- ences are not always the poorest constituents in tries have the discretion to remove countries and developing countries. When rents do accrue to the products from the program, creating uncertainty developing country, they tend to accrue to the own- and discouraging investment in developing coun- ers of the most intensively used factors. With tries to exploit available opportunities. Recently, agricultural preferences, the main beneficiaries are however, the European Union introduced the Every- the owners of land. Preferences could have a strong thing but Arms program for the least-developed impact on poverty if the landowners are poor or, countries, introducing an element of permanency when they are not poor, if policies for redistribu- into preference schemes for the first time. tion are in place. So even when preferences create 58 Global Agricultural Trade and Developing Countries substantial transfers for producers in developing ences to a particular market and the potential countries, they may not necessarily stimulate the impact of preferences is much greater. The actual long-term growth of exports or reduce poverty, and utilization of preferences, from very low rates to full they can lead to a less-diversified export base. utilization, also varies substantially across coun- tries. Also of importance is that utilization rates The Nature and Impact of tend to be lower for processed products. Preferences Offered by the United States, the European Union, and Japan The Scope of Preferences The impact of a particular scheme of trade prefer- Whether trade preference schemes assist the inte- ences on individual countries is determined by gration of developing countries into world markets several factors: depends on the breadth of the preferences offered in terms of the number and importance of eligible · The scope of preferences in terms of the range of products. products covered. Products subject to tariff quotas complicate the · The importance of products eligible for prefer- assessment of the impact of trade preferences. Dur- ences in the export and production structure of ing the Uruguay Round of world trade negotia- the beneficiary country. tions, industrial countries agreed to reduce tariffs · The margins of preference, determined by the on a range of sensitive agricultural products but height of the MFN tariff and the size of the only for limited quantities of imports, often creat- preference. ing two or more tariff lines for each product: the · Actual utilization of preferences. To a large duty on in-quota quantities and the duty (often extent this reflects the costs of satisfying the very high) on additional out-of-quota imports. rules, mainly the rules of origin, governing pref- Preferences are offered on the in-quota quantities erences. If the costs of compliance exceed the only, and once the quota is reached, preferences are margin of preference, the preference will not no longer available. Quotas can be global (available be used. to all eligible countries) or bilateral (limits are spec- · The extent to which preferences facilitate diver- ified for a particular country). With bilateral limits, sification into a broader range of products. This quantities may not be sufficient to induce invest- is determined by the coverage of the scheme, the ments in raising capacity, whereas for preferences margins of preference on products not currently based on global tariff quotas, uncertainty over exported, and the rules of origin relating to when the quota will be filled dampens interest in these products. investment or even in exporting. Thus the lack of preferences for out-of-quota Whether such preference opportunities are quantities is important in assessing the impact of actually exploited depends on the domestic invest- preference schemes. The analysis here includes out- ment environment in the beneficiary country and of-quota rates in calculating the average duty on the extent to which legal characteristics of the pref- products not covered by preferences even if the quo- erence scheme constrain investment decisions. The tas are not exceeded, because of the discouraging economic impact of the preferences offered by the impact of the tariff quotas. This approach differs United States, the European Union, and Japan vary from that of preference-giving countries, which typ- enormously across beneficiary countries. For some ically assume that if exports from a preference- countries exports are dominated by products that receiving country or group of countries do not do not receive preferences, and there has been little exceed the preferential quota, the product is fully success in diversification. This is especially the case covered by the scheme and the out-of-quota rates for countries dependent on products that are cur- are not relevant. Of course, the obvious response is rently subject to zero import duties in developed that if the out-of-quota rates are not relevant, there countries, such as coffee and cocoa. For other coun- is no reason not to offer full duty- and quota-free tries, however, all exports are eligible for prefer- access. The Impact of Agricultural Trade Preferences, with Particular Attention to the Least-Developed Countries 59 U.S. Preferences under the GSP and AGOA little impact on the LDCs, reducing the average tariff by just 0.2 percentage points, but it offers The United States has offered preferences under the non-LDCs enhanced preferences similar to those Generalized System of Preferences (GSP) since the available to LDCs under the GSP. All these average mid-1970s, with a significant increase in coverage tariffs include the out-of-quota duties for tariff for low-income countries in 1997. The current GSP quota products. program expires at the end of 2006. In 2003, 143 Comparing MFN duties on the products cov- developing countries were eligible for preferences ered by preferences and average duties on products under the GSP. There are no partial preferences, so excluded from preferences shows that the average the preferential rate on all included products is margin of preference on products under the GSP is zero. However, the preferences can be withdrawn at 3.6 percent for non-LDCs and 5 percent for LDCs any time. In addition, the GSP contains safeguards (table 4.2). AGOA enhances the preferences avail- in the form of benefit ceilings for each product and able for non-LDCs by including products subject to country, known as competitive need limitations an average duty of 7 percent. However, the average (these do not apply to LDCs). A country loses its tariff on products excluded from preferences is GSP eligibility for a product if it supplies more than more than 30 percent. The GSP and AGOA do not 50 percent of U.S. imports of that product or if its exports exceed a certain dollar value.4 affect the maximum duty that can be applied to imports from LDCs (more than 160 percent for The African Growth and Opportunity Act of groundnuts, an important product for a number of 2000 offers improved market access to 48 Sub- African countries). Saharan African countries subject to certain criteria The duties shown in tables 4.1 and 4.2 are regarding basic human rights and the rule of law. unweighted averages. They do not capture actual The competitive needs limitations of the GSP do duties being levied on developing-country exports not apply to AGOA preferences. The current but rather the duties that would apply if developing scheme expires in 2015. So far, 38 countries have countries exported a completely diversified bundle been granted eligibility for AGOA preferences. of agricultural products. In practice, the duties The average duty on agricultural goods from actually levied on many countries are close to zero countries that do not receive preferences in the since these countries export a bundle of exports United States is 7.3 percent (table 4.1). The prefer- ences available under the GSP for non-LDCs TABLE 4.2 Average Unweighted MFN reduce the average tariff to 6.2 percent. The impact Tariffs on Agricultural Products of the GSP on LDCs is more substantial, reducing Covered by GSP and AGOA in the average tariff to just under 4 percent. AGOA has the United States, 2003 (percent) TABLE 4.1 Average Unweighted Tariffs on Category Non-LDC LDC Agricultural Products in the United States, 2003 Total GSP 3.6 5.4 (percent) GSP 3.6 3.6 GSP LDC -- 7.0 Category Non-LDC LDC Total AGOA 7.0 9.4 Excluded lines 32.5 32.8 MFN rates 7.3 7.3 GSP beneficiaries 6.2 3.9 -- Not available AGOA beneficiaries 3.8 3.7 Note: Data for calculated duties and customs value for the GSP group were used to derive ad Note: Because of the potential effect on decisions valorem equivalents for specific duties. When to export and invest, average tariffs include out- there are zero duties from the GSP group of of-quota tariffs on tariff quota products even if countries, data for total imports were used to quotas are not filled. calculate the ad valorem equivalent. Source: Calculated using data from U.S. Interna- Source: Calculated using data from U.S. Interna- tional Trade Commission dataweb. tional Trade Commission dataweb. 60 Global Agricultural Trade and Developing Countries TABLE 4.3 Number of Agricultural Tariff Lines Liberalized under GSP and AGOA Programs in the United States, 2003 Category Non-LDCs LDCs Total tariff lines 1,723 1,723 Total GSP 519 1,038 GSP LDC 519 (38) 547 (158) AGOA 541 (120) 26 Duty-free lines 440 440 Dutiable lines (MFN) 223 219 Note: The numbers in parentheses are the number of product lines relating to in-quota duty rates for products subject to tariff quotas. Source: Calculated using data from U.S. International Trade Commission dataweb. concentrated on zero or low-duty products. Devel- sugar products, chocolate, prepared foodstuffs, and oping countries in Africa currently export almost tobacco products. no products that are subject to tariff quotas--the main exceptions are sugar and tobacco. But this EU Preferences under the GSP and Cotonou may simply reflect the fact that very high duties can be levied once the quota is reached.5 Liberalization Agreement of many of these products under AGOA or GSP is The current GSP scheme of the European Union, unlikely to have a substantial impact on trade in the which runs to the end of December 2004, has two short term, but it could encourage investment in categories of products: nonsensitive, for which future capacity in certain countries. duties are suspended; and sensitive, which face a Data on the number of tariff lines liberalized flat rate reduction of 3.5 percentage points from the under U.S. preference programs show that a quar- MFN rate. A number of products, including meats, ter of tariff lines already have zero MFN duties dairy products, certain vegetables, cereals, some (table 4.3). For the LDCs, AGOA liberalizes only an prepared foodstuffs, and wine are entirely excluded additional 26 agricultural tariff lines, or less than from the scheme. Among eligible products, propor- 2 percent of the total number of agricultural lines tionate reductions are high for most industrial and just under 12 percent of the remaining dutiable products, for which the average MFN tariff is 4 per- lines (those lines for which the MFN duty is not cent, but relatively low for many agricultural prod- zero). The main products liberalized under AGOA ucts, for which the average MFN tariff is almost have already been liberalized for LDCs under the 20 percent. The EU tariff structure for agricultural GSP. For non-LDCs, AGOA adds 541 products to products is extremely complicated, with more than the 519 products already eligible for duty-free pref- 45 percent of product lines subject to non-ad- erences for developing countries under the GSP. valorem duties. This complexity is reflected in sim- Hence, the potential impact is much greater for ilar complexity in preferences granted. non-LDCs.6 Specific duties, those based on physical rather Under AGOA more than 200 agricultural tariff than monetary values, are reduced by 30 percent, lines have MFN duties but no preferences. These except when they are combined with ad valorem amount to 17 percent of the number of dutiable duties (as in a range of processed agricultural prod- agricultural tariff lines in the U.S. schedule, ucts of interest to developing countries), when they although they protect much more than 17 percent are not reduced. Typically, the specific duties pro- of U.S. agricultural production. More than 150 of vide the greatest part of the protection on these these lines relate to the over-quota rates for prod- products. For a number of products, primarily ucts subject to tariff rate quotas. These products fruits and vegetables, the European Union applies a include certain meat and dairy products, many system of minimum reference prices that vary by The Impact of Agricultural Trade Preferences, with Particular Attention to the Least-Developed Countries 61 season, despite the dubious compatibility of the TABLE 4.4 Average Unweighted Tariffs approach with WTO rules. This can lead to a very on Agricultural Products in complex structure of preferences (box 4.1). Mini- the European Union, 2002 (percent) mum duties specified in the European Union's Common Customs Code do not apply to products Category Average Tariff covered under the GSP. Within the GSP, the European Union discrimi- MFN rates 17.3 nated in favor of the least-developed countries. All GSP beneficiaries 15.3 imports of industrial products and a range of agri- ACP beneficiaries 6.9 cultural products from these countries entered duty free, but a significant number of agricultural Note: For seasonal rates, the duty applied on July 1 is used, the high season for most fruits and products still faced some market access barriers. vegetables. For products for which it is not These were removed under the Arms initiative possible to calculate an ad valorem equivalent introduced in 2001, which grants duty-free access, of the complex duties that are applied (for without any quantitative restrictions, to imports of example, for chocolate the duty depends on the milk and sugar content), an ad valorem duty of all products from the least-developed countries, 30 percent was assigned for the MFN rate and except arms and munitions. Liberalization was 20 percent for the ACP rate. These are conserva- immediate except for three products (fresh tive assumptions since many of these complex duties are likely to be prohibitive. In 2002 there bananas, rice, and sugar), for which tariffs gradu- were 161 lines for which the ad valorem ally will be reduced to zero (in 2006 for bananas equivalent could not be computed. and in 2009 for rice and sugar). The effect of the Source: Calculated from EU Commission data and Arms initiative will be limited in the short run since World Trade Organization Integrated Data Bank. the LDCs were not exporting the products that were immediately liberalized (Brenton 2003). of the value of ACP preferences in the late 1990s, Because preferences for the least-developed including industrial products, which are all eligible countries are granted for an unlimited period and for duty-free access (McQueen 1999). are not subject to periodic review, the Everything The average duty in the European Union is very but Arms program should provide greater certainty high, at more than 17 percent. Countries eligible of market access and therefore stimulate a greater for GSP benefits on agricultural products face a production response by existing products and a slightly lower average duty of 15.3 percent, and conducive environment for exports of a wider ACP countries face a much lower average duty of range of products. This is a crucial aspect of the about 7 percent (table 4.4). The average duty that program. The challenge for developing countries is would be levied on products covered by the GSP if to create a climate that allows investment to take those preferences were removed is about 14 percent place in activities in which a comparative advantage (table 4.5). Full preferences tend to be granted on can be sustained in the long run. agricultural products with lower MFN rates, However, these changes may be partly under- whereas those with higher MFN rates tend to mined by the inclusion of a new reason for sus- receive only partial preferences. Products not pending preferences: "massive increases in imports granted preference under the GSP scheme tend to of products relative to the usual levels of production be very-high-duty products, with an average tariff and export capacity" (our emphasis). This could of more than 25 percent. constrain large-scale investment to transform the The average duty on products covered by the production capacity in a particular country and Cotonou Agreement is more than 21 percent, and discourage diversification into new products.7 the preferences available are much deeper than The European Union offers enhanced prefer- those under the GSP. And while very-high-duty ences beyond those of the GSP to Sub-Saharan products tend to receive only partial preferences, African, Caribbean, and Pacific countries (the ACP many high-duty products excluded from the GSP countries) under the Cotonou Agreement. There receive preferences under Cotonou. The average are individual protocols for bananas, beef, veal, and duty on excluded products is just under 10 percent. sugar. These products accounted for three-quarters Nevertheless, preferences do not reduce the 62 Global Agricultural Trade and Developing Countries Box 4.1 The EU System of Entry Prices: The Example of Tomatoes The EU entry price system for imports of vegeta- are granted under the GSP. For African, bles such as tomatoes consists of two sets of Caribbean, and Pacific (ACP) countries, prefer- tariffs that vary according to the price and sea- ences are granted from January to April and in the son. If the import price is higher than a specified final 10 days in December. The preference takes level (which varies by season), only an ad valorem the form of a reduction in the ad valorem duty; duty is applied. If the import price is lower than there is no reduction in the specific duty. This this level, then a specific duty is applied as well, form of entry price system is applied to 34 agri- which varies by price and season. No preferences cultural products, mainly fruits and vegetables. Entry Prices and Duties for Tomatoes Entry Price Time (euros per 100 kg) MFN ACP GSP GSP LDC Jan­Mar 84.6 8.8 3.5 No preference 0 April 112.6 8.8 3.5 No preference 0 May 1­14 72.6 8.8 No preference No preference 0 May 15­31 72.6 14.4 No preference No preference 0 June­Sept 52.6 14.4 No preference No preference 0 Oct 62.6 14.4 No preference No preference 0 Nov 1­ Dec 20 62.6 8.8 No preference No preference 0 Dec 21­Dec 31 67.6 8.8 3.5 No preference 0 Specific Duties When Import Price Falls below Set Levels Import Price (euros per 100 kg) MFN Duty ACP Duty January 82.9 to 84.6 8.8% + 1.7 euro/100 kg 3.5% + 1.7 euro/100 kg 81.2 to 82.9 8.8% + 3.4 euro/100 kg 3.5% + 3.4 euro/100 kg 79.5 to 81.2 8.8% + 5.1 euro/100 kg 3.5% + 5.1 euro/100 kg 77.8 to 79.5 8.8% + 6.8 euro/100 kg 3.5% + 6.8 euro/100 kg 0 to 77.8 8.8% + 29.8 euro/100 kg 3.5% + 29.8 euro/100 kg July 51.5 to 52.6 14.4% + 1.1 euro/100 kg 14.4% + 1.1 euro/100 kg 50.5 to 51.5 14.4% + 2.1 euro/100 kg 14.4% + 2.1 euro/100 kg 49.4 to 50.5 14.4% + 3.2 euro/100 kg 14.4% + 3.2 euro/100 kg 48.4 to 49.4 14.4% + 4.2 euro/100 kg 14.4% + 4.2 euro/100 kg 0 to 48.4 14.4% + 29.8 euro/100 kg 14.4% + 29.8 euro/100 kg For example, an ACP exporter of tomatoes that even with preferences, is higher than the duty- tries to sell in the EU market in January at a price inclusive price of the high-cost supplier. Hence, of, say, 67 euros per 100 kilograms would face specific duties act as an implicit preference an ad valorem equivalent duty of 49.8 percent toward high-cost suppliers and against lower- with the MFN rate being 53.3 percent. A higher cost developing countries, although in this case, cost non-ACP producer who sells at 80 euros per if sufficient information is available, there is an 100 kilograms would face a duty of 13 percent. opportunity for the low-cost ACP supplier to The duty-inclusive price of the low-cost supplier, raise its price and pay a lower duty. The Impact of Agricultural Trade Preferences, with Particular Attention to the Least-Developed Countries 63 maximum duty that can be applied (a duty of more put. Cotonou provides full preferences (100 per- than 200 percent on milk and cream). cent duty reduction) for 50 percent of the total Cotonou preferences cover 81 percent of agri- number of tariff lines and partial reductions for cultural tariff lines (table 4.6). Of the remaining 31 percent of products (typically removal of the ad lines, 14 percent have zero MFN duties, and 5 per- valorem component but not the specific duty). cent cover products excluded from preferences. Most of the products are highly sensitive and highly Again, this 5 percent of lines will be protecting taxed imports. The ad valorem equivalent of these much more than 5 percent of EU agricultural out- specific duties is often very high (see table 4.5). Japan's GSP Scheme TABLE 4.5 Average Unweighted MFN Tariffs on Agricultural Products Japan offers GSP preferences to 164 developing Covered by GSP and Cotonou countries. The current scheme expires in 2011. The Agreement, 2002 scheme provides enhanced preferences for LDCs, (percent) with partial preferences deepened to 100 percent Category Average Tariff cuts and (since April 2003) greater product coverage (which is not captured here). There are no explicit Total GSP 14.1 quantitative ceilings on preferences, although there Lines with full preferences 7.0 are safeguard mechanisms and a country's exports Lines with partial preferences 15.2 are excluded if they exceed 25 percent of Japan's Excluded GSP lines 26.3 total imports and 1 billion yen in value. Total Cotonou 21.3 The average MFN tariff on Japanese imports of Lines with full preferences 13.7 agricultural products in 2002 was 15.6 percent Lines with partial preferences 33.0 (table 4.7). GSP preferences reduced this to Excluded Cotonou lines 9.6 15.1 percent for non-LDCs, an average preference margin of 0.5 percentage point. The average mar- Source: Calculated using data from EU Commis- gin for the slightly deeper preferences for LDCs was sion and World Trade Organization Integrated Data Bank. 1.4 percentage points. Again, it must be noted that TABLE 4.6 Number of Tariffs Lines Liberalized under EU Preferences for ACP Countries, 2002 Tariff Lines Share of Total Category Number (percent) Total lines 2,354 MFN duty-free 334 14 Total ACP 1,905 81 Full reduction preferences 1,181 50 Partial reduction preferences 724 31 Dutiable MFN lines 115 5 Main sectors containing products excluded from preferences Wine Main sectors containing products subject to partial preferences Meat, dairy, fruits and vegetables, grains and flour, prepared food stuffs Source: Calculated using data from EU Commission. 64 Global Agricultural Trade and Developing Countries TABLE 4.7 Average Unweighted Tariffs on TABLE 4.8 Average Unweighted MFN Agricultural Products in Japan, Tariffs on Agricultural Products 2002 Covered by GSP in Japan, 2002 (percent) (percent) Category Non-LDCs LDCs Category Non-LDCs LDCs Average MFN 15.6 15.6 Total GSP 10.4 9.8 Average applied 15.1 14.2 Full preference 7.3 9.8 preferential duty Partial preference 12.0 Excluded lines (MFN) 20.8 21.5 Note: Specific duties were converted to ad valorem equivalents based on the total value Source: Calculated using data from United and quantity of imports from developing Nations Conference on Trade and Development countries. When that information was not (TRAINS). available, the value and quantity of imports from all sources was used. For tariff lines for which there were no imports, an ad valorem equivalent of 30 percent was assumed-- probably an underestimate since these duties are likely to be prohibitive. Source: Calculated using data from United Nations Conference on Trade and Development TRAINS. TABLE 4.9 Tariffs Lines Liberalized under Japan's GSP Preferences, 2002 Non-LDCs LDCs Number of Share of Total Number of Share of Total Category Tariff Lines (percent) Tariff Lines (percent) Total lines 2,014 2,014 MFN duty-free 393 20 393 20 Total GSP 221 11 298 15 Full preferences 80 4 298 15 Partial preferences 141 7 0 Dutiable lines (MFN) 1,400 70 1,323 66 Main sectors containing products Meat, fish, dairy, cereals, prepared meat and fish, sugar, cocoa, excluded from preferences prepared food products Source: Calculated using data from the United Nations Conference on Trade and Development (TRAINS). these are not average duties paid since few imports under the GSP cover 11 percent of agricultural are in the high-duty categories. products for non-LDCs and 15 percent for LDCs The average duty on products covered by the (table 4.9). The 2003 reform of the GSP added an GSP was 10.4 percent for non-LDCs and 9.8 per- additional 198 products (or 10 percent of total tar- cent for the LDCs (table 4.8). The duty on products iff lines) to preferences for LDCs. For non-LDCs excluded from preferences is high relative to duties most products under preferences receive only a on products covered by preferences, even when partial reduction in duties, and 71 percent of agri- conservatively estimated, at about 21 percent (see cultural products are excluded from preferences, note to table 4.7). while preferences were not available for 67 percent Some 20 percent of agricultural tariff lines in of tariff lines for LDCs in 2002 (falling to 57 per Japan are subject to zero duties, while preferences cent in 2003). The Impact of Agricultural Trade Preferences, with Particular Attention to the Least-Developed Countries 65 Proportion of Trade Covered preferences (table 4.11) and show the following: by Preferences · The value of agricultural exports to the Euro- U.S. Preferences pean Union is much larger than that of exports Examination of the proportion of developing- to the United States, for both processed and pri- country exports covered by U.S. preference pro- mary products. Again, the value of exports is grams shows that (table 4.10) smaller for processed products than for primary products. · Exports of processed agricultural products are · A much larger proportion of exports are eligible much smaller than exports of primary agricul- for preferences than in the United States because tural products. fewer export products have MFN duties of zero. · For a large proportion of primary product Two-thirds or more of exports are eligible for exports (more than 70 percent for the three preferences. groups of countries), there are no preferences · Products not eligible for preferences constitute a since the MFN duty is zero. A much larger very small proportion of current exports. proportion of processed exports is eligible for preferences. Japanese Preferences · Preference use is high for primary products An examination of the proportion of developing (more than 80 percent) and higher than the rate country exports covered by GSP exports to Japan of preference use for processed products. shows the following: (table 4.12) · Products not eligible for preferences constitute a small proportion of current exports. · As a market for the exports of agricultural prod- ucts of African LDCs, Japan is smaller than the European Union and about the same size as the EU Preferences United States. Several findings stand out in an examination · Exports from other LDCs, including those in of the proportion of exports covered by EU Asia, are considerably smaller than those from TABLE 4.10 Exports to the United States under AGOA and by other LDCs under the GSP, 2002 (US$ millions) GSP and AGOA Preferences Category GSP+AGOA Non-LDCs GSP+AGOA LDCs GSP LDCs Basic agricultural commodities Total exports to United States 600 247 122 Exports duty free 431 (72) 190 (77) 114 (93) Exports for which preferences requested 149 (25) 47 (19) 7 (6) Exports eligible, but preferences not requested 15 (1) 11 (4) 1 (1) Exports not eligible for preferences 6 (1) 0 (0) 0 (0) Processed agricultural products Total exports to United States 133 51 2.3 Exports duty free 55 (41) 9 (18) 0.9 (41) Exports for which preferences requested 61 (46) 31 (61) 0.6 (29) Exports eligible, but preferences not requested 11 (8) 10 (20) 0.7 (30) Exports not eligible for preferences 5 (4) 0 0 Note: Numbers in parentheses are shares of exports for each category of agricultural exports. Source: Calculated using data from U.S. International Trade Commission dataweb. 66 Global Agricultural Trade and Developing Countries TABLE 4.11 Exports to the European Union from ACP Beneficiaries, 2002 (US$ millions) Africa Category Africa LDC Africa Non-LDC Caribbean Pacific Basic agricultural commodities Total exports to the European Union 1,904 5,159 1,018 310 Exports duty free 533 (28) 2,065 (40) 55 (5) 68 (22) Exports for which preferences requested ACP+GSP 1,188 (62) 2,623 (51) 874 (86) 223 (72) Exports eligible, but preferences not requested 183 (10) 471 (9) 89 (9) 19 (6) Exports not eligible for preferences 0.2 (0) 0.5 (0) 0.18 (0) 0 (0) Processed agricultural products Total exports to the European Union 303 1,414 455 15 Exports duty free 30 (10) 16 (1) 8 (2) 1 (10) Exports for which preferences requested ACP+GSP 235 (78) 1,186 (84) 416 (92) 8 (57) Exports eligible, but preferences not requested 37 (12) 212 (15) 30 (7) 5 (34) Exports not eligible for preferences 0.1 (0) 0 (0) 0.1 (0) 0 (0) Note: Numbers in parentheses are shares of exports for each category of agricultural exports. Source: Calculated using data from EU Commission. TABLE 4.12 Exports to Japan from LDCs in 2002 (US$ millions) Category All LDCs African LDCs Other LDCs Basic agricultural commodities Total exports to Japan 381 241 140 Exports duty free 131 (34) 124 (51) 8 (5) Exports for which preferences requested 62 (16) 52 (22) 10 (7) Exports eligible, but preferences not requested 3.7 (1) 3 (1) 0 (0) Exports not eligible for preferences 184 (48) 62 (26) 122 (87) Processed agricultural products Total exports to Japan 40.8 39.4 1.4 Exports duty free 36.7 (90) 36.2 (92) 0.5 (35) Exports for which preferences requested 3.5 (9) 2.8 (7) 0.6 (44) Exports eligible, but preferences not requested 0.3 (0.8) 0.3 (0.7) 0.1 (3.8) Exports not eligible for preferences 0.3 (0.7) 0.1 (0.2) 0.2 (16.5) Note: Numbers in parentheses are shares of exports for each category of agricultural exports. Source: Calculated using data from Ministry of Finance, Japan. African LDCs. This may reflect the structure of remaining exports to Japan, 23 percent are eligi- protection and preferences in Japan. ble for preferences, and 26 percent are excluded · For African LDCs, more than 50 percent of from preferences. For other LDCs, only 5 per- exports of basic agricultural products enter the cent of exports of basic agricultural products Japanese market at zero duty MFN rates. Of the enter duty free under zero percent MFN rates, The Impact of Agricultural Trade Preferences, with Particular Attention to the Least-Developed Countries 67 just 7 percent are eligible for preferences and for preferences made use of those preferences, 87 percent are excluded from preferences. while 85 percent of eligible exports to the United · Exports of processed products to Japan are a States did. Botswana used 99 percent of available very small share for African LDCs and are non- preferences in the European Union but only 22 per- existent for non-African developing countries. cent of those in the United States. Within the overall figures for each preference- The Value of Preferences granting market, there are large variations across countries. For example, while the value of exports An estimate of the value of trade preferences to the eligible for preferences exceeds 20 percent of total exporting countries was also calculated, using the exports to the European Union for non-LDC amount of exports actually receiving preferences African countries, it is less than 20 percent for 11 and the margin of preference to derive the tariff LDC African countries and higher than 80 percent revenue that would have been paid without prefer- for 11 other LDC African countries. There are also ences. This overstates the actual transfers to devel- important differences across schemes for the same oping countries because some of the rent will be country. For example, 90 percent of Guinea- acquired by importers in the preference-granting Bissau's exports to the European Union are eligible country, especially if there is a single buyer, and for preferences, yet none of its current exports to because of the administrative costs incurred by the United States receive preferences because the exporters, such as compliance with rules of origin.8 exports are subject to an MFN rate of zero. For Average transfers to LDCs under AGOA and the Mozambique, by contrast, 97 percent of exports to GSP amount to less than 1 percent of their agricul- the European Union and 86 percent of exports to tural exports to the United States in 2002 (figure 4.1). the United States are eligible for preferences. For most countries, preferences have a negligible There are also substantial variations across impact under the current structure of exports. Pref- countries in their use of available preferences. For erences of this magnitude will not encourage addi- example, in 2002, only 10 percent of Ethiopia's tional investment in these countries and will do little exports to the European Union that were eligible to mitigate the high transaction costs these countries FIGURE 4.1 The Value of Preferences Requested under GSP and AGOA Programs of the United States, as a Share of Agricultural Export to the United States (percent) Percent 8 7 AGOA LDC Group AGOA US Non-LDC Group UN LDCs Group 6 Processed products 5 4 3 2 1 0 Niger Rep. Mali Chad Benin Laos Haiti Leone Verde Faso Togo Congo Malawi Eritrea Gabon Kenya Ghana Congo Africa Nepal Kiribati Liberia Sudan Gambia Angola Rwanda Principe Lesotho Uganda Guinea Ethiopia Zambia Nigeria Yemen Bhutan Samoa Tanzania Average Djibouti Namibia d'Ivoire Senegal Somalia Guinea Islands Mauritius Average Burundi Comoros Myanmar Maldives Vanuatu Average Mauritania Seychelles Cameroon Botswana Swaziland Cambodia The African and Madagascar Sierra Cape Dem. Mozambique Côte Rep. Eq. South Afghanistan Bangladesh Burkina Group Guinea-Bissau Group Group Solomon Tomé Central São Source: Calculated using data from USITC dataweb. 68 Global Agricultural Trade and Developing Countries face in accessing world markets. Malawi may be an amounts to 6 percent of the value of their exports exception. It receives a transfer equivalent to just to the European Union. A large number of coun- over 7 percent of the value of exports to the United tries receive little or no benefit from EU preferences States, thanks largely to exports of processed prod- on agricultural products. For 10 of the LDCs, ucts (mainly tobacco). Haiti is the only other LDC to including Chad, Niger, and Rwanda, the value of receive significant preferences. It is granted more EU preferences amounts to less than 2 percent of favorable treatment than the GSP under the the value of exports. Caribbean Basin Initiative, mainly for exports of As with the U.S. and EU programs, Japanese tropical fruits such as mangoes. Preferences for non- preferences for a few countries under the GSP pro- LDCs in the United States are also small, with the gram in 2002 are substantial, primarily for fish average transfer being less than 1 percent of the value products (figure 4.3). For the majority of LDCs, of exports to the United States. however, transfers due to preferences are zero. Only Under the Cotonou and GSP preference 6 of the 46 LDCs receive a transfer greater than 1 schemes, the highest transfers go to non-LDCs, percent of the value of agricultural exports to Japan mainly as rents on sugar exports (figure 4.2). Mau- in 2002. ritius, for example, is a major beneficiary receiving transfers in 2002 equivalent to more than 52 per- Preferences and Export cent of the value of its agricultural exports to the Diversification European Union in that year. The value of prefer- ences for sugar accounted for more than 30 percent A key problem for the least-developed countries of the value of exports for Fiji, the Republic of has been their export reliance on a small number of Congo, Swaziland, and a number of Caribbean agricultural commodities. This export concentra- countries. Among LDCs, preferences on sugar tion leaves them vulnerable to external shocks and resulted in substantial transfers to Burkina Faso, the downward trend in commodity prices. Prefer- Malawi, and Mozambique. ences could provide incentives for investment in While transfers to a small number of LDCs sectors in which countries have a comparative under the Cotonou Agreement are substantial, the advantage but that are not being exploited because average transfer across all LDC beneficiaries of difficulties in accessing export markets. FIGURE 4.2 The Value of Preferences Requested under Cotonou and GSP Programs of the EU, as a Share of Agricultural Export to the EU (percent) Percent 70 60 LDC Group Non-LDC Group 50 40 Value of preferences from sugar exports 30 20 10 0 Fiji Chad Haiti Mali Togo Benin Faso Niue Rep. Tuvalu Islands Leone Niger Verde Palau Nauru Lucia Belize Nevis Burundi Guinea Congo Samoa Eritrea Liberia Sudan Angola Malawi Kiribati Islands Tonga Nigeria Ghana Kenya Islands Gabon Congo Rwanda Republic Principe Somalia Djibouti Guinea d'Ivoire Vanuatu Uganda Ethiopia Gambia Guinea Comoros Average Zambia Senegal Lesotho Vincent Tobago Tanzania Bahamas Grenada Barbuda Namibia Surinam St Jamaica Guyana Non-LDC Dominica Mauritius Eq. Mauritania teô Zimbabwe Botswana Micronesia Cameroon St Swaziland Barbados Seychelles and Sierra and Dem. The Madagascar Cape Burkina New and é LDC Mozambique Cook C and Guinea-Bissau Kitts Solomon African Marshall Dominican St Tom Papua oã Antigua Trinidad Central S Source: Calculated using data from EU Commission. The Impact of Agricultural Trade Preferences, with Particular Attention to the Least-Developed Countries 69 FIGURE 4.3 The Value of Preferences for LDCs under the GSP Program of Japan, as a Share of Agricultural Export to Japan (percent) Percent 8 7 6 5 4 3 2 1 0 Laos Haiti Mali Faso Niger Chad Togo Guinea Liberia Congo Benin Verde Sudan Nepal Burundi Uganda Zambia Samoa Kiribati Bissau Guinea Eritrea Tuvalu Yemen Malawi Bhutan Islands Angola of Vanuatu Rwanda Principe Republic Somalia Lesotho Ethiopia Senegal Gambia Tanzania Cambodia Myanmar Maldives Mauritania Afghanistan and Bangladesh Burkina Mozambique Madagascar Cape The é Guinea African Solomon Republic Equatorial Tom oã S Central Source: Calculated using data from Ministry of Finance, Japan. Preferenceshavedonelittletoincreasethediversi- possibly contributing are the administrative rules fication of agricultural exports. Only 4 of the 38 governing the granting of preferences, described countries eligible for preferences under AGOA have below. However, the main factor constraining diver- significantly diversified their exports of agricultural sification is likely to be the poor domestic invest- products over the last 20 years (Ghana, Nigeria, ment climate in most of the beneficiary countries. South Africa, and Tanzania). For the other 34 coun- tries, five or fewer products accounted for more than Constraints on Preferences and 90 percent of their agricultural exports in both 1982 Diversification: Rules of Origin and 2002.9 The same pattern holds for African exports to the European Union.In 1982,for 37 of the Rules of origin are essential to ensure that prefer- 44 African ACP countries, five or fewer products ences are granted only to exporters from eligible accounted for more than 90 percent of exports to the countries. The nature of the rules of origin, how- European Union.By 2002 only two of these countries ever, are a key element determining the extent to had diversified their exports to the European Union which countries are able to take advantage of the to reduce the importance of the main five export preferences available to them. For a product pro- products. None of the LDCs had diversified exports duced in a single stage or wholly obtained in one to Japan. In 2002 the five top products accounted for country, origin is relatively easy to establish. Pri- 90 percent or more of every country's exports. mary agricultural products typically fall into this Much of this failure to stimulate export diversifi- category. Proof that the product was produced or cation likely results from several features of the pro- obtained in the preferential trade partner is nor- grams. The uncertain duration of the preferences mally sufficient. The process of proving conform- granted, the exclusion of many products with the ity, however, may incur costs that reduce the value largest preference margins, and the inadequacy of of the preferences. preference margins for making investments in new For processed manufactured products, rules of activities attractive, given the high transaction costs origin stipulate how much or what kind of domes- of operating in the least-developed countries. Also tic processing must take place. The U.S. GSP scheme 70 Global Agricultural Trade and Developing Countries has a value-added requirement of 35 percent for all Cooperation), but it excludes the ACP countries. products. The U.S. scheme also allows for cumula- Hence LDC members of the ACP that are eligible to tion between selected countries, so that value added export to the EU under Everything but Arms may in those countries can be counted toward the over- often prefer to continue exporting under the Coto- all value-added requirement for the product nou Agreement because of the more liberal rules of exported to the United States. AGOA permits such origin (Brenton 2003). cumulation among all Sub-Saharan preference The rules of origin for the Japanese GSP require trade partners. In practice, many processed food a change of tariff heading to demonstrate that a products are excluded from the GSP and AGOA. substantial transformation has taken place, EU rules of origin are product specific and although there is a list of products for which spe- sometimes complex. Some products require a cific criteria are defined. Thus, for example, flour or change of tariff heading, some have a value-added similar products cannot be produced from requirement, and some are subject to a specific imported grains. Cumulation is allowed among a manufacturing process requirement. In some cases limited group of Southeast Asian countries these methods are combined. For certain industrial (Indonesia, Malaysia, Philippines, Thailand, and products, a choice among alternative methods is Vietnam). permitted--for example, either a change of tariff An important feature of these preferential trade heading or satisfaction of a value-added require- schemes is the requirement of direct consignment ment. This more flexible approach is not available or direct transport. This stipulates that goods for for agricultural products. which preferences are requested be shipped directly For many products, the EU rules require a to the destination market. If they are in transit change of chapter, which is even more restrictive through another country, documentary evidence than a change of heading. Some of the EU rules may be required to show that the goods remained exclude some changes in tariff classification by pro- under the supervision of the customs authorities of scribing the use of certain imported inputs. For the country of transit, did not enter the domestic example, the rule of origin for bakery products market there, and did not undergo operations such as bread, pastry, cakes, and biscuits requires a other than unloading and reloading. In practice, it change of tariff heading except for any heading in can be very difficult to obtain the necessary docu- chapter 11 (products of the milling industry), mentation from foreign customs. meaning that bakery products cannot use imported In general, preferences are more effective when flour, a restrictive requirement for countries with- the rules of origin are simple and easy to apply. Fur- out a competitive milling industry. Products that ther, the value of OECD (Organisation for Eco- include sugar have to demonstrate that the value of nomic Cooperation and Development) preferences any imported sugar does not exceed a certain pro- would be enhanced by greater uniformity in the portion of the price of the product. way given products are treated in the different While the European Union has sought to har- schemes. Thus a product that qualifies for prefer- monize the processing requirements for each ences in one market should be granted preferential product across preference programs, a number of access to all other OECD countries. The WTO general rules vary substantially across different would be an appropriate forum for discussing and schemes, particularly those on the nature and agreeing on a common set of rules of origin. extent of cumulation and the tolerance rule. There are important differences in the rules of origin Preference Erosion by Multilateral among the Everything but Arms program, the GSP, Tariff Reductions and the Cotonou Agreement. For example, the Cotonou Agreement permits full cumulation. The As multilateral tariff reductions are negotiated at GSP has more limited partial cumulation that can the WTO, the margins of preference available to take place only within four regional groupings developing countries decline. Whether developing (Association of South-East Asian Nations, Central countries lose overall from multilateral liberaliza- American Common Market, the Andean Commu- tion depends on the extent of negotiated tariff nity, and South Asian Association for Regional reductions on products that currently receive The Impact of Agricultural Trade Preferences, with Particular Attention to the Least-Developed Countries 71 preferential access and on the importance of prod- trade liberalization contributes to this outcome by ucts excluded from preferences or not using current ensuring that preferences have a short "half-life"11 preferences. The analysis here makes clear that the and that inefficient, high-cost industries with impact of reducing tariffs will vary substantially entrenched lobbies do not constrain flexibility and across countries but that most countries will not adjustment. Multilateral liberalization is also lose because they currently gain very little from important for limiting the long-term trade divert- preferences.10 For the countries that receive sub- ing impact of preferences on other countries (typi- stantial transfers from preferences, the commodity cally these will be other developing countries). impact of tariff liberalization is crucial. For exam- In practice, only a small number of countries ple, significant reductions in EU tariffs and internal receive large transfers as a result of preferences in prices for sugar would have a significant impact on OECD markets. The values of preferences are a number of countries, especially if existing quotas largest in the EU market, driven by a narrow range were maintained. of products and the very high EU price for sugar. In The impact of reducing tariffs on products a few countries, such as Mauritius, preferences excluded from preferences will tend to be positive appear to have contributed to a relatively strong for the least-developed countries. These products economic performance and economic diversifica- have very high tariffs, and a reduction in protection tion (Subramanian 2001). In other countries, even would stimulate exports from countries with a though preferences have led to large transfers, comparative advantage in these products. Whether domestic industries have experienced rising costs the least-developed countries would gain more and declining output and have accumulated large from the inclusion of these very-high-duty prod- debts.12 Nevertheless, the majority of beneficiaries ucts under preferences and the continuation of of U.S., EU, and Japanese preferences have experi- high levels of protection is difficult to assess, but enced little or no impact. Preferences have done the uncertain duration of many nonreciprocal nothing to stimulate the export of a broader range schemes and the difficulties of satisfying rules of of products. origin are likely to limit the value of preferences on The key issues for improving trade preference these products. schemes are as follows: Wainio and Gibson (2003) estimate that, as a group, countries receiving nonreciprocal trade · How to enhance the value of preferences under preferences on agricultural products in the United current export structures, which would be facili- States would gain from multilateral trade liberaliza- tated by tion because losses from preference erosion would ­Extending coverage to all agricultural products. be exceeded by gains on products on which these ­Liberalizing the rules of origin and simplifying countries pay the MFN tariff. Within this group, a the process of certifying compliance. country will tend to lose on balance from multilat- ­Removing sources of uncertainty concerning eral liberalization only if more than 80 percent of product and country coverage and the its exports to the United States receive preferences, duration of preference schemes. while it would tend to gain if less than 50 percent of · How to strengthen the impact of existing prefer- its exports benefit from preferences. For countries ences on developing countries, which would be in between, whether there was a net gain or a loss facilitated by from reducing MFN tariffs would depend on the ­Improving the domestic investment particular tariff-cutting formula and the structure environment. of exports. ­Addressing the internal barriers that raise the costs of trade for developing countries-- inadequate and high-price transport services, Conclusion reflecting lack of infrastructure and lack of In principle, trade preferences can assist develop- effective competition in many countries, ineffi- ment if they provide temporary margins of prefer- cient and corrupt customs practices, and lack ence to enable industries to adjust and compete of trade-supporting financial and telecommu- more effectively in global markets. Multilateral nications services.13 72 Global Agricultural Trade and Developing Countries · How to ensure that preferences do not interfere are not granted the preferential access for which they are in prin- with multilateral liberalization, which would be ciple eligible) cannot reflect the inability to meet other require- ments to access the relevant market, such as health and safety or facilitated by sanitary requirements or deficiencies in their infrastructure, as is ­Developing mechanisms for helping countries sometimes suggested. Lack of infrastructure might explain a that incur significant losses from preference muted response from trade to preferences but cannot explain why, at the border, some products that are eligible for prefer- erosion adjust. ences do not request those preferences. ­Not using fear of preference erosion to main- 4. For a comprehensive description of U.S. preferences for tain high levels of protection in industrial agricultural products, see Wainio and Gibson (2003) 5. Many of the tariff quota products are also subject to safe- countries. guard measures. Once quantities exceed the quota, exports to the United States are subject to both the high MFN duty and an The challenge is to find preference schemes that additional, often high, safeguard duty. complement the domestic reforms that developing 6. A number of lines shown as AGOA products are likely to be economically meaningless. These are lines that refer to Gen- countries must undertake to improve the returns to eral Note 15 of the U.S. tariff schedule, which excludes from the exports without stifling diversification and multi- in-quota quantity for a product subject to a tariff rate quota and lateral trade liberalization. Trade preferences are to safeguard amounts that are imported by the U.S. government, not a panacea for success but rather should be seen by individuals in quantities of less than five kilograms, and sam- ples for exhibition or for display at trade fairs. If such products as just one part of a strategy to boost export-led are imported from AGOA countries, they are eligible for zero growth and development. Realizing the full poten- duty access. In the 2002 tariff schedule, 85 agricultural lines des- tial of trade also requires improving customs clear- ignated as AGOA products referred to General Note 15, or 14 percent of AGOA-designated agricultural tariff lines. In 2002 ance procedures, reducing the costs of transporta- imports from AGOA countries were recorded in only one of tion and other trade-related services, ending these categories, and the amount was negligible. For a more corruption, and removing other disincentives to accurate representation of the impact of AGOA, these lines are excluded from the analysis. investment. Addressing these issues will permit 7. This clause was initially discussed in the context of com- broad-based export growth and will ensure that as bating fraud. However, this is not made clear in the legislation, preferences decline with multilateral liberalization, and it appears that the clause could be invoked in more general the economic structure needed for continued circumstances. 8. The value of preferences will also be overstated for prod- export expansion is in place. The Integrated Frame- ucts for which there are no nonpreferential imports and for work for Trade-Related Technical Assistance, when which the duty exceeds the prohibitive level--the gap between incorporated into poverty reduction strategies, the internal price in the importing country and the world price. 9. The analysis was undertaken at the 5-digit level of the provides a vehicle for addressing these issues, defin- Standard International Trade Classification (SITC). There are ing appropriate policy responses, and mobilizing around 250 agricultural products in this classification. The 34 relevant resources. countries include 3 that did not export any agricultural products to the United States in 2002. 10. Stevens (2003) notes that the flip side of the preference Notes coin is agricultural protectionism in OECD countries, which has led to cheaper imports for African countries of a number of 1. By providing a stimulus to increased trade, preferences agricultural products, such as cereals. There are two impacts: can lead to lower transportation costs, which in turn lead to a preferences increase export receipts to pay for imports, and further trade impact. Hummels and Skiba (2002) discuss how OECD protectionism reduces the prices of those imports. Even economies of scale in transportation can lead to a virtuous circle countries that gain little from preferences may lose from multi- involving increased trade and lower transportation costs. lateral trade reform. 2. For example, the share of Sub-Saharan African countries 11. Taken from Schott (2004), who presented the notion in in U.S. imports of agricultural products fell from 4.3 percent in terms of free trade agreements. 1982 to 2.5 percent in 2002. Similarly, the share of the African, 12. Mitchell (2004) shows that the sugar industry in the Caribbean, and Pacific countries in total EU agricultural Caribbean is dominated by high-cost producers, few of which imports fell from 11.7 percent in 1982 to 7.8 percent in 2002. can profitably export to the European Union, even at four times And the share of low-income countries in Japanese imports fell world prices. Sugar production has been declining, and efforts to from 1.2 percent to 0.5 percent over the same period. More sys- diversify away from sugar have generally been unsuccessful. A tematic empirical studies of the impact of trade preferences serious problem in a number of countries is the high level of are rare and seldom separate out the impact on agricultural accumulated debt of the state-owned sugar industries, which products. can amount to a substantial proportion of gross domestic 3. The rate of utilization of preferences is the proportion of product. exports from developing countries to the European Union, the 13. These issues are highlighted in diagnostic trade studies United States, and Japan that are recorded at the border as undertaken in the context of the Integrated Framework for requesting preferences. Therefore, the underutilization of pref- Trade-Related Technical Assistance for the least-developed erences (the fact that some exports do not request and therefore countries (see www.integratedframework.org). The Impact of Agricultural Trade Preferences, with Particular Attention to the Least-Developed Countries 73 References Schott, J. 2004. "Free Trade Agreements: Boon or Bane of the World Trading System?" In J. Schott, ed., Free Trade Agree- Brenton, P. 2003. "Integrating the Least Developed Economies ments: US Strategies and Priorities. Washington, D.C.: Insti- into the World Trading System: The Current Impact of EU tute for International Economics. Preferences under Everything but Arms." Journal of World Stevens, C. 2003. "Agricultural Reform and Erosion of Prefer- Trade 37: 623­46. ences." Institute of Development Studies, Brighton, U.K. Brenton, P., and H. Imagawa. 2004. "Rules of Origin, Trade and Subramanian, A. 2001. "Mauritius: A Case Study." Finance and Customs." In J. Sokol and L. de Wulf, eds., The Customs Mod- Development 38(4): 22­25. ernisation Handbook. Washington, D.C.: World Bank. Topp, V. 2001. "Trade Preferences: Are They Helpful in Advanc- Hummels, D., and Alexandre Skiba. 2002. "A Virtuous Circle? ing Economic Development in Poor Countries?" Australian Regional Tariff Liberalization and Scale Economies in Trans- Bureau of Agricultural and Resource Economics, Canberra. port." Purdue University, West Lafayette, Ind., www.mgmt UNCTAD. 2001. "Improving Market Access for LDCs." .purdue.edu/faculty/hummelsd. UNCTAD/DITC/TNCD/4, http://www.unctad.org/en/docs// McQueen, M. 1999. "After Lome IV: ACP-EU Trade Preferences poditctncd4.en.pdf. in the 21st Century." Intereconomics 34: 223­32. Wainio, J., and P. Gibson. 2003. "The Significance of Nonrecip- Mitchell, D. 2004. "Sugar in the Caribbean: Policies and Diversi- rocal Trade Preferences for Developing Countries." Paper fication Strategies to Cope with Declining Preferences." presented at the International Conference "Agricultural Pol- World Bank, Washington, D.C. icy Reform and the WTO: Where Are We Heading?" June Ozden, C., and M. Olarreaga. 2003. "AGOA and Apparel: Who 23­26, Capri, Italy. Captures the Tariff Rent in the Presence of Preferential Mar- ket Access?" World Bank, Washington, D.C. 5 EXPERIENCE WITH DECOUPLING AGRICULTUR AL SUPPORT John Baffes and Harry de Gorter For most of the past half century industrial coun- deem necessary. Perhaps the only effective way to tries have had high levels of agricultural protection, achieve socially acceptable and politically feasible provided by import tariffs, quantitative restric- reform is to decouple payments from current pro- tions, and domestic subsidies. Among the many duction levels, input use, and prices. Thus, the rele- claimed objectives of these policies, boosting the vant question is how support can be given without income of small family farms is by far the most fre- creating negative effects for the rest of the world-- quently cited (Winters 1989­90). Because most of how to increase farmers' incomes without distort- this support is based on current output, input use, ing production and consumption. and prices, it also induces overproduction. Given This chapter analyzes the experience with decou- the weight of industrial countries in the global pling, making a clear distinction between decou- trading system, the aggregate effect of such support pling that replaces domestic support and is to depress world commodity prices, reducing the decoupling that replaces border support. It reviews export shares of countries that do not protect their a number of one-time buyouts, the best form of agricultural sectors. Such support is costly and decoupling, and looks at the externalities of decou- often goes to unintended recipients, thus exacer- pling, especially for middle- and low-income coun- bating rather than eliminating the presumed tries, in reducing poverty, instituting land title income inequalities that justified support in the reform, and providing credit. first place. Considering the harmful effects of such support What Is Decoupling? on world markets and the mismatch between stated objectives and ultimate outcomes, its outright Decoupling has different meanings to economists, elimination is sometimes advocated. But societies policymakers, and trade negotiators. Some see it as a have the right to transfer income to groups as they transition mechanism to a fully competitive sector. The authors would like to thank Ataman Aksoy, John Beghin, Uri Dadush, Gaston Gohou, Bernard Hoekman, John Nash, and Marcelo Olarreaga for helpful comments on earlier drafts. Comments received from the participants at the 2003 Annual World Bank Conference on Development Economics­Europe and several World Bank seminars are greatly appreciated. 75 76 Global Agricultural Trade and Developing Countries Others see it as another support program, with purpose would not be to support income at arti- fewer production- and trade-distorting effects. ficially high levels but to prevent a severe tempo- Some use decoupling only to refer to programs for rary decline in individual income; (3) the right transferring income to producers; others use the to benefits would attach to the person, not to term much more broadly, to include, for example, farm land or the farm enterprise, and would programs to improve the environment. Sometimes accordingly not be transferable; (4) the benefit decoupling is assessed according to the policy's to be enjoyed by any individual would not long-run impact on output through such factors as exceed a modest maximum; (5) benefits would uncertainty, investment, and expectations. not be conditioned upon the production of par- Decoupling was discussed in the literature as ticular commodities or even upon continued early as 1945, when the American Farm Economic employment in agriculture. . . . Association announced 18 awards for papers on "a price policy for agriculture, consistent with eco- Another early decoupling proposal was put for- nomic progress, that will promote adequate and ward by Nash in Europe (1961, p. 188): more stable income from farming." Nicholls and Johnson (1946)--recipients of the first- and second- Instead of obstructing the withdrawal of farmers place awards--summarized the main findings of the from an industry which cannot adequately award-wining papers. Several recommendations reward them, . . . an unconditional payment closely resembled decoupled support. For example [could be made] to all those at present engaged (p. 281), in farming, or to those of them deemed to be in need of compensation, calculated by the refer- Cochrane presents a special formula for pro- ence to the difference between the incomes now gressively smaller income payments for aiding earned under the protective system and those producers in adjusting their operations from a capable of being earned under a system of free support level to a free market situation. These market prices. An annuity calculated in this way declining payments would be based on produc- and payable for life to all engaged in farming, tion during 1939­41, so that the producer would but not transferable to their successors, would, not be "tied to commodity in surplus to receive a in theory at least, make it possible to bring the payment benefit." Thus, he could shift to some protective system to an end while fully making other product during the payment period with- good the loss of income to its present beneficiar- out losing the specified payments. ies. There is no doubt that compensation of this kind is feasible. Perhaps the first analyst to explicitly advocate decoupled support in U.S. agriculture was Swerling The proceedings of the workshop "Decoupling: (1959). Two characteristics of Swerling's proposal The Concept and Its Future in Canada" contains are especially interesting. First, he advocated a numerous definitions of decoupling (Finkle and safety net mechanism for agriculture, similar to Cameron 1990). Consider the following two rather safety nets in other sectors of the economy (such as contrasting views. Van Donkersgoed (1988, p. 40), unemployment insurance). Second, he proposed of the Christian Farmers Foundation of Ontario, linking the benefits of the decoupled support to defined decoupling as "a program in which eligibil- income declared in tax returns during the recent ity is not linked to production, the production past (not to historical production or area). Specifi- potential of resources or the production effort of a cally (pp. 179­80), he wrote: farm entrepreneur; rather eligibility is linked to stewardship farming practices, marketing, the Removal of this price stimulus is long over- maintenance of rural communities, diversified due. . . . An income-insurance plan for farm- ownership of the assets of production, moderate- operators [should be in place] that include[s] sized family enterprises and other rural, non- the following elements: (1) . . . benefits will production valuables that add to the quality of be related to income experience of the particular Canadian life." Spriggs and Sigurdson (1988, p. 93), individual during the recent past; (2) the in contrast, simply stated: "In fact, a program to Experience with Decoupling Agricultural Support 77 eliminate subsidies would be the ultimate in decou- Trade Organization (WTO) policy in being mini- pling. It is the only truly decoupled program that mally trade distorting (as Swerling originally sug- there is." gested in 1959). It is very difficult to end farm sub- Cahill (1997, p. 351) defines as fully decoupled sidy programs, however; there is always a bias to from production a policy that "does not influence maintain current policies because politicians lose production decisions of farmers receiving pay- more support if they take away subsidies than they ments, and that permits free market determination gain if they introduce new ones. Furthermore, of prices (facing all farmers, whether or not receiv- governments like to concentrate the benefits of ing income support)." A policy is effectively fully subsidies and diffuse the costs to as many people as decoupled if "the provision of the compensatory possible in order to maximize political support. payment package results in production that, for any And small groups are better able to organize and crop, does not exceed that level that would exist control free riders. All this appears to make it without compensation." The Organisation for Eco- inevitable that governments will favor commodity- nomic Co-operation and Development (OECD or sector-based policies over all other forms of 2000a) defines decoupling in a similar way. agricultural support. So, fully decoupled payments Hennessy (1998) includes as decoupling pay- and one-time buyouts, even more than universal ments triggered by ex post market or production programs, have the political disadvantage of not conditions, as long as the payment level is not con- being able to continue to favor incumbent farmers. ditioned on an individual's specific level of produc- They also look like corporate welfare, whereas tion. Disaster relief measures, for example, would trade barriers and price supports reduce visible be considered decoupled because they are not taxpayer costs and hide the fact that large farms get affected by the individual's level of production. most of the transfers. Politicians also lack the Goodwin and Mishra (2002) argue that a fully commitment mechanism to keep such policies in decoupled payment must be fixed and guaranteed place--politicians are tempted to reintroduce sup- and thus is not influenced by ex post realizations of port later in its original form or with new distort- market conditions (such as low prices or area ing programs. yields). This is the narrowest definition because Under many current systems, a complex web of neither payments nor the rules of eligibility and the policies, including payments not to produce, subsi- base criteria can be changed. If a time limit is added dies, and production controls, help to obfuscate the to this definition, then decoupling simply implies a policies' true nature in terms of who benefits and to number of annual payments to producers. Where what extent. Another class of subsidy programs, the financial markets function efficiently, these bonds whole farm insurance program used in Canada and can be converted into a single payment. In such a the revenue stabilization programs used in the past, setting, decoupling would consist of an administra- has the economic advantage of not singling out tive decision to remove distortions followed by a specific sectors. And because all farms are eligible, single payment--a radical policy initiative. In fact, taxpayer constraints dilute the per farm benefits a number of analysts have advocated a fundamental thereby reducing the political support for such pro- reform of the European Union's Common Agricul- grams. Thus Canada has eliminated perfectly func- tural Policy (CAP), with the last step consisting of tioning revenue insurance programs, and other payment of a bond (see, for example, Beard and countries are not rushing to adopt wholesale farm Swinbank 2001, Swinbank and Tangermann 2001, income insurance programs. and Tangermann 1991). Politicians and farm lobbies capitalize on the fact that most voters know of at least one farmer, often a family member, who experienced severe adjust- The Politics of Decoupling ments in the past 50 years. Thus it is much easier to Politicians are reluctant to subsume all agricultural maintain the status quo of subsidies in agriculture. policies into a universal social welfare program, Politicians also play on insecurities related to food including job retraining and the like, even though self-sufficiency in case of war, food safety issues these types of programs to help small farmers arising because of new technologies and genetically would be ideal and truly compatible with World modified organisms, and the multifunctional 78 Global Agricultural Trade and Developing Countries benefits of farms in providing landscape amenities from decoupling may be illusory. The issue here and rural livelihoods. is whether the costs of distortions in commodity markets are necessarily greater than the costs of distortions introduced elsewhere in the econ- The Economics of Decoupling omy to finance "decoupled" transfers. Decoupling can be viewed as two distinct transition mechanisms: one replacing domestic support and Moschini and Sckokai (1994) claim that the wel- one replacing border measures. The key variable fare losses of raising new taxes to finance decou- driving this distinction is the source of financing pling are unlikely to be larger than the welfare gains for the original support measures: consumers, tax- from decoupling. Beghin, Bureau, and Park (2003) payers, or a combination. estimate that in the Republic of Korea it costs Replacing domestic support measures such as taxpayers $1.61 for every $1 transferred to produc- production subsidies with decoupled support is ers. Using a general equilibrium model, Parry straightforward in the small country case and can (1999) finds that the efficiency cost of taxpayer- be shown to be Pareto improving. Instead of pro- financed lump-sum transfers to agriculture equals viding output-based subsidies, the government 27 percent of the amount of the income transfer. makes lump-sum payments to producers based on Since most of the support is at the border, some historical criteria without any constraint or decoupling is likely to be a complicated exercise requirement on the current use of their resources. with mixed outcomes. Although the costs of Under the lump-sum scheme, producers can taxpayer-financed programs are shown to be signif- receive higher payments because welfare losses (the icant, welfare gains depend on how decoupled pro- so-called Harberger triangles) disappear. Taxpayers grams are financed. But the general result from the can also be better off if part of the efficiency gains is public finance literature is that trade taxes have translated into lower taxes. Because both producers much higher inefficiencies relative to other forms and taxpayers can be made better off, decoupling in of taxation or sources of revenue for farmers. the production subsidy case is clearly a Pareto- improving move. Experience with Broad Decoupling in the case of an import tariff, how- Decoupling Attempts ever, is more complicated as it involves eliminating tariffs, raising additional taxes, and distributing the Early attempts at decoupling failed. The 1949 tax revenues to producers. Producers are no worse Brannan Plan in the United States, which proposed off (they receive the same amount of support), con- cash payments to farmers whose overall income fell sumers are better off (they pay lower prices), but below a certain level, was defeated in the U.S. taxpayers are worse off because they lose the tariff Congress. Similarly in Europe, the Mansholt Plan revenue and must finance the decoupled support. of 1968, which advocated support in order to Assuming that welfare losses arising from border finance mandatory retirement for older farmers, measures are higher than welfare losses arising also failed. from domestic subsidies, decoupling of border The first attempt at decoupling came in the measures is welfare improving. It is not a Pareto United States with the 1985 Farm Bill, which improvement, however. Furthermore, while the shifted the base of support from current yields to removal of the import tariff implies welfare gains, historical yields (see timeline in table 5.1). The introduction of the tax to finance decoupled sup- European Union (EU) partially replaced interven- port implies welfare losses. Alston and Hurd (1990, tion prices with direct payments following the p. 155) contend the following: Common Agricultural Policy reform of 1992. Mexico replaced price supports with direct pay- Currently it is fashionable to argue for "decou- ments in 1994 with the introduction of the pling" farm programs in the sense that income National Program for Direct Assistance to Rural transfers should be achieved with minimal Areas (Programa de Apoyos Directos al Campo, or consequences for commodity markets. Along Procampo). The United States replaced deficiency with the benefits from transparency, the benefits payments with decoupled support in the 1996 Experience with Decoupling Agricultural Support 79 TABLE 5.1 Chronology of Broader Decoupling and Recoupling Episodes, 1985­2004 Year Country Policy change 1985 United States 1985 Farm Bill introduces "frozen" government payment yields per acre. 1992 European Union Mac Sharry reforms of the Common Agricultural Policy reduce price supports and introduce direct payments linked to historical area planted (with "frozen" government payments for the output per hectare) or number of animals (but farmers still need to produce to receive payments). 1994 Mexico Procampo introduces payments based both on historical acres and yields up to 2008 with a phase-out of import barriers under the North American Free Trade Agreement, input subsidies, and activities of the state trading monopoly. 1996 United States 1996 Farm Bill eliminates target prices, replacing them with decoupled historical entitlements, the so-called production flexibility contract payments, to end in 2002. 1996 Mexico Base acres can be switched to other crops or enterprises, and rural development policy is launched to foster productivity. 1998 United States Emergency market loss assistance payments effectively reverse the 1996 Farm Bill. 2000 European Union Agenda 2000 extends, deepens, and widens the Mac Sharry reforms. 2001 Turkey Direct income support program reduces some administered prices and input subsidies. Only minor changes in border policies. 2002 United States 2002 Farm Bill extends production flexibility contracts, formalizes emergency payments as countercyclical payments, adds new crops to production flexibility contracts program, allows base acres and payment yields to be updated, increases price supports for coupled subsidies, and introduces three new crops to the coupled subsidy program. 2002 Mexico Target prices and input subsidies are reintroduced. Procampo remains largely unchanged. 2002 European Union Mid-term review, resulting in June 2003 agreement to switch most direct payments to decoupled payments, with entitlements sold with or without land; level of payments and some support prices to decline in 2005­07. 2004 European Union Decoupled payments are introduced for the so-called Mediterranean products (cotton, olive oil, and tobacco). Source: Authors' compilations. Farm Bill. More recently, Turkey replaced some through farm bills. There have been 20 such bills price support and input subsidies with direct pay- since the first one in 1929. The central feature of the ments. In addition to broad decoupling programs, New Deal farm programs of the 1930s was price there have been numerous one-time buyouts, supports achieved through taxpayer-funded pro- including New Zealand's exit grant in 1984, the duction subsidies and supply controls (acreage set- buyout of Canada's grain transportation subsidy in asides, accumulation, maintenance, and disposal of 1995, and the buyout of the U.S. peanut marketing public stocks). Payments were based on the differ- quota under the 2002 Farm Bill. ence between the target price set by the government and the higher of the market price or the price at which the government would value crops used as Decoupling Efforts in the United States collateral for loans made by a public corporation. The budgetary outlays for most U.S. commodity The total payment was equal to the yield per acre programs are authorized by Congress (and subse- multiplied by a farm's eligible payment acreage (the quently approved by the president) every few years amount of land devoted to cultivation of the crop 80 Global Agricultural Trade and Developing Countries TABLE 5.2 Composition of Agricultural Support in the United States, 1986­88 to 1999­2001 (US$ millions) Category 1986­88 1989­92 1993­95 1996­98 1999­2001 Value of production 113,537 168,615 184,239 199,990 192,417 Total support estimate 68,540 72,779 79,060 81,715 95,455 Producer support estimate 41,839 34,326 31,091 36,384 51,256 Market price support 19,533 17,825 16,969 17,864 18,662 Budgetary support 22,306 16,501 14,123 18,519 32,594 Output 2,919 510 241 1,644 9,285 Input use 6,516 6,574 6,003 6,088 6,877 Area 11,313 6,897 5,396 1,247 2,722 Historical entitlements 0 0 0 6,647 10,085 Input constraints 637 1,776 1,963 1,940 1,844 Overall farm income 912 743 520 954 1,780 Source: OECD database. in question). This portfolio of policy instruments The effect of the 1996 Farm Bill on the structure was the primary means of price support for the of budgetary outlays is shown in table 5.2. It breaks major field crops for decades until the 1980s. the producer support estimate down into market The Food Security Act of 1985 set a new trend price support (a measure of border protection) and for major field crops by reducing the role of acreage budgetary support (a measure of domestic sup- set-asides and public stockholding and moving port). Budgetary support is further decomposed toward decoupling, with a "freeze" on payment into support based on output and input use (con- yields (farmers were paid on the basis of fixed out- sidered as having a large impact on production and put per acre regardless of what was actually pro- trade, or fully coupled support) and support based duced). Payment yield was established for each on area, historical entitlements, input constraints, farm by the Department of Agriculture, based on and overall farm income (considered as having a average yields in 1981­85. smaller impact on production and trade, or par- Acreage set-asides and public stockholding were tially decoupled support; for further details and largely abandoned by the mid-1990s and elimi- definitions, see OECD 2000b). nated soon thereafter with the introduction of the Historical entitlements, which did not exist Federal Agricultural Improvement and Reform before 1996, represented more than a third of total (FAIR) Act in 1996. FAIR also banished the target budgetary support in 1996­98. They are exempt price used in calculating deficiency payments but from disciplines in the WTO (they are in the Green maintained the lower fixed price, called the loan Box; see chapter 3). Area payments declined from rate, which had triggered public stock purchases in $5.4 billion in 1993­95 to $1.2 billion in 1996­98 the past. In place of the links between support, and are also exempt from reduction commitments prices, and production, production flexibility con- in the WTO (they are in the Blue Box). During tract payments were introduced. Participating pro- these two periods, output payments under disci- ducers received payments in proportion to what pline in the WTO (in the Amber Box) also they had received during 1990­95 or would have increased, from $0.2 billion to $1.6 billion, a reflec- received had they been enrolled. These historical tion primarily of declining commodity prices and benefits were in turn determined by a farmer's his- consequently increased loan rate payments. torical production levels. Each participating pro- Although payments were made on a crop-by- ducer received a fixed schedule of payments, which crop basis, planting was not required or restricted was to decline gradually through 2002. Although to any particular crop. Payments were tied to not specifically stated, it was implicitly assumed 85 percent of the fixed-base area (average of acres that the payments would end by 2002. planted or prevented from being planted for Experience with Decoupling Agricultural Support 81 covered crops of wheat, feed grains, rice, and cot- Following the Spaak Report of 1956, which sug- ton) and fixed-payment yields. Because the pay- gested that agriculture requires special treatment, ments were independent of current production, the Stresa Conference of 1958 outlined CAP's three farmers had far greater flexibility to make planting guiding principles: free flow of agricultural com- decisions (or to not plant at all). Farmers were free modities within the common market, preference to allocate their land to any crops on the "contract to member states, and common financing. CAP, acres" except fruits and vegetables, but they had to formally put in place in 1962, had multiple objec- maintain their land in "agricultural use." Thus pro- tives: increase agricultural production, ensure a fair ducers were to depend more heavily on the market standard of living for the agricultural community, and also bear greater risk from increased price stabilize markets, guarantee a regular supply of variability. agricultural commodities, and ensure reasonable The FAIR Act was meant to be a transition prices for consumers. The objectives were to be toward a new policy environment with a dimin- achieved through domestic price supports, export ished government role in commodity markets. subsidies, and common trade barriers. The first and Commodity prices declined sharply in the late last objectives were fully met within a few years, but 1990s, however, triggering three major policy concerns were soon raised about excess production events that reversed much of what had been and the unsustainable level of CAP budgetary accomplished by the FAIR Act. First, emergency requirements if policies did not change. payments were introduced, approximately equal to Reform of the CAP was attempted in 1972, fol- 50 percent of decoupled payments in 1998 and 100 lowing the recommendations of the 1968 Mansholt percent of decoupled payments in 1999, 2000, and Plan. The plan proposed, among other reforms, 2001. These were designated as non-product- lump-sum transfers to 5 million farmers to retire specific support and so escaped reduction under them from farming and reduce active farmland by the de minimis proviso of the WTO. Second, when 5 percent. The Mansholt Plan, the first attempt to market prices fell below the loan rate, the govern- decouple, was never implemented. ment extended the marketing loan program by The first major reinstrumentation of the CAP issuing loan deficiency payments, which had the took place in 1992. The reform, known as the Mac same economic effects as the previous deficiency Sharry reform after the EU's Commissioner for payment scheme. Third, the 2002 Farm Bill was Agriculture, together with the Blair House Accord introduced, increasing several loan rates, introduc- of the United States, paved the way for the signing ing three more crops into the loan rate scheme, and of the Uruguay Round Agreement on Agriculture allowing base acres and payment yields to be in 1994. For cereal, oilseed, and protein crops updated and soybean acreage to be added to the and for beef and veal, price supports provided by base. The bill formalized the emergency payments import levies or export refunds were reduced, and into a new countercyclical scheme in which pay- farmers were compensated with direct payments. ments vary with price but not with quantity. For crops, payments were based on 85 percent of The emergency measures introduced in 1998 historical plantings (with a paid minimum area (and later the 2002 Farm Bill) changed the structure set-aside requirement, a further paid voluntary of the budgetary outlays considerably. Between set-aside of up to 30 percent of historical area, 1996­98 and 1999­2001, historical entitlements and a base acre limit for payments set at the increased by more than 50 percent (from $6.6 bil- national or regional level). The area-payment rates lion to $10.1 billion, area payments increased varied by crop type, and the set-aside payments twofold, and payments based on output increased were initially higher but are now equal. The only more than fivefold (see table 5.2), implying that requirement is the land had to be set aside or support is less decoupled now than it was after 1996. planted in crops or temporary grass. Small-scale farmers producing less than 92 tons of cereals annually are exempt from set-asides and receive Decoupling Efforts in the European Union "all cereals" payments irrespective of crop planted The principal vehicle of support in the European (representing 25 percent of area but 70 percent of Union has been the Common Agricultural Policy. farmers). 82 Global Agricultural Trade and Developing Countries TABLE 5.3 Composition of Agricultural Support in the European Union, 1986­88 to 1999­2001 (US$ millions) Category 1986­88 1989­92 1993­95 1996­98 1999­2001 Value of production 214,849 275,770 286,658 291,427 237,990 Total support estimate 109,654 138,927 133,050 129,328 112,628 Producer support estimate 93,719 117,097 116,519 111,966 99,343 Market price support 80,257 93,282 76,084 64,989 60,863 Budgetary support 13,446 23,327 40,279 47,468 38,693 Output 5,009 6,769 2,999 3,945 3,644 Input use 5,025 7,135 8,133 8,446 6,540 Area 2,701 6,987 24,326 29,419 24,733 Historical entitlements 0 559 1,466 1,007 597 Input constraints 711 1,877 3,356 4,650 3,178 Overall farm income 0 0 0 2 0 Source: OECD database. Between 1986­88 and 1993­95, budgetary sup- importance of rural development seems to follow port in the EU increased threefold, from $13.4 bil- from the official reference to it as the "second pillar lion to $40.3 billion, while border support declined of the CAP." from $80 billion to $76 billion (table 5.3). Most of The European Union now has greater flexibility the increase in budgetary support was attributable to overhaul any policy element in light of changes to area payments and, to a more limited extent, to in market developments, costs, enlargement, WTO historical entitlements and input constraints (des- (and other) trade negotiations, food crises, and ignated Blue Box payments and so exempt from other pressure for reform. The budget for Agenda reductions in the WTO). 2000 did not include any provision for extending Following the 1992 reforms, the level of support direct payments to farmers in Eastern Europe, remained unchanged, but its structure changed making reform a requirement. Meanwhile, the considerably. For example, while estimated pro- European Union has launched free trade negotia- ducer support averaged $117 billion for 1989­92 tions with Mercosur, and it established the and 1993­95, border protection support declined Everything but Arms initiative with low-income from $93 billion in 1989­92 to $76 billion in developing countries. Because Mercosur includes 1993­95. Support based on output declined from some major agricultural exporting nations and the $7 billion to $3 billion, and area payments Everything but Arms program will increase increased from $7 billion to $24 billion. Thus imports, especially for sugar, rice, and bananas, fur- the 1992 CAP reform was a good step toward ther reform of the CAP is necessary. decoupling. Recent food crises underline the need for Under Agenda 2000, price support to crops reform, sometimes for more regulation and con- declined, direct payments increased and were trols over production practices, including animal realigned across all crops, and reference yields were welfare. Against this background, the European changed in some countries. A push toward more Commission's midterm review of Agenda 2000 investment in rural development was also made. A proposed a set of reforms that include further large transformation has occurred away from bor- decoupling, continuing set-asides, and more cross- der protection and input subsidies to direct pay- compliance rules with statutory environmental, ments. Total support has been declining, especially food safety, and animal health and welfare in grains and oilseeds. More than the increase in standards. budgetary allocations, which remains moderate Current EU compensatory payments still influ- compared with other expenditures, the growing ence farmers' decisions on how much land to plant. Experience with Decoupling Agricultural Support 83 This results not only because farmers are obligated decoupling within the limits for each sector. They to produce cereals on the base acres to receive the may also give up to 10 percent of the payments for payments, but also because area payments in the environmentally friendly farming and restrict enti- European Union are made on an aggregate, fixed- tlement trading within a region. All payments are area basis that is set at the national or regional level. to be reduced 3 percent in 2005, 4 percent in 2006, Individual farmers do not have a base area--just and 5 percent in 2007. Support prices will also eligible acres for which they receive payments and decline. Payments will be conditional on compli- have area set-asides. If the regional base area is ance with various measures, including environ- exceeded, the per-unit subsidy is prorated down- mental and acreage set-asides. ward proportionately for all farmers in the region. Because the prorating occurs on the total area Decoupling Efforts in Mexico planted ex post, farmers have an incentive to over- plant in order to maximize their share of fixed About a quarter of Mexico's population depends on budget outlays or to defend against share erosion agriculture, which contributes 5 percent to gross due to overplanting by other producers. This domestic product (GDP), down from 9 percent in means that the area payments are fully coupled to the early 1980s. According to the OECD, total plantings because individual farmers are not penal- transfers to agriculture averaged $7 billion annually ized for their own decisions to overplant. Area pay- during 1999­2001, $5.7 billion of which went to ments with a national base area are therefore not a producer support. This support corresponds to limit on total acres planted. $1,000 per full-time farmer equivalent and $53 per For EU cattle, the headage payments under hectare, both considerably lower than the OECD "production-limiting" arrangements are anything averages of $11,000 per farmer and $192 per but production limiting because farmers are hectare. About 29 percent of producer support allowed to keep more cattle than are eligible for went to maize, 21 percent to milk, and 13 percent to payments, so there is no absolute production con- sugar. trol, and the number of eligible animals is not lim- Traditionally, Mexico's state agricultural enter- ited to the number on farms prior to the introduc- prise, Conasupo (Compania Nacional de Subsis- tion of payments in 1992. Where numbers of tencias Populares), has been heavily involved in the animals are below the maximum that could be marketing, transportation, storage, and processing claimed per farm, farmers have an incentive to of most agricultural commodities. Maize, beans, expand their stock up to the limits on which pay- and wheat, by far the most important agricultural ments are made. Thus incentives in the program commodities, have been heavily subsidized through have been to encourage initial expansion of animal a system of guaranteed prices. The government also numbers and then to lock production in at around set prices, which were usually announced before the levels that are consistent with the maximum planting decisions were made and were uniform number of animals eligible for payments. Those across the country and across seasons. Conasupo numbers reflect the very high levels of support for bought unlimited quantities at the guaranteed several decades as well as the incentives inherent in prices. Hence, producers knew in advance the price the headage payments. they would receive and shifted production to crops The CAP reform agreement of June 2003 with the highest degree of relative protection rather requires decoupling at least 75 percent of payments than with the highest profitability according to in the arable sector and at least 50 percent in the world prices. The poorest peasants did not benefit beef and sheep sectors. Dairy premiums will be from guaranteed prices since they formally mar- added into the single farm payment after 2007. The keted little or none of their production. decoupled single farm payment will be based on In 1994 Mexico introduced Procampo, a decou- average payments claimed over the three-year refer- pled support program to provide income support ence period, 2000­02, and will be paid per eligible to grain and oilseed producers--about 90 percent hectare of land. Entitlements can be sold with or of all Mexican farmers. Procampo replaced the old without land. Member states are offered some flex- scheme of guaranteed prices. By supporting farm- ibility in the year they begin and in fully or partially ers' incomes rather than production of specific 84 Global Agricultural Trade and Developing Countries commodities, Procampo was expected to make Few small farmers benefited from that system production and trade less distorted. It is also distri- because they were often net buyers, sold products at butionally more attractive than the earlier guaran- distress prices at harvest, or could not take advan- teed price program because poor subsistence farm- tage of price supports because they were not inte- ers are eligible for payments and there is a ceiling of grated with market price centers because of high 100 hectares on the amount of land that a single transaction costs. farmer can use to claim payments. Just as the United States did, however, Mexico Government credibility became a major issue reintroduced its price support in 2002. New coun- for Procampo. Initially, some producers did not tercyclical payments, similar to those that the believe that the government would actually imple- United States introduced in its 2002 Farm Bill, took ment the program. Fearing increased taxation, they effect with the 2002­03 marketing year. The pay- underreported land allocated to eligible commodi- ments were to equal the difference between the tar- ties. The government's turnaround, requiring that get price and the sum of the market price and land be allocated to eligible crops after initially Procampo payments. The payments would apply to delinking payments from the current use of land, eight commodities. In addition, a new common likely further discredited the government. (In 1996 subsidized price for electricity used for agricultural the government increased the number of eligible production was introduced (estimated to cost $0.6 crops.) The macroeconomic environment also billion annually.) played an important role. When Procampo was in The most visible change in Mexican agricul- the design phase, most commodities were highly tural policies has been the move from support protected, but the 1994 devaluation of the peso based on input use to support based on historical sharply reduced protection rates. entitlements, under Procampo (table 5.4). Border Despite these shortcomings the program has at measures are still the dominant component of least two features that improve income distribution support, accounting for 64 percent of producer (sometimes at the cost of more inefficiency). First, support during 1999­2001. decoupled area payments are given for a minimum Mexico's decoupled payment program encoun- of one hectare, even if the actual size of a farm is tered several problems. The program was less than one hectare. Second, land reforms allow announced well in advance of the registration of small farms to rent approximately 10 percent of eligible producers. The lag allowed many farmers to their land to larger farmers. These features can have increase the amount of land in production of the a significant positive impact on income distribu- eligible commodities and thus to increase their tion compared with historical guaranteed prices. future payments. So rather than moving resources TABLE 5.4 Composition of Agricultural Support in Mexico, 1986­88 to 1999­2001 (US$ millions) Category 1986­88 1989­92 1993­95 1996­98 1999­2001 Value of production 15,412 25,209 26,186 27,033 30,328 Total support estimate 1,287 8,121 7,558 4,858 6,999 Producer support estimate -266 5,718 5,060 3,190 5,694 Market price support -1,710 4,025 2,918 1,495 3,625 Budgetary support 1,444 1,692 2,142 1,695 2,068 Output 1 26 52 4 110 Input Use 1,442 1,663 1,308 676 721 Area 0 3 6 62 61 Historical entitlements 0 0 776 925 1,112 Input constraints 0 0 0 0 0 Overall farm income 0 0 0 27 63 Source: OECD database. Experience with Decoupling Agricultural Support 85 to more efficient uses, the scheme, initially at least, Responding to the high cost of support and its moved more resources into production that was distortionary effects, Turkey embarked on a major already inefficient. Moreover, because land rights agricultural policy reform program in 2001 with among landowners, tenants, and sharecroppers World Bank assistance (World Bank 2001). A were unclear, it was difficult to determine who was main component of the reform was to replace entitled to payment. administered prices and input subsidies with annual direct income support payments. In addi- tion, farmers were granted a one-time payment to Decoupling Efforts in Turkey cover the cost of transition from overproduced The agricultural sector in Turkey employs 43 per- and highly subsidized commodities to other cent of the labor force and contributes 16 percent commodities. to GDP, down from 26 percent in 1980. Total agri- Income support payments were set at $100 per cultural support in Turkey reached an annual aver- hectare, but even this low level of transfer implied age of $9.7 billion during 1999­2001, $6.5 billion of an eventual annual expenditure of $1.9 billion. The it in direct producer support, according to the upper limit, initially set at 20 hectares, was raised to OECD (table 5.5). Of that amount, $5.1 billion was 50 hectares in 2002. As in Mexico, to allow small transferred through border measures, the domi- subsistence farmers (who otherwise received no nant component of agricultural support in Turkey. support) to benefit from the program, a minimum At 5.1 percent of GDP, Turkey's agricultural sup- payment was set for farmers cultivating below a port rate is the highest of all OECD countries and certain threshold. almost four times the OECD average of 1.3 percent. A number of hard choices had to be made fol- This support corresponds to $162 per hectare, lowing the decision to implement direct income compared with the $192 per hectare average for support payments. A key decision related to records OECD. Sugar accounts for 13 percent of estimated (as was the case in Mexico). A pilot program was set producer support, milk for 11 percent, and wheat up in several districts in four provinces to test two for 10 percent. The main policy instruments for methods of developing a registry for producers. agricultural support have been border measures, One method, applied in two provinces, used the administered prices, input subsidies, and budgetary existing land registry records. A second method, payments. With a per capita GDP of a little over applied in the other two provinces, was based on $3,000, this support imposes considerable budget- certifications by the chief of the village, the council, ary strains on the economy. and the local farmers associations. Payments were TABLE 5.5 Composition of Agricultural Support in Turkey, 1986­88 to 1999­2001 (US$ millions) Category 1986­88 1989­92 1993­95 1996­98 1999­2001 Value of production 18,343 26,859 29,158 34,068 29,458 Total support estimate 3,092 7,212 6,027 10,705 9,649 Producer support estimate 2,779 6,127 4,675 7,791 6,522 Market price support 1,884 4,784 2,712 5,710 5,093 Budgetary support 895 1,344 1,962 2,081 1,429 Output 11 30 242 104 337 Input Use 885 1,314 1,720 1,978 957 Area 0 0 0 0 0 Historical entitlements 0 0 0 0 136 Input constraints 0 0 0 0 0 Overall farm income 0 0 0 0 0 Source: OECD database. 86 Global Agricultural Trade and Developing Countries made on a per hectare basis in two installments for both quantity and acreage allotments to be eligible up to two hectares. for payments. The acreage allotment was aban- While the pilot benefited 9,681 farmers includ- doned in 1982. Quantity quotas were tradable, with ing many small farmers, at a cost of $2.3 million, some exceptions. Imports were banned. numerous problems were encountered during The program again ran into trouble as the costs implementation. Land registries contained unclear of the program grew enormously. Peanut manufac- descriptions, shared titles did not specify the turers pressed for reforms because they wanted amount of land that each person owned, and many access to lower-priced peanuts, while the introduc- landowners who had inherited their land did not tion of the North American Free Trade Agreement possess deeds. Registration procedures were also (NAFTA) allowed peanut products to enter duty unclear, and various "producer certificates" were free from Mexico and Canada. issued without uniform standards. Many share- Some modifications were made in 1996, but the croppers were declared ineligible for participation biggest change came with the 2002 Farm Bill, with because they lacked official documents. There were the government deciding to buy out the marketing also cases of false claims, for nonfarm land or land quotas created in 1978. Eligible quota holders are to not in agricultural use. be compensated for the lost value of the marketing Other problems were related to the design and quota during fiscal years 2002­06. Quota holders implementation of the pilot. Farmers received can elect to receive payment in five equal install- inadequate information about the program, and ments of $0.11 a pound per year times the actual consequently many failed to apply for benefits quota allotment for the 2001 marketing year or to (especially in remote villages). The agencies receive the undiscounted sum of all the payments in involved in the pilot also received inadequate train- the first year, equal to $0.55 a pound. Given that an ing and information. And farmers were not given average effective quota for 1998­2000 is 5.6 million enough time to apply for the program. tons, the buyout is expected to cost $181 million a year, or $1.4 billion for the five-year period. During the same period, the annual value of U.S. peanut Experience with One-Time Buyouts production was $3.1 billion (8.79 million tons In addition to broad decoupling attempts, coun- times $355 per ton). In addition to the quota buy- tries have conducted numerous one-time buyouts out, peanut producers will be compensated by in the last two decades. These buyouts have been receiving support from the other provisions of the much more successful than the broader decoupling 2002 Farm Bill (decoupled and countercyclical pay- efforts. ments). Several factors led to this change in the existing peanut program: pressure from imports under NAFTA, opposition by other industry The 2002 U.S. Peanut Quota Buyout groups, and enormous increases in the fiscal costs The U.S. peanut program goes back to 1934, when of the program (see chapter 12 in this volume). peanut producers agreed to reduce their acreage in return for payments. The program failed to reduce Canada's Buyout of the Railway Subsidy ("Crow output and was revised in 1941 by introducing Rate") for Grain Shippers individual acreage allotments and penalties for farmers who exceeded the allotments. The allot- Canada's Crow Rate program (named for ments were not enforced, however. The Agricul- Crowsnest Pass in the Rocky Mountains) goes back tural Act of 1949 established support prices for to 1897, when Canadian Pacific Railway was given a peanuts, and until 1978 all peanuts from approved subsidy of $3.4 million to build a line between allotments were guaranteed the support price. The Alberta and British Columbia. In exchange for the program again ran into financial difficulties prima- subsidy, Canadian Pacific agreed to charge grain rily because of the introduction of high-yielding farmers 20 percent less than the (then) prevailing varieties. Beginning in 1978 peanut quotas were set rates. The 1925 Railway Act made the subsidized annually and producers received support for quota rates statutory. Over the years the Crow subsidies peanuts only. During 1979­82 farmers had to have were extended to numerous commodities. Because Experience with Decoupling Agricultural Support 87 of the higher prices received by western grain farm- annual income. Farmers with extremely low ers created by the transportation subsidies, value- incomes were temporarily entitled to social welfare added industries (especially livestock production), income support. Farmers were also offered limited moved to central and eastern Canada where grain financial advice. There was no substantive effort to prices were lower (Klein and Kerr 1995). soften the effects of the change. Despite early pre- In 1995 the Canadian government decided to dictions that large numbers of farmers would leave terminate the program, which was becoming the land, only 1 percent of farms failed, with signif- fiscally unsustainable. To ease the transition, a one- icant adjustments occurring in the form of off- time payment of C$1.6 billion was made to eligible farm employment and changes in input use and farmers. An additional C$300 million was invested output mix. in a more efficient grain handling and transporta- Land prices, which had been kept artificially tion system. The one-time payment was spread high by the subsidies, plummeted with their over two fiscal years and made to owners of prairie removal. Marginal land reverted to bush, and farmland with eligible crops grown in 1994 and subsidy-driven land management problems ended. summer fallow land in 1993, adjusted for a produc- Now farmland values have more than recovered as tivity factor, distance factor, and provincial alloca- farm profitability has been restored. Farmers tion factor. Eligible crops were those that were reduced costs and focused on producing higher- eligible for subsidies under the Western Grains value products, where profitable. Many farmers Transportation program. There were no restric- restructured their debts and continued farming, tions on how the payments were used, and they adjusting farm practices to reduce input costs. With were treated as a capital gain rather than as current investment decisions now subject to commercial income, a concession valued by the OECD at an and good farming disciplines, agricultural input estimated $0.6 billion. suppliers were forced to become more competitive, The outcome has been positive overall. The also improving the competitiveness of the agricul- lower grain prices lifted a constraint on value- tural sector. added industries, encouraging entrepreneurship Since 1986­87 the value of economic activity in and innovation; led to diversification into specialty New Zealand's farm sector has grown by more than crops; lowered land prices; and exposed the indus- 40 percent in constant dollar terms, and agricul- try to trade challenges. The change also brought ture's contribution to the economy has risen from Canada into compliance with international trade 14.2 percent of GDP in 1986­87 to 16 percent in agreements. 1999­2000. With the removal of farm subsidies, GDP growth went from 1 percent in 1986 to the current annual average of 5.9 percent. New Zealand The 1984 New Zealand Exit Grant has around 80,000 farm holdings. Sheep and beef Before 1984 New Zealand's farmers were receiving farms account for 20 percent of the number of generous support--in some years as high as 40 per- farms, and dairy farms for 18 percent. Horticulture, cent of the value of production. In 1984 the govern- forestry, cropping, and rural tourism also con- ment abolished the subsidies. With the economy tribute to the rural sector, which employs 11.4 per- almost on the brink of bankruptcy and facing dete- cent of the work force. About 80 percent of New riorating external markets, inflation, and histori- Zealand's farm outputs are exported, accounting cally high interest rates, the government eliminated for more than half of New Zealand's merchandise almost 30 different production subsidies. Although exports. the end of agricultural subsidies took place in con- junction with overall deregulation of the economy Assessing Decoupling and reduced input costs, currency appreciation and low commodity prices during 1985­87 made the The movement toward decoupled agricultural poli- transition stressful. cies is undeniably a step in the right direction, To ease the transition, the government provided reducing trade distortions and increasing world one-time exit grants to farmers leaving the land, prices for developing countries' exports. But how equivalent to about 66 percent of their previous much movement has actually occurred? And what 88 Global Agricultural Trade and Developing Countries have been the net effects on resource use, efficiency, tial. Large payments can have risk reduction effects and trade distortions? that lead to increased output. Direct payments also The rate of agricultural protection in OECD help cover fixed costs, allowing farmers to cross- countries has declined, while the share of domes- subsidize production at market prices. Direct tic support has increased. Total direct support payments can affect farmers' investment and exit to agricultural producers as measured by esti- decisions if they are facing constraints in capital mated producer support averaged $235 billion in and labor markets. Direct payments allow banks to 2000­02, 63 percent of it from border measures. make loans that they otherwise would not and Most support is concentrated in a few sectors allow farmers with specialized skills to stay in (milk, meats, and sugar). agriculture. Although the absolute level of producer support The primary motivation for decoupling is to has remained fairly constant, taxpayer-financed compensate farmers for the move to free markets subsidies paid directly to farmers have increased by providing transitional adjustment assistance. significantly. From 1986­88 to 2000­02, domestic This also makes the programs politically more subsidies to farmers rose 60 percent, with large- palatable and transparent. Ideally, compensation impact programs (output and input subsidies) programs would be universal (open to all sectors in increasing moderately compared with the substan- the economy, not just agriculture) or at least non- tial increases in so-called smaller-impact programs sector-specific within agriculture. A simple and (subsidies for land area and number of animals, minimally distorting scheme would be a one-time decoupled historical entitlements, and payments unconditional payment to everyone engaged in based on input use restrictions and overall farm farming or deemed in need of compensation that is income). Payments based on area planted and nontransferable, along the lines of the one-time number of animals have increased the most, fol- buyouts discussed earlier. lowed by historical entitlements. Several countries, However, because a one-time buyout is an however, have made little progress in reforming the unlikely outcome (unless it is well-targeted in one composition of support away from border support sector), specific attention should be given to time to domestic support (among them Japan and limits, harmonization with other support pro- Switzerland), while others have not needed sub- grams, government credibility, and constraints on stantial reforms (many members of the CAIRNS input use (Baffes and de Gorter 2003 provide a Group). detailed discussion of these conditions along with As for reductions in trade distortions, experi- WTO's potential role on decoupling). Unless these ence in the decoupled programs described above aspects are properly addressed, decoupled pro- has been mixed. The few countries studied here grams are likely to have the same detrimental have moved away from border support to domestic effects as other subsidy programs. support and to less distorting domestic support. Most important, programs should be strictly Although there is evidence of a reallocation of limited in duration. The European Union and resources across agriculture as a result, the decline Turkey have no limit: the United States had (at least in total output and increase in world prices have implicitly) one in the 1996 Farm Bill but violated it been modest. three years later. Mexico's reform had a time limit, In addition to the uneven distribution of "cou- which so far has not been extended. A time limit pled" subsidies (less in major field crops, more in helps to ensure that payments are made for adjust- sugar and livestock), other factors help to explain ment purposes only. the lack of significant reductions in output. Eligi- If there are other (coupled) support programs, bility rules have changed, and expectations about the decoupled program may not eliminate the future policies and dynamic considerations affect incentives to overproduce. All four decoupling current production decisions because producers cases examined here either left other coupled sup- develop expectations about future assistance based port programs in place or added new ones. on past government actions. Experience shows that To maintain government credibility and reduce imperfect decoupled programs still distort trade, uncertainty, eligibility rules need to be clearly especially when decoupled payments are substan- defined and not allowed to change. The time period Experience with Decoupling Agricultural Support 89 on which payments are based, the level of payments, Finkle, Peter, and Duncan Cameron. 1990. Decoupling: The and the sectors covered should all remain fixed. Concept and Its Future in Canada. Ottawa: Agriculture Canada. Updating bases and adding crops create a govern- Goodwin, Barry K., and Ashok K. Mishra. 2002. "Are ment credibility problem, making the decoupling `Decoupled' Farm Program Payments Really Decoupled? policy inconsistent over time. If governments have An Empirical Evaluation." Ohio State University, Columbus, http://departments.agri.huji.ac.il/economics/kenes- the discretion to change eligibility criteria and pay- goodwin2.pdf. ments as market conditions change, these commit- Hennessy, David A. 1998. "The Production Effects of Agricul- ments will not be viewed as binding. Farmers, tural Income Support Policies under Uncertainty." American Journal of Agricultural Economics 80(1): 46­57. meanwhile, will change their production decisions Klein, K. K., and W. A. Kerr. 1995. "The Crow Rate Issue: A to reflect this, thus undermining decoupling. Retrospective on the Contributions of the Agricultural Eco- Support to specific sectors within agriculture nomics Profession in Canada." Canadian Journal of Agricul- should be in the form of taxpayer-funded pay- tural Economics 44(1): 1­18. Moschini, G., and P. Sckokai (1994), "Efficiency of Decoupled ments. There should be no requirement of produc- Farm Programs under Distortionary Taxation," American tion. Land, labor, and any other input should not Journal of Agricultural Economics 76. have to be in "agricultural use." Nash, Eric Francis. 1961. "Agriculture." In Arthur Selton, ed., Agenda for a Free Society: Essays on Hayek's "The Constitution Experience shows the difficulty of designing of Liberty." London: The Institute of Economic Affairs. effective decoupling schemes. But strict criteria are Nicholls, William H., and D. Gale Johnson. 1946. "The Farm required to minimize direct trade distortions Price Policy Awards, 1945: A Topical Digest of the Winning Essays." Journal of Farm Economics 27(1): 267­83. because sector-specific decoupled support can still OECD (Organisation of Economic Co-operation and affect output indirectly, through wealth effects and Development). 2000a. "Decoupling: A Conceptual lessened constraints in credit and labor markets. Overview." COM/AGR/APM/TD/WP 14. Paris. One way to improve the performance of decou- ------. 2000b. "Methodology for the Measurement of Support and Use in Policy Evaluation." Paris. pling schemes might be to have the WTO specify Parry,IanW.H.1999."Agricultural Policies in the Presence of Dis- the conditions; this approach would avoid counter- torting Taxes."American Journal of Agricultural Economics 81. vailing duties by other countries. Spriggs, John, and Dale Sigurdson. 1988. "The Economics of Decoupling and Targeting Canada's Grain and Livestock Stabilization Programs." In Peter Finkle and Duncan Cameron, eds., Decoupling: The Concept and Its Future in References Canada. Ottawa: Agriculture Canada. Swerling, Boris G. 1959. "Income Protection for Farmers: A Alston, J. M., and B. H. Hurd. 1990. "Some Neglected Social Possible Approach." Journal of Political Economy 64(2): Costs of Government Spending in Farm Programs." 173­86. American Journal of Agricultural Economics 72. Swinbank, Alan, and Stefan Tangermann. 2001. "The Future of Baffes, John, and Harry de Gorter. 2003. "Decoupling Support Direct Payments under the CAP: A Proposal." EuroChoices to Agriculture: An Economic Analysis and of Recent 1(1): 28­35. Experience." Paper presented at the Annual World Bank Tangermann, Stefan. 1991. "A Bond Scheme for Supporting Conference on Development Economics­Europe, May Farm Incomes." In J. Marsh, B. Green, B. Kearney, L. Mahé, S. 15­16, Paris. Tangermann, and S. Tarditi, eds., The Changing Role of the Baffes, John, and Jacob Meerman. 1998. "From Prices to Common Agricultural Policy: The Future of Farming in Incomes: Agricultural Subsidization without Protection?" Europe. London: Belhaven. World Bank Research Observer 13(2): 191­211. Van Donkersgoed, Elbert. 1988. "Decoupling: A Family Farm Beard, Nick, and Alan Swinbank. 2001. "Decoupled Payments to and Stewardship Approach." In Peter Finkle and Duncan Facilitate CAP Reforms." Food Policy 26: 121­46. Cameron, eds., Decoupling: The Concept and Its Future in Beghin, John C., Jean-Christophe Bureau, and Sung Joon Park. Canada. Ottawa: Agriculture Canada. 2003. "Food Security and Agricultural Protection in South Winters, L. Alan. 1989­90. "The So-Called `Non-Economic' Korea." American Journal of Agricultural Economics 85: Objectives of Agricultural Support." OECD Economic Studies 618­32. 13: 238­66. Cahill, Sean, A. 1997. "Calculating the Rate of Decoupling for World Bank. 2001. "Turkey--Agricultural Reform Implementa- Crops under CAP/Oilseeds Reforms." Journal of Agricultural tion Project." Project Appraisal Document. Europe and Economics 48: 349­78. Central Asia Region, Turkey Country Unit, Washington, D.C. 6 AGRO-FOOD EXPORTS FROM DEVELOPING COUNTRIES: THE CHALLENGES POSED BY STANDARDS Steven M. Jaffee and Spencer Henson Food safety and agricultural health standards can Standards: Barrier or Catalyst? impede trade, especially for developing countries, The expansion of global trade in perishable agricul- through explicit bans on imports of particular tural products and high-value foods has high- products or through the high cost of compliance lighted the great divergence in national standards with stringent standards, which can diminish com- for food safety and animal and plant health and in petitiveness. In certain circumstances, however, the the capacities of public authorities and commercial new landscape of proliferating and increasingly supply chains to manage the risks associated with stringent food safety and agricultural health trade in these products. For many higher-value standards can be a basis for the competitive reposi- foods, including fruits and vegetables, fish, beef, tioning and enhanced export performance of poultry, and herbs and spices, the challenges of developing countries. Key to this is the ability of international competitiveness have moved well developing countries to upgrade capacity and make beyond price and basic quality to food safety and necessary adjustments in the structure and opera- agricultural health concerns.1 There is increasing tion of their supply chains. In an attempt to rebal- attention to the risks associated with microbial ance much of the dialogue in this area, this chapter pathogens; residues from pesticides, veterinary explores the nature of the new standards landscape medicines, and other agricultural inputs; and envi- and the related capacity requirements, before look- ronmental or naturally occurring toxins. And there ing at the impacts on trade. In addition to the tradi- is greater scrutiny of the production and processing tional approach using quantitative measures of techniques employed along supply chains (Buzby changes in trade that are related to the evolution of 2003). standards, the chapter presents a number of illus- There are several reasons why food safety and trative case studies that relate losses or gains in agricultural health standards, referred to as sanitary trade to food safety and agricultural health require- and phytosanitary measures within the World ments within the context of wider supply chain Trade Organization (WTO), differ across countries challenges. 91 92 Global Agricultural Trade and Developing Countries (Unnevehr 2003; Henson 2004). Differences in developing-country consumers, especially those tastes, diets, income levels, and perceptions influ- with higher incomes and in urban areas. ence people's tolerance of these risks. Differences in The proliferation and enhanced stringency of climate and available technology (from refrigera- food safety and agricultural health standards are a tion to irradiation) affect the incidence of food growing concern among many developing coun- safety and agricultural health hazards. Standards tries and those promoting their increased integra- reflect the feasibility of implementation, itself tion into the world trading system. Reflecting wider influenced by legal and industry structures as well changes in the trade regime for various agricultural as technical, scientific, administrative, and finan- and food products, there is a presumption that cial resources. Some food safety risks tend to be food safety and agricultural health measures will be greater in developing countries because of weak- used as protectionist tools, providing "scientific" nesses in physical infrastructure and the higher justification for prohibiting certain imports or incidence of certain infectious diseases. Tropical applying higher standards to imports than to and subtropical climates may be more conducive to domestic supplies. Even if standards are not inten- the spread of certain pests and diseases that pose tionally used to discriminate, their growing com- risks to health. plexity and lack of harmonization could still Thus the intrinsic risks associated with the pro- impede the trading efforts of developing countries. duction, transformation, and sale of high-value There is also a concern that many developing and perishable food products, combined with dif- countries lack the administrative, technical, and ferent standards and institutional capabilities, can scientific capacities to comply with emerging pose major challenges for international trade. And requirements. The investment and recurrent costs food safety and agricultural health standards of compliance could undermine the competitive are changing rapidly, along with increased public position of developing countries or otherwise com- awareness of food safety in high-income countries press the profitability of high-value food exports. following a series of highly publicized food scares The combined effects of institutional weaknesses or scandals (bovine spongiform encephalopathy, or and rising compliance costs could contribute to the BSE, in beef in the United Kingdom, E. coli in ham- further marginalization of weaker economic play- burgers in the United States, dioxins in animal feed ers, including poor countries, small businesses, and in Belgium). In response, there have been signifi- smallholder farmers. cant institutional changes in food safety oversight A less pessimistic view emphasizes the opportu- and reform of laws and regulations. For long-held nities provided by evolving standards, which some concerns such as pesticide residues, there has been developing countries can use to their competitive a tightening of standards in many countries. And advantage. Many of the emerging public and new standards are being applied to address previ- private standards can serve as a bridge between ously unknown or unregulated hazards, such as increasingly demanding consumers and distant BSE, genetically modified organisms, and environ- suppliers. The standards can provide a common mental contaminants. language within the supply chain and promote As official standards and public oversight have consumer confidence in food product safety. changed, the private sector has moved rapidly to From this standards-as-catalyst perspective, address food safety risks and the concerns and pref- food safety and agricultural health standards may erences of consumers, resulting in a proliferation of provide a powerful incentive for modernizing private codes of practice and other forms of supply developing-country export supply chains and giv- chain governance. Private systems of food safety ing greater clarity to the management functions of governance are also being applied more widely in government. Further, there may be spillovers into middle-income and some low-income countries, in domestic food safety and agricultural health, to the part through investments by multinational super- benefit of the local population and domestic pro- market and restaurant chains and competitive ducers. Part of the costs of compliance could be responses by local firms (Reardon and Berdegue considered necessary investments, while an array of 2002). In addition, new food safety standards in foreseeable and unforeseeable benefits might arise industrial countries are shaping the expectations of from the adoption of different technologies and Agro-Food Exports from Developing Countries: The Challenges Posed by Standards 93 management systems. Rather than degrading the the International Office of Epizootics for animal comparative advantage of developing countries, health, and the International Plant Protection Con- enhancement of capacity to meet stricter standards vention for plant health. The agreement protects could create new forms of competitive advantage, the rights of countries to choose their own "appro- providing the basis for more sustainable and prof- priate level of protection," yet guides members to itable trade over the long term. "take into account the objective of minimizing neg- This rather crude dichotomy between standards ative trade effects." as barriers and standards as catalysts suggests a Important underlying objectives are minimiza- complex reality in which close attention is needed tion of protectionist and unjustified discriminatory to the specifics of particular markets, products, and use of standards and promotion of greater trans- countries to understand how the changing food parency and harmonization. In both regards, expe- safety and agricultural health standards environ- rience has been mixed. The difficulties encountered ment is providing challenges and opportunities for are probably due less to specific shortcomings of developing countries. This chapter draws on the the Sanitary and Phytosanitary Agreement itself literature and work in progress to examine the than to the intrinsic complexities of managing food underlying evidence on the changing standards safety and agricultural health protection and the environment and its implications for developing- rapidly evolving markets for agricultural and food country exporters of high-value agricultural and products. Further, it is evident that WTO members food products. Drawing on both systematic and vary widely in their understanding of the agree- anecdotal evidence, the chapter presents a varied ment and their ability to take advantage of the picture, partially supporting both perspectives. rights and responsibilities it defines. The agreement has not brought an end to the differential application of standards--nor should it. The Sanitary and Phytosanitary Differentiation is a necessary part of any risk-based Agreement: An End to Disguise and Discrimination? food safety and agricultural health control system. The hazards to be monitored and the control meas- During the Uruguay Round of multilateral trade ures to be implemented need to be prioritized at the negotiations, agricultural exporters voiced con- country, industry, and enterprise levels. Political cerns that sanitary and phytosanitary measures factors as well as scientific evidence influence prior- were being used to restrict foreign competition and ities, focusing, for example, on issues of greatest that such protectionist measures would likely concern to consumers and other interest groups increase as the use of more traditional trade barri- (Henson 2001). As resources are limited and imple- ers declined. The Agreement on the Application of mentation may be costly, an effective risk manage- Sanitary and Phytosanitary Measures provided a set ment system will go beyond prioritizing potential of multilateral rules recognizing the need of coun- hazards to differentiate explicitly among alternative tries to adopt such measures and creating a frame- sources of supply based on conditions of produc- work to reduce their trade distorting effects. tion, experience, and assessments of risk manage- The agreement, built on the Standards Code of ment capabilities in the supply chain. the 1947 General Agreement on Tariffs and Trade, permitted measures "necessary to protect human, Separating Legitimate and Illegitimate Standards animal, or plant life and health," yet required that Differentiation regulators base measures on a scientific risk assess- ment, recognize that different measures can achieve When regulators and others have wide discretion equivalent safety outcomes, and allow imports and differentiation is required for cost-effective from particular regions in an exporting country management of food safety and agricultural health, when presented with evidence of the absence or low there remains ample scope for mischief. Yet separat- incidence of pests or diseases. The agreement ing legitimate differentiation from illegitimate dis- encouraged the adoption of international stan- crimination is problematic. Even more difficult is dards, making explicit reference to those of the clearly attributing standards to protectionist inten- Codex Alimentarius Commission for food safety, tions,considering that in most circumstances at least 94 Global Agricultural Trade and Developing Countries partially legitimate food safety or agricultural health controls than do domestic suppliers in certain issues are involved. For example, in two widely refer- industrial countries. But this intensive oversight enced cases of assumed protectionist motivation-- and monitoring often come from private entities, restrictions on exports of Mexican avocados and especially supermarkets and their buying agents, Argentine citrus fruits to the United States--there rather than from official systems. And the methods was scientific justification for measures to prevent of control that exporters face are more visible in the spread of plant disease, though less-trade- their effects, in that compliance for exporters is restricting measures were available (Roberts and assessed at the border, with entry possibly denied Orden 1997). In other cases, trade partners have on this basis, whereas domestic suppliers are regu- different perspectives on the state of scientific lated through inspection of processing facilities, knowledge and the need to make allowance for with a focus on system-based controls and market uncertainty.A prominent case is the dispute between surveillance. the European Union (EU) and United States over the Yet, there is anecdotal evidence that regulatory use of hormones in beef cattle (Pauwelyn 1999; oversight is substantially more stringent on domestic Bureau, Marette, and Schiavina 1998). supplies in certain products and markets. For exam- Thus, questions remain about whether there is ple, there is no official requirement in the United systematic discrimination against imports in the States for border testing of cereals or nuts for the application of food safety and agricultural health presence of aflatoxin. Private-sector testing for afla- controls. One question is whether foreign suppliers toxin levels in cereals is commonplace in the domes- must comply with higher standards than domestic tic market, however, with frequent price discounts suppliers. No systematic research has been done on being applied by buyers. Over a typical three-year this subject, although a great deal of anecdotal evi- period the U.S. Food and Drug Administration's dence is presented by those who purport to have (FDA) Center for Food Safety and Applied Nutrition been adversely affected by such discrimination. inspects all domestic firms that produce low-acid And WTO members raised 241 complaints in the canned foods, yet only 3 percent of foreign facilities Sanitary and Phytosanitary Committee from 1995 that export such products to the United States mar- to 2002 (Roberts 2004). ket undergo such inspection.2 The FDA inspects only General impressions suggest that many coun- 1­2 percent of the more than 6 million consignments tries, both industrial and developing, have a lower of food (and cosmetic products) imported each year. tolerance for certain animal and plant health risks For relatively high-risk products (for example, fish from imports than from domestic sources. There and meat products),a higher proportion of domestic have been cases when countries have restricted than imported supplies is inspected. In both the imports from countries experiencing a plant pest United States and the European Union compliance or animal disease that is also prevalent domesti- monitoring for pesticide residues pays consider- cally. Similar observations can be made for some ably more attention (absolute and proportional) to food safety controls. For example, the United States domestic suppliers than to imports.3 has long argued that, like itself, a broad array of There is also little research comparing the inten- countries has a near-zero tolerance for salmonella sity with which private buyers and distributors in imported poultry products yet this pathogen is enforce their own standards among domestic sup- widely present in domestic supply chains. Coun- pliers and foreign suppliers, especially in develop- tries can also apply discriminatory measures to ing countries. With less opportunity to observe different importing countries. For example, the directly the food safety and agricultural health Philippines complained that Australia prohibited control systems employed by developing-country imports of Philippine sauces containing benzoic suppliers, private buyers would likely emphasize acid while permitting imports from New Zealand end-product testing or third-party certification of of similar products containing that additive. quality management systems. This is certainly a clear trend among buyers in the United Kingdom and the Netherlands, for example, yet it is doubtful Private and Public Oversight and Monitoring that such requirements are being imposed on High-value food exporters in developing countries developing-country suppliers at the same rate as on frequently claim that they face more rigorous their industrial-country competitors.4 Agro-Food Exports from Developing Countries: The Challenges Posed by Standards 95 Increased Complexity of the Standards imported fresh fruit and vegetables, wide variations Environment remain in operative standards due to countries' dif- ferent approaches to surveillance and enforcement. The overall picture for food safety and agricultural Variations in standards are also common in health requirements in trade is becoming increas- other sectors. Henson and Mitullah (2004) note the ingly complex and fast moving as standards are varied standards that developing countries must promulgated in multiple spheres at both public and meet to gain and maintain access to the U.S., EU, private and national and international levels. The and Japanese markets for fish products. While some complexity of this issue stems from the variability requirements overlap, differences remain in both of the standards themselves and from differences in regulatory and technical requirements. Likewise, how and with what intensity standards are moni- Mathews, Bernstein, and Buzby (2003) highlight tored and enforced, which is also changing over the range of product and process standards coun- time. tries require to minimize the risk of salmonella The transparency of official regulatory measures contamination in poultry products. Dohlman in the application of food safety and agricultural (2003) and Otsuki, Wilson, and Sewadeh (2001) health requirements has clearly improved since the discuss the significant differences in the maximum Sanitary and Phytosanitary Agreement entered into permitted level for aflatoxin in cereals and nuts and force. Some 85 percent of WTO members have in the sampling methods used. This lack of harmo- established an "enquiry point" for obtaining infor- nization of standards and conformity assessment mation on proposed food safety and agricultural procedures raises production and transaction costs health requirements. Between 1995 and 2002 WTO for developing-country suppliers, necessitating members submitted some 3,220 notifications indi- duplicative testing and reducing their ability to cating the nature and objectives of proposed meas- achieve economies of scale in production and in ures, the products they applied to, whether they food safety and agricultural health management were based on an international standard, and when functions. the measure was to come into force. These notifica- Also contributing to the increased complexity of tions provide advance warning of new or modified the standards environment is the expansion of risk- measures and an opportunity for trading partners based process standards relating to production, to raise questions about the proposed measures, postharvest, and other procedures, and the prolifer- both bilaterally and through the Sanitary and Phy- ation of private standards. Roberts (2004) notes tosanitary Committee. An increasing proportion of that the major international standards organiza- WTO members, including developing countries, tions have devoted more of their attention and has been taking advantage of this opportunity to resources over the past decade to the development raise concerns (Roberts 2004). of common approaches to risk identification, While the transparency of many food safety and assessment, and management than to international agricultural health measures has increased, consid- standards themselves. This reflects both the ineffi- erable variation remains in standards across coun- ciency and the inefficacy of end-product testing, tries. And there is widespread uncertainty about particularly in view of the levels of risk deemed how certain countries are implementing their stan- acceptable today and the emergence of new or dards. Roberts, Josling, and Orden (1999) note the newly prominent food-borne pathogens. paucity of international standards for many agro- With respect to private standards, there have food products and indicate that the vast majority been attempts to harmonize standards formerly of food safety and agricultural health measures applied by individual private companies, yet a notified to the WTO during 1995­99 had no inter- national standard.5 With specific reference to horti- plethora of private standards are still simply com- municated through individual supply chains and cultural products, Roberts and Krissoff (2003) can vary widely in their specific requirements. found that over the same period two-thirds of Examples of private protocols that have been codi- notifications involved measures for which there was fied and are available to the public include food no recognized international standard and that safety and food hygiene protocols, such as the many involved maximum pesticide residue levels. British Retail Consortium Technical Food Standard Jaffee (2003) notes that despite EU efforts to har- and the EUREPGAP Fruit and Vegetable Standard, monize maximum pesticide residue levels in 96 Global Agricultural Trade and Developing Countries which combines food safety, environmental, and hazards and having an information system to social dimensions. Other standards focus on social inform decision-making processes. or environmental issues, such as Social Account- · Employing emergency procedures in the event ability 8000, the Ethical Trading Initiative, and the of emerging hazards or outbreaks. Marine Stewardship Initiative. · Certifying that traded products meet established Variations in food safety and agricultural health food safety risks. requirements together with the progressive shift · Undertaking scientific analysis of hazards in toward process-based measures have enhanced the agricultural inputs and food products. importance of "equivalence" of national standards · Establishing and maintaining the identity of and systems. Currently, there is no systematic agricultural products through the supply chain. recording of equivalence agreements. Most appear · Establishing and maintaining systems for to be between industrial countries. Certain develop- hygienic practices in agro-food product han- ing countries, including successful agricultural dling and transformation. exporters, have highlighted the difficulties in gain- · Registering the production, distribution, and ing recognition for the equivalency of their food use of agricultural inputs that may pose risks to safety and other controls to those of their major human, animal, or plant health. trading partners (WTO 2001). A successful and wide-ranging example of equivalence, however, is Administrative and technical capacities for food the recognition by the European Union that many safety and agricultural health management are developing and industrial countries have established embodied in institutional structures and proce- systems of hygienic control for fish and fishery dures, physical infrastructure, and human capital. products that offer a level of protection at least com- It is frequently assumed that managing food safety parable to its own legislation (see discussion below). and agricultural health is predominantly a public- A parallel trend, reflecting the proliferation of sector responsibility. While some crucial regulatory, private standards, is the heightened importance of research, and management functions are normally certification of compliance with defined standards, carried out by governments, and importing coun- which is typically undertaken by a third-party tries may require that certain functions be per- agency that the buyer recognizes as "competent." formed by a designated public-sector "competent A crucial issue for developing countries is the authority," the private sector also has important establishment of certification capacity and parallel roles: institutions for accrediting certification bodies. · Because it is typically well informed about tech- Exporters in countries that lack an accreditation nical options and hazard management systems, certification system may be forced to use the ser- it should contribute to standard setting. vices of an accredited body in another country, at · Compliance with food safety and agricultural considerable cost (El-Tawil 2002). health standards requires specific actions by individual producers and processors. What Capacity Is Needed? · Capacity building in the private sector can com- Countries frequently require guarantees that plement (or substitute for) public-sector capac- imports come from areas that are free of certain ity, as through investment in accredited labora- pests or diseases; that minimum standards of tory testing facilities. hygiene have been applied in manufacture, packag- ing, and distribution; and that products are free of Development of food safety and agricultural excessive residues of pesticides, medicines, and health management systems is closely related to the other contaminants. The exporting country must availability of wider technical, administrative, and have the capacity to comply with these require- scientific capacities that reflect broader patterns ments and to demonstrate compliance. Among the of economic development as much as specific required capacities are: demands for food safety and agricultural health controls. Unnevehr and Hirschhorn (2001) high- · Detecting the presence or demonstrating the light the capacity needs for food safety manage- absence of biological, chemical, and physical ment at different stages of economic development. Agro-Food Exports from Developing Countries: The Challenges Posed by Standards 97 Export-oriented agriculture gives rise to a new set partners. These costs can come in various forms, of challenges because foreign food safety and agri- including fixed investments in adjusting produc- cultural health requirements may differ sharply tion and processing facilities and practices, from domestic requirements, especially in the case recurrent personnel and management costs to of low-income countries (Dong and Jensen 2004). implement food and other control systems and the Some regulatory, technical, and administrative public- and private-sector costs of conformity capacities represent a greater constraint on assessment. developing-country exports of agricultural and Typically there are a variety of technological and food products than do others.6 In general, weak- administrative ways in which to achieve compli- nesses in the management of plant and animal ance with a certain standard. For this and other health issues are more likely to be an absolute bar- reasons the level and relative significance of com- rier to trade than is lack of food safety controls. For pliance costs can vary enormously from industry to many food safety hazards there is an array of effec- industry and between countries.7 Important vari- tive technologies or approaches, some of which do ables include the structure and conditions of the not require sophisticated equipment or expertise. supply chain, the extent of administrative and sci- Even where management of food safety hazards is entific capacities, the degree of cooperation within well within the capacities of producers and proces- industries and between the public and private sec- sors, systems of conformity assessment require test- tors, and the strength of technical service indus- ing and certification of food safety management tries. Where the export industry is mature and systems and end products. Many developing coun- reasonably well developed, changes in food safety tries lack the capability to undertake the rigorous and agricultural health standards may require epidemiological surveillance and risk assessments only incremental adjustments by producers and demanded by trading partners. They lack the exporters and modest changes in public-sector accredited laboratories and internationally recog- oversight arrangements. Where the supply chain is nized systems for certification (El-Tawil 2002). makeshift, however, or uses multipurpose facilities Thus, regardless of private-sector capacity to meet and when new requirements (or levels of enforce- the food safety and quality requirements of foreign ment) necessitate major upgrades, some firms may customers, the country as a whole will be unable to need to redirect their products to less-demanding gain market access. markets, while others will need to undertake signif- While many developing countries have wide- icant fixed investments. spread weaknesses in food safety and agricultural Consider the differences in adjustment costs health management capacity, there is evidence that associated with investments in the upgrading of even low-income countries can establish the regu- hygiene controls in the shrimp industries in latory, technical, and administrative arrangements Bangladesh and Nicaragua (table 6.1). In to meet demanding standards in high-income Bangladesh major investments had to be made in export markets. The European Commission has the mid-1990s to upgrade fish-processing facilities, recognized a relatively large--and growing-- product-testing laboratories, and other areas in number of low- and lower-middle-income coun- response to repeated quality and safety detentions tries as having standards of hygiene in the capture, of products entering the United States and a ban in processing, transportation, and storage of fish and 1997 on shrimp imports into the European Union. fishery products that are at least equivalent to those These investments equaled 2.3 percent of the value of the European Union. Their shipments benefit of the country's shrimp exports in 1996­98. Annual from reduced physical inspection at the border. maintenance costs for hazard analysis and critical control point (HACCP) and regulatory systems equaled 1.1 percent of exports. The Nicaraguan How Significant Are Compliance shrimp industry needed to make adjustments dur- Costs? ing 1997­2002 to hygiene controls to ensure com- Developing countries can incur significant "costs pliance with modified U.S. fish safety regulations, of compliance" whenever changes are made in including requirements to implement a HACCP international standards or those of their trading program. But many Nicaraguan factories were 98 Global Agricultural Trade and Developing Countries TABLE 6.1 Costs of Compliance with Export Food Safety Requirements in the Shrimp Processing Industry in Bangladesh and Nicaragua (millions of US$) Bangladesh Nicaragua Cost 1996­98 1997­2002 Industry facility upgrading 17.55 0.33 Government 0.38 0.14 Training programs 0.07 0.09 Total 18.01 0.56 Annual maintenance of HACCP program 2.43 0.29 Shrimp exports during focal periods 775.00 92.60 Average annual shrimp exports 225.00 23.20 Ratio of upgrade costs to focal year export (%) 2.3 0.61 Ratio of maintenance costs to annual exports (%) 1.1 1.26 Source: Based on Cato and Lima dos Santos 2000, and Cato, Otwell, and Coze 2003. relatively new and modern, so only modest incre- The enhancement of food safety capacity can mental investments were needed, equivalent to also have more dynamic and wide-ranging impacts 0.6 percent of the value of exports. on private-sector suppliers. For example, imple- Many technological and organizational changes menting an HACCP system and gaining third- involve shifts in levels and structures of operating party certification can send positive signals to costs. The costs associated with these changes are existing and potential customers, enabling firms to often controversial. The changes are sometimes reposition themselves in the marketplace or access perceived to be unjustified, because they lack scien- new markets. Indian fish-processing plants that tific basis or replace simpler, less costly procedures have invested in sophisticated systems of hygiene that might provide similar outcomes. Another control are seeking to access higher-value markets complaint is that suppliers obtain little or no bene- for processed and semi-processed products. Some- fit beyond continued market access, while the times, when problems are experienced in comply- opportunity cost of the required investments can ing with requirements in a particular market, be considerable. This complaint is more difficult to producers and exporters will shift to markets with sustain, often reflecting a lack of appreciation of the lower or different food safety requirements. Kenyan frequently intangible or indirect benefits that can fish exporters, which have been highly dependent result from enhancement of food safety controls, on European markets, have attempted to diversify for example. Improved control systems can reduce their exports to Australia, Japan, and the United waste, improve product-cost accounting, and States. enhance staff morale. Thus, changes in product and But even where the administrative, technical, process technologies can generate substantial and financial burdens of compliance are manage- increases in efficiency, reducing production costs able at the country or industry level, the burdens and promoting competitiveness. may be too great at the firm level. There is a general The expenditures related to standards compli- concern that the challenge of rising standards is ance can have other beneficial, multiplier effects. marginalizing smaller players, especially producers, Some of the needed investments may require labor, traders, and processors, as well as smaller industries especially skilled and supervisory workers, creating as a whole. There is, however, little empirical evi- additional job opportunities. Other expenditures dence to support this argument. In part, this is may go toward building materials, contractors, and because of the difficulties of disentangling the spe- technical services, much of which could be sourced cific role of standards compliance in the consolida- locally. Only where upgrading relies primarily on tion processes of agro-food systems. imported equipment or expertise would there be In many cases, compliance requirements exacer- few multiplier effects. bate other factors that threaten the status quo in Agro-Food Exports from Developing Countries: The Challenges Posed by Standards 99 established supply chains. For example, both the A frequent presumption when discussing the Indian and Kenyan fish-processing sectors were marginalization of suppliers is that standards com- facing longer-term challenges when forced to com- pliance is a do-or-die situation. In reality, however, ply with enhanced hygiene standards for exports. there is rarely just a single market for a particular Indian exporters were facing intense price competi- product. Suppliers need to seek out markets (and tion from other suppliers such as Thailand and market segments) where they have advantages Vietnam. The Kenyan fish-processing sector suf- rather than disadvantages. For example, there may fered from chronic excess capacity because of raw be opportunities in domestic or regional markets material shortages. In both cases the costs of com- for the same or similar products, with lower prices pliance with stricter food hygiene standards have offset by the absence of compliance challenges and induced consolidation of the industry that likely costs. Directing attention to these markets may be would have occurred anyway, albeit over a longer one way to avoid marginalization. Thus, the devel- period. In Nicaragua the decline in the production opment of high-value agricultural and food share of small-scale shrimp producers had more to products sectors in the future is likely to be do with Hurricane Mitch and its aftermath than bimodal, with some firms upgrading and adapting with the tightening of standards (Cato, Otwell, and and others targeting other markets and raising their Coze 2003). capacity at a slower pace. A particular concern is that smaller players can be disadvantaged where there are economies of What Impact Are Standards scale or scope in the implementation of particular Having on Exports of High-Value technologies or administrative systems. Studies of Agro-Food Products? compliance with labor and environmental stan- dards in the United States suggest that costs are The application of food safety and agricultural proportionately higher for smaller firms (Crain health standards by governments and the private and Johnson 2001). In some cases the necessary sector can significantly affect international trade. investments have elements of lumpiness, for exam- While most standards are designed in pursuit of the ple, in laboratory equipment and cold-storage legitimate goals of maintaining human, plant, and facilities, which are economically viable only for animal health, they can also serve as technical bar- large-scale operations or require collective action. riers to trade. Roberts, Josling, and Orden (1999) Likewise, smaller firms may find it more difficult to classify technical trade barriers associated with hire certain types of skilled personnel. More gener- agricultural and food products into three cate- ally, smaller firms can be overwhelmed by the sheer gories: full or partial import bans; technical specifi- number of changes needed to comply with new cations, including product and process standards; food safety requirements, even when the cash and information remedies, including packaging investments required are not substantial. and labeling requirements and controls on volun- Sometimes certifying that the standards have tary (health and other) claims. Full or partial bans been met is more difficult for small producers than are the most trade restricting. Total bans are typi- complying with food safety and agricultural cally used when great risks are associated with cer- health requirements. For example, Kenyan veg- tain plant and animal health problems and where etable exporters face considerable oversight costs in cost-effective measures are not available. Partial demonstrating compliance to their major European bans may permit trade only in certain seasons or buyers.8 In turn, this generates pressure for rational- from certain countries or regions. Technical specifi- izing supply chains. Changes in the product compo- cations and informational remedies will normally sition of trade may also affect structural patterns.9 apply to both imports and domestic supplies. Their Further, in a competitive environment, exporters effects on trade will derive from the relative abilities find it difficult to control the volume and continuity of different suppliers to comply with these meas- of smallholder supplies due to side-selling by farm- ures, the incidence of compliance costs, and how ers. Where export supply commitments are firm each affects the relative competitiveness of different and specific, exporters need more effective control, suppliers. and this can induce backward integration into While there is general agreement that food production. safety and agricultural health measures strongly 100 Global Agricultural Trade and Developing Countries affect international agro-food trade, there is no were identified as constraining exports valued at consensus on the importance of individual meas- $5 billion, equal to around 7 percent of U.S. agri- ures, their impact compared with other trade- cultural, food, and forestry trade in 1996. Two- distorting measures, or their aggregate net effect. thirds of the identified measures, including nearly Testing the empirical impact of such standards on all full or partial import bans, addressed risks for trade is enormously difficult. First, it requires animal or plant health (Roberts, Josling, and Orden assumptions about how the broad array of meas- 1999). ures is actually enforced and how enforcement This type of broad estimate of trade effects has deters or encourages potential export suppliers, not been made for any other country. Other depending on whether suppliers need to make approaches have provided insights into the subject, major or modest adjustments. This variable cannot however. Most commonly, researchers have looked be aggregated and differs across countries and to the only two available multicountry sources of industries. Second, food safety and agricultural data on the subject, official listings of agricultural health standards may have secondary effects, for and food product detentions and rejections by example, leading to shifts in sourcing, the produc- industrial countries and the growing number of tion of complementary and competitive goods, and complaints recorded by the Sanitary and Phytosan- the spread of regulations and restrictions to other itary Committee. Though incomplete, both are countries. Third, a specific measure may not be a useful proxies for the trade-inhibiting effects of dominant or even important determinant of food safety and agricultural health standards. observed trade flows. There is a risk of ascribing agro-food standards to shifts in trade that are Border Detentions and Rejections of Agricultural driven by other economic or technical factors. and Food Products Fourth, there are problems in defining the counter- factual. Without the measure, would trade have Information is available for a limited number of been unimpeded, or would distributors and con- countries (through periodic reports and web-based sumers have sought the product from other suppli- databases) on the incidence of detention or rejec- ers instead? In the absence of a (trade-restricting) tion of imported agricultural and food products for measure, might overall demand have declined for a reasons associated with quality, safety, labeling, or product for which certain problems were identi- other technical issues. The most widely available fied? Finally, many food safety and agricultural and cited data are for the European Union and the health measures will affect domestic suppliers as United States.10 The data provide a reasonable pic- well, with varied outcomes in terms of shifts in the ture of the incidence of product rejections over relative competitiveness and market share of the time by country of origin but do not specify the different players. volume or value of rejected consignments. These and other empirical problems have led Several patterns emerge from product rejection researchers to devote more attention to specific data for the European Union and United States: cases and to attempt to highlight the role played by (changing) food safety and agricultural health · Rising incidence. In the European Union the requirements on bilateral or broader multicountry number of notifications or alerts increased more patterns of trade. Some of the cases are discussed than sixfold between 1998 (230 cases) and 2002 below. Only one study, however, has attempted to (1,520). This increased incidence of rejections provide an aggregate measure of the level of agri- reflects a combination of factors, including the cultural and food trade constrained or blocked by tightening or harmonizing of standards, appli- technical barriers. In 1996 the U.S. Department of cation of standards for formerly unregulated Agriculture, drawing on the expert opinions of staff hazards, and substantially increased capacity and other regulatory personnel, found that "ques- for inspection and enforcement. In the United tionable" technical barriers (measures judged to States there was a sixfold increase in the number have no scientific basis) were inhibiting U.S. of product inspections by the FDA, in part exports of agricultural and food products to some because of heightened concerns about bioter- 62 countries. More than 300 market restrictions rorism.11 Agro-Food Exports from Developing Countries: The Challenges Posed by Standards 101 · Product concentration. Most detentions and down from 41 percent. This pattern reflects rejections occurred in a few product categories: the growing harmonization of EU standards fish and crustaceans (35 percent of rejections in for an array of chemical contaminants and the 2002), meat products, and fruits and vegetables. increased political and technical attention to For the European Union there was also a high these issues within the European Union. For the incidence of rejections for nuts, while for the United States a large proportion of border rejec- United States there were many rejections of low- tions in the late 1990s was due to the presence of acid canned foods. Comparatively few rejections filth or foreign bodies (32 percent), microbial were issued on quality or safety grounds for bev- pathogens (17 percent), or problems associated erage crops, cereal products, feedstuffs, or spices. with the packaging or labeling of canned food · Country of origin concentration. A few countries products for which botulism is a risk (13 per- accounted for the bulk of rejections. Among cent). A smaller proportion of rejections was developing countries most of the rejections due to chemical contaminants (12 percent). were from countries that have been dominant suppliers of "sensitive" products for many years Neither the European Union nor the United (for example, Brazil, Mexico, Thailand, and States systematically reports on the volume or value Turkey) or newly emerging large exporters of of trade that is affected by border inspections and such products (for example, China, India, and rejections. To obtain a rough notion of the value of Vietnam). In 2002 five countries (Brazil, China, trade interrupted by technical measures, data were Thailand, Turkey, and Vietnam) accounted for collected (from official sources and consultations nearly 60 percent of EU rejections of agricul- with private traders) on the proportion of trade in tural and food products from outside Europe. particular products that was likely to have been Some of these countries, however, are simulta- detained or rejected in 2000­01. These estimates neously increasing their EU market share for were then applied to overall trade in these prod- such products, suggesting that border rejections ucts to estimate the value of interrupted trade are more of an irritant than a major problem (table 6.2). For simplicity, the proportion of trade for larger exporters. for particular products that is subject to rejections · Minimal interception of products from low- is assumed to be the same for products flowing income countries. Exports from low-income between low-, middle-, and high-income countries. countries account for a very small proportion of This is unlikely to be so in practice, but data are not product rejections. For example, in 2002 the available to provide more refined estimates. European Union rejected only 26 consignments The value of world agro-food trade affected by from low-income Sub-Saharan African coun- official product rejections at the import level is esti- tries, with most countries experiencing only one mated at $3.8 billion in 2000­01.12 This is almost or two rejections. Most of these countries are certainly an overestimate since similar levels of exporting less sensitive products in terms of rejection are assumed for products entering devel- food safety or agricultural health risks, or they oping countries as for those entering industrial have been recognized as being fully harmonized countries, even though levels of standards and with EU requirements for more sensitive prod- enforcement capacities are typically lower in devel- ucts such as fish and thus are subject to lower oping countries. Reflecting the dominant share of levels of border inspection. high-income countries in certain product groups · Leading reasons for product rejections. For the for which detention or rejection levels are high (for European Union the largest (and growing) pro- example, meat and dairy products, other processed portion of rejections concerns chemical and foods, and processed fruit and vegetables), these other contaminants in food, especially veteri- countries are estimated to account for 53 percent nary drug residues, pesticide residues, and of rejected exports, while they account for some mycotoxins. Chemical contaminants accounted 63 percent of world agricultural and food product for nearly two-thirds of rejections in 2002, up exports. The estimated value of developing- from 55 percent in 2000. Microbial pathogens country agro-food border rejections is $1.8 billion, were implicated in 30 percent of rejections, 74 percent of it accounted for by middle-income 102 Global Agricultural Trade and Developing Countries TABLE 6.2 Estimated Value of World Agricultural and Food Trade Directly Affected by Import Border Rejections Based on Technical Standards, 2000­01 (millions of US$) Estimated High- Middle- Low- Proportion Income Income Income Total Trade Product Group of Trade (%)a Countries Countriesb Countries China Affected Meat and dairy products 1.25 811 142 8 21 982 Fish and fishery products 1.00­2.00 232 417 145 90 884 Fruit and vegetables 0.75­1.50 367 439 44 61 911 Grains 0.50 160 40 6 8 214 Animal feed 0.50 65 39 4 2 110 Tropical beverages 0.25 25 18 16 0 59 Nuts and spices 0.75­1.50 16 33 30 1 80 Other processed food 1.00­2.00 122 53 3 6 184 All other categoriesc 0.25 199 112 19 6 307 Total 1,997 1,332 275 195 3,799 Proportion of trade affected 0.70 1.10 0.93 1.25 0.84 a. Where there are two numbers the first relates to exports of high-income countries and the second to those of middle- and low-income countries. b. Excluding China. c. Includes oilseeds, textile fibers, drinks, tobacco/cigarettes, and sugar/confectionery. Source: Authors' computation based on official data and consultations with private traders. countries. For low-income countries the estimated tially lower than the proportion of sales that are $275 million in agricultural and food product trade subjected to price discounts by private buyers rejected at the importing-country border repre- because of quality defects, lack of timeliness, and sents less than 1 percent of their agricultural and poor presentation. The products with the highest food exports. The product composition of the estimated proportion of rejections are also those rejected exports is broadly consistent with the data with the highest rates of growth in international presented earlier on EU and U.S. rejections. For agricultural trade. middle-income countries, the dominant products Thus while undoubtedly an irritant to exporters, are fruits and vegetables and fish, followed by live- border rejections are not a major impediment to stock products. For low-income countries, fish is trade. Still, they are costly, both in the value of lost the dominant category, accounting for more than product and in adverse reputation effects on the half the estimated rejections. supplier and the country of origin. Some import- Until recently, border rejections for food safety ing countries will list for automatic detention par- or related technical reasons have had only a modest ticular suppliers or the entire country following impact on overall trade in agricultural and food repeated violations of food safety and other products, including that of developing countries. standards. Subsequent shipments are detained, An estimated 1 percent of this trade was directly inspected, and tested at the expense of the exporter affected in 2000­01. Further, only a small propor- or importer until a record of compliance has been tion of rejected consignments is actually destroyed (re)established. This can take a long time, and the at the point of import. Some (perhaps significant) costs can be considerable (Lamb, Velez, and Barclay proportion of the product is reshipped, recondi- 2004). Further, during this period exporters may tioned, or otherwise managed for sale whether in lose customers who are unwilling to incur the costs the domestic market of the exporter or in some and delays associated with enhanced border for- other international market. Indeed, for most food malities. categories the proportion of agro-food trade that In addition, there are some indications for cer- encounters official rejection is probably substan- tain high-income countries that increased attention Agro-Food Exports from Developing Countries: The Challenges Posed by Standards 103 is being given to border inspections of products Because of an inability to meet a broad array of deemed "sensitive" in relation to new regulatory food safety and agricultural health requirements concerns about food safety and agricultural health pertaining to livestock disease and hygiene con- risks. If the patterns described above are indicative, trols, most low-income countries are restricted to an increasing level of border interceptions of prod- trade in live animals rather than livestock products. ucts would be expected in coming years. This will This avoids the need for attention to hygienic either increase the transaction costs for certain slaughter in an abattoir, meat inspection, and developing-country suppliers or induce them to refrigerated transport.15 Even if animal disease and make adjustments in production, postharvest, and hygiene capacity could be enhanced, however, low- product monitoring and testing arrangements. income countries would need to compete with Border rejections attributable to food safety well-established livestock product exporters such as concerns represent only a small part of the con- Argentina and Australia, which are more reliable straint on international trade in agricultural and producers with fewer animal health problems and food products associated with food safety and agri- more standardized production. However, the bene- cultural health measures. For example, although fits from access to high-value markets could be meat and dairy products may be subject to the considerable for developing countries that invest in highest level of rejections in global trade, these animal disease controls, as a case study of foot and are not significant for low-income countries and mouth disease controls in Zimbabwe shows (Perry are probably of secondary importance for most and others 2003). middle-income countries. In terms of the impact on aggregate trade, far more inhibiting are the Disputes and Complaints through the WTO broad array of measures related to animal and plant health that render large numbers of countries ineli- Complaints and counter-notifications made gible to supply many livestock products and food through the Sanitary and Phytosanitary Committee crops to other countries (Sumner 2003). within the WTO also provide an indication of While this pattern undoubtedly reflects tradi- the nature and breadth of the standards challenge tional trade protections and subsidies in industrial for developing countries (table 6.3). While the countries that distort world trade, animal disease counter-notification database and the information controls exclude many developing countries from provided in most counter-notifications do not per- world markets for these products altogether.13 mit quantifying the levels of developing-country In part this reflects the prevalence of endemic trade that has or might be affected by the contested infectious diseases of animals in many low- and measures, it does provide some insights. A sum- middle-income countries. Indeed, the high costs of mary of complaints by regulatory goal and country establishing and maintaining disease-free areas can group suggests that (at least some) developing be beyond the means of many of the poorest coun- countries have actively used this formal review and tries. Many developing countries lack the surveil- complaint process to register their concerns about a lance and risk assessment capacity to demonstrate significant number of notified measures by both that they have areas that are disease-free and to get industrial and developing countries. A more these areas recognized as such by the International detailed look at the individual complaints, indicates Office of Epizooties.14 And even where developing that: countries have established disease-free areas, they face the risk that trade will be disrupted should · Complaints by developing countries are domi- outbreaks of disease occur. A recent example is the nated by a handful of countries--Argentina, restrictions applied to exports of poultry from Brazil, Chile, and Thailand. Each of these coun- Thailand and Vietnam because of an outbreak of tries has issued or supported more than a dozen avian flu. The overall impact of animal disease complaints, with Argentina being involved in issues, therefore, is to enhance the risks associated more than a quarter of all developing-country with trade in livestock products and put a great complaints. Only a handful of other countries, onus on public authorities to invest in disease con- including Ecuador, India, the Philippines, South trols and to ensure their continued efficacy. Africa, and Uruguay, have been involved in 104 Global Agricultural Trade and Developing Countries TABLE 6.3 Number of Counter-Notifications to the Sanitary and Phytosanitary Committee Relating to Reported Measures, 1995­2002 Regulatory Goal of Contested Measure Plant Animal Human Complaints Health Health Health Othera Total By industrial countries against: Industrial countries 16 7 44 3 70 Developing countries 17 11 41 4 73 Multiple countries -- 2 -- -- 2 Subtotal 33 20 85 7 145 By developing countries against: Industrial countries 12 12 34 2 60 Developing countries 8 17 7 2 34 Multiple countries -- 2 -- -- 2 Subtotal 20 31 41 4 96 Total 53 51 126 11 241 -- Not available a. Includes complaints about horizontal regulations (such as those regulating products of modern biotechnology) that reference human, animal, and plant health as objectives. Source: Roberts 2004. multiple cases. This pattern of participation dealing with foot and mouth disease and beef reflects the prominence of certain countries in products or bovine spongiform encephalopathy trade in a few product categories, especially beef and animal by-products for pet food, animal and horticultural products, rather than the over- feed, and cosmetics. Similarly, most complaints all structure of developing country agricultural about plant health issues relate to claims of and food trade. overly restrictive measures for plant diseases or · These data alone provide little information pests or for horticultural products. Complaints about the extent to which food safety and agri- related to food safety are a mixture of specific cultural health measures are inhibiting exports concerns, with no large clustering around partic- of low-income countries. Low-income countries ular themes. Surprisingly, given the huge impor- are weakly represented in counter-notifications, tance for developing-country trade, there are issuing or supporting complaints in only five few complaints about measures governing fish cases. This could reflect the structure of their products. exports (concentrated in commodities for which · The reasons for developing-country complaints food safety and agricultural health measures are are varied, yet most involve concerns about the of lesser importance) or their limited capacity to lack of scientific evidence in relation to food participate in the formal review process. This safety, the absence of risk assessments in relation lack of formal complaints does not mean that to plant health, and inconsistencies between they have been able to resolve their concerns country and international standards in animal bilaterally. health. · Among the seemingly large number of · Among industrial countries, the European developing-country complaints are a limited Union has been the subject of the largest num- number of repeated concerns, with slight varia- ber of complaints by developing countries. tions. Most complaints about animal health There were more than three times as many com- issues relate to what are claimed to be overly plaints against the European Union as against restrictive (and nonscientifically based) measures the United States. Several factors might account Agro-Food Exports from Developing Countries: The Challenges Posed by Standards 105 for this. Harmonization of food safety and agri- analysts are using case studies to examine the cultural health measures within the European effects of specific standards on the trade of particu- Union has often resulted in the adoption of the lar countries and products. Earlier work empha- most stringent standards previously applied by sized the potential disruptive impact of food safety individual member states. The European Union and agricultural health measures on exports from has more frequently and most visibly embraced developing countries (Otsuki, Wilson, and the precautionary principle when adopting cer- Sewadeh 2001; Wilson and Otsuki 2003). More tain standards, sometimes giving rise to contro- recent work by the U.S. Department of Agriculture versies over the scientific basis for the measures. (Buzby 2003) and the International Food Policy And because of the complex administrative Research Institute (Unnevehr 2003) point to more structure of the European Union, some coun- varied experiences. Other case study analyses have tries find it difficult to resolve concerns through been undertaken by the United Nations Conference bilateral discussions and therefore more readily on Trade and Development (UNCTAD), the United turn to the Sanitary and Phytosanitary Commit- Nations Environment Programme (UNEP), and tee for concerns related to the European Union the World Bank.16 than for other countries. This section draws on selected examples to illus- trate the complex ways in which sanitary and Thus, the growing number of recorded com- phytosanitary measures can affect developing- plaints or counter-notifications by developing country exports. Three of the most prominent countries provides only a crude indicator of the concerns raised in the literature are emphasized in extent to which food safety and agricultural health the selection of case studies: fish trade bans and measures impede their trade in high-value agricul- their wider supply chain effects, limits on myco- tural and food products. These complaints proba- toxins as trade barriers, and the strengthening of bly represent only the tip of the iceberg as most horticultural product and process standards. concerns and disputes are raised bilaterally, and the majority of negotiations are handled by technical Fish Bans and Their Wide Supply-Chain Effects organizations rather than country trade representa- tives. Some of the complaints have occurred in the Since 1990 developing-country exports of fish and context of expanding trade ties, which can increase fishery products have increased at an average the seriousness of previously minor effects. annual rate of 6 percent (Delgado and others 2003). There is little basis for associating the growing A major challenge faced by developing countries in number of complaints with deliberate protection- seeking to maintain and expand their share of ism. Many of the concerns seem to be related more global markets is the progressively stricter food to inadequate (scientific) information rather than safety requirements, particularly in major indus- to discrimination. Further, the apparatus of formal trial countries. Previous studies suggested that complaints relates only to mandatory standards set some exporters experienced considerable problems by public agencies. A growing array of standards in complying with these requirements. are being set privately, either through consensus The European Union lays down harmonized within particular industries or by the gatekeepers of requirements governing hygiene throughout the the dominant supply chains. While many such supply chain for fish and fishery products. Process- standards are ostensibly voluntary, they are becom- ing plants are inspected and approved individually ing the de facto standards to gain or maintain by a specified "competent authority" in the coun- access to particular buyers or market segments. try of origin, whether an EU member state or a third country, to ensure compliance. Imports from third countries are required to have controls that Some Illustrative Case Studies are at least equivalent to those of the European Because the data on agricultural and food product Union.17 Exports from countries for which local rejections and disputes related to food safety and requirements have been recognized as equivalent agricultural health measures provide an incomplete are subject to reduced physical inspection at the picture of the effects on developing countries, border. Countries that have not yet met these 106 Global Agricultural Trade and Developing Countries requirements but that have provided assurances consignments of Nile perch from the region. that their controls are at least equivalent to those of Following an outbreak of cholera across East the European Union are permitted to export but Africa, testing was extended to all fish and to cover are subject to higher rates of border inspection. Vibrio cholerae and Vibrio parahaemoliticus. These The current deadline for all countries to be fully requirements were lifted in June 1998. In March harmonized with the EU's hygiene standards is 1999 a suspected case of fish poisoning with pesti- December 31, 2005. cide was identified in Uganda. The European Kenya is an example of longer-term efforts to Union subsequently imposed a ban on exports of comply with the European Union's food safety Nile perch in April 1999 that was not lifted for requirements, overlaid with the need to overcome Kenya until December 2000. In each case, the restrictions on trade relating to immediate food impact was immediate. Exports declined, although safety concerns. Kenya's major fish export is Nile over time declines were partially offset by increased perch from Lake Victoria. Until the mid-1980s this sales to other markets. Fish-processing plants, most was a relatively minor species in the Lake Victoria already operating at less then 50 percent capacity, fishery, but with a shift in focus from local to export reduced their production, and some closed. In turn, markets, Nile perch came to account for more than the landed price of Nile perch fell. 90 percent of Kenya's exports of fish and fishery Both the Kenyan government and the private products by the mid-1990s, with a value of around sector tried to upgrade food safety controls. $44 million in 1996. Most exports were destined for Responsibility for regulatory controls was split the European Union. Through the 1980s there was between the Ministry of Health and the Fisheries significant investment in industrialized fish- Department of the Ministry of Agriculture and processing facilities, and 15 facilities were in opera- Rural Development, creating significant coordina- tion by the mid-1990s. At the landing beaches, tion problems. To improve compliance, the Fish- however, there was little or no change in fishing eries Department was made the sole "competent methods or marketing facilities. authority," and legislation was quickly revised in Initially, Nile perch exports were extremely prof- line with the European Union's requirements. itable. Processing capacity soon exceeded the sup- Fish-processing plants upgraded their facilities ply of fish, however, a situation that sets the com- and implemented an HACCP system, at an esti- petitive environment in which all levels of the chain mated average cost per plant of about $40,000 and operate. Although food safety requirements in their a total cost of $557,000. These costs were prohibi- major export markets were evolving, most proces- tive for several processing facilities, which closed, sors made little attempt to upgrade their facilities helping to reduce excess capacity. Simultaneously, and systems of procurement, processing, and mar- fish-processing companies began to cooperate to keting. Likewise, the legislative framework of food present a united voice to the government and Euro- safety controls remained largely unchanged, despite pean Commission. The Kenya Fish Processors and the fact that the structure and focus of the supply Exporters Association (AFIPEK), formed in 2000, chain had shifted to exports. The picture was of a has developed a code of good manufacturing prac- supply chain that had not been upgraded in line tice for the sector. with the growth in exports and was unable to A remaining weakness in the Nile perch supply implement effective controls within the context of chain is standards of hygiene at landing beaches. rapidly evolving standards overseas. Thus, both the Most attempts by the government to implement public authorities and exporters were in the posi- effective management of the fishery resource and tion of continuous problem solving. marketing arrangements have failed. Only recently In recent years exporters of Nile perch have have efforts been made to provide toilets, paved faced a catalogue of restrictions on trade with the and fenced landing areas, potable water, and cov- European Union. In 1996 salmonella was detected ered markets. This is the biggest compliance issue in a number of consignments of Nile perch from facing the sector in the short to medium term for Kenya (and Tanzania and Uganda) at the Spanish access to EU markets. border, and Spain immediately prohibited imports. The efforts of the Kenyan government and pri- In April 1997 the European Commission intro- vate sector eventually paid off, and in December duced a requirement for salmonella testing of all 2003 the European Commission recognized the Agro-Food Exports from Developing Countries: The Challenges Posed by Standards 107 controls in place as equivalent to those in the in 77 countries to protect consumers (Egmond European Union. The European Union's hygiene 1999). There are wide differences in national stan- requirements for fish and fishery products have had dards, however, linked to different susceptibilities profound effects on the Nile perch sector in Kenya. and different perceptions of acceptable health risks. Whereas the export supply chain had developed For example, acceptable tolerances for aflatoxin in with a sole focus on EU markets, today most food range from zero to 50 parts per billion. exporters have diversified their export base and There are indications that mycotoxin problems have major markets in Australia, Japan, and the have disrupted developing-country trade. Thailand United States. Compliance with EU requirements was once a leading world exporter of corn. Because helped Kenyan exporters to access and maintain of persistent aflatoxin problems, however, Thai these markets. corn regularly sold at a discount, costing the coun- This case illustrates the significant impact that try an estimated $50 million a year in reduced stricter food safety requirements can have on a sup- export revenue.19 Similarly, India was historically a ply chain that is almost entirely export oriented and significant supplier of peanut meal to the European dependent on a single market. It also demonstrates Union, but this trade declined sharply in the early how such requirements can exacerbate pressures 1980s because of problems meeting stricter stan- for restructuring and reform, while prevailing sup- dards for aflatoxin. Otsuki, Wilson, and Sewadeh ply and capacity issues constrain how various levels (2001), in a widely cited study, examine the process of the chain are able to respond. The case also illus- of harmonization of the European Union's stan- trates the interdependencies between levels of the dards for aflatoxin and the potential impact on supply chain and between the public and private exports of selected products, including cereals and sectors in meeting food safety requirements in dried fruit and nuts, from African countries.20 In export markets. And it demonstrates the impor- 1997 the European Union proposed a set of harmo- tance of responding quickly to emerging food nized standards for aflatoxin for member states, safety and agricultural health standards. The peri- which had developed their own standards, and ods of restrictions faced by Kenyan exporters of a uniform sampling procedure for testing. In Nile perch very much reflect the fact that little had response to the European Union's notification to been done in response to the implementation of the WTO, developing countries raised a series of stricter food safety requirements in the country's objections to the proposed standards and sampling most important export market. Rather, most of the methods. The proposed standards were to be far concerted effort to comply with these requirements more stringent than the proposed Codex standard, was stimulated by the sudden loss of market access, without proper scientific justification.21 in very much a crisis management mode of Otsuki, Wilson, and Sewadeh (2001) argue that operation. these standards are unnecessarily stringent given the estimated risk reduction that would be achieved. Their work is widely cited for its econo- Limits on Mycotoxins as Trade Barriers metric estimation of the potential loss of African Mycotoxins are toxic by-products of mold infesta- trade that could be attributed to the change in the tions, affecting as much as a quarter of global food European Union's standard. Using a gravity model, and feed crop output (Dohlman 2003; Reddy and which incorporates a number of variables assumed others 2002). They commonly occur in the produc- to affect bilateral trade flows, they compare existing tion of corn, wheat, and peanuts, causing consider- levels of African exports to the European Union able crop losses (Bhat and Vasanthi 1999). Their with likely levels following implementation of the incidence is affected by weather and insect infesta- new standards and likely levels had the European tion, although proper production and postharvest Union adopted the Codex standard (15 parts per (especially storage) practices can strongly mitigate billion) across all product categories. They estimate occurrence.18 Consuming foods with very high lev- that annual African exports to the European Union els of mycotoxins can be fatal, and long-term con- of cereals and nuts and dried fruit would decline sumption of foods with lower levels has been linked from $770 million to $372 million following adop- to liver cancer. Since the discovery of mycotoxins in tion of the EU standard but would rise to slightly the 1960s, regulatory limits have been established more than $1 billion under the Codex standard. 108 Global Agricultural Trade and Developing Countries Hence, the decision by the European Union to What about groundnuts? Africa's groundnut adopt the more stringent standards was estimated exports are dominated by South Africa, although to have cost Africa some $667 million. Egypt, The Gambia, Sudan, and Senegal have main- The conclusions of this work were headline tained small exports of confectionery nuts. Various grabbing but widely misinterpreted. Many subse- supply-side constraints have inhibited the competi- quent commentators have mistakenly referred to tiveness of many African countries in the interna- the estimates as if they were actual losses rather tional market for groundnuts (see chapter 12).24 In than the results of an econometric simulation. Sev- 2002 South Africa had 12 consignments of ground- eral shortcomings of this method need to be taken nuts rejected by EU member states because of afla- into account when interpreting the results.22 The toxin. Only 3 of the 12 would have met the less major focus here relates to the value of exports stringent Codex standard or the standards applied before and after the adoption of the standard previously by the individual member states. The and the lessons that stakeholders take from the rejected consignments were returned to South example. Africa, presumably for sale elsewhere, rather than The trade data used to establish the baseline put destroyed. Probably a few hundred thousand dol- African exports to the European Union (in 1998) at lars of business was affected, although the probable $472 million for dried fruit and nuts and $298 mil- sale of these nuts in other markets would have sub- lion for cereals, with the bulk of this trade occurring stantially mitigated these losses. No evidence was with France. These figures seem implausible, espe- found that Africa's limited exports to the European cially for cereals, given Africa's lack of competitive- Union of either cereals or tree nuts have been ness in this sector relative to Europe. Statistics from adversely affected since the adoption of the new the United Nations COMTRADE database show standards. Thus the near-term loss of African trade much lower European imports from Africa in 1998 because of the more stringent EU standards has of $104 million for dried fruit, $45 million for likely been in the hundreds of thousands of dollars groundnuts, $27 million for other edible nuts, and rather than in the hundreds of millions. less than $14 million for cereals and cereal prod- While the case for significant African trade losses ucts.23 This suggests that the baseline against which is weak, compliance with the EU aflatoxin standards the impact of the standards should have been remains a challenge for some developing countries. assessed was $190 million (c.i.f.--cost, insurance, Between 2000 and 2002 the number of border rejec- and freight--value) rather than $770 million. tions of nuts, nut products, and other snacks What about the evidence on impact? Most of the increased threefold (from 92 to 251). In 2002 some region's dried fruit trade is accounted for by two 235 consignments of nuts and dried fruit were North African countries--Tunisia and Algeria-- rejected on grounds of excessive levels of aflatoxin. whose exceptionally dry climate contributes to a Most of the rejected shipments were from Turkey very low incidence of aflatoxin. The only other (77 cases involving hazelnuts and dried fruit), Brazil African country with any history and recent (51 cases, mainly Brazil nuts), and Iran (50 cases, strength in exports of dried fruit is South Africa. mainly pistachios). Other countries with more than The new EU standards came into full force in April a few rejections were China (18), South Africa (12), 2002. Both in the year proceeding and the year fol- the United States (7), and Argentina (5). lowing that date there were no cases of dried fruit Although the data are incomplete, the EU notifi- consignments from Africa being detained on entry cations and alerts database reports the actual test to the European Union. In fact, while total EU results for levels of aflatoxin for many months. In imports of dried fruit declined somewhat in 2002, most cases of rejection, the measured levels of afla- imports from Africa increased, boosting Africa's toxin are substantially higher (sometimes many share from 9.8 percent in 2001 to 10.3 percent in times higher) than the Codex standard and also sig- 2002. Competing countries with more humid con- nificantly above the domestic standards of the ditions (especially Turkey) incurred higher levels of exporting countries. For example, of the 15 nut and product rejections during 2003. For dried fruit, the dried fruit consignments rejected in January 2002, more stringent EU standards and enforcement at only 3 were above the EU standard but below the the border worked to the competitive advantage of Codex standard. In October 2002, one country Africa's leading suppliers. source of nuts had 38 individual consignments Agro-Food Exports from Developing Countries: The Challenges Posed by Standards 109 rejected, 15 involving aflatoxin levels of 100 parts raspberries from Guatemala to the United States per billion or more.25 This suggests that suppliers, during months when U.S. domestic supplies were especially those producing in humid conditions, limited (Calvin 2003; Calvin, Flores, and Foster are having considerable difficulty controlling afla- 2003). By 1996 these exports had reached $3 mil- toxin contamination. lion, with some 85 growers participating. In that It is still unclear, however, how much EU year, however, the U.S. Centers for Disease Control standards on aflatoxin have affected developing- and Prevention and Health Canada received country trade. For example, Iran has been experi- reports of some 1,465 cases of food-borne illness encing problems with aflatoxin for several years. Its associated with the parasite Cyclospora. Raspberries exports of edible nuts declined from $452 million from Guatemala were identified as the most likely in 1996 to less than $210 million in 2002. Further source of the contamination. analysis is needed to determine how much of this The FDA sent a team to Guatemala to investi- decline can be attributed to problems with afla- gate, amid considerable scientific uncertainty and toxin contamination and of that, how much to reg- great difficulty identifying the likely source of the ulatory measures rather than to a more general loss contamination. The association of Guatemalan of buyer confidence. Exports from Turkey, however, growers (GBC) remained unconvinced that its seem to have been little affected by the increased raspberries were the source of the problem. It stringency of EU standards and enforcement. In attempted to put in place a limited program to 2002 the volume of products rejected by the screen out potentially high-risk farms, but the pro- European Union constituted less than 1 percent of gram had no effective enforcement mechanism. Turkish exports of nuts and dried fruit to that mar- After another large outbreak of Cyclospora-related ket. Any rejected product is reexported to countries illnesses in the spring of 1997, the GBC voluntarily with less strict standards (or enforcement) or sold agreed to stop exports of raspberries to the United domestically, reducing losses. States. Despite the fact that the Guatemalan gov- ernment created a food safety commission with enforcement powers in late 1997, the FDA was Proliferation of Horticultural Product Standards unconvinced and essentially imposed an import The regulatory and private governance systems for ban on Guatemalan raspberries. international fresh produce markets are becoming During the next two years many organizations increasingly complex. This changing regulatory in the United States and Canada worked with the environment appears to be raising the bar for new Guatemalans to solve the problem. A Model Plan of entrants and throwing new challenges in the path of Excellence was put in place in 1999, involving the existing developing-country suppliers. Concern is application of food safety practices by growers, mounting about the ability of small and low-income mandatory inspection by government, and a sys- countries to meet rising public and private stan- tem of product traceability back to individual dards and thus their ability to remain competitive growers. The United States lifted the ban on in international fresh produce markets (Dolan and imports of Guatemalan raspberries. In 2000, how- Humphrey 2000; Chan and King 2000; Buurma and ever, there were two further Cyclospora outbreaks, others 2001). High-profile food scares and highly which were traced back to a single Guatemalan publicized instances of pesticide residue violations farm. The grower was removed from the program, have created an impression of extreme vulnerability and there have been no further outbreaks. of developing-country suppliers. Yet experiences are While the Model Plan of Excellence was techni- mixed, and most countries and industries that have cally successful, it came too late to save the industry. run into standards-related barriers have also been Facing consumer concerns, several supermarkets in struggling with other supply-chain problems that the United States sought alternative sources of have inhibited their profitability and competitive- raspberries. Recognizing the enormous challenge of ness. Consider the contrasting experiences of two rehabilitating the reputation of Guatemalan rasp- low-income countries, Guatemala and Kenya. berries in the eyes of both consumers and distribu- tors, several leading firms (both Guatemalan and Guatemalan raspberries: a cautionary tale? In international firms) shifted their operations to the late 1980s several firms began exporting Mexico. By 2001 only four growers of raspberries 110 Global Agricultural Trade and Developing Countries for export remained in Guatemala, with exports of For years the industry functioned with simple less than $200,000. Meanwhile, Mexico's exports of supply chains, involving little investment in infra- raspberries grew from $2.9 million in 1998 to $8.9 structure, product development, or management million in 2002 and now account for the largest systems. Around a dozen medium-size firms plus share of an expanding U.S. import market. large numbers of small, part-time operators han- Although the Guatemalan raspberry industry dled the exports, frequently trading with relatives never recovered, other parts of the fresh produce or similar small-scale companies in Europe. Fresh industry built on the institutional capacity building produce was purchased from large numbers of that had taken place. For example, the inspection growers. Produce was generally collected in card- agency, the Integrated Program for Agricultural board boxes from farms or along roadsides and and Environmental Protection (PIPAA), has been brought to a central warehouse, sifted through and working closely with local blackberry growers, a regraded if necessary, cooled a little, and trucked to leading local supermarket chain, and others to the airport for evening shipment. Ministry of Agri- enhance food safety management systems. PIPAA culture officials at the airport conducted limited is also collaborating with the Animal and Plant inspections. This was the model from the 1960s to Health Inspection Service and the FDA in the the mid-1980s. The industry remained competitive United States on a program for Guatemalan in some markets and for some products, but not for exports of mangoes and papayas. others. The Kenyan fresh produce export trade Calvin, Flores, and Foster (2003) draw several grew slightly in the 1970s but stagnated in the lessons from Guatemala's experience. Delays in 1980s. addressing food safety and agricultural health Since the early 1990s the industry has been problems may hurt an industry's exports and repu- reshaped and transformed in response to--and in tation. An effective traceability system allows focus- anticipation of--commercial, regulatory, and pri- ing on particular growers or exporters rather than vate governance changes within its core external needing to enhance standards in an entire industry. markets. Commercial pressures came from satu- Finally, strong grower organizations can improve rated markets for certain products and increased an industry's ability to respond to food safety competition from suppliers that had improved challenges. There is also a wider lesson from their supply capabilities and had less expensive sea Guatemala's experience. Small countries and niche or air-freight costs than did Kenya. Commercial products are probably far more vulnerable to loss changes within Europe also required a shift in the of markets and collapse of reputation in the face of Kenyan approach. In many countries large super- food safety problems than are larger countries and market chains were in ascendancy while wholesale more mainstream or generic products. Both inter- markets were declining. Consolidation was also national buyers and consumers would likely be occurring among importers, packers, and distribu- more tolerant and patient with core, long-standing tors. The growing segments of the fresh produce suppliers that have established a "brand" in which market were being managed by fewer players. On they have confidence. the regulatory front a steady wave of activity was geared toward strengthening and harmonizing EU Kenyan fresh produce exports: some success. and member state regulations and monitoring sys- Kenya's fresh produce trade dates to the mid-1950s, tems for food safety, quality conformity, and plant when small quantities of temperate-climate vegeta- health. Also emerging were progressively refined bles and tropical fruits were supplied in the Euro- private-sector standards or codes of practice gov- pean winter off-season to up-market stores in erning food safety, plant health, and other issues. London.26 This off-season trade was later joined by Several leading Kenyan exporters caught an year-round supplies of high-quality green beans early glimpse of this new fresh produce environ- and a broad array of vegetables that are part of the ment and began to reorient their operations. With traditional diets of UK immigrant populations the encouragement of several UK supermarkets, from South Asia. Most of the products were air- they began to experiment with new crops, new con- freighted in two-kilogram boxes for sale through sumer packaging, and new combinations of vegeta- wholesale markets or to distributors and caterers. bles. An increasing proportion of products was Agro-Food Exports from Developing Countries: The Challenges Posed by Standards 111 directed to selected supermarket chains, which dards have become an increasingly important began to send "audit" teams to Kenya to check influence on the international competitiveness of hygiene and other conditions on farms and pack- developing countries, especially for high-value houses. Improvements and investments were rec- agricultural and food products. ommended and in some cases required. With The evidence presented in this chapter, while renewed confidence in the future of the industry, admittedly incomplete, suggests that the picture for several exporters invested heavily in new or developing countries as a whole is much less upgraded pack-houses and related food safety pessimistic than that widely presented by the management systems for packing ready-to-eat and standards-as-barriers perspective. Indeed, rising semi-prepared products. Mixed salads, stir-fry standards accentuate underlying supply-chain mixes, vegetable kebabs, and other value-added strengths and weaknesses and thus affect the com- products now account for more than 40 percent of petitive positions of countries and distinct market what has been a burgeoning trade over the past participants, making it important to view the decade. Between 1991 and 2003 Kenya's fresh effects of food safety and agricultural health meas- vegetable exports increased from $23 million to ures in the context of wider capacity constraints. $140 million.27 The key question for developing countries is how to Rising public- and private-sector standards have exploit their strengths and overcome their weak- posed challenges to the Kenyan fresh produce nesses to emerge as gainers rather than losers. industry, yet they have also thrown a lifeline to the Still, by raising the bar for new entrants and industry. Because of Kenya's location and relatively placing a premium on effective safety management high air-freight costs, its fresh produce sector and logistical coordination, higher official and pri- cannot compete with many other players on a unit- vate standards can weaken the competitive position cost basis. Margins have been squeezed in the mar- of small and poorer countries and the ability of ket for mainstream, commodity-type vegetables. small enterprises and farmers to remain active and With rising labor costs in Europe, however, the profitable in export supply chains. But food safety Kenyan industry has positioned itself as a slicer, and agricultural health standards are here to stay, dicer, and salad-maker, all labor-intensive func- and there is no slowing down their rate of change tions. Thus far, this market segment has grown or applying for special and differential treatment. fastest in the United Kingdom, although there is Much of the impetus for standards comes from increased buyer interest and consumer demand on consumer and commercial interests, magnified the European continent as well. This suggests that by advances in technology and added security well-organized industries in low-income countries concerns. can use stricter standards as a catalyst for change-- The answer for developing countries is to and profit in the process. develop and improve food safety and agricultural health management systems. This requires simulta- neous attention to legal systems, human capital, and Conclusions physical infrastructure, among other things. Man- There are now a number of documented cases in agement capacity is required not only to comply which developing countries have faced restrictions with different requirements in different markets, because of their inability to meet food safety or but also to demonstrate compliance with standards. agricultural health requirements. In some of these, Although many countries have struggled to meet well-established export-dependent sectors have ever stricter standards, even some very poor coun- been compromised by the implementation of new, tries have managed to implement the necessary stricter standards, with negative repercussions for capacity. This has most commonly occurred where the livelihoods of those involved. At the same time, the private sector is well organized and the public other countries have managed to gain access to sector is well focused and supports the efforts of high-value markets in industrial countries despite exporters. To meet the challenges posed by stan- the exacting standards. Clearly, the situation is not dards in international markets for high-value agri- as black and white as some commentators suggest. cultural and food products, developing countries What cannot be disputed, however, is that stan- need institutional frameworks to help them 112 Global Agricultural Trade and Developing Countries overcome the problems associated with being poor other business functions and is thus a joint product with those or small. These can include outgrower programs for functions. Thus, there are questions over what investments and which management systems are put in place strictly to ensure smallholder farmers, systems of training and over- compliance with particular standards and which service a multi- sight for smaller enterprises organized through plicity of functions. In practice, it is often difficult to make this associations and other groups, and twinning and separation. For example, cold-store facilities may be needed to prevent the multiplication of bacteria in fresh produce, yet such regional networking for small countries. facilities are also critical for achieving a quality characteristic or An overarching message is the need for develop- extended shelf-life. ing countries (and their exporters) to be proactive 8. One leading Kenyan firm estimated that the costs of its small farmer oversight arrangements represented about 12 per- on food safety and agricultural health issues. It is cent of its costs of raw materials. These transaction costs repre- important not to be pushed into action by a major sent 6 percent of the f.o.b. (free on board) value of French beans, crisis. By thinking strategically, countries, produc- which is equivalent to the exporter's profit on the product and ers, and exporters can program capacity enhance- about 60 percent of the grower's profit. 9. Some new products may not require as much farm ment into wider and longer term efforts to enhance labor as previously traded products or may require more capital domestic food safety and agricultural health man- investment. In either scenario the comparative advantage of agement systems and export competitiveness. The smallholders may be reduced. 10. Data for the United States can be found at www.fda.gov/ alternative is that large investments will be required ora/oasis/ora_oasis_ref.html, and for the European Union at over a long period just to "put out fires." In all of www.europa.eu.int/comm/food/fs/sfp/ras_index_en.html. The this, there is a need for the public and private sec- data exclude certain agricultural and food products for which the FDA has no jurisdiction, most notably meat and poultry. tors to work together to identify the most efficient Until 2002 these data referred to border detentions regardless of and effective ways to develop capacity, viewing food whether the product was eventually permitted to enter. Since safety and agricultural health controls as a collabo- then they have recorded border rejections. The European Union rative effort. has made disaggregated data on import alerts available only since 2002, although annual reports with broad summary statis- tics were published previously. Notes 11. Between 2002 and 2003 the number of ports at which the FDA has assigned inspection staff increased from 40 to 90. Dur- 1. For the more traditional food exports of developing coun- ing this period, a $96 million increase in the FDA's budget for tries, such as beverage crops, fiber crops, tobacco, and sugar, food security work enabled it to hire 655 new field personnel. In international trade is still largely governed by price and quality the Bush administration's proposed 2005 budget, the FDA and by traditional forms of trade protection and preferences (see would receive a 9 percent increase in funding to expand its "food chapters 3 and 4). defense" program. The fiscal 2005 budget calls for 97,000 import 2. According to a source at the Food and Drug Administra- inspections, seven times the number undertaken in 2001. Simi- tion, while 99 percent of domestic facilities are found to be in larly large increases were proposed for the Department of compliance, some 30 percent of inspected foreign facilities have Agriculture's work on food safety. significant system defects. 12. To put this number into perspective, the estimated total 3. See www.cfsan.fda.gov and europa.eu.int/comm/food/fs/ costs to the United Kingdom alone from BSE-related market inspections/fnaoi/reports/annual_eu/index_en.html. losses and for the various cull and disposal schemes was more 4. For example, as of August 2003, two countries--the than $5 billion (Mathews, Bernstein, and Buzby 2003). This does Netherlands and the United Kingdom--accounted for more not take any account of the adverse impact on the country's than two-thirds of the area certified as being EUREPGAP- tourism industry. compliant. EUREPGAP is a set of good agricultural practices 13. For example, the United States currently permits imports (GAP) based on accepted standards and promoted by the Euro- of beef from only 33 countries and imports of chicken from only pean retailer produce group (EUREP). Only a small proportion 4 countries. of the area in developing countries on which fresh produce is 14. Currently, the International Office of Epizooties recog- grown for the European market was so certified, the bulk of it in nizes only 57 countries as being totally free of foot and mouth South Africa. Recognizing these constraints, extended deadlines disease without vaccination, of which 26 are developing coun- have been given to many developing-country exporters and pro- tries--only 3 of them low-income countries. For further infor- ducers to adopt and gain certification against the EUREPGAP mation, see www.oie.int. protocol. 15. Indeed, more widespread cases of both new and well- 5. Only 20 percent of the notifications by low-income coun- established animal diseases have led to heightened concerns tries and 22 percent of those of high-income countries involved about the role of international trade in the spread of such dis- applications of international standards. eases. In the case of BSE, widespread restrictions have been 6. An array of capacity assessment instruments are used to applied to trade in live animals, meat, animal feed, and an array gauge strengths and weaknesses of food safety and agricultural of by-products used in the cosmetics, pharmaceutical, and other health management capacity. Some instruments focus on spe- industries. cific dimensions of capacity, while others provide a broader 16. For the case studies produced by UNCTAD, see overview. r0.unctad.org/trade_env/test1/openF1.htm. For the case studies 7. In practice, it is rather difficult to measure "costs of com- produced by UNEP, see www.unep.ch/etu/publications/ pliance." Food safety is very often achieved in combination with Ctry_studies.htm. Agro-Food Exports from Developing Countries: The Challenges Posed by Standards 113 17. The European Commission has presented its controls on the industry has transformed itself. Some 25 smaller exporters hygiene for imports of fish and fishery products as a practical lack the financial resources to invest in modern pack-houses and example of the application of equivalence (WTO 2001). Thus, continue to supply loose produce to commission agents and rather than laying down specific requirements, the European others in European wholesale markets and the Middle East. Commission focuses on the conditions under which products will be equivalent to those produced in the European Union. 18. Although the growth of mycotoxin-producing molds is References an endemic problem in humid areas, management of this prob- lem need not involve very sophisticated or costly measures. See Beghin, J., and J. C. Bureau. 2001. "Quantitative Policy Analysis Boutrif (1997), Park, Njapau, and Boutrif (1999), and of Sanitary, Phytosanitary and Technical Barriers to Trade." Dimanche and Kane (2002) for examples of practical and low- Economie Internationale 87(3): 107­30. cost measures. Bhat, Ramesh, and S. Vasanthi. 1999. "Mycotoxin Contamina- 19. There, most of the problem occurred during postharvest tion of Foods and Feeds." Paper presented at the Third Joint as the harvested maize was typically stored in moist if not wet FAO/WHO/UNEP International Conference on Mycotox- conditions for one to two months before sale and processing ins, March 3­6, Tunis. (Tangthirasunan, T. n.d.). Boutrif, E. 1997. "Aflatoxin Prevention Programmes." Food and 20. This is probably the most widely cited study on the poten- Agriculture Organization, Rome. tial or actual impact of rising food safety standards on exports of Bureau, J. C., S. Marette, and A. Schiavina. 1998. "Non-Tariff agricultural and food products from developing countries. Trade Barriers and Consumers' Information: The Case of the 21. In response to objections, the European Union revised EU­US Trade Dispute over Beef." European Review of Agri- some of its proposed measures. In its 1998 Directive, it estab- cultural Economics 25(4): 437­62. lished a limit for total aflatoxin in groundnuts subject to further Buurma, J. S., M. J. B. Mengelers, A. J. Smelt, and E. Muller. 2001. processing at 15 parts per billion, and a limit for aflatoxin B1 at 8 "Developing Countries and Products Affected by Setting parts per billion, which was consistent with the proposed Codex New Maximum Residue Limits of Pesticides in the EU." standard. For other nuts and dried fruit subject to further pro- Agricultural Economics Research Institute, The Hague. cessing, more stringent limits were set at 10 parts per billion for Buzby, J., ed. 2003. International Trade and Food Safety: Economic total aflatoxin and 5 parts per billion for aflatoxin B1. There was Theory and Case Studies. Agricultural Economic Report 828. no equivalent Codex standard. The strictest standard was set for U.S. Department of Agriculture. Washington, D.C. cereals, dried fruits and nuts intended directly for human con- Calvin, L. 2003. "Produce, Food Safety, and International Trade: sumption with maximum levels of 4 parts per billion for total Response to U.S. Foodborne Illness Outbreaks Associated aflatoxin and 2 parts per billion for aflatoxin B1. Again, there with Imported Produce." In J. Buzby, ed., International Trade was no equivalent Codex standard. and Food Safety: Economic Theory and Case Studies. Agricul- 22. See, for example the discussions about gravity models tural Economic Report 828. U.S. Department of Agriculture. and other approaches to estimating the trade impacts of stan- Washington, D.C. dards in Beghin and Bureau (2001), OECD (2003), and Wilson Calvin, L., L. Flores, and W. Foster. 2003. "Case Study: (2003). Guatemalan Raspberries and Cyclospora." In L. Unnevehr, 23. In that year, African exports of cereals totaled $105 mil- ed., Food Safety in Food Security and Food Trade. Washing- lion, with Egypt accounting for $70 million. The vast majority of ton, D.C.: International Food Policy Research Institute. this trade was conducted with countries of the Near East and Cato, J., and C. Lima dos Santos. 2000. "Costs to Upgrade the Middle East. Bangladesh Frozen Shrimp Processing Sector to Adequate 24. In the 1960s and 1970s Africa was a major world supplier Technical and Safety Standards and to Maintain a HACCP of groundnuts, with large exporters in Malawi, Nigeria, Senegal, Program." In L. Unnevehr, ed., HACCP: New Studies of Costs and other countries. For reasons unrelated to aflatoxin, these and Benefits. St. Paul: Eagen Press. exports lost their international competitiveness, and most pro- Cato, J., S. Otwell, and A. Coze. 2003. "Nicaragua's Shrimp Sub- duction went to serve domestic markets or for use in oil crush- sector: Developing a Production Capacity and Export Mar- ing. Over the years, research activity and the commercial trade in ket during Rapidly Changing Worldwide Safety and Quality Africa moved away from confectionery-type varieties preferred Regulations." Case study prepared as part of a World Bank in world markets, and recent attempts to revive confectionery program on The Challenges and Opportunities Associated nut exports have encountered major problems attributable to with International Agro-Food Standards. Washington, D.C. inadequate seed, basic quality control and price incentives for Chan, M., and B. King. 2000. "Review of the Implications of farmers. Changes in EU Pesticides Legislation on the Production and 25. Moonen (2004) reports on testing results from the Export of Fruits and Vegetables from Developing Country Dutch import control authority. It is common that groundnut Suppliers." Natural Resources and Ethical Trade Programme, shipments from developing countries have levels of aflatoxin London. contamination of between 50 and 800 parts per billion. Also Crain, W., and J. Johnson. 2001. "Compliance Costs of Federal cited by Moonen are toxicological surveys in Senegal for Workplace Regulations: Survey Results from U.S. Manufac- groundnuts sold in the domestic market. Some 90 percent of turers." George Mason University, Regulatory Studies Pro- sampled groundnuts were contaminated with aflatoxin with the gram, Arlington, Va. average level being 230 parts per billion. Delgado, C. L., N. Wada, M. W. Rosegrant, S. Meijer, and M. 26. The discussion in this section draws on Jaffee (2003). Ahmed. 2003. Fish to 2020: Supply and Demand in Changing 27. Systems for crop procurement have also been trans- Global Markets. Washington, D.C.: International Food Policy formed, with many leading companies investing in their own Research Institute. farms or inducing changes in the practices of outgrowers. There Dimanche, P., and A. Kane. 2002. "Senegal's Confectionery has been an array of joint public-private initiatives to train Peanut Supply Chain: The Challenge of Controlling growers in all aspects of good agricultural practice. But not all of Aflatoxin Levels." In Food Safety Management in Developing 114 Global Agricultural Trade and Developing Countries Countries. Proceedings of the International Workshop, Park, D., H. Njapau, and E. Boutrif. 1999. "Minimizing Risks CIRAD-FAO, December 11­13, Montpellier, France. Posed by Mycotoxins Utilizing the HACCP Concept." Paper Dohlman, E. 2003. "Mycotoxin Hazards and Regulations: presented at the Third Joint FAO/WHO/NEP International Impacts on Food and Animal Feed Crop Trade." In J. Buzby, Conference on Mycotoxins, March 3­6, Tunis. ed., International Trade and Food Safety: Economic Theory Pauwelyn, J. 1999. "The WTO Agreement on Sanitary and Phy- and Case Studies. Agricultural Economic Report 828. U.S. tosanitary (SPS) Measures as Applied in the First Three SPS Department of Agriculture. Washington, D.C. Disputes." Journal of International Economic Law 2(94): Dolan, C., and J. Humphrey. 2000. "Governance and Trade in 641­49. Fresh Vegetables: The Impact of UK Supermarkets on the Perry, B. D., T. F. Randolph, S. Ashley, R. Chimedza, T. Forman, African Horticulture Industry." Journal of Development Stud- J. Morrison, C. Poulton, L. Sibanda, C. Stevens, N. Tebele, ies 37(2): 147­76. and I. Yngstrom. 2003. "The Impact and Poverty Reduction Dong, F., and H. Jensen. 2004."The Challenge of Conforming to Implications of Foot and Mouth Disease Control in Sanitary and Phytosanitary Measures for China's Agricul- Southern Africa." International Livestock Research Institute, tural Exports," MATRIC Working Paper 04-MWP 8, Iowa Nairobi. State University, Ames, Iowa. Reardon, T., and J. A. Berdegue. 2002. "The Rapid Rise of Egmond, Hans. 1999. "Worldwide Regulations for Mycotoxins." Supermarkets in Latin America: Challenges and Opportuni- Paper presented at the Third Joint FAO/WHO/UNEP Inter- ties for Development." Development Policy Review 20(4): national Conference on Mycotoxins, March 3­6, Tunis. 371­88. El-Tawil, A. 2002. "An In-Depth Study of the Problems by the Reddy, D., K. Thirumala-Devi, S. Reddy, F. Waliyar, M. Mayo, K. Standardizers and Other Stakeholders from Developing Devi, R. Ortiz, and J. Lenne. 2002. "Estimation of Aflatoxin Countries." Paper presented at the ISO/WTO regional work- Levels in Selected Foods and Feeds in India." In Food Safety shops: Part 1. International Organization for Standardiza- Management in Developing Countries, Proceedings of the tion, Geneva. International Workshop, CIRAD-FAO, December 11­13, Henson, S. J. 2001."Appropriate Level of Protection: A European Montpellier, France. Perspective." In K. Anderson, C. McRae, and D. Wilson, eds., Roberts, D. 2004. "The Multilateral Governance Framework for The Economics of Quarantine and the SPS Agreement. Uni- Sanitary and Phytosanitary Regulations: Challenges and versity of Adelaide, Centre for International Trade Studies, Prospects." Paper prepared for a World Bank training semi- Australia. nar on Standards and Trade, January 27­28, Washington, ------. 2004. "National Laws, Regulations, and Institutional D.C. Capabilities for Standards Development." Paper prepared for Roberts, D., and B. Krissoff. 2003. "The WTO Agreement on the a World Bank training seminar on Standards and Trade, Jan- Application of Sanitary and Phytosanitary Barriers." U.S. uary 27­28, Washington, D.C. Department of Agriculture, Economic Research Service, Henson, S. J., and W. Mitullah. 2004. "Kenyan Exports of Nile Washington, D.C. Perch: Impact of Food Safety Standards on an Export- Roberts, D., and D. Orden. 1997. "Determinants of Technical Oriented Supply Chain." Case study for a World Bank pro- Barriers to Trade: The Case of U.S. Phytosanitary Restric- gram on the Challenges and Opportunities Associated with tions on Mexican Avocados, 1972­1995." In D. Orden and D. International Agro-Food Standards. Policy Research Work- Roberts, eds., Understanding Technical Barriers to Trade. ing Paper 3349. World Bank, Washington, D.C. Minneapolis-St. Paul: University of Minnesota, Interna- Jaffee, S. 2003. From Challenge to Opportunity: Transforming tional Agricultural Trade Research Consortium. Kenya's Fresh Vegetable Trade in the Context of Emerging Food Roberts, D., T. Josling, and D. Orden. 1999. "A Framework for Safety and Other Standards in Europe. Agriculture and Rural Analyzing Technical Trade Barriers in Agricultural Markets." Development Discussion Paper 1. World Bank, Washington, U.S. Department of Agriculture, Economic Research Service. D.C. Washington, D.C. Lamb, J., J. Velez, and R. Barclay. 2004. "The Challenge of Com- Sumner, D. A., ed. 2003. Exotic Pests and Diseases: Biology and pliance with SPS and Other Standards Associated with the Economics for Biosecurity. Ames, Iowa: Iowa State Press. Export of Shrimp and Selected Fresh Produce Items to the Tangthirasunan, T. n.d. "Mycotoxin Economic Aspects." Food United States Market." Paper submitted to the World Bank. and Agriculture Organization. Rome. Washington, D.C. Unnevehr, L. 2003. "Food Safety in Food Security and Food Mathews, K, J. Bernstein, and J. Buzby. 2003. "International Trade: Overview." In L. Unnevehr, ed., Food Safety in Food Trade of Meat/Poultry Products and Food Safety Issues." In Security and Food Trade. Washington, D.C.: International J. Buzby, ed., International Trade and Food Safety: Economic Food Policy Research Institute. Theory and Case Studies. Agricultural Economic Report 828. Unnevehr, L., and N. Hirschhorn. 2001. "Designing Effective U.S. Department of Agriculture, Washington, D.C. Food Safety Interventions in Developing Countries." World Moonen, I. 2004. "The Aflatoxin Case: Process and Impact Eval- Bank, Washington, D.C. uation of the EU Regulation on Aflatoxin." Paper submitted Wilson, J., and T. Otsuki. 2003. "Balancing Risk Reduction and to the Netherlands Ministry of Foreign Affairs, Policy Benefits from Trade in Setting Standards." In L. Unnevehr, Coherence Unit, The Hague. ed., Food Safety in Food Security and Food Trade. Washing- OECD (Organisation for Economic Co-operation and Develop- ton, D.C.: International Food Policy Research Institute. ment). 2003. "Trade Effects of the SPS Agreement." Joint Wilson, N. 2003. "A Review of Empirical Studies of the Trade Working Party on Agriculture and Trade, Paris. and Economic Effects of Food Safety Regulations." In A. Otsuki, T., J. Wilson, and M. Sewadeh. 2001. "Saving Two in a Velthuis, ed., New Approaches to Food Safety Economics. Billion: Quantifying the Trade Effect of European Food Boston: Kluwer Academic Publishers. Safety Standards on African Exports." Food Policy 26(5): WTO (World Trade Organization). 2001 "Equivalence: Note by 495­514. the Secretariat." G/SPS/W/111. Geneva. 7 GLOBAL AGRICULTURal Reform: WHAT IS AT STAKE? Dominique van der Mensbrugghe and John C. Beghin This chapter uses a global, dynamic applied general decomposes the aggregate results by looking at the equilibrium model (LINKAGE) to assess how the impacts of partial reforms--both regionally and multifarious trade and support policies in agricul- across instruments--to identify what share of the ture affect income, trade, and output patterns at the global gains derives from reform in industrial coun- global level.1 Such models have become a standard tries and what from reform in developing countries tool for assessing policy reforms because they cap- and what share is driven by border protection and ture linkages across sectors and regions (through what by domestic support. The second set of simu- trade) and because, by their nature, they have lations addresses issues raised by critics of trade adding-up constraints so that supply and demand reform--notably that the predicted gains for devel- are in equilibrium in all markets. The analysis pro- oping countries are too optimistic and that the tran- vides order-of-magnitude estimates of the poten- sition costs for industrial-country farmers are high tial consequences of policy changes, rather than and too often ignored. Concerns have also been a single point or "best" estimate. It also looks at raised about the ability of developing countries to the induced structural changes, including cross- respond to reforms and to achieve consistently high regional patterns of output and trade, which tend productivity gains. To answer the questions about to be much larger than the more familiar gains the impacts on developing countries, three assump- to real income. Whereas income gains typically tions are explored: the consequences of assuming amount to 1 percent of base income or less, struc- differential and lower agricultural productivity in tural changes--for example, in sectoral output or some developing countries, the impacts of con- trade--can be greater than 50 percent. straining output supply response in selected Two sets of simulations are used to create a low-income countries, and estimates of trade elas- deeper picture of what drives the key results.One set ticities. The chapter also assesses the impacts of With great appreciation from the authors, the underlying work in this chapter has benefited over several years from useful comments and suggestions from the following colleagues: Ataman Aksoy, Jonathan Brooks, Bill Cline, Uri Dadush, Bernard Hoekman, Jeff Lewis, Will Martin, John Nash, Richard Newfarmer, David Roland-Holst, Josef Schmidhuber, and Hans Timmer as well as participants at seminars at the OECD and the University of California at Berkeley. 115 116 Global Agricultural Trade and Developing Countries slower exit by industrial-country farmers and how ing countries include some of the large countries this would affect transition adjustments. that are important in agricultural markets as pro- Some of the main findings: ducers or as consumers (Argentina, Brazil, China, India, and Indonesia). The remaining developing · Reform of agricultural and food trade policy countries are grouped into regional aggregations.2 provides 70 percent of the global gains from The sectoral decomposition is concentrated in the merchandise trade reform--$265 billion of a agricultural and food sectors (15 of the 22 sectors). total of $385 billion. The LINKAGE model is dynamic, with scenarios · The global gains are shared roughly equally spanning 1997 to 2015. The dynamics include between industrial and developing countries, exogenously given labor and land growth rates, but developing countries gain significantly more savings-driven investment and capital accumula- as a share of initial income. Significant income tion, and exogenous productivity growth. Struc- gains occur in developing-country agriculture, tural changes over time are driven by differential where poverty tends to be concentrated. growth rates and supply and demand parameters. · Developing countries gain more from reforming Trade is modeled using the Armington assumption. their own support policies than from improved Goods are differentiated by region of origin using a market access in industrial countries. Likewise, two-nested structure (domestic absorption first industrial countries also gain relatively more allocated across domestic and aggregate import from their own reform. goods, then aggregate imports allocated across · Notwithstanding the overall benefits from different regions of origin). greater openness, structural changes are impor- tant and transition adjustments need to be Overview of Baseline Simulation addressed. · Productivity and supply assumptions affect Assessing the impacts of policy reforms requires impact assessment, but their influence is small, two steps in the dynamic framework of the and they do not alter the main aggregate LINKAGE model, a baseline (or reference) simula- findings. Trade elasticities, however, are key in tion and a reform simulation. The baseline involves determining the overall level of the income running the model forward from its 1997 base year gains. Higher elasticities dampen terms-of-trade to 2015, with exogenous assumptions about labor effects and increase trade and real income gains and population growth rates, productivity, and more than proportionally, while the opposite is demand behavior parameters including savings, true for lower elasticities. These effects can be which determines the rate of capital accumulation very large for individual countries. (adjusted exogenously for depreciation). The baseline simulation can also incorporate changes in base year policies--to take into account The Modeling Framework known changes in policies (between 1997 and the The LINKAGE model is based on a standard neo- present) or anticipated changes. The baseline classical general equilibrium model with firms described below assumes no changes in base year maximizing profit in competitive markets and con- policies, however; they are held at their 1997 levels. sumers maximizing well-being under a budget con- Thus the reform simulations reflect changes from straint. The model has added features related to its their 1997 levels, not changes that would be antici- dynamic nature. It is global, with the world decom- pated from 2004 levels.3 It is unclear in which posed into 23 regions, and multisectoral, with direction some past and anticipated changes would economic activity aggregated into 22 sectors (see affect the global trade reform results. Some changes annex A in the report on the CD-ROM). Seven of clearly reflect further opening--for example, the 23 regions are classified as high-income (or China's accession to the World Trade Organization industrial) including Canada, Western Europe (WTO) and some bilateral free trade agreements. (European Union-15 plus the European Free Trade Others would go in the opposite direction--for Association countries), Japan, and the United example, the changes to the U.S. farm support States--the so-called Quad countries. The develop- programs. Global Agricultural Reform: What Is at Stake? 117 Agriculture and Food Trends in the Baseline a much higher 2.9­3.4 percent in developing coun- Scenario tries (tables 7.1 and 7.2 summarize the results; tables on the CD-ROM provide details for individ- Trends in agriculture and food supply and demand ual countries). On a per capita basis there is more across the globe as determined in the baseline sce- demand growth in developing countries, largely nario are driven in part by the macroeconomic because of higher income elasticities for food. Thus environment (as described in annex B on the the baseline assumes that demand growth will be CD-ROM). But they are also driven by microeco- lower than output growth in industrial countries nomic assumptions about the mobility of factors, and higher than output growth in developing production technologies, income and price elastici- countries. ties, and trade elasticities, among others. With higher output growth than demand, For agriculture and food between 2000 and industrial countries will see an increase in their 2015, both demand and production grow at exportable surplus. On aggregate their net 1.0­1.2 percent a year in industrial countries and at TABLE 7.1 Trends in Agriculture, 2000­15 Average Annual Growth Net Trade (percent) (billions of 1997 US$) Country Grouping Output Demand Imports Exports 2000 2015 High-income countries 1.2 1.1 1.9 3.0 -24.3 -3.1 Low-income countries 3.6 3.5 4.4 5.5 9.9 21.6 Middle-income countries 3.2 3.3 8.3 5.4 14.4 -18.5 Low-income countries, 3.7 3.4 3.6 6.6 7.2 22.4 excluding India Middle-income countries, 3.2 3.4 8.3 5.1 17.1 -19.3 including India Developing countries 3.3 3.4 7.8 5.4 24.3 3.1 World total 2.6 2.6 4.4 4.2 0.0 0.0 Note: Net trade is measured at f.o.b. prices (imports exclude international trade and transport margins). Source: World Bank simulations with LINKAGE model, based on release 5.4 of the GTAP data. TABLE 7.2 Trends in Processed Foods, 2000­15 Average Annual Growth Net Trade (percent) (billions of 1997 US$) Country Grouping Output Demand Imports Exports 2000 2015 High-income countries 1.2 1.0 1.3 2.4 7.7 53.5 Low-income countries 3.3 3.4 3.9 2.2 3.6 1.8 Middle-income countries 2.9 3.1 4.5 2.0 -11.3 -55.3 Low-income countries, 3.1 3.3 3.8 2.2 1.8 -0.2 excluding India Middle-income countries, 2.9 3.1 4.5 2.0 -9.5 -53.4 including India Developing countries 2.9 3.2 4.5 2.1 -7.7 -53.5 World total 1.8 1.8 2.4 2.3 0.0 0.0 Note: Net trade is measured at f.o.b. prices (imports exclude international trade and transport margins). Source: World Bank simulations with LINKAGE model, based on release 5.4 of the GTAP data. 118 Global Agricultural Trade and Developing Countries agricultural and food trade will improve dramati- in world prices induced by trade reform could lead cally from a deficit of $17 billion in 2000 to a sur- to a negative terms-of-trade shock since the agri- plus of $50 billion in 2015 (at 1997 prices). The cultural commodities they tend to export--for opposite occurs in developing countries, where a net example, coffee and cocoa--already have relatively positive balance in agriculture and food turns into a free access. large deficit of $50 billion, due mostly to a balloon- With relatively low demand growth in industrial ing deficit in processed food. Agriculture and food countries and relatively high output growth, the balances are positive for low-income countries in exportable agricultural surplus will increase sub- 2000 and 2015. stantially, particularly from North America and Developing a baseline of the future world econ- Oceania. Europe and Japan are the exceptions, with omy requires nuanced analysis. The country and output growth expected to be anemic. regional growth rates used here are in line with consensus views, given stronger demographic The Impacts of Agricultural trends and income elasticities for agriculture and Reform food in developing economies. World and regional totals may be skewed by several factors. The weights The impacts of agricultural trade reform are exam- are biased toward industrial countries because of ined first in the context of global merchandise trade the use of base year (1997) value shares. Volume reform, and then the results are decomposed by shares would yield different figures. Demand type of reform and region, to assess the relative growth in developing countries may be overstated importance for developing countries of reforms in because income elasticities are held constant at industrial countries and in developing countries. their base year levels. It is plausible to argue that income elasticities would converge toward those of Results of Global Merchandise Trade Reform high-income countries as developing countries grow. The growth numbers are also broadly consis- Global reform involves removing protection in all tent with Food and Agriculture Organization (nonservice) sectors, in all regions, and for all (FAO) historical trends. The discrepancy between instruments of protection (leaving other taxes agricultural growth and food processing originates unchanged, although lump-sum taxes (or trans- in the growth in intermediate demand for agricul- fers) on households adjust to maintain a fixed gov- tural products as food processing grows. A more ernment fiscal balance). The model contains six meat-intensive future world would also exhibit a instruments of protection: slight acceleration in agricultural growth relative to food because of the feed input in the livestock · Import tariffs, eliminated only if they are sector. So, while the baseline scenario is plausible, positive aggregate growth rates should be used with caution · Export subsidies, eliminated only if they are for all these reasons. negative5 The biggest mover among developing countries · Capital subsidies, with direct payments con- is China, where the food deficit of $8 billion in 1997 verted into subsidies on capital would swell to somewhere around $120 billion by · Land subsidies, with some payments also con- 2015. Demand is expected to outpace output by verted to subsidies on land about 1 percentage point a year.4 In agriculture this · Input subsidies provides new opportunities for Sub-Saharan Africa · Output subsidies and Latin America, with both seeing a large rise in agricultural surplus (on an aggregate basis). Sub- The overall measure of reform, referred to as Saharan Africa will nonetheless see a slight deterio- real income, measures the extent to which house- ration in its processed-food balance. The aggregate holds are better off in the post-reform scenario net trade balances may mask more detailed sectoral than in the baseline scenario in the year 2015.6 The shifts. For example, Sub-Saharan Africa will con- world gain (measured in 1997 U.S. dollars) is tinue to be a net importer of grains through the $385 billion, an increase from baseline income of baseline scenario time horizon, and therefore a rise some 0.9 percent (table 7.3). The gains are relatively Global Agricultural Reform: What Is at Stake? 119 TABLE 7.3 Real Income Gains and Losses from Global Merchandise Trade Reform: Change from 2015 Baseline Export Capital Land Input Output All Tariffs Subsidies Subsidies Subsidies Subsidies Subsidies Country Grouping Instruments Only Only Only Only Only Only Change in value (billions of 1997 US$) High-income countries 188.3 160.4 1.4 1.1 -4.8 -0.3 9.0 Low-income countries 31.9 34.6 -1.1 -0.1 -0.7 -0.3 0.2 Middle-income countries 164.7 187.7 -7.0 -1.2 -7.3 -3.8 -6.4 Low-income countries, 19.9 21.5 -0.9 -0.1 -0.6 -0.2 0.9 excluding India Middle-income countries, 176.7 200.8 -7.3 -1.2 -7.4 -3.9 -7.0 including India Developing countries 196.5 222.3 -8.2 -1.3 -8.1 -4.1 -6.2 World total 384.8 382.7 -6.8 -0.2 -12.8 -4.4 2.8 Percentage change High-income countries 0.6 0.5 0.0 0.0 0.0 0.0 0.0 Low-income countries 1.6 1.7 -0.1 0.0 0.0 0.0 0.0 Middle-income countries 1.8 2.0 -0.1 0.0 -0.1 0.0 -0.1 Low-income countries 1.9 2.1 -0.1 0.0 -0.1 0.0 0.1 excluding India Middle-income countries 1.7 1.9 -0.1 0.0 -0.1 0.0 -0.1 including India Developing countries 1.7 1.9 -0.1 0.0 -0.1 0.0 -0.1 World total 0.9 0.9 0.0 0.0 0.0 0.0 0.0 Source: World Bank simulations with LINKAGE model, based on release 5.4 of the GTAP data. evenly divided between industrial countries ipated changes, such as the elimination of the Multi- ($188 billion) and developing countries ($197 bil- fiber Arrangement, are not taken into account.8 lion), but developing countries are considerably Third, changes to some key assumptions or better off as a share of reference income, with a gain specifications could generate higher benefits. For of 1.7 percent compared with 0.6 percent for indus- example, raising the trade elasticities--as some trial countries. have argued--dampens the negative terms-of- trade effects. Increasing returns to scale can gener- Caveats. A few caveats about the basic global ate greater efficiency improvements, depending reform scenario. First, there are known deficiencies on the structure of product markets and scale in the base year policies,which are taken from release economies to be achieved. Reform of services could 5.4 of the Global Trade Analysis Project (GTAP) have economywide impacts to the extent that database. Most preferential arrangements-- cheaper and more efficient services can lower pro- including the Generalized System of Preferences duction costs as well as improve real incomes. and some regional trading agreements--are not Changes in investment flows--not modeled here-- incorporated.7 Alternative scenarios could be under- have proven to be as important (sometimes more) taken to test their overall importance especially as lowering trade barriers in many regional agree- regarding the utilization rates of the preferences. ments. In a global model, the net change would be Second, the reference scenario assumes no changes zero. Therefore, any reallocation of capital would in the base year policies between the base and termi- lead some countries to be better off, all else remain- nal years. Thus changes in trading regimes since ing the same, and others worse off (abstracting 1997,such as China's accession to the WTO,or antic- from the benefits of future repatriated profits). 120 Global Agricultural Trade and Developing Countries Gross flows could have a greater impact than net general equilibrium effects inherent in multisec- capital flows to the extent that they raise productiv- toral global models. ity if they are associated with technology-laden While the aggregate measure of gain often gar- capital goods. Finally, dynamic effects can also lead ners the most attention--at least from policy mak- to a boost in the overall gains from reform. ers and the media--more relevant for most players The global scenario captures some of the inher- are the detailed structural results. By and large, it is ent dynamic gains, notably changes from savings the structural results that influence the political and investment behavior. These can sometimes economy of reforms, particularly since the losers have a substantial impact, to the extent that from reforms tend to be concentrated and a well- imported capital goods are taxed. Assuming that identified pressure group, whereas the gainers are savings rates are unchanged, a sharp fall in the price typically diffuse and harder to identify. For exam- of capital goods can lead to a significant rise in ple, a 10 percent decline in the price of wheat could investment (more bang per dollar invested). The have a major impact on a farmer's income, but scenario does not incorporate changes to produc- an almost imperceptible effect on the average tivity, however. The channels and magnitudes of consumer. trade-related changes to productivity are as yet With reform, aggregate agricultural output of poorly validated by solid empirical evidence, and industrial countries declines--by more than attempts to incorporate these effects are by and 11 percent when all forms of protection are elimi- large simply illustrative of potential magnitudes. nated (table 7.4). Removal of tariff protection gen- Recent World Bank reports suggest that these erates the greatest change to production in indus- effects could be large, but the reports are really an trial countries, but unlike the case with the welfare appeal for more empirical research.9 impacts, the other forms of protection have meas- urable, if smaller, impacts on output. Removal of Decomposition by instrument. The key finding output subsidies results in the next greatest change on instruments of protection is the predominant in agricultural output, driven largely by the nearly role of tariffs. Removal of tariffs accounts for virtu- 5 percent output decline in the United States-- ally all of the gains. The other instruments have although land and export subsidies have nearly the much smaller impacts on real income--slightly same aggregate impact. The detailed results for the positive on average for industrial countries and Quad countries confirm several points of common negative in aggregate for developing countries. For wisdom regarding the patterns of protection. First, example, elimination of export subsidies negatively the United States makes more use of output subsi- affects Africa--both North and Sub-Saharan--and dies than do Europe and Japan. Europe makes the Middle East, although it provides a positive greater use of export subsidies and direct payments benefit for Europe. Elimination of domestic protec- (capital and land subsidies). Japanese protection is tion also tends to be negative for developing mostly in the form of import barriers. countries, and at times for industrial countries as well. The rest of Sub-Saharan Africa is a notable Results of Agricultural Reform exception, with an income gain of 0.6 percent. This could reflect the removal of significant output sub- Full merchandise trade reform provides a bench- sidies on cotton in some of the major producing mark from which to judge the maximal effects from countries (China and the United States). reform. This section focuses on the agricultural and The ambiguity of the welfare impact is in part food sectors. driven by the nature of partial reforms. Removal of one form of protection may exacerbate the negative Real income gains. If all regions remove all pro- effects of other forms of protection. For example, tection in agriculture and food, the global gains in removal of output subsidies may worsen the impact 2015 amount to $265 billion--nearly 70 percent of of tariffs if removal of the subsidies leads to a the gains from full merchandise trade reform reduction in output and an increase in imports. (table 7.5). This is remarkable considering the There are no robust theoretical arguments to deter- small size of agriculture and food in global output mine which is more harmful. There are also other (figure 7.1).10 Agriculture represents less than Global Agricultural Reform: What Is at Stake? 121 TABLE 7.4 Agricultural Output Gains and Losses from Global Merchandise Trade Reform: Change from 2015 Baseline Export Capital Land Input Output All Tariffs Subsidies Subsidies Subsidies Subsidies Subsidies Country Grouping Instruments Only Only Only Only Only Only Change in value (billions of 1997 US$) High-income countries -109.7 -56.2 -9.5 -1.6 -10.4 -7.4 -12.0 Low-income countries 14.8 11.5 1.1 0.0 0.7 0.4 2.0 Middle-income countries 41.8 18.1 8.2 -0.2 8.5 0.5 9.3 Low-income countries, 13.7 10.5 0.9 0.0 0.5 0.3 2.8 excluding India Middle-income countries, 42.9 19.2 8.4 -0.2 8.7 0.6 8.6 including India Developing countries 56.6 29.7 9.3 -0.1 9.2 0.9 11.3 World total -53.1 -26.6 -0.2 -1.7 -1.2 -6.5 -0.7 Percentage change High-income countries -11.1 -5.7 -1.0 -0.2 -1.1 -0.7 -1.2 Low-income countries 2.4 1.8 0.2 0.0 0.1 0.1 0.3 Middle-income countries 2.4 1.0 0.5 0.0 0.5 0.0 0.5 Low-income countries, 4.1 3.1 0.3 0.0 0.2 0.1 0.8 excluding India Middle-income countries, 2.1 0.9 0.4 0.0 0.4 0.0 0.4 including India Developing countries 2.4 1.2 0.4 0.0 0.4 0.0 0.5 World total -1.6 -0.8 0.0 -0.1 0.0 -0.2 0.0 Source: World Bank simulations with LINKAGE model, based on release 5.4 of the GTAP data. TABLE 7.5 Real Income Gains from Agricultural and Food Trade Reform: Change from 2015 Baseline (billions of 1997 US$) Agricultural and Global Food Trade Agricultural Merchandise Reform Trade Reform Trade Reform Only High-Income Country Grouping Global Global Countries High-Income High-income countries 188.3 136.6 92.0 29.3 Low-income countries 31.9 10.3 3.0 1.1 Middle-income countries 164.7 118.2 6.9 -4.9 Low-income countries, 19.9 8.4 3.6 1.6 excluding India Middle-income countries, 176.7 120.1 6.4 -5.3 including India Developing countries 196.5 128.6 10.0 -3.8 World total 384.8 265.2 102.0 25.5 Source: World Bank simulations with LINKAGE model, based on release 5.4 of the GTAP data. 122 Global Agricultural Trade and Developing Countries FIGURE 7.1 Output Structure in Base Year, 1997 Percent 100 90 80 70 60 50 40 30 20 10 0 Industrial Low-income Middle-income Agriculture Processed foods Manufacturing Other goods and services Source: GTAP release 5.4. 2 percent of output for industrial countries and income countries, where the gains from their own 10.5 percent for developing countries, while agricultural and food reform would be quite sub- processed foods represent 4.5 percent for industrial stantial. On a percentage basis, this is less so for countries and 7.5 percent for developing countries. low-income countries. The industrial countries Agriculture is still a relatively high 19 percent of reap gains of $92 billion, implying that agricultural output in the low-income developing countries. reform in developing countries could generate Clearly, protection tends to be higher in agriculture gains of about $45 billion for the industrial and food than in other sectors, particularly in countries. industrial countries, but in middle-income coun- The final decomposition scenario is to assess the tries as well. Protection is more uniform in low- impacts of reform in agriculture alone in industrial income countries. countries--leaving protection unchanged for For low-income countries the gains from global processed foods. This lowers the gains substantially free trade in agriculture and food amount to for industrial countries--from $92 billion to $29 around one-third of the gains from global free billion (see table 7.5). Protection is high in both trade in all merchandise. This is a consequence of sectors, and the processed foods sector is more than their dependence on imports of the most protected twice as large as the agricultural sector. Further- food items--such as grains--while they are net more, in a partial reform scenario, the efficiency exporters of commodities with little or no protec- gains in agriculture could be offset to some extent tion. The middle-income countries gain 71 percent by further losses in processed foods. Output will from global free trade in agriculture and food, expand in the processed food sector as resources nearly as much as industrial countries, which gain are moved around--and the lower costs of inputs 72 percent as compared with full merchandise will also provide incentives to increase output. trade reform. Middle-income countries could lose from an If reforms are limited to high-income agriculture-only reform in industrial countries. countries--a super-version of special and differen- They would benefit little from improved market tial treatment--with perhaps an agreement by access in agriculture, and in a partial reform sce- middle-income countries to bind at existing levels nario, expansion of their protected domestic agri- of protection, global gains drop to $102 billion, culture and food production leads to efficiency indicating that a significant portion of the global losses that are not compensated elsewhere. gains is generated by removal of agricultural barri- To conclude--global agricultural trade reform ers in developing countries (see table 7.5).11 The generates a huge share of the gains to be made from drop in gains is particularly striking for middle- merchandise trade reform. Market access into Global Agricultural Reform: What Is at Stake? 123 industrial countries provides significant gains, but Exports in agriculture and food from developing a greater share of the gains for developing countries countries would jump $300 billion, an increase of comes from agricultural trade reform among more than 115 percent, with industrial-country developing countries. Finally, reform in agriculture exports increasing $220 billion, or 50 percent. On alone provides few benefits. It needs to be linked to the flip side, imports from both industrial and reform in the processed food sectors. developing countries would rise substantially. The net trade position of industrial countries would Structural implications. Accelerating integration deteriorate marginally--from $50 billion in the is one of the key goals of trade reform. Beyond the baseline in 2015 to $48 billion after global reform efficiency gains that come from allocating resources of agriculture and food. The marginal improve- to their best uses, integration is expected to bring ment for developing countries decomposes into a productivity increases--scale economies, greater boost of nearly $12 billion for low-income coun- competitiveness, ability to import technology- tries and deterioration for middle-income coun- laden intermediate goods and capital, greater mar- tries of nearly $10 billion. ket awareness, and access to networks. If the reform is limited to industrial countries, The potential changes in trade from global the picture is modified significantly. First, the reform of agriculture and food are large. World change in imports for industrial countries is almost trade in these two sectors could jump by more than identical under the two scenarios--$223 billion half a trillion dollars in 2015 (compared with the with full reform and $205 billion with industrial- baseline), an increase of 74 percent (table 7.6). country reform only (see table 7.6). Developing TABLE 7.6 Impact of Global Agricultural and Food Reform on Agricultural and Food Trade: Change from 2015 Baseline Exports Imports Net Trade Country Grouping Global Industrial Global Industrial Global Industrial 2015 Baseline Change in value (billions of 1997 US$) High-income countries 221.2 63.4 223.3 205.3 -2.1 -141.9 50.4 Low-income countries 41.0 20.9 29.2 -0.3 11.8 21.2 23.4 Middle-income countries 260.1 120.5 269.8 -0.2 -9.7 120.7 -73.8 Low-income countries, 33.8 17.5 21.9 0.1 11.8 17.5 22.2 excluding India Middle-income countries, 267.3 123.9 277.1 -0.5 -9.8 124.4 -72.7 including India Developing countries 301.1 141.4 299.0 -0.4 2.1 141.9 -50.4 World total 522.3 204.9 522.3 204.9 0.0 0.0 0 Percentage change High-income countries 50 14 57 52 Low-income countries 74 38 92 -1 Middle-income countries 125 58 96 0 Low-income countries, 70 36 84 0 excluding India Middle-income countries, 125 58 96 0 including India Developing countries 115 54 95 0 World total 74 29 74 29 Note: The columns labeled Global refer to the impacts from global agriculture and food reform. The columns labeled Industrial refer to industrial-country only reform of agriculture and food. Source: World Bank simulations with Linkage model, based on release 5.4 of the GTAP data. 124 Global Agricultural Trade and Developing Countries countries see a significant rise in exports, but to would gain additional market access from develop- industrial countries only, with little or no change in ing countries and reinforce their comparative their own imports. Thus industrial countries would advantage over more highly protected countries in witness a much sharper deterioration in their net East Asia. The biggest beneficiary in net terms food bill, with net imports registering a change of would be China. While its (small) exports would $142 billion instead of $2 billion, as under the not change much, removal of its own protection global reform scenario. The United States and induces a huge shift in imports. The lack of reform Europe bear the brunt of the adjustment, with under the partial reform scenario means that Canada, Australia, and New Zealand seeing little instead of its net food position deteriorating by difference between the global and partial reform $74 billion in the global reform, it sees a small scenarios. In other words, these three countries improvement of $6 billion. On aggregate for devel- reap much of the trade benefits from greater mar- oping countries the partial reform would gener- ket access within industrial countries. Opening up ate an improvement in net trade of food of of markets in developing countries significantly $142 billion. dampens the adjustment process for the United The structural impacts described above are asso- States and Europe, and the United States would ciated with global changes in the distribution of reinforce its net exporting status significantly farm income. With global agriculture and food under a global reform scenario. reform, farm incomes barely change at the global Most developing countries see a greater im- level (a loss of perhaps $10 billion,12 or 0.6 percent provement in their net food trade with industrial- of baseline 2015 farm income). Changes are much country-only reform than with global reform-- more significant at the regional level (figures 7.2 but not all countries. Argentina, Brazil, and the rest and 7.3). The largest absolute gains in farm income of East Asia improve their net food trade more are in the Americas, Australia and New Zealand, with global reform than with partial reform. They and developing East Asia excluding China. Latin FIGURE 7.2 Change in Rural Value Added from Baseline in 2015 (billions of 1997 $US) United States Rest of Latin America Argentina Brazil Australia and New Zealand Rest of East Asia Rest of Sub-Saharan Africa Canada Mexico Turkey Indonesia EU accession countries Southern African Customs Union Vietnam Rest of the world Rest of Europe and Central Asia India Rest of South Asia Korea, Rep. of, and Taiwan (China) Japan Western Europe China 100 80 60 40 20 0 20 40 Source: World Bank simulations with Linkage model, based on release 5.4 of the GTAP data. Global Agricultural Reform: What Is at Stake? 125 FIGURE 7.3 Percentage Change in Rural Value Added from Baseline in 2015 Australia and New Zealand Canada Vietnam Argentina Southern African Customs Union Rest of East Asia Rest of Latin America Rest of Sub-Saharan Africa Brazil United States EU accession countries Turkey Mexico Indonesia Rest of the world Rest of Europe and Central Asia India Rest of South Asia China Korea, Rep. of, and Taiwan (China) Western Europe Japan 40 20 0 20 40 60 80 Source: World Bank simulations with Linkage model, based on release 5.4 of the GTAP data. America would receive 40 percent of the total posi- labor and capital returns are determined essentially tive gains; Australia, Canada, and New Zealand on national markets.14 Thus wage changes are 18 percent, and the United States 15 percent. modest overall, with generally greater impacts The relative position of regional gainers is some- in developing countries, where more labor is what different, however (see figure 7.3). Farmers in employed in agriculture (table 7.7). For example, Australia, Canada, and New Zealand gain the most wages for unskilled labor increase 8 percent in from global free trade in agriculture and food, with Argentina and Vietnam, and 5­6 percent in the rest income gains of 50­65 percent. Farmers in a num- of Latin America and the rest of Sub-Saharan ber of developing regions have gains of more than Africa. Unskilled workers in Australia and New 25 percent--Vietnam, Argentina, countries of the Zealand also benefit from these reforms. Unskilled Southern Africa Customs Union (SACU), the rest workers in developing countries generally do better of East Asia (which includes Thailand, Malaysia, in relative terms than skilled workers, largely as a and the Philippines), and the rest of Latin America. result of their concentration in agricultural sectors. The farmers who lose most are in China, with China is a significant exception. Removal of its potential losses of $75 billion in 2015 compared agricultural protection lowers demand for with the baseline scenario.13 The next biggest losers unskilled workers, and their wages decline. The are farmers in Western Europe and the developed impact on wages in the European Union and Japan East Asian economies--Japan, the Republic of is negligible, as agriculture employs a very small Korea, and Taiwan (China). In percentage terms the share of the national labor force. biggest losses occur in Japan (30 percent) and West- As in the labor markets, the returns in capital ern Europe (24 percent), with China's losses down market are determined mainly at the national level to about 15 percent because of its huge rural (table 7.8). Thus changes to income will largely be economy. reflected in volume changes, not price changes. Most of the impact on rural incomes is gener- Direct payments to farmers, however, are imple- ated by volume changes, not factor returns. Both mented as an ad valorem subsidy on capital (and 126 Global Agricultural Trade and Developing Countries TABLE 7.7 Impact of Global Agriculture and Food Reform on Agricultural Employment and Wages: Change from 2015 Baseline (percent) Total Agriculture Cereals and Sugar Livestock and Dairy Wages Wages Wages Employ- Employ- Employ- Country or Region ment Unskilled Skilled ment Unskilled Skilled ment Unskilled Skilled Canada 8.5 1.0 0.8 30.4 1.0 0.8 -15.5 1.0 0.8 United States 0.4 0.6 0.6 -12.4 0.6 0.6 3.3 0.6 0.6 European Union -23.7 -0.6 0.4 -57.7 -0.6 0.4 -28.0 -0.6 0.4 with EFTA* Australia and 18.2 3.4 2.3 25.6 3.4 2.3 31.1 3.4 2.3 New Zealand Japan -26.8 -0.9 -0.1 -28.9 -0.9 -0.1 -46.2 -0.9 -0.1 Korea, Rep. of and -13.8 -0.2 0.7 -3.9 -0.2 0.7 8.2 -0.2 0.7 Taiwan (China) Argentina 13.3 7.9 5.5 25.8 7.9 5.5 14.3 7.9 5.5 Brazil 12.5 3.4 3.0 25.8 3.4 3.0 11.7 3.4 3.0 China -6.6 -3.1 0.0 -26.6 -3.1 0.0 8.6 -3.1 0.0 India -0.3 0.0 0.2 0.7 0.0 0.2 1.1 0.0 0.2 Indonesia 4.3 1.4 -0.3 6.1 1.4 -0.3 -2.0 1.4 -0.3 Mexico 5.0 1.3 -0.2 1.3 1.3 -0.2 -4.8 1.3 -0.2 Southern African 13.8 1.3 1.1 31.7 1.3 1.1 8.8 1.3 1.1 Customs Union Turkey 5.2 3.0 0.5 -15.3 3.0 0.5 -18.7 3.0 0.5 Vietnam 17.0 7.8 3.0 63.1 7.8 3.0 -15.4 7.8 3.0 Rest of East Asia 11.6 2.7 0.9 72.0 2.7 0.9 -9.1 2.7 0.9 Rest of South Asia -1.3 -0.2 0.0 1.1 -0.2 0.0 0.7 -0.2 0.0 EU accession countries 6.9 1.6 0.9 12.8 1.6 0.9 13.3 1.6 0.9 Rest of Europe and -0.4 -1.0 -0.3 0.3 -1.0 -0.3 -2.4 -1.0 -0.3 Central Asia Rest of Sub-Saharan 6.2 6.0 1.9 17.9 6.0 1.9 1.2 6.0 1.9 Africa Rest of Latin America 6.2 5.4 3.4 17.9 5.4 3.4 42.6 5.4 3.4 Rest of the World -0.1 -0.3 0.9 2.6 -0.3 0.9 -4.2 -0.3 0.9 including Middle East and North Africa *European Free Trade Association, (Austria, Finland, Iceland, Norway, Sweden, and Switzerland). Source: World Bank simulations with LINKAGE model, based on release 5.4 of the GTAP data. land), thus creating a wedge between the cost to fixed factor in agriculture, with some allowance for farmers and the returns to owners. Removal of the movements up and down the supply curve and for capital subsidy has little effect on owners since the cross-sectoral shifts in land usage.15 In Europe the return is determined at the economywide level, but average return to land drops 66 percent, with the it raises the costs to farmers. For example, the cost supply of land falling 9 percent following global of capital net of subsidies increases by almost 1 per- reform. Farmers gain some benefit in lower unit cent in the European Union, but the average cost to costs because of falling land prices. But removal of farmers increases by 22 percent--and even more the direct subsidy does not allow farmers to reap for livestock producers (43 percent). Note that the full cost gains from falling land prices. The these capital subsidies are used mainly in industrial average cost for farmers drops 57 percent, lower countries, so for most developing countries there is than the drop in the rental price of land (66 per- no difference between the owner return and the cent). And the change in the cost structure is cost to farmers. highly sector specific. Thus cereal and grain farm- The changes in the contribution of land to agri- ers see a small drop in their net cost of land (5 per- cultural incomes are driven largely by price cent); however, the drop in the price of land does movements--contrary to the case for labor and not compensate for removal of the subsidies since capital income (table 7.9). Land is essentially a the returns to owners falls by 74 percent. This is Global Agricultural Reform: What Is at Stake? 127 TABLE 7.8 Impact of Global Agricultural and Food Trade Reform on Agricultural Capital: Change from 2015 Baseline (percent) Total Agriculture Grains and Sugar Livestock and Dairy Owners' Farmers' Owners' Farmers' Owners' Farmers' Country or Region Volume Return Cost Volume Return Cost Volume Return Cost Canada -4.9 -0.5 4.1 7.2 -0.5 3.1 -17.0 -0.5 7.1 United States 0.8 0.7 2.6 -19.2 0.7 2.6 4.5 0.7 6.5 European Union with EFTA -32.9 0.7 21.8 -67.1 0.7 21.7 -29.2 0.8 43.1 Australia and New Zealand 40.2 0.6 1.2 3.0 0.7 1.3 123.5 0.6 1.7 Japan -22.9 1.7 4.9 -25.0 1.7 7.6 -47.0 1.7 12.2 Korea, Rep. of and -4.3 0.7 12.0 8.9 0.8 15.4 17.5 0.8 103.8 Taiwan (China) Hong Kong (China) 9.8 0.7 0.7 75.4 0.7 0.7 -4.3 0.7 0.7 and Singapore Argentina 6.0 4.2 4.2 9.0 4.2 4.2 17.9 4.2 4.2 Brazil 10.1 3.1 3.1 21.9 3.1 3.1 9.8 3.1 3.1 China -2.7 3.2 3.2 -17.5 3.2 3.2 5.8 3.2 3.2 India 0.0 0.1 0.1 0.8 0.1 0.1 1.2 0.1 0.1 Indonesia 0.7 -0.2 -0.2 1.0 -0.2 -0.2 -0.9 -0.2 -0.2 Mexico 4.3 -0.1 3.7 2.3 -0.1 4.4 -7.5 -0.1 9.1 Southern African 19.5 -0.6 -0.6 39.4 -0.6 -0.6 25.4 -0.6 -0.6 Customs Union Turkey 0.2 -0.4 -0.2 -15.8 -0.4 0.5 -15.1 -0.4 -0.4 Vietnam 2.4 1.8 1.8 28.7 1.8 1.8 -13.3 1.8 1.8 Rest of East Asia 20.9 0.2 0.2 36.5 0.2 0.2 -8.6 0.2 0.2 Rest of South Asia 0.1 1.3 1.3 2.3 1.3 1.3 0.7 1.2 1.2 EU accession countries -0.2 0.5 21.6 7.7 0.5 18.9 -6.3 0.5 67.6 Rest of Europe and -2.5 1.6 7.7 -1.9 1.6 8.3 -5.8 1.6 9.3 Central Asia Rest of Sub-Saharan Africa 0.5 -1.1 -1.1 5.6 -1.1 -1.1 4.0 -1.1 -1.1 Rest of Latin America 6.2 1.8 1.8 15.9 1.8 1.8 41.1 1.8 1.8 Rest of the World 0.3 -0.2 -0.2 2.9 -0.2 -0.2 -3.7 -0.2 -0.2 including Middle East and North Africa Source: World Bank simulations with LINKAGE model, based on release 5.4 of the GTAP data. not the case in the livestock sector, where subsidy Sensitivity Analysis payments are linked to capital (the herds) and not This section uses sensitivity analysis to explore how to land. The impacts in the United States are results change when some of the basic assumptions muted, with the overall return to landowners of the model change. It focuses on four areas: changing slightly--a decline of 5 percent--but costs to farmers increasing substantially--22 per- · The agricultural productivity assumptions of cent on average and more than 42 percent for the standard baseline scenario. Agricultural cereal and sugar producers. productivity is cut by 1 percentage point in In most developing countries land prices developing countries and the results from global increase substantially, except in China and in a few agriculture and food reform are compared with other regions. This may reduce to some extent the results using the default productivity the positive distributional impacts from relatively assumptions. In a separate analysis, productivity higher wages for unskilled labor since land owner- is increased for middle-income developing ship may not necessarily be congruent with the countries. unskilled labor working the land. There are some · The impacts of the mobility of agricultural capi- interesting sectoral shifts. For example, China tal. Agricultural capital is more closely tied to would see more land devoted to livestock and dairy the sector, making it more difficult to shed and and less to cereals, which would be imported from leading to a different transition when reform is lower-cost sources. undertaken. 128 Global Agricultural Trade and Developing Countries TABLE 7.9 Impact of Global Agriculture and Food Reform on Agricultural Land: Change from 2015 Baseline (percent) Total Agriculture Cereals and Sugar Livestock and Dairy Price Price Price Country or Region Land Owner Farmer Land Owner Farmer Land Owner Farmer Canada -6.4 69.5 133.8 6.6 76.9 192.8 -25.2 56.8 83.5 United States 2.4 -5.1 22.1 -19.0 -12.5 42.1 12.3 -0.2 9.1 European Union with EFTA -9.4 -66.3 -57.0 -58.9 -74.1 -4.7 -3.5 -65.0 -59.7 Australia and New Zealand 6.2 197.8 219.1 1.9 197.0 224.0 34.8 219.6 252.4 Japan -21.0 -44.9 -41.5 -24.0 -45.5 -34.6 -34.1 -48.9 -48.9 Korea, Rep. of and -11.4 -27.6 -27.1 -0.2 -25.3 -24.6 4.1 -23.0 -20.9 Taiwan (China) Argentina 4.5 56.2 56.2 11.4 59.5 59.5 12.0 60.0 60.0 Brazil 9.9 18.0 18.0 23.8 22.9 22.9 8.6 17.6 17.6 China -0.9 -25.7 -25.7 -21.1 -31.1 -31.1 7.6 -23.6 -23.6 India 0.0 -1.8 -1.8 0.8 -1.5 -1.5 1.4 -1.3 -1.3 Indonesia 0.7 10.9 10.9 2.1 11.4 11.4 -1.8 10.0 10.0 Mexico 2.7 0.6 13.1 -8.9 -3.6 52.1 -1.6 -0.6 0.8 Southern African 8.0 86.4 86.4 26.4 95.2 95.2 4.5 84.9 84.9 Customs Union Turkey 0.8 47.3 47.3 -14.9 39.0 39.0 -20.2 36.1 36.1 Vietnam -0.3 44.6 44.6 33.2 60.3 60.3 -16.0 38.1 38.1 Rest of East Asia -1.5 34.1 34.1 43.7 53.8 53.8 -9.6 32.7 32.7 Rest of South Asia -0.1 -6.0 -6.0 3.2 -5.0 -5.0 1.3 -5.4 -5.4 EU accession countries 2.6 2.0 6.1 4.6 2.8 10.8 7.5 3.5 8.8 Rest of Europe and -1.5 -2.4 -2.4 -1.1 -2.3 -2.3 -1.2 -2.2 -2.2 Central Asia Rest of Sub-Saharan Africa -0.3 62.8 62.8 9.0 67.9 67.9 -2.4 61.7 61.7 Rest of Latin America 1.0 55.4 55.4 5.0 58.6 58.6 40.3 74.9 74.9 Rest of the World 0.0 0.1 0.1 2.7 0.8 0.8 -4.3 -1.2 -1.2 including Middle East and North Africa Source: World Bank simulations with LINKAGE model, based on release 5.4 of the GTAP data. · Sensitivity of the results to supply rigidities in baseline was constructed with agricultural produc- developing countries. tivity improving at a slower 1.5 percent for devel- · Sensitivity of the results to the key trade oping countries, but remaining at 2.5 percent for elasticities. industrial countries. Trade impact. Under the standard baseline, high- Agricultural Productivity income countries go from a position of net food Agricultural productivity is assumed to grow importers in 1997 to net food exporters in 2015 2.5 percent a year globally in the standard baseline (table 7.10). Low-income countries improve their scenario based on existing evidence (Martin and position significantly, going from a positive food Mitra 1996, 1999). This may be too optimistic for balance of $12.5 billion in 1997 to $23 billion in developing countries, particularly for low-income 2015. The position of middle-income countries countries. This assumption may have an impact on deteriorates, however. Under the low-productivity long-term self-sufficiency rates, particularly of sen- baseline, the net food trade position of industrial sitive commodities. The more trade reform raises countries increases substantially--jumping to $151 the world price of food, the more net food billion in 2015 compared with only $50 billion in importers will be adversely affected by negative the standard baseline. Low-income countries still terms-of-trade shocks. To test the sensitivity of the maintain a positive balance, but the balance is trade results to agricultural productivity, a different much closer to zero than it was in the previous Global Agricultural Reform: What Is at Stake? 129 TABLE 7.10 Net Trade Impacts Assuming Lower Agricultural Productivity in Developing Countries (billions of 1997 US$) Standard Productivity Low Productivity Baseline Reform Baseline Reform Country Grouping 1997 2015 2015 2015 2015 High-income countries -23.1 50.4 48.4 151.2 181.6 Low-income countries 12.5 23.4 35.2 0.9 4.5 Middle-income countries 10.5 -73.8 -83.6 -152.0 -186.1 Low-income countries, 7.4 22.2 34.1 8.5 17.2 excluding India Middle-income countries, 15.6 -72.7 -82.4 -159.7 -198.9 including India Developing countries 23.1 -50.4 -48.4 -151.2 -181.6 Source: World Bank simulations with LINKAGE model, based on release 5.4 of the GTAP data. TABLE 7.11 Impacts on Output Assuming Lower Agricultural Productivity for Developing Countries Growth in 2000­ Baseline Difference Difference between Baseline 15 (percent) in 2015 and Reform Scenario in 2015 Low Standard Value Percentage Low Standard Low Standard Country Grouping Baseline Baseline ($ billions) Change ($ billions) ($ billions) (percent) (percent) High-income countries 1.9 1.2 122.6 12.4 -100.0 -107.7 -9.0 -10.9 Low-income countries 2.8 3.6 -71.6 -11.4 8.7 12.1 1.6 1.9 Middle-income 2.6 3.2 -166.2 -9.4 27.0 37.2 1.7 2.1 countries Low-income countries, 3.0 3.7 -39.4 -11.6 10.3 12.3 3.4 3.6 excluding India Middle-income 2.6 3.2 -198.4 -9.7 25.4 37.0 1.4 1.8 countries, including India Developing countries 2.6 3.3 -237.8 -10.0 35.7 49.4 1.7 2.1 World total 2.4 2.6 -115.2 -3.4 -64.3 -58.3 -2.0 -1.7 Source: World Bank simulations with LINKAGE model, based on release 5.4 of the GTAP data. baseline. And the net food trade situation of but by a more modest $3.6 billion rather than the middle-income countries shows a greater depend- nearly $12 billion using the standard productivity ence on world markets. assumptions. Whereas reform in the standard baseline posi- tions low-income countries as net food exporters Output impact. Average annual agricultural out- and has only a mild negative effect on the food bal- put growth in developing countries slows from ance of high- and middle-income countries, under 3.3 percent in the standard baseline to 2.6 percent the low-productivity assumption, the food trade in the low-productivity baseline (table 7.11). In balance of the high-income countries improves industrial countries higher productivity provides substantially--by $30 billion--largely because of an opportunity to gain market share, and higher an increased dependence on food imports by world prices relative to the original baseline pro- middle-income countries. The low-income coun- vide greater incentives to produce. World output tries still see an improvement in their trade balance, under the alternative scenario declines 3.4 percent 130 Global Agricultural Trade and Developing Countries (higher prices lead to reduced demand), with a drop of $1.4 billion (0.08 percentage point) and reallocation between industrial and developing middle-income countries a drop of $17.8 billion countries. Industrial countries benefit from a (0.19 percentage point). 12 percent increase in output in 2015 compared with the standard baseline, whereas developing A high-productivity assumption. Many middle- country output is reduced by some 10 percent. income countries such as Argentina, Brazil, and With respect to output impacts following the Thailand have experienced rapid growth in agricul- trade reform scenario, the qualitative results of the ture, suggesting the potential for higher productiv- different baseline assumptions of agricultural pro- ity growth than assumed in the standard baseline. ductivity are identical--trade reform of agriculture To explore this, agricultural productivity growth and food lead to a shift in agricultural production was raised from 2.5 percent to 4.0 percent for from industrial to developing countries. In the middle-income countries (China, India, Indonesia, standard baseline, developing-country agricultural rest of East Asia,Vietnam,Argentina, Brazil, Mexico, output increases more than 2 percent, whereas in rest of Latin America, the EU accession countries, the low-productivity baseline the increase is only rest of Europe and Central Asia, and Turkey). 1.7 percent. The decline in industrial countries Changes are as expected. Agricultural supply drops to 9 percent, from 11 percent in the standard and exports expand for natural exporters such as baseline. The changes in output patterns across Argentina and Brazil. China, the largest middle- regions are identical, although the magnitudes income importer, reduces its deficit by about differ. $18 billion (table 7.12). The middle-income group including India experiences a net surplus of $30 bil- Aggregate welfare. The change in the agricul- lion in 2015, whereas under the standard baseline it tural productivity assumption translates into mod- has a deficit of $19 billion. High-income countries est changes in aggregate welfare (figure 7.4). Indus- experience a deterioration of their net agricultural trial countries see an improvement of $18 billion in trade of about $50 billion, compared with $3 billion 2015, a jump in gains of some 0.05 percentage in the standard baseline, and Europe's deficit point. Developing countries see a reduction in their increases to nearly $60 billion. Results for the food welfare gains, with low-income countries seeing a sector are qualitatively similar, but smaller in size, FIGURE 7.4 Welfare Impacts of Productivity Changes $ billion Percent 180 1.4 160 1.2 140 1.0 120 100 0.8 80 0.6 60 0.4 40 0.2 20 0 0.0 High- Low- Middle- High- Low- Middle- income income income income income income Low productivity Default productivity Source: World Bank simulations with Linkage model, based on release 5.4 of the GTAP data. Global Agricultural Reform: What Is at Stake? 131 TABLE 7.12 Baseline Trends in Agriculture with Higher Agricultural Productivity in Middle-Income Countries Average Annual Growth, 2000­15 (percent) Net Trade (billions of 1997 US$) Country Grouping Output Demand Imports Exports 2000 2015 High-income countries 0.6 1.0 2.7 0.9 -24.3 -50.2 Low-income countries 3.9 3.8 4.0 6.2 9.9 25.2 Middle-income countries 3.7 3.6 7.5 7.2 14.4 24.9 Low-income countries, 3.8 3.5 4.0 6.2 7.2 19.1 excluding India Middle-income countries, 3.7 3.7 7.4 7.1 17.1 31.1 including India Developing countries 3.7 3.7 7.0 7.0 24.3 50.2 Note: Net trade is measured at f.o.b. prices (imports exclude international trade and transport margins). Source: World Bank simulations with LINKAGE model, based on release 5.4 of the GTAP data. TABLE 7.13 Baseline Trends in Food Processing with Higher Agricultural Productivity in Middle-Income Countries Average Annual Growth, 2000­15 (percent) Net Trade (billions of 1997 US$) Country Grouping Output Demand Imports Exports 2000 2015 High-income countries 1.1 1.0 1.4 2.0 7.7 36.2 Low-income countries 3.5 3.5 3.6 3.2 3.6 4.4 Middle-income countries 3.1 3.3 4.1 2.6 -11.3 -40.6 Low-income countries, 3.2 3.3 3.7 2.2 1.8 -0.1 excluding India Middle-income countries, 3.1 3.3 4.1 2.8 -9.5 -36.2 including India Developing countries 3.2 3.3 4.1 2.7 -7.7 -36.2 Note: Net trade is measured at f.o.b. prices (imports exclude international trade and transport margins). Source: World Bank simulations with LINKAGE model, based on release 5.4 of the GTAP data. with an increase in competitiveness of food pro- among developing countries--particularly middle- cessing in middle-income countries and a decrease income countries--and could lead to a different in net trade by high-income countries relative to assessment of the direction of food self-sufficiency the standard baseline (table 7.13). These large in the aftermath of reform. changes show how sensitive baseline trajectories are to changes in assumptions about the future. Mobility of Agricultural Capital and the Transition They do not, however, affect the impact of the in Industrial Countries reform scenario measured in deviations from the baseline. The focus so far has been mainly on the long-term In conclusion, the baseline assumptions regard- impact of the removal of protection, with little ing productivity are important, although changes in attention to the transitional impacts. A key mecha- the assumption would not yield substantially differ- nism of the model is the vintage structure of capi- ent results from agriculture and food trade reform tal. Sectors in decline have excess capital that will for developing countries in terms of net benefits not readily be used in other sectors. This is certainly and agricultural output.16 However, lower produc- the case with agricultural capital, although some tivity will reduce the level of food self-sufficiency could be used for nonagricultural purposes, and 132 Global Agricultural Trade and Developing Countries other equipment could be used in nonprotected over all industrial countries in any given year, and agricultural sectors. at most 0.3 percent for developing countries, but in Excess capital is released to other sectors follow- the opposite direction. There are no discernible ing an upward-sloping supply curve. The value for impacts on welfare. the supply elasticity in the standard model is 4. To In conclusion, lowering the supply elasticity will test the importance of this elasticity, the reform sce- draw out the supply response during the transition nario is simulated again, but with a supply elasticity phase but will have no discernible long-term of 0.5. This makes excess supply much less mobile impact on the results. and, all else equal, will tend to increase supply rela- tive to the same simulation with a higher supply Supply Response in the Low-Income Countries elasticity. Consider the case for the sugar sector in Europe. This section evaluates the impact of lowering the The starting point is 2004, since the trade reform land supply response in three regions--rest of starts in 2005. Under the baseline, sugar output in South Asia, the Southern Africa Customs Union Europe increases modestly between 2004 and 2015 region, and the rest of Sub-Saharan Africa--to (figure 7.5). With the start of reform, output drops examine whether low-income countries, with their rapidly, and by 2015 output has fallen from about potentially low supply response, will benefit from $42 billion to about $11 billion. The supply elastic- greater market access. This involves three param- ity has an impact on the rate of decline of sugar eters. First, the base year land supply elasticity was output, but the final level is more or less identical. reduced from 1 to 0.25. Second, the land supply Thus with a low supply elasticity, the transition is asymptote was reduced from 20 percent of the ini- drawn out over a longer period. The rate of decline tial land supply to 10 percent.17 These two param- between 2004 and 2010 is 18.4 percent using the eters determine aggregate land supply. A third standard elasticity and 16.5 percent with the lower parameter moderates the degree of land mobility elasticity. across sectors. The allocation of land across sectors There are only a handful of sectors in industrial is governed by a constant elasticity of transforma- countries where the supply elasticity has any tion function.18 The standard transformation elas- noticeable impact: wheat and sugar in the United ticity is 3, a relatively elastic value. In the sensitivity States; rice, wheat, other grains, oil seeds, and sugar simulation, the transformation elasticity for the in the European Union; and wheat and oil seeds in three regions is set to 0.5. Japan. The aggregate impacts on agricultural The lower land supply elasticities affect the base- production are negligible, at less than 1 percent line scenario. For the three regions where changes FIGURE 7.5 Sugar Output in Europe (US$ billions) $ billion 45 Baseline 40 35 Low supply 30 elasticity 25 20 Standard supply 15 elasticity 10 5 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Source: World Bank simulations with Linkage model, based on release 5.4 of the GTAP data. Global Agricultural Reform: What Is at Stake? 133 were made to supply elasticities, the overall rate of Saharan Africa, and remains the same for SACU at growth of agricultural output between 2000 and 2.1 percent (table 7.14). In all three regions, the 2015 declines from 3.4 to 3.1 percent in rest of most affected crop is plant-based fibers. These South Asia, from 4.0 to 3.8 percent in rest of Sub- three regions have a sizable market share at the TABLE 7.14 Impact of Lower Land Supply Elasticities in Rest of South Asia and Sub-Saharan Africa (percent) Impact of Trade Baseline Growth Rates 2000­15 Reform Standard Standard Baseline Standard Supply Low Supply Difference Supply Low Supply Commodity Elasticity Elasticity in 2015 Elasticity Elasticity Rest of South Asia Rice 2.8 2.7 -2.1 3.4 2.4 Wheat 2.7 2.6 -3.8 34.4 19.6 Other grains 3.8 3.6 -3.6 -2.4 -1.5 Oil seeds 4.1 3.5 -9.1 -10.0 -6.8 Sugar 3.8 3.3 -8.9 -17.2 -12.4 Plant-based fibers 4.5 3.7 -13.2 19.2 6.2 Other crops 3.6 3.2 -7.3 -8.0 -5.4 Cattle 4.0 3.7 -4.9 1.7 1.5 Other meats 4.1 3.6 -8.3 -1.0 -1.8 Raw milk 3.9 3.5 -6.1 1.3 1.3 Total 3.4 3.1 5.6 0.2 0.6 Southern African Customs Union Rice 2.3 2.4 -1.7 8.8 8.4 Wheat 1.9 1.9 -0.5 0.0 0.5 Other grains 1.1 1.3 0.8 29.5 19.9 Oil seeds 1.6 1.8 -1.8 9.2 8.6 Sugar 1.3 1.4 -0.2 87.7 50.6 Plant-based fibers 6.0 3.8 -35.9 3.4 3.6 Other crops 2.4 2.4 -8.7 7.2 4.3 Cattle 2.2 2.2 0.0 24.2 23.0 Other meats 2.2 2.2 0.1 5.0 5.1 Raw milk 2.2 2.2 0.0 -2.7 -2.6 Total 2.1 2.1 2.9 18.4 14.0 Rest of Sub-Saharan Africa Rice 3.2 3.2 -0.1 -1.2 -0.9 Wheat 3.4 3.5 0.4 0.3 3.0 Other grains 3.2 3.2 0.4 -0.1 3.0 Oil seeds 3.9 3.8 -0.8 51.0 37.7 Sugar 3.2 3.2 1.5 48.1 40.3 Plant-based fibers 8.1 6.5 -23.2 42.8 24.9 Other crops 4.5 4.2 -5.8 -3.6 0.0 Cattle 3.5 3.4 -1.4 4.6 3.5 Other meats 3.7 3.6 -1.9 -0.7 0.3 Raw milk 3.3 3.3 -1.1 1.7 1.2 Total 4.0 3.8 4.1 5.6 4.9 Source: World Bank simulations with LINKAGE model, based on release 5.4 of the GTAP data. 134 Global Agricultural Trade and Developing Countries global level in 1997 of 11.4 percent for plant-based The impacts of the agriculture and food trade fibers and 15 percent for rice. The demand for rice, reform were reassessed using two alternative elastic- however, is much less elastic than for plant-based ities.A low scenario uses trade elasticities 50 percent fibers. The lower supply elasticity would make land lower than the standard, and a high scenario uses relatively more costly, all else equal, and given the trade elasticities 50 percent higher than the stan- higher demand elasticities, the higher land prices dard (the standard values used in this study are will be reflected in lower demand from these three shown in table A3 on the CD-ROM). Each set of regions. assumptions requires two simulation runs. A new The impact of trade reform on agricultural out- baseline is constructed each time--with all assump- put using both the standard and the lower land tions identical except for the trade elasticities--and elasticities is broadly the same qualitatively, the reform scenario is simulated. Thus the compar- although lower in magnitude in general. Consider isons are between each individual baseline and each sugar again. Output increases 88 percent in SACU associated reform scenario. and 48 percent in the rest of Sub-Saharan Africa Within this range of trade elasticities the model using the standard supply elasticity. Sugar output exhibits some modest nonlinearity, particularly on expansion drops to 51 percent in SACU and 40 per- the upside (figure 7.6). For all three regions the cent in the rest of Sub-Saharan Africa when lower 50 percent higher elasticities lead to a greater than land supply elasticity is assumed. 50 percent rise in real income gains--particularly The welfare impacts are modest, but measura- for developing regions, where the rise is almost ble, and the results reflect only some of the possible 75 percent. On the downside, both high- and low- supply constraints in low-income countries. For income regions see an equiproportionate fall in the the three regions under question, aggregate welfare real income gains relative to the elasticities, with a would decline $1.1 billion compared with the stan- fall to 40 percent of the standard gains in the case of dard assumption and would drop from 1.2 percent the middle-income countries. The higher elastici- to 1.1 percent of baseline income. ties dampen the adverse terms-of-trade shocks from reforms, leading to the higher income gains. The global gains vary from a low of $126 billion to Trade Elasticities a high of $438 billion, with the gains at $265 billion The most critical parameter in trade reform scenar- using the standard elasticities. ios is trade elasticities. There is ongoing debate For some countries and regions the range of about their size. Most econometric evidence sug- results is much broader than at the aggregate level. gests that the Armington elasticities (measuring the For example, Mexico would lose some $1.2 billion degree of substitutability between domestic and with the low elasticities and gain $3 billion with the imported goods) are low, in the range of 1 to 2.19 high elasticities compared with a gain of 0.9 with The studies are riddled with data problems-- the standard elasticities. Several other regions show particularly the evaluation of unit values--and similar variation. The standard deviation of the many trade economists downplay the empirical index across all developing countries is 130 in the evidence, for two main reasons. First, low Arming- case of the high elasticities, whereas the weighted ton elasticities lead to implausible terms-of-trade average is 170. effects. And second, low elasticities would suggest The impacts on trade are similar to the impacts on high optimal tariffs. Trade studies fall into three income but exhibit more nonlinearity (figure 7.7). groups--those with relatively low elasticities (1­3), At the global level, exports increase 80 percent those with middling elasticities (3­6), and those using the high elasticities and decline 60 percent with very high elasticities (20­40). Examples of the using the low elasticities (with export increases first are the MONASH model (Dixon and Rimmer ranging from a low of $216 billion to nearly $1 tril- 2002) and the standard GTAP model (Hertel 1996). lion). There is also less variability across regions of Recent World Bank work has been using the mid- the model than with the income results. In isolation dling elasticities. High elasticities are mainly associ- the trade elasticities appear to have the greatest ated with the work of Harrison, Rutherford, and impact in determining the overall outcomes of Tarr (for example, see Harrison, Rutherford, and trade reform, although other model changes--both Tarr 2003). in specification and in elasticities--combined may Global Agricultural Reform: What Is at Stake? 135 FIGURE 7.6 Real Income and Trade Elasticities Index relative to default elasticities in 2015 200 175 150 125 100 75 50 25 0 High-income Low-income Middle-income Low Default High Source: World Bank simulations with Linkage model, based on release 5.4 of the GTAP data. FIGURE 7.7 Exports and Trade Elasticities Index relative to default elasticities in 2015 200 175 150 125 100 75 50 25 0 High-income Low-income Middle-income Low Default High Source: World Bank simulations with Linkage model, based on release 5.4 of the GTAP data. be at least as important in determining overall out- be much larger than the more familiar gains to real comes. This is an area of active research to better income. A decomposition of the aggregate results determine the bounds on the possible ranges for across policy instruments and regions shows that these elasticities. Better data would help, but there reforms in agriculture and food account for a large are still issues relating to model specification and share of the global gains of reforms of total mer- aggregation that need to be thought through. chandise trade. This result is driven by the relatively low protection levels in manufacturing sectors. Another major finding is that developing countries Conclusions have more to gain from reforming their own sup- This quantitative assessment of the impact of agri- port policies than from reforms in high-income cultural and food market distortions on incomes, countries. Symmetrically, high-income countries welfare, trade, and output shows that the changes in would experience larger welfare gains from their cross-regional patterns of output and trade tend to own reforms than from developing countries' 136 Global Agricultural Trade and Developing Countries reforms. These dimensions of the debate are often gains in rural value-added as do SACU and the rest overlooked but are crucial. Global reform leads to of Sub-Saharan Africa. Wages for unskilled labor in additive results with aggregate gains close to the developing countries are moderately influenced by gains from reforms in each group. A third key find- major policy reforms such as in China, where they ing is that agricultural reform alone in high- decrease, but more significantly in Argentina, income countries would create moderate gains, where they increase. about 10 times smaller than those of a combined reform of food and agricultural markets. Develop- Notes ing countries would be negatively affected as a group, because their own distortions would be 1. The model is based at the World Bank and uses the GTAP release 5.4 dataset (see van der Mensbrugghe 2003 for details). exacerbated by the agricultural reforms in high- The details of the modeling and the results are given in the income countries. attached CD-ROM. The results are broadly robust to changing 2. East Asia is divided into four economies--China, assumptions on future agricultural productivity in Indonesia, Vietnam, and the rest. South Asia has two compo- nents--India and the rest. Latin America has four economies-- developing countries, supply constraints, and level Argentina, Brazil, Mexico, and the rest. Europe and Central Asia of the trade elasticities, but the levels of the trade is split into three components--the European Union accession elasticities remain of foremost importance. The countries, Turkey, and the rest. Sub-Saharan Africa has two components--the Southern African Customs Union countries trade effects of reforms are also sensitive to and the rest, and the Rest of the World region has all other coun- assumptions about agricultural productivity gains tries including those in the Middle East and North Africa. in developing countries. Assuming low productiv- 3. Agricultural policies derived from the Agricultural Market Access Database (AMAD) reflect 1998­99 levels of support, ity gains leads to a reversal in the estimated impact except for cotton, for which International Cotton Advisory of global liberalization for industrial countries, Committee data were used (see chapter 14 in this volume). increasing their net food trade surplus as middle- 4. Income elasticities are held more or less constant over the income countries become much larger importers time horizon. With China's rapid growth, one might anticipate a convergence of income elasticities toward levels in higher- of food and agricultural products. Low-income income countries and thus a dampening of food growth over countries experience an increase in net food trade time relative to incomes. surplus that is much smaller than under the higher 5. Textile and apparel quotas that generate quota rents for exporters are converted to export taxes (for the country of ori- productivity assumption. Hence, variations in pro- gin). In the current simulations, these have not been eliminated. ductivity could lead to a different assessment of the 6. Technically, it is a measure of the Hicksian equivalent direction of food self-sufficiency after reform. variation. When comparing aggregate welfare measures across studies, it is important to convert them to similar scales. Thus Supply constraints do not qualitatively affect the $350 billion in 2015 is more or less equivalent to $250 billion in estimated impact of trade reform on agricultural 2004 and $200 billion in 1997--assuming an average annual output, although estimated changes tend to be global growth rate of 3 percent in gross domestic product (all in smaller. Higher trade elasticities dampen the 1997 US$, the base year of release 5 of the GTAP data set). Assuming a world inflation rate of 2.5 percent over the entire adverse terms-of-trade shocks from reforms, lead- period, the measured $250 billion in 2004 in 1997 dollars ing to higher income gains. The global gains vary becomes $300 billion in 2004 dollars. from a low of $126 billion with low elasticities to a 7. The Mercosur preferential agreement is not incorporated in the standard GTAP dataset but is included in the dataset used high of $438 billion with high elasticities, with the for these simulations. Efforts were made to minimize distortions gains at $265 billion using the standard elasticities. to the original social accounting matrix (SAM) while adjusting There is also higher variation at the individual the original dataset. 8. There is also an issue regarding whether bound or applied country level. tariffs are liberalized. Most developing countries have bound The changes in agricultural value added and fac- their tariffs at rates much higher than applied rates. Negotiations tor prices are considerable in several cases. The esti- concern the bound tariffs; the reforms described here are relative mated loss of rural value-added is large in Japan to the applied tariffs. For a full reform scenario, it is not much of an issue, but for analyzing potential outcomes of a negotiation, it and the European Union, the Republic of Korea, could be. Taiwan (China), and China. Thus, considerable 9. See Global Economic Prospects 2002 and 2004 (World Bank adjustment and displacement of resources would 2001, 2003). The 2002 report notes dynamic gains of $830 billion compared with static gains of $350 billion, with a range take place to reflect these changes. Cairns Group of up to $1,340 billion depending on some key parameters countries and the United States experience sizable (table 6.2, page 100). Global Agricultural Reform: What Is at Stake? 137 10. Figure 7.1 shows output shares in the base year. One 18. The elasticity measures the ease of shifting land from one would assume that the agricultural and food shares are declining activity to another when the relative price of these two activities over time as income elasticities for food tend to be lower than changes. for other goods and services. 19. More recent econometric work is resulting in higher esti- 11. While the model is highly nonlinear, the results to a close mates for the trade elasticities, and these are now being reflected approximation are relatively additive. in the forthcoming release of the GTAP dataset. 12. Nominal values are measured with respect to the model's numéraire--the average export price of manufactured exports from industrial countries. References 13. This should be considered an upper bound on China's potential loss since the baseline scenario does not include the Dixon, P. B., and M. T. Rimmer. 2002. Dynamic, General Equilib- impacts of China's accession to the WTO. Thus the reform sce- rium Modelling for Forecasting and Policy: A Practical Guide nario is capturing the combined gains from global reform and and Documentation of MONASH. Boston: North-Holland. China's WTO accession, which include the gains to be had from Harrison, G. W., T. F. Rutherford, and D. Tarr. 2003. "Trade reforming from 1998­99 base agricultural policies. Liberalization, Poverty and Efficient Equity." Journal of 14. Sector-specific capital returns may be possible during the Development Economics 71(1): 97­128. transition phase, as sectors in decline shed unwanted capital. Hertel, T. W. 1996. Global Trade Analysis: Modeling and Applica- The most mobile equipment will be shed first, and the return to tions. Cambridge, U.K.: Cambridge University Press. the remaining capital may be priced lower than the national rate Martin, W., and D. Mitra. 1996. "Productivity Growth in of return to capital. Agriculture and Manufacturing." World Bank, International 15. In the default version of the model, cross-sectoral trans- Economics Department, Washington, D.C. formation elasticities are set to 3. Thus a 10 percent rise in the ------. 1999. "Productivity Growth and Convergence in return in one sector (relative to the others) will lead to a 30 per- Agriculture and Manufacturing." Country Economics cent shift of land into that sector. Because of the finite transfor- Department Working Paper 2171. World Bank, Washington, mation elasticity, land prices are sector specific. D.C. 16. Given the aggregate nature of the model, the impacts on van der Mensbrugghe, D. 2003. "LINKAGE Technical Reference vulnerable countries or sectors are harder to assess. In particular, Document." Working Paper. World Bank, Economic Sub-Saharan Africa is a heterogeneous subcontinent that is not Prospect Group, Washington, D.C. reflected in the level of aggregation of this study. World Bank. 2001. Global Economic Prospects and the Developing 17. The land supply function is governed by a logistic curve. Countries 2002: Making Trade Work for the Poor. Washington, It is calibrated in the base year to an exogenously given elasticity D.C.: World Bank. and the value of the asymptote relative to the base supply level. ------. 2003. Global Economic Prospects 2004: Realizing the Thus if the asymptote is set to 1.2, land supply can increase by at Development Promise of the Doha Agenda. Washington, D.C.: most 20 percent above its base level. World Bank. Part II The Commodity Studies 8 SUGAR POLICIES: AN OPPORTUNITY FOR CHANGE Donald O. Mitchell Sugar protection dates back to at least the 1800s. It fallen from 2.5 million tons to 1.5 million tons over has been greatest in countries of the northern hemi- the past two decades. Thus, the three largest sphere that produce sugar beets. That is because markets for sugar imports in the 1970s have been sugar from beets is nearly twice as expensive to pro- closing to competition after becoming largely self- duce as sugar from cane, and most beet producers sufficient, at least compared with the early 1970s, cannot survive without high protection. Over the when their combined net imports accounted for years, high protection has led to lower consump- half of the world's exports (figure 8.1). tion, reduced imports, and surplus production, which is disposed of in the world market at Background subsidized prices. Many other countries have been pressured by their producers for protection from Sugar occurs naturally in most foods, but it is eco- heavily subsidized exports and depressed world nomically extracted from only a few crops such as market prices. The cycle of protection, subsidies, sugar beets, sugar cane, and corn. Sugar beets are an and more protection has run for decades. annual root crop grown in temperate climates, The European Union (EU), Japan, and the while sugar cane is a tall perennial grass grown in United States are among the areas with the highest tropical and semitropical climates. About 55 coun- level of protection and therefore the most distorted tries grow sugar beets, and 105 grow sugar cane. import patterns. Since the early 1970s, U.S. sugar The process of producing sugar (sucrose) from imports have declined from more than 5 million sugar beets or sugar cane requires that the juice be tons per year to slightly more than 1 million tons extracted and processed in a factory near where the per year. The European Union was a net importer beet or cane is grown. The by-products of sugar of about 2.5 million tons of sugar in the early cane are bagasse and molasses. Bagasse is the 1970s, compared with net exports of about 5 mil- residue of cane, after the juice is extracted. It has lion tons in recent years. Japan's sugar imports have some industrial uses and is often used to fuel the The author wishes to express thanks for useful comments to Ataman Aksoy, John Beghin, Margaret Blamberg, Uri Dadush, Harry de Gorter, Stephen Haley, Steven Jaffee, Will Martin, John Nash, and Albert Viton. 141 142 Global Agricultural Trade and Developing Countries FIGURE 8.1 World Sugar Exports and Net Imports of Selected Countries (million of tons) Million metric tons 50 40 World exports 30 20 10 Net imports of European Union, Japan, and United States 0 10 1965 1970 1975 1980 1985 1990 1995 2000 Source: FAOSTAT. boilers in the sugar factory (also called a sugar mill). cane and sugar beets. Dextrose is a sugar derived Molasses is an edible by-product as well as an ani- synthetically from starch (most commonly corn mal feed. The by-products of sugar beets are beet starch). Fructose is a very sweet sugar derived from tops, the leafy portion of the beet used for animal dextrose. High-fructose corn syrup (HFCS) is pro- feed, and molasses, which is also used primarily as duced by the enzymatic conversion to fructose of a an animal feed. Once harvested, sugar cane is portion of the dextrose in corn syrup. It is chemi- highly perishable and must be processed quickly. cally similar to sugar used in soft drinks, which is a Sugar beets are less perishable than sugar cane but mixture of equal parts of dextrose and fructose. still must be processed soon after harvest. The high The fact that identical or nearly identical sugars can cost of transporting sugar beets or cane makes it be produced from different crops provides produc- impractical to locate the factory far from the pro- ers and consumers with a wide range of substitu- ducing areas. tion possibilities. It also means, however, that sugar Sugar growers and processors are economically policies are often complex, as the different indus- interdependent and normally share in the value of tries vie for support. For example, sugar producers total sugar and molasses sales according to a con- in the European Union have been able to get legis- tractual agreement. Both can influence the value of lated quotas on HFCS production. Japan also limits total output since the volume and sugar content of HFCS production to prevent it from further erod- sugar beets or cane is affected by input use and pro- ing sugar's market share. In the United States, duction practices, and the recovery of sugar from HFCS producers benefit from high sugar prices and beets or cane is dependent on the technology and support current sugar policies. operation of the sugar factory. Various ownership High protection has led to the emergence of arrangements exist in the industry--ranging from HFCS as a substitute for sugar in the United States ownership by a single company of the factory and and Japan. Because it is a nearly perfect substitute producing lands to independent growers who con- for sugar in uses such as soft drinks, HFCS and tract production with a factory. Some growers are other corn syrups now account for 40 percent of members of cooperatives, which own and operate a caloric-sweetener use in Japan and more than half sugar factory. State ownership of factories and lands of U.S. caloric sweetener consumption (figure 8.2). is still common in developing countries, but sub- The technique for commercial production of high- stantial privatization has taken place in recent years. fructose corn syrup was discovered in the late 1960s Common sugar is sucrose. It is extracted in and made profitable by high sugar prices in the nearly pure, chemically identical form from sugar protected Japanese and U.S. markets. But now, Sugar Policies: An Opportunity for Change 143 FIGURE 8.2 U.S. Sugar and HFCS Consumption Million metric tons 12 10 Sugar 8 6 HFCS 4 2 0 1970 1975 1980 1985 1990 1995 2000 Source: USDA. economies of scale, improvements in production renewable sources such as corn. The legislation techniques, and large installed production capacity provided tax incentives for ethanol, amounting to (financed under high prices), have made corn $0.54 cents a gallon when blended with gasoline at syrups competitive with sugar from cane and less a 10 percent rate. Some Midwestern states provide costly than sugar from beets. additional tax incentives. Another product that can be produced from The cost of ethanol production from corn is sugar beets, sugar cane, and corn (and some other about $1.10 per gallon, but because ethanol contains crops) is ethanol--a clear colorless, flammable, less energy than gasoline, the comparable energy- oxygenated hydrocarbon that can be used for a equivalent cost is $1.65 per gallon (Oregon Office of number of purposes, including as a vehicle fuel, a Energy 2002). Thus, with the $0.54 tax incentive, use that accounts for about two-thirds of world ethanol is competitive with regular gasoline. In ethanol consumption. It is normally more costly response to the incentive, U.S. ethanol production than petroleum-based fuels, however, and is used has been growing by about 6 percent per year (Berg only when special incentives, such as environmen- 2001). Both HFCS and ethanol can be produced in tal regulations or government subsidies encourage the same facility by adding an ethanol unit to an its production and use. Ethanol can be produced HFCS facility, so the tax incentive on ethanol partly from crude oil, ethylene, and coal, or from agricul- finances the facilities that produce HFCS. A seasonal tural products. Roughly 60 percent of global complementarity between ethanol and HFCS pro- ethanol production comes from sugar cane and duction is also possible because ethanol is used for sugar beets. In Brazil, half of sugar cane production fuels primarily during the winter months, whereas is used for ethanol production; government ethanol the demand for HFCS in soft drinks increases dur- policies mandate the share of ethanol to be blended ing summer months. The U.S. ethanol policy con- with gasoline. tributes to production capacity, which also can be The United States is a major producer and con- used for HFCS production, thereby reducing HFCS sumer of ethanol from corn. Ethanol has environ- production costs and making HFCS more competi- mental advantages when used as a fuel, because it tive with sugar. burns cleaner than gasoline and does not produce greenhouse gases. In 1990 amendments to the U.S. Estimates of Production Costs Clean Air Act required certain U.S. regions to use oxygenated, reformulated gasoline during certain Although the costs of producing sugar vary among high-smog months and stipulated that a certain countries for a variety of reasons, it is cheaper percentage of oxygenates must be derived from to produce it from cane than from beets in all 144 Global Agricultural Trade and Developing Countries countries.LMC International,a London-based con- on board), while costs are exfactory; thus the prices sulting firm, periodically estimates production costs should be higher. for cane sugar, beet sugar, and HFCS. The firm's The average cost of producing raw cane sugar by most recent estimates cover 41 beet-producing major exporters was 10.39 U.S. cents per pound in countries, 63 cane-producing countries, and 1994­99, while the average cost of refined cane 19 HFCS-producing countries (table 8.1). LMC sugar was 14.25 cents per pound. Thus the raw-to- bases its estimates on an engineering cost approach white spread averaged 3.86 cents per pound. that accounts for the physical inputs of labor, Refined sugar from beets cost an average of machinery, fuel, chemicals, and fertilizers used in 25.31 cents per pound--78 percent more than field and factory. The estimates are of actual average refined cane sugar. Among low-cost producers, the costs and include the impact of policies that protect difference between refined cane and beet sugar was producers in certain countries. Such cost estimates even wider. The average production cost for low- do not represent the supply curve normally esti- cost producers of refined cane sugar was 11.44 cents mated by economists, since they are not estimates of per pound, compared with 22.29 cents per pound marginal costs. Nevertheless, they are useful for for refined beet sugar--a difference of 95 percent. comparing the average costs of production of dif- Based on this comparison, sugar from beets was not ferent products. Actual raw cane sugar prices are competitive with sugar from cane by either major provided for comparison. The prices are f.o.b. (free exporters or low-cost producers. However, the wide TABLE 8.1 Average Costs of Producing Cane Sugar, Beet Sugar, and High-Fructose Corn Syrup by Categories of Producers, and Actual Sugar Prices, 1994­1999 (nominal U.S. cents per pounda) Category 1994­95 1995­96 1996­97 1997­98 1998­99 Raw cane sugar Low cost producersb 7.43 8.10 8.18 7.78 7.58 Major exportersc 10.37 10.60 10.72 10.52 9.73 Cane sugar, white equiv. Low cost producersb 11.02 11.75 11.84 11.41 11.19 Major exportersc 14.23 14.48 14.61 14.38 13.53 Beet sugar, refined Low cost producersd 21.31 23.16 23.09 21.21 22.67 Major exporterse 25.47 26.87 25.90 23.56 24.75 High-fructose corn syrupf Major producersg 13.45 16.78 13.57 12.86 11.76 Actual market prices Raw cane sugarh 13.53 12.23 11.21 10.71 7.05 a. Exfactory basis. b. Average of 5 producing regions (Australia, Brazil­Center/South, Guatemala, Zambia, and Zimbabwe). c. Average of 7 countries (Australia, Brazil, Colombia, Cuba, Guatemala, South Africa, and Thailand). d. Average of 7 countries (Belgium, Canada, Chile, France, Turkey, United Kingdom, and United States). e. Average of 4 countries (Belgium, France, Germany, and Turkey). f. HFCS-55, dry weight. g. Average of 19 countries (Argentina, Belgium, Canada, Egypt, Finland, France, Germany, Hungary, Italy, Japan, Mexico, Netherlands, the Slovak Republic, the Republic of Korea, Spain, Taiwan (China), Turkey, United Kingdom, and United States). h. Raw cane sugar price is U.S. cents per pound, July-June average of monthly prices, f.o.b. Caribbean ports. Source: LMC International (1999). Actual market prices are from World Bank databases. Sugar Policies: An Opportunity for Change 145 margin between refined sugar from beets and cane industry employee for those countries is estimated is partly a reflection of protection to sugar beet pro- to range from 16.3 tons to 19.9 tons. In contrast, ducers in the European Union and United States, countries known to be high-cost producers such as which encourages production in marginal areas Fiji, Kenya, and Mauritius have production of 7.0 to and contributes to higher average costs. 8.3 tons of raw sugar per industry employee. Thus, Production costs for HFCS-55 (55 percent fruc- one can reasonably conclude that an additional tose) averaged 13.68 cents per pound and were million tons of sugar production from a low-cost lower than white sugar from cane produced by sugar-producing country would generate about major exporters in four of the five years. They 55,500 direct employment jobs. If the exports came exceeded the cost of cane sugar only when corn from a high-cost producer, the same million tons of prices rose sharply in 1995­96. Thus, HFCS-55 can production would generate about 128,000 direct compete with refined cane sugar in the current pol- employment jobs. Additional indirect employment icy environment, and perhaps even in a fully liber- jobs would also be generated in transportation and alized market environment, since many studies related industries, but no attempt was made to esti- have suggested that raw sugar prices would rise mate these jobs. more than corn prices under liberalization. The World Sugar Market Employment Brazil, the European Union, and India are the Data on employment in developing countries' largest sugar producers, each accounting for sugar industries are not readily available but can be roughly 14 percent of world production during estimated from reports, surveys, and industry state- 1999­2001 (table 8.3). They are followed by China ments. Such estimates (table 8.2) show consider- and the United States, which each produce about able cross-country consistency among high- and 6 percent of the world's sugar. Sugar trade is domi- low-cost producers. For example, Brazil, Guyana, nated by Brazil and Russia, with Brazil accounting and South Africa are known to be among the for about one-quarter of world net exports and lowest-cost producers; raw-sugar production per Russia accounting for about 14 percent of world TABLE 8.2 Raw Sugar Produced Annually per Sugar Industry Employee, Selected Developing Countries Tons of Raw Tons of Raw Direct Employment Sugar Produced, Sugar Produced Country (Growers and Factory) Average 1999­2001 Per Employee Low-cost producers Brazil 1,100,000 19,485,000 17.7 Guyana 18,000 293,072 16.3 South Africa 130,000 2,589,667 19.9 High-cost producers Fiji 40,500 336,333 8.3 Kenya 69,000 485,333 7.0 Mauritius 65,000 529,299 8.1 Other producers Malawi 17,000 200,667 11.8 Mexico 300,000 5,069,233 16.9 Note: Production is the three-year average of raw sugar production during 1999­2001 from FAOSTAT. Source: Employment figures are derived from various sources and include total direct employment in sugar factories and plantations. Employment data for Brazil, Mexico, and South Africa are from OECD (2002a); Fiji, Guyana, and Mauritius data are from F. O. Licht (2002); Kenya data are from the Kenya Sugar Board; Malawi data are from the Malawi Ministry of Commerce and Industry. 146 Global Agricultural Trade and Developing Countries TABLE 8.3 Major Sugar Producers, Net Exporters, and Net Importers, 1999­2001 Average Producers Net Exporters Net Importers Millions Millions Millions Country/Region of Tons Country/Region of Tons Country/Region of Tons India 19.4 Brazil 9.3 Russia 5.2 European Union 18.6 European Union 4.2 Indonesia 1.7 Brazil 18.5 Australia 3.8 Japan 1.6 United States 7.9 Thailand 3.6 United States 1.4 China 7.8 Cuba 3.2 Korea, Rep. of 1.2 Thailand 5.4 South Africa 1.3 Canada 1.2 Mexico 5.1 Guatemala 1.1 Iran 1.0 Australia 4.9 Colombia 1.0 Malaysia 1.0 Cuba 3.8 Turkey 0.6 Algeria 0.9 Pakistan 3.0 Mauritius 0.5 Nigeria 0.7 All other 38.9 All other 10.3 All other 20.7 World 133.3 World 38.9 World 36.6 Note: Data are in raw sugar equivalents. Source: USDA Production Supply and Distribution (PS&D) 2002. net imports during 1999­2001. The European tries have reformed their policies. As the share of Union is the second largest net exporter, followed those countries in global consumption and imports by Australia, Cuba, and Thailand, which each has increased with population and income growth, export about 8­10 percent of the world total. Net the reformed policies have led to greater price imports are widely dispersed after Russia, with the responsiveness by sugar producers and consumers, next largest net importer accounting for less than likely reducing the severity of future price spikes. 5 percent of world imports. India is the largest The collapse of the former Soviet Union also led to sugar consumer, with about 15 percent of world the abandonment of dedicated sugar imports from consumption, followed by the European Union Cuba and increased trade at world market prices. with 10 percent, and Brazil with 7 percent. Many developed countries still maintain highly World HFCS production averaged 11.7 million protected sugar sectors and thus contribute to the tons (dry weight basis) during 1999­2001. Produc- likelihood of price spikes, but they now account for tion in the United States alone averaged 9.2 million only one-third of consumption and one-half of tons--79 percent of the world total. Japan was the imports--compared with slightly more than half of second-largest producer, with an average of .78 mil- consumption and 60 percent of imports when the lion tons, followed by Argentina, Canada, European last sugar price spike occurred in 1980. Union, Mexico, and Republic of Korea with be- Despite some liberalization of sugar policies, tween .3 and .4 tons each. HFCS is considered roughly 80 percent of world sugar production and equivalent to sugar on a dry weight basis when used 60 percent of world sugar trade is at subsidized or to produce products such as soft drinks. protected prices. Only three major producers World sugar prices have historically been char- (Australia, Brazil, and Cuba) have sugar sectors that acterized by periodic sharp increases followed by produce and operate at world market price levels.1 long periods of low or declining prices. This pattern These three producers account for a combined has been caused, in large part, by policies in both 20 percent of world production and 40 percent of developed and developing countries that isolated world trade. The remaining 80 percent of world consumers and producers from international prices production and 60 percent of world trade relies on and diminished their price responsiveness. Since production subsidies, export subsidies, or preferen- the early 1980s, however, some developing coun- tial access to protected markets. The European Sugar Policies: An Opportunity for Change 147 Union, Japan, and the United States account for exports are viewed as the third alternative for 20 percent of world production; their average Brazilian sugar cane after production of fuel producer prices are more than double the world ethanol and sugar for the large domestic market. market. China and India account for another Only half of Brazil's sugar cane is used to produce 20 percent of world production and protect pro- sugar; the other half goes into ethanol for automo- ducers with prices that are higher than world mar- tive fuel. Sugar cane can easily be divided between ket prices. The remaining 40 percent of production sugar and ethanol production depending on mar- is in countries that either produce for preferential ket conditions and government policies. If all of markets (as is the case with Fiji, Mauritius, the Brazil's sugar cane were used to produce sugar, pro- Philippines, and many African and Caribbean duction could roughly double (an increase of countries) and thus receive prices higher than those roughly 18.5 million tons per year). Most of the of the world market, or they protect their domestic increase could be exported, subject to port and producers with policies that restrict imports to milling capacity. provide above-market prices. The Brazilian government has pursued a biofuel The value of world sugar exports has remained policy since the 1970s, when concerns about the relatively constant in nominal dollars ($11.8 billion adequacy of petroleum supplies were high. These during 1980­85; $11.6 billion during 1995­2000), policies included tax incentives and direct subsidies and sugar has remained an important source of for ethanol production and use, sugar price con- export earnings for some developing countries. trols, and restrictions on sugar exports. Lower However, the share of developing countries in total petroleum prices during the 1980s led to reduced sugar exports declined from 71 percent during ethanol subsidies and the removal of export and 1980­85 to 54 percent in 1995­2000, as developed- price controls on sugar beginning in 1990. Other country exports increased and the share of higher- controls on sugar were eased during the 1990s, and valued refined-sugar exports by developed countries sugar exports increased from 1.5 million tons in increased. Twelve countries received 10 percent or 1990­91 to 11.3 million tons in 2000­01. Some more of their total export earnings from sugar subsidies remain on ethanol production and use, during 1995­2000, and an additional five received and the future of such subsidies can strongly influ- 5­10 percent. In contrast, during 1980­85, ten ence the use of sugar cane for ethanol versus sugar countries received 20 percent or more of total production. Government mandates on the share exports from sugar, and nine additional countries of ethanol to be included in gasoline (currently received from 5­20 percent. 20­24 percent) can strongly influence demand for ethanol as automotive fuel and the supplies of sugar cane directed to sugar production. The future Sugar Policies in Selected of the biofuel program depends on international Developing Countries petroleum prices as well as Brazilian policy. Although this chapter focuses on prospects for pol- Recently marketed flex-fuel automobile engines icy reform in the European Union, Japan, and the that run equally well on gasoline or pure hydrous United States, it is useful to examine policies in alcohol are expected to boost ethanol demand and other major sugar-producing and -trading coun- direct some sugar cane production away from sugar tries to see how they would be affected by such production and exports. reforms. Sugar cane production has increased rapidly in Brazil--the world's largest sugar exporter the center-south region of Brazil, where the climate and generally considered to be its lowest-cost is favorable, land is available, and sugar cane yields producer--would be a major beneficiary of good returns relative to other crops. Further expan- increased world sugar trade and higher prices sion of sugar cane production in the center-south because it has the capacity to increase sugar pro- region is possible and expected by most industry duction and exports substantially. The devaluation experts, but milling capacity will need to be ex- of the Brazilian real by 65 percent relative to the panded to allow significantly more sugar produc- dollar since 1998 has contributed to the country's tion. Sugar is also produced in the northeast competitiveness. Despite its dominance, however, region, where high-cost growers receive a small 148 Global Agricultural Trade and Developing Countries subsidy. The central government allocates Brazil's India's sugar industry, heavily regulated under total annual quota of premium-priced U.S. imports the Essential Commodities Act of 1955, is very to this region. politicized because of the large number of sugar China was an occasional large sugar importer cane growers (reportedly as many as 5 million) and and exporter in the 1990s, but average net imports the importance of sugar in Indian diets. The indus- were about 400,000 tons during 1990­2000. Most try is largely self-sufficient, with occasional imports of these imports came from Cuba under a long- to offset domestic shortfalls. An import duty (cur- term trade agreement. The government has fol- rently 60 percent) is varied to maintain domestic lowed a policy aimed at self-sufficiency by provid- prices above those of the world market. Large ing strong price incentives to producers, controlling stocks of sugar currently burden the industry and imports, and accumulating and releasing govern- can only be exported with substantial subsidies or ment stocks to maintain high internal market at substantial losses. India provides an internal prices. About 90 percent of China's sugar produc- freight reimbursement and ocean freight subsidy to tion comes from sugar cane and the remainder help export surplus production. State controls limit from sugar beets. A "guidance price" is provided to internal sugar movements, and licensing and stock- sugar refiners for sugar cane and beet, but market holding requirements for mills and shops con- forces largely determine prices (Sheales and others tribute to industry inefficiencies. Sugar mills are 1999). The policy and strong demand growth kept small and inefficient, and high internal transport sugar prices high during most of the 1990s, but costs would limit export potential even if world prices fell sharply after the record 1998­99 crop, prices were to rise above internal prices. Sugar remaining low through 2000. Prices increased in millers and importers are required to sell a portion 2001, with white wholesale sugar prices averaging of their supplies to the Public Distribution System about $0.22 per pound during the first half of 2001 at below-market prices for resale to low-income (F. O. Licht 2002), more than double the world consumers. Sugar-cane production is more prof- market price and similar to U.S. domestic prices. A itable than most other crops, with prices that are record 2002­03 harvest caused prices to fall again. about 50 percent higher than world market prices Artificial sweeteners, mainly saccharin, are an due to minimums established by the central gov- important competitor to sugar in China and substi- ernment and higher prices advised by the states. tute for as much as 2.4 million tons. When China India has a small ethanol program, and there are entered the World Trade Organization (WTO) in government proposals to require ethanol to be 2001, it agreed to a tariff rate quota of 1.6 million blended with gasoline to reduce pollution. The gov- tons of sugar at a tariff rate of 20 percent, with an ernment has announced plans to liberalize the sec- over-quota tariff of 76 percent. The quota is sched- tor, but past efforts at liberalization have been uled to increase to 1.945 million tons, and the over- unsuccessful. Decades of regulation have also cre- quota rate to fall to 65 percent, by 2004. If China ated complicated political interdependencies that were to import the full amount specified by the tar- will be difficult to disentangle. It is unlikely that iff rate quota, imports would increase substantially India would emerge as a significant exporter even if over the levels of recent years. China's WTO tariff policies in the European Union, Japan, and the quota does not commit the country to import all of United States were changed to allow greater imports. the quota tonnage, however, and China can choose Mexico privatized its sugar mills and partially among a number of different methods of adminis- deregulated its sugar industry in reforms that con- tering the quota to influence its fill rate (Jolly 2001). cluded in 1992 (Escandon 2002). It has maintained For example, actual imports during 2001­02 were strong government regulation of the sector, how- 1.15 million tons, according to the International ever, by setting sugar-cane prices for its 150,000 Sugar Organization (2002), despite the tariff rate growers. Mexico liberalized pricing and production quota of 1.6 million tons. The Chinese sugar indus- of sugar in 1995 but simultaneously increased pro- try would undergo substantial adjustment if it were tection by increasing tariffs from 65 percent to opened to international competition. A large num- 136 percent on raw sugar and from 73 percent to ber of small, high-cost sugar mills would become 127 percent on refined sugar. This led to a 60 per- unprofitable, and production would likely decline. cent increase in domestic sugar prices, a 50 percent Sugar Policies: An Opportunity for Change 149 increase in production, and a doubling of exports marketing year are estimated to total 650,000 tons, from 1992 to 2002. of which 148,000 tons were exported to the United The North American Free Trade Agreement States duty-free. Beginning in 2009, Mexico will (NAFTA) came into force on January 1, 1994. A have unlimited duty-free access to the U.S. sugar 15-year adjustment period ending in 2008 was to be market and will likely increase exports substantially. followed by free trade in sugar between Mexico and The Russian Federation is by far the world's the United States. The implementation of NAFTA largest sugar importer, with average annual imports has been contentious because of a last-minute side- of 5.2 million tons during 1999­2001, three times letter agreement on sugar added to ensure approval the amount of the next largest importer. Following by the U.S. Congress.Although the side-letter agree- the breakup of the Soviet Union, the Russian sugar ment was never ratified by Mexico's Congress and is sector faced an uncertain future, an unstable and not recognized as valid by Mexico, the U.S. govern- confused policy structure, and a technically weak ment administers NAFTA in accordance with its industry. Sugar production, all from beets, declined terms. Under NAFTA, the amount of Mexico's duty- by about 45 percent from 1992 to 2000, while con- free access to the U.S. sugar market depends on sumption declined by 17 percent and sugar imports whether Mexico is a surplus sugar producer (sugar increased by 35 percent. Low beet yields, poor production minus sugar consumption). The side- factory recovery rates, outdated technology, and letter agreement changed the definition of surplus shortages of fuel and replacement parts hampered producer to include HFCS consumption as well as the adjustment of the Russian sugar industry to sugar consumption. Using this definition, Mexico privatization. With trade policy changing fre- could export up to 25,000 tons per year of surplus quently, high perceived risks discouraged foreign sugar duty-free during the first 6 years of NAFTA. direct investment and slowed the modernization of Beginning in year 7 (the 2000­01 marketing year), the industry. and until the end of the 15-year adjustment period, The government uses high tariffs to protect the Mexico could export up to 250,000 tons of surplus domestic industry. To protect domestic sugar refin- sugar duty-free. ers, tariffs on white sugar are higher than on raw High prices for sugar in Mexico led to large sugar. Seasonal tariffs are added during periods of imports and increased production of HFCS, which peak domestic production to protect local produc- quickly displaced sugar in the soft-drinks industry ers and support prices. The import duty on raw and left Mexico with large sugar stocks that could cane sugar for 2003 has been set at $95 per ton not be exported duty-free to the United States ($.043 per pound). Russia is expected to remain a because of the 25,000 ton limit. After the United large importer as long as the investment climate States rejected a request to allow increased duty- remains uncertain and foreign companies are free exports, Mexico charged that the United States reluctant to invest. Even with foreign investment, was dumping HFCS in Mexico and initiated Russia will likely remain a high-cost producer antidumping duties. Negotiations are continuing to because its industry is based on beets. resolve the trade and duties on HFCS. Thailand is the world's fourth-largest sugar Caught between the high prices that the govern- exporter, with net exports of 3.6 million tons dur- ment had established for sugar cane and the weak ing 1999­2001 (annual average). Thailand's sugar domestic and world market prices for sugar, many policy is patterned after that of the European of Mexico's 60 sugar mills became insolvent. The Union, with high internal sugar prices maintained government expropriated 27 mills with large and by quotas and import tariffs. The government uses unpayable debts in September 2001. Public invest- production quotas, tax incentives, and subsidized ments are being made to prepare these mills for credit to encourage exports. The tariff rate quota resale to private investors. agreed under the WTO Agreement on Agriculture Among the measures in Mexico's national sugar was 65 percent for in-quota imports in 1999 and policy for 2002­2006, which is designed to make the 99 percent for outside-quota imports (Sheales and sector profitable, is the formation of an export others 1999). Despite high protection, Thailand's cooperative of all private and government-owned costs of production are among the lowest in the sugar mills. Mexico's sugar exports in the 2001­02 world, roughly comparable to those of Australia 150 Global Agricultural Trade and Developing Countries (Borrell and Pearce 1999). High protection and low Sugar Policies in Selected OECD costs have led to rapid growth of production and a Countries more than tripling of exports over the past two More than half of the value of sugar production in decades. OECD (Organisation for Economic Co-operation This selective review of policies in major sugar- and Development) countries during 1999­2001 producing and -trading countries illustrates the came from government support or transfers from significance of policy distortions in the world sugar consumers. Such high support typically limits con- market. India, the largest sugar producer, has a sumption through high prices and encourages pro- heavily regulated domestic sugar market and high duction even when a country does not have a com- import tariffs to protect local producers. China's parative advantage in sugar production. Support to import restrictions keep domestic sugar prices OECD sugar producers during 1999­2001 totaled nearly as high as those in the United States. Russia, $6.35 billion, more than half the value of world the largest net importer, has high tariffs to protect sugar trade (about $11.6 billion) and nearly equal sugar-beet producers and additional tariffs on to developing-country exports of about $6.5 bil- white sugar to protect local refiners. Brazil, the lion. The European Union provided the largest largest sugar exporter, has a sugar policy that is annual support, with $2.71 billion, while the partly driven by its biofuel policies; until recently it United States provided $1.30 billion, and Japan restricted sugar exports. Thailand, the fourth- provided $0.44 billion. Several developing coun- largest net exporter and a low-cost producer, has tries also provided high levels of support to sugar used high domestic prices, tax incentives, and sub- producers, including Mexico, Poland, and Turkey sidized credit to increase exports. Mexico's high (table 8.4). Much of that support is provided domestic prices have stimulated production in through border protection. anticipation of unlimited duty-free access to the The benefits of more liberalized trade in sugar U.S. sugar market beginning in 2009. and reduced domestic support, especially in OECD TABLE 8.4 Government Support to Sugar Producers, 1999­2001 Producer Support Producer Nominal Support from Border Country/Region (million US$) Assistance Coefficient Protection (percent) OECD 6,351 2.11 n.a. Australia 51 1.11 0.0 Czech Republic 16 1.25 47.6 European Union 2,713 2.11 91.7 Hungary 12 1.20 41.5 Japan 437 2.17 88.7 Mexico 713 2.10 83.9 Poland 176 2.28 92.9 Slovak Republic 16 1.94 54.7 Switzerland 86 4.36 73.0 Turkey 749 3.02 95.8 United States 1,302 2.37 84.3 n.a. Not applicable. Note: Producer support was converted from local currency to U.S. dollars using period average annual exchange rates from the IMF's International Financial Statistics, May 2002. Producer nominal assistance coefficient is an indicator of the nominal rate of assistance to producers measuring the ratio between the value of gross farm receipts including support and gross farm receipts valued at world market prices without support. No calculations were made for Canada, Iceland, New Zealand, Norway, or the Republic of Korea. Source: OECD 2002b. Sugar Policies: An Opportunity for Change 151 countries, are substantial, according to several stud- Europe; 25 percent in Eastern Europe, Indonesia, ies (Borrell and Pearce 1999; Devadoss and Kropf Mexico, and the United States; and 10 percent in 1996; Elbehri and others 2000; El-Obeid and China, the Philippines, and Ukraine. Lower prices Beghin 2004; USGAO 1993 and 2000; Sheales and would lead to higher consumption, lower produc- others 1999; USITC 2002; van der Mensbrugghe, tion, and increased imports of sugar in those coun- Beghin, Mitchell 2003; Wohlgenant 1999). Study tries that had trade protection. World prices would results differ because of different assumptions, increase by 38 percent, and lower-cost producers methodologies, and scenarios, but the general con- would increase production and exports--however clusion is that reduced support to OECD sugar consumption decreased from the higher prices. In producers would result in lower production in countries with the highest protection (Europe, those countries, lower domestic prices, increased Indonesia, Japan, the United States), net imports consumption, and increased net imports. World would increase by 15 million tons per year. Japan's sugar prices would increase and exports from production would drop by 44 percent, that of the developing countries, and some developed-country United States by 32 percent, and Western Europe's exporters, would rise. According to Sheales and by 21 percent. Among low-cost producers and others (1999), full liberalization of the world sugar exporters, Australia and Thailand would increase market would result in a 41 percent increase in production by 25 percent, and Brazil, Cuba, and world sugar prices. Sugar imports would increase other Latin American countries (excluding Brazil, by 44 percent in the United States. Exports would Mexico, and Cuba) would increase production by decline by 34 percent in the European Union. Low- about 15 percent. cost sugar-producing countries would increase Global welfare gains from full liberalization are exports, with Australia's exports rising 16 percent, estimated by Borrell and Pearce (1999) to total Brazil's 23 percent, and Thailand's 22 percent. $4.7 billion per year based on historical supply Removal of government support from domestic responses; gain could go as high as $6.3 billion per producers in the European Union, Japan, and the year if higher supply responses occur. Brazilian United States would save consumers $4.8 billion producers would gain the most from liberalization, per year. A study by the U.S. General Accounting at around $2.6 billion per year, but this would be Office (USGAO 2000) concluded that the U.S. sugar offset by a loss of $1 billion to Brazilian consumers program resulted in a net loss to the U.S. economy who would pay higher prices after liberalization-- of $1 billion in 1998. Elbehri and others (2000) leaving a net gain of $1.6 billion for Brazil. Japan used the global trade analysis project (GTAP) would enjoy a net gain of about $0.4 billion from multisectoral, multiregional general-equilibrium lower consumer prices that would more than offset model to examine the impacts of partially liberaliz- lower producer prices on the 40 percent of sugar ing sugar tariff-rate-quota regimes, concluding that that is domestically produced. The United States cutting the European Union's over-quota tariff by would have a small net gain of about $0.5 billion one-third would yield a global welfare gain of $568 from full liberalization, with consumer gains million. Coordinated global reforms would result slightly larger than producer losses. Western in the greatest benefits. Wohlgenant (1999) esti- Europe would gain about $1.5 billion as consumer mated that global sugar-trade liberalization would gains of about $4.8 billion exceeded producer losses result in a 43 percent increase in world price. of about $3.3 billion. Borrell and Pearce (1999) used a 24-region The exporting countries that now enjoy prefer- model of the global sweetener market to examine ential access to European and U.S. sugar markets consumption, production, trade, price, and welfare gain about $0.8 billion per year from prices that are effects for seven classes of sweeteners. A baseline more than twice world market prices on sales to the projection that continued current protection was European Union, and 80 percent more than the compared with a fully liberalized market with no world market price for sales to the United States. trade protection or domestic support in any coun- The value of the preferential access is less than it try or region. Under the fully liberalized scenario, appears, however, because many of these producers sugar prices were projected to fall from the baseline have high production costs and would not produce by 65 percent in Japan; 40 percent in Western as much at world market prices. Further, world 152 Global Agricultural Trade and Developing Countries FIGURE 8.3 Sugar Prices, 1970­2003 U.S. cents per pound 35 30 25 EU intervention 20 15 10 World 5 0 1970 1975 1980 1985 1990 1995 2000 market prices would rise by an estimated 38 per- (subsidies) to support producer prices at levels well cent after full liberalization, partially offsetting the above international prices. The program is financed loss of high prices in preferential markets for pro- primarily by the European Union's consumers, who ducers. Borrell and Pearce (1999) estimate the net pay high prices for sugar. The sugar policy began in loss to these exporting countries from full liberal- 1965 as part of the Common Agricultural Policy ization at $0.45 billion. The cost to taxpayers in the (CAP). Under the CAP, intervention sugar prices European Union and United States of providing have been constant in nominal terms since each $1 of preferential access is estimated to be 1984­85; however, they vary with exchange rates more than $5. In a recent study of the Fijian econ- when expressed in U.S. dollars (figure 8.3). They omy (Levantis, Jotzo, and Tulpule 2003), alternative have been more than double world market prices forms of aid were found to deliver much greater during most of the past 20 years. Import duties are economic benefits and growth prospects. used to prevent lower-priced imports from the Caribbean sugar producers are among the world market, and export refunds are paid to largest group of countries having preferential exporters to cover the gap between the EU price access to the European and U.S. sugar markets. A and the generally lower world market prices when recent study (Mitchell 2004) found that most of the commodities are sold from intervention stocks. Caribbean producers cannot export profitably even Production quotas limit the amount of sugar to the European Union, which pays prices that are eligible for price support. Quotas are divided into more than triple those of the world market. Many categories A and B, with different levels of price of these countries have abandoned their U.S. quo- support. Sugar production in excess of quota is tas because they do not produce enough to satisfy classed as C sugar and is not supported, but it can both their EU and U.S. quotas, and EU quotas have be carried over for use as quota sugar in the next higher prices. year or exported at world market prices. The total While the benefits to reform are not widely dis- of A and B quota sugar was 14.592 million tons in puted, the opposition to reform within certain 2000­01, of which 11.983 million tons was for A countries has been strong. The remainder of this quota and 2.611 million tons was for B quota section examines the sugar policies of the European (USDA 2003). The quotas have been declining to Union, Japan, and the United States with an eye to meet WTO commitments. The surplus of A and B the prospects for reform. quota sugar above domestic consumption is about 1.5 million tons; it is exported with subsidy. Excess quota (C sugar) averaged 1.59 million tons (white The European Union's Sugar Policy equivalent) between 1995­96 and 2000­01. Thus, The European Union's sugar policy uses produc- the EU sugar program results in about 3.1 million tion quotas, import controls, and export refunds tons of sugar exports per year (about 10 percent of Sugar Policies: An Opportunity for Change 153 world exports), and half of this is subsidized.2 Crit- this amount has remained constant, with realloca- ics of EU policy charge that A and B quota sugar is tion of quotas among existing members when a subsidizing the production and export of C sugar. country did not fulfill its quota. The sugar imported Australia, Brazil, and Thailand have filed a com- under the Lomé Convention is known as"preference plaint with the WTO to that effect. sugar." An additional import allocation of between Production levies are applied to all quota sugar 200,000 and 350,000 tons of sugar was made to ACP production to cover the costs of export refunds. countries (primarily) in 1995. This allocation of The levy on A quota sugar is 2 percent, whereas the "special preference sugar" is not permanent, and the levy on B quota varies from 30 percent to 37 per- quantity can vary based on import needs. The price cent depending on world market prices. An addi- specified for special preference sugar was 85 percent tional levy can be collected in the next marketing of the guaranteed price for the permanent prefer- year to recover any shortfall in export refunds. ence sugar. In addition, the European Union took Quotas are also set for some alternative sweeteners over the WTO import commitments of the new such as HFCS (known as isoglucose within the members joining the European Union in 1995. European Union) and inulin (produced from These included a tariff quota of 85,500 tons, mainly chicory and Jerusalem artichoke). The quota for from Brazil, with an in-quota tariff rate of 98 ECU production of HFCS is 303,000 tons; that for inulin (European currency unit) per ton. The European is 323,000 tons. Union has also granted several countries in the The Uruguay Round commitments had little ini- Balkans temporary access to its sugar market. tial impact on the European Union's sugar regime. Imports under this program totaled about 100,000 The variable import levy was replaced by a fixed tons in 2001­02. In total, the EU permanent import duty plus a safeguard clause allowing for a variable commitment is 1.39 million tons (white sugar additional duty with minimal impact on protection equivalent) plus additional quantities of up to to sugar beet producers. The European Union 450,000 tons of temporary imports. agreed to reduce both the amount spent on export The European Union's Everything But Arms refunds and the volume of sugar exported with sub- initiative (EBA), approved in 2001, allows duty- sidy. Export refunds are also payable on sugar free access to the EU sugar market by the 48 least- exported in the form of processed goods such as developed countries (39 are ACP countries). It sugar confectionery, chocolate, biscuits, cakes, ice could become the largest of the European Union's cream, soft drinks, and so on. The European Union commitments. Initially EBA imports will be limited amended legislation to allow changes in sugar- by quotas, and the sugar imported will be counted production quotas on an annual basis (rather than against the quota of special preference sugar. The the previous five-year basis) to ensure that the limits EBA quota will increase annually until full duty- on exports were met. The WTO commitment was to free access for white and raw sugar is allowed in reduce only the subsidized exports net of preferen- 2009. Safeguard clauses in the EBA initiative could tial imports. This is a small proportion of total be used to limit imports, but these would be diffi- exports, and amounted to just 34 million tons from cult for the European Union to invoke because the 1986­90 base of 1.612 million tons per year. doing so would be seen in the least-developed Preferential access to the European Union's sugar countries as a policy reversal. market and its high prices are granted to the Imported sugar will eventually displace domes- 46 countries from Africa, the Caribbean, and the tic EU production and could severely strain the EU Pacific (ACP) that signed the first Lomé Convention sugar regime. The European Commission esti- in 1975. The Lomé sugar protocol provided for mated the possible impact of the EBA on the EU imports of specified quantities of raw or white cane sugar regime in 2000, concluding that sugar sugar originating in the ACP states at guaranteed imports could increase by an additional 2.4 million prices. Unlike most articles of the Lomé Convention, tons and cost the EU budget about 1.05 billion the sugar protocol does not expire and cannot be euros. These imports would have to be offset by changed unilaterally. The original quantities speci- reduced domestic production quotas or used for fied were 1,294,700 tons of white-sugar equivalent, ethanol (European Commission 2003). with an additional amount allotted to India. The A longer-term threat to the EU sugar program total import commitment was for 1,304,700 tons; is the Commission's plan to offer, all 77 ACP 154 Global Agricultural Trade and Developing Countries countries the same conditions as the EBA countries as is done by current EU producers, and export it at under the Economic Partnership Agreements world market prices. A 1998 EU Commission study (EPAs). Negotiations, begun in September 2002, are of the 10 accession countries concluded that the expected to take five years. Under the EPAs, all ACP group would add at least 200,000 tons to the countries would have duty-free access to the EU European Union's export surplus. market for all goods except arms. These countries The current EU sugar regime runs until June currently produce 6.2 million tons of sugar. They 2006, and the European Commission opened could provide all of it to the European Union on discussions on reform on September 23, 2003. short notice while covering their own demand from However, unlike the other commodities sched- the world market. Taken together, EBA and ACP uled for reform discussions--cotton, olive oil, supplies could total 8.6 million tons. This is 60 per- and tobacco--specific reform proposals were not cent of current EU production and would force offered for sugar. Instead, three scenarios for major changes to the EU sugar program. reform were offered, ranging from an extension of Enlargement of the European Union may also the current sugar regime beyond 2006 to complete create new problems for its sugar regime. The liberalization of the current regime. Complicating 10 countries that joined in mid-2004 were Cyprus, the reform discussions is an investigation launched the Czech Republic, Estonia, Hungary, Latvia, by the WTO in August 2003 in response to the Lithuania, Malta, Poland, the Slovak Republic, and complaint by Australia, Brazil, and Thailand that Slovenia. Bulgaria and Romania will likely join in the EU sugar regime illegally subsidizes the indus- the next several years; the last of the current round try and depresses world prices. A negative finding of accession countries, Turkey, may join several against the EU by the WTO dispute-settlement years later. Poland is the largest sugar producer of body could force changes to the EU sugar regime. the 10 countries that joined in 2004, with nearly 60 percent of the group's total production. The first Japan's Sugar Policy 10 accession countries produce about one-fifth as much sugar as the European Union, have higher per Japan is the third-largest net sugar importer, after capita consumption, lower yields, and lower recov- Russia and Indonesia, with average annual net ery rates than the European Union. They agreed to imports of about 1.6 million tons of raw sugar dur- an A and B quota of 2.958 million tons, with ing 1999­2001 (figure 8.4). Imports supply about 2.829 million tons of A quota and 0.129 million two-thirds of domestic consumption; the remain- tons of B quota (European Commission 2003). ing one-third is supplied by highly subsidized beet Acceding producers will likely also produce C sugar, and cane production. Domestically produced FIGURE 8.4 Japanese Sugar Trends, 1970­2000 Million metric tons (raw equivalent) 4.0 3.5 Consumption 3.0 2.5 Imports 2.0 1.5 Production 1.0 0.5 0.0 1970 1975 1980 1985 1990 1995 2000 Source: FAOSTAT. Sugar Policies: An Opportunity for Change 155 HFCS accounts for about 40 percent of total caloric tic market price and the "target price." In marketing sweeteners. The government intervenes in the sugar year 2001, the target price for raw sugar was market by establishing guaranteed minimum prices 151,800 yen per ton ($1,168 per ton or $0.53 per for sugar beets and cane, controls on raw sugar pound), while the resale price on imported raw imports, prohibitive duties on refined sugar cane sugar was about $0.22 per pound (Fukuda, imports, high tariffs on imported products con- Dyck, and Stout 2002). The difference was made up taining sugar, and quotas, tariffs, and other controls by a subsidy financed by a surcharge on imported on sugar substitutes. The system results in retail sugar, other surcharges, and funds from Japan's sugar prices that are among the highest in the world national budget. The current subsidy to refiners is ($.89 per pound in Tokyo in 2000) and producer 90 billion yen ($692 million) (Fukuda, Dyck, and prices for sugar beets and sugar cane that are Stout 2002). The government regulates the produc- roughly 10 times world market levels. Sugar con- tion and price of HFCS to limit competition with sumption is gradually declining due to competition sugar and obtain funds to partially pay for the high from HFCS, high sugar prices, slow economic support to sugar beet and cane producers. growth, and dietary changes away from sweeteners. Full liberalization of Japan's sugar and sweetener Consumption may actually be higher than re- market would likely reduce domestic sugar produc- ported, however, because sugar contained in im- tion drastically--perhaps completely eliminat- ported products is not reported and is estimated to ing domestic production. Consumption would account for as much as an additional 10 percent of increase as consumers faced lower sugar prices. sugar consumption. Imports would increase to meet consumer Japan's Ministry of Agriculture, Forestry, and demand. HFCS consumption would likely increase Fisheries (MAFF) sets guaranteed minimum prices without current controls but would not necessarily for sugar cane and sugar beets according to the increase under full liberalization of the sugar and Sugar Price Stabilization Law of 1965 and the sweetener markets because of competition from Revised Sugar Price Adjustment Law of 2000 imported sugar. The Australian Bureau of Agricul- (Fukuda, Dyck, and Stout 2002). The minimum ture and Resource Economics (see Sheales and oth- producer prices are set based on a formula com- ers 1999) estimated that sugar imports would rise paring current agricultural input prices and consu- by 500,000 tons if Japan eliminated its tariffs, sur- mer goods relative to prices that prevailed in 1950 charges, and levies on sugar imports. The study and 1951. The minimum producer price for sugar assumed that domestic production would decline beets during 1990­95 averaged $149 per ton, while by just 22 percent because of other means of gov- the minimum producer price for sugar cane was ernment support--this is probably an underesti- $174 per ton. By comparison, U.S. sugar beet and mate. The Economic Research Service of the U.S. cane producers received an average of $29 and $40 Department of Agriculture (USDA) estimated that per ton, respectively, during the same period. Thus production would decline by 40 percent if Japan Japanese beet and cane producers received at least were to eliminate all border protection and trade- 10 times the world market prices.3 For the 2001 distorting domestic support. Consumer and pro- marketing year, the minimum price was 17,040 yen ducer prices in Japan would fall by 70 percent per ton ($131 per ton) for sugar beets and under the scenario, and imports would rise by as 20,370 yen per ton for sugar cane ($157 per ton). much as 735,000 tons (Fukuda, Dyck, and Stout Australian sugar cane producers, which receive no 2002). Borrell and Pearce (1999) estimated that government price supports, received $16 per ton in sugar prices would decline by 65 percent, produc- the 2001 marketing year (Sheales 2002). The MAFF tion would decline by 44 percent, and net imports also sets the raw sugar price for domestic refiners, would increase by about 1.5 million tons. known as the "domestic sugar rationalization target price," at a level intended to allow restructured The United States' Sugar Policy sugar refining firms to pay the guaranteed mini- mum price to sugar cane and beet producers and U.S. sugar policy provides for a loan program for still recover costs. A subsidy is provided to sugar sugar beets and cane.4 The nonrecourse loan pro- refiners to cover the difference between the domes- gram is reauthorized through fiscal 2007 at 18 cents 156 Global Agricultural Trade and Developing Countries per pound for raw cane sugar and 22.9 cents per through 2000 to 16.21 cents per pound. The over- pound for refined beet sugar. A Refined Sugar Reex- quota tariff will remain prohibitive at a world price port Program allows sugar cane refiners to pur- of about 5 cents per pound (assuming a U.S. raw chase raw sugar at world prices, without duty, and sugar market price of 22 cents per pound and a export a like amount within 90 days. A similar pro- transportation cost of 1.5 cents per pound). gram exists for manufacturers of sugar-containing Under NAFTA most trade barriers between products. A no-cost provision of the policy requires Canada, Mexico, and the United States were to be the secretary of agriculture to make every effort to eliminated by 2009. As described previously, the operate the sugar program in a way that avoids for- treaty's sugar provisions were altered by a side- feiture under the loan program. To avoid forfei- letter agreement prior to the start of NAFTA. But tures, it is necessary to keep the domestic sugar the side-letter agreement did not change other price above the world market price. This is done by NAFTA provisions--such as the phased reduction restricting sugar imports, first by quotas introduced in the United States over-quota tariff of 16 cents in May 1982, and then by tariff rate quotas begin- per pound by a total of 15 percent during the first ning in 1990 following a successful GATT (General six years, and then in a straight line to zero in calen- Agreement on Tariffs and Trade) challenge. dar year 2008. The over-quota tariff on raw sugar, Minimum import levels were approved in 1990 7.6 cents per pound in 2003, drops about 1.5 cents to allay concerns of quota-holding countries and per pound each year. If the world raw sugar prices cane processors. It provided for marketing allot- are in the range of 7 cents per pound, and U.S. raw ments on domestically produced sugar if estimated sugar prices are about 18 cents per pound, Mexican imports were less than 1.25 million tons, raw value. producers would benefit from exporting to the The secretary of agriculture has the authority to United States instead of to the world market impose marketing allotments in order to balance (USDA 2002). Currently, Mexico does not have a markets, avoid forfeitures, and comply with the large surplus of sugar to export, but increased pro- U.S. sugar-import commitments under WTO and duction or reduced consumption could change NAFTA. The allotments can be used only when that. In future years, the over-quota tariff will con- sugar imports, excluding imports under the reex- tinue to decrease and could lead to large imports. A port program, are less than 1.532 million tons.5 The provision of the U.S. sugar legislation removes pro- USDA announced flexible marketing allotments for duction quotas if imports exceed 1.5 million sugar for the 2002­03 marketing year (Haley and tons--a free-for-all if imports increase beyond Suarez 2002). certain limits. Under this alternative, the U.S. gov- In the Uruguay Round Agreement on Agricul- ernment could end up holding large stocks ture (URAA), the United States agreed to maintain defaulted under the sugar loan program, and the minimum imports of 1.139 million metric tons of sugar system would become more difficult to man- raw-value sugar imports (1.256 short tons). Of this, age because of the no-net-cost provision. Mexico 22,000 metric tons were reserved for refined sugar. has increased sugar production from about 3.5 mil- The tariff rate quota on raw cane sugar was allo- lion tons during 1989­91 to 5.2 million during cated to 40 quota-holding countries based on their 2000­02, while consumption has increased from export shares during 1975­81, when trade was rela- 4.0 to 4.5 million tons. Following the end of the tively unrestricted. The duty of 0.625 cents per NAFTA phase-in period, Mexico can ship unlim- pound, raw value, continues on quota imports. ited quantities of sugar to the United States duty- Most countries continue to avoid the duty because free without the condition of being a net surplus of programs under the Generalized System of Pref- producer. This will likely force changes to the U.S. erences and the Caribbean Basin Initiative. The sugar program. For example, Mexico could increase duty on raw sugar above the tariff rate quota was imports of HFCS for use in the soft drink industry, 17.62 cents per pound beginning in January 1995 freeing sugar for export to the United States. and declined by 0.45 cents per pound each year The effectiveness of the U.S. sugar program at until it reached 15.36 cents per pound in 2000. The keeping domestic prices above world prices since over-quota rate for refined sugar was 18.62 cents 1980 can be seen in figure 8.5. During this period, per pound in 1995 and declined by 0.48 per year world prices have fallen sharply, but U.S. producers Sugar Policies: An Opportunity for Change 157 FIGURE 8.5 U.S. Sugar Loan Rates, U.S. Prices, and World Prices, 1980­2002 U.S. cents per pound, raw cane sugar 30 U.S. raw sugar price 20 U.S. loan rate 10 World market price 0 1980 1985 1990 1995 2000 Source: USDA. were protected. The sugar program, however, faces that program domestic prices were supported at new challenges in the near future that could bring about double world prices, with quotas to limit into conflict the no-cost provision of the sugar pro- production and tariff rate quotas to limit imports gram, the minimum import commitment under (see Diop, Beghin, and Sewadeh in this volume). the WTO, and the duty-free access provision to And, like sugar, the edible peanut program faced Mexican imports in 2009. The rapid growth of the threat of increased imports due under WTO sweetener production compared with consumption agreements and NAFTA. In the 2002 U.S. Farm Bill, could also destabilize the program. The growth rate the loan rate for edible peanuts was cut by half, of sweetener production during 1985­2000 was compared with the mid-1990s, production quotas 3.2 percent, compared with consumption growth of were eliminated, and direct cash payments were 2.1 percent over the same period. If these growth made to producers. The payments consisted of rates are extended into the future, marketing allot- deficiency payments if prices fell below the new ments would be needed to prevent stock building. lower loan rates, decoupled direct payments, and The problem is further exacerbated by the agree- countercyclical payments. In addition, quota hold- ment under the URAA to import 1.139 million tons ers were compensated with direct payments for of sugar per year. their loss of quota rights. A similar program for The U.S. sugar program, like that of the sugar would be complicated by the loss of benefits European Union, almost certainly will have to by HFCS producers, who benefit from high sugar change. But although it benefits just 9,000 sugar prices. Reform of the sugar program may also beet producers and 1,000 sugar cane producers require compensating the industries that now (Orden 2003), opposition to policy reform is depend on distorted sugar policies. strong, especially from sugar-cane producers, who average nearly 3,000 acres per producer (compared Conclusions with 200 acres per beet producer). Florida accounts for one-quarter of U.S. sugar production, and two Sugar cane is an almost ideal commodity for some large corporations account for nearly 80 percent of developing countries to grow for domestic con- the cane acreage in Florida. Such concentration of sumption and export. It can be produced efficiently production suggests that reforming the U.S. sugar in tropical climates under a wide range of technolo- program will likely require compensation to exist- gies, from low-input labor-intensive to high-input ing producers. fully mechanized. Sugar is locally consumed in all A model to consider is the recent reform of the producing countries and provides a substantial U.S. edible peanut program (Orden 2003). Under part of total calories in many countries. Processing 158 Global Agricultural Trade and Developing Countries can be varied to meet the needs of low-income liberalization is estimated to total $0.45 billion domestic or high-income foreign consumers. Raw per year. cane sugar stores well after initial processing. There The nature of reforms can have very different are few problems in meeting sanitary and health consequences for developing countries. If existing standards because sugar cane juice is boiled during EU and U.S. polices are adjusted to accommodate initial processing and raw cane sugar is boiled again higher imports from countries in the EBA and when refined to produce white sugar. The biggest NAFTA systems, low-cost producers such as Brazil problems for producers are limited export oppor- will lose. Full multilateral liberalization of the tunities and low world prices--caused partly by world sugar market would allow efficient producers policies in OECD countries. to expand production and exports, thereby benefit- Support for current OECD sugar policies ing consumers in protected markets. Coordinated among beneficiaries is obviously strong, but prob- multilateral liberalization also offers the advantage lems are emerging that make change inevitable. The of somewhat higher world prices to soften the benefits of sugar policy reform are substantial, and adjustment for producers in protected markets the gains are greatest under multilateral reform. such as the European Union, Japan, and the United According to recent studies of the global sugar and States. sweetener markets, the global welfare gains of removing all trade distortions and domestic sup- Notes port are estimated to total as much as $4.7 billion per year. In countries with the highest protection 1. Brazil's policies on ethanol indirectly affect sugar, but the government provides no direct subsidies to sugar producers. (Europe, Indonesia, Japan, and the United States), Other small sugar producers that produce at world market net imports would increase by 15 million tons per prices include Canada and Malaysia. year. World sugar prices would increase about 2. An additional 1.8 million tons of sugar is imported under 40 percent, while sugar prices in countries that the sugar protocol between the EU and the member countries of ACP (Africa, the Caribbean, and the Pacific) and reexported heavily protect their markets would decline. The with subsidy after processing. greatest price decline would occur in Japan, where 3. There appears to be an anomaly between the OECD's esti- sugar prices would fall 65 percent, followed by a mate of producer support in table 8.4 and the prices received by sugar beet and sugar cane producers in Japan. If sugar beet and 40 percent decline in Western Europe and a 25 per- cane producers in Japan receive five times the prices in the cent decline in the United States. Brazilian produc- United States, then it appears the Producer Support Estimate ers would gain the most from liberalization, around (PSE) in percentage form should be higher rather than lower as reported in table 8.4. $2.6 billion per year, offset by a loss of $1 billion to 4. Nonrecourse commodity loans are used by the govern- Brazilian consumers who would pay higher prices ment to support prices of many crops. Under the program, under liberalization. Employment in developing farmers who comply with the provisions of each commodity countries would increase by approximately 1 mil- program are allowed to pledge their commodity as collateral and obtain a loan from the USDA's Commodity Credit Corporation lion workers if the 15 million ton increase in net at the specified loan rate per unit for the commodity. The bor- imports that accompanied the removal of all trade rower may elect to repay the loan with interest within a specified distortions and domestic support were supplied by period and regain control of the commodity, or default on the loan as payment of the loan and interest. The farmer will nor- developing countries. mally default on the loan if the market price is below the level The exporting countries that currently have necessary to repay the loan and interest. Thus, the loan rate preferential access to European and U.S. sugar mar- becomes the effective floor price. 5. This seems the opposite of what is required, but the logic kets gain about $0.8 billion per year through prices is apparently that if imports exceed this amount then the sugar that are more than double world market price. The program has lost its ability to control imports and U.S. produc- value of the preference is less than it appears, how- ers should be given unrestricted freedom to produce. ever, because many of these protected producers have high production costs and would not produce at world market prices. Further, world market References prices would rise by about 40 percent after full mul- Berg, C. 2001. World Ethanol Production 2001. Ratzeburg, tilateral liberalization, partially offsetting the loss to Germany: F.O. Licht. Borrell, B., and D. Pearce. 1999. "Sugar: The Taste Test of producers of high prices in preferential markets. Trade Liberalization." Center for International Economics, The net loss to these exporting countries from full Canberra and Sydney, Australia. September. Sugar Policies: An Opportunity for Change 159 Devadoss, S., and J. Kropf. 1996."Impact of Trade Liberalization Agricultural Competitiveness and World Trade Liberaliza- under the Uruguay Round on the World Sugar Market." tion: Implications of the WTO and FTAA, Fargo, North Agricultural Economics 15: 83­96. Dakota, May 29. Diop, N., J.C. Beghin, and M. Sewadeh. 2004."Groundnuts: Poli- Oregon Office of Energy. 2002. "Biomass Energy: Cost of Pro- cies, Global Trade Dynamics, and the Impact of Trade Liber- duction." Salem, Oregon. alization." In M.A. Aksoy and J.C. Beghin, eds., Global Agri- Organisation for Economic Co-operation and Development cultural Trade and the Developing Countries. World Bank, (OECD). 2002a. Directorate for Food, Agriculture and Chapter 12. Fisheries "Working Party on Agricultural Policies and European Commission. 2003."Reforming the European Union's Markets, Background Information on Selected Policy Issues sugar policy. Summary of impact assessment work." Com- in the Sugar Sector." Paris, June. mission Staff Working Paper. Brussels. ------. 2002b. Agricultural Policies in OECD Countries. Moni- Elbehri, A., T. Hertel, M. Ingco, and K. Pearson. 2000. "Partial toring and Evaluation. Paris, June. Liberalization of the World Sugar Market: A General Equi- Sheales, T.C. 2002. "Australia's Sugar Industry: Operating in a librium Analysis of Tariff-Rate Quota Regimes." U.S. Free-Market Environment,"in A. Schmitz, T.H. Spreen, W. A. Department of Agriculture, Economic Research Service, Messina and C.B. Moss, eds. Sugar and Related Sweetener Washington, D.C. August 25. Markets: International Perspectives. CABI Publishing, El-Obeid, A., and J.C. Beghin. 2004."Multilateral Trade and Agri- Wallingford, Oxfordshire. cultural Policy Reforms in Sugar Markets," CARD Working Sheales, T.C., S. Gordon, A. Hafi, and C. Toyne. 1999. "Sugar Paper 04-WP 356, Iowa State University, Ames, Iowa. International Policies Affecting Market Expansion," ABARE Escandon, J. 2002. "Mexico: Sugar and Ethanol Don't Mix Well." Research Report 99.14. Australian Bureau of Agricultural F. O. Licht's International Sugar and Sweetener Report. and Resource Economics, Canberra. November. Ratzeburg, Germany, June 27. USDA. 2002. "Production, Supply, and Distribution" Foreign F. O. Licht. 2002. "India to Liberalize Its Sugar Market," F. O. Agricultural Service (FAS). Washington D.C. Licht's International Sugar and Sweetener Report, March 4. ------. 2002. "Briefing Room Data for Sugar." Economic Fry, J. 1985. "Aspects of a Complex Commodity." World Bank Research Service. Washington, D.C. Working Paper. Commodities Division, Washington, D.C. ------. 2003. "European Union Sugar Annual", Global Agricul- Fukuda, H., J. Dyck, and J. Stout. 2002. "Sweetener Policies ture Information Network Report, Foreign Agricultural Ser- in Japan." Electronic Outlook Report SSS-234-01. U.S. vice (FAS). Washington D.C. Department of Agriculture, Economic Research Service, U.S. General Accounting Office (USGAO). 1993. "Sugar Pro- Washington, D.C. September. gram: Changing Domestic and International Conditions Haley, S., and N.R. Suarez. 2002. Sugar and Sweeteners Outlook. Require Program Changes." Report GAO/RCED-93-84, U.S. Department of Agriculture, Economic Research Service, April, Washington, D.C. Washington, D.C. September 26. ------. 2000. "Sugar Program: Supporting Sugar Prices Has International Sugar Organization. 2002. "Quarterly Market Increased Users' Cost While Benefiting Producers." Report Outlook," September, London. GAO/RCED-00-126, June, Washington, D.C. Jolly, L. 2001."Will China Ever Fill its WTO Sugar Tariff Quota?" U.S. International Trade Commission (USITC). 2002. "The International Sugar Association, London. Processed. Economic Effects of Significant U.S. Import Restraints, Levantis, T., F. Jotzo, and V. Tulpule. 2003. "Ending of EU Sugar Third Update 2002." Publication 3519. Washington, D.C. Trade Preferences: Potential Consequences for Fiji." Current June. Issues 3 (2). Australian Bureau of Agricultural and Resource van der Mensbrugghe, D., J.C. Beghin, and D. Mitchell. 2003. Economics, Canberra. "Modeling Tariff Rate Quotas in a Global Context: The Case LMC International. 1999. "Pricing Policy Reform--A Sum- of Sugar Markets in OECD Countries." CARD Working mary." Sweetener Analysis. Paper 03-WP 343. Iowa State University, Ames, Iowa. Mitchell, D. 2004."Sugar in the Caribbean: Adjusting to Eroding Wohlgenant, M.K. 1999. "Effects of Trade Liberalization on the Preferences." Policy Research Working Paper (forthcoming). World Sugar Market." United Nations Food and Agriculture World Bank, Washington, D.C. Organization, Rome. Orden, D. 2003. "Sugar Policies and Implications: Integration into an FTAA." Paper presented at the conference on 9 DAIRY: ASSESSING world MARKETS AND POLICY REFORMs: IMPLICATIONS FOR DEVELOPING COUNTRIES Tom Cox and Yong Zhu World dairy markets exhibit an extreme case of dis- tion in industrialized countries are also important tortions traceable to a complex system of domestic sources of economies in procurement, processing and international trade barriers--including sur- and logistics, and foreign direct investment. plus disposal in the Quad countries (Canada, Policy reforms leading to free markets would European Union, Japan, and the United States) and pitch consumers against producers in most coun- the Republic of Korea. Oceania (Australia and New tries because of the large transfers implied by cur- Zealand)--which, with the Quad, dominates the rent policies and their removal. In importing coun- export market--is a competitive exporter with few tries with high barriers (Asia), consumers' gains are distortions. However, dairy interest groups in the larger than producers' losses because the dairy sec- Quad are entrenched, and prospects for policy tor is small. In competitively producing countries, reforms appear dim. Domestic price discrimina- consumers currently benefit from depressed world tion schemes in the Quad (minus Japan) rely heav- dairy prices and low trade barriers and have much ily on the ability to close borders, suggesting that to lose in undistorted markets, whereas producers the emphasis in the Doha negotiations should be have much to gain. The largest net welfare gains on commitments to lower border protection to would accrue in the Quad, however, because large force domestic reforms. consumer gains and reduced budgetary costs for Despite the quagmire of distortions, dairy is a support policies would be much larger than pro- dynamic sector with much growth potential, espe- ducers' losses. In most other countries, net effi- cially in Asia, where dairy consumption has been ciency gains would remain small because the gains propelled upward by income growth, urbanization, to one group would be offset by losses to the other. and westernization of diets. Dairy is also experienc- Our simulations also show the production gains ing innovations in food processing, with value- from trade liberalization are captured by dynamic added opportunities in traditional products and reformers attracting foreign direct investment and new dairy-based protein ingredients facing few overcoming supply constraints and technology trade barriers. Concentration and vertical integra- transfers.1 161 162 Global Agricultural Trade and Developing Countries Background on the World ment to local economies. Derivative dairy products, Dairy Sector or manufactured products, can be either in liquid form (standardized milk, pasteurized milk, cream, Milk Production and Dairy Product Manufacturing partly or totally skimmed milk, buttermilk), or products no longer liquid (cheese, butter, cream, Milk and dairy products are expensive to produce. condensed and evaporated milk, milk powder, Production of animal feed uses three to nine times casein). more land than production of food plants that Development of the dairy industry requires good produce the same amount of protein (Bender infrastructure. A good transportation system, avail- 1992). When food sufficiency is a problem in a ability of low-cost refrigeration technology, and country, dairy is probably not an appropriate way good packing technologies are all prerequisites for to produce food, because the nutritional conver- an advanced dairy manufacturing sector. Most sion rate from grains to animals is low. In most developing countries have poor conditions for dairy developed countries, milk is produced by feeding manufacturing. In those that lack an adequate milk animals concentrates made from grains--these can supply and infrastructure to process and distribute be used directly as human foods. Livestock feed milk, people reconstitute milk powder and butter oil requires on average 7 kilocalories input for each back to fluid form to meet daily consumption kilocalorie generated. The range extends from 16 needs. In many countries, commercial milk combi- for beef production to 3 for broiler chickens, with nation, which reestablishes the product's specified milk somewhere in between (Bender 1992). fat-to-nonfat solids ratio and solids-to-water ratio, Animal food production does not always com- is widely used. pete with other food production, however. Some Overall, approximately one-third of world milk animals, including sheep and cows, can be fed on is consumed in fluid form. About one-fourth is inedible agricultural and industrial by-products used in cheese making. The joint production of (with limited alternative uses) to produce highly butter, milk powder, and casein uses roughly one- nutritional foods, or they can be grazed on mar- fifth of all milk. The remainder is processed into ginal land. Marginal land suitable for grazing is soft or frozen products, condensed and evaporated often on small parcels in remote areas with low milk, or other dairy products. population density. Shipping perishable dairy Derivative products satisfy specific consumption products to urban consumers is expensive, as many needs. Simple technologies for separating and developing countries face production and distribu- recombining nutritional components of milk have tion challenges that constrain milk production, lowered the cost of processing and made it possible such as poor infrastructure and limited refrigera- to adjust fat content to different dietary needs. tion facilities. Cheese and butter do not require advanced technol- In Argentina, Australia, Ireland, and New ogy. The production of milk protein concentrates Zealand, milk production occurs generally on large and whey and lactose milk fractionations, however, pastures within reach of relatively efficient trans- is relatively new technology. Milk protein concen- portation systems and with the support of better trates and whey and lactose products are important human capital and technology. These factors pro- in the world dairy markets, where most buyers are vide considerable advantage to these countries in developing countries, countries with low self- producing milk and dairy products in free-trade sufficiency in dairy production, and developed environments. However, pasture-based milk farm- economies with relatively low trade barriers on ing is seasonal and vulnerable to weather and natu- these products. ral disasters. Skim milk powder, whole milk powder, butter Although raw animal milk is nutritious, only 5 oil, and even butter are important inputs in a to 10 percent of milk is consumed in raw form in special dairy processing practice called milk recon- most developed countries. Most raw milk is stitution, a technology that converts milk powder, processed into derivative products in these coun- milk fat products, and other dairy products back to tries (FAO various years). These processing sectors fluid milk for consumption or for making other can be significant sources of income and employ- dairy products. With the rapid development of Dairy: Assessing World Markets and Policy Reforms: Implications for Developing Countries 163 dairy processing technology, milk reconstitution regions (China, Korea, other Southeast Asia, becomes commercially practical and desirable, even Middle East and North Africa, and Central and in more advanced dairy-producing countries where South America) increased their combined share fresh milk is readily available. from just over 10 percent in 1990 to almost 20 per- Fluid milk is a major product of milk reconstitu- cent in 2000. These trends indicate significant tion in developing countries. In developed countries changes in the production of milk in the develop- and countries with dairy foods in the traditional ing regions. For example, growth in South Asia and diet, "hard" dairy products, such as cheese, are pro- the Middle East and North Africa was 3.8 percent duced through milk reconstitution. The recent annually, in sharp contrast to the negative growth growth of trade in milk powder and butter oil is of ­3.5 percent recorded in Eastern Europe and partially attributable to the improvement in milk the FSU. reconstitution technology in many dairy-importing About 25 percent of world milk was produced in countries. Milk reconstitution overcomes high Western Europe before the 1994 Uruguay Round transportation and storage costs. Trade distortions Agreement on Agriculture (URAA) (FAO various and investment decisions in the dairy manufactur- years). After 1994 Western Europe's share of world ing sector also make reconstitution desirable. Many milk production began to dip slowly, falling to developing countries reconstitute fluid milk based 23 percent in 2000. Nevertheless, Western Europe on cheap milk powder available in world markets. remains a key factor in world dairy markets. The For example, Mexico makes cheese by reconstitut- share of world milk production of the United States ing imported skim milk powder and adding veg- and Canada remained relatively constant at around etable oil (filled milk). 14 percent before (1989­1994) and after (1995­ Besides technical difficulties and additional 2000) the URAA. The share of Oceania in world costs involved in milk reconstitution, other prob- milk production increased slightly, reaching 4 per- lems limit the practice of this technology. The cent in 2000. Japan's share of world milk produc- reconstituted dairy products are often considered tion remained relatively constant at around 1.5 per- to be inferior substitutes for fresh-milk-based cent over the same periods. products. In addition, milk reconstitution from In China, Korea, and the rest of Southeast Asia, dried dairy ingredients often induces a loss of a steady increase in share occurred both before nutrients due to the heat required to condense and and after the URAA, reaching 2.8 percent in 2000. dry milk. Much of the growth occurred in China and Korea. India, as well, showed a steady growth in share of world milk production before and after the URAA World Milk Production Trends periods. That share stood at 13.8 percent in 2000. In the last decade, world milk production was India appears to be expanding its activity in world between 445 million and 470 million metric tons. export markets, particularly for butter fat products The Quad and Oceania's share of this production (butter, ghee/anhydrous milk fat). The share of was around 42 percent, while the share from East- other countries in South Asia reached 5.1 percent ern Europe and the Former Soviet Union (FSU) in 2000. fell steadily from 27 percent to 16.8 percent. There Central and South America showed strong is some reason to expect that the decline in the growth in share of world milk production both share of Eastern Europe and the FSU in world before and after the URAA. Its share is now 10.5 per- milk production may be reversed as several key cent. Expansion in this region is dominated by milk-producing countries (Poland, Hungary, and Argentina and, to a lesser extent, Uruguay, Brazil, Baltic countries join the European Union (EU) in and Mexico. Sub-Saharan Africa, including South 2004. Africa, has held relatively steady at around 3 per- South Asia (India and other South Asian Coun- cent of world milk production, suggesting that milk tries) and Sub-Saharan Africa increased their share production and dairy processing are not expanding of world milk production slightly from just under as rapidly in southern Africa as in other developing 20 percent in 1990 to more than 20 percent by economies. Disparity in regional income growth 2000. Similarly, other developing countries and likely plays a key role in this trend. 164 Global Agricultural Trade and Developing Countries Trends in the World Dairy Trade rather than URAA-induced trade liberalization, are likely responsible for most of these changes. Dairy products exhibit two-way trade because they Other developing economies (China, Korea, are differentiated products.2 Over the 1989­2000 other Southeast Asia, Middle East and North period, world dairy exports increased more or less Africa, and Central and South America almost steadily from 25.6 million to 39.1 million metric doubled their share of world dairy exports from tons, with no difference between the pre- and post- 5.3 percent in 1989 to 9.6 percent by 2000; much URAA periods. World dairy exports are dominated of the increase came from the Southern Cone by the developed economies (Quad and Oceania), (Argentina, Chile, and Uruguay), where exports but their share of the markets shrank from 87 per- represented 1.9 percent of total milk production in cent in 1989 to 78 percent in 2000. Developed 1989 and 3.6 percent by 2000. countries' share of world dairy exports grew an In South Asia (India and other South Asian average 2.9 percent annually in 1989­2000. Exports countries) and Sub-Saharan Africa (including represented 9.7 percent of the developed countries' South Africa), which are primarily importers, total milk production in 1989; that share had exports represented less than 1 percent of total grown to 12.4 percent by 2000. World dairy milk production in 1989­2000. The share of world imports increased more or less steadily, from dairy imports for these two regions was 9­10 per- 26 million to 36 million metric tons (FAO, various cent (3 million metric tons) over the same period. years). World dairy imports in the pre-URAA The dominant countries in these regions (India period (1989­94) averaged 27 million metric tons, and South Africa) are either self-sufficient or net increasing to 33 million metric tons in 1995­2000 exporters. Overall, the participation of South period. Asian countries in world dairy trade has been lim- Comparison of dairy exports as a share of total ited, because most countries in the region strive for milk production before and after the URAA pro- self-sufficiency in food; both imports and exports vides some informal evidence that the dairy trade of dairy products are restricted by the government. liberalization during the Uruguay Round increased Imports consist of intermediate products such as the importance of exports in the world's dairy butter oil, milk powder, and condensed milk, economies. Developed countries' (Quad and Ocea- mostly obtained from food aid programs from nia) import market shares--16 percent (4 million Western countries. metric tons) in 1989 and 20 percent (7 million met- The Middle East and North Africa are signifi- ric tons) in 2000--grew an average 4.8 percent cant importers of dairy products, accounting for annually, but average annual growth was much 21 percent of total world imports of dairy products. faster after the URAA (7.3 percent). Japan is a large Lacking natural resources to expand their milk pro- importer of high value-added dairy products, such duction, the countries of this region will continue as cheese and casein. Its demand for dairy products to rely on imports to meet their increasing con- is driven by high incomes, more Westernized diets, sumption needs. and the inability of domestic supply to satisfy the Import substitution policies and economic demand growth despite protectionist policies. hardship in Central and South America have pre- Despite their decreased share of world milk vented these countries from fully exploiting their production, Eastern Europe and the FSU increased significant comparative advantage in agriculture, their share in world dairy exports from 7.6 per- though some South American countries increas- cent to 11.6 percent during the period under study. ingly participate in world and regional dairy trade. Those exports represented 1.3 percent of total milk Under more stable macroeconomic environments production in the region in 1989 and increased and regional trade agreements, Latin American more than three-fold, to 4.8 percent, by 2000. countries, have increased trade volume consider- Eastern Europe and the FSU are relatively small ably. Latin America imports about 18 percent of importers, with imports fluctuating in the 3 million­ world dairy trade while exporting 2 percent. Mexico 4 million metric ton range. Import growth slowed in is the world's largest importer of milk powder, the years after the URAA (­5.9 percent). Low accounting for some 10 percent of total world trade. growth rates in gross domestic product (GDP), Brazil is a significant importer of dairy products, Dairy: Assessing World Markets and Policy Reforms: Implications for Developing Countries 165 most of which are imported from its Mercosur part- equilibrium price, more milk is generated, less pre- ners, Argentina and Uruguay. The European Union mium milk is consumed (due to the higher admin- is also a big player in the Brazilian market, but istered prices), and the prices for manufactured Brazil's dependence on imports will likely change milk are depressed relative to a competitive, non- with the rapid adoption of new technology in distorted equilibrium. In a sense, these classified domestic milk production and dairy processing. pricing schemes generate consumption cross- Other African countries are net importers subsidies to manufactured products (and the (mostly through food aid programs), accounting consumers and processors who purchase these for about 6 percent of total world trade. Reduced products) at the expense of the consumers of government intervention in agriculture, propelled premium products. To the extent that these cross- by the URAA and other factors, will result in subsidized manufactured products are exported, reducing food aid to traditional recipients, includ- there is an open question as to whether the implicit, ing the Sub-Saharan countries. A significant pro- consumption cross-subsidies are in fact export portion of food aid has been redirected to transi- subsidies. tional economies. The affordability of commercial In the case of Canada, several of the dairy classifi- imports of dairy products (without current export cations targeted to compete on export markets (the subsidies) is questionable for many African world or the U.S. market) have been deemed export countries. subsidies generated by government intervention-- hence countable against URAA export-subsidy Trade and Domestic Policy commitments. Regimes in Key Producing and Canada, the European Union, and the United Consuming Countries States also use a variety of other subsidies in pro- duction, marketing, and export financing. Among Developed Economies: Quad and Oceania these are the European Union's consumption subsi- Most developed countries intervene in their dies on butter (60 percent of EU butter is subsi- domestic dairy sectors with a wide variety of policy dized for use by the bakery sector) and skim milk instruments. Intervention prices (price support powder (45 percent is subsidized for animal feed). programs) establish minimum domestic prices that Market access under the URAA is controlled pri- are generally well above world market levels. This is marily by tariff rate quotas, a system of in-quota true for butter and skim milk powder (SMP) in tariffs up to a negotiated limit, and a series of out- Canada and the European Union; for butter, of-quota tariffs that are generally quite prohibitive. cheese, and SMP in the United States, although In addition, a variety of sanitary, phytosanitary, and SMP is priced near world-market levels. technical trade restrictions (such as country-level Canada and the European Union use a system of standards of identity) act as nontariff barriers and milk production and marketing quotas to limit the impede global trade in dairy products. production of milk and reduce the cost of protect- Developed countries agreed in the URAA to ing the domestic milk- and dairy-processing sec- increase their import quotas to 5 percent of con- tors. These policies have generated substantive sumption by 2000. Similarly, in- and out-of-quota "quota rents" to the holders of the quotas. tariffs were to be reduced 16 percent over six years, Canada and the United States use classified pric- by 2000. Table 9.1 summarizes the URAA increases ing schemes to enhance market returns for dairy in dairy import quotas for the developed countries. farmers based on how their milk is used. Generally Table 9.2 summarizes the URAA dairy product these are price-discrimination schemes that admin- tariff reductions. ister higher prices to less elastic, higher-value- Under the URAA, the European Union and added, and more perishable product markets (bev- United States were obliged to make the greatest erage milks, soft and frozen products). To the increases in access to their domestic markets, par- extent that these premium markets are nontrad- ticularly for cheese, butter, butter oil, and skim milk able, such schemes can help insulate domestic mar- powder. Australia, Canada, and Japan endured less kets from world market forces. In addition, because market access change under the URAA since many they increase milk prices above the competitive of their dairy product imports were already in 166 Global Agricultural Trade and Developing Countries TABLE 9.1 Dairy Import Quotas for the Developed Countries under the URAA (1000 tons) Butter, Skim Milk Whole Milk Country/Region Policy Regime Cheese Butter Oil Powder Powder Western Europe BASE 37.0 79.5 41.2 0.7 GATT 2000 123.1 91.3 69.2 1.1 GATT 2005 194.7 101.0 92.6 1.5 Eastern Europe BASE 6.9 12.6 10.5 3.0 GATT 2000 8.8 20.9 19.0 5.0 GATT 2005 10.4 27.8 26.1 6.7 Japan BASE na 3.5 99.8 0.0 GATT 2000 na 3.5 99.8 0.0 GATT 2005 na 3.5 99.8 0.0 Australia BASE 11.5 na na na GATT 2000 11.5 na na na GATT 2005 11.5 na na na Canada BASE 20.4 2.0 0.9 0.0 GATT 2000 20.4 3.3 0.9 0.0 GATT 2005 20.4 4.4 0.9 0.0 United States BASE 116.4 7.5 1.3 0.5 GATT 2000 136.4 13.1 5.3 3.4 GATT 2005 153.1 17.7 8.6 5.9 na -- Not available Data source: International Dairy Arrangement, Fifteenth Annual Report. November 1994. BASE and GATT 2000 follow the URAA of the GATT (General Agreement on Tariffs and Trade), assume linear changes. GATT 2005 projects the Uruguay Round Agreement linearly to 2005. excess of the requirement of 5 percent of domestic protection has been radically reduced. Currently, consumption (by 2000). The level of most out-of- Australian dairy exports are without explicit export quota tariffs is prohibitive, with the result that tariff subsidies. rate quotas act as pure import quotas for many products in developed countries. Eastern Europe and Baltics The heavy intervention in domestic dairy mar- kets by the developed countries often generates sur- Among World Trade Organization (WTO) members plus production relative to domestic consumption in Eastern Europe, only the Czech Republic and requirements. Because the domestic policies usu- Estonia use Green Box policies to support their ally keep domestic prices above world market domestic dairy industry. These policies are limited. levels, many developed countries--particularly Under its Domestic Food Aid Plan, the Czech Canada, the European Union, and the United Republic donates milk to schools. The program's States--are forced to export these surpluses with monetary value reached $0.74 million in 2000.At the considerable subsidy. Those export subsidies same time, the Czech government paid $4.7 million depress world market prices, making cost-effective in support to dairy cow herds through structural dairy production more difficult in many develop- adjustment assistance provided through an invest- ing countries. While frustrating potential milk pro- ment aids plan. The monetary value of Estonia's ducers, consumers in these developing countries school milk program value was small ($0.6 million gain substantively from the transfer of wealth, in 2001) (Megli, Peng, and Soufi 2002). which comes in the form of cheaper dairy imports.3 The dairy industry is important in Eastern Since Australia substantially deregulated its Europe. Many countries in the region use Amber domestic market in 2000­01, domestic market Box policies to support its development. Some even Dairy: Assessing World Markets and Policy Reforms: Implications for Developing Countries 167 TABLE 9.2 Tariff Reductions for Dairy Products under the URAA, by Region (US$/ton for specific duties, percent for ad valorem tariffs) Butter, Skim Milk Whole Milk Cheese Butter Oil Powder Powder Country/ Policy Policy Region Regime Instrument In-Q Over-Q In-Q Over-Q In-Q Over-Q In-Q Over-Q Western BASE Specific duties 547 3,643 1,189 3,971 639 1,956 1,773 4,102 Europe GATT 2000 768 2,362 1,225 2,572 632 1,561 1,760 3,486 GATT 2005 952 1,295 1,255 1,406 626 1,232 1,750 2,972 Eastern BASE Ad valorem 53% 181% 39% 187% 67% 200% 40% 160% Europe GATT 2000 59% 133% 39% 142% 66% 164% 40% 102% GATT 2005 65% 92% 39% 105% 65% 134% 40% 54% Japan BASE plus Ad valorem 50% 50% 35% 35% 1.3% 15% 30% 30% Specific duties 0 0 0 13,406 0 4,954 0 10,228 GATT 2000 Ad valorem 32% 32% 35% 30% 13% 13% 30% 26% plus Specific duties 0 0 0 11,397 0 4,210 0 8,691 GATT 2005 Ad valorem 17% 17% 35% 25% 13% 11% 30% 22% plus Specific duties 0% 0% 0% 9,723 0 3,500 0 7,411 Australia BASE Specific duties 71 1,068 74 1% 37 1% 37 1% GATT 2000 or ad valorem 71 905 0 1% 0 1% 0 1% GATT 2005 71 769 0 1% 0 1% 0 1% Canada BASE Specific duties 56 3,794 193 3,483 48 1,720 48 3,315 GATT 2000 24 3,231 83 2,915 21 1,462 21 2,820 GATT 2005 0 2,761 0 2,442 0 1,247 0 2,408 United BASE plus Ad valorem 10.5% 0.0% 5.0% 5.0% 0.0% 0.0% 0.0% 0.0% States Specific duties 0 1,924 62 2,004 33 1,018 68 1,320 GATT 2000 Ad valorem 10.5% 0.0% 4.3% 4.3% 0.0% 0.0% 0.0% 0.0% plus Specific duties 0 1,636 62 1,703 33 865 68 1,122 GATT 2005 Ad valorem 10.5% 0.0% 3.6% 3.6% 0.0% 0.0% 0.0% 0.0% plus Specific duties 0 1,395 62 1,453 33 738 68 957 Mexico BASE Ad valorem 50% 95% 35% 35% 0% 139% 0% 139% GATT 2000 49% 89% 31% 31% 0% 131% 0% 131% GATT 2005 47% 84% 27% 27% 0% 124% 0% 124% South BASE Ad valorem 56% 66% 59% 60% 65% 68% 49% 82% America, GATT 2000 49% 58% 56% 57% 59% 31% 46% 75% North GATT 2005 44% 51% 53% 54% 54% 56% 43% 69% South BASE Ad valorem 37% 37% 37% 37% 36% 36% 36% 36% America, GATT 2000 36% 36% 35% 35% 35% 35% 34% 34% South GATT 2005 36% 36% 34% 34% 33% 33% 32% 32% Data source: International Dairy Arrangement, Fifteenth Annual Report, November 1994 BASE and GATT 2000 follow the URAA of the GATT (General Agreement on Tariff and Trade) and assume linear annual changes. GATT 2005 projects the Uruguay Round Agreement linearly to 2005. "In q" and "over q" are in quota and above quota tariff rates respectively. increased their support to the dairy industry dur- milk (increasing to $6.6 million in 1997 and 1998) ing the WTO implementation period. Hungary and $9 million on ice cream in 1997 (decreasing to increased its support for cow milk from $17.7 mil- $0.8 million in 1998). Compared with negligible lion in 1996 to $58.5 million in 1998 (nonspecific support before 1998, Poland increased its market component of the aggregate measure of support). support for butter to $6.5 million and $1.9 million In 1995 and 1996 Slovenia spent $5.5 million on in 1999 and 2000, respectively. 168 Global Agricultural Trade and Developing Countries Some countries apply tariff rate quotas (TRQ) The implementation of a set of common exter- on dairy imports to protect their domestic markets. nal tariffs and a substantive (if not total) elimina- The Czech Republic applies TRQs to milk, cream, tion of internal tariffs and nontariff barriers in yogurt, butter, and ice cream. While the fill rate for Mercosur countries began in 1995. Certain prod- ice cream and yogurt is relatively high, it is low for ucts or regions are exempted from these regulations butter and even less for milk and cream (less than until 2006. Mercosur has 11 different tariff levels 1 percent). Similarly, the fill rates for butter in bounded by 20 percent; exceptions can be greater Slovenia and the Slovak Republic are quite low as than 20 percent but no more than 35 percent. As a they are for milk and cream in Hungary and result of this common market, trade among the Poland. The resulting reduction in market access member countries increased from $4.7 billion in contributes to distortions in world dairy markets, 1991 to $18 billion in 1998. but the levels are small, reflecting the modest levels Other trade agreements exist between countries of both trade and support. in South America. The Andean Group consists of The countries of Eastern Europe use different Bolivia, Chile, Colombia, Ecuador, and Peru, while methods to administer TRQs. The Czech Republic the Group of Three is made up of Colombia, uses a first-come-first-served method to allocate Mexico, and Venezuela. Bilateral agreements are them. Hungary uses licenses on demand. For milk also common (Bolivia-Mexico; Brazil-Argentina). and cream Poland applies licenses on demand; for Chile's tariff rates are around 8 percent for most other dairy products it uses the applied-tariff dairy products. Peru has rates of 20 percent on method. Slovenia and the Slovak Republic apply a most dairy products, with a surcharge of 5 percent. license-on-demand method to allocate TRQs. No Both countries are members of the Asia Pacific matter which method is used, additional import Economic Cooperation (APEC). Argentina's agri- costs are incurred, and imports are restricted. cultural sector enjoys export rebates that range Milk powder, butter, cheese, casein, yogurt, from 1.4 to 10 percent. Brazil has a system of export creams, and some other dairy products receive credits and cash advances for exported products. export subsidy from the Czech Republic, Poland, and the Slovak Republic. The actual level of export East Asia and South Asia subsidies is much lower than the commitment lev- Income, price, tastes, age, and geography are main els, however, and these levels have been decreasing factors affecting dairy consumption in Asia.4 China, over time. India, Indonesia, Japan, the Republic of Korea, Malaysia, Philippines, Thailand, and Singapore-- all WTO members--are the main dairy producers Latin America and consumers in the region. Because dairy prod- A major question regarding the future of dairy pol- ucts are not necessary products in Asia, income is icy and trade agreements in South America is the major factor affecting their consumption.When whether the Mercosur countries will join with the income goes up, dairy consumption will increase, as NAFTA countries in creating a Free Trade Area of can be seen from the experience of China, Japan, the Americas, or FTAA. Initiated in 1994 with nego- and the Republic of Korea. Although the three tiations to be finished by 2005, the FTAA project is countries show similar patterns of food consump- considered a single undertaking: nothing is agreed tion behavior, Japan's per capita GDP is much until all is agreed. If implemented, major changes higher than those of Korea and China, and its per can be expected for trade policies in South America, capita consumption of dairy products is 66 and especially for the Mercosur countries (Megli 2002). 23 kilograms more, respectively. Current trade policies in South America revolve In developed countries and cities of developing mainly around the Mercosur policies for the four countries, dairy products represent a small portion main economies of Argentina, Brazil, Paraguay, and of total expenditure. Thus, consumers' reaction to Uruguay. Negotiations are under way to bring Chile price change is not highly sensitive. In rural areas of and Bolivia into full member status, and future developing countries, however, where dairy prod- negotiations are expected to take place for Colombia, ucts are a luxury, consumption is highly sensitive to Peru, and Venezuela. price changes. Dairy: Assessing World Markets and Policy Reforms: Implications for Developing Countries 169 Most of the dairy products consumed in Asia are the highest-price bidders at quota auctions held by fluid milk, yogurt, and milk powder. In 1999, for the livestock products marketing organization. In example, Asia's consumption of fluid milk and Japan, the tariff rate quota fill rates for skimmed milk powder accounted for 97 percent of dairy milk powder, whey, and butter are around 50 per- consumption. cent. South Korea and Malaysia have higher fill People in dairy-producing areas tend to con- rates, but real imports are still lower than tariff rate sume more dairy products than in other regions, quotas. Indonesia has out-of-quota imports, and due to easy and convenient access. This may explain the tariff rate quota fill rate is 100 percent. why per capita dairy consumption in South Asia is World dairy trade liberalization would increase higher than in East Asia, even though its per capita the region's imports. GDP is much lower. From countries' notifications to the WTO Agri- The Middle East and North Africa culture Committee, we find that only Japan uses Green Box policies to support domestic dairy mar- Dairy tariff levels in the Middle East and North ket in school-lunch programs. Another potential Africa vary greatly among countries and products. user of this policy may be China, which has begun a They appear to be relatively high for nonconcen- program to provide subsidized milk for school chil- trated milk, cream, and yogurt, for example, and dren in some cities and expects to expand it to the lower for milk powder and butter.5 In addition to rest of the country. tariffs, countries implement regulations aimed at With the exception of Japan and Korea, most protecting dairy consumers from fraud. Such regu- countries in East Asia show a negative or de mini- lations impose technical requirements related to mus aggregate level of support for agriculture product composition and associated customs pro- (Amber Box policy). Japan uses price support pro- cedures such as sanitary certifications. There are grams for certain dairy products (mainly butter various bilateral trade arrangements between the and skimmed milk powder), and also makes defi- European Union and the countries of the region. ciency payments for calves and milk manufactur- Import tariffs directly affect the supply of dairy ing. In 2000 Japan's milk-producer support esti- products in the Middle East and North Africa. For mate (PSE) reached $4.7 billion; the nominal instance, relatively high tariffs on milk powder tend protection coefficient for milk in 1999 was 364 per- to increase raw milk supply, while low tariffs on raw cent (OECD 2001). Korea's milk producers receive milk handicap local production. Also, low import more than three times the world price for milk. The tariffs on raw materials and equipment stimulate PSE for Korean milk reached $0.8 billion in 2000; the production of processed dairy products and the the nominal protection coefficient in 1999 was derived demand for raw milk. Subsidies are not the 225 percent. No Asian country uses blue-box poli- main incentive tool used to encourage raw milk cies to support their dairy sectors. supply in the Middle East and North Africa. Where East Asia protects its dairy markets, and entry of subsidies exist, they support production or trans- many imports is governed by tariff rate quotas and portation, but their level is not always effective. high tariffs. The scale and scope of tariffs differ widely, however. China, India, Japan, and Korea The Impact of World Dairy impose relatively higher tariffs on dairy products Policy Reforms than do Indonesia, the Philippines, Malaysia, and Singapore. Ten dairy exports to Japan, five to South Considerable scope remains for further removal of Korea, two to Malaysia, and just one to Indonesia trade and domestic support policy distortions in are subject to tariff rate quotas. Japan's and Korea's the Doha Round. Even after full implementation of tariff rate quotas are allocated on a global basis. For the URAA provisions by developed countries, whey and skim milk powder, Japan's tariff rate quo- almost 60 percent of world dairy trade will still be tas are allocated to producers and producer organi- exported with subsidies (U.S. Dairy Export Council zations or sellers of mixed feed. For skim milk pow- 1999). Market access provisions allow for tariff rate der, whole milk powder, and other milk and cream, quotas with prohibitively high rates of out-of- Korea's tariff rate quotas are allocated according to quota duty (Griffin 1999). Also, special safeguards, 170 Global Agricultural Trade and Developing Countries low minimum-access requirements, and small tariff tion combines two other scenarios: free dairy trade, reduction requirements for individual commodi- and no domestic support. The free-dairy-trade sce- ties undermine the market access provisions of the nario considers the elimination of all trade distor- URAA (Coleman 1998). Thus, even after full tions for 2001 through 2005. All export subsidies implementation, world dairy markets will continue and import tariff rate quotas (quotas, in- and out- to be characterized by highly subsidized exports, of-quota tariffs) are eliminated. Domestic support limited market access, and heavy government policies are maintained as in the base scenario. This intervention. should increase world trade, increase world market As part of the URAA, countries agreed to begin prices, and put considerable strain on several new agricultural negotiations by the beginning of domestic support policies (intervention price pro- 2000, and dairy groups in several countries have grams, in particular) in the protected dairy sectors. detailed their policy objectives and positions for the The no-domestic-support scenario eliminates all Doha Round. U.S. dairy industry representatives domestic supports from 2001 to 2005. These meas- outlined their negotiating priorities early on (U.S. ures include intervention prices for the European Dairy Export Council 1999). Those priorities Union (SMP), Canada (butter and SMP), the include gradual elimination of export subsidies, United States (butter, SMP, cheese), and other reduction and harmonization of high tariffs, and countries; the elimination of classified pricing in tightening disciplines on domestic supports. By the United States and Canada (modeled as a price eliminating export subsidies and reducing import premium for residual--fluid, soft, and frozen-- barriers, it is assumed that world prices will rise products over manufactured products); and elimi- sufficiently for the United States to be competitive nation of production and marketing quotas in the in world markets (Kirkpatrick 1998). Countries of European Union and Canada. the Cairns Group (with the exception of Canada), Because the United States incurred large costs in which represents small- and medium-sized agricul- the base year (2000) through its intervention/price tural exporters, are pushing for measures that go support program (about $500 million in SMP pur- even further toward freer markets and liberalized chases), domestic deregulation would have strong trade (Cairns Group Farm Ministers 1998). While impacts on U.S. milk prices. Similarly, given the the negotiating goals of the European Union are large levels of rents from milk-production quotas not yet articulated, their priorities will likely in the European Union and Canada (35 percent involve minimizing increases in import access and and 40 percent of the domestic milk prices, respec- reductions in export subsidies, as well as maintain- tively), elimination of these policies would sharply ing the Blue Box and the Peace Clause (Oxford increase these countries' competitiveness (no quota Analytica 1998). constraints and sharply reduced production costs). The implications of alternative proposals on Hence milk production would increase sharply developed versus developing countries are not well even as milk prices and revenues drop. researched. This chapter addresses these questions Domestic deregulation would lower prices in the by simulating various dairy policy liberalization protected dairy economies and thus lower world scenarios using the University of Wisconsin- dairy prices, but it would not necessarily widen Madison World Dairy Model (UW-WDM).6 The access to competitive exports--unless out-of-quota results of the simulations provide insights into the tariffs became less prohibitive at the lower market tradeoffs between the heavily protected developed prices. Moreover, the increased milk production for economies and the developing economies, provid- the European Union and Canada would need to ing quantitative measures of the impact of those find a market, potentially beyond domestic con- tradeoffs on economic welfare and world trade. sumption, and so would likely displace base-level imports by these dairy sectors and reduce other World Dairy Deregulation Scenarios countries' potential for export-market growth. A first scenario, discussed here in detail, contem- We focus our presentation of the simulation plates full dairy sector liberalization: all trade and results on the main scenario (full liberalization) domestic support policies are removed between and refer readers to the annex tables for the sepa- 2001 and 2005. Full world dairy sector liberaliza- rate second and third scenarios. Dairy: Assessing World Markets and Policy Reforms: Implications for Developing Countries 171 Full world dairy sector liberalization. Devel- nario, Canada's milk prices would drop by 44 per- oped economies with dairy sectors characterized by cent and production by 4 percent, significantly strong protection from domestic and trade policies more than under the no-domestic-support sce- would experience large changes from full liberaliza- nario. Dairy exports would fall by 6 percent, while tion, with large transfers from producers to con- imports would increase 215 percent (versus 80 per- sumers. In the absence of rents from milk produc- cent export expansion and 5 percent contraction of tion quotas, EU milk prices would fall 23 percent by imports, under the no-domestic-support scenario). 2005, generating a moderately competitive EU milk Producer surpluses would be cut in half (to sector and expanded production (approximately $1.4 billion) by 2005, but consumer welfare gains 8 percent at prices roughly 20 percent less than base would be even greater at $1.6 billion (up 14 per- levels by 2005). The expansion implies a potentially cent). Total government revenues would fall slightly radical restructuring of the EU milk sector toward (lost tariff revenues being not quite offset by gains more efficient farms. Dairy exports would increase from elimination of export subsidies, the interven- 16 percent, while imports would fall 50 percent by tion price program, and production and marketing 2005, suggesting that lower domestic prices (inter- subsidies). Consumer welfare gains would offset vention price floors having been eliminated) and producer and treasury losses, yielding a net welfare larger domestic milk availability at sharply lower gain of 2.7 percent ($385 million). prices (due to quota elimination) would cut In the United States, milk production (­7 per- imports. Currently competitive exporters, there- cent), prices (­12 percent), and producer surplus fore, would suffer. The current producer surplus (­17 percent, ­$2.7 billion) would fall sharply by would take a massive hit of $8.1 billion by 2005, 2005 under full liberalization, about three times and the social and political costs of the implied rad- more than under the no-domestic-support sce- ical restructuring of the milk production sector nario. These relative impacts indicate that U.S pro- would be nontrivial. Consumers would be the big ducers enjoy substantive protection from current gainers from deregulation (due to falling prices), trade-policy distortions. U.S. exports would fall with large welfare gains of $8.1 billion. Total gov- 61 percent (down 331,000 metric tons), while ernment costs would fall slightly (lost tariff rev- imports would more than double (130 percent, enues offset by reduced costs of domestic support 510,000 metric tons) by 2005. U.S. consumers and export subsidies). Consumer and treasury would gain $3.4 billion (4 percent); government gains would offset producer losses, yielding net costs would be reduced by $147 million (lost tariff welfare gains of $1.1 billion. revenues net of gains from eliminating intervention The scenario would work similarly in Japan. By price and export subsidy costs). These gains would 2005 Japanese milk production would fall by exceed producer losses by $2.7 billion to generate 23 percent, milk prices by 54 percent, and the pro- net welfare gains of $729 million (0.7 percent) by ducer surplus by 61 percent (or $3.2 billion). The 2005. concurrent removal of domestic regulations would As expected, Oceania's dairy producers and have little effect because trade barriers sustain most processors would gain under full liberalization, of the domestic programs. Imports would increase despite giving up large quota rents (especially New by 134 percent, bringing consumers a surplus of Zealand) associated with current preferential $4 billion. Net government revenues would fall by (quota) access to the protected developed-economy $21 million (lost tariff revenues net of smaller markets. As low-cost exporters, Australia and New domestic policy savings). Consumer gains would Zealand would be able to fully exploit their com- offset producer and treasury losses to generate net parative advantage in undistorted world dairy mar- welfare gains of $1.1 billion. kets, increasing milk production by 6 percent, pro- While the dairy sectors in Canada and the ducer prices by 22 percent, and the producer United States use both trade policies and domestic surplus by 42 percent, or $1.1 billion, by 2005. support programs, they derive more protection Their exports would rise by 21 percent, or 429,000 from the former (subsidized exports and limited metric tons, by 2005. The production and trade market access due to import quotas and higher out- gains would be less than under the free-trade-alone of-quota tariffs). Under the full liberalization sce- scenario, because that scenario would not increase e Change 0.1- 0.5- 0.1- 0.2- 0.1 0.2- 0.1- 0.0 0.2 0.7- 0.0 0.5- 1.3- 0.9- 7.8- 2.2 0.2- 0.3- 1.3 0.2- 0.4 elfar % W otalT M 9 1 (95) (13) (18) (33) (22) (16) 129 (81) (51) 80 (55) $US (173) (284) (197) (387) (861) 4,176 (1,037) 3,139 Gov M) (2) (2) (0) 1 (4) (36) (74) (47) (34) (114) (443) (133) (248) (163) (138) (107) (363) (611) otalT Rev/Costs ($US (1,680) 1,184 (1,795) Change 1.3- 2.9- 2.8- 0.4- 1.9- 7.0- 1.9- 4.0- 3.3 0.5 0.6 2.2 0.4- 5.0 5.3- 1.5 0.2 6.8 14.4 0.5- 2.5 Surplus % Baseline M Consumer (66) (57) (993) (516) (236) (692) 171 160 300 (48) 237 (219) 416 372 521 $US (2,560) (2,574) 1,707 17,464 (2,039) 15,425 2005 from 6.0 1.7 9.3 11.4 17.4 13.1 37.0 23.1 Change 18.0- 0.2- 0.5- 12.4- 1.5- 11.9- 43.3- 15.0 1.2- 0.7 4.1 25.3- 9.3- Surplus % Changes oducer Pr M 55 60 972 498 247 669 (12) (32) (41) 80 (64) (334) (125) (302) 298 -WDM). $US ,5012 2,562 Support: (1,445) 2,797 (14,472) (11,675) (UW Model Domestic Price 1.2 4.4 8.6 -- centage 14.3 10.4 24.2 1.0- 1.0- 0.1- -- 2.7 y change) Exporters 30.4- 8.3- 8.9- 10.6 13.0- 18.6 0.3- 20.7- 7.8- Dair and Milk (per orld W radeT of Competitive 1.4 4.3 6.5 0.4 3.3 3.8 2.6 4.0 Milk centage 5.7- 0.1- 0.1- 0.3- 2.9- 0.5- 3.3- 6.7- 3.5 0.0 1.1 1.1 1.1 change) Production Importers (per Potentially Net isconsin-Madison W Liberalization South North Rep.s of Europe Republic Union Asia Full Economies, Economies, of, Asia orld y/Region People' Eastern America, Mongolia Soviet East African America, South W America, Africa Rep. available University 9.3 of Korea ce: Countr Developing India Other South China, Poland South tal:oT Caribbean Dem. of Developing Former South Other Middle Rest Mexico North Central Korea, Southeast tal:oT orld Not Developed Developing W -- Sour ABLET 172 Dairy: Assessing World Markets and Policy Reforms: Implications for Developing Countries 173 production in Canada and the European Union. generate lower prices and large consumer gains Consumer losses would pale in comparison to the ($1.7 billion, or 3.3 percent). Several regions would substantive producer gains, generating net total show substantive increases in production, price, welfare gains of 8.8 percent, or $1 billion, by 2005. and producer surplus, notably the FSU (with a pro- Developing-country exporters would enjoy the ducer surplus gain of $2.6 billion). Treasuries in all same benefits from full dairy sector liberalization countries would suffer from lost tariff revenues. as Oceania, but at slightly lower levels of gain and Aggregate treasury losses would amount to $1.7 bil- with larger transfers from consumers to producers lion exceeding modest aggregate producer gains of (table 9.3). Wider access to developed-economy $298 million and consumer gains of $521 million. markets and elimination of export subsidies would Net welfare losses would be $861 million by 2005. generate aggregate increases in milk production Under full liberalization aggregate world milk (2.6 percent), prices (1­24 percent), and producer production would rise by 1.1 percent by 2005. surpluses ($2.5 billion, or 9.3 percent), suggesting Average milk prices would decrease by 7.8 percent substantial import-substitution and export opportu- overall, falling 20.7 percent in the developed coun- nities available in some of these countries. However, tries, while rising 2.7 percent in the developing the aggregate consumer surplus in the developing countries, reflecting the modest loss to consumers in countries would fall by $2.6 billion (1.9 percent) the latter countries on average. World dairy trade because of the loss of subsidized imports and higher would expand by more than 2 million metric tons by domestic prices (except in Eastern European coun- 2005 as the impacts of domestic deregulation tries). Together with the loss of tariff revenues (chiefly quota removal) reinforced the impacts from ($114 million), aggregate consumer and taxpayer the elimination of trade barriers. World producer losses would slightly dominate producer gains, surpluses would fall sharply in the developed coun- generating modest welfare losses ($173 million, or tries (­$14.5 billion,­25 percent) while increasing in 0.1 percent) by 2005. The political economy of dairy the developing countries ($2.8 million, 4.1 percent). reform is complex even in developing countries, Developed-country losses would be due primarily to because consumer and producer interests are dia- the loss of quota value in the European Union and metrically opposed. The poverty implications are Canada, and to the removal of substantive domestic also stark, pitching poor consumers (who benefit supports (in Japan and the United States). from the current regime) against the rural dairy Savings from elimination of domestic and sector, which would gain under free markets. export subsidies would exceed lost tariff revenues Consumers in net-importing regions would in the developed countries, generating a net treas- gain or lose depending on the tradeoffs between ury savings of $1.2 billion by 2005. In developing increased world import prices (a negative impact) countries, where domestic supports are generally and increased dairy trade (a positive impact). The much smaller, their elimination would not offset loss of previously subsidized imports can be offset the loss of tariff revenues, generating net increases by gains from broadly expanding trade depending in treasury costs of $1.8 billion, which could be an on the size, composition, and direction of import issue in some developing countries with few alter- price increases. Many governments currently tax native fiscal sources. Aggregate world treasury rev- their consumers--removing those taxes could off- enues would fall nearly $611 million by 2005. Con- set price increases. Although there may be some sumer welfare would increase by $17.5 billion in opportunity to expand domestic production to the developed countries, while falling $2 billion in substitute for previously subsidized imports, the the developing regions. In the developed countries, cost-competitiveness of scale-efficient exporters gains by consumers and taxpayers would exceed makes this less viable for many of these countries producer losses, generating $4.2 billion in net wel- that would experience negative impacts on milk fare gains by 2005. Just the opposite would occur in production, prices, and producer surpluses under the developing regions, where producer gains full liberalization. These producer surplus losses would fail to offset consumer and treasury losses on could be offset by consumer gains, notably in South average, yielding net welfare losses of $1 billion. America (dominated by Brazil), where dismantling Because the markets of the developed countries Mercosur common external import tariffs would are so much larger than those of the developing 174 Global Agricultural Trade and Developing Countries world, aggregate consumer gains would be larger by strong consumer welfare gains. Total govern- (at $15.4 billion) than aggregate producer losses of ment costs would fall, yielding substantive net wel- $11.7 billion and treasury losses of $611 million, fare gains of $4 billion, much larger than under full yielding net welfare gains for the world of $3.1 bil- liberalization results ($1.1 billion). Similar forces lion by 2005. These aggregate patterns hide the would apply in Canada. Dairy sectors that were variability in individual country impacts and the more protected by trade barriers than domestic large transfers at work between consumers and subsidies (such as Japan) would not experience producers within many countries. such losses. The United States would bear the full brunt of domestic policy deregulation, as U.S. milk Other scenarios. The free-trade scenario models prices, production, and producer surpluses all the elimination in 2001 of export subsidies and all would fall sharply. Reduced government costs and tariff rate quota barriers, while keeping substantive massive consumer gains would offset producer domestic supports in place. In the absence of losses, however, leading to a net welfare gain. In changes in domestic support programs, free trade Oceania, New Zealand would gain and Australia would decrease welfare overall, pointing to the fis- would lose under the no-domestic-support sce- cally unsustainable nature of domestic programs nario. Milk production, prices, and producer sur- under conditions of free trade. Free trade would plus would rise across most of the developing have the effect of stimulating domestic supports, world, but consumers in unprotected markets possibly leading to violations of WTO ceilings on would face higher prices. Impacts on net importers aggregate measure of support. in the developing world would be quite similar to Under the scenario of free trade alone, trade- those for potential developing-country exporters, protected producers in developed economies with the notable exception of the FSU, where con- would suffer substantive losses as their domestic sumer gains would barely exceed producer losses to consumers enjoyed world prices. In the protected generate a breakeven net welfare impact. developed countries, milk prices, exports, and pro- ducer surpluses would fall, while imports would Conclusions climb. Elimination of export subsidies would not offset the loss of tariff revenues in Canada or the The world dairy sector is complex and character- European Union, yielding net treasury and welfare ized by multifaceted domestic and trade policy dis- losses in these countries. Exporting developing tortions. The results of our simulation model countries and Oceania's dairy producers and (detailed by commodity, policy, and region) pro- processors would realize strong gains due to free vide a quantitative measure of the economic and access to higher-priced, protected markets. Con- welfare impacts of those distortions across regions, sumers in these countries would lose, but not as producers, consumers, and governments. While the much as producers gain, resulting in net welfare usual limitations of sectoral simulation studies gains. Consumers in net-importing dairy regions should be kept in mind, the simulations confirm would gain or lose depending on the tradeoffs what most standard economic policy analyses between increased import prices (a negative suggest--that the numerous and sizeable distor- impact) and increased trade (a positive impact) tions induced by most developed economies to from elimination of tariffs on imports into these protect their domestic dairy sectors have large and regions. generally negative spillover effects on competitive In the last scenario, that in which domestic sup- exporters and developing countries. Liberalization ports are eliminated but trade barriers remain, would lessen those spillovers, creating opportuni- dairy producers in developed countries with strong ties for growth in the domestic and potentially domestic support and production controls would export-oriented portions of the dairy sectors in suffer as production quotas and associated rents developing countries, but several caveats must be were eliminated. Lower prices would lead ineffi- noted. cient producers to exit the market and nearly elim- Liberalization would also cut into the large inate imports to these markets. In the European benefits that now accrue to consumers who enjoy Union, producers would take a massive hit, offset access to subsidized dairy products on world Dairy: Assessing World Markets and Policy Reforms: Implications for Developing Countries 175 markets thanks to the current protection regimes European Union is absorbed in its expansion to the of the developed countries. Millions of these con- East and the new 2003 CAP reforms, which leave sumers live in poorer countries with low border dairy relatively unchanged. protection and limited capacity to develop a dairy U.S. dairy policy as articulated in the 2002 Farm sector. The interests of these developing countries Bill increases domestic subsidies through the Milk diverge starkly from those of others having an Income Loss Contract program. Meanwhile, low- actual or potential dairy industry. cost dairy exporters (Australia, Argentina, and Eastern Europe) will likely continue pushing hard for additional market access through lower tariffs, World Dairy Sector Growth: A Component lower export subsidies, and increased import Perspective quotas. World product markets are increasingly driven by The fundamental question is, "Who has the bar- milk components (milk fat and fat fractionations; gaining power in dairy issues?" The WTO meeting casein, whey, and other protein fractionations; and in Cancún in 2003 changed this calculus by creating lactose). Current world growth trends are domi- strong opportunities for expansion of regional nated by "industrial" demand for dairy-based trade agreements (as opposed to a difficult global ingredients--intermediate products, not consumer agreement) that will limit access by nonmembers. products. This growth in demand is driven by Expansion of the European Union will provide advances in food processing, both on the input side protected access to new members, benefiting the (fractionations of milk components) and the pro- dairy sectors in several Eastern European countries. duction side (processes that optimize cost and However, managing the EU's structural milk sur- functionality using the evolving dairy-based ingre- plus will remain challenging in the face of existing dients), and by consumer demand for the final WTO commitments, the integration of Eastern processed products. Shaping a competitive dairy Europe, and the relatively strong entrenchment of sector in a world context will require producers to protectionist farm lobbies. The interests and influ- have component-based marketing plans, to organ- ence of EU dairy processors and consumers, both ize incentive structures rewarding such plans, and of whom would benefit in a liberalized market, to meet quality standards regimes. Making use of compete directly with the established interests of new dairy-based ingredients demands a moderately the milk producers. sophisticated food-processing sector and technol- ogy. Size and scale economies are important in Prospects for Developing Economies many of these processes, suggesting differential advantages to larger firms and to firms with foreign The potential for domestic market growth is driven direct investment backed by knowledge, expertise, by population and GDP. Population growth stimu- and ready capital. lates consumption of traditional dairy products; whereas increased incomes favor growth in new value-added products. Slow GDP growth will stall Prospects for World Dairy Policy Liberalization consumption of both types of products. Trends in dairy product development and markets What firms will supply the demand of growing occur in the context of current and evolving populations of more affluent consumers in the WTO agricultural trade negotiations. Short-term developing world? Will they be local or multina- prospects for further dairy trade liberalization in tional firms? Will they use local milk supplies, developed markets may be somewhat limited, how- imported dairy ingredients, or some combination ever. The heavily protected dairy sectors of Canada, of the two? Industry structure and infrastructure the European Union, Japan, and the United States are crucial to answering these questions. Scale effi- are not likely to open their markets before reducing cient (low-cost) and innovative processing firms subsidy levels. While the United States and Canada are likely to have competitive advantages in meet- would likely support liberalization in grains, ing these potential growth markets. Local versus oilseeds, and livestock products, dairy remains multinational ownership will be influenced by an especially sensitive industry. Meanwhile, the access to and the cost of capital and by the firms' 176 Global Agricultural Trade and Developing Countries marketing and procurement strategies. Foreign 6. The model is described on the CD-ROM that accompa- direct investment is often used to overcome market nies this volume. See also Zhu, Cox, and Chavaz (1999) and Cox and others (1999). access limitations allowed by the current WTO agreement and regulations by countries that permit only domestically owned firms to import dairy References products. Bender, Arnold E. 1992. Meat and Meat Products in Human Export potential into the developed economies Nutrition in Developing Countries. Rome: Food and Agricul- will be closely linked to further dairy trade liberal- ture Organization. Cairns Group Farm Ministers. 1998. "Communiqué following ization characterized by increased market access talks on the upcoming WTO talks on agriculture." Sydney, and lower domestic subsidies. In this context, opti- Australia, April 2­3. mal world supply and demand will remain a crucial Coleman, J. R. 1998."World Dairy Trade Overview: Trade Policy and the WTO." Paper presented at the Workshop for Dairy determinant of export prices and hence will define Economists and Policy Analysts, Cornell Program on Dairy the competitive context of world trade. If recent Markets and Policy, Baltimore, October 26­27. trends continue, export markets should remain rel- Cox, T. L., J. R. Coleman, J. P. Chavas, and Y. Zhu. 1999."An Eco- atively competitive, with lower production costs nomic Analysis of the Effects on the World Dairy Sector of Extending Uruguay Round Agreement to 2005." Canadian and prices, but also with some structural weakness Journal of Agricultural Economics 47: 169­183. in demand due to macroeconomic factors. Discern- FAO (Food and Agriculture Organization of the United ing the differential potential for market growth in Nations). Various years. FAO Yearbook: Production. Rome. ------. "FAOSTAT PC: User Manual." Rome. value-added products (which are sensitive to con- Griffin, M. 1999. "The Effects of Agricultural Trade Liberaliza- sumer income) versus bulk commodities (which tion on Global Dairy Markets." Paper presented at the 4th are more responsive to price) will require careful Interleite Symposium on Intensive Dairy Production, Caxambu, Brazil, July 22­24. consideration. Kirkpatrick, E. 1998."Editorial comments."Cheese Reporter, July 10. Overall, countries that are actual or potential Megli, K., T. Peng, and W. Soufi. 2002. "Eastern Europe and dairy producers and exporters stand to gain from Former Soviet Union Dairy: A Look at Policy, Country an unfettered market, but as liberalization occurs, Groupings, and Income Elasticity." Agriculture and Applied Economics Working Paper, University of Wisconsin­ special consideration should be given to poor con- Madison. sumers who are likely to suffer from higher con- Megli, K. 2002. "South America in the UW World Dairy Model: sumer prices. Poor consumers in such countries Policy, Grouping, and Elasticity." Agriculture and Applied Economics Working Paper, University of Wisconsin­- will be hurt, at least in the short run, by a move to Madison. global free trade in dairy products unless special Organisation for Economic Co-operation and Development measures are taken. (OECD). 2001. Agricultural Policies in OECD Countries. Monitoring and Evaluation 2001. June. Oxford Analytica. 1998. "European Union: Farm Trade." Daily Brief, June 2. Notes Peng, T. 2002. "A Study of Income Effects on Dairy Consump- tion and Milk Supply Elasticity in Asia." Agriculture 1. The CD-ROM included with this volume contains and Applied Economics Working Paper, University of an annex for this chapter presenting detailed market data Wisconsin­Madison. and policy information by country, a description of the model Soufi, W. 2002."North Africa and Middle East Dairy Supply and used here, and additional tables of results of policy-reform Consumption: Economic Policy, Disaggregation of the Area simulations. and Econometric Modeling." Agriculture and Applied Eco- 2. In this section all dairy products are expressed as total nomics Working Paper, University of Wisconsin­Madison. solids, milk equivalent, to facilitate comparisons. U.S. Dairy Export Council, International Dairy Food Association, 3. The European Union and, to a lesser extent, Canada, and National Milk Producers Federation, American Dairy Prod- the United States have had substantive export subsidy ucts Institute. 1999. "White Paper: Preparing for the Seattle allowances under the URAA and used them--a major impedi- Round: U.S. Dairy Industry Negotiating Priorities." July 21. ment to the expansion of dairy production in many developing Zhu, Y., T. L Cox, and J. P. Chavas. 1999. "An Economic Analysis countries. of the Effects of the Uruguay Round Agreement and Full 4. This discussion draws on Peng (2002). Trade Liberalization on the World Dairy Sector." Canadian 5. See Soufi (2002) for further details. Journal of Agricultural Economics 47: 187­200. 10 RICE: GLOBAL TRADE, PROTECTIONIST POLICIES, AND THE IMPACT OF TRADE LIBER ALIZATION Eric J. Wailes Rice is one of the most important food grains in the markets. Clearly, however, domestic price stabiliza- world, accounting for more than 20 percent of tion policies have been pursued by restricting im- global calories consumed and 29 percent in low- ports, in turn contributing substantially to interna- income countries (table 10.1). Thus, policies that tional market thinness. Therefore, it is difficult to affect rice prices, production, and trade have a large ignore the effect of domestic stabilization policies impact on the poor. achieved through import and export restrictions Despite the importance of rice as a basic staple, as a significant cause of international rice price global trade accounts for only 6.5 percent of con- instability. sumption. That means that most countries are self- In addition to the thinness of rice trade, another sufficient in rice and face increased price volatility important structural characteristic is the geo- in times of production shortfalls. By contrast, graphic concentration of production and con- wheat trade accounts for 18 percent of consump- sumption in Asia. More than 90 percent of produc- tion, corn for 12 percent, and soybeans for 35 per- tion and consumption occur in Asia--nearly cent (USDA PS&D 2003). The thinness of trade for two-thirds of it in just three countries (China, rice stems primarily from the use of protectionist India, and Indonesia). With as much as 40 percent mechanisms to achieve national policy objectives of Asian rice cultivated under rain-fed systems, of domestic food security and support for producer the monsoon weather effects are magnified on prices and incomes in major rice-producing and rice trade. -consuming countries (box 10.1). Finally, there is substantial market segmentation Jayne (1993) argues that the link between by rice type and quality. A key structural dimension domestic stabilization policies and instability in is the degree of end-use differentiation. Substitu- world rice prices has been exaggerated, empha- tion among rice types and qualities is limited by sizing instead the role of thin and fragmented differences in taste preferences. Low substitutability The author acknowledges the assistance of Frank Fuller and Alvaro Durand-Morat for model computations and expresses appreciation to John Beghin, Don Mitchell, and Kenneth Young for useful comments. The annex for chapter 10, on the CD-ROM, contains more extensive tables and figures on market data, policies, and simulation results. 177 178 Global Agricultural Trade and Developing Countries TABLE 10.1 Share of Calories from Rice by Region and Income Level, 2000 Total Calories Rice Calories Share of Calories Region Per Capita Per Capita from Rice (%) World 2,805 576 20.5 Developed countries 3,260 118 3.6 Developing countries 2,679 703 26.2 Low-income countries 2,405 702 29.2 Low-income food-deficit countries 2,625 732 27.9 Africa 2,434 178 7.3 Asia 2,713 856 31.6 Sub-Saharan Africa 2,226 174 7.8 South America 2,838 315 11.1 North and Central America 3,411 117 3.4 Europe 3,250 45 1.4 Source: FAOSTAT. BOX 10.1 Definitions of Rice Trade Flows in This Study The international rice trade is differentiated by kernels. Paddy rice refers to rice as it is harvested type, quality, degree of processing, and degree in the field before the husk and bran layer are of milling. Long-grain varieties are typically removed. Brown rice, also referred to as cargo or longer than 6.2 millimeters , while medium- and husked rice, has had the husk removed but short-grain varieties are 6.2 millimeters or less. retains the bran layer. Milled rice, also referred to Many factors enter into the designation of qual- as white rice, has had both the husk and bran ity for long-grain rice, including share of broken layers removed. The fragrant rice varieties, bas- kernels, seeds, chalkiness, and color. In this study mati and jasmine, are generally considered long- high quality refers to grain that contains 10 per- grain types but are marketed and priced in cent or less of broken kernels and low quality to global markets differently from unscented long- rice that contains more than 10 percent broken grain varieties. for rice exists on both demand (mill and end-use) production area, they also limit the production of and supply sides. On the demand side, the closest other crops that cannot withstand flood condi- substitute is wheat, particularly important in South tions. Development of rice varieties that will be Asia (India and Pakistan). In many Asian nations much less dependent on water will have the poten- rice has become an inferior good, so that as tial to greatly expand production areas suitable for incomes rise it is replaced by meats, fruits, and cultivation, changing costs of production and geo- vegetables. graphic areas of comparative advantage and disad- On the supply side, different rice varieties vantage. As the first major food crop to have its require different climatic conditions and produc- genomic structure fully described, rice genomics tion and milling technologies. This limits the ability and biotechnology are progressing rapidly (Khush of producers to respond to price incentives by and Brar 2002). switching the type of rice produced. Production Thus, the combination of high levels of domestic benefits greatly from access to plentiful supplies of protection, geographic concentration, erratic surface or ground water and soils with poor weather, inelastic price responses in production and drainage that can maintain a flood condition. end-use markets, and relatively thinly traded volumes While these characteristics limit the potential rice results in volatile prices and trade (Wailes 2002). Rice: Global Trade, Protectionist Policies, and the Impact of Trade Liberalization 179 Rice Trade and Policies in the The major rice-producing countries are also the Major Producing and Consuming major rice-consuming countries and leading rice Nations exporters and importers (tables 10.2 and 10.3 and annex table 1 on the CD-ROM). Because rice has been so highly protected in both industrial and developing nations, trade liberaliza- China tion under the Uruguay Round Agreement on Agriculture is having a profound impact on the China, the largest rice-producing and -consuming international rice market (Wailes 2002). Trade has country, accounts for nearly a third of the global essentially doubled in volume and as a share of con- rice economy. Rice has been an important compo- sumption since the 1970s and 1980s (figure 10.1). nent of China's food grain security objectives and The changes in protection have been modest, has been managed through procurement support however, and rice remains one of the most protected prices to ensure stable supplies. Government rice food commodities in world trade. stocks increased in the late 1990s to about 100 mil- As a result of the more limited and longer mar- lion metric tons, 73 percent of domestic use. In ket access reforms required for developing coun- 1999 the government eliminated purchases of low- tries under the Uruguay Round, rice policies in quality early season rice and lowered the procure- developing countries have not changed signifi- ment prices for its rice purchases. The area planted cantly since the early 1990s. This lack of rice policy with rice has declined (USDA PS&D 2003), and rice reforms has intensified price volatility, placing a stocks were reduced by more than 30 percent by the heavy burden on poor consumers and on govern- end of 2002, to 67.6 million metric tons. In some ments to provide food distribution programs for coastal provinces the government has since elimi- the poor. The coefficient of variation of domestic nated its procurement policy entirely, leaving pro- prices in real terms over the past 20 years was 0.43 ducers to sell their rice in the open market (Wade in India and 0.26 in Indonesia; it was 0.37 in China and Junyang 2003). The government policy now over the past 16 years. However, some price stability emphasizes quality over quantity, and rice produc- was achieved in these Asian countries in the 1990s ers are quickly adopting improved quality varieties. because real world prices had fallen dramatically The rice tariff rate quota negotiated by China during this period as well as variability. was initially 2.66 million metric tons in 2002, FIGURE 10.1 World Rice Trade and Share of Total Use Million metric tons Percent 30.0 8.0 7.0 25.0 6.0 20.0 5.0 15.0 4.0 3.0 10.0 2.0 5.0 1.0 0.0 0.0 1973­741975­761977­781979­801981­821983­841985­861987­881989­901991­921993­941995­961997­981999­002001­02 Trade Percent of use Source: USDA PS&D 2003. 180 Global Agricultural Trade and Developing Countries TABLE 10.2 Leading Rice-Producing, -Consuming, -Exporting, and -Importing Countries Rank Producing Consuming Exporting Importing 1 China China Thailand Indonesia 2 India India India Nigeria 3 Indonesia Indonesia Vietnam Bangladesh 4 Bangladesh Bangladesh United States Iran 5 Vietnam Vietnam China Philippines 6 Thailand Japan Pakistan Brazil 7 Japan Thailand Uruguay Iraq 8 Myanmar Myanmar Argentina Saudi Arabia 9 Philippines Philippines Egypt European Union 10 Brazil Brazil Myanmar Senegal 11 United States Korea, Rep. of Australia China 12 Korea, Rep. of United States Japan South Africa 13 Pakistan Nigeria European Union Côte d'Ivoire 14 Egypt Egypt Guyana Malaysia 15 Cambodia Iran Ecuador Cuba Source: USDA PS&D 2003. equally divided between long-grain and medium- India and short-grain or other rice (WTO 2001). Only As the second-largest rice producer, consumer, and 10 percent of the long-grain tariff rate quota and exporter, India plays an important role in the global 50 percent of the medium-short grain quota are rice economy. India is a major supplier of low- designated for private firms. The tariff rate quota quality long-grain rice and fragrant basmati rice. rose to 3.78 million metric tons in 2003 and will Like China, India views rice as a strategic commod- increase to 5.32 million metric tons by 2004 (Sun ity for food security based on grains (rice and and Branson 2002; Zhang, Matthews, and Branson wheat). Consequently, the government intervenes 2002). Nearly all rice imports are fragrant jasmine in the market through grain procurement, price rice, primarily from Thailand. Domestic produc- supports, and export subsidies. In recent years the tion of fragrant rice is increasing, however, and government has procured some 25 percent of the displacing imports. Unless there is a significant annual harvested crop to replenish government adverse weather event, China is not expected to fill stocks. Since April 2001 the government has its rice tariff rate quota. In-quota tariffs are 1 per- actively subsidized rice exports at 50 percent cent for grains (including milled rice) and no more of procurement prices, underselling Pakistan, than 10 percent for partially processed grain prod- Thailand, and Vietnam in low-quality long-grain ucts. Over-quota tariffs will be 76 percent initially, markets by $15­$20 a metric ton. Major markets reduced to 65 percent in 2004 (WTO 2001). for India's low-quality parboiled and regular China is a significant exporter of low-quality long-grain rice include Bangladesh, Côte d'Ivoire, long-grain rice, with principal markets in Côte Indonesia, Nigeria, Philippines, and South Africa. d'Ivoire, Cuba, and Indonesia. Medium-grain rice Major markets for basmati rice include the is exported competitively into Russia, Japan, the European Union (EU), Iran, Kuwait, Saudi Arabia, Republic of Korea, and the Democratic Republic of and United Arab Emirates. Korea (Hansen and others 2002). While the state India bound its rice tariffs under the Uruguay trading agency handles most rice exports, export Round at zero percent. Until May 1997 all rice was subsidies are not considered necessary for China's imported through the Food Corporation of India. rice export shipments (except for out-of-condition Under an agreement to privatize the rice trade, the stock liquidation). government negotiated higher import tariffs that Rice: Global Trade, Protectionist Policies, and the Impact of Trade Liberalization 181 TABLE 10.3 Net Rice Trade, 1982­2002 (thousands of metric tons) Countries by Trade Status 1982­83 1990­91 1995­96 2001­02 Net exporters 9,167 12,041 17,415 24,522 Thailand 4,111 4,432 5,249 7,468 India -255 643 2,900 5,275 Vietnam -65 1,481 3,183 3,685 United States 2,222 2,070 2,302 2,699 China 631 731 15 1,738 Pakistan 1,159 1,347 1,733 1,225 Myanmar 739 181 140 1,000 Uruguay 191 302 619 550 Egypt 43 184 269 350 Australia 286 534 567 310 Argentina 106 140 440 222.5 Net importers 4,735 5,460 9,789 15,129 Indonesia 744 366 960 3375 Nigeria 774 260 325 1803 Iran 715 938 1431 1,250 Iraq 364 408 489 1175 Philippines -16 175 829 1,050 Saudi Arabia 507 553 716 919 Brazil 187 708 801 650 South Africa 152 339 466 650 Côte d'Ivoire 329 216 402 637.5 Malaysia 386 333 483 616.5 Japan -135 18 461 525 European Union 321 33 492 525 Mexico 53 280 290 472 Bangladesh 249 25 593 387.5 Eastern Europe 10 0 163 311 367.5 Hong Kong (China) 369 398 345 320 Turkey 44 227 301 272.5 Canada 102 182 233 242.5 Taiwan (China) -452 -150 -193 -54 Korea, Rep. of 44 -8 58 -55 Rest of the World 4,432 6,581 7,626 9,380 Source: USDA PS&D 2003. become effective April 2000. Current tariffs are management of imports by the state monopoly, the 80 percent on paddy, brown rice, and broken rice National Logistics Agency (BULOG). In late 1998, and 70 percent on milled rice. Indonesia agreed to liberalize the rice trade to pri- vate traders, but unable to sustain the domestic floor price, the government restored market powers Indonesia to BULOG. The third-largest rice-producing and -consuming Following Indonesia's financial collapse and po- country, Indonesia is also the largest rice importer. litical instability in the late 1990s, the government Rice policy, particularly price stabilization policy, sought to stabilize and support producer rice prices was historically implemented through quantitative through a specific rice tariff of 430 rupiahs (Rp) 182 Global Agricultural Trade and Developing Countries per kilogram (equivalent to a 30 percent ad val- a result, the government withdrew the 10 percent orem tariff). Nontariff barriers and trader response regulatory duty in 2002 and more recently reduced to risks and regulation (including a 2002 require- letter-of-credit margins from 100 percent to 25 per- ment for an import license and redlining) have cent. Import restrictions that remained in 2003 raised the effective rate of protection to 100 percent include a tariff of 22.5 percent, an advance income (Timmer 2002). Average border prices of milled tax of 3 percent, and a development surcharge of rice were $200 per metric ton in 2002, while 3.5 percent. Bangladesh imposes no quantitative monthly retail prices in Jakarta averaged $377 per restrictions. metric ton (Katial-Zemany and Alam 2003). It is believed that a significant share of imports in 2002 Vietnam was smuggled into the country, thanks to a porous border and this large difference between world and Vietnam produces the fifth-largest rice crop and is domestic prices. also the fifth-largest rice-consuming country. Fol- The tariff policy is currently under review, and lowing the adoption of the Doi Moi reform pro- producers are pressuring for an increase to Rp 510 gram in late 1986, Vietnam's rice economy recov- per kilogram, equivalent to a 36 percent tariff but ered, and by the mid-1990s Vietnam had become well below the WTO (World Trade Organization) the world's second-largest rice exporter. Vietnam bound rate of 160 percent until 2004 (Katial- exports both high- and low-quality long-grain rice. Zemany and Alam 2003). Floor prices for paddy Important export destinations include Cuba, and milled rice were increased by 13 percent in Indonesia, Iraq, Malaysia, and several African 2003. In early 2003 BULOG's status was changed countries. Rice exports and prices are under the from a state agency to a state trading enterprise. It control of the Ministry of Trade and Vietnam's continues to distribute subsidized rice to low- Food Association (Vinafood) (Young, Wailes, income consumers. Current import and domestic Cramer, and Tri Khiem 2002). price support policies clearly have negative conse- Vietnam has no significant production support quences for Indonesia's consumers, especially poor policies or export subsidy programs. Vietnam and consumers, and negative consequences on real the other major Asian rice exporters (China, India, wages and therefore economic growth. Pakistan, and Thailand) have discussed the forma- tion of a rice export cartel in response to the low world prices for rice since 1999. India rejected the Bangladesh idea, but the others are developing the concept. Bangladesh is the fourth-largest rice-producing and -consuming nation and an important but Thailand highly variable rice import market. Since much of the rice production in Bangladesh is dependent on Thailand has been the world's leading rice exporter monsoon weather, production can fluctuate greatly. for the past several decades. Private export compa- In 1998 Bangladesh was the world's second largest nies supply world markets with a wide range of importer at 2.5 million metric tons, but since 1998 long-grain rice, including the fragrant jasmine rice. it has imported an average of only 500,000 metric The primary government rice policy is the paddy tons annually. mortgage scheme, a loan program operated under In 2000 Bangladesh imposed an import tariff of the Bank for Agriculture and Agricultural Coopera- 5 percent on rice. The rate was raised to 25 percent tives (BAAC). Participating farmers can obtain in 2001, and a 10 percent regulatory duty was loans from BAAC using their crop as collateral. The added mid-year, along with an advance income tax loan price is set at 95 percent of a government- of 3 percent and a development surcharge of 2.5 per- determined target price. In 2002 loan rice prices cent. These import protections along with a crop were $8 to $10 per metric ton higher than market shortfall in 2001 and a policy shift to distributing prices (a 10 percent price support). Nearly a third money instead of food grains in the national food of the Thai crop was pledged to the loan price sup- distribution program resulted in a higher domestic port program. Government stocks increased as price and a rise in smuggled imports from India. As farmers defaulted on their loans. The government Rice: Global Trade, Protectionist Policies, and the Impact of Trade Liberalization 183 procured rice is milled and then exported through made through an auction at which importers sell government-to-government arrangements. rice to the Food Agency and simultaneously buy it back. The Food Agency selects bids that maximize the markup. They have averaged ¥100­¥200 per kilo- Japan gram ($1,000 to $2,000 per metric ton). Over-quota Japan's rice economy is supported by the high tariffs are ¥341 per kilogram, ($2,842 per metric ton prices paid by consumers. Japan controls rice in 2003). The average successful bid price in Decem- imports through a tariff rate quota with a prohibi- ber 2002 was $318 per metric ton. Summary meas- tive over-quota tariff. As the traditional staple food, ures of protection from the Organisation for Eco- rice dominates the government's agricultural policy nomic Co-operation and Development (OECD (Fukuda, Dyck, and Stout 2003). 2003) indicate that the average producer support In 1996 the government ended regulation of rice estimate in Japan for 2000­02 was 86 percent and the marketing, freeing up wholesale and retail markets nominal protection coefficient was 6.89. from government supervision and licensing Republic of Korea requirements. With market liberalization, farm- gate prices have declined. In 1998 the government The Republic of Korea also protected its rice sector adopted the Rice Farming Income Stabilization with an import ban until 1995, when it agreed to a Program. When prices fall below a seven-year, minimum market access import commitment in moving-average standard rice price, producers are the Uruguay Round. In 2004, the final year of com- paid 80 percent of the difference between the cur- mitment, Korea will import 205,000 metric tons, rent year price and the standard price. Payments are 4 percent of domestic consumption in the 1986­88 made from the Rice Farming Income Stabilization base period. Consumption has been declining and, Fund, with 25 percent of contributions from rice coupled with rising minimum market access producers and 75 percent from the government. imports, this has resulted in excessive stocks. Participation is voluntary, but participants must In April 2002 the government released "A Com- also enroll in the Production Adjustment Promo- prehensive Plan on the Rice Industry" to cope with tion Program, which diverts land from rice to other the structural problem of oversupply and to pre- crops (wheat, barley, soybeans, forages, vegetables, pare for future restructuring. The government had and fruits). Since stabilization fund payments are relied on a procurement program to support farm tied to a diversion program, Japan claims Blue Box prices. In 2002 it procured 789,000 metric tons of a treatment (see chapter 3). Income stabilization pay- total production of 4.9 million metric tons at 2,097 ments to rice producers in 1999, the most recently won per kilogram ($1,667 per metric ton). Under reported year, were $815 million. Payments under the proposed comprehensive plan the government the diversion program were $1.03 billion. intends to decouple payments, moving from price Before the Uruguay Round Agreement on Agri- supports to income support. In 2002 the govern- culture, Japan had banned rice imports for 30 years, ment made a direct payment of 500,000 won except following the devastating production short- ($398) per hectare in agricultural promotion areas fall in 1993. Japan now imports 682,000 metric tons and 400,000 won ($319) per hectare in nonpromo- annually under a tariff rate quota, 7.2 percent of tion areas. The program is similar to Japan's domestic consumption in the base period 1986­88. income stabilization program in that it will be In-quota purchases are controlled exclusively by linked to a production adjustment system to shift the Food Agency, for which a markup of up to ¥292 rice areas to other crops (soybeans, forages, and fal- ($2.41 in 2001) per kilogram is allowed. low) and therefore will claim Blue Box WTO status. Imports are purchased through either ordinary In 2003 the government announced plans to keep market access or the simultaneous-buy-sell system. rice land fallow by paying producers 3 million won Under ordinary market access, which accounted for ($2,531) per hectare on 27,500 hectares--2.6 per- 85 percent of imports in 2001, the Food Agency cent of total rice area. OECD (2003) estimates an imports rice and resells it into Japan's domestic mar- average producer subsidy equivalent to 82 percent ket or donates it to food assistance programs. Under and a nominal protection coefficient of 5.35 per- the simultaneous-buy-sell system, purchases are cent for Korean rice producers in 2000­02. 184 Global Agricultural Trade and Developing Countries Imports are guided by the Uruguay Round min- territories at zero percent duty. Beginning in 2007, imum market access agreement and are assessed a tariffs on imports from the 48 least-developed 5 percent tariff under this tariff rate quota agree- countries will be progressively reduced to zero by ment. Imports, strictly controlled by the Ministry 2009 under the Everything but Arms agreement of Agriculture, have generally been of low-quality negotiated in 2001. rice and are made available to end-users through The export regime for rice is based on Uruguay controlled channels. Round agreement commitments, which limit refunds to 133,400 metric tons of milled rice equiv- European Union alent and a subsidy expenditure of no more than 36.8 million ($39.4 million). Export refunds are The European Union maintained an intervention price on paddy rice of 298.35 per metric ton. set by type of rice and destination. In 2003 export subsidies ranged from 111 to 165 ($119 to $177) Since 1996 the European Union has accumulated per metric ton. The OECD (2003) estimates the intervention stocks as a result of increased produc- producer subsidy equivalent at 31 percent and tion and imports. Direct payments were introduced nominal protection coefficient at 1.24 for 2000­02. in 1997, with payments up to a maximum guaran- teed area of 433,123 hectares. The current direct payment rate is 325.70 per hectare. Based on aver- United States age yields, the direct payment is equivalent to 52.65 per metric ton. Total support to rice pro- The United States is the world's fourth-largest rice ducers, taking into account the intervention price exporter, exporting nearly 45 percent of its produc- and direct payment, is 351 per metric ton (Com- tion. Under the 2002 Farm Bill, the U.S. govern- mission of the European Communities 2002). ment provides price supports through a market Under the Uruguay Round agreement, the Euro- loan rate of $143 per metric ton of paddy rice. A pean Union agreed to convert variable levies to market loan deficiency payment is made if the fixed tariffs and to reduce them by 26 percent by world reference price falls below the market loan 2000. Current tariff levels are 211 per metric ton rate. The 2002 crop received an average payment of for paddy, 264 per metric ton for brown rice, and $73 per metric ton. 416 per metric ton for milled rice. Import prices Producers also receive income support through of brown rice were approximately 250 per metric two payment programs, a fixed decoupled direct ton in 2003, so the 264 tariff provides a protection payment of $51.80 per metric ton and a decoupled rate of 105 percent. Tariff escalation makes the tar- countercyclical payment when the direct payment iff rate on milled rice prohibitive. plus the market price or market loan rate (whichever A variety of tariff concessions and preferences is higher) are below a target price of $231.48 per for EU rice imports exist. Brown basmati imports metric ton.1 When the market price is below the from India and Pakistan are given a 250 per metric market loan rate, the maximum countercyclical pay- ton reduction, resulting in a tariff of 14 per metric ment is $36.68 per metric ton. Both direct payment ton. With the accession to the European Union of and countercyclical payment are made on 85 percent Austria, Finland, and Sweden in 1995, a tariff rate of a fixed historical production level. quota was negotiated with zero tariff per metric ton Rice imports are subject to tariffs of $14 per met- on imports of 63,000 metric tons of milled rice, ric ton for milled rice, 11.2 percent ad valorem for 88 per metric ton on imports of 20,000 metric parboiled, $21 per metric ton for brown, $8.30 per tons of brown rice, and 28 per metric ton on metric ton for basmati brown, and $18 per metric imports of 80,000 metric tons of broken rice. Egypt ton for paddy rice. In 2002, 10 percent of exports has an import concession for 39,000 metric tons at (380,000 metric tons) were funded by government a 25 percent tariff reduction, and Bangladesh has a programs, all food aid shipments. Export subsidies 4,000 metric ton concession for brown rice at a under the Export Enhancement Program have not 50 percent tariff reduction. Preferences are given been used for U.S. rice exports since 1996. The through a 110,000 metric ton quota to the African, OECD (2003) estimates a producer subsidy equiva- Caribbean, and Pacific countries at a 35 percent lent of 50 percent and a nominal protection coeffi- tariff reduction and overseas countries and cient of 1.77 for 2000­02. Rice: Global Trade, Protectionist Policies, and the Impact of Trade Liberalization 185 Magnitude of Policy Distortions in tral and South America and the European Union. Key Rice Markets EU tariffs are 46 percent for brown rice but 80 per- cent for milled rice (table 10.4). In Mexico, paddy The major distortions in world rice markets are rice imports pay a 10 percent tariff while brown caused by import tariffs and tariff rate quotas in and milled rice pay a 20 percent tariff. key importing countries and price supports in key The effect of tariff escalation is seen in trade exporting countries. The global trade-weighted flows. Most of the trade in milled high-quality average tariff on all rice was 43.3 percent in 2000: long-grain rice goes to countries with low tariffs, 217 percent for medium- and short-grain rice and while most of the trade in brown and paddy rice 21 percent for long-grain rice. Medium-grain rice goes to countries with high tariff escalation. Trade- markets are far more distorted than long-grain rice weighted average tariffs for high-quality long-grain markets because of tariff rate quotas and quotas rice are estimated at 4.3 percent for milled rice, in the major medium-grain rice importing coun- 31.4 percent for brown rice, and 16.9 percent for tries of Japan, the Republic of Korea, and Taiwan paddy rice. Simple non-trade-weighted averages (China). OECD countries are a major source of dis- are 13.7 percent for milled rice, 18.7 percent for tortions, with average annual producer support brown rice, and 25.4 percent for paddy. reaching $25 billion in 2000­02. The greatest degree of protection in world rice Trade protection is also provided for domestic trade is in medium- and short-grain rice. Protec- milling industries. This protection is expressed in tion by Japan, the Republic of Korea, and Taiwan tariff escalation and is especially prevalent in Cen- TABLE 10.4 Schedule of Tariffs, Tariff Rate Quotas, and Quotas in Rice, 2002­03 Levels (percent) Long-Grain Tariff Rate Milled Medium­Short Country or Quota (1,000 Region Nonfragrant Fragrant Brown Paddy Milled Brown MetricTons) Bangladesh 22.5 22.5 22.5 22.5 22.5 22.5 Brazil 15.0 15.0 13.0 13.0 15.0 13.0 Canada 0.0 0.0 0.0 0.0 0.0 0.0 China 1.0 1.0 1.0 1.0 1.0 1.0 5,320 Costa Rica 35.0 35.0 35.0 20.0 35.0 35.0 Côte d'Ivoire 32.0 32.0 12.0 7.0 32.0 12.0 European Union 80.0 71.0 46.0 146.0 75.0 64.3 India 70.0 70.0 80.0 80.0 70.0 80.0 Indonesia 21.0 16.1 25.0 35.0 14.3 15.6 Japan (yen/kg) 341 341 341 341 341 341 682 Korea, Rep. of 5.0 5.0 5.0 5.0 5.0 5.0 204* Malaysia 0.0 0.0 0.0 0.0 0.0 0.0 Mexico 20.0 20.0 20.0 10.0 20.0 20.0 Nigeria 50.0 50.0 50.0 50.0 50.0 50.0 Philippines 50.0 50.0 50.0 50.0 50.0 50.0 Russia 5.0 5.0 5.0 5.0 5.0 5.0 Senegal 12.7 12.7 12.7 12.7 12.7 12.7 Taiwan (China) 0.0 210.0 0.0 0.0 210.0 229.4 Turkey 35.0 27.0 35.0 27.0 35.0 35.0 United States 14 14 21 18 14 21 ($/metric ton) * The Republic of Korea uses a quota rather than a Tariff Rate Quota (TRQ). Sources: AMAD (Agricultural Market Access Database), USDA, FAS GAIN reports. 186 Global Agricultural Trade and Developing Countries (China) lowers world export prices by some 100 per- rice trade and large price adjustments. An earlier cent. Currently, very few rice exporting countries version of the model was used to assess trade produce medium- and short-grain rice. The clear liberalization prior to the Uruguay Round (Cramer, beneficiaries of trade liberalization in medium- and Wailes, and Shui 1993; Cramer and others 1991). short-grain rice will be countries, especially China, For the current study RICEFLOW was more with a competitive advantage in production costs completely disaggregated by rice type and degree of and logistics relative to such other export competi- milling, and the baseline trade flows and elasticity tors as Australia, Egypt, and the United States. estimates were updated through 2000. The results Trade liberalization would be expected to stimu- reflect the effects of trade liberalization applied to late production of medium- and short-grain rice in year 2000 trade flows and prices. Detailed analysis other countries, but current varieties are suitable by quantities traded and prices are presented in only for temperate climates. Thus South American table 10.5. exporters such as Argentina and Uruguay could Complete liberalization in 2000 would have develop adapted varieties more quickly. Many other resulted in a significant expansion in global rice developing countries have tropical or subtropical trade of nearly 3.5 million metric tons, a 15 percent climates and would require a decade or more to increase in trade. Trade-weighted average export develop varieties that would be competitive in lib- prices would be 32.8 percent higher and trade- eralized medium- and short-grain rice markets. weighted import prices would be 13.5 percent lower. Production capacity in Australia and the United Trade in medium- and short-grain rice, where States and to some degree in China is increasingly initial protection was highest, would increase by constrained by lack of water. 73 percent. Producer export prices would rise 91 per- Long-grain rice markets are far less protected. cent and import prices would decline 27 percent.2 Tariffs in major low-quality rice-importing nations In the most protected medium- and short-grain such as Indonesia and Bangladesh are estimated to brown rice markets, trade would increase 141 per- reduce world prices by as much as 30 percent com- cent, export prices would increase 200 percent, and pared with full liberalization. The major impact is import prices would decrease 41 percent. Trade on consumers in these low-income developing would expand 59 percent in milled medium- and countries and on producers of low-quality long- short-grain rice markets, with export prices 71 per- grain rice in exporting countries such as India, cent higher and import prices 25 percent lower. Pakistan, Thailand, and Vietnam. While tariffs Because trade in high-quality, long-grain mar- are lower than on medium- and short-grain rice, kets is subject to much less protection, trade liberal- tariff escalation is substantial, particularly in the ization results in only slight increases in volume European Union and several Central and South traded--4 percent more for paddy rice, 7 percent American countries. This pattern of protection for brown rice, and 3 percent for milled rice. Export depresses world prices for milled high-quality long- prices increase only 2 percent but import prices fall grain rice relative to brown and paddy rice, creating 18 percent (10 percent for paddy, 31 percent for economic hardship for the milling industry in brown rice, and 4 percent for milled rice), improv- high-quality long-grain exporting countries such as ing consumer welfare in rice-importing countries. Thailand, Vietnam, and the United States. Protec- Most of the expansion in trade occurs in the low- tion in high-quality long-grain milled rice markets quality markets, such as Bangladesh, Indonesia, and is estimated to reduce prices by 10­20 percent. the Philippines. Traded volumes increase 13 per- cent and import prices fall 14 percent, improving consumer welfare in many low-income developing Trade Flow and Price Impact of countries. Removing protection in these markets Rice Trade Liberalization also improves producer welfare in developing coun- Estimates of the impact of the elimination of tries as export prices rise 7 percent. In the fragrant import tariffs and export subsidies using a spatial rice market liberalization results in a 41.5 percent equilibrium model, RICEFLOW (Durand-Morat lower import price but only slight increases in the and Wailes 2003), show a significant expansion of volume traded and the export price. Rice: Global Trade, Protectionist Policies, and the Impact of Trade Liberalization 187 TABLE 10.5 Simulation Results for Rice Trade Liberalization Using RICEFLOW, 2000 Rice Type Baseline Free Trade Change (%) Long grain High-quality Paddy Quantity (metric tons) 1,035,320 1,081,254 4.4 Export price ($/metric ton) 149.21 154.67 3.7 Import price ($/metric ton) 185.51 166.89 -10.0 Brown Quantity (metric tons) 856,798 916,721 7.0 Export price ($/metric ton) 223.75 219.25 -2.0 Import price ($/metric ton) 363.32 250.64 -31.0 Milled Quantity (metric tons) 7,495,594 7,704,482 2.8 Export price ($/metric ton) 225.97 225.58 -0.2 Import price ($/metric ton) 262.06 252.16 -3.8 Low-quality Milled Quantity (metric tons) 8,084,093 9,149,728 13.2 Export price ($/metric ton) 177.05 188.70 6.6 Import price ($/metric ton) 248.19 213.09 -14.1 Fragrant Milled Quantity (metric tons) 2,449,711 2,467,502 0.7 Export price ($/metric ton) 265.24 267.07 0.7 Import price ($/metric ton) 511.20 299.07 -41.5 All long grain Quantity (metric tons) 19,921,516 21,319,687 7.0 Export price ($/metric ton) 206.87 210.68 1.8 Import price ($/metric ton) 287.45 236.43 -17.7 Medium and short grain Brown Quantity (metric tons) 483,063 1,162,478 140.6 Export price ($/metric ton) 271.80 814.47 199.7 Import price ($/metric ton) 1438.54 842.75 -41.4 Milled Quantity (metric tons) 2,487,760 3,946,170 58.6 Export price ($/metric ton) 367.71 628.92 71.0 Import price ($/metric ton) 855.89 645.69 -24.6 All medium and short grain Quantity (metric tons) 2,970,823 5,108,648 72.0 Export Price ($/metric ton) 352.11 671.14 90.6 Import Price ($/metric ton) 950.63 690.53 -27.4 All rice Quantity (metric tons) 22,892,339 26,428,335 15.4 Export Price ($/metric ton) 225.71 299.69 32.8 Import Price ($/metric ton) 373.51 322.97 -13.5 Source: Durand-Morat and Wailes 2003. 188 Global Agricultural Trade and Developing Countries Welfare Impact of Rice Trade and $23 million to exporters. Most of these gains Liberalization are for high-quality milled rice ($69 million) and brown rice ($124 million). Liberalization of paddy Global rice trade liberalization results in a total eco- nomic surplus gain of $7.4 billion annually.3 rice trade improves the welfare of exporters by $2.4 million and of importers by $22.4 million. Importing countries have a net gain of $5.4 billion Liberalization of low-quality rice trade improves and exporting countries a net gain of $2 billion. the welfare of importing countries by $315 million Gains vary considerably by country, rice type, and and exporters by $52 million. degree of milling. Again these small net figures hide the large transfers at work between sellers and buyers of rice. Impact on Price Importers and Exporters Nearly all of the net gains are captured by develop- ing countries. Reform of fragrant rice trade is In most rice-importing countries, consumers gain estimated to improve the welfare of importers by ($32.8 billion for all importers), but producers lose $547 million, a result due primarily to Japan. ($27.2 billion). In some countries with large but Exporting countries (India, Pakistan, and Thailand) not prohibitive tariffs, significant tax revenues gain marginally in the net, although their producers evaporate under free trade ($2.9 billion in aggre- do gain substantially. gate), while significant public savings occur with the removal of domestic support ($2.7 billion in aggregate). Impact by Country In rice-exporting countries producers gain from higher prices and expanded output ($70.3 billion), Results by country or region depend on the type of while consumers lose ($68.8 billion). Among rice and degree of protection. The results discussed exporters, China accounts for the bulk of the pro- here are for some key countries that are highly pro- ducer gains and consumer losses. Behind the net tectionist or large traders of rice. gains are much larger transfers between producers and consumers. When these transfers are normal- Asian importers. Among Asian importers, Japan, ized by population to account for the large number the most protectionist country in rice trade, would of producers and consumers in China and some gain the most from liberalization. Medium-grain other countries, the transfers are much smaller and white rice prices would decline from $3,098 per less daunting than they appear. Many households metric ton to $656 per metric ton, while the volume are involved in both production and consumption. of trade would increase from 392,000 metric tons The net buyers detached from production activities to 2.18 million metric tons. This results in a welfare are the largest losers. gain of $3.6 billion per year, with producers losing $19.2 billion and consumers gaining $24.2 billion. Savings from removing farm programs more than Impact by Type of Rice offset the loss in tariff revenue. This logic of large transfers between consumers The patterns for the Republic of Korea and and producers holds on examination of the impact Taiwan (China) are similar. Border reform would of reforms by rice type. Reform of trade in triple Korean imports of medium-grain rice to medium-grain milled rice accounts for more than 306,000 metric tons. Prices would decline from 60 percent of the total global welfare improvement, $1,952 per metric ton to $840 per metric ton. Korea at $4.3 billion, with importers benefiting by $3.4 bil- also imports fragrant rice. With liberalization, the lion and exporters by $905 million. A breakdown by fragrant rice price would fall from $2,003 per metric milling stage reveals that importers of medium- ton to $288 per metric ton. Consumers would expe- grain brown rice benefit by $1 billion and exporters rience a net gain of $6.2 billion, while producers by $449 million. Liberalization of long-grain rice would lose $5.9 billion. Taiwan (China) shows trade trade generates improvements of $1.14 billion, with and welfare patterns similar to those of Korea, but importers gaining $1.06 billion and exporters just the welfare gains are of magnitude smaller. $80 million. High-quality rice trade yields welfare The Philippines is a major low-quality, long- gains of $218 million--$195 million to importers grain rice importer. Elimination of import tariffs Rice: Global Trade, Protectionist Policies, and the Impact of Trade Liberalization 189 would result in an increase of imports from tons to 3.4 million metric tons, milled low-quality 787,000 metric tons to 1.02 million metric tons, long-grain exports from 1.6 million metric tons to induced by price declines to $215 per metric tons. 1.8 million metric tons, and fragrant rice exports Consumers would gain $701 million annually, and from 1.21 million metric tons to 1.23 million met- producers lose from lower prices by $629 million. ric tons. Price increases would be modest and lead The largest rice importer, Indonesia, would ben- to small gains to producers of $123 million annu- efit from tariff reform of its low-quality long-grain ally, while consumers would lose $101 million. rice imports. The volume of imports would in- In India producers of long-grain rice would gain crease from 1.3 million metric tons to 1.7 million substantially ($973 million) but the gains would be metric tons. Prices would decline from $228 per almost offset by losses to consumers ($967 mil- metric ton to $196 per metric ton. Producers would lion). These figures are the results of moderate lose $1.02 billion annually while consumers would price changes applied to large volumes and repre- gain $1.07 billion. sent moderate impacts per producer or consumer. Asian exporters. China is the world's largest pro- Other exporters. Among other exporters, the ducer of medium-grain rice and would therefore be United States would be the next most important the largest export beneficiary of medium-grain rice beneficiary of rice trade liberalization after China. trade liberalization. Exports would more than dou- Milled medium-grain rice exports would increase ble, from 614,000 metric tons to 1.47 million metric from 226,000 metric tons to 383,000 metric tons, tons. Export prices would increase from $270 per with prices rising from $270 per metric ton to metric ton to $647 per metric ton. Brown medium- $617 per metric ton. Brown medium-grain exports grain rice trade would increase from 113,000 metric would increase from 292,000 metric tons to tons to 403,000 metric tons, with prices rising from 594,000 metric tons, and prices would rise from $233 per metric ton to $834 per metric ton. China $296 per metric ton to $803 per metric ton. The is also a significant exporter of low-quality long- United States is also a major exporter of high- grain rice. With trade liberalization, exports would quality long-grain rice. Summing across all rice increase from 1.9 million metric tons to 2.3 million imports and exports, the net gain to the United metric tons, and prices would increase from $178 States would be $326 million annually, a result of per metric ton to $190 per ton. Producers would higher total gains to producers of $2.2 billion and gain $64.2 billion in aggregate, and consumers losses to consumers of $1.9 billion annually. would lose $63.6 billion, a large aggregate loss but Australia is the third largest medium-grain rice less so when normalized by population. producer and exporter and would also benefit Vietnam is the major low-quality long-grain rice greatly from rice trade liberalization. Exports of exporter. Therefore tariff reform by importers of milled medium-grain rice would increase from this type of rice would mostly benefit Vietnam. 475,000 metric tons to 756,000 metric tons, with Exports would be expected to increase from 2.7 mil- prices rising from $271 per metric ton to $615 per lion metric tons to 3.1 million metric tons, and metric ton. The net welfare gain for milled medium- prices would rise moderately to $185 per metric grain rice is $211 million. Brown medium-grain rice ton. Vietnam has steadily increased its volume of export prices would increase from $235 per metric high-quality milled long-grain rice. Trade reform ton to $805 per metric ton. Producers would gain would increase this volume moderately as well as its $1.03 billion from higher prices, while consumers price. In aggregate, Vietnamese consumers would lose $745 million. lose from higher prices by $210 million annually, The fourth major medium-grain exporter is but producers would gain $229 million. Egypt. Trade reform would result in an increase in Thailand is the world's dominant rice-exporting exports of milled medium-grain rice from 326,000 nation. All Thai exports are long grain, which is the metric tons to 448,000 metric tons and an increase least protected rice type in world trade. As a result, in prices from $298 per metric ton to $629 per met- the benefits of rice trade liberalization are small for ric ton. Producers would gain $1.39 billion and Thailand. Milled high-quality long-grain rice consumers lose $1.26 billion, with a moderate exports would increase from 3.3 million metric aggregate net gain of $128 million. 190 Global Agricultural Trade and Developing Countries Other importers. Among importers outside These are countries that have been importing rice Asia, the European Union would have an overall without trade barriers. In a sense they have benefited net welfare gain from rice trade liberalization of from protection by other importers since this pro- $145 million annually. As an importer of high- tection has depressed export prices. Removing trade quality long-grain brown rice, the European Union barriers would boost export prices for all rice types would increase imports from 451,000 metric tons by degree of milling. This has negative consequences to 588,000 metric tons, and prices would fall from for countries that have had little or no import pro- $496 per metric ton to $260 per metric ton. The tection in rice. Most seriously affected would be aggregate welfare gain for high-quality long-grain Turkey, a major importer of medium-grain rice, brown rice imports would be $138 million annu- which faces much higher export supply prices after ally. This gain is offset by the higher prices that the global trade reform. The estimated net welfare loss European Union would pay for medium-grain for Turkey is $137 million.All importers of medium- imports, up from $372 per metric ton to $624 per grain rice, except Japan, the Republic of Korea, and metric ton. The volume of medium-grain imports Taiwan (China) lose as a result of significantly higher would decline from 645,000 metric tons to 595,000 import prices after global reform. The same situa- metric tons. The aggregate welfare change would be tion holds for long-grain rice importers that have a gain to consumers of $254 million annually and a little or no import protection. This includes Middle loss to producers of $109 million. Eastern countries such as Iran, Iraq, and Saudi Arabia. Brazil, Canada, Hong Kong (China), Africa. Nigeria became a major rice importer Malaysia,Singapore,and South Africa would also not when it relaxed quantitative restrictions to rely pri- be expected to benefit from rice trade liberalization. marily on tariffs. Nigeria imports milled high- and low-quality parboiled rice. High-quality imports Dynamic Analysis of Rice Trade would increase from 36,000 metric tons to 144,000 and Domestic Policy Reforms metric tons, and low-quality imports would in- crease from 682,000 metric tons to 877,000 met- The RICEFLOW model used for the trade and wel- ric tons. Prices would fall substantially for both. fare analysis presented above is a static spatial equi- Rice producers would lose $186 million annually librium framework of excess supply and demand while consumers would gain $271 million. Several equations. It does not allow for analysis of domestic smaller African developing nations would gain farm policies. For that, the Arkansas Global Rice similarly, with large gains to consumers partially Model (AGRM) was used. The AGRM is a partial offset by losses to producers. equilibrium nonspatial dynamic econometric model of the global rice economy (Fuller, Wailes, North and Central America. Central American and Djunaidi 2003; Wailes, Cramer, Chavez, and paddy rice importers would capture most of the Hansen 2000). The AGRM structure is based on gains associated with liberalization of paddy rice. equations for supply (expressed for estimated area On the export side the analysis does not change harvested and yields) and demand (domestic con- current rules in most countries, which ban paddy sumption, exports, imports, and ending stocks). export. Only Argentina and the United States cur- Rice prices are endogenized, with world reference rently export paddy. The net gain to these two equilibrium prices for long-grain and medium- countries would be $2.4 million. Paddy rice grain rice. The AGRM is used to generate baseline importers--Costa Rica, El Salvador, Guatemala, estimates for domestic and international rice for Honduras, Mexico, and Nicaragua--would have a the FAPRI outlook (FAPRI 2004). net gain of $22.4 million from lower import prices For this analysis, policy interventions in rice and increased imports. The expanded trade would supply that are trade distorting (Amber Box in benefit the domestic milling industries in the WTO parlance) were removed. The model was also importing countries and rice consumers at the simulated for the removal of import tariffs and expense of rice producers. export subsidies, to provide perspective. Finally AGRM was used to examine the net effect of com- Other countries. Several developing countries plete policy reform including domestic support, and regions would lose from rice trade liberalization. import protection, and export subsidies. Rice: Global Trade, Protectionist Policies, and the Impact of Trade Liberalization 191 The AGRM baseline global trade (sum of all 70­80 percent, a result similar to the findings using exports) projections are 27.9 million metric tons in the RICEFLOW model. 2005 increasing to 33.6 million metric tons by 2012.4 Long-grain rice prices in the baseline begin Policy Implications and Conclusions at $232 per metric ton in 2005 and increase to $277 per metric ton by 2012. Medium-grain prices rise Despite the importance of rice as a basic food sta- from $332 per metric ton in 2005 to $406 per met- ple, especially in developing countries, rice trade ric ton by 2012. accounts for only 6.5 percent of consumption. Such The removal of tariffs dominates all policy limited trade is due partly to preferences for spe- reform scenarios. Global rice trade increases by cific types and grades of rice, but also to protection- 3.5 million metric tons in 2005 and continues to ist policies based on food security objectives or expand to 5.3 million metric tons above the base- price and income support for producers. The trade- line. The removal of export subsidies reduces weighted average import tariff on rice was 43 per- global rice trade in the short term by 720,000 met- cent in 2000, and tariff escalation is common, to ric tons, but the long-term effect is negligible. protect rice milling industries. Taken together, the tariff effects swamp the export Several market and production characteris- subsidy effects, and global trade is higher by tics make rice prices more volatile than the prices 2.7 million metric tons in 2005 and by 5.2 million of most other commodities. Much of Asian rice metric tons in 2012. Elimination of domestic production is subject to monsoon climates, re- supports in the United States, the European Union, sulting in uncertain yields. Global rice trade is and Japan reduces trade very slightly in the short highly segmented by rice type (long and medium), term and not at all over the longer term. The com- degree of processing (milled, brown, and paddy), bined effect of the removal of tariff barriers, export and quality (generally pertaining to the percent of subsidies, and domestic supports increases trade broken kernels). As a staple food, the demand for by 2.4 million metric tons in 2005 and by 4.9 mil- rice is not very responsive to price and income lion metric tons in 2012. The 15 percent expansion changes. of global rice trade given by the more aggregated The combination of a high degree of protection, but dynamic AGRM model is remarkably similar geographic concentration, market segmentation, to the static results generated by the RICEFLOW inelastic supply response to price, and inelastic model. demand response to price and income results in The impact on global export prices follows the volatile prices and volumes traded. Distortions in impact on trade, with the dominant impact on rice trade occur throughout the world. State trad- prices resulting from removal of import tariffs. The ing enterprises are pervasive in rice trade, most long-grain export price is higher by $23 per metric notably in China, Indonesia, India, Japan, Republic ton in the short term and by $43 per metric ton in of Korea, Vietnam, and Australia. State trading the longer term relative to the baseline level. In the tends to result in a lack of transparency in pricing more highly protected medium-grain rice market, and trade competitiveness. Thailand is a clear tariff removal boosts prices by $291 per metric ton exception, as rice trade is managed by a very com- in 2005 and by $340 per metric ton in 2012. The petitive group of export companies. impact of removal of export subsidies is important Domestic policy distortions exist in a number of only in the short term, with long-grain rice export major rice trading nations, including Japan, the prices 6 percent higher and medium-grain prices European Union, and the United States. In the 5 percent higher. The effect of removal of domestic United States and the European Union, domestic support is negligible throughout the projection support results in implicit or direct export subsi- period. The aggregate effects of policy reforms, dies. In Japan the government's commitment to including tariffs, export subsidies, and domestic support rice prices is based on an aggressive rice supports, is significant for both long-grain rice and land diversion program and a tightly managed tar- medium-grain rice prices. Long-grain rice export iff rate quota. prices are 18­22 percent higher. This result differs Policy reforms to eliminate protection in the from the RICEFLOW model result. Medium-grain global rice economy are estimated to boost rice prices are higher than baseline projections by economic welfare by more than $7.4 billion per 192 Global Agricultural Trade and Developing Countries year. But the real story is the large transfers between protection would experience higher prices as a consumers and producers that lead to these net result. The benefits of removing moderate levels of gains. Most of the gains can be achieved by elimi- tariff protection, as in the case of Turkey, are nating tariffs on imports. In importing countries swamped by the price effect of free and expanded consumers gain $32.8 billion, while producers lose trade in medium-grain rice. $27.2 billion. Governments lose $2.9 billion in tar- Domestic policy reforms in the United States iff revenue but gain $2.7 billion by eliminating and the European Union are estimated to reduce domestic supports. The net welfare gain to rice- rice exports by less than 5 percent in the initial importing countries is estimated at $5.4 billion. In years and to have little or no impact on trade in the exporting countries producers gain $70.2 billion, longer term. Prices are estimated to be 5­10 percent while consumers lose $68.8 billion. Imports by the higher initially, but the effect diminishes to zero exporting countries result in a loss of tariff revenue over the longer term. of $5.3 million while elimination of domestic sup- The multilateral and regional trade policy ports saves $598 million. The net welfare gain in reforms adopted since the early 1990s have con- exporting countries is $2 billion. tributed to an expansion in rice trade and more sta- With global policy reform, rice trade is estimated ble prices. The achievements of the Uruguay Round to increase by 10­15 percent. Prices received by Agreement on Agriculture include the opening of exporters would be 25­35 percent higher. Prices the previously closed Japanese and Korean markets. paid by importers would be 10­40 percent lower, But the limits on domestic support and export sub- depending on the type of rice. Rice trade, despite the sidies have yet to have a significant impact. Regional expansion, would remain relatively thin. Complete agreements such as NAFTA (North American Free policy reform would result in an increase in rice Trade Agreement) and Mercosur have increased rice trade from the current level of 6.5 percent of con- trade in the Western Hemisphere. The prospects for sumption to 8.4 percent by 2012. Thus, one of the the success of the Doha Round of the WTO hinge to major sources of world rice price instability is likely a great extent on continuing the expansion of mar- to remain after liberalization. Global rice stocks ket access, reductions of tariffs, and limits on export have declined by 30 percent between 2000 and 2003. subsidies required to achieve the benefits estimated Thus, the ability of stocks to buffer supply shocks has here from global trade liberalization. been markedly reduced. Global rice trade liberaliza- tion would make low-income, net-rice-importing Notes countries more reliant on world rice trade, likely 1. The direct payment is paid on a historical base produc- reducing political and food security. tion, decoupled from both current production and current mar- Medium-grain rice is the most protected rice ket conditions. The countercyclical payment is also paid on a type. Consequently, policy reform would have its historical base production, and although payment is decoupled from current production, it is triggered by current market price biggest impact on countries that export and import conditions. The government claims both payments as Green Box medium-grain rice. Japan is estimated to capture in the WTO. nearly 70 percent of the global economic welfare 2. The large increase in the export prices for short- and medium-grain rice does not consider the likely supply responses gains. Other industrial countries, such as Australia, by less-competitive producers that could enter the market and the European Union, and the United States, that survive at that high price. Hence, this is an upper-bound esti- export medium-grain rice would also be significant mate of the likely price increase. 3. Consumer and surplus gains and losses are estimated beneficiaries of trade policy reform. using the results of the baseline and free trade results of the Countries that had little or no protection before RICEFLOW model. The welfare estimates for producers and reform are likely to be harmed by global policy consumers are detailed in annex table 3 on the CD-ROM. The reforms. This result is due to the large country results are reported for the major importing and exporting countries or regions by rice type and degree of milling. Annex impacts that increased imports in countries like table 4 on the CD-ROM includes the producer and consumer Japan would have, increasing the demand for welfare estimates with the impact on government revenues lost medium-grain rice and thereby boosting world due to tariff elimination and government expenditures elimi- nated because of the elimination of domestic support programs. prices. Countries like Turkey and Russia that have 4. Results are presented in annex table 5 and annex figures 1­3 imported medium-grain rice with moderate or no on the CD-ROM. Rice: Global Trade, Protectionist Policies, and the Impact of Trade Liberalization 193 References Khush, G. S., and D. S. Brar. 2002. "Biotechnology for Rice Breeding: Progress and Potential Impact." Presented at the Agricultural Market Access Database. (AMAD). Online: 20th Session of the International Rice Commission, http://www.amad.net/index.htm. July 23­25, Bangkok, Thailand. www.fao.org/DOCREP/ Childs, N. W., and L. Hoffman. 1999. "Upcoming World MEETING/004AC347E/AC347E00.HTM. Trade Organization Negotiations: Issues for the U.S. Rice Nielsen, C. P. 2002. "Vietnam in the International Rice Market." Sector." Rice Situation and Outlook. RCS-1999. U.S. Rapport 132. Fødevareøkonomisk Institut. Copenhagen. Department of Agriculture, Economic Research Service, OECD (Organisation for Economic Co-operation and Develop- Washington, D.C. ment). 2003. Agricultural Policies in OECD Countries: Commission of the European Communities. 2002. "Rice: Mar- Monitoring and Evaluation. Paris. kets, CMO and Medium Term Forecasts." Commission Staff Sun, X., and A. Branson. 2002. "Food and Agricultural Import Working Paper SEC(2002) 788. Brussels. Regulations and Standards, Interim Rules and Regulations Cramer, G. L., E. J. Wailes, and S. Shui. 1993. "The Impacts of of Tariff Rate Quota 2002." Foreign Agricultural Service Liberalizing Trade in the World Rice Market." American Jour- GAIN Report CH2007. U.S. Department of Agriculture, nal of Agricultural Economics 75 (February): 219­26. Washington, D.C. Cramer, G. L., E. J. Wailes, J. Goroski, and S. Phillips. 1991. "The Timmer, C. P. 2002. "Food Security and Rice Price Policy in Impact of Liberalizing Trade on the World Rice Market: Indonesia: The Economics and Politics of the Food Price A Spatial Model Including Rice Quality." Arkansas Experi- Dilemma." Indonesian Food Policy Program Working Paper ment Station Special Report 153. University of Arkansas, 14. BAPPENAS/Department Pertanian/USAID/DAI Food Fayetteville. Policy Advisory Team. Jakarta. Durand-Morat, A., and E. J. Wailes. 2003. "RICEFLOW(R): A USDA PS&D (U.S. Department of Agriculture, Production, Spatial Equilibrium Model of World Rice Trade." Staff Paper Supply and Distribution). 2003. PS&D Online. www.fas. SP 02 2003. University of Arkansas, Department of Agricul- usda.gov/psd. tural Economics and Agribusiness, Division of Agriculture, USDA, FAS. "Rice. Annual". Various years and countries. Global Fayetteville. Agricultural Information Network (GAIN) Reports, Foreign Food and Agricultural Policy Research Institute (FAPRI). 2004. Agricultural Service (FAS), Washington, D.C. FAPRI 2004 World Agricultural Outlook, Iowa State Univer- Wade, J., and J. Junyang. 2003. "China Grain and Feed Annual sity and University of Missouri, Ames, Iowa, Staff Report 1- 2003." Foreign Agricultural Service GAIN Report 3010. U.S. 04, ISSN 1534-4533. Department of Agriculture, Washington, D.C. Fukuda, H., J. Dyck, and J. Stout. 2003. "Rice Sector Policies in ------. 2002."Trade Liberalization in Rice." In P. Lynn Kennedy, Japan." RCS-0202-01. U.S. Department of Agriculture, Eco- ed., Agricultural Trade Policies in the New Millennium. nomic Research Service, Washington, D.C. New York: Haworth Press. Fuller, F., E. J. Wailes, and H. Djunaidi. 2003. "Revised Arkansas Wailes, E. J., G. Cramer, E. Chavez, and J. Hansen. 2000."Arkansas Global Rice Model." Staff Paper SP 01 2003. University of Global Rice Model: International Baseline Projections for Arkansas, Department of Agricultural Economics and 2000­2010." Arkansas Agricultural Experiment Station Agribusiness, Division of Agriculture, Fayetteville. Special Report 200. University of Arkansas, Fayetteville. Hansen, J., F. Fuller, F. Gale, F. Crook, E. Wailes, and M. Moore. WTO (World Trade Organization). 2001. "Report of the Work- 2002. "China's Japonica Rice Market: Growth and Competi- ing Party on the Accession of China." WT/MIN(01)/3. Pre- tiveness." Rice Situation and Outlook Yearbook. RCS-2002. sented at the Fourth Session of the Ministerial Conference, U.S. Department of Agriculture, Economic Research Service, November 9­14, Doha, Qatar. Market and Trade Economics Division, Washington, D.C. Young, K. B., E. J. Wailes, G. L. Cramer, and N. Tri Khiem. 2002. Jayne, T. S. 1993. "Sources and Effects of Instability in the World "Vietnam's Rice Economy: Developments and Prospects." Rice Market." MSU International Development Paper 13. Arkansas Agricultural Experiment Station Research Report Michigan State University, Department of Agricultural Eco- 968. University of Arkansas, Fayetteville. nomics, East Lansing, Michigan. Zhang, J., R. Matthews, and A. Branson. 2002. "Food and Agri- Katial-Zemany, A., and N. S. Alam. 2003. "Indonesia Grain cultural Import Regulations and Standards, Implementation and Feed Annual, 2003." Foreign Agricultural Service GAIN Measures for 2002 TRQ Allocation." Foreign Agricultural Report ID3008. U.S. Department of Agriculture, Washington, Service GAIN Report CH2008. U.S. Department of Agricul- D.C. ture, Washington, D.C. 11 WHEAT: THE GLOBAL MARKET, POLICIES, AND PRIORITIES Donald O. Mitchell and Myles Mielke Wheat is one of the most important food crops, pro- local production. Competing exporters such as viding nearly one-fifth of the world's calorie sup- Argentina, Kazakhstan, the Russian Federation, and plies. About 19 percent of the world's production is Ukraine are also harmed because they receive lower traded internationally, primarily as exports from the prices for their wheat. The surplus disposal pro- countries of the Organisation for Economic Co- grams are reduced during periods of low stocks and operation and Development (OECD)--including relative wheat shortages, thus contributing to Australia, Canada, the European Union (EU), and global price volatility. In addition, many exporting the United States--to developing countries to supply countries have resorted to export restrictions to basic food needs and the growing demand for prod- protect domestic consumers when prices are high, a ucts made from wheat flour,such as bread,pasta,and practice that further adds to global price volatility. noodles.1 Wheat is also the food crop most com- Such policies make it very difficult for importing monly stored as a buffer against production short- countries to rely on the world wheat market to ful- falls, with an average of 30 percent of the world's fill a significant portion of their needs because of wheat production carried over from one crop year to the uncertainty of world supply. Consequently, the next. The global wheat situation and wheat poli- many countries follow policies aimed at self- cies of major actors are thus central to the food secu- sufficiency and thus are deprived of the benefits of rity and dietary preferences of many countries. trade. Policy reforms that reduced global volatility Major OECD wheat exporters, such as the Euro- in wheat prices, cut production subsidies, and pean Union and the United States, support domes- improved access to exports during periods of high tic production. The support policies often lead to prices would reduce food security concerns. surpluses, which are then exported with subsidies This chapter discusses major trends and devel- or donated as food aid that is not emergency opments in the world wheat market and their related. Developing countries sometimes benefit impact on trade and food security. We begin by from such surplus disposal programs because they looking at the characteristics of wheat and trends in pay lower import prices or receive food aid. How- wheat production, use, trade, stocks, and prices. We ever, countries are also harmed by such programs then examine the policy environment, focusing because they depress world prices and discourage especially on trade policy and domestic support. 195 196 Global Agricultural Trade and Developing Countries Wheat Characteristics and Trends 1990s, compared to just 2 in 1980 (Carter and Wilson 1999). Wheat is produced in 120 countries and accounts for about 19 percent of the world's calorie supplies. Production and Yields It is used primarily as flour for making bread, pas- try, pasta, or noodles. It is also used to feed livestock, Wheat is produced under a variety of climatic con- with feed use accounting for about 17 percent of ditions using technologies ranging from fully global wheat consumption. In addition, the by- mechanized production and harvesting on large products from milling wheat into flour are used as tracts to manual planting and harvesting on small feed. Wheat stores for several years without deterio- plots. About 61 percent of wheat is produced in ration under proper conditions, making it well non-OECD countries (table 11.1); this share has suited for use as a buffer against food shortages. been increasing over time as production has grown The many varieties of wheat have different pro- more rapidly in developing countries than in tein levels and varying milling and baking charac- OECD countries. The European Union, China, and teristics. The protein levels range from about 8 to India are the largest producers, with 18, 16, and 18 percent. High-protein wheat is better suited to 13 percent of global production, respectively. bread and pasta making, while lower protein wheat Wheat yields have increased significantly since is better suited for pastry and noodles. There is sub- the middle of the 20th century. From 1961 (when stitution between wheat varieties and blending of data on many countries first became available) to different varieties to produce flour with specific 2000, world wheat yields increased by an average of characteristics. The demand for high-quality wheat 2.4 percent each year. The increase in yields in and wheat with specific characteristics is increas- developing countries came from using more inputs ing, as buyers become more sophisticated. Protein (such as fertilizer) and high-yielding semi-dwarf premiums have steadily increased since the early seed varieties developed at the International Maize 1980s, as responsibility for import decisions has and Wheat Improvement Center (CIMMYT) in shifted to the private sector, which is better able Mexico and released to developing countries in the to evaluate quality and more willing to pay mid-1960s in what is now known as the Green Rev- premiums (Wilson and Dahl 1999). There has also olution. These varieties, adapted to local condi- been greater specificity in purchasing contracts. tions, were quickly adopted (Dalrymple 1974). As For example, the Australian Wheat Board offered shown in figure 11.1, average yields in India are 34 different segregations of wheat in the mid- now very similar to those in the United States, at TABLE 11.1 Wheat Production, Trade, and Growth Rates 1989­91 to 1999­2001, by Region Millions of Tons Growth Rates (percent) Region Production Imports Exports Production Imports Exports World 583 107 108 0.5 0.4 0.4 OECD 227 19 80 0.6 6.2 -1.1 Non-OECD 356 88 28 0.4 -0.5 5.4 Africa 17 26 0 1.3 3.3 9.7 Americas 111 25 58 -0.4 6.8 -0.3 Asia 199 27 4 1.7 -2.1 15.1 Europe 130 8 17 -1.0 -4.3 -2.2 FSU 75 6 9 -1.5 -12.2 4.8 Middle East 29 12 4 0.5 3.2 -6.3 Oceania 24 1 17 6.6 5.4 5.4 Note: Production, imports, and exports are the average for 1999­2001 crop years, which begin with harvest and vary by country. Growth rates are for the average of 1999­2001 compared with 1989­91. Source: USDA PSD online database and USDA 2003. Wheat: The Global Market, Policies, and Priorities 197 FIGURE 11.1 Wheat Yields, U.S. and India, 1900­2000 Tons per hectare 3.0 2.5 2.0 1.5 United States India 1.0 0.5 0.0 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 Source: USDA. TABLE 11.2 Global Wheat and Wheat Products Exports, Selected Periods (millions of US$) Average Value Annual Percentage Increase Product 1970* 1980* 1990* 2000* 1970­80 1980­90 1990­2000 Wheat 3,146 15,502 15,572 14,399 17.3 0.0 -0.8 Bakery 227 1,362 3,913 8,108 19.6 11.1 7.6 Flour 408 1,889 1,748 1,763 16.6 -0.8 0.1 Pasta 39 285 841 1,508 21.9 11.4 6.0 Other 113 894 1,237 972 23.0 3.3 -2.4 Total wheat products 787 4,430 7,738 12,352 18.9 5.7 4.8 *Data is three-year average centered on year shown. Note: Values are in nominal U.S. dollars. Bakery products include bread and pastry. Other products include gluten feed and meal, bran, germ, and whole meal bulgur. Source: FAOSTAT. about 2.8 tons per hectare, thanks largely to the Soviet Union (FSU), reduced production because Green Revolution. The annual increase in yields of reduced input use and lower domestic demand. from 1950 to 2000 was 1.89 percent in the United States and 2.95 percent in India.2 Trade While improvements in wheat yields have con- tinued along historical trends, the growth of global Trade of wheat is primarily from OECD to non- wheat production has slowed to just 0.5 percent per OECD countries, with about three-quarters of year over the last decade (see table 11.1), largely global wheat exports coming from OECD countries because of slower consumption growth and the and 82 percent of imports absorbed by non-OECD corresponding adjustment in production. Area countries (see table 11.1). Trade of wheat grew only planted with wheat declined by 5.5 percent from 0.4 percent per year during the 1990s, while trade 1989­91 to 1999­2001, mostly as a result of land- in processed products made from wheat (bakery diversion policies of major exporters such as the products, flour, pasta, and other products) ex- United States. Certain regions, such as Oceania, panded more rapidly (table 11.2). This increase in increased production and exports during this wheat product trade has occurred despite tariff period because of favorable exchange rates and low escalation with higher levels of processing. Most of production costs, while others, such as the former the trade in processed products has been between 198 Global Agricultural Trade and Developing Countries developed countries. During 1999­2001 about tries tended to substitute wheat for other grains 85 percent of global exports and 77 percent of when prices were advantageous; this was particu- global imports of processed wheat products were larly true for those Asian countries that are sensitive by developed countries. Developing countries pri- to prices of grain imports for feed rations. The marily import wheat rather than products, with demand for other uses of wheat, such as industrial about 80 percent of total expenditures on wheat uses and as a food additive (gluten, starches, and so going for grain imports during 1999­2001. At the on), also has been steady during the past four same time, developing countries have increased decades, growing at an average of 1.6 percent their wheat product exports from one-third of the per year.3 average value of total wheat and product exports during 1979­81 to one-half the average value in Stocks 1999­2001. Wheat carryover, or ending-stocks, provide a buffer against wheat shortages during years of low produc- Use tion or rapid increases in demand. When stocks are Wheat use has grown faster than population--at high, prices tend to be low, and vice-versa. The level 2.5 percent per year since 1961, compared with of global ending-stocks as a percentage of con- population growth of 1.7 percent. More recently, sumption and real wheat prices are shown in fig- however, use has slowed, and per capita food con- ure 11.3. The inverse relationship is readily apparent. sumption has remained nearly constant for more The share of global wheat stocks held by the five than a decade in both developed and developing major exporters (Argentina, Australia, Canada, the countries (figure 11.2). Feed use has shown strong European Union, and the United States), which growth among both developed and developing together account for three-quarters of net exports, countries, but this growth has been offset by the declined from 80 percent in 1960 to about 20 percent dramatic drop in feed use in the countries of the in 2002. This dramatic shift occurred for two main FSU following the breakup of the Soviet Union. reasons. First, the share of global production of the Global use of wheat for feed rose hardly at all dur- five major exporters declined from a high of 46 per- ing 1981­2001, compared with average annual cent in 1963 to 33 percent in 2002 as production in growth of 7 percent during 1961­81. Among devel- developing countries increased more rapidly than oping countries, feed use grew by 2.4 percent per among major exporters. Second, policy changes in year in 1981­2001, compared to 6 percent per year the major exporters reduced government-held during the previous two decades. Developing coun- stocks.The consequence has been a shrinking supply FIGURE 11.2 Per Capita Food Consumption of Wheat Kilograms per year 150 120 90 60 30 0 1960 1965 1970 1975 1980 1985 1990 1995 2000 Developed Developing Source: FAO. Wheat: The Global Market, Policies, and Priorities 199 FIGURE 11.3 Wheat Ending-Stocks vs. Prices Stocks as percent of use Real U.S. fob price 50 500 40 400 30 300 20 200 10 100 0 0 1960 1965 1970 1975 1980 1985 1990 1995 2000 Ending stocks U.S. fob prices Source: USDA and World Bank. FIGURE 11.4 U.S. Wheat Price Constant 2002 $s per ton 600 500 400 300 200 100 0 1900 1920 1940 1960 1980 2000 Farm season average fob calendar year Source: USDA. of exportable wheat supplies that could lead to more sharp real price increases, but these were quickly volatile prices in the future. reversed as production increased to meet the rising imports. By the end of the 20th century, real U.S. Prices producer prices had declined by about 75 percent from the highs of the early 1900s (figure 11.4). During most of the 20th century, real wheat prices fell (by about 75 percent since 1900 and by about Overall Trends half since 1970), while the policies of the major exporters (with the exception of Argentina) were The overall trends in wheat show that production aimed primarily at supporting prices, expanding has grown more rapidly than population since exports, and restricting production through vari- 1961, and that, in recent years, production and ous schemes. A surge in wheat exports during the trade growth have slowed significantly because of 1970s, combined with the oil-price shock, led to slower consumption growth. Part of the recent 200 Global Agricultural Trade and Developing Countries slowdown in consumption has been due to the col- from foreign competition. Applied tariffs were lapse of the FSU, but, in addition, per capita con- often set high, and bound tariffs even higher, leav- sumption of wheat as food has stopped increasing ing open the possibility of future increases in ap- in both developed and developing countries. Trade plied tariffs. Wheat export subsidies were reduced in wheat products has grown more rapidly than by the European Union and the United States--the grain trade, especially among developed countries. countries with the largest export subsidies--but The steady decline in real wheat prices and the that was attributable more to budget constraints decline in stocks held by the major exporting coun- than to the URAA. Implementation of minimum tries as a share of world stocks could lead to greater access and tariff reductions have stalled as coun- price volatility. tries have introduced new measures to offset agreed commitments or to prevent them from taking Policy Environment effect. Under the Uruguay Round Agreement on Agricul- Market Access ture (URAA), member countries of the World Trade Organization (WTO) had to convert quanti- Most countries met the minimum market access tative restrictions on imports into bound tariffs, requirements of the URAA by establishing tariff reduce those tariffs over an implementation period, rate quotas (TRQs), which provided for reduced open their markets to imports under minimum tariff rates on a specified volume of imports access provisions, limit and reduce the most trade- (table 11.3). Imports above these quotas faced distorting forms of domestic support, and cap and higher tariffs. However, regional trading agree- reduce subsidized wheat exports. Despite these sig- ments often have provided even lower tariffs or nificant achievements, the amount of trade liberal- duty-free access to regional trading partners. For ization achieved in wheat was modest because of example, Mexico established a TRQ for wheat of the way the reforms were implemented. Many 605,000 tons at an in-quota tariff of 67 percent. countries applied the Uruguay Round provisions so Meanwhile, Canada and the United States receive a that they could protect producers in key sectors preferential tariff of 4.5 percent under the North TABLE 11.3 Tariffs and Tariff Rate Quotas, by Country (percent and millions of metric tons) Bound Tariffs Applied Tariffs Year of Regional Trade Country Report Final TRQ In-Quota Above-Quota Average Preferential Agreements Brazil 2001 750 0.0 55.0 12.5 0.0 Mercosur Canada 2001 227 0.7 62.8 1.3 0.0 NAFTA China 2001 7,884 1.0 74.0 -- -- -- Colombia 2001 692 124.0 130.0 12.5 -- Andean Ecuador 1999 480 19.0 23.6 9.2 -- -- European 2001 350 0.0 58.9 12.8 0.0 Central Union Europe Israel 2000 450 92.0 137.8 -- 0.0 EU, U.S. Japan 2001 5,740 249.2 414.3 -- -- -- Mexico 2001 605 50.0 67.0 67.0 4.5 NAFTA Morocco 2001 1,555 144.0 198.4 30.1 2.5 EU Poland 2000 280 25.0 64.0 20.0 0.0 EU South Africa 2001 108 20.0 93.0 -- -- -- Tunisia 2000 900 17.0 86.7 20.0 -- EU Venezuela 2000 1,317 24.0 117.0 11.0 -- -- -- Not available. Source: WTO, FAO, USDA, and ABARE 2002. Wheat: The Global Market, Policies, and Priorities 201 American Free Trade Agreement (NAFTA). Conse- Korea produced less than 1 percent of its consump- quently, virtually all of Mexico's wheat imports tion but used high tariffs to protect flour millers. come from Canada and the United States. Brazil's Japan has specific duties on wheat, flour, and pasta; wheat imports are mostly supplied by Argentina, Mexico had specific duties on bakery products. The which has a comparative advantage in geographic relatively large imports of bakery products despite proximity and in preferential treatment under the high tariffs suggest high demand or lack of compet- Mercosur regional trade agreement (Diaz-Bonilla itiveness of local bakers. 1999). Indonesia and Malaysia, which had low tariffs Turkey did not establish a TRQ for wheat, but for wheat and all wheat products, provide an inter- instead relied on tariff-only protection with a esting case of imports without much distortion. bound tariff of 188 percent and an applied tariff of There is still some tariff escalation, but the maxi- 55 percent. Because imports from the European mum tariff was a relatively low 6.3 percent on pasta Union receive a zero tariff under a regional trade imports in Malaysia and 5.0 percent on bakery agreement, however, most of Turkey's imports products in Indonesia. Imports reflect these low come from the European Union. The Russian Fed- tariffs, with wheat products accounting for almost eration, which is not a member of the WTO, pro- half of the value of imports in Malaysia and for vides preferential access to several FSU countries. one-third in Indonesia. This suggests that without The European Union has protected its producers tariff escalation, wheat product trade would have through variable import levies for many years as increased significantly, benefiting consumers. Trade part of the Common Agricultural Policy, but it in processed wheat products is concentrated within more recently resorted to TRQs on low- to free trade areas such as within the European Union medium-grade wheat to slow wheat imports from and NAFTA. The shares of global trade occurring Ukraine and the Russian Federation. within these two regions alone during 2000­01 Tariff escalation is a common practice that were 23, 36, 50, and 66 percent, respectively, for encourages trade in wheat grain rather than in wheat, flour, pasta, and bakery products. wheat products (table 11.4). Tariffs generally esca- Some countries have used nontariff barriers late with the degree of processing of wheat prod- (NTBs) to protect their domestic wheat markets. ucts. Brazil, the largest wheat importer during the For example, the United States resorted to phy- 1990s, imposed an average tariff on wheat imports tosanitary standards during the 1980s to block of 6.3 percent, but the tariff on wheat flour was wheat imports from Mexico. At that time, durum 13.5 percent, and those for pasta and bakery prod- wheat producers in the southwestern United States ucts were 18.5 and 20.5 percent, respectively. used the existence of Karnal bunt as a reason to Because of the high tariffs on value-added prod- block wheat imports from Mexico (Beattie and ucts, most imports were in the form of grain or Biggerstaff 1999).4 In an ironic twist, wheat imports flour. Bangladesh, Costa Rica, Egypt, Guatemala, from four southwestern states of the United States Jordan, the Philippines, and Uganda showed simi- are currently banned by Mexico because of con- lar patterns, with most tariffs escalating with cerns about Karnal bunt (USDA 2004). U.S. wheat greater processing and larger imports of the prod- was also barred from three major wheat markets ucts with the lower tariffs. Kenya, the Republic of during the second half of the 1990s because of Korea, Japan, and Mexico had different patterns phytosanitary concerns. Brazil, China, and India that may reflect stronger protection to producers or banned the import of U.S. wheat, in particular specific processors. For example, Kenya had a from the Pacific Northwest, based on the possible 35 percent tariff on wheat imports, but a 25 percent presence of tilletia controversa kuhn (TCK) fungus tariff on flour, pasta, and bakery products. Korea and mycotoxins. While all three cases were resolved had prohibitive tariffs on wheat flour imports, but by the end of the decade, U.S. wheat exports to much lower tariffs on wheat grain, pasta, and these markets have not recovered due to changing bakery products. Consequently, there have been market conditions. almost no flour imports, while wheat, pasta, and Several kinds of NTBs have also been adminis- bakery products have had large imports. The very tered by governments to control wheat imports-- low tariffs on wheat probably reflect the fact that in many cases these have been lessened or 202 Global Agricultural Trade and Developing Countries TABLE 11.4 Average Tariff Rates and Imports for Wheat and Wheat Products Tariffs (percent) Imports (million dollars) Country Year Wheat Flour Pasta Bakery Wheat Flour Pasta Bakery Australia 2001 0.0 0.0 4.0 7.0 0 0 -- -- Bangladesh 1999 5.0 8.3 37.5 37.5 361 19 0 1 Brazil 2001 6.3 13.5 18.5 20.5 872 36 11 21 Bulgaria 2001 23.1 25.0 39.1 54.4 3 1 -- -- China 2001 74.0 98.8 24.1 24.0 319 20 -- -- Costa Rica 2001 0.0 6.0 14.0 14.0 36 3 1 8 Egypt, Arab Rep. 1998 1.0 8.3 40.0 40.0 816 25 1 3 European Union 2001 65.7 45.1 20.8 22.7 844 9 -- -- Guatemala 1999 0.0 10.0 15.0 15.0 56 1 3 19 Hungary 2001 25.0 38.4 38.4 34.7 0 0 -- -- India 2000 41.7 38.5 38.5 38.5 1 1 -- -- Indonesia 1999 0.8 2.5 5.0 5.0 404 68 3 8 Jordan 2000 0.0 12.5 26.7 25.6 68 2 1 5 Kenya 2001 35.0 25.0 25.0 25.4 46 2 3 2 Korea, Rep. of 2001 2.7 151.4 8.0 8.0 530 1 34 34 Malawi 2000 0.0 16.7 25.0 22.9 3 4 0 1 Malaysia 2001 0.0 0.0 6.3 3.3 206 4 19 24 Mexico 2001 67.0 11.7 12.0 10.0 423 8 11 56 Morocco 1997 33.6 71.3 59.5 50.0 366 0 3 3 Pakistan 2001 5 25 70 27 0 -- -- Philippines 2001 4.8 8.0 15.0 15.0 427 6 14 24 Romania 1999 232.9 206.3 261.5 225.0 1 5 -- -- Russia 2001 5.0 10.0 102 19 -- -- South Africa 2001 0.0 20.0 25.0 27.6 32 0 -- -- Togo 2001 5.0 13.3 20.0 20.0 23 0 3 1 Uganda 2001 0.0 15.0 15.0 15.0 11 1 0 0 United States 2001 1.1 0.7 4.6 0.8 300 56 -- -- Zimbabwe 2001 5.0 30.0 40.0 40.0 5 0 0 0 SITC Code -- 410 460 483 484 410 460 483 484 -- Not available. Note: Unless otherwise noted, the duties applied to 100 percent of the tariff lines. The exceptions are as follows: Mexico 83 percent of pasta tariff lines covered; South Africa 50, 67, 75, and 88 percent of tariff lines covered for wheat, flour, pasta, and bakery, respectively; Turkey had only 25 percent of pasta and 14 percent of bakery tariff lines covered; the European Union had 25, 0, 9, and 0 percent of tariff lines covered for wheat, flour, pasta, and bakery, respectively; the United States had only 33 percent of wheat and flour tariff lines covered. Source: For Egypt's tariff data, TRAINS database. For Pakistan, all data from WTO tables. For all others: tariffs from WTO Integrated Database, MFN (most-favored-nation) Applied Duties; imports from FAOSTAT and COMTRADE. eliminated following liberalization of domestic of global wheat imports. Among large wheat markets and international trade. Government importers, examples of greater private sector import controls include the issuing of import involvement can be found in Indonesia, Pakistan, licenses, quantity and quality restrictions, state the Philippines, and Turkey in Asia; Algeria, the trading, and bureaucratic red tape in general.5 State Arab Republic of Egypt, and Morocco in Africa; trading is still practiced by many governments, but and Brazil and Mexico in Latin America. However, the private sector is responsible for a growing share many governments--among them those of China, Wheat: The Global Market, Policies, and Priorities 203 the Islamic Republic of Iran, India, and Japan--still 85 percent of global wheat food aid during control wheat imports through various schemes. 1990­2000 was provided by four of the world's major wheat exporters. The United States provided 54 percent of world wheat food aid, the European Export promotion Union 20 percent, Canada 8 percent, and Australia The two largest providers of wheat export subsi- 3 percent (FAOSTAT). In total, wheat food aid dies, the European Union and the United States, accounted for about 6 percent of wheat trade dur- had largely eliminated export subsidies by 2001 as ing 1990­2000. Over a longer period, the United global prices rose and the European Union reduced States provided an average of 3.3 million tons of intervention prices under its 1992 and Agenda 2000 wheat food aid from 1970 to 2000, averaging policy reforms. The European Union agreed to 10 percent of U.S. wheat exports. The level of food reduce subsidized exports to 14.4 million tons by aid varied with price (figure 11.5), which suggests 2000 and thereafter under the URAA and has not that it was partly surplus disposal.7 reported export subsidies since the 2001­02 sea- son.6 The United States agreed to reduce subsidized Food Security and Global Wheat Trade wheat exports to 14.5 million tons by 2000 and thereafter under the URAA, although the U.S. pri- Since wheat is the food grain most often used as a mary export subsidy facility, the Export Enhance- buffer against food shortages, any disruption of ment Program, has not been used for wheat since trade flows or sharp increases in prices causes food 1995. Under URAA, however, both countries could security concerns for importing countries. When revive their export subsidy programs and together real wheat export prices doubled from 1970 to 1974, could subsidize nearly one-quarter of global wheat for example, policymakers in wheat-importing exports. Export credits were still used by Australia, countries were quick to raise concerns. These were Canada, the European Union, and the United States further heightened when the United States imposed as recently as 1998, the most recent period for grain export embargoes in 1974 and 1975 to protect which complete data were available (OECD 2000). its own consumers from high prices.8 Such actions Food aid has been provided by many countries likely contributed to the strong desire for food self- to respond to emergencies or persistent food short- sufficiency in many food-importing countries. The ages. It has often been charged, however, that food United States again embargoed grain sales to the aid is partly used as a way to dispose of surplus pro- Soviet Union in 1980 as a foreign policy action duction. This charge is supported by the fact that motivated by the USSR's invasion of Afghanistan. FIGURE 11.5 U.S. Wheat Food Aid vs. Prices Food aid million tons Price US$ per ton 8 250 200 6 150 4 100 2 50 0 0 1970 1975 1980 1985 1990 1995 2000 Total aid U.S. fob price Source: USDA and World Bank. 204 Global Agricultural Trade and Developing Countries The embargo lasted nearly 16 months and included Domestic support a wide range of products. The European Union pro- Because domestic support commitments under the voked anxiety over food security again in 1995, after URAA apply to the whole of agriculture rather than it had become a wheat exporter, when it imposed an to individual commodities, countries have been export tax of $35 per ton on wheat during 1995­96 to protect its consumers from high prices.9 The able to protect their most politically sensitive sec- tors by keeping support high. According to OECD Russian Federation recently took similar action by estimates, domestic support to OECD wheat pro- imposing an export tax on wheat of 25 euros per ducers averaged $17.3 billion per year during ton during the period from January 15 to May 1, 1999­2001 (table 11.5), compared with $18.7 bil- 2004. Several other wheat exporters also imposed lion per year during the 1986­88 base period of the restrictions on wheat exports, including Hungary, URAA. Domestic support in 1999­2001 accounted India, and Ukraine. As with the previous actions, for 41 percent of the value of wheat production at the policies were intended to protect domestic sup- farm-gate prices. The European Union and the plies and control prices. All came at times when United States provided the largest absolute support world prices were rising and the availability of sup- to wheat, amounting to $9.6 billion and $4.9 billion plies was uncertain. Such actions send the signal per year, respectively. However, support represent- that access to wheat exports cannot be relied upon ing a higher percentage of the value of wheat during periods of shortages and high prices. TABLE 11.5 Support to OECD Wheat Producers, 1999­2001 Producer Support OECD Country/Region (millions of dollars) PSE Percentage Producer NPC Producer NAC OECD 17,331 41 1.16 1.70 Australia 119 5 1.01 1.05 Canada 358 15 1.01 1.17 Czech Republic -10 -3 0.87 0.97 European Union 9,565 48 1.15 1.95 Hungary 53 13 1.06 1.15 Japan 822 86 6.38 7.20 Mexico 237 39 1.44 1.64 Norway 67 71 2.63 3.58 Poland 251 21 1.22 1.27 Slovak Republic 5 5 0.86 1.06 Switzerland 191 63 2.33 2.82 Turkey 551 26 1.30 1.41 United States 4,928 46 1.12 1.86 Note: Producer support was converted from local currency to U.S. dollars using period average annual exchange rates from International Monetary Fund's International Financial Statistics, 2002 Yearbook. PSE percentage is producer support estimate, an indicator of the annual monetary value of gross transfers from consumers and taxpayers to support agricultural producers, measured at farm-gate level as a percentage of the value of production arising from policy measures, regardless of their nature, objectives, or impact on farm production or income. Producer NPC is producer nominal protection coefficient, an indicator of the nominal rate of protection for producers measuring the ratio between the average price received by producers (at farm gate), including payments per ton of current output, and the border price (measured at farm-gate level). Producer NAC is producer nominal assistance coefficient, an indicator of the nominal rate of assistance to producers measuring the ratio between the value of gross farm receipts including support and gross farm receipts valued at world market prices without support. No calculations were made for Canada, Iceland, New Zealand, Norway, or the Republic of Korea. Source: OECD 2002 Wheat: The Global Market, Policies, and Priorities 205 TABLE 11.6 Major Wheat Exporters' Shares of Global Wheat Net Exports (percent) Exporter 1970­79 1980­89 1990­99 United States 40.9 36.1 27.1 Canada 19.5 18.4 17.9 Australia 12.1 12.8 12.0 European Union 0.0 7.8 9.9 Argentina 3.9 5.6 6.2 Total 76.4 80.7 73.1 Note: The average export shares were calculated on net exports because the European Union was both a large importer and exporter and net exports capture the net trade situation better than gross exports. When net exports were negative, a zero was assigned. The European Union was defined as the current 15 members even though not all of these countries were members during the entire period. Source: FAOSTAT. production was provided by Japan (86 percent), TABLE 11.7 Producer Support Estimates, Norway (71 percent), and Switzerland (63 percent). 1986­88 and 2000­02 (percent) Major Wheat Exporters Exporter 1986­88 2000­02 Five countries accounted for about three-quarters of Australia 9 5 net global wheat exports in 1970­1999 (table 11.6). Canada 45 16 The United States was the largest net exporter over European Union 52 46 this period, but its share declined from nearly United States 49 40 41 percent during the 1970s to 27 percent during the 1990s. Canada was the second-largest net exporter Source: OECD 2003. with an 18­19 percent share. Australia maintained its 12 percent share throughout the period, and Argentina expanded its share from about 4 to 6 per- and 2000­02, with total producer support declin- cent of global exports. The European Union ex- ing from 45 percent of the value of production to ported 10 percent of global net exports during the 16 percent. The country largely abandoned direct 1990s after being a net importer of about 10 percent price support to individual commodities in favor of of global trade during the 1970s. The emergence of income support in the early 1990s (Gardiner 1999). the European Union as a major exporter was attrib- This led to reduced wheat production and reduced utable to highly subsidized production and exports net exports by 13 percent between 1990­95 and under the Common Agricultural Policy (CAP) and 1996­2001 (table 11.8). was largely at the expense of U.S. exports. The European Union sharply reduced wheat Wheat support policies of the major exporters intervention prices in the CAP reform of 1992 and have changed since the early 1980s, with Australia implemented a mandatory land set-aside policy and Canada significantly reducing support to their (Rayner and others 1999). Further reforms were wheat producers, while the European Union and the taken in 2000 and 2003. However, total support did United States have decreased support more moder- not decline significantly (see table 11.7), and pro- ately according to OECD estimates (table 11.7). duction continued to increase (see table 11.8). Argentina, which is not an OECD country and does Consumption increased because of the lower inter- not have producer support estimates comparable to vention prices, allowing net exports to fall by the other major exporters, has historically taxed 41 percent between 1990­95 and 1996­2001. rather than supported its wheat producers. The United States undertook major reforms in Canada made the largest reductions in wheat the 1980s, with the reduction in wheat loan rates support among major exporters between 1986­88 and the introduction of the Conservation Reserve 206 Global Agricultural Trade and Developing Countries TABLE 11.8 Percentage Change of Wheat Production, Area Harvested, Yields, and Net Exports of Major Exporters from 1990­95 to 1996­2001 Exporter Production Area Harvested Yields Net Exports Argentina 47.2 28.0 14.7 53.8 Australia 66.2 37.5 21.7 61.9 Canada -10.0 -13.4 4.0 -13.3 European Union 12.4 3.0 9.1 -40.8 United States -2.0 -9.3 8.6 -16.7 Major exporters 10.2 1.4 8.6 -6.3 Source: Authors' calculations. Program, which removed about 10 million acres of caused by major economic and policy changes in wheat land from production (15 percent of wheat several regions and countries. These include area) (Hoffman, Schwartz, and Chomo 1995). Net increased wheat imports by OPEC (Organization of exports declined by nearly 17 percent from Petroleum Exporting Countries), large net imports 1990­95 to 1996­2001, as area declined 9.3 percent. by the FSU and Eastern Europe because of poor pro- The 2002 U.S. Farm Bill somewhat reversed previ- duction, and policy changes in China that led to ous reforms but did continue the large wheat-land large net imports. Most of these changes have either diversion program begun in the 1980s. been reversed or had more moderate influences on The changes in Argentina led to large investments wheat imports since 1980. in the Argentine grain marketing system, more Imports by the FSU accounted for one-quarter intensive input use, and a 50 percent increase in net of global imports at their peak in 1984 and then exports from 1990­95 to 1996­2001 (Schneph, steadily declined to only 3 percent by 1995. Imports Dohlman, and Collins 2001). The financial crisis of by Eastern European countries, while much smaller 2002 contributed to the profitability of exportable than those of the FSU, declined by 75 percent agriculture, as the peso was devalued by 70 percent during the same period (FSU+Eastern European after being fixed to the U.S. dollar for 10 years. countries in figure 11.6). China's imports peaked at Export taxes of 20 percent were reinstated to offset 16 million tons in 1989 (16 percent of global trade) windfall profits from the currency devaluation. and declined to 2 million tons in 2000 due to rap- The combined impacts of reduced production idly increasing domestic production following pol- support, lower export subsidies, and land set-asides icy changes (see figure 11.6). are reflected in lower production and reduced net OPEC's import increases slowed during the exports by Canada, the European Union, and the 1980s as oil prices fell. Imports have only recently United States (see table 11.8). These declines were begun to increase with the recovery in oil prices mostly offset, however, by larger exports from that began in 1999. Thus the countries that fueled Argentina and Australia. On balance, the five major the large increase in wheat trade during the 1970s exporters reduced net wheat exports from 80 mil- largely accounted for its stagnation after 1980. Off- lion tons during 1990­95 to 75 million tons during setting these declines have been steady increases in 1996­2001, a decline of 6.3 percent. The largest imports by other developing countries (shown as decline in producer support came in Canada, where Developing minus China in figure 11.6), but the lower support led to lower area planted, produc- increases were not large enough to raise global tion, and net exports. trade significantly. Imports by the developed coun- tries have remained largely constant since 1980, with lower imports by Western Europe offsetting Major Wheat Importers increases from high-income Asia. Global wheat imports have grown by just 1.2 percent Brazil, China, Egypt, Japan, and the Russian per year since 1980, compared with nearly 6 percent Federation were the largest importers during per year between 1970 and 1980 (figure 11.6). 1990­2000, each with 5­7 percent of global imports The rapid increase in imports during the 1970s was (figure 11.7). They were followed by Algeria, Wheat: The Global Market, Policies, and Priorities 207 FIGURE 11.6 Global Wheat Imports Million tons 120 100 80 60 40 20 0 1970 1975 1980 1985 1990 1995 2000 World Developing China China OPEC FSU Eastern European Countries Source: FAO. FIGURE 11.7 Wheat Net Imports, Average for 1990­2000 Million tons 7 6 5 4 3 2 1 0 of Iran Egypt Brazil Japan China Russia Algeria Rep. IndonesiaMorocco Korea, Source: USDA. Indonesia, Iran, and Korea, each with 2­5 percent of consumption. It is also the major staple crop pro- global imports. The 10 largest net importers during duced in the country, occupying about one-third of 1990­2000 accounted for 46 percent of global the total winter crop area. The government's policy imports. Many of these large importers undertook objectives in the agricultural sector have been to policy reforms during the late 1980s or 1990s that provide an adequate supply of food to all income removed government monopolies on imports. groups, to promote greater self-sufficiency in crop Egypt was the largest importer during production, and to increase farm income. 1990­2000, with average imports of slightly more In the mid-1980s the widening food gap, stagna- than six million tons per year. Wheat is considered tion of the agricultural sector, and the rising costs a strategic commodity in Egypt, providing more of the food subsidy system encouraged the govern- than one-third of the daily caloric intake of ment to reform agriculture and the wheat sector Egyptian consumers and 45 percent of protein under the Agricultural Reform Program, initiated 208 Global Agricultural Trade and Developing Countries in 1987. In the first phase of the reform, prices, production include subsidized loan programs for quotas, and marketing controls were partially farmers and processors to borrow against their liberalized. Import subsidies were reduced, and products at below-market-interest rates while markets were opened to private investment. The holding their products as collateral in accredited second phase of the reform coincided with the warehouses. Small producers are eligible for financ- launching of the Economic Reform and Structural ing of production costs at subsidized interest rates Adjustment Program in 1991, which sought to shift under a program to strengthen family farms. Egypt from a state-controlled economy into a more Longer-term support for production and process- market-oriented economy in which the private ing of agricultural products is available from the sector could play a major role (Kherallah and Brazilian Bank for Economic and Social Devel- others 2000). opment and the Special Agency for Industrial In 1992 the Egyptian government also began to Financing. liberalize the wheat-milling sector, which up to that Japan was the third-largest wheat importer dur- time had maintained a monopoly over the impor- ing 1990­2000, but unlike other importers it has tation of all types of wheat grain and flour. In addi- not reduced import controls or significantly tion, around 80 percent of all industrial wheat mills reduced producer support. Japan's agricultural pol- in the country belonged to the public sector--the icy is strongly influenced by concerns for food rest were licensed to mill for the government under security and self-sufficiency. In addition, postwar specific arrangements. The partial liberalization of land reforms created a very small-scale farm struc- wheat trading started in the flour market in 1992, ture that is inefficient by global standards; thus when the government freed the prices of flour and income support for farmers is also a high priority. bread and allowed the private sector to import Wheat producers receive about $1,200 per ton for wheat for the production of flour. Resale of wheat wheat--about 10 times the U.S. f.o.b. (free on in excess of the milling needs for flour was not board) price and 6 times the c.i.f. (cost, insurance, permitted. The government also allowed private and freight) import price. Domestic wheat produc- traders to import flour directly. All the remaining tion is about 10 percent of domestic consumption, restrictions on flour production and trading were and the Japanese Food Agency imports about removed in 1993, allowing both the public and pri- 6 million tons of wheat per year. Import policy has vate sectors to freely import, produce, distribute, focused on food security and diversification of sup- and sell flour at free-market prices. The quotas of plies in an effort to ensure guaranteed supplies government-milled flour going to food-processing rather than low import prices. Domestic producers factories, shops, and bakeries were also eliminated, are paid an administered purchase price for wheat, thus allowing these outlets to purchase their flour which is then resold at higher prices to the domes- freely in the market. tic milling industry. Imported wheat is resold to Brazil was the second-largest importer during millers at prices that are about double the import 1990­2000, with imports averaging 6 million tons price. A margin between the resale of domestic and per year. Before 1991 the Brazilian government imported wheat is necessary to adjust for quality heavily subsidized wheat flour, but consumer subsi- differences between Japanese and imported wheat. dies were removed in late 1991 along with other A new wheat policy was introduced in 1998 by price controls, and the mill-quota system was elim- the Japanese Food Agency, with implementation inated (Brandão and Salazar 2003). Brazil now occurring during the 2000 to 2002 crop years. The obtains nearly all of its wheat from Argentina with Japanese Food Agency retained control over the a zero import duty because of its membership in pricing and marketing of domestic wheat, as well as the regional trade group, Mercosur. In 2002, to the importing and pricing of foreign wheat. The reduce wheat imports, the government introduced new policy allows the private sector to import a program to expand domestic wheat production to wheat, whereas the Japanese Food Agency had been 50 percent of total consumption by 2004. The the exclusive importer under the previous system. government operates a minimum-support-price The new policy also introduced a new compensa- system for wheat and other commodities. Other tion system for domestic wheat producers. Other policies and programs to support domestic wheat programs to improve quality allow continuous Wheat: The Global Market, Policies, and Priorities 209 importation of wheat for food use by the Japanese imported wheat uncompetitive in some years. Dur- Food Agency and a simultaneous-buy-and-sell sys- ing the phase-in period after China's entry into the tem for imported feed wheat. WTO in 2001, the volume of imports was regulated China was the fourth-largest importer during by a tariff rate quota system. The initial TRQ the 1990­2000 period, but it has undertaken major for wheat was 7.3 million tons in 2001; it rose to reforms in the past few years, lowering support 9.64 million tons in 2004. China also agreed to prices to near world market levels and reducing expand the role of private traders after WTO acces- imports by 90 percent (Crook 1996 and 1997; sion, but state trading enterprises would still con- USDA 1998 and 2001a). The government's long- trol 90 percent of the wheat TRQ. Notwithstanding standing policy has been to approximate self- China's WTO commitments, however, the fill rate sufficiency in food staples, including wheat. This of the wheat TRQ has been minimal--8 percent in began in the 1950s with producer quotas, but the 2002 and 5 percent in 2003, according to periodic system changed significantly with the introduction reports from the Global Agricultural Information of the Household Responsibility System in 1978. Network of the U.S. Department of Agriculture. Under this system, local leaders began contracting While many of the major importing countries production quotas with small work units and fam- have reformed policies, high protection is still evi- ily farms instead of large collectives. China initiated dent in many countries, as shown by disparities the Governors' Grain-Bag Responsibility System in between producer prices and U.S. fob prices (fig- 1995, whereby provincial authorities were given the ure 11.8). Japan, has the highest producer prices, task of stimulating production, stabilizing prices, but several other countries also have high prices. making provisions for adequate grain stocks, Consumers in these countries have the most to gain reducing imports, and ensuring supplies for urban from more liberal trade policies. areas and the military. These reforms led to an increase in wheat production of about 22 percent Emerging Wheat Exporters between 1990­92 and 1997­99 and to huge stocks by the end of the 1990s. Since the mid-1990s, the China, India, Pakistan, and several countries of the state grain procurement program has not been FSU have emerged as wheat exporters. This is a altered to any significant degree, although there shift from the past when these countries were either have been revisions in some procurement proce- large regular importers or occasional importers. dures and efforts to improve wheat quality. In 2000 the government introduced new wheat standards to Former Soviet Union upgrade the average quality of the crop. Protected prices were removed from spring wheat in the Several countries of the FSU have the potential to north and winter wheat south of the Yangtze River. become large exporters--among them Kazakhstan, In 2001 market reforms eliminated protected prices the Russian Federation, and Ukraine, which collec- in many provinces but not in the major producing tively moved from being net importers of 15 mil- regions. lion tons of wheat in 1992 to net exporters of One of the most significant consequences of 23 million tons in 2002 (figure 11.9). China's domestic policies has been the shift in its The emerging exporters of the FSU have many cereal trade balances. The accumulation of large common problems, including weak marketing sys- grain stocks caused average annual wheat imports tems; inefficient storage, transport, and grain han- to fall from 10 million tons in the early 1990s to dling systems; lack of credit; and the challenge of below 1 million tons since 2000. China became a making the transition from collective farms to pri- net exporter of wheat in 2002. vate production systems. Policy reforms have been Other factors contributed to the decline in slow and only partially effective in stimulating wheat imports, including local market conditions private-sector initiative. Despite these problems, all and government actions. China assessed a 13 per- of these countries have large land areas well suited cent value-added tax on imported wheat (while not to wheat production and low production costs. collecting the tax on most domestic wheat produc- They also have an advantage in transporting wheat tion), and a 1 percent import duty, thus making to importers in the Middle East compared with 210 Global Agricultural Trade and Developing Countries FIGURE 11.8 Wheat Producer Prices in 2001 for Selected Countries U.S. dollars per ton 1200 1000 800 600 400 U.S. fob price 200 0 of Japan Syria Iran Chile India Nigeria Rep. Sudan Egypt Mexico ChinaTurkey Brazil MoroccoColombia Pakistan Switzerland Korea, Source: FAO. FIGURE 11.9 Emerging Wheat Net Exports of Emerging Exporters in the FSU Million tons 30 20 10 0 10 20 1990 1992 1994 1996 1998 2000 2002 Source: FAOSTAT. major exporters such as Australia, Canada, and the sector activity. Production is mostly high-protein United States. spring wheat, which, with quality improvements, Kazakhstan, a major wheat producer and could compete with the best wheat from other exporter during the Soviet era, used intensive farm- exporters. Large investments during the Soviet era ing methods that relied on subsidized inputs. These left the country with considerable infrastructure intensive farming methods, which are not prof- for grain transport and exports. If costs can be con- itable without large subsidies, have been aban- trolled and production increased, Kazakhstan doned. The country has a large land area, however, could become a major wheat exporter (Longmire with considerable potential for expanding wheat and Moldashev 1999). production and exports using low-input farming Russia accounted for one-quarter of world methods. Production remains constrained by high wheat exports in the early years of the 20th century, domestic marketing and transport costs traceable when its yields were only slightly less than those of to a lack of competition and insufficient private the United States. Whether the Russian Federation Wheat: The Global Market, Policies, and Priorities 211 can return to its former role as a major exporter implemented, but specific agricultural and institu- depends largely on policy developments. Input sub- tional reforms were only partially implemented. sidies and price supports were largely dismantled in Land reform has been slow to develop in the years January 1992, when the transition toward a market since transfer and ownership legislation was passed economy began. Decreases in real incomes and in 1994 and 1995. The result has been a slow recov- changes in food prices led to substantial changes in ery of production (Debatisse and Chabot 2000). food consumption and declines in feed use. Gov- ernment procurement of wheat declined from Asia nearly all before the reform to just 21 percent by 1995--in part because of lack of funds by the state Several Asian countries have become net wheat procurement agency. Private grain-trading compa- exporters in recent years because of large crops and nies have largely replaced the state. Simultaneously, stocks. It remains to be seen, however, whether they subsidies have been reversed and the agricultural can sustain exports or will revert back to being net sector (including wheat) are taxed. Grains are still importers. Domestic policies in Asia remain aimed produced on large farms and under the same man- at self-sufficiency and self-reliance rather than on agement as before the reforms, but many of the promoting surpluses for export. farms have been converted into private stock com- India has followed a policy of self-sufficiency panies. Regional authorities use wholesale and since independence, increasing crop output by retail price controls, subsidies, and barriers to inter- expanding irrigation,improving crop yields through regional trade to regulate prices and food stocks. high-yielding varieties, and increasing land-use The emerging private sector must deal with the intensity with multiple cropping. Better yields were high transaction costs of these interregional trade possible because of the Green Revolution, which restrictions. Producers have considerable potential provided the high-yielding varieties that were to increase yields if economic incentives improve. adopted with support from production and price Efficient transportation and marketing systems policies. To increase yields, the public sector pro- could make the Russian Federation a net exporter vided agricultural inputs, such as fertilizers, power, (Goodwin and Grennes 1999). and water for irrigation at subsidized prices. The Ukraine emerged from a decade of adjustment government also established a system of minimum following the end of the Soviet era to become a sig- support prices to procure wheat from farmers. Sub- nificant wheat exporter. In 2002­03 the country sequently, India made substantial gains in food grain exported about 6.6 million tons of wheat before a production, and over the past 30 years wheat pro- production slump in 2003­04 forced a return to duction has grown by about 3.5 percent per year. imports. However, low production costs and shifts In the early 1980s, India cautiously began to lib- in resource use since the Soviet era suggest that eralize trade, but only since 1991 has the process of wheat exports will likely increase in the future. liberalization picked up speed. In July 1991 India During the Soviet era, Ukraine concentrated on introduced radical policy reforms in various eco- livestock and poultry production, but soon after nomic sectors, but trade restrictions on agricultural independence in 1991 poultry and livestock num- products were left largely untouched. Subsequent bers declined by more than half. Wheat production changes in trade policy gradually lifted restrictions fell as well (in part because of lower demand for on agricultural products. Bumper wheat crops animal feed) until 2000, rebounding in the two starting in 1999 led to large accumulations of pub- years before the severe drought of 2003. Production lic stocks, which eventually prompted wheat costs are estimated by the U.S. Department of Agri- exports, making India a net wheat exporter in 2000. culture (USDA) to be $50 per ton (Thursland and In order to be competitive in the Asian markets, the Prikhodko 2002), offsetting inefficiencies in han- government subsidized exports, displacing sales in dling, storage, and transport. the region by the United States and other tradi- Policy reforms began with price liberalization in tional suppliers. Exports reached a record 6 million 1992, but many agricultural subsidies were contin- tons in 2002­03. Wheat exports were halted in ued, resulting in budget deficits and inflation. August 2003 as stocks diminished following a poor Economywide price and trade reforms were fully crop in 2003. 212 Global Agricultural Trade and Developing Countries China has been exporting wheat since 1992. The FAPRI results show that wheat prices (U.S. Exports reached 1 million tons in 2002­03, making fob Gulf) would rise from the baseline by an aver- China a significant net wheat exporter for the first age of 4.8 percent under the full liberalization time. Despite stagnant domestic demand, however, scenario and by 7.6 percent in the trade-only sce- domestic surpluses have been falling in recent years nario. The price increase is lower under full liberal- in line with policy-driven production declines. ization because set-asides would be removed in the Rozelle and Huang (1999) argue that China will European Union and the United States, resulting in remain a net importer at the levels of the early to a substantial increase in production and exports mid-1990s. that would dampen the price effect. Global wheat Pakistan has been a net wheat exporter since trade would increase by 7.9 percent under full lib- 2000­01, reaching 1 million tons in 2002­03, but eralization and by 5.0 percent under the trade-only reverted to being a net importer during the scenario, with the largest export increase going to 2003­04 season. the European Union once set-asides were removed. China would reduce imports under both scenarios because it would face slightly higher prices on The Impact of Liberalization wheat allowed under the low in-quota tariff in its Various studies have estimated the impact of liber- wheat tariff rate quota. India would reduce exports alizing trade and reducing domestic support on the and become a net importer, because export subsi- world wheat market. Results vary and do not always dies would no longer be allowed. consider the full range of reforms. The Food and A USDA study (2001b) found a larger wheat Agricultural Policy Research Institute (FAPRI price increase from elimination of all policy distor- 2002) at Iowa State University and the University of tions. It concluded that wheat prices would rise by Missouri recently evaluated the impacts of liberal- 18.1 percent from elimination of all policy distor- ization of agricultural markets using a multimarket tions. Removal of global tariffs would raise prices global agricultural model. The results for wheat are 3.4 percent; elimination of OECD domestic subsi- reported here. The study considered two scenarios. dies would raise them 12.0 percent; and global The first, full liberalization, explored the probable elimination of export subsidies would raise them effects of removing all agricultural distortions-- 2.0 percent. A recent FAO study (Poonyth and domestic farm programs and border measures-- Sharma 2003) concluded that wheat prices would including all TRQ schemes, tariffs, and direct rise by 11.9 percent under the U.S.-proposed WTO export subsidies such as the European Union's reform, which is similar to the USDA's full- CAP. The second investigated the effects of remov- liberalization scenario. ing only border measures. The two scenarios allow The three studies provide a range of estimates-- the impact of domestic programs to be evaluated from 4.8 to 18.1 percent--of the increase in world separately from border measures. The question of prices that would result from eliminating all pro- how to examine domestic programs without bor- ducer support and trade distortions. der measures was addressed by assuming that gov- ernment payments would be used to provide pro- Conclusions ducers with the difference between current domestic price floors and the lower prices that The global wheat market has become less distorted would result without border measures. The simula- since the early 1990s, as several countries have tions did not include the 2002 U.S. Farm Bill but undertaken reforms unilaterally or as a conse- instead used an extension of the previous farm bill quence of commitments under the URAA. Govern- in the baseline simulation. Nor did they include the ments in OECD countries still provide substantial reforms to the European Union's CAP in mid-2003. support to producers, however. The effects of that The URAA was assumed to extend after 2004, when support have been partially offset by land set-aside the final provisions are to be implemented. The programs and by the way in which support is pro- results are presented as average percentage changes vided; however, support policies still distort trade relative to the baseline simulation for the period and depress world prices. The European Union and from 2002 to 2011. the United States have not used export subsidies in Wheat: The Global Market, Policies, and Priorities 213 recent years but still use other surplus disposal pro- 2. Note that the yields in India and the United States are not grams, such as nonemergency food aid and export strictly comparable because a larger share of wheat area in India is irrigated than in the United States. credits. These programs make it more difficult for 3. About 9 percent of global grain-based starch production emerging exporters to compete with established comes from wheat (80 percent comes from maize). "Starch-- exporters. Versatile and in Demand," World Grain, January 2004. 4. Karnal bunt is a wheat fungus that occurs during cool, Most importing countries have reduced wheat rainy growing conditions. It is named after the city in India near tariffs or allowed duty-free imports from regional where it was first reported in the 1930s. It is not harmful to trading partners, thus benefiting from lower prices. humans or animals, but it causes an unpleasant odor in wheat flour. A few countries, such as Japan, continue to apply 5. Exchange rate controls are another means to regulate extreme protection, with internal prices more than imports, but these are not normally commodity-specific. five times global market levels. Tariff escalation is a 6. Based on the use of wheat subsidies as reported to the major problem for countries trying to diversify WTO. Source: USDA, ERS WTO Agricultural Trade Policy Com- mitments Database. production and exports, with tariffs on flour well 7. Regressing wheat food aid (FA) on wheat prices (WP) above those on wheat grain, and tariffs on bakery shows a statistically significant relationship between the quan- and pasta products even higher. Consequently, tity of food aid and prices; as prices fall the quantity of food aid increases. The OLS regression estimated was FA = 6.487 - trade in wheat products is largely confined to free .023*WP. The R2 = .42 and the coefficient on prices was statisti- trade areas such as within the European Union or cally significant at the 1 percent level of significance with t = NAFTA. -4.6. 8. The United States suspended grain exports in 1974 and A major concern for wheat-importing countries again in 1975 because of low stocks, poor crop production is the lack of assured access to wheat export mar- prospects, and concern about the inflationary impacts of high kets during periods of supply shortages and high grain prices. However, the action was directed only at the USSR prices. Policies such as the U.S. grain export in 1974 and at the USSR and Poland in 1975 since these coun- tries were major buyers and perceived to be disrupting the mar- embargo of the 1970s, designed to protect domestic kets (USDA 1986). consumers, contribute to higher global wheat 9. "EU Hits Grain Exporters, Steep New Tax Aims to Protect prices and increase the uncertainty of wheat- Supplies," International Herald Tribune, December 8, 1995. Most recently, however, the European Union declined to impose importing countries. The threat of such policy wheat export taxes following the 2003 production shortfall, but actions continues, with the European Union did temporarily suspend grain export licenses in August 2003. imposing export taxes on wheat in 1995 and the Reuters, July 31 and September 2, 2003. Russian Federation imposing export taxes in 2004--also to protect domestic consumers. Such References actions reinforce the calls for food security through ABARE. 2002. "Global Grains Policy and WTO Agricultural self-sufficiency in importing countries and deprive Negotiations." Current Issues 02.2 (January). those countries of the benefits from trade. Beattie, B. R., and D. R. Biggerstaff. 1999."Karnal Bunt: A Wimp Future reforms of the global wheat market of a Disease But an Irresistible Political Opportunity." should focus on reducing producer support in Choices (2nd Quarter). Brandão, A., and P. Salazar. 2003. "Brazil." In FAO, WTO Agree- OECD countries, reducing protection in the ment on Agriculture, The Implementation Experience. Devel- remaining highly protected markets, reducing tariff oping Country Case Studies, Rome. escalation on wheat products, and ensuring access Carter, C. A., and W.W. Wilson. 1999. "Emerging Differences in State Grain Trading: Australia and Canada." In J. M. Antle to exportable supplies during price spikes. Elimina- and V. H. Smith, eds., The Economics of World Wheat tion of production subsidies and trade distortions Markets. New York: CABI. could raise world wheat prices by 5­18 percent Crook, F. 1996. "China's Agricultural Policy Developments." China Situation and Outlook Series WRS-96-2, U.S. Depart- according to recent studies, but the large surplus ment of Agriculture, Economic Research Service, Washing- capacity among major wheat exporters could boost ton, D.C. June. production under policy reforms and prevent ------. 1997. "Current Agricultural Policies Highlight Con- prices from rising further. cerns About Food Security." China Situation and Outlook Series WRS-97-3, U.S. Department of Agriculture, Eco- nomic Research Service, Washington, D.C. June. Dalrymple, D. G. 1974. "Development and Spread of High- Notes Yielding Varieties of Wheat and Rice in the Less Developed Nations." Economic Research Service, U.S. Department of 1. More detailed trade flow tables are given in the attached Agriculture, in Cooperation with the U.S. Agency for Inter- CD-ROM. national Development, Washington, D.C. 214 Global Agricultural Trade and Developing Countries Debatisse, Michel, and Philippe Chabot. 2000. "A Review of the Poonyth, Daneswar, and Ramesh Sharma. 2003. "The Impact of Grain Marketing Sector in Kazakhstan and Ukraine." Envi- the WTO Negotiating Modalities in the Areas of Domestic ronmentally and Socially Sustainable Development Working Support, Market Access, and Export Competition on Devel- Paper 25, World Bank, Washington, D.C. June. oping Countries: Results from ATPSM." Paper prepared for Diaz-Bonilla E. 1999. "South American Wheat Markets and the International Conference Agricultural Policy Reform MERCOSUR" in J. M. Antle and V. H. Smith, eds., The and the WTO, Capri, Italy, June 23­26. Economics of World Wheat Markets. New York: CABI. Rayner, Anthony, Robert Hine, Timothy Lloyd, Vincent H. FAOSTAT. Nutritional data: Food aid shipments. On line: Smith, and Robert Ackrill. 1999. "The European Union http://faostat.fao.org/faostat/. Common Agricultural Policy under the GATT." In J. M. FAPRI (Food and Agricultural Policy Research Institute). 2002. Antle and V. H. Smith, eds., The Economics of World Wheat "The Doha Round of the World Trade Organization: Markets. New York: CABI. Appraising Further Liberalization of Agricultural Markets." Rozelle, Scott D., and Jikun Huang. 1999."Wheat in China: Sup- Iowa State University CARD Working Paper 02-WP 317. ply Demand and Trade in the 21st Century." In J. M. Antle November. Ames, Iowa. and V. H. Smith, eds., The Economics of World Wheat Mar- Gardiner, Bruce L. 1999. "Canadian/U.S. Farm Policies and the kets. New York: CABI. Creation of a Single North American Grain Market." In J. M. Schneph, Randall, Erik Dohlman, and Chris Collins. 2001."Agri- Antle and V. H. Smith, eds., The Economics of World Wheat culture in Brazil and Argentina: Developments and Markets. New York: CABI. Prospects for Major Field Crops." Agriculture and Trade Goodwin, Barry K., and Thomas J. Grennes. 1999. "Russian Report WRS-01-3. U.S. Department of Agriculture, Eco- Agriculture and World Grain Trade: Lessons from the Past nomic Research Service, Washington, D.C. November. and Implications for the Future." In J. M. Antle and V. H. Thursland, Marget E., and Dmitri Prikhodko. 2002. "Ukraine Smith, eds., The Economics of World Wheat Markets. New Grain and Feed: How Is Ukraine Grain Competitive?" U.S. York: CABI. Department of Agriculture, Global Agricultural Information Hoffman, Linwood A., Sara Schwartz, and Grace V. Chomo. Network, Washington, D.C. August 2. 1995. "Wheat Background for 1995 Farm Legislation." AER- USDA (U.S. Department of Agriculture). 1986. "Embargoes, 712. U.S. Department of Agriculture, Economic Research Surplus Disposal, and U.S. Agriculture." AER 564. Economic Service, Washington, D.C. April. Research Service, Washington, D.C. December. Kherallah, Mylene, Hans Lofgren, Peter Gruhn, and Meyra M. ------. 1998. "China: International Agriculture and Trade Reeder. 2000."Wheat Policy Reform in Egypt, Adjustment of Report." WRS-98-3. Economic Research Service, Washing- Local Markets and Options for Future Reforms." IFPRI ton, D.C. August. Research Report 115. International Food Policy Research ------. 2001a. "China: Agricultural in Transition." WRS-01-2. Institute, Washington D.C. Economic Research Department, Washington, D.C. Longmire, Jim, and Altynbeck Moldashev. 1999. "Changing November. Competitiveness of the Wheat Sector of Kazakhstan and ------. 2001b."The Road Ahead: Agricultural Policy Reform in Sources of Future Productivity Growth." Economics Work- the WTO--Summary Report."AER 797. Economic Research ing Paper 99-06. International Maize and Wheat Improve- Department, Washington, D.C. January. ment Center (CIMMYT), Mexico. ----. 2003. Wheat Situation and Outlook Yearbook. WHS-2003. OECD (Organisation for Economic Co-operation and Develop- Economic Research Service, Washington, D.C. March. ment). 2000. "An Analysis of Officially Supported Export ------. 2004. "Mexico Grain and Feed Annual Report." Credits in Agriculture." Joint Working Party of the Commit- MX4033. Global Agricultural Information Network Foreign tee for Agriculture and the Trade Committee, Directorate for Agriculture Service, Washington, D.C. March. Food, Agriculture and Fisheries, Paris. December 20. Wilson, William W., and Bruce L. Dahl. 1999. "Grain Quality ------. 2002. Agricultural Policies in OECD Countries, Monitor- and North American Hard Wheat Exports." In J. M. Antle ing and Evaluation. Paris. June. and V. H. Smith, eds., The Economics of World Wheat ------. 2003. Agricultural Policies in OECD Countries, Monitor- Markets. New York: CABI. ing and Evaluation. Paris. 12 GROUNDNUT POLICIES, GLOBAL TRADE DYNAMICS, AND THE IMPACT OF TRADE LIBERALIZATION Ndiame Diop, John C. Beghin, and Mirvat Sewadeh Since the mid-1990s all major groundnut-export- Sub-Saharan Africa heavy producer taxation has ing countries have been gradually liberalizing their ended, and unilateral liberalization efforts are con- groundnut sectors, in part to fulfill their commit- tinuing, although significant protection of process- ments under World Trade Organization (WTO) ing remains. agreements.1 The results have been mixed, and The current situation raises many questions trade in groundnut products remains heavily dis- about the future of the sector and the prospects for torted. Both China and India have removed some various players. How will multilateral groundnut import restrictions and allowed wider private- trade liberalization affect the competitive positions sector participation in importing groundnuts. of different players? Which countries are likely to However, tariffs on groundnut products remain gain and which are likely to lose? How will small very high in India and high in China. The large Sub-Saharan African producers be affected? market size of both countries exacerbates these dis- tortions and their effects on the world market. Groundnut Production In the United States the 2002 Farm Bill elimi- nated many unsustainable features of previous Groundnuts are a valuable source of protein, fat, groundnut policies (such as the high support price energy, and minerals, and they generate cash and production quotas), but it introduced new dis- income to many poor farmers in the developing tortions, such as countercyclical payments and the world, especially in Sub-Saharan Africa and Asia. In floor price mechanism. These policies subsidize Senegal, for instance, 70 percent of the rural labor U.S. groundnut exports when world prices are low, force is employed in groundnut production, which with the potential to depress world market prices. accounts for 60 percent of households' agricultural Argentina still selectively subsidizes some income. Groundnut production and processing processed groundnut products and exports and represent about 2 percent of gross domestic prod- applies moderate export taxes on groundnuts. In uct and 9 percent of exports in Senegal. This chapter is an abbreviated version of Diop, Beghin, and Sewadeh (2004), which is reproduced on the CD-ROM. 215 216 Global Agricultural Trade and Developing Countries TABLE 12.1 Production, Use, and Export of Groundnuts, Average 1996­2001 Area Domestic Edible Crushed for Net Harvested Yield Production Use Groundnuts Oil and Cake Exports Country (1,000 ha) (mt/ha) (1,000 mt) (1,000 mt) (1,000 mt) (1,000 mt) (1,000 mt) World 21,452 1.4 29,997 29,896 12,416 14,590 169 Main producers and exporters China 4,234 2.9 12,204 11,777 4,753 6,140 427 India 7,902 0.9 7,176 7,082 534 5,581 94 United States 569 3.0 1,701 1,428 978 280 220 Argentina 280 1.5 403 191 21 155 213 Main producers in Africa Nigeria 1,187 1.1 1,340 1,340 636 427 0 Senegal 690 1.1 722 730 317 304 -6 South Africa 98 1.7 161 123 72 32 33 Malawi 117 0.9 103 101 78 18 2 Gambia, The 89 1.0 95 80 26 54 15 Main importers European Union 1 1.0 1 454 433 17 -449 Canada 0 0.0 0 115 115 0 -115 Japan 12 2.3 28 129 121 2 -103 Korea, Rep. of 7 2.2 15 30 30 0 -15 Note: The difference between production plus net exports and domestic utilization reflects stock variation and feed and seed use. Ending stocks are negligible for all countries except the United States, which had ending stocks of 28 percent of total production during 1996­2001. Source: U.S. Department of Agriculture. China is the world's largest groundnut producer, Driven by tremendous growth in China, global with 40 percent of world production in 2001. India shelled production grew 34 percent between accounts for 23 percent of worldwide production, a 1981­85 and 1996­2000. Growth has been uneven group of Sub-Saharan African countries produces across countries (figure A12.1 on the CD-Rom). 8.4 percent, and the United States produces 5.6 per- China doubled its production between 1992 and cent (table 12.1). Malawi, Nigeria, South Africa, and 2000 by increasing its use of high-yielding varieties the United States produce shelled groundnuts, and agricultural inputs, including fertilizers, pesti- which are used as seed, consumed raw, or consumed cides, mechanization, and irrigation (Colby and after having been transformed into prepared others 1992). In India production exhibited signifi- (roasted, salted, flavored) groundnuts or groundnut cant fluctuations,increasing between 1987 and 1998 butter or paste. In contrast, Argentina, China, India, before returning to the production levels of the and Senegal devote more than 60 percent of their 1970s (about 6 million tons). Production in Sub- production to crushing groundnuts for oil and Saharan Africa picked up in the early 1990s, after a meal.2 long period of decline. Production has been stable Groundnut production conditions vary consid- since the early 1970s in the United States,which pro- erably across countries, reflecting differences in duces about 2 million tons a year, and in Argentina, technological development, access to modern which produces 300,000 tons. inputs and irrigation, and farm management prac- The economic costs of groundnut production tices. Yields are highest in the United States and vary significantly across countries. In 1993 the China and lowest in Sub-Saharan Africa (except average cost per acre was $694 in the United States, South Africa) and India. The low yields in Africa more than three times the average cost in China of and India are the result of limited use of modern $164 per acre (table 12.2). The higher economic inputs, including high-yielding seed varieties, and costs per acre for U.S. groundnuts were attributed heavy dependence on rainfall. chiefly to production quota rent, land value, and Groundnut Policies, Global Trade Dynamics, and the Impact of Trade Liberalization 217 TABLE 12.2 Costs of and Revenues from Groundnuts in China and the United States (US$ per acre) 1992 1993 Item United States China United States China Variable costs Seed 70.32 43.83 71.18 45.96 Fertilizer 43.27 25.03 42.40 26.13 Chemicals 89.70 3.40 92.57 3.68 Labor 89.14 71.51 86.17 75.86 Other expenses 212.84 41.43 188.54 12.82 Subtotal 505.27 185.20 480.86 164.45 Fixed costs Land value 92.58 -- 97.77 -- Quota rent 113.38 -- 115.40 -- Total costs 711.23 185.20 694.03 164.45 Yield (pounds per acre) 2576 2520 1940 2135 Revenue (producer price times yield) 753.66 323.69 570.58 280.83 -- Not available. Note: More recent data are not available. Source: Chen and others 1997. the costs of using and maintaining farm equip- groundnuts are the most traded, with a volume of ment, fuel, electricity, repair, and capital replace- 1.2 million tons in 2001; trade in groundnut oil was ment. Quota rent and land value are not costs for 250,000 tons (table 12.3). The global export of edi- farmers in China, since there is no production ble groundnuts has increased 2.2 percent a year quota and land is considered public property, since the early 1980s, while exports of groundnut belonging to local communities organized in oil declined 1 percent and meal exports fell 2.5 per- groups of 30­40 households (Chen and others cent, despite growing global consumption of both 1997).3 Net returns for China and the United States products. are not significantly different if quota rent (irrele- China is the world's largest exporter of edible vant since the 2002 Farm Bill) is excluded from U.S. groundnuts, accounting for 32 percent of world production costs. The U.S. cost disadvantage is, exports. The United States is the second-largest however, compensated for by higher producer exporter, with 19 percent of the world market, fol- prices brought about by the groundnut program lowed by Argentina, at 10.5 percent. Sub-Saharan and the higher quality of U.S. groundnuts. The Africa (The Gambia, Malawi, Nigeria, Senegal, elimination of quota production (and thus quota South Africa, and Sudan) has lost market share rent) in the 2002 Farm Bill reduces U.S. production in international edible groundnut markets and costs. This development, as well as the high quality accounts for only 5 percent of world trade. Senegal of U.S. groundnuts, which earn a high price pre- is the world's largest supplier of groundnut oil, but mium in international markets, may well maintain this market has declined as other vegetable oils are U.S. competitiveness with China.4 increasingly used as substitutes. The European Union (EU) is the single largest groundnut market, accounting for 43 percent of Global Trade in Groundnuts world imports. The total value of net groundnut Domestic consumption of groundnuts is high, and imports in the European Union was just below only 5 percent of world production is sold in inter- $500 million a year in 1996­2000. Canada, with national markets. Of the three major groundnut 9 percent of world imports, is the second-largest products traded internationally (edible ground- market, followed by Japan, which imports 8.2 per- nuts, groundnut oil, and groundnut meal), edible cent of world groundnuts. 218 Global Agricultural Trade and Developing Countries TABLE 12.3 Value of Net Exports, by Groundnut Product, 1996­2000 (US$ millions) Country Edible Groundnut Groundnut Oil Prepared Groundnut Total EU-15 -378.47 -115.12 -4.54 -498.13 Japan -44.00 -1.85 -71.46 -117.31 Canada -76.67 -1.19 -3.31 -81.18 Korea, Rep. of -4.55 0.01 -14.31 -18.86 Malawi 0.77 0.00 0.00 0.77 Nigeria -3.29 4.64 0.00 1.35 Gambia, The 4.49 1.09 0.05 5.63 South Africa 16.01 4.68 0.27 20.95 Senegal 3.34 48.99 0.60 52.92 India 86.85 -0.13 7.27 93.99 United States 126.43 -12.77 28.26 141.92 Argentina 160.98 51.52 25.82 238.32 China 193.79 2.82 111.06 307.68 Note: Prepared groundnuts are roasted, salted, or flavored groundnuts. Peanut butter is not included. Source: FAOSTAT. Consistent with growth in world consumption, Figures 12.1e and 12.1f show the trends in exports of raw edible groundnuts and prepared exports and market shares of groundnut oil. Many groundnuts have expanded rapidly since the mid- countries, including Brazil and China, have exited 1980s (figures 12.1a­12.1d). Exports of edible the groundnut oil market since 1976 to focus on groundnuts increased 8 percent in the 1990s, after a edible groundnuts and other vegetable oils. Other dramatic increase of more than 20 percent during countries, such as the United States, have chosen the 1980s. The pattern of growth in prepared to enter the market only when the quality of groundnut exports broadly mirrors that of edible groundnuts harvested is too low for the nuts to be groundnuts, signaling the highly integrated nature sold in the edible groundnut market. Senegal and of these markets (figure 12.1c). Argentina remain the world's leading exporters China has been the major beneficiary of this of groundnut oil. The market has become sig- expansion (figure 12.1b). From barely 1 percent in nificantly fragmented, however. Argentina, Brazil, 1976, its global market share in exports of edible Senegal, and the United States jointly supplied just groundnuts rose to 32 percent in 2001. During the 52 percent of total exports in 2001, down from same period, the U.S. market share dropped from 85 percent in 1976. 32 percent to 19 percent. The emergence of China The decline in African countries' shares in global as a leading exporter in the prepared groundnut groundnut markets has significantly reduced the market is even more impressive (figure 12.1d). contribution of groundnut products to the export While the international edible groundnut mar- earnings of many Sub-Saharan African countries. ket has become more concentrated (with 61 per- The importance of groundnut products as a source cent of exports controlled by China, the United of export earnings has declined dramatically in States, and Argentina in 2001), the market for pre- Malawi, Senegal, and South Africa since the early pared groundnuts has become more fragmented. 1980s (table 12.4). The importance of ground- Concentration in the edible groundnut market nut products increased significantly only in The partially reflects the significant decrease in Sub- Gambia, where they accounted for 84 percent of Saharan Africa's share of prepared groundnuts, total merchandise exports in 2000­02. from 17 percent in 1976 to 5 percent in 2001. The volume of raw edible groundnuts exported Africa's market countries (including The Gambia, decreased significantly in Malawi, Nigeria, and Malawi, and Nigeria) enter the edible groundnut South Africa and stagnated in The Gambia and export market intermittently, depending on their Senegal (table 12.4). As a result of declining crop quality and world market demand. and almost stagnant volumes, export earnings for Groundnut Policies, Global Trade Dynamics, and the Impact of Trade Liberalization 219 FIGURE 12.1 Global Groundnut Consumption, Exports, and Market Shares (a) Global Exports and Consumption of Raw Edible (b) Market Shares of Global Raw Edible Groundnuts, 1976­2002 Groundnut Exports, 1976­2002 (1,000 metric tons) (percent) World consumption World export Percent 1,800 90 80 1,600 36,000 Consumption 70 1,400 60 32,000 1,200 50 28,000 40 1,000 Export 24,000 30 800 20 20,000 10 16,000 0 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 World edible export World groundnut consumption Argentina China United States Sub-Saharan Africa (c) Global Exports of Prepared Groundnuts, (d) Market Shares of Global Prepared 1976­2002 Groundnut Exports, 1976­2002 (tons) (percent) Metric tons Percent 350,000 100 90 300,000 80 250,000 70 200,000 60 50 150,000 40 100,000 30 20 50,000 10 0 0 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 United States China Netherlands (e) Global Exports of Groundnut Oil (f) Market Shares of Global Groundnut (1,000 metric tons) Oil Exports (percent) Metric tons Percent 550 90 Post-NAFTA 500 and -URAA 80 450 70 60 400 50 350 40 300 30 250 20 200 10 150 0 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 Senegal Argentina United States Brazil Source: FAOSTAT. 220 Global Agricultural Trade and Developing Countries TABLE 12.4 Share of Groundnut Products in Total Merchandise Exports Item Period The Gambia Malawi Nigeria Senegal South Africa Shelled groundnut exports Volume (metric tons) 1980­82 27,333 14,867 1,026 2,725 41,333 2000­02 27,939 662 412 2,915 34,830 Value (US$ thousands) 1980­82 11,743 12,333 400 2,145 29,730 2000­02 5,763 436 204 1,371 22,875 Groundnut oil exports Volume (metric tons) 1980­82 7,651 0 0 82,693 22,667 2000­02 8,633 0 1,287 98,879 1,519 Value (US$ thousands) 1980­82 6,400 0 0 60,285 14,071 2000­02 6,333 0 797 63,007 1,053 Share of groundnuts in 1980­82 59.62 4.65 0.003 16.17 0.21 total exports (percent) 2000­02 84.64 0.10 0.006 8.16 0.08 Note: The share of groundnuts in total goods exports includes exports of groundnut meal. Source: Production and groundnut exports data, FAOSTAT; total goods exports, World Bank. shelled groundnuts dwindled. The extent of the In sharp contrast with the 1970s, groundnut decline suggests that unit values also decreased in prices over the past 20 years have been stable, The Gambia, Malawi, and Senegal. constantly reverting to their mean values following shocks (tables A12.4 and A12.5 on the CD-ROM). Two subperiods can be distinguished. Before 1994 International Prices of Groundnuts prices of groundnuts displayed a higher level of Prices declined sharply in The Gambia and, to a volatility. The coefficient of variation of prices lesser degree, in Senegal, while prices in South stood at 20 percent between 1980 and 1994, almost Africa remained higher (figure 12.2). The discount three times the 7 percent level witnessed between on groundnuts of Gambian and Senegalese origin 1995 and 2001. reflects both their lower quality and stricter EU What are the main causes of this price variabil- quality and technical standards. The European ity? Is the change in price variability permanent? Union has become more demanding, from both a Revoredo and Fletcher (2002b) analyze both public health and a technical standpoint (size, production instability (originating in exporting uniformity).5 Nigeria and Senegal increased the countries) and consumption instability (originat- volume and value of their exports of groundnut oil, ing in importing countries). They find that the while South Africa exited the groundnut oil market. steady expansion of Chinese exports, which are International prices of edible groundnuts and negatively correlated with exports from Argentina groundnut oil in the Rotterdam market (the refer- and United States, was a stabilizing force in the sec- ence for groundnut trade) have exhibited two ond half of the 1990s. This stabilization occurred distinct patterns since 1970 (figure 12.3). During despite the fact that Argentina, India, and South 1970­81 the prices of both products were increas- Africa now transmit a higher proportion of their ing. Tests show no cointegration between edible supply shocks to the world market. groundnut and groundnut oil during this period Substitution between Chinese and U.S. ground- (tables A12.1­A12.3 on the CD-ROM). Prices were nuts appears to have increased in recent years, high, and the world market was dominated by although detailed data on substitution in world the United States, which supplied 45 percent of markets are not available. exports, and Sub-Saharan Africa, which supplied In the groundnut oil market, the influence of 18 percent. China exported no edible groundnuts Senegal on world prices remains significant. Senegal or groundnut oil. exported about 100,000 metric tons of groundnut Groundnut Policies, Global Trade Dynamics, and the Impact of Trade Liberalization 221 FIGURE 12.2 Unit Price of Raw Edible Groundnuts Produced in The Gambia, Senegal, and South Africa (US$ per ton) US$ per ton 1,200 1,000 800 600 400 200 0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 The Gambia Senegal South Africa Source: FAOSTAT. FIGURE 12.3 Rotterdam Prices of Groundnuts, 1970­2000 (cif, US$ per metric ton) US$ per metric ton 1,400 Prices are nonstationary Edible groundnut (U.S. and not integrated shelled runner 40/50) 1,200 1,000 800 Groundnut oil (any origin) 600 400 Period of stationary prices (tendency to revert to their mean values) 200 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 Source: Oil World, various issues. oil in 2000­01, representing one-third of world seems to be much stronger for other oilseeds than exports and more than 60 percent of demand from for groundnuts, however. the European Union, Senegal's main export market. While 2000­01 was an exceptional production year, Domestic Groundnut Policies of econometric tests strongly indicate that variations Major Countries in World Markets in Senegal's exports were transmitted into the vari- ability in international prices and that the reverse Domestic producer support and taxation and trade was not true. policies determine excess supply and trade flows. Groundnut oil markets are broadly integrated It is important to examine these policies--in with markets for other vegetable oils (soy oil, rape- Argentina, China, India, Sub-Saharan Africa, and seed oil, palm oil, and sunflower oil). Integration the United States--in some detail to anticipate the 222 Global Agricultural Trade and Developing Countries potential implications of policy changes on the dis- The 2002 Farm Bill eliminated groundnut pro- tribution of gains and losses across countries. duction quotas (with a quota buyout), converted the price support program to a system of direct and Groundnut Policies in the United States countercyclical payments, and set a price floor with a production subsidy (nonrecourse loans, with Groundnut products are a minor sector nationally, marketing loan provisions). The key features of the but they are a key component of agriculture and new program include the following: rural development in the southern part of the United States. Based on the U.S. Department of · All current groundnut producers have equal Agriculture (USDA) Census of Agriculture, many access to a marketing loan program under which counties in the South derive 50­70 percent of their producers can pledge their crops as collateral to agricultural income from groundnuts. Shelling is obtain a marketing loan rate equal to $355 per performed locally, as are many groundnut product short ton. Producers may repay the loan at a rate manufacturing activities (Fletcher 2001). Ground- that is the lesser of the USDA­set repayment nut policies have played a major role in maintain- rate plus interest or the marketing loan rate plus ing rural income in these counties. interest, or they can forfeit the loan. The foundation of U.S. groundnut policy is the · For producers with a history of groundnut pro- U.S. peanut program, which traces its roots to the duction, a new direct and fixed payment of $36 1930s. Until the enactment of the 2002 Farm Bill, per short ton is available. Historic producers are the pillars of the system were production regulation those who were engaged in groundnut produc- through quotas, high producer support prices, and tion between 1998 and 2001. Eligible production import control. The groundnut support program equals the product of average yields in the base existed as a two-tier price support program. The period and 85 percent of base-period acres. support price for edible groundnuts was $610 per These so-called decoupled payments are made short ton paid for production under quota. Other regardless of current prices or the actual crop groundnuts ("additionals"), which could be either planted, as long as the farm remains in approved exported at world prices or sold to the domestic agricultural uses. crushed groundnut industry, were eligible for a · Producers with a history of groundnut produc- lower support price ($132 in 2001). The quota tion are also eligible for a new countercyclical farm-gate price tended to be higher than the pre- payment when market prices are below an estab- vailing export prices (table 12.5). lished target price of $495 per short ton minus The average annual aggregate measure of sup- the $36 per ton direct payment. The payment port for U.S. groundnuts was estimated at $330 mil- rate is the difference between the target price lion during 2000­01, $31 million more than in ($495 per short ton) minus the direct fixed pay- 1996­2001. The average cost of aggregate support ment ($36 per ton) and the higher of the 12- in 1996­01 stood at $206 per metric ton of ground- month national average market price for the nuts produced in the United States (table 12.6). marketing year for groundnuts or the marketing TABLE 12.5 U.S. Producer Support Prices for Groundnuts, 1993­94 to 1998­99 (US$ per pound) Item 1993­94 1994­95 1995­96 1996­97 1997­98 1998­99 Quota price 0.388 0.339 0.339 0.305 0.305 0.305 "Additional" price 0.660 0.660 0.660 0.660 0.660 0.660 Average farm price 0.304 0.289 0.293 0.281 0.283 0.280 CCCa export price 0.200 0.200 0.200 0.200 0.200 0.200 Export unit value 0.330 0.286 0.292 0.322 0.327 0.328 Rotterdam cif price 0.371 0.292 0.336 0.316 0.360 0.290 a. Commodity Credit Corporation. Source: Skinner 1999. Groundnut Policies, Global Trade Dynamics, and the Impact of Trade Liberalization 223 TABLE 12.6 U.S. Aggregate Support to Groundnuts, 1986­88 to 2000­01 Aggregate Measure of Support Aggregate Measure of Support per Period (millions of US$) Metric Ton of U.S. Production (US$) 1986­88 347.2 203.3 1995­96 414.6 264.1 1996­97 299.0 180.0 1997­98 305.8 190.5 1999­2000 300.0 172.7 2000­01 330.0 222.8 Source: Skinner 1999; Hart and Babcock 2002 for 2000­01 aggregate measure of support; USDA database for production data. TABLE 12.7 U.S. Edible Groundnut Tariff Rate Quota Allocation, 1995­2008 (metric tons) Argentina Mexico Other Total Uruguay Round NAFTA Tariff Uruguay Round NAFTA + Uruguay Year Tariff Rate Quota Rate Quota Tariff Rate Quota Round 1995 26,341 3,478 4,052 33,871 1996 29,853 3,582 5,043 38,478 1997 33,364 3,690 6,034 43,088 1998 36,877 3,801 7,024 47,702 1999 40,388 3,915 8,015 52,318 2000 43,901 4,032 9,005 56,938 2001 43,901 4,153 9,005 57,059 2002 43,901 4,278 9,005 57,184 2003­07 43,901 4,278 9,005 -- 2008 43,901 unrestricted 9,005 -- -- Not available. Source: USDA. assistance loan rate ($355 per short ton). The Beginning in 1994, under the Uruguay Round total countercyclical payment to each eligible and NAFTA (North American Free Trade Agree- producer is calculated as the product of the pay- ment), the United States gradually increased the ment acres (85 percent of base acres), the base- quantities of groundnut imports through a tariff year average yield, and the payment rate. rate quota system. For edible groundnuts the total · Owners of groundnut quotas under the previ- tariff rate quota in 2001 was 57,059 metric tons, or ous legislation receive compensation payments 4 percent of domestic consumption, allocated first for the loss of quota asset value. Payments may to historical importers and then on a first-come, be made in five annual installments of $220 per first-served basis (table 12.7). In-quota tariffs for short ton during fiscal years 2002­06, or the edible and prepared groundnuts range between quota owner may opt to take the outstanding $.066 and $.0935 per kilogram. Out-of-quota payment due in a lump sum. These payments tariffs are very high (131.8­163.0 percent under the are based on quota owners' 2001 quota, as long Uruguay Round; table 12.8).6 as they owned a farm eligible for the groundnut The phase-out of groundnut trade barriers quota (Wescott, Young, and Price 2002). under NAFTA and the Uruguay Round is limited in 224 Global Agricultural Trade and Developing Countries TABLE 12.8 U.S. Over-Quota Tariffs, 1994­2008 (percent) Uruguay Round NAFTA Prepared Groundnuts and Year Edible and Prepared Groundnuts Peanut Butter Basea 123.1 155.0 1994 120.0 -- 1995 116.9 151.1 1996 113.9 147.3 1997 110.8 143.4 1998 107.7 139.5 1999 104.6 135.7 2000 93.0 131.8 2001 81.4 131.8 2002 69.8 131.8 2008 0.0 -- a. Note: This indicates the base tariff levels from which the agreed cuts will be made. -- Not available. Note: Prepared groundnuts are roasted, salted, or flavored groundnuts. Source: USDA. TABLE 12.9 U.S. Imports of Edible Groundnuts Argentina Mexico Over-the- Total Total Quota Year Imports Quota Imports Quota Imports Quota Imports 1996 38,270 29,853 4,710 3,583 57,000 38,478 18,522 1997 40,622 33,365 6,148 3,690 64,000 43,088 20,912 1998 34,465 36,875 4,834 3,801 70,000 47,702 22,298 1999 39,494 40,388 4,916 3,915 82,000 52,318 29,682 2000 72,230 43,901 4,864 4,032 97,000 56,938 40,062 2001 37,557 43,901 3,611 4,153 81,000 57,059 23,941 2002 29,927 43,901 4,406 4,278 46,795 57,184 Not filled 2003 4,692 43,901 292 4,278 5,698 57,184 Not filled Source: USDA. scope, but it continues to have a dramatic impact decreased the price paid by U.S. processors, increas- on U.S. imports. Edible groundnut imports by the ing domestic consumption of groundnuts (fig- United States--which were almost zero before ure 12.4). It also removed the incentive to import 1994--have increased dramatically (table 12.9). edible groundnuts (Fletcher and Revoredo 2003; Argentina's average fill rate was 87 percent and Revoredo and Fletcher 2002a). Mexico's, 77 percent, but out-of-quota imports were quite important, averaging 25,000 metric tons annually during 1996­2001. Edible groundnut Groundnut Policies in Argentina, China, and India imports represented 6 percent of U.S. groundnut consumption in 2001. Since the mid-1990s Argentina, China, and India The initial impacts of the 2002 Farm Bill are also have gradually reduced potentially market-distorting reflected in the collapse of imports in 2003 (see direct government intervention in the production table 12.9). The elimination of production quotas and marketing of groundnut products (table 12.10). Groundnut Policies, Global Trade Dynamics, and the Impact of Trade Liberalization 225 FIGURE 12.4 U.S. Domestic Groundnut Prices, 1993­2003 (US$ per metric ton) 1,100 Rotterdam price 1,000 900 800 700 US domestic price 600 500 400 New farm bill 300 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Source: USDA. TABLE 12.10 Groundnut Trade Policy Distortions in Argentina, China, and India (percent) Country Product Description 1999/2002 Argentina Edible groundnuts Import tariff 5 Groundnut oil Import tariff 13 Groundnut meal Import tariff 8 Edible groundnuts Export tax 3.5 Groundnut meal Export rebate 3.2 Groundnut oil (refined) Export rebate 2.3 China Raw edible groundnuts Import tariff 15 Processed edible groundnuts Import tariff 30 Groundnut oil Import tariff 10 Groundnut meal Import tariff 5 VAT on edible groundnuts and groundnut oil VAT 17 India Edible groundnuts Import tariff 45 Groundnut oil Import tariff 35 Groundnut meal Import tariff 45 Note: India raised its tariff on oil to 65 percent in 2002 and 85 percent in 2003. Raw edible groundnuts are raw, not roasted or cooked, in shell or shelled groundnuts. Processed groundnuts are bleached, preserved, or otherwise prepared groundnuts, including roasted, salted, and groundnut butter. Source: WTO; WITS; USDA GAIN Report. Argentina. Argentina's groundnut trade policy increased to 20 percent. This export tax may coun- contrasts sharply with that of India and China, as tervail the positive signal sent to groundnut almost all the distortions are associated with exporters through the peso devaluation. Argentina exports, which are subject to a 3.5 percent tax on maintains import tariffs on groundnut products, raw groundnuts. With the peso devaluation of which exhibit some escalation (5 percent on edible 2001, export tax retention on groundnut exports groundnuts, 8 percent on groundnut meal, and 226 Global Agricultural Trade and Developing Countries 13 percent on groundnut oil). These tariffs are Tariffs on edible groundnuts and groundnut often redundant, since the country is a net exporter meal stood at 45 percent, while the tariff on ground- of groundnut products. nut oil was 35 percent in 2001 (see table 12.10). Since 2002 India has reversed its trade liberalization China. Like India, China liberalized groundnut course on vegetable oil, increasing applied tariffs trade to some degree in recent years. Before 1999 on groundnut oil to 65­75 percent in 2002­03 six public agencies were the only importers of and 85 percent in 2003­04 (Gulati, Pursell, and groundnut products; today, private firms are free to Mullen, 2003; Pursell, 2003). The bound tariff is import groundnuts. However, while the govern- 100 percent. ment has committed to cap and reduce trade- Regulatory burdens increase domestic costs and distorting domestic subsidies as part of its WTO prices. Producers are obligated to sell and purchase accession commitments, guaranteed prices and groundnuts only in the agricultural produce government procurement schemes remain in wholesale market.9 The "small-scale reservation" place.7 Furthermore, border protection remains policy in groundnut processing sets limits on fixed high for processed groundnuts (30 percent). The assets in plant and machinery, preventing the tariff on raw groundnuts was only 15 percent in domestic processing industry from realizing 2001, and many regions of China are natural economies of scale. exporters of groundnuts, making the tariff redun- dant. In-quota tariffs on groundnut oil (10 percent) Groundnut Policies of Key African Exporters and groundnut meal (5 percent) were much lower. USDA attaché reports have repeatedly raised the After decades of extensive intervention in the issue of the uneven application of the Chinese groundnut sector, to varying degrees African coun- value-added tax (VAT) on imported and domestic tries underwent market reforms in the 1980s under products. The VAT is significant, ranging from 13 structural adjustment plans. One of the main to 17 percent, depending on the product, and there objectives of market reforms was to eliminate is ample room for tax evasion (USDA FASa, USDA direct and indirect taxation of farmers that had FASb). The lack of uniformity in application pre- undermined production incentives in the 1970s vents a more accurate measure of the impact. The and early 1980s and led to underutilized processing quantitative policy analysis presented in the next capacities in many groundnut producing countries section examines several cases with and without (Badiane and Kinteh 1994). the VAT. Reforms have been piecemeal and partial. Gov- China's state trading imposes quantitative ernments have generally withdrawn from input restrictions through quotas and licenses on ground- markets, making it difficult for producers to obtain nut oil imports, and it imposes tariff barriers on certified seeds and fertilizer in countries such as seeds, meal, and oil. These barriers create a wedge The Gambia and Senegal, where there are market between domestic and world market prices. failures (in the credit market and elsewhere) and Domestic prices of most oils, including groundnut high transactions costs (Akobundu 1998). Govern- oil, are significantly higher than international mar- ments have been reluctant to liberalize groundnut ket prices. Tariffs and rents on import licenses processing, for which privatization efforts started explain the price differentials.8 only recently (in The Gambia and Senegal). In The Gambia and Senegal producer prices are still set by India. India removed most restrictions on the government. domestic trade, storage, and export of groundnuts African governments have traditionally used by 1998 and permitted trading in groundnut pricing policies as convenient levers to tax or subsi- futures. However, import tariff levels remain very dize farmers based on their industrial policies and high for all three groundnut products. Moreover, in political circumstances.10 Taxation of groundnut response to declines in prices, India has intensified farmers was high in the 1970s but has been reversed its use of trade policy measures to protect its pro- since the early 1990s in most countries, while real ducers and processors. India is now the largest world prices have trended downward (Badiane and source of distortions in these product markets. Kinteh 1994). In The Gambia and Senegal the main Groundnut Policies, Global Trade Dynamics, and the Impact of Trade Liberalization 227 rationale for state intervention in the groundnut TABLE 12.11 Tariffs on Groundnut sector has been to safeguard the viability of state- Products in The Gambia, owned processing mills. Consequently, the share of Malawi, Nigeria, Senegal, and South Africa the export price to groundnut farmers has consis- (percent) tently been less than 60 percent in these countries. This policy has been counterproductive for the Country Product Tariff 1999­2002 state-owned enterprises, since farmers have bypassed large public processing companies, creat- Gambia, The ing excess capacity and financial difficulties. Edible groundnuts 0 Groundnut meal 0 Trade policies vary widely among traditional Groundnut oil (refined) 0 groundnut exporters in Africa. Malawi and Senegal Malawi apply high tariffs on processed groundnuts to Raw edible groundnuts 5 encourage domestic processing or oil production Processed edible 25 (table 12.11). In contrast, The Gambia and Nigeria groundnuts have liberal trade policies, with no tariffs or export Groundnut meal 0 taxes. South Africa's tariff structure exhibits a slight Groundnut oil (refined) 20 escalation, with processed groundnuts subject to a Nigeria tariff of 6 percent while unprocessed groundnuts Edible groundnuts 0 enter duty free. In The Gambia and Senegal unoffi- Groundnut meal 0 Groundnut oil (refined) 0 cial cross-border trade is significant, with farmers Senegal frequently crossing the border depending mainly on Raw edible groundnuts 5 producer prices and domestic supply levels in the Processed edible 20 two countries. Groundnut oil imports face a 20 per- groundnuts cent tariff in Malawi, Senegal, and South Africa. Groundnut meal 0 African countries are facing difficulties meeting Groundnut oil (refined) 20 EU standards on aflatoxin and stricter product and South Africa quality standards. In The Gambia and Senegal Raw edible groundnuts 0 groundnut varieties were originally selected for oil Processed edible 6 groundnuts production, which can accommodate lower-quality Groundnut meal 0 seeds, and raw groundnuts. A seed variety in Groundnut oil (refined) 20 Malawi proved successful in producing better yields, but it lacked commercially viable character- Note: Raw edible groundnuts are raw, not roasted istics. Groundnuts exported from most African or cooked, in-shell or shelled groundnuts. countries are sold at a discount relative to the high- Processed groundnuts are bleached, preserved, or quality groundnuts sold in the European Union. otherwise prepared groundnuts, including roasted and salted groundnuts and groundnut butter. African producers may be able to shift out of Source: World Trade Organization database. groundnut oil and upgrade the quality of their edible groundnuts. Unlike demand for groundnut groundnuts. Were The Gambia able to upgrade oil and meal, demand for confectionery groundnut 50 percent of its 10,000 tons of exports from crush- (the higher-quality edible groundnut) has been ris- ing to edible groundnuts, it would increase its ing and is expected to continue to increase in the revenues by $1.5 million. medium term. Confectionery groundnuts receive a price premium of as much as 100 percent over Groundnut Trade Policies of High-Income grades used for oil and meal. In Senegal 1 ton of Importers first-grade confectionery groundnuts sells for $800­$900, equivalent to the price of groundnut Despite a general pattern of tariff escalation, trade oil. It takes three tons of unshelled groundnuts to barriers are not a major obstacle to high-income produce 1 ton of oil. However, fob (free on board) groundnut importers: the European Union and prices of Gambian groundnuts are about $300 for Canada have a zero tariff for unprocessed ground- crushing, $450 for birdfeed, and $600 for edible nuts and low tariffs for processed groundnut for 228 Global Agricultural Trade and Developing Countries TABLE 12.12 Average Tariffs on Edible Unprocessed and Processed Groundnuts (percent) Most-Favored-Nation Generalized System of Low-Income Developing- Tariffs Preferences Tariffs Countries Tariffs Importer Unprocessed Processed Unprocessed Processed Unprocessed Processed Canada 0 4 0 4 0 3.2 European Union 0 13 0 9 0 0 Japan 3.7 19 3.7 19 3.7 15 Korea, Rep. of 243 65 243 65 243 65 Source: WTO. generalized system of preferences (GSP) and low- tion of the U.S. peanut program (FMTL&US income developing countries (table 12.12). Assess- and FMTL) ment of market access in these countries must, · Multilateral trade liberalization of groundnuts, however, take into account the strict quality with and without elimination of the U.S. peanut standards. program (GMTL&US and GMTL) In contrast to the European Union and Canada, · Full trade liberalization in the two largest and Japan and the Republic of Korea have high tariff most distorted groundnut markets, China and regimes for groundnuts. Japan applies a high tariff India (CIFTL) on processed groundnuts and offers a very limited preference margin of 4 percent for groundnut Summary results of these five scenarios are pre- exports from low-income developing countries. sented in tables 12.13 and 12.14. Detailed results for Korea has very high tariffs on both raw and pro- each scenario are presented in the longer report on cessed groundnuts, with tariffs on raw groundnuts of the CD-ROM. more than 200 percent. This high tariff may reflect the government's desire to stimulate production, Analysis Results which has plummeted since the mid 1980s. In con- trast to edible groundnuts, groundnut oils and meal In countries with high groundnut protection, the enter all of these high-income countries duty free. combined effect of the world price increase and elimination of their own protection is beneficial to Impact of Groundnut Product final users of groundnuts, other things being equal. Policy Reforms on World Prices, For countries with moderate or no protection Trade Flows, and Welfare before reform, tariff elimination and changes in the Several key findings emerge from the quantitative terms of trade result in an increase in domestic analysis of distortions in groundnut markets. (For a groundnut prices, handicapping groundnut users full description of the model, see Beghin and others (consumers and crushers). A similar logic holds for 2003.) The main results obtained under the most groundnut oil and meal, for which the combined plausible assumptions underlying the model are effect of world price increases and the elimination presented first, followed by sensitivity analysis test- of tariffs has to be assessed. These substantial ing the effects of U.S. policy and uncertainty about terms-of-trade effects have a significant impact on protection by China (VAT and protection of trade and welfare. Allocative efficiency gains in processed groundnuts). domestic markets can be offset by large price increases originating in postreform world markets. Policy Reform Scenarios In countries with high protection of groundnut oil or meal (such as India), tariff elimination, net of Several scenarios are analyzed: the world price hike, induces lower domestic prices · Full multilateral trade liberalization for ground- and reduces margins on crushed groundnuts. As a nuts, meal, and oil, with and without elimina- result, the domestic excess supply of groundnuts Groundnut Policies, Global Trade Dynamics, and the Impact of Trade Liberalization 229 TABLE 12.13 Welfare Effects of Policy Scenarios, 1999­2001 Average (millions of 1995 US$) Country FMTL&US FMTL GMTL&US GMTL CIFTL Argentina 16.07 15.94 9.97 9.84 12.66 EU-15 -51.83 -51.27 -34.40 -33.82 -58.87 China 666.25 668.76 650.65 653.33 716.25 India 213.27 214.11 196.57 197.79 228.59 Rest of the world -126.69 -127.06 -4.21 -4.86 -71.06 Canada -5.94 -5.87 -4.88 -4.81 -4.59 Mexico -7.43 -7.34 -6.11 -6.01 -5.73 Senegal 41.03 40.96 21.93 21.86 21.39 Nigeria 15.93 15.77 7.22 7.07 13.45 South Africa 2.30 2.28 2.19 2.17 0.53 Malawi 7.45 7.45 7.60 7.61 -1.06 The Gambia 0.43 0.42 0.24 0.24 0.36 USA 20.18 16.70 21.71 18.40 12.39 Africa-5 total1 67.14 66.89 39.18 38.95 34.67 Total 791.01 790.87 868.48 868.79 864.32 1. Denotes the aggregate of Senegal, Nigeria, South Africa, Malawi, and the Gambia. Source: USDA. crushed into oil and meal decreases, increasing crush margin reduces groundnut oil production imports. In contrast, countries with moderate or no and meal production; lower consumer prices for all protection in their oil and meal markets face a net groundnut products stimulate groundnut oil pro- price increase for oil and meal after full trade liber- duction and increase demand and eventually alization. Their final consumption of these value- imports. African producers expand their exports of added products decreases, and crushing increases, value-added products. Malawi and Senegal as the crush margin improves. Their excess supply decrease their exports of groundnuts because of of these products increase, and they increase increased domestic use, while India experiences a exports. trade reversal, becoming a large importer of The two full trade liberalization scenarios with groundnut oil and meal. Aggregate trade in and without the elimination of the U.S. farm policy groundnuts increases 16 percent, and trade in (FMTL&US and FMTL) induce strong price value-added products more than doubles. increases for all three products: 10 percent for The aggregate net welfare effects of FMTL&US groundnuts, 18 percent for groundnut meal, and and FMTL amount to about $791 million at 1995 27 percent for groundnut oil (see table 12.13). The prices in each scenario (see table 12.14). China and welfare impact of the FMTL&US and FMTL India experience the largest welfare gains--not sur- reforms is influenced by the change in the ground- prisingly, since they have the two largest and most nut oil price, which affects the crush margin. Crush distorted groundnut product markets. China's wel- margins narrow in the European Union and India, fare gains are about $666 million, India's are about decreasing supply, but they may increase in China, $213 million. The "moderate" magnitude of global The Gambia, Nigeria, Senegal, South Africa, and welfare gains first comes from offsets--some coun- the United States. tries gain in aggregate whereas some others, chiefly Trade patterns change dramatically. China the European Union-15, lose. For many countries expands its exports of the three products. The large other than China and India, individual net gains increase in the price of groundnut oil improves the and losses are moderate, mostly because of the crush margin, stimulating crushing in China. small size of the groundnut markets and their Higher prices for groundnut oil in world markets price-inelastic nature, which produces large translate into larger exports. In India the lower transfers but small deadweight losses. Indeed, 230 Global Agricultural Trade and Developing Countries TABLE 12.14 Impact of Different Liberalization Scenarios on Groundnut Trade and Prices (percent) FMTL&US FMTL GMTL&US GMTL CIFTL (percent except welfare) for 3 years for 3 years for 3 years Peanuts Trade (1,000 metric tons) Net Exporters Argentina 7 6 22 22 -6 China 36 34 42 41 13 Gambia, The 11 11 31 30 3 India -62 -64 -556 -557 -94 Malawi -80 -82 -93 -95 84 Nigeria 3667 3564 7470 7358 1776 Senegal -287 -298 -8 -20 -708 South Africa 22 22 20 20 14 United States 8 15 48 55 2 Total net exports 16 16 -17 -16 -4 Net Importers Canada -5 -5 -4 -4 -4 European Union -3 -3 -3 -3 -2 Mexico -8 -8 -7 -7 -6 Rest of the world 63 65 -47 -45 -7 Residual 0 0 0 0 0 Total net imports 16 16 -17 -16 -4 Peanuts Price US Run. 10 10 8 8 8 40/50 cif Rotterdam $ per mt Peanut Meal Trade (1,000 metric tons) Net Exporters Argentina 13 13 -6 -6 18 China 741 739 -144 -146 759 Gambia, The 22 21 -2 -2 22 India -1702 -1703 344 342 -1690 Malawi 9 8 1 1 46 Nigeria 2867 2862 -193 -196 2867 Senegal 5 5 -1 -1 14 South Africa 95 95 -7 -7 139 United States 484 487 -380 -376 563 Rest of World 499 499 -70 -69 385 Total Net Exports -9 -9 0 0.4 -9 Net Importers European Union -12 -12 1 1 -13 Residual 0 0 0 0 0 Total net imports -9 -9 0 0 -9 Meal Price 48/50 cif 18 18 0 0 18 Rotterdam $ per mt Peanut Oil Trade (1,000 metric tons) Net exporters Argentina 11 11 -6 -6 16 Groundnut Policies, Global Trade Dynamics, and the Impact of Trade Liberalization 231 TABLE 12.14 (Continued) FMTL&US FMTL GMTL&US GMTL CIFTL (percent except welfare) for 3 years for 3 years for 3 years China 3469 3459 -705 -713 3354 Gambia, The 589 587 -5 -6 567 India -24288 -24304 4591 4558 -24481 Malawi 43 43 0 0 185 Nigeria 123 123 -6 -6 120 Senegal 5 5 -1 -1 19 South Africa 49 49 -4 -4 224 United States 288 290 -194 -192 326 Rest of world 861 864 -103 -99 665 Total net exports -6 -6 0 0 -8 Net importers European Union -9 -9 0 0 -12 Residual 0 0 0 0 0 Total net imports -6 -6 0 0 -8 Peanut Oil Price cif 27 27 0 0 26 Rotterdam $ per mt welfare(million dollars) 791 791 868 869 864 Argentina 16.07 15.94 9.97 9.84 12.66 EU-15 -51.83 -51.27 -34.40 -33.82 -58.87 China 666.25 668.76 650.65 653.33 716.25 India 213.27 214.11 196.57 197.79 228.59 Rest of world -126.69 -127.06 -4.21 -4.86 -71.06 Canada -5.94 -5.87 -4.88 -4.81 -4.59 Mexico -7.43 -7.34 -6.11 -6.01 -5.73 Senegal 41.03 40.96 21.93 21.86 21.39 Nigeria 15.93 15.77 7.22 7.07 13.45 South Africa 2.30 2.28 2.19 2.17 0.53 Malawi 7.45 7.45 7.60 7.61 -1.06 Gambia, The 0.43 0.42 0.24 0.24 0.36 United States 20.18 16.70 21.71 18.40 12.39 Africa-5 totala 67.14 66.89 39.18 38.95 34.67 Total 791.01 790.87 868.48 868.79 864.32 Note: Table totals are average changes for three years totaled in each column. Results are percentage changes from the baseline. Baseline and simulations were run for three years (1999­2001) and averaged. a. Totals are three-year averages. Source: computed by the authors a. Denotes the aggregate of Senegal, Nigeria, South Africa, Malawi, and The Gambia. substantial offsetting transfers occur between con- gram, groundnut producers gain $34 million in sumers, crushers, and producers. quasi-rents, consumers experience welfare losses of Price effects induced by the reforms have a simi- $65 million (because of higher oil and processed lar impact large welfare transfer (rectangles) and groundnut prices), crushers gain $51 million, and small net welfare effects (triangles), even in coun- meal users (feed users) lose about $3 million. In tries with undistorted markets. In Nigeria, for aggregate the country is better off by $16 million. example, following full multilateral trade liberaliza- Under multilateral trade liberalization for all tion without elimination of the U.S. peanut pro- three products, elimination of the U.S. program 232 Global Agricultural Trade and Developing Countries affects trade flows, terms of trade, and welfare. The scenarios. The rest of the world fares much better strong price effects of trade liberalization invalidate under the groundnut trade liberalization scenarios the price floor established by the U.S. loan rate. The than under the full liberalization scenarios. In con- only remaining production-distorting element is trast, African economies do much better under the the fixed payment (fully coupled to production in full liberalization scenarios than with groundnut the model), which is small. Results under both sce- trade liberalization reforms. The potential welfare narios (full trade liberalization with and without gains for Africa-5 (The Gambia, Malawi, Nigeria, elimination of the U.S. peanut program) are quali- Senegal, and South Africa) nearly double by mov- tatively identical, except for the United States, ing from groundnut trade liberalization to full which experiences additional welfare gains of $3.5 liberalization scenarios. million (gains to U.S. taxpayers net of losses by U.S. If China and India liberalized alone (the CIFTL producers) by eliminating its domestic distortions. scenario), the qualitative results of the full liberal- The world price impacts of the FMTL scenario ization scenarios hold. What is striking in this last are identical to those of FMTL&US (a 10 percent scenario is the importance of India's--and to a increase for groundnuts, an 18 percent increase for lesser extent China's--distortions and market size meal, and a 27 percent increase for groundnut oil). on welfare, trade, and price effects. As suggested by Trade flows are barely affected by the elimination of table 12.13, FMTL really hinges on the elimination the U.S. domestic program under free trade. U.S. of distortions in India and China. With liberaliza- groundnut exports are about 15,000 metric tons tion in India and China, world prices would rise lower in the FMTL&US scenario than in the FMTL 8 percent for groundnuts, 18 percent for meal, and scenario. Given that the parameterization of U.S. 26 percent for oil. The major welfare differences farm policy assumes full coupling to production for occur in the rest of the world, where consumers are payments received by producers, the assessment worse off than they would be under the multilat- provides an upper bound on the effect of the cur- eral groundnut trade liberalization scenario, since rent U.S. peanut program.11 groundnut oil prices are higher. Africa-5 improves Many agricultural negotiations during the Doha its lot in aggregate but by less than it would under Round of the WTO revolve around narrow issues of the full liberalization scenario, since groundnut substantial importance to developing countries. prices are lower and distortions within Africa-5 Hence it is useful to assess what a narrow agricul- remain in place. tural liberalization encompassing the value-added Two key assumptions in the model--the prevail- products of groundnut oil and meal would achieve ing groundnut market price in the U.S. market and relative to full trade liberalization. the level of protection of the groundnut market in The GMTL&US and GMTL scenarios consider China--were investigated. The model was cali- these reforms and their impacts. Much is achieved brated on 2002­03 U.S. prices ($389 per metric by groundnut trade liberalization alone, but with a ton) to see if the new U.S. policy under the 2002 large second-best component, since distortions are Farm Bill would have had a stronger impact on the present in the value-added markets. In these world market under lower prevailing prices (farm groundnut liberalization scenarios, the prices of prices in the United States were 25 percent lower in meal and oil are little affected, and crush margins 2002­03 than in 2001­02). The loan rate, counter- are driven primarily by changes in groundnut cyclical payments, and fixed payments were elimi- prices. Margins improve in India but deteriorate in nated (countercyclical payments and fixed pay- countries with limited groundnut distortions. ments are assumed to be fully coupled to provide Consumer welfare implications are also differ- an upper bound on the effect of the U.S. program), ent in these trade scenarios. In highly protected while all distortions in all other countries were groundnut oil markets, prices are higher under the retained. The price floor provided by the loan rate groundnut trade (GMTL) scenarios than they are is effective under the lower 2002­03 farm price. under all-product trade liberalization (FMTL sce- U.S. output decreases 7 percent under the new narios). In countries with no oil distortions, prices prices, and U.S. exports decrease 52 percent, induc- remain roughly at their baseline level, and con- ing a 0.9 percent increase in the world price of sumers do better under the groundnut trade groundnuts and negligible price impacts in the liberalization than under the full liberalization other markets. The aggregate net welfare effect is Groundnut Policies, Global Trade Dynamics, and the Impact of Trade Liberalization 233 negligible and negative. Higher world prices exac- States the 2002 Farm Bill eliminated some unsus- erbate distortions in other markets or increase tainable features of earlier policies, but it intro- import costs in net importing countries. The duced new distortions that have some limited United States gains about $22 million (program potential to depress world market prices and subsi- cost savings net of producer loss). dize U.S. groundnut exports. India and China have The same change was also tested with all other succeeded in stimulating production and exports, distortions removed in all countries (FMTL&US capturing a growing share of the international mar- scenario). In this scenario the world price of ket. In India these gains have been artificial, groundnuts was 0.5 percent higher than under free because the groundnut industry relies on heavy trade plus the 2002 U.S. Farm Bill. The results sug- protection. In contrast, in Sub-Saharan Africa and gest that removing the 2002 Farm Bill incentives in Argentina, government intervention has hurt a free trade world would decrease U.S. production the sector. about 4 percent and exports about 31 percent. The Unlike U.S. policy for cotton, dairy, rice, and aggregate welfare gains vary by less than $1 million. sugar, the current U.S. domestic peanut program is Hence the conclusion that the new U.S. groundnut now largely a domestic support program with some policy is more benign than its predecessor remains distortive impact. The scenario analysis suggests unaltered under much lower prices. that developing countries would gain little by try- The sensitivity analysis on China's protection ing to negotiate further U.S. groundnut policy structure is more pivotal to the conclusions reached, reform, as these changes would prove ineffective especially the protection of the groundnut sector. unless groundnut prices fell to very low levels. Only Protection of groundnut producers is now assumed then would the countercyclical U.S. policy further to be 15 percent (the tariff is redundant in the origi- destabilize world prices, sending the wrong market nal model). Without assistance the Chinese farm signal to U.S. producers. Under prevailing market sector is no longer assumed to be a net exporter. conditions, U.S. producers would actually bene- Under this new assumption and following full trade fit from multilateral trade liberalization in ground- liberalization (FMTL&US), China becomes a net nut product markets. Hence it would be rational importer of groundnuts, because demand for edible for the United States to support foreign groundnut and crushed groundnuts increases. China's welfare producers in their attempt to liberalize. As a gains are $1,029 million; aggregate welfare gains are bloc most OECD (Organisation for Economic Co- $1,160 million. World prices increase 18 percent operation and Development) countries would for groundnuts, 19 percent for meal, and 29 percent experience welfare losses after trade liberalization, for oil. with moderate gains in the United States offset by A second sensitivity analysis examines the effect losses in Canada, Mexico, and the European Union- of providing baseline protection of processed 15. Mexico, Canada, and the European Union-15 groundnuts with a 15 percent ad valorem tariff (the would lose from trade liberalization because there original tariff was 30 percent and the VAT was 17 are few distortions in these markets, so consumers percent). Under this assumption welfare gains from would be penalized by price increases for ground- FMTL are only $266 million in China, and aggre- nut products. gate gains are just $388 million. The world price of Elimination of trade distortions by the two groundnuts increases 9 percent in this modified largest developing economies, India and China, is scenario, down from 10 percent under the original essential. The size of their markets--and the huge run. The major change in welfare occurs in China, distortions in India--substantially depress world because Chinese consumers gain much less from prices of the three globally traded groundnut prod- trade liberalization relative to the initial situation. ucts. Following elimination of these distortions, net buyers of these products would be worse off, but most Sub-Saharan African countries that export Conclusions and Policy Implications groundnuts would gain. The groundnut market has historically been dis- Full trade liberalization would increase world torted by heavy government intervention in both market prices about 10 percent for groundnuts, industrial and developing countries. In the United 18 percent for groundnut meal, and 27 percent for 234 Global Agricultural Trade and Developing Countries groundnut oil. Trade in groundnuts would increase created by the expansion of the edible groundnut 16 percent, and trade in oil and meal would more exports market because of inadequate quality. than double. Although the net world welfare effects of liberal- Notes izing markets in the United States, China, and India 1. Groundnuts are also known as peanuts, earthnuts, are moderate, they remain significant for small goobers, pinders, and Manila nuts. The groundnut plant is a agrarian economies in Sub-Saharan Africa. Liberal- hairy, tap-rooted, annual legume that measures 1­1.5 feet in ization would thus produce welfare gains in the height. countries in which they are most needed. 2. Groundnut oil is an excellent cooking oil, with a high smoke point and neutral flavor and odor. Groundnut meal is The simulations show that liberalization of the used as animal feed. value-added markets is essential to achieve larger 3. Any grower in the group is eligible to farm a certain num- welfare gains in African countries. Although the ber of acres of land. Farmers who use the land are obligated, however, to pay agricultural tax in kind and sell a certain bulk of the world welfare gains occur with ground- amount of their products to the state government at regulated nut trade liberalization, elimination of distortions prices. in value-added markets doubles net welfare gains 4. Export markets reflect relatively high-quality premiums and discounts. Prices of U.S. groundnuts set a ceiling for inter- in Africa by yielding larger profits to producers national prices, because the quality is high. Edible U.S. ground- and exporters of groundnuts and groundnut oil. nuts commanded a 40 percent premium on world markets over African countries modeled in the trade liberaliza- shelled Chinese groundnuts in 2000 (FAO 2002). tion analysis would experience aggregate welfare 5. In 1998 the EU harmonized country regulations on the maximum permissible level of aflatoxin, setting levels at the low- gains of $67 million, with Senegal and Nigeria est possible level (0.002 milligrams for B1 type aflatoxin for edi- reaping most of these gains. Groundnut and ble groundnuts). Aflatoxin is a cancer-causing chemical pro- groundnut oil consumers in Africa tend to be duced by species of aspergillus molds that can contaminate groundnuts. The spores of these molds, present anywhere in the urban, whereas groundnut production generates air and the soil, require specific temperature, moisture, and income in rural areas as a cash crop. African nutrient substrates to germinate. Aflatoxin contamination of groundnut producers modeled in the analysis gain groundnuts can occur during cultivation in the field, as well as during harvesting, postharvesting, storage, or processing. While $50 million­$150 million of farm income, depend- aflatoxin disappears with crushing, it remains in edible ground- ing on the assumptions underlying the model. nuts and groundnut meal. Technical processes exist to reduce These figures are significant for small African aflatoxin contamination (with ammoniac, for example, which economies and represent an important opportu- Senegal uses on groundnut meal), but the best method is to improve farm practices through use of best-quality and resistant nity to expand rural development in these areas. In seeds, proper management of farms, and appropriate storage to the scenarios tested, the rest of the world would fare avoid exposure to high temperature and humidity. See chapter 6 worse under full trade liberalization, because con- in this volume for a discussion of food safety and agricultural health standards. sumers are required to pay higher groundnut prod- 6. The levels of quota and tariff for the period after 2003 are uct prices. Groundnuts are not without substitutes, currently under negotiation. however, and the price increase may induce 7. According to the FAO (Food and Agriculture Organiza- tion), these policies provide little incentive to expand produc- increases in demand for other oils. tion due to unattractive administrative price levels and greater Recent changes present both challenges and involvement of the private sector in marketing operations. Data opportunities for major countries in the market. on the magnitude of domestic support are not available. The United States is likely to continue to dominate 8. The international price of groundnut oil in Hong Kong (China) was $728 a ton in 1998, and the wholesale price in the high end of the international confectionery China was 67.8 percent higher (Fang and Beghin 2002). market under its new peanut program. Argentina 9. This regulation imposes a significant burden on farmers and China have established strong groundnut sec- and processors. Even if they are located very close to one another, they have to travel to the wholesale market and pay an tors that can compete favorably under free market "agent commission" and other marketing fees before the trans- conditions. Chinese exports played a stabilizing action is processed. role in world markets in the 1990s. 10. Taxation of producers was direct (that is, marketing boards or similar agencies captured the rent, equal to the differ- All developing countries except Argentina face ence between the net world price and the producer price) or the challenge of meeting the quality requirements indirect (through appreciation of the real exchange rate). This of the expanding confectionery markets. This is taxation was generally mitigated by input subsidies and border particularly so for African countries, which are protection. 11. Eliminating U.S. distortions under existing trade distor- missing out on the opportunities and rewards tions produces a 0.13 percent increase in the world price of Groundnut Policies, Global Trade Dynamics, and the Impact of Trade Liberalization 235 groundnuts and virtually no increase in world cake and oil FAO (Food and Agriculture Organization of the United prices. U.S. groundnut exports decrease by 10 percent, or about Nations) 2002. FAO Yearbook. Trade. Rome. 20,000 metric tons a year. Thus in contrast to the effect of U.S. Fletcher, S. M. 2001."Peanuts: Responding to Opportunities and subsidies on rice, cotton, and sugar, the impact of the current Challenges from an Intertwined Trade and Domestic Poli- U.S. farm program on the world price of and trade in ground- cies." University of Georgia, National Center for Peanut nuts is negligible. Competitiveness, Griffin, Ga. Fletcher, S. M., and C. L. Revoredo. 2003."Does the United States Need the Groundnut Tariff Rate Quota Under the 2002 U.S. References Farm Act?" Paper presented at the International Conference on Agricultural Policy and the WTO: Where Are We Head- Akobundu E. 1998. Farm-Household Analysis of Policies Affecting ing, June 23­26, Capri, Italy. Groundnut Production in Senegal. MS Thesis, Virginia Poly- Hart, C. and B. Babcock. 2002. "U.S. Farm Policy and the WTO: technic and State University. How Do They Match Up?" The Estey Centre Journal of Inter- Badiane, O., and Kinteh, S. 1994. "Trade Pessimism and Region- national Law and Trade Policy 3:119­39. alism in African Countries: The Case of Groundnut Pursell, G. 2003. Private e-mail correspondence on India's Exporters." IFPRI Research Report 97. International Food groundnut protection, September. Policy Research Institute, Washington, D.C. Revoredo, C. L., and S. Fletcher. 2002a. "The U.S. 2002 Farm Act Beghin, J., N. Diop, H. Matthey, and M. Sewadeh. 2003. "The and the Effects on U.S. Groundnut Exports." University of Impact of Groundnut Trade Liberalization: Implication for Georgia, Griffin, Ga. the Doha Round." Presented at the 2003 American Agricul- ------. 2002b. "World Peanut Market: An Overview of the Last tural Economics Association Annual Meeting, July 27­30, 30 Years." University of Georgia, Griffin, Ga. Montreal. Skinner, R. 1999."Issues Facing the United States: Peanut Indus- Chen, Changping, Stanley M. Fletcher, Ping Zhang, and Dale H. try during the Seattle Round of the World Trade Organiza- Carley. 1997. "Competitiveness in Peanuts: United States tion." U.S. Department of Agriculture, Economic Research versus China." Research Bulletin 430. University of Georgia, Service, Washington, D.C. Georgia Agricultural Experiment Station. USDA FAS (U.S. Department of Agriculture, Foreign Agricul- Colby, W. H., F. W. Crook, and S. Webb. 1992. Agricultural Statis- ture Service). GAIN Reports, various years a. "China, Peo- tics of the People's Republic of China, 1949­1990. SB-844, ple's Republic of. Oilseeds and Products. Oilseeds Update." United States Department of Agriculture, Economic Washington, D.C. Research Service, December. ------. Various years b. "China, People's Republic of. Oilseeds Diop, N., J. Beghin, and M. Sewadeh. 2004."Groundnut Policies, and Products. Annual." Washington, D.C. Global Trade Dynamics and the Impact of Trade Liberaliza- Wescott, P. C., C. E. Young, and J. M. Price. 2002. "The 2002 tion," World Bank Policy Research Working Paper 322. Farm Act Provisions and Implications for Commodity Mar- Washington, D.C. kets." Agriculture Information Bulletin Number 778. U.S. Gulati, A., G. Pursell, and K. Mullen. 2003. "Indian Agriculture Department of Agriculture, Economic Research Service, since the Reforms: Performance, Policy Environment, and Washington, D.C. Incentives." World Bank, Washington, D.C. Fang C., and J. Beghin. 2002. "Urban Demand for Edible Oils and Fats in China: Evidence from Household Survey Data." Journal of Comparative Economics 30(4): 732­53. 13 FRUITS AND VEGETABLES: GLOBAL TR ADE AND COMPETITION IN FRESH AND PROCESSED PRODUCT MARKETS Ndiame Diop and Steven M. Jaffee Trade in fruit and vegetable products has been exports of sugar, and seven times exports of textile among the most dynamic areas of international fibers. agricultural trade, stimulated by rising incomes and This chapter highlights major global, regional, growing consumer interest in product variety, fresh- and product-specific trends in the trade in fruit and ness, convenience, and year-round availability. vegetable products, and examines the major policy Advances in production, postharvest handling, pro- and other factors that have affected this trade over cessing, and logistical technologies--along with the past two decades.1 Particular attention is given increased levels of international investment--have to the performance and position of developing played a facilitating role. For developing countries, countries in this trade and the policies, institutions, trade in these products has been attractive in the and infrastructure they need to succeed. face of highly volatile or declining long-term trends in the prices for many traditional export products. Fruit and Vegetable Production Although many developing-country suppliers have and Trade Growth entered the field, relatively few have achieved signif- icant, sustained success, reflecting the fact that the For the purpose of this study, we group fruits and industry is highly competitive and rapidly changing. vegetables in four main categories: fresh fruits, Still, the aggregate picture is favorable. Fresh and fresh vegetables, processed fruits, and processed processed fruit and vegetable products accounted for vegetables. These categories comprise all SITC 16.7 percent of total agricultural exports from devel- (Standard International Trade Classification) Revi- oping countries in 1980­81. By 2000­01, this share sion 1, Chapter 5 items except nuts, roots, and had increased to 21.8 percent. Only for one other tubers. They correspond to most products in product category--fish and fisheries products--are Chapter 7 (edible vegetables and certain roots developing countries more significant exporters (see and tubers), Chapter 8 (edible fruits and nuts; chapter 1). Fruit and vegetable exports from devel- peel of citrus fruits or melons) and Chapter 20 oping countries are now more than double exports (preparations of vegetables, fruit, nuts, or other for tropical beverages, three times exports of grains, parts of plants) of the Harmonized System (HS) three times exports of livestock products, five times nomenclature. 237 238 Global Agricultural Trade and Developing Countries Trends in World Production and Trade followed even more robust growth in the 1980s, when trade in fruits and vegetables doubled. World World production of fruit and vegetables grew by trade in all categories of fruits and vegetables 30 percent between 1980 and 1990 and by 56 per- has grown strongly, with only slight changes in cent between 1990 and 2003, reaching 1,274 mil- its broad composition. In 2001 fresh produce lion tons by 2003. Much of this growth occurred in accounted for 63 percent of the total, whereas China, where production grew by 134 percent in processed products accounted for 37 percent. The the 1980s and by 200 percent in the 1990s. China is complexity of these definitions must be kept in currently the world's largest producer of fruits and mind, however. Both in Europe and the United vegetables, with a share of 34 percent, followed by States, one of the fastest-growing product segments Latin America and the Caribbean (11 percent), is semi-prepared and packed fresh produce, includ- India (10 percent), and Africa and the European ing preassembled salads, vegetable dips, and sliced Union (EU) (both at 9 percent) (figure 13.1). or mixed fruit products. The structure of world trade in fruits and veg- Taking all fruit and vegetable products com- etables does not fully mirror that of production. bined, the value of world imports grew at 2­3 per- Many of the largest producers are not significant cent a year during the 1990s, a sharp deceleration traders due to a combination of domestic demand from the 7­8 percent annual growth during the and geographical and logistical factors. For exam- previous decade (figures 13.2­13.5). As elaborated ple, in China and India, where strong domestic below, the slower growth in world imports during demand is fueled by growing income and a large the 1990s reflects two primary factors: a decline in and rapidly growing urban population, only a small world prices for many important fruit and veg- percentage of fruit and vegetable production is etable products in the latter half of the 1990s, and exported. In contrast, Latin American countries stagnation in EU import demand due to market (such as Mexico, Chile, and Costa Rica) are among saturation. the world's leading exporters of fruits and vegeta- Within the fresh fruit category, the deceleration bles, mainly because of their proximity to the large has been sharpest for apples, grapes, and citrus (fig- U.S. market. ure 13.2). Comparatively more dynamic trade has World trade in fruits and vegetables, fresh and remained for various tropical fruits (especially processed, has increased by 30 percent since 1990, papaya, mango, and pineapple), with average reaching $71.6 billion in 2001 (table 13.1). This annual growth in the 1990s remaining at 8 percent. FIGURE 13.1 Production of Fruit and Vegetables by Region (million of tons) Millions of tons 500 450 400 350 300 250 200 150 100 50 0 Africa EU-15 LAC China India U.S. Rest of World 1980 1990 2003 Source: United Nations Food and Agriculture Organization. Fruits and Vegetables: Global Trade and Competition in Fresh and Processed Product Markets 239 TABLE 13.1 World Fruit and Vegetable Imports, 1980­2001 (US$ millions) Percent Share Percent Share Percent Share Category 1980­81 in Total 1990­91 in Total 2000­01 in Total Fresh fruits 10,971 40 20,981 38 27,978 39 Processed fruits 4,441 16 9,916 18 13,176 18 Fresh vegetables 6,805 25 13,315 24 16,914 24 Processed vegetables 5,424 20 10,883 20 13,577 19 Total 27,641 100 55,094 100 71,644 100 Source: COMTRADE. FIGURE 13.2 Annual Growth Rates of World Imports of Selected Fruits (Value) Percent 12 10 8 6 4 2 0 Fresh Citrus Bananas Apples Grapes Other Tropical fruits total fruit fruit 1980­81 to 1990­91 1990­91 to 2000­01 Source: COMTRADE. Still, these latter products represent a relatively is tomatoes, which account for 17 percent of the small proportion of world fresh fruit trade (7 per- total. The category of beans, peas, and lentils cent in 2001), which is still heavily concentrated in accounts for another 14 percent. Other relatively particular lines, including bananas (25 percent of major commodities in the fresh vegetable trade the total), citrus fruit (20 percent), grapes (11 per- include onions, potatoes, asparagus, mushrooms, cent), and apples (10 percent). A large number of and various types of sweet and pungent peppers. other fresh fruits, not separated in the statistics, col- The evolution of trade in processed fruit and lectively represent 28 percent of world fresh fruit vegetable products mirrors that for fresh produce. imports. Prominent items in this category include The annual growth rate in trade value was 8.3 per- melons, various types of berries, and other temper- cent a year during the 1980s, yet only 3 percent ate fruits. during the 1990s (figure 13.4). All categories of World import values for fresh vegetables grew at processed products saw a deceleration in trade 6.9 percent a year during the 1980s, yet decelerated expansion, although fruit and vegetable juices and to 2.4 percent a year in the 1990s (figure 13.3). The preserved fruits and jams fell most sharply. deceleration affected most individual commodities. Processed vegetables (such as canned mushrooms, World vegetable trade is fragmented among a large dried mushrooms, and tomato paste) account for number of individual items. The largest single item 55 percent of world trade in all these products, fruit 240 Global Agricultural Trade and Developing Countries FIGURE 13.3 Annual World Import Growth Rates of Selected Vegetables Percent 7 6 5 4 3 2 1 0 Fresh Other Beans, Tomatoes vegetables total vegetables peas, lentilsa 1980­81 to 1990­91 1990­91 to 2000­01 a. The corresponding value for 1990­91 to 2000­01 has the value of 0%. Source: COMTRADE. FIGURE 13.4 Annual Growth Rates of World Import of Major Processed Fruits and Vegetables Percent 10 9 8 7 6 5 4 3 2 1 0 Total Dried fruits Preserved Fruit juices, Processed processed fruits, jams, vegetable vegetables fruits and and flour juices vegetables 1980­81 to 1990­91 1990­91 to 2000­01 Source: COMTRADE. and vegetable juices for some 20 percent, and sev- fruit and vegetable imports during the 1990s. eral smaller categories for the balance. Indeed, for each of the most important traded fresh Price played a role in the observed trends. The fruits and vegetables, the rate of import volume unit values of fresh fruit and of fresh and prepared growth was modestly higher in the 1990s than in vegetables dropped sharply in the second half of the 1980s (table 13.2). the 1990s after an extended period of increase dat- Part of the decline in world prices is a statistical ing from the early 1980s (figure 13.5). These trends matter. The data above are recorded in U.S. dollars. suggest that price factors played a very significant During the latter half of the 1990s, the U.S. dollar role in the declining rate of growth in the value of appreciated vis-à-vis the yen and most European Fruits and Vegetables: Global Trade and Competition in Fresh and Processed Product Markets 241 FIGURE 13.5 World Unit Values for Fresh Fruits, Fresh Vegetables, and Prepared Vegetables (US$ per metric ton) US$ 1,400 1,200 1,000 800 600 400 200 0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 Prepared vegetables Fresh fruit Fresh vegetables Source: FAOSTAT. TABLE 13.2 Average Annual Growth Rates leading importer); competitive and structural in World Import Volumes, changes in fruit and vegetable distribution systems, 1980­2001 which put downward pressures on trader and manufacturer margins; and greater availability of Commodity 1980­1991 1990­2001 product and intensified international competition. Bananas 3.45 3.96 Oranges 1.11 1.68 Sources and Destinations of Apples 1.42 2.33 Exported Fruits and Vegetables Grapes 4.31 4.34 Tomatoes 3.11 4.45 The European Union, NAFTA (North American Onions 3.50 4.17 Free Trade Agreement), and a few middle-income Green beans 4.59 5.98 countries dominate trade in fruits and vegetables (table 13.3). Eight categories of countries are dis- Source: FAOSTAT. tinguished. China, India, Japan, and the European Union are singled out. The United States, Canada, currencies, deflating Japanese and European and Mexico are grouped together in the NAFTA import values upon conversion into U.S. dollars. category. The developing-country group minus For some commodities, however, the unit import China and India is split between low-income coun- values into Japan, Europe, and elsewhere actually tries and middle-income ones. did decline in local currency terms--in some cases substantially. For example, from 1996 to 2001, the Global Trade Patterns average unit value of Japanese fresh vegetable imports fell by 25 percent, while that of processed The European Union is the world's largest market vegetable imports fell by 8 percent.2 A major factor and supplier of fresh and processed fruits and veg- in this decline was the rapidly expanding supply of etables. In 2001 its 15 member countries accounted low-cost production in China. During the 1990s for $37 billion in imports, or 51 percent of world China accounted for virtually all of the incremental imports, while exports stood at $28 billion, or expansion in Japan's vegetable trade, taking market 39.5 percent of world exports.3 EU trade in fruits share from other suppliers. Declining unit import and vegetables is, however, largely intraregional. values for various products in Europe can be attrib- Intra-EU imports represent 64 percent of EU uted to at least three factors: the slow economic imports, while 83 percent of EU export trade occurs growth in the region (especially in Germany, the among its 15 member states. Still, with its affluent 242 Global Agricultural Trade and Developing Countries TABLE 13.3 Import and Export of Fruits and Vegetables by Region or Country (US$ billions) Exporters Low- Middle- European Other Total Importers Year Income Income Union Japan NAFTA Industrial China India Imports Low-income 1980­81 0.05 0.14 0.09 0.00 0.03 0.01 0.02 0.02 0.37 1990­91 0.02 0.12 0.05 0.00 0.02 0.05 0.04 0.04 0.34 2000­01 0.08 0.19 0.10 0.00 0.13 0.09 0.14 0.05 0.78 Middle- 1980­81 0.10 2.24 0.88 0.18 0.84 0.21 0.39 0.09 4.94 income 1990­91 0.14 2.77 0.91 0.10 1.12 0.33 0.79 0.12 6.27 2000­01 0.44 4.69 2.22 0.05 1.88 0.58 1.04 0.17 11.08 European 1980­81 0.42 4.36 8.63 0.02 0.75 1.08 0.24 0.02 15.52 Union 1990­91 0.75 8.86 20.59 0.01 1.07 1.71 0.50 0.02 33.51 2000­01 1.08 9.19 23.45 0.01 1.19 1.09 0.55 0.11 36.67 Japan 1980­81 0.02 0.68 0.05 0.00 0.49 0.07 0.13 0.00 1.43 1990­91 0.01 1.49 0.17 0.00 1.21 0.25 0.56 0.00 3.69 2000­01 0.04 1.49 0.25 0.00 1.69 0.36 1.93 0.00 5.76 NAFTA 1980­81 0.04 1.64 0.29 0.05 1.97 0.09 0.05 0.00 4.12 1990­91 0.07 3.94 0.75 0.06 3.59 0.23 0.13 0.01 8.80 2000­01 0.12 5.20 1.15 0.05 7.10 0.27 0.36 0.10 14.35 Other 1980­81 0.01 0.31 0.61 0.00 0.13 0.11 0.02 0.00 1.20 industrial 1990­91 0.03 0.55 1.16 0.00 0.17 0.21 0.04 0.01 2.17 2000­01 0.03 0.56 1.10 0.00 0.23 0.21 0.08 0.01 2.22 China 1980­81 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.01 1990­91 0.02 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.06 2000­01 0.06 0.32 0.01 0.00 0.12 0.03 0.00 0.01 0.56 India 1980­81 0.02 0.04 0.00 0.00 0.00 0.00 0.00 0.00 0.06 1990­91 0.08 0.10 0.00 0.00 0.02 0.04 0.00 0.00 0.24 2000­01 0.10 0.03 0.00 0.00 0.08 0.02 0.00 0.00 0.23 Total 1980­81 0.65 9.41 10.55 0.25 4.21 1.58 0.85 0.13 27.64 exports 1990­91 1.12 17.86 23.62 0.18 7.22 2.83 2.06 0.20 55.09 2000­01 1.95 21.66 28.29 0.12 12.41 2.65 4.10 0.45 71.64 Source: COMTRADE. and aging population, its high factor costs, and its market. Trade between Mexico, Canada, and the cold winters, this region represents one of the largest United States accounted for 49 percent of NAFTA's fruit and vegetable markets for non-EU countries imports and 53 percent of its exports in 2001. ($13.2 billion), especially for low- and middle- Intra-NAFTA trade is most important for fresh income countries, which exported $1.08 and $9 bil- vegetables. For this commodity group, 90 percent lion, respectively, to the European Union in 2001. of exports and 86 percent of imports occur within Major middle-income suppliers to the European the trade group (Huang 2004). Still, middle- Union market include banana-exporting coun- income countries (excluding Mexico) have a strong tries (mainly Colombia, Costa Rica, Côte d'Ivoire, foothold in this market. By securing 71 percent of Ecuador, and Panama4) and counterseasonal- the $7.25 billion extra-NAFTA import market in supplying countries such as Argentina, Chile, and 2001, middle-income countries are major players. South Africa. Led by South Africa, the latter three Interestingly, thanks to growing incomes in the countries dominate exports of apples, grapes, and 1990s, middle-income countries have seen their pears to the European Union. own market become a major destination of fruit and Intraregional trade is also significant in NAFTA, vegetable exports from other countries, with import the world's second-largest fruit and vegetable demand totaling $11 billion in 2001. South-South Fruits and Vegetables: Global Trade and Competition in Fresh and Processed Product Markets 243 trade--trade between developing countries, exclud- 1.4 percent annually over the 1990s, reflecting its ing China and India--totaled $5.4 billion in 2001, slow pace of economic growth during this decade.5 accounting for 45 percent of developing countries' Income and population composition and imports. Japan has also emerged as a significant dynamics are the major drivers of import demand. market for fruits and vegetables over the 1990s, with Demand for fruits and vegetables--derived from a import demand culminating at $5.8 billion in 2001. combination of broad demand dynamics, domestic China has emerged as a major supplier of horticul- supply trends, and trade policies--is relatively ture to Japan with its market share doubling from income elastic. Higher incomes typically induce 16 percent in the 1980s to 33 percent in the 1990s. increased expenditures on a broader array of fresh While import penetration increased in the and processed fruit and vegetable products. In United States and other regions, EU import addition to income, other important factors demand grew little in the 1990s.As table 13.3 shows, include the size, age, ethnic composition of the a salient feature of world import dynamics is the population, cultural and religious factors, lifestyle sharp increase in imports of developing countries factors (including work patterns and urbanization), (6 percent a year) and NAFTA (5 percent a year), and consumer education about health matters. and the stagnation of EU imports (1 percent a year) Although not all of these factors can be exam- over the 1990s. Focusing on the United States, the ined statistically, we attempt here to quantify the import penetration ratio has increased steadily over importance of most of the factors that explain time for fresh produce (figure 13.6). In contrast, the observed cross-country differences in growth extra-EU import demand grew by just 0.2 percent a in imports of fresh fruits, fresh vegetables, and year between 1990 and 2001, indicating that the processed fruits over the 1990s. The analysis, based bulk of the small increase in the European Union's on a sample of 49 major importers, uses the stan- horticultural trade, shown in table 13.3, occurred dard imperfect substitutes model (Goldstein and internally. This is a major change from the 1980s, Khan 1985), which assumes that imports are not when EU imports almost doubled. Closer examina- perfect substitutes for domestic goods for the coun- tion of the data shows that the European Union's tries under consideration.6 import deceleration is largely driven by Germany, It is well known that economic growth strongly which represents 25 percent of the EU's fruit and stimulates imports of fruits and vegetables, whereas vegetable market. Germany's imports dropped by inflation reduces them (Goldstein and Khan 1985).7 FIGURE 13.6 Import Penetration Ratios in U.S. Fruit and Vegetable Markets, 1970­2001 (percent) Percent 50 45 40 35 30 25 20 15 10 5 0 1970 1975 1980 1985 1990 1995 1996 1997 1998 1999 2000 2001 Fresh fruit Fruit juice Fresh vegetables Source: USDA. 244 Global Agricultural Trade and Developing Countries TABLE 13.4 Concentration of Fresh Fruit and Vegetable Exports among Developing Countries, 2001 Joint Percentage of Product Leading Suppliers World Exports (value) Asparagus Peru, Mexico, Thailand 94 Mangoes Brazil, Mexico, Philippines 62 Pineapples Costa Rica, Côte d'Ivoire 61 Bananas Ecuador, Colombia, Costa Rica 60 Avocados Chile, Mexico 53 Tomatoes Mexico, Syria 52 Grapes Chile, China, Mexico 38 Green beans Jordan, Kenya, Mexico 49 Green peas Guatemala, Kenya, Zimbabwe 38 Source: FAOSTAT. But it is urbanization, not growth per se, that exer- middle-income countries have succeeded on a sus- cises the strongest and most significant influence on tained basis. The average shares of developing imports of fresh fruits (with the nuance that it plays countries in world exports of fresh fruits and veg- a major role in developing countries only given the etables hide the heavy domination of trade by just a population shifts that are still occurring in these handful of middle-income countries. Between 1997 countries, see the Result Table A1 on the CD-ROM). and 2001, just four Latin American countries-- Several factors could explain this relationship. First, Chile, Costa Rica, Ecuador, and Mexico-- urbanization helps to reduce the logistical and accounted for 43 percent of developing-country transaction costs to service demand from interna- exports of fresh fruit (FAO 2003). These countries tional sources, especially if the major cities are are leading players in the most internationally located in close vicinity to major ports or along effi- traded fruit products (table 13.4). While exports of cient transport nodes. Second, with urbanization, vegetables are similarly concentrated, the geo- there is greater demand for convenience in meeting graphical distribution of exporters is wider. Mexico food needs. Many fresh and processed fruit and veg- is the world's leading exporter of tomatoes, Kenya etable products can be consumed with little or no supplies 25 percent of the world's green beans, further household preparation. Plus, these products while Guatemala and Kenya jointly lead the world feature heavily in menus of restaurants and catering market for green peas.8 Between 1997 and 2001, services, most of which are in urban areas. Third, four suppliers--Argentina, China, Mexico, and the urban populations tend to be more heterogeneous Syrian Arab Republic--accounted for 67 percent of in ethnic and other composition than is typical in fresh vegetable exports by developing countries any single rural setting. This mixing of populations (FAO 2003).9 increases consumer exposure to new or even exotic A small number of medium-income countries products, some of which may only be sourced in have been successful in the processed segment of large quantities from abroad. the export market, but as a group, developing As expected, tariffs negatively affect fresh fruit countries account for a relatively low share in world trade, but the elasticity is not significantly different exports of these products (36 percent in 2001). from zero. In sharp contrast, tariffs have a negative Chile, China, Thailand, and Turkey account for and statistically significant effect on processed fruit 58 percent of developing countries' exports of trade, highlighting the high degree of tariff escala- processed fruit and vegetable products (FAO 2003). tion affecting trade in fruits and vegetables. Secondary, yet still significant exporters include Argentina, Indonesia, Mexico, and the Philippines (a combined 14 percent of developing countries' Developing Countries' Performance exports). Trade by developing countries in specific Although many countries have entered the fresh processed products is relatively highly concentrated fruit and vegetable export markets, only a few (table 13.5). Fruits and Vegetables: Global Trade and Competition in Fresh and Processed Product Markets 245 TABLE 13.5 Concentration of Processed Fruit and Vegetable Exports among Developing Countries Joint Percentage of Product Leading Suppliers World Exports (value) Orange juice concentrated Brazil 91 Canned pineapples Indonesia, Philippines, Thailand 74 Canned mushrooms China 52 Dried mushrooms China 52 Dried fruits Chile, China, Thailand 35 Tomato paste Chile, China, Turkey 35 Apple juice, concentrated Argentina, Chile, Turkey 31 Source: FAO 2003. TABLE 13.6 Export Subsidy Expenditures for Horticultural Products Export Total Horticultural Export Expenditures as a Expenditure Exports Percent of Total Horticultural Country (US$ millions) (US$ millions) Export Value European Union 40 5,301 0.8 Switzerland 14 69 20.6 Colombia 13 521 2.4 Turkey 11 2,348 0.0 Source: WTO 2000b and subsequent WTO notification updates. Export data from FAOSTAT. Policy Factors Shaping International producers at least the minimum price fixed each Trade Patterns year (for dried figs and certain prunes, for exam- ple). Japan reported price support for starches Domestic Support and Export Subsidies ($179 million for potatoes and sweet potatoes in Government interventions for fruits and vegetables 1998) and direct payments to the vegetable and are significantly lower than in other agricultural fruit sectors, but the aggregate measure of support sectors. Consistently, domestic subsidies to produc- was below the de minimis level. Similarly, in ers are relatively low in OECD (Organisation for Canada, only its support for dry beans was above Economic Co-operation and Development) coun- the de minimis levels (WTO 2000a). tries. Of the major industrialized regions, only the Unlike in many other agricultural sectors, the European Union reports an aggregate measure of use of export subsidies is not pervasive in horticul- support related specifically to several fruits and ture. The export subsidy expenditures notified to vegetables, while Japan and Canada indicate mod- the WTO in 2000 (WTO 2000b) (table 13.6) are erate levels of aggregate support for a few com- well below those reported for other agricultural modities.10 The European Union's budgetary out- categories. In only one country, Switzerland, were lays for fruits and vegetables totaled $1.55 billion in expenditures large relative to horticultural exports, 1999. Those expenditures covered compensation with export subsidies accounting for 21 percent of for surplus withdrawals; production aid to produc- that country's exports. The European Union's ers of bananas, peas, lentils, beans, pineapples, export subsidies represented less than 1 percent of grapes, and stone fruits; and aid to producers of the value of its total exports. Although it did not certain products intended for processing (toma- supply information to the WTO, the United States toes, peaches, pears) and to processors who pay indirectly subsidized horticultural exports, albeit 246 Global Agricultural Trade and Developing Countries to a very limited degree, through export credit opacity of protection. Tariff peaks in the European guarantees. Union, for example, can reach as high as 128 percent for fresh fruits and 132 percent for fresh vegetables. Viewed in closer detail, the protection structure Tariffs and Other Import Restrictions of several OECD countries is opaque. Canada, Regulating market access is the main instrument Japan, and the United States have the lowest tariffs, used to protect the fruit and vegetable sector. The with, for instance, 85 percent of U.S. tariffs under European Union, Japan, and the United States use, 10 percent. In sharp contrast, the Republic of Korea to varying degrees, similar protection tools: low but and the European Union apply high tariffs on many highly dispersed ad valorem tariffs, specific duties, products. For instance, in the Republic of Korea, seasonal tariffs, tariff escalation, and preferential 59 percent of fresh fruit and vegetable tariff lines lie access along with tariff-rate quotas. Tariffs for a between 20 and 50 percent and 37 percent of the specific range of products depend on numerous lines are over 100 percent. Protection of fruits factors, including the date of entry (seasonality fac- remains relatively nontransparent, as well, especially tor), the degree of processing (escalation phenome- in Canada and the European Union (table 13.8). non), and the relationships with exporting coun- The percentage of fresh fruit tariff lines that are spe- tries (preferential agreements and regional and cific, compound, or mixed stands at 31 percent in bilateral free trade agreements--FTAs). Canada and 25 percent in the European Union. Average applied most-favored-nation (MFN) Fresh vegetable exports face, in general, higher tariffs are very low in all countries of the Quad-- levels of protection, reflecting the fact that tariffs on Canada, the European Union, Japan, and the temperate horticultural commodities are higher United States. These tariffs range between 0.9 per- than they are for tropical commodities, dominated cent for fresh fruits in Canada to 9.2 percent in the by fruits. The EU tariffs are particularly high for European Union for the same product category many commodities, as 60 percent of vegetable tariff (table 13.7). The average tariffs, however, do not lines are in the 20­50 percent range and 23 percent accurately reflect the level of actual protection of the latter are greater than 50 percent. This caused by the wide dispersion in tariffs and the reflects the large number of items that are subject prevalence of high peaks. Closer examination of the to ad valorem tariffs (including seasonal) aug- tariff structure highlights the importance of tariff mented by specific tariffs under the European peaks, especially in the European Union and the Union's minimum entry price scheme (box 13.1). TABLE 13.7 Applied MFN Tariffs for Fresh Fruit and Vegetables in the Quad Countries, 1999 and 2001 (percent) Average ad Percentage of Number of Valorem All Rates Lines Standard Total Country Rate Average Covered Maximum Deviation Lines European Fruits 7.3 9.2 75.0 127.6 15.4 89 Union Vegetables 5.5 6.8 98.0 131.8 10.7 200 (1999) United States Fruits 6.1 4.6 100.0 29.8 7.0 70 (2001) Vegetables 4.1 3.1 98.0 24.3 5.0 189 Japan (2001) Fruits 8.7 8.7 100.0 32.0 6.8 56 Vegetables 3.9 3.9 94.0 40.0 5.6 185 Canada Fruits 0.9 0.9 72.0 8.5 2.5 67 (2001) Vegetables 1.1 1.1 74.0 16.0 2.7 216 Source: WTO Integrated Data Base at the original tariff line level (6- to 11-digit tariff line depending on the country). Fruits and Vegetables: Global Trade and Competition in Fresh and Processed Product Markets 247 TABLE 13.8 Percentage of Tariff Lines on Fresh Fruits and Vegetables in Selected OECD Countries by Tariff Levels Tariff Levels Canada United States European Union Japan Korea, Rep. of (percent) (2001) (2001) (1999) (2001) (2001) Fresh fruits Duty free 60 19 2 7 0 1­10 9 66 49 64 0 11­20 0 9 22 20 0 21­50 0 6 0 9 59 Over 50 0 0 1 0 37 Specific, compound, 31 1 25 0 5 mixed Fresh vegetables Duty free 47 16 0 10 3 1­10 7 59 11 69 42 11­20 1 14 1 11 47 21­50 0 1 60 7 1 Over 50 0 0 23 0 1 Specific, compound, 45 9 5 2 5 mixed Note: Average applied out-of-quota ad valorem and ad valorem equivalent of non-ad-valorem tariffs for those equivalent data reported. Source: WTO IDB database. BOX 13.1 The European Union's Entry Price Scheme: Hindering Cost-Based Competition in the EU Market Current basic rules governing trade in fruits and specific tariff (called the maximum tariff equiva- vegetables were defined as part of the European lent) is levied against the shipment, most likely Union's 1996 Common Market Organization prohibiting its importation. If the entry price (CMO) reform. One of the most prominent fea- stands between 92 and 100 percent of the entry tures of this reform is the "minimum entry price" price, an additional specific duty is levied. system. This complex tariffication system applies Through this system, applied tariffs are actu- to imports of a large number of fruits and veg- ally linked to the delivered price and the season. etables, including fresh or chilled tomatoes, For instance, fresh tomatoes imported between courgettes, cucumbers, apples, grapes, pears, June 1 and October 30 and priced 8 percent peaches, plums, apricots, cherries, and citrus below the reference price of 52.6 per100 kilo- fruits. Under the system, the European Union cal- grams face tariffs amounting to 57 percent of culates an entry price for each of the commodi- the import price (Sallyards 2001). The entry ties covered by the program. The tariffs levied for prices are generally highest during the EU pro- each item depend on its import price compared duction season and lowest during the off- with the calculated price.11 Fruits and vegetables season, and the difference can be very large. The imported at prices equal to or greater than the entry price for courgettes, for instance, increases established entry price are charged an ad valorem from a base level of 450 a metric ton to 730 a duty only. Commodities valued below the entry metric ton in April and May. This system strongly price are charged a specific tariff in addition to restricts an exporter's ability to increase market the ad valorem duty. In the latter case, two situa- shares in the European Union based on lower tions are distinguished: if the import price is more prices and efficiency, especially during the than 8 percent below the entry price, a large European production season. 248 Global Agricultural Trade and Developing Countries TABLE 13.9 Percentage of Tariff Lines on Processed Fruits and Vegetables in Selected OECD Countries by Tariff Levels Tariff Levels Canada United States European Union Japan Korea, Rep. of (percent) (2001) (2001) (1999) (2001) (2001) Processed fruits Duty free 35 1 0 7 1 1­10 43 14 0 62 2 11­20 20 43 0 24 44 21­50 0 34 47 4 40 Over 50 0 0 53 4 2 Specific, compound, 2 7 0 0 10 mixed lines Processed vegetables Duty free 22 3 0 8 3 1­10 31 39 6 65 6 11­20 40 50 0 24 76 21­50 0 6 88 3 5 Over 50 0 0 3 0 3 Specific, compound, 7 2 3 0 8 mixed lines Note: Average applied out-of-quota ad valorem and ad valorem equivalents of non-ad-valorem tariffs for those equivalent data reported. Source: WTO IDB database. The tariff structures of the European Union, countries analyzed above. Ninety-four percent of Japan, the Republic of Korea, and the United States MFN tariff lines for fresh fruits in India are also feature a high degree of escalation. All EU between 21 and 50 percent, while all "MFN" fruits processed fruit tariffs are above 20 percent, and the entering Morocco face a tariff of more than 50 per- majority of processed fruit products entering the cent. (In contrast, Indonesia and South Africa have European Union face a tariff of greater than 50 per- tariff structures similar to those of Japan and the cent (table 13.9). There is also tariff escalation in the United States, with most tariff lines falling between European Union for processed vegetables, with zero and 10 percent.) The potential hindrance of 88 percent of these products facing a tariff in the these high tariffs to developing-country exports range of 21­50 percent. The European Union's esca- should not be underestimated. As seen above, lating tariffs for tomato and apple-based products South-South trade in fruits and vegetables is grow- inhibit a potentially large level of trade by nonmem- ing rapidly and now represents about one-fifth of ber countries. Tariffs facing most processed fruit developing countries' exports. and vegetable products entering Canada, Japan, Korea, and the United States are below 20 percent. Preferential Market Access and Magnitude These low tariffs do not exclude the use of high of Preference levels of protection for particular products, as illus- trated by U.S. protection of its own orange juice The protection structure just described does not industry (box 13.2). apply equally to all exporting countries. Many In most middle-income countries, the tariff high-income countries maintain a complex system structure is more transparent than in the Quad, but of preferential access (that is, better-than-MFN average tariff levels are higher (table 13.10), posing access) designed to provide privileged partners a challenge to would-be external suppliers. Average with favorable entry without undermining the applied MFN tariffs in Brazil, India, and Morocco, protection of domestic producers. The product for example, are far higher than in the high-income coverage of preferential access schemes is wide, but Fruits and Vegetables: Global Trade and Competition in Fresh and Processed Product Markets 249 BOX 13.2 The U.S.-Brazilian Trade Dispute over Orange Juice The world market for orange juice is basically a combined tax and duty accounts for nearly duopolistic market structure, with only two play- 50 percent of the cost of a ton of Brazilian con- ers, the United States (mainly Florida) and Brazil, centrate. This discrimination between imported supplying roughly 85 percent of the world mar- and domestic products has prompted Brazil to ket. Over 95 percent of Brazil's production is initiate a dispute settlement process at the WTO. exported, whereas more than 95 percent of U.S. In March 2002 the government of Brazil orange juice is consumed domestically. Of the requested bilateral consultations under WTO aus- U.S. imports of concentrated orange juice, some pices with the United States regarding the 90 percent comes from Brazil. Imported orange "equalizing excise tax." Brazil argued that the juice is mixed with U.S. juice to improve its color incidence of the tax on imported processed citrus and make up for seasonal supply shortfall. This products and not on domestic products consti- trade pattern reflects Brazil's production cost tutes a de facto violation of most-favored-nation advantages, which in turn mirror lower labor and national treatment provisions of GATT (Gen- costs in Brazil, reinforced in recent years by the eral Agreement on Tariffs and Trade) (Articles II: devaluation of the real. The United States levies a 1(a), III.1 and III:2, GATT 1994). tariff of 7.85 cents per liter on Brazilian orange Interestingly, the private sector has already juice. In addition, an antidumping order remains taken pragmatic actions to deal with these in effect, with dumping duties ranging from problems, through joint investments and joint 2 percent to 27 percent on imports of the production. In an increasingly common tariff- Brazilian product. jumping tactic, the Brazilian producers in the Furthermore, Brazilian exporters pay a "Florida early 1990s began to invest directly in the Florida equalizing excise tax" on frozen orange juice con- industry. It is estimated that foreign--mainly centrate, from which domestic producers in Brazilian--companies own as much as 40 per- Arizona, California, and Texas whose juice is also cent of the Florida processing industry. The U.S. blended with Florida orange juice are exempt. presence in Brazil's citrus industry started in the The proceeds of the tax are allocated by statute 1960s, when winter freezes prompted U.S. to the exclusive promotion of Florida-grown growers to seek out Brazil for planting. citrus products. According to one estimate, the Source: Thunder Lake Management 2002. entry is often limited by quotas for "sensitive prod- Trade Preference Act (ATPA), and free trade agree- ucts" such as those put in special protocols (such as ments with Israel and Chile. Since NAFTA was bananas). signed, tropical fruits shipped from Mexico (man- The major EU preferential access schemes rele- goes, guavas, avocados, and papayas) have been vant to trade in fruits and vegetables include the subject to steadily reduced tariffs. Since January Everything but Arms (EBA) initiative that benefits 2003 they have entered the U.S. market free of duty. the 48 UN-defined least-developed countries While these different agreements are not always (LDCs); the EU-ACP Lomé Conventions, under directly comparable, it is clear that they provide which the European Union grants unilateral prefer- varied degrees of preference to the suppliers ential access to 75 African, Caribbean, and Pacific involved. Figure 13.7, adapted from Stevens and (ACP) countries, bilateral agreements such as the Kennan (2000), highlights the hierarchy of prefer- Euro-Med Agreements between the European ences within the European Union's fruit and veg- Union and many Mediterranean countries, and the etable import regime as of 2003. The major changes EU-South Africa free trade area; and the general- since 2000 include the promulgation of the EBA ized system of preferences (GSP). initiative and the multiplication of bilateral agree- A large number of countries also enjoy preferen- ments, which erode the preferences of those on top tial access to the United States through formal of the pyramid. The ranking of preferences regional, bilateral, and preferential trade agree- depends on the difference between preferential ver- ments. These include NAFTA, the African Growth sus MFN tariffs (that is, the margin of preference), Opportunity Act (AGOA), the Caribbean Basin the breadth of product coverage, the extent of Economic Recovery Act (CBERA), the Andean quota limitations, and the degree of certainty of 250 Global Agricultural Trade and Developing Countries TABLE 13.10 Percentage of Tariff Lines at Different Levels in Selected Developing Countries, 2001 Tariff Levels (percent) Brazil India Indonesia Morocco South Africa Fresh fruits Duty free 0 0 0 0 15 1­10 0 0 100 0 47 11­20 100 3 0 0 29 21­50 0 94 0 0 9 Over 50 0 3 0 100 0 Specific, compound, 0 0 0 0 0 mixed lines Fresh vegetables Duty free 24 0 2 0 46 1­10 0 0 98 0 6 11­20 76 85 0 13 30 21­50 0 15 0 19 13 Over 50 0 0 0 68 0 Specific, compound, 0 0 0 0 6 mixed lines Note: Average applied out-of-quota ad valorem and ad valorem equivalents of non-ad-valorem tariffs for those equivalent data reported. Source: WTO IDB database. FIGURE 13.7 The Hierarchy of Preferences in the European Fruits and Vegetables Market EBA countries (1 percent of EU imports) ACP countries (8 percent of EU imports) Bilateral agreements GSP (14 percent of EU imports) MFN Source: Authors' calculations using COMTRADE data and based on Stevens and Kennan 2000. preferences. The key characteristics of the Euro- third-country imports. Just a few countries-- pean Union's preference system are: including Cameroon, Côte d'Ivoire, and Kenya-- account for the bulk of this trade. · Duty-free and quota-free access for LDCs under · ACP access for bananas is limited by quotas. the EBA initiative.12 The beneficiary countries ACP countries enjoy, within the allocated quota generally lack the capacity to provide reliable of 850,000 tons, duty-free access to the EU supplies, in part because of poor infrastructure banana market (until 2008), whereas third- and other behind-the-border constraints. Even country suppliers face a duty of 75/ton.13 with EBA, they accounted in 2002 for only 1 per- · Preferential access for many countries comes cent of the European Union's imports of fruits through bilateral agreements. The concessions and vegetables from outside the EU. granted under these agreements are typically · Generous access for ACP countries, account- restricted to certain tariff quotas or to certain ing for 8 percent of the European Union's periods of the year, depending on the EU season. Fruits and Vegetables: Global Trade and Competition in Fresh and Processed Product Markets 251 Tensions between opening markets for privi- market (Stevens and Kennan 2000). We examine leged partners and protecting domestic producers here the effectiveness of ACP and AGOA prefer- have led to a widespread use of tariff rate quotas ences for fruits and vegetables exported to the (TRQs) in fruit and vegetable trade. In 2000 devel- European Union and United States, respectively. oped countries applied 355 TRQ schemes to These indicators are preferred over changes in imported fruits and vegetables, compared with 56 MFN tariffs, which do not capture the variety of for tropical beverages and processed agricultural specific trade regimes in the European Union and products (Jabati 2003). Quotas are typically set at the United States that are relevant for many devel- low levels with low in-quota tariffs and prohibitive oping countries. over-the-quota tariffs. A good example is that of The value of an ACP preference can be defined winter seedless grapes, a product exported by some as the product of the value of exports for which Southern Hemisphere suppliers, including South preferences have been requested and the preferen- Africa and Namibia. Namibia may export only tial margin (figure 13.8). The value of the ACP pref- 900 tons per year to the European Union from erence in fruits and vegetables represented 12 per- November to end of January (Jabati 2003). Any cent of ACP country exports, with a great deal of over-the-quota export is subject to an import tariff variation around the average.14 The ACP "rent" of 16.4 percent. The tariff and period restrictions stood at less than 5 percent for Ethiopia, Madagas- clearly constitute a constraint for Namibia if it car, and Namibia, but between 28 and 42 percent wants to increase its exports. for major banana producers such as Cameroon, the Analysis of the value of ACP and AGOA prefer- Caribbean islands, and Côte d'Ivoire. ences show heterogeneous situations among ACP In terms of scope, about 82 percent of ACP fruit countries, while South Africa stands out as the only and vegetable exports to the European Union are country taking significant advantage of the AGOA eligible for preference. Use of this preference is preference. It has been argued that preferential quite high, with an average 75 percent of eligible treatment has contributed to the successful pene- exports requesting preference. Use rates vary widely tration of some developing countries into the EU across ACP countries, however. FIGURE 13.8 Value of Fruit and Vegetable Preference for Major ACP Exporters, 2002 St. Vincent St. Lucia Jamaica Dominica Dominican Rep. Belize Swaziland Namibia Zimbabwe Zambia Madagascar Kenya Ethiopia Cameroon Ghana Côte d'Ivoire Senegal 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 Value of ACP preference requested (percent of exports) Source: World Bank staff calculations based on the EU Commission Database. 252 Global Agricultural Trade and Developing Countries In sharp contrast with the ACP scheme, the use As more and more countries are enjoying better- of the AGOA preference to export fruits and vegeta- than-MFN access to the EU market, preferences are bles to the United States is not yet a widespread eroding and competition is stiffer--as illustrated by practice in Africa, since only 14 of 38 AGOA coun- the experience of large middle-income exporters tries exported fruits and vegetables to the United (figure 13.10). Morocco, South Africa, and Turkey States in 2002. Among those having done so, how- are large exporters of fruits and vegetables that enter ever, the use of the AGOA preference was high, the European Union under bilateral agreements averaging 73 percent (see annex table A3). The very with limitations on some products.15 Morocco, small number of countries that have exported fruits among the first countries to sign preferential agree- and vegetables under AGOA is associated with the ments with the European Union in the late 1960s, constrained logistics on African­U.S. trade in has lost ground in the EU market to Turkey since perishable products, the very limited degree of U.S. Turkey's free-trade agreement with the European private investment in Africa in this field, and Union was signed in 1998. South Africa recently stringent U.S. phytosanitary requirements. unseated Turkey in the EU market and is now the Not surprisingly, countries that have more largest third-country supplier of fruits and vegeta- advanced logistical systems and stronger inter- bles to the European Union (with a 31 percent national marketing ties (for example, South share). Turkey holds a slightly lower share (29 per- Africa) are better placed to benefit from AGOA. cent), while Morocco lags far behind (22 percent, In fact, South Africa is the only AGOA country down from 37 percent in 1980). In several products for which the preference "rent" represents a signifi- (citrus fruit, tomato products, dried fruit, and fruit cant share of export values (74 percent) (fig- juice), these suppliers have competed directly. ure 13.9). Few other African countries have a com- parative advantage in servicing the U.S. market and, Determinants of Success in Fruit and Vegetable Export Markets given comparative freight costs and availability, find the European market a more attractive outlet. We Many developing countries have sought to take analyze these issues further in the next section. advantage of emerging international markets for FIGURE 13.9 Value of Fruit and Vegetable Preference under AGOA Zambia Uganda Tanzania Swaziland South Africa Sierra Leone Nigeria Mozambique Malawi Madagascar Kenya Ghana Ethiopia Côte d'Ivoire 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 Value of AGOA preference requested (percent of exports) Source: World Bank staff calculations based on USITC Database. Fruits and Vegetables: Global Trade and Competition in Fresh and Processed Product Markets 253 FIGURE 13.10 EU­Third Country Imports and the Share of Key Exporters Percent 40 35 30 25 20 15 10 5 0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 Morocco South Africa Turkey Source: COMTRADE. fresh and processed fruit and vegetable products. some of the developing world's leading fresh and Yet, as noted above, relatively few have achieved processed fruit and vegetable success stories-- dominance in a range of such products. In Sub- among them Brazil, Chile, Kenya, Mexico, and Saharan Africa several dozen countries have Taiwan (China). In each case, the initial take-off participated in this trade, yet just three--Côte occurred during a period of stable macroeconomic d'Ivoire, Kenya, South Africa--accounted for conditions and the presence of a favorable invest- nearly 90 percent of the region's trade in recent ment climate. Important initial catalysts for export decades. Only a few other countries in the region growth included sudden shortfalls in major over- have been able to sustain growth in their horticul- seas markets, new foreign direct investment or tural trade over time; none has emerged as a major strategic partnerships, and improvements in inter- player in the international market. national logistics capacity. International technical and marketing partnerships provided a vehicle for Success Factors the transfer of technology, for new market penetra- To succeed in exporting fruits and vegetables, a tion, and for creating an identity for the products country must have important assets that provide an from the exporting country. initial comparative advantage. Among those assets Many countries have experienced short-term are favorable agroclimatic conditions and ample spurts in horticultural exports; few have been able and accessible land and water resources; a physical to consolidate their early gains. Those that have location on the sea or close to a major market; done so invested in research and adopted interna- ample and relatively inexpensive labor; and a class tional technologies, expanded and upgraded logis- of entrepreneurs with commercial experience. tical facilities, strengthened vertical supply chains, Many countries may possess some or even most of developed industry organizations for collective these assets. But translating them into a competi- action, and built credible systems for quality assur- tive horticultural industry that maintains or im- ance and food safety management. Industry expan- proves its competitiveness over time requires a sion induced the development of associated indus- distinctive set of investments and institutional tries, such as packaging, equipment supply, and structures, a range of facilitative government poli- technical consulting, which in turn contributed to cies, and, usually, a bit of luck (Jaffee 2003; Gabre- the underlying competitiveness of the industries. Madhin and Minot 2003; FAO 2003; Huang 2004). Further investments were made in the industries' Jaffee (1993) examines the ingredients common underlying assets, for example, through irriga- to the initial growth and subsequent maturation of tion development and worker and management 254 Global Agricultural Trade and Developing Countries training. Synergies have generally developed tion within the industry and in critical support serv- between export horticulture and complementary ices, negotiation of favorable international market industries such as domestic catering and tourism. access, and resolution of trade disputes. With certain historical exceptions, in most of the long-standing industries the private sector domi- Explaining Intercountry Differences in Export nates the commercial dimensions of the business, Performances while governments play a substantial and multidi- mensional facilitative role. In the early stages of Focusing on the factors identified in the literature as industry development, the public sector has been influencing export performance, this section critical in improving transportation and port/ attempts to quantify their importance in determin- airport infrastructure,investing in research and farm ing the value of fruit and vegetable exports across a advisory services, facilitating access by investors and sample of 45 developing countries. The theoretical farmers to suitable land, helping to transfer tech- anchor of the empirical investigation is Redding nologies and skills, and advancing the broad array andVenables'geographic and trade model (Redding of policies that make for a conducive investment and Venables 2002).16 climate (box 13.3).Over time,other important func- The variables assumed to have a significant tions for government have emerged,notably sanitary impact on the value of fresh and processed fruit and and phytosanitary control, promotion of competi- vegetable exports are grouped into supply-capacity BOX 13.3 Peruvian Asparagus Exports--A Standard Success Story? Once the leaders of Peru's asparagus industry and incident helped motivate the industry and gov- government specialists realized that it was in the ernment to take action, by reinforcing the fact best interest of the country, they worked together that one careless exporter could disrupt markets. to bring national standards in line with interna- Beginning in 1998, officials of the Peruvian tional norms. The industry--and Peru--have Commission for Export Promotion (PROMPEX) greatly benefited as a result. Over the past convinced the asparagus industry to implement decade, Peru has quickly risen to become one of the Codex code of practice on food hygiene, not the world's largest exporters of asparagus. In because it was the easiest but because it was the 2002 earnings reached $187 million, represent- most appropriate. PROMPEX specialists worked ing nearly 25 percent of Peru's total agricultural closely with industry leaders and production exports. Peru is able to produce quality asparagus managers to ensure proper implementation of year-round, yet because of high transportation good hygiene standards. The industry soon saw costs, its exporters are unable to match prices improved production methods, greater worker with inexpensive asparagus from some other efficiency, and better product quality. countries. Nonetheless, they have continued to Thus, when the national fresh asparagus increase exports and gain market share by grow- norms were published in early 2001, because ing asparagus of consistently higher quality that the industry was already familiar with the con- can be internationally certified. By meeting inter- cept of national standards, producers quickly national standards, Peruvian exporters have complied with little argument. The first national increased production and worker efficiency, norms--for fresh asparagus--established a qual- gained access to industrialized country markets, ity and performance baseline for the industry built customer loyalty, and drastically reduced the that allowed many to generate the skills and industry's risk of trade disruptions caused by poor experience needed to voluntarily certify under quality, food safety hazards, and plant disease. more stringent international standards, includ- In 1997 Spanish health authorities asserted ing HACCP (hazard analysis and critical control that two cases of botulism had been caused by point), traceability systems, and Good Agricul- consumption of canned Peruvian asparagus. tural Practice (GAP) certification. Many large Despite assurances from the Peruvian govern- exporters have reached the level where they ment and companies, press coverage of the can now be certified under the even stricter botulism scare left an unfavorable impression EUREPGAP protocol. among consumers in European markets, causing Source: Tim O'Brien, Interamerican Institute sales to slump in Peru's leading market. The for Cooperation in Agriculture. Fruits and Vegetables: Global Trade and Competition in Fresh and Processed Product Markets 255 variables and market-access variables. Supply- · Remoteness (that is, being landlocked) has a sig- capacity variables include: nificant adverse effect on fruit and vegetable exports, corroborating the literature on geogra- · Domestic market size, captured by the size of phy and trade--for example, Frankel and Romer urban population. (1999), who showed that countries that are land- · Infrastructure, proxied by the percentage of locked or remote from major markets tend to paved roads and access to telephones. trade less than those that are not. · Institutional quality and setting, captured by · Domestic market size comes out with a negative two variables: the number of days to enforce a sign in almost all estimations (although at a sta- contract and whether or not the country is a sig- tistically insignificant level), apparently contra- natory of the International Plant Variety Protec- dicting the usual argument that exporting fruits tion Convention (a dummy variable). and vegetables requires the prior or parallel · Human capital, captured by two separate vari- development of domestic markets and experi- ables: availability of semi-skilled labor, captured ence in brand name merchandising (see, for by the adult literacy rate, and managerial capac- example, Jaffee 1993).18 ity, proxied here by the level of manufacturing exports. There are sharp differences in the factors explaining the intercountry performance in exports Market-access variables include: of fresh versus processed fruits and vegetables. Holding all else constant, the economies of scale · Geographic variables: landlocked status obtained through large volumes of air freight are a (dummy), which increases distance and cost to key success factor for fresh fruit and vegetable reach markets. exports but do not appear significant for processed · Volume of air transport freight (in millions of horticultural exports (annex table A4 on the tons per kilometer). This variable attempts to CD-ROM). This is because economies of scale capture freight space availability and the translate into lower transportation costs, which economies of scale in international transport claim a larger share of final value for fresh products (Clark, Dollar, and Micco 2002). The higher the than for processed products. Clearly, higher volume of freight, the higher the economies of spoilage and handling costs make fresh produce scale realized by shippers and the lower the much more expensive to transport, explaining why transport cost.17 more countries tend to import fresh produce from · Existence of a preferential agreement with the the closest producers (Huang 2004). In contrast, European Union or the United States (dummy because they are easier to handle and are almost variable). universally shipped by sea, transport costs are smaller for processed products, making the geo- Table 13.A2 (on the accompanying CD-ROM) graphical outreach of processed trade much larger. reports our estimations of the equation for fresh The other factor that has a differentiated impact fruit and vegetable exports and for processed fruits on fresh versus processed exports is the level of pro- and vegetables. Common factors--notably distance tection. While preferential access to the European (landlocked status) and the level of human capi- Union or the United States has a positive and sig- tal--explain success in both fresh and processed nificant impact on export of processed horticul- export markets. tural products, it is not statistically significant at 5 percent for fresh product exports. This result is · Literacy and managerial capabilities exert a consistent with the structure of tariffs analyzed strong, robust, and statistically significant impact above, which features a high degree of escalation in on export of fresh and processed fruits and veg- OECD countries. etables. This result reflects two facts. First, horti- In conclusion, our estimations show the critical culture is a knowledge-intensive business. Sec- importance of proximity to major export markets ond, success in world markets requires the and availability of human capital as common fac- availability of a skillful class of entrepreneurs. tors explaining success in exports of both fresh and 256 Global Agricultural Trade and Developing Countries processed horticultural products. They also indi- degrees, complex protection tools--among them cate that countries wishing to boost exports of fresh highly dispersed ad valorem tariffs, specific duties, products should invest in high-quality logistics, seasonal tariffs, tariff escalation, and preferential whereas those seeking success in processed markets access along with tariff-rate quotas. A complex sys- need to develop or tie in with leading product tem of preferential access in many rich countries brands and circumvent tariff escalation through provides privileged partners with favorable entry preferential agreements with major trading without undermining protection of domestic pro- partners or within the framework of multilateral ducers. The product coverage of preferential access negotiations. schemes is wide but quotas for "sensitive products" often limit entry. Tariff escalation for processed fruit and vegetable products is widespread, Conclusion although its extent varies significantly between This chapter has highlighted the major global, regions. regional, and product-specific trends in fruit and Because horticultural sectors throughout the vegetable products trade and examined major world have traditionally seen a low level of direct structural and policy factors that have affected this government interventions, changes in domestic trade over the past two decades. Growth in world support cannot affect the sector broadly or signifi- imports was 2­3 percent a year in the 1990s, repre- cantly. Reductions in tariffs and other import senting a deceleration in the rate of growth from restrictions, however, are critical in determining the the 1980s. This slower growth in world trade in the impact of trade agreements and policies on world 1990s was strongly affected by the European Union, horticultural trade. Further tariff liberalization which experienced relatively low growth in popula- efforts would need to reduce tariff peaks, especially tion and income during the decade and had many in the European Union and the Republic of Korea. mature and saturated product markets. Adverse Past trends suggest that the main beneficiaries from price movements for fresh and processed products such reforms will be a limited number of middle- from the mid-1990s onward also contributed sub- income countries that have developed strong pro- stantially to the overall deceleration in the growth duction, postharvest processing, logistical market- of trade values. Robust growth in trade has contin- ing, and sanitary and phytosanitary management ued among NAFTA countries, for sales to high- systems over the years and that continue to attract income Asian countries, and between developing new investment. With only a few exceptions, low- countries. income countries still face enormous supply-side For developing countries, fresh and processed challenges in taking advantage of existing and fruits and vegetables is now one of the most impor- future international market opportunities. tant categories for agro-food trade, accounting for about 22 percent of their exports in 2000­01. This is far larger than their current level of trade in many Notes traditional commodities. Still, although many 1. More detailed tables on trade flows and tariffs are pre- developing-country suppliers have entered this sented on the attached CD-ROM. 2. "Statistics on Foreign Trade of Vegetables." Vegetable Sup- field, relatively few countries have achieved sus- ply Stabilization Fund. Tokyo. October 2002. tained success at a high level--testimony that the 3. This reflects the heavy influence of Spain, which is the industry is highly competitive and rapidly chang- global leader in fresh fruit exports (mainly oranges and clemen- tines). Italy is also a significant exporter of grapes, apples, and ing, and that it requires sustained investments in peaches. infrastructure, human capital, technology, and 4. Bananas accounted for more than 80 percent of the fresh good governance. fruit imported by the European Union from these countries Unlike the situation in many other agricultural (Huang 2004). 5. Between 1993 and 2003, gross domestic product growth in sectors, production and export subsidies are not Germany averaged 1.4 percent, the lowest among EU member common in horticulture. Instead, domestic fruit states. and vegetable producers are protected through reg- 6. See box A1 in annex to this chapter in the CD-ROM. 7. Lacking a better alternative, food prices are considered ulation of market access. The European Union, here as a proxy for the prices of fruits and vegetables. The main Japan, and the United States use, to varying caveat associated with this is that the range of goods covered in Fruits and Vegetables: Global Trade and Competition in Fresh and Processed Product Markets 257 the food-price index is wider than what a proper (and exoge- Clark, X., D. Dollar, and A. Micco. 2002. "Port Efficiency, Mar- nous) fruits and vegetables import price index would cover. itime Transport Costs and Bilateral Trade," NBER working 8. Mexico remains the world's top exporter of many smaller paper 10353. National Bureau of Economic Research, Inc, vegetable products--among them asparagus, eggplant, and Cambridge, MA. onions. FAO (United Nations Food and Agriculture Organization). 9. The bulk of Syria's trade is with other Middle Eastern 2003."The Market for Nontraditional Agricultural Exports." countries; Argentina's is targeted primarily to other Latin Commodity and Trade Division. Rome. American countries. FAOSTAT 2003. FAO Statistical Databases. Accessible at 10. The aggregate measure of support was defined in the http://apps.fao.org/default.jsp. Uruguay Round Agreement on Agriculture as an aggregate sub- Frankel, J., and D. Romer. 1999. "Does Trade Cause Growth?" sidy measure, designed to quantify and compare countries' American Economic Review 89(3), June, 379­399. annual levels of domestic support. It aggregates the effects of all Gabre-Madhin, E., and N. Minot. 2003. "Successes and Chal- trade-distorting policies (direct subsidies plus implicit subsidies lenges for Promoting African Horticulture Exports." World from border measures) into a single measure of support. Bank, Washington, D.C. August. 11. An importer can choose one of the three following meth- Goldstein, M., and M. S. Khan. 1985. "Income and Price Effects ods to calculate entry price: the standard import value (SIV), in Foreign Trade." In R. W. Jones and P. B. Kenen, eds., Hand- calculated daily by product and by origin and published in EU's book of International Economics, vol. 2. New York: Elsevier Official Journal; the f.o.b. price of the product in the country of Science Publications. origin; the effective resale value of the shipment concerned. Huang, S. W., ed. 2004."Global Trade Patterns in Fruits and Veg- 12. Bananas are the only exception in the fruit and vegetable etables." Agriculture and Trade Report WRS-04-06. U.S. category. For this product, duty-free access is phased in between Department of Agriculture, Washington, D.C. 2002 and 2006 by a 20 percent yearly tariff reduction. This is Jabati, M.C. 2003. "Market Access for Developing Countries of unlikely to have a significant impact in the short run, however, Africa: The Reality." AGSF Occasional Paper No. 1. United as all LDCs producing bananas of export quality belong to the Nations Food and Agriculture Organization. Rome. ACP group, which enjoys duty-free access. Jaffee, S. 1993. "Exporting High-Value Food Commodity: Suc- 13. This preferential access has allowed countries like cess Stories From Developing Countries." World Bank Dis- Cameroon, Côte d'Ivoire, and Dominican Republic to compete cussion Paper 198. Washington, D.C. with lower-cost Latin American suppliers (Costa Rica, Ecuador). ------. 2003. "From Challenge to Opportunity: Transforming 14. For more details regarding the definitions and estima- Kenya's Fresh Vegetable Trade in the Context of Emerging tions of these indicators, see chapter 4 of this volume. Food Safety and Other Standards in Europe." Agriculture 15. A recent study has identified three categories of products and Rural Development Discussion Paper 1. World Bank, that receive different treatments under the European Union and Washington, D.C. Mediterranean countries trade agreements: first, the products Redding, Stephen, and A. Venables. 2002. "The Economics of for which the preferential margin (tariff difference with MFN) is Isolation and Distance." Nordic Journal of Political Economy granted without quantitative restriction but with seasonal 28, Conference Volume, no. 2, pp. 93­108. restrictions (such as tomatoes); second, the products for which Sallyards, M. 2001. "European Union Agricultural Situation: EU the tariff reduction applies solely if the entry price is higher than Fruits and Vegetables Regime." GAIN Report E21058. U.S. a reference price in EU; and a third group for which the impact Department of Agriculture, Washington, D.C. of tariff reductions are severely limited by quota restrictions Stevens, C., and Jane Kennan. 2000. "Will Africa's Participation (such as bananas and olives). Chahed and Drogué (2002). in Horticulture Chains Survive Liberalization?" Working 16. The empirical model and its derivation from the Redding Paper 106. University of Sussex, Institute for Development and Venables' theoretical model is described in box A2 of the Studies, Brighton, U.K. annex to this chapter in the attached CD-ROM. Thunder Lake Management, Inc. 2002. "Florida-Brazil Citrus 17. Lack of data prevented us from using the volume of mar- Case." Trade Policy Monitor. April. Online: http://www. itime freight. thunderlake.com/brazil_oj.html. 18. As noted earlier, a large and rapidly growing urban pop- Vegetable Supply Stabilization Fund. 2002."Statistics on Foreign ulation (as in China and India) can absorb very large quantities Trade of Vegetables." Tokyo. of fruit and vegetables and lead entrepreneurs to focus on serv- WTO (World Trade Organization). 2000a. "Domestic Support: icing the domestic market. Background Paper by the Secretariat." G/AG/NG/8. May 26. ------. 2000b. "Export Subsidies: Background Paper by the Secretariat." G/AF/NG/S/1. April. References Chahed, Y., and S. Drogué. 2002. "Echanges agricoles et Accords Euromed, l'érosion des préférences en question." Depart- ment of Public Economics, INRA-INAPG, Paris-Grignon. 14 COTTON: MARKET SETTING, TR ADE POLICIES, AND ISSUES John Baffes Cotton--by far the most common natural fiber of rable to cotton's since the early 1970s. Between the 19th and 20th centuries--has been used as a 1960 and 1972, the polyester price indicator raw material for clothing for at least 5,000 years. Its declined from $12 to $2.50 per kilogram, mainly a use expanded significantly after 1793, when the reflection of the technological improvements (and invention of the cotton gin introduced mechanical consequent cost reductions) that took place in the separation of lint from seed. The industrial revolu- chemical fiber industry. After reaching parity with tion, which reduced the cost of producing textiles, cotton prices in 1972, the ratio of polyester to cot- accelerated cotton's progress.1 ton prices has increased at an average rate of 1 per- Cotton's most important competitors are natural cent per year, implying that while cotton and and synthetic man-made fibers such as rayon and polyester are priced at similar levels, polyester has polyester.2 Although large-scale commercial pro- made small pricing gains (see figure 14.1). duction of man-made fibers did not begin until after World War II, experimentation was taking place as The Global Cotton Balance early as the late 1800s. In 1925, rayon, a natural man- made fiber produced from cellulose, accounted for Cotton is produced in many countries, but the 1.6 percent of the world's total fiber consumption. Northern Hemisphere accounts for 90 percent of Twenty years later, this share had increased to global output. More than two-thirds of the world's 11.8 percent. The share of all man-made fibers in cotton is produced by developing countries. total fiber consumption reached 22 percent in 1960 Between 1960 and 2001, global cotton output and now stands at about 57 percent. doubled--from 10.2 million to 20 million tons. As production of man-made fibers expanded, Most of this growth came from China and India, cotton's share fell (figure 14.1). Between 1960 and which tripled and doubled their production, 2002, man-made fiber consumption grew at an respectively, during this 40 year period. Other annual rate of 4.7 percent, compared to just countries that significantly increased their share 1.8 percent for cotton. of cotton production were Turkey, Greece, and Synthetic (noncellulose) man-made fibers such Pakistan. Some new entrants also contributed. as polyester and nylon have traded at prices compa- Australia, which produced only 2,000 tons of cotton 259 260 Global Agricultural Trade and Developing Countries FIGURE 14.1 Cotton's Share in Total Fiber Consumption and Polyester to Cotton Price Ratio, 1960­2002 Cotton's share (percent) Price ratio 80 4.5 4.0 70 3.5 3.0 60 2.5 50 2.0 1.5 40 1.0 30 0.5 1960 1963 1966 1969 1972 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 Cotton's share Price ratio Source: International Rayon and Synthetic Fibers Committee. in 1960, averaged 650,000 tons a year during the three-quarters of global cotton consumption. late 1990s. Francophone Africa, which produced Several East Asian countries have emerged recently less than 100,000 tons in the 1960s, now produces as important cotton consumers. For example, almost 1 million tons. The two dominant producers Indonesia, Thailand, the Republic of Korea, and during the 1960s, the United States and the Central Taiwan (China) consumed only 130,000 tons in Asian republics of the Soviet Union, have main- 1960 (1.2 percent of global consumption), but tained their output levels at about 3.5 million and 1.5 million tons in 2002 (7.2 percent of global 1.5 million tons, respectively, thereby halving their consumption). shares. Several Central American countries that Growth in the demand for cotton has been slow. used to produce almost 250,000 tons of cotton Between 1960 and 2000, cotton demand grew at the annually now produce almost none. The share of same rate as population (1.8 percent a year), imply- East African cotton producers, too, has declined ing that per capita cotton consumption has considerably during this period. remained stagnant. The two largest cotton producers, China and the Stocks, which historically have fluctuated United States, each account for approximately between 20 and 50 percent of global output, have 20 percent of world output, followed by India affected the cotton market considerably, especially (12 percent), Pakistan (8 percent), and Uzbekistan in the area of price variability. The stockholding (5 percent). Other significant cotton producers are policies of the United States and China, the two the countries of Francophone Africa, Turkey, major holders of cotton stocks, have affected the Brazil, Australia, and Greece, which account for a level and volatility of prices. Two major cotton combined 18 percent of global output. destocking episodes are associated with periods of The consumption pattern of cotton is primarily considerable price variability: the 1985 shift in U.S. determined by the size of the textile industries of policy from stockholding to price support and the the dominant cotton consumers. China, the leading 1999 reforms in China. textile producer, absorbed more than a quarter of One-third of cotton production is traded inter- global cotton output during the late 1990s. Other nationally. The four dominant exporters--United major textile producers (and hence major cotton States, Uzbekistan, Francophone Africa, and consumers) are India, the United States, and Australia--account for more than two-thirds of the Turkey, which, together with China, account for world's exports. Four major producers, China, Cotton: Market Setting, Trade Policies, and Issues 261 India, Pakistan, and Turkey do not export cotton Price Trends and Variability and occasionally import to supply their textile Real cotton prices over the last two centuries have industries. Imports of cotton are more uniformly followed a declining pattern showing temporary distributed than exports. spikes and troughs. The reasons for the long-term During the 2000­01 season, the eight largest decline are similar to those behind the price importers (Indonesia, India, Mexico, Thailand, declines in most primary commodities--reduction Turkey, Russia, Italy, Korea) accounted for more in the costs of production due to technological than one-half of world cotton imports. Apart improvements, slow demand growth, and strong from Russia (which before 1990 was considered a competition from substitutes (in this case, chemical major producer but not an importer because Cen- fibers). The declining pattern of cotton prices has tral Asian cotton production was considered not been smooth, and it appears that a structural internal trade), most of the remaining cotton break took place in the mid-1980s. Between 1960 importers are new in the sense that they have been and 1984 real cotton prices averaged $2.62 per kilo- importing cotton to supply newly developed gram. Following a sharp decline in 1984 (from textile industries. For example, four East Asian $2.45 per kilogram in 1984 to $1.83 in 1985 and textile producers (Indonesia, Thailand, Taiwan, $1.27 in 1986), they have been fluctuating around and Korea) accounted for less than 3 percent of $1.49 per kilogram. Between 1985 and 2002, prices world cotton imports in 1960, compared to declined 0.9 percent a year (as opposed to just 22 percent in 2002. 0.2 percent a year during 1960­84). The International Cotton Advisory Committee Reductions in the costs of production stem pri- (ICAC) collects data comparing costs of production marily from yield increases--from 300 kilograms among cotton producers. Its most recent 2001 sur- per hectare in the early 1960s to 600 kilograms per vey, based on a questionnaire of 28 cotton- hectare in the late 1990s. The phenomenal yield producing countries, suggests that West Africa growth is attributable to the introduction of im- (especially Benin, Mali, and Burkina Faso), Uganda, proved cotton varieties, expansion of irrigation, use and Tanzania are among the lowest-cost cotton pro- of chemicals and fertilizers, and mechanical har- ducers. High-cost producers are the United States, vesting. To these improvements one should add Israel, and Syria. The two European cotton produc- developments in genetically modified seed technol- ers, Greece and Spain, are probably the world's ogy and precision farming during the late-1990s, highest-cost cotton producers, although they did which are expected to further reduce the costs of not participate in the survey. Calculating and com- production. Innovations in transportation and in- paring the costs of producing cotton in various formation technology have lowered costs of trans- countries is, admittedly, a difficult task, involving porting cotton and reduced the need to hold large assumptions about the cost of land and capital as inventories. Substantial technological improve- well as various hidden subsidies and distortions. ments in the textile sectors have made it possible to The ICAC (2001) warns that its data must be used obtain high-quality fabric from lower-quality cot- carefully:"Differences in production practices, vari- ton, a trend that holds for many products whose ations in the input supply among countries, and main input is a primary commodity. direct and indirect technical and financial support The prime movers behind the 1984­85 decline to farmers in the form of free seed, technical advice, in cotton prices were the structural shift in the sup- etc., make comparisons difficult among countries." port policy of the United States and the shift in Population growth for the current decade is China's trade policy (MacDonald 1997). During the projected at 1.2 percent a year. In the absence of 1950s the U.S. Commodity Credit Corporation policy reforms by major players, ICAC (2003a and bought and sold most American cotton. For exam- 2003b) projects that annual consumption growth ple, between 1962 and 1966, it accounted for almost during the decade will be about 1.8 percent, imply- two-thirds of cotton stocks. Although its role was ing that by 2010 world cotton consumption will be reduced after 1970, the United States still accounted 23.6 million tons. That may be optimistic, however, for 35 percent of world cotton stocks (exclusive of considering that for the last 15 years cotton con- Chinese stocks). Following enactment of the 1985 sumption grew at an annual rate of just 0.7 percent. 262 Global Agricultural Trade and Developing Countries Farm Bill, support prices for cotton (that is, "loan as variability from one year to another shows that rates," the equivalent of a floor price) were substan- during 1985­2002 volatility was 2.5 times higher tially reduced, and most of the U.S. stocks were than in 1960­72 but only half of the level in released to the market, depressing the world prices. 1973­84. Note that 1973 reflects the commodity 1985 also marked the beginnings of large exports price boom, while 1985 coincides with the U.S. by China, which for the previous 20 years had been change in cotton policy and the subsequent dis- a net importer. In fact between 1980 and 1985, posal of large cotton stocks. In summary, cotton China went from the world's largest importer to the prices were very stable before 1974, highly volatile world's largest exporter. until 1985, and then less volatile, but not as stable as Visual inspection of the 1984­85 price decline before 1974. suggests a structural break in the series, something also supported by statistical tests. However, it may Cotton and the Developing be argued that if the policy shift in the United Countries States, which caused massive destocking, had been the main reason behind the price decline, a new Although cotton trade is insignificant on a global stock equilibrium level would have brought a price scale--accounting for just 0.12 percent of total increase, making the 1984­85 decline temporary. In merchandise trade--it is an important cash crop reality, the U.S. policy shift accelerated a price for several developing countries at both the farm decline that would have taken place even without it. and national levels. Cotton accounted for between Real cotton prices did rise somewhat after the shift 30 and 44 percent for total merchandise exports but never reached pre-1984 levels. in five West African countries (Burkina Faso, While falling, cotton prices have been volatile. Benin, Chad, Mali, Togo) during 1998­99 Admittedly, measuring volatility is a difficult (and (table 14.1). The corresponding figures for often tricky) task precisely because prices have Uzbekistan, Tajikistan, and Turkmenistan are 32, shown a long-term, nonlinear decline, making it 15, and 12 percent. Cotton's contribution to the difficult to isolate a meaningful average around gross domestic product (GDP) of these countries which variability can be defined. Defining volatility has been substantial, ranging between 3.6 percent TABLE 14.1 Cotton's Importance to Developing and Transition Economies, 1989­99 Average Cotton Exports Percent of Merchandise Millions of Merchandise Percent of Exports (millions Per Capita Countrya Dollars Exports GDP of dollars) GDPb Burkina Faso 127 43.9 5.1 289 249 Benin 164 39.1 7.1 419 398 Uzbekistan 1,038 32.2 6.5 3,227 467 Chad 76 32.2 4.7 236 224 Mali 180 29.5 6.7 611 285 Togo 67 21.3 4.7 315 341 Tajikistan 97 15.1 8.2 643 352 Turkmenistan 110 12.3 3.6 891 1,126 Tanzania 44 7.6 0.5 576 185 Syrian Arab Rep. 214 6.7 1.4 3,177 858 Sudan 41 6.0 0.4 688 290 a. Countries ranked by decreasing order of cotton exports in merchandise exports; b. Constant 1995 U.S. dollars. Source: FAOSTAT, and World Bank, World Development Indicators, various years. Cotton: Market Setting, Trade Policies, and Issues 263 (Turkmenistan) and 8.2 percent (Tajikistan). With cotton (sometimes referred to as Bt cotton) has not the exception of Turkmenistan and Syria, the per faced the degree of opposition faced by genetically capita annual GDP in these countries is well below modified food crops, allowing more rapid adop- $500. In most (especially in Africa), cotton is typi- tion. Organic cotton has been embraced enthusias- cally a smallholder crop and the main cash crop. It tically by environmental activists but not by con- is grown in rain-fed land with minimal use of pur- sumers. Hence, while there is plenty of room for chased inputs such as chemicals and fertilizers. expanding genetically modified cotton, the scope According to FAO (Food and Agriculture Orga- for expanding organic cotton appears to be limited. nization) estimates, as many as 100 million rural Genetically modified cotton, a result of techno- households may have been involved in cotton pro- logical developments of the 1990s, has the potential duction during 2001. In China, India, and Pakistan of reducing the cost of production and hence about 45, 10, and 7 million rural households, respec- increasing profitability of the early adopters of tively, were engaged in cotton production. The total this technology. Like other genetically modified number of rural households depending on cotton products, it provides insurance against pests, insects, in major African producing countries, including and weeds. Growers pay a premium for the resistant Nigeria, Benin, Togo, Mali, and Zimbabwe, was seed, as they would when buying insurance. 6 million. Genetically modified cotton was first grown in The high dependence on cotton in these coun- the United States in 1996. Among the cotton- tries has important ramifications for poverty, espe- producing countries that have introduced it since cially when prices change suddenly. In a study of then are China, India, and Mexico in the Northern Benin, Minot and Daniels (2002) estimated that a Hemisphere and Argentina, Australia, and South 40 percent reduction in farm-gate cotton prices-- Africa in the Southern. Other countries are in the equivalent to the price decline that occurred process of approval or at the trial stage, including between December 2000 and May 2002--implied a Brazil, Indonesia, Israel, Pakistan, and Turkey. 7 percent reduction in rural per capita income in Major producers that had not used or approved the short run and a 5­6 percent reduction in the genetically modified cotton as of 2003 were the long run. They also estimated that the incidence of European Union, Central Asia, and Francophone poverty among cotton growers will rise in the short Africa (except Burkina Faso, which is conducting run from 37 percent to 59 percent, while the aver- trials). age incidence of rural poverty (among cotton It is estimated that about 22 percent of the growers and other farmers) will increase from world's cotton plantings are now in genetically 40 percent to 48 percent. modified varieties, up from 2 percent in 1996­97. In terms of policy interventions, the cotton sec- The largest user of such cotton is the United States, tor in developing countries has been traditionally which during the 2003­04 season is estimated to taxed either explicitly through export taxes or have sown 70 percent of its cotton area with geneti- implicitly through price-fixing arrangements or cally modified varieties. In Australia about 44 per- exchange-rate misalignments. The pattern, how- cent of cotton area was sown to such varieties ever, changed somewhat during the 1990s, as a in 2002­03, up from 40 percent two years earlier. In number of cotton producers undertook policy China, which adopted the new technology at an reforms. However, several African and all Central experimental stage in 1996, more than 20 million Asian cotton producers still tax their cotton sectors. hectares were planted with genetically modified varieties in 2002, corresponding to more than 20 percent of cotton acreage. In addition to the Nonconventional Cotton imported genetically modified varieties, China has Production developed 11 of its own varieties. According to Pray Recent trends in growing cotton focus on cost and others (2001), the major share of the benefits reductions through less intensive use of inputs, from growing Bt cotton in China went to farmers especially chemicals. These include the use of (most of whom are smallholders). In contrast, most genetically modified seed technology and organic of the benefits associated with genetically modified methods of production. Genetically modified products in the other cotton-producing countries 264 Global Agricultural Trade and Developing Countries go to biotech and seed companies. They also found process compares unfavorably with much simpler that the increased use of genetically modified cot- labeling for, say, coffee or tea where something like ton in China was associated with considerable pos- "Organically grown from Costa Rica" or "Organic itive health effects--notably fewer hospitalizations of Kenyan origin" is likely to suffice. Third, organic from pesticide poisoning. Farmers who did not use products are typically associated with health- Bt cotton had to spray 12 times on average, whereas related benefits that do not apply to nonfood prod- farmers who used Bt cotton had to spray only ucts such as cotton. 3­4 times. If the conversion to genetically modified cotton varieties continues at rates experienced dur- Distortions in the Cotton Market ing the last few years, as much as half of the world's Cotton has been subject to various marketing and cotton (from 40 percent of total cotton acreage) trade interventions. Townsend and Guitchounts will be of genetically modified origin within five (1994) estimated that in the early 1990s, more than years. two-thirds of cotton was produced in countries The second trend, organic cotton, may be a that had some type of government intervention, small market niche to be exploited by developing including taxation and subsidization policies. The countries. Many developing countries can be classi- ICAC (2002 and 2003), which has been monitoring fied as "organic" cotton producers without altering the level of assistance to cotton production by their current production practices because of their major producers since 1997­98, found that eight low reliance on chemicals and fertilizers. The countries provided direct support to cotton potential for organic cotton appears to be limited, production--Brazil, China, Arab Republic of however. Organic cotton initiatives have taken Egypt, Greece, Mexico, Spain, Turkey, and the place in many countries, including in Africa, but United States (table 14.2). For 2001­02, direct gov- the scale is still insignificant compared to global ernment assistance to U.S. cotton producers production of conventional cotton. Myers and reached $3.9 billion; China's support totaled Stolton (1999) reported that in 1997, about 8,150 $1.2 billion; and the European Union's was almost tons of certified organic cotton fiber was produced $1 billion. Producers in Brazil, Egypt, Mexico, and worldwide--2,600 tons was produced in the United Turkey received a combined total of $150 million in States, 1,175 in India, 1,800 in Turkey, 1,570 in support. India also supported its cotton sector in Africa, and 845 in Latin America. 2001­02 with an estimated $0.5 billion. Significant expansion of organic cotton faces In addition to domestic support, some border difficulties on both the supply and demand sides. restrictions apply, mainly in the form of import On the supply side, the certification process (espe- tariffs. Most countries that impose import quotas cially in African cotton-producing countries where are cotton exporters, some with large textiles sec- the majority of growers are smallholders) is costly tors. Import tariffs rates for 2003 were: Argentina to implement and monitor. On the consumption (7.5 percent); Brazil (7.5 to 10 percent); China side, demand for organic cotton is not as strong as (3 percent within quota, 90 percent outside it is for other commodities such as coffee and tea. quota3); Egypt (5 percent); India (10 percent); There are three reasons for this. First, there is a "dis- United States (4.4 cents per kilogram within quota tance" in the eyes of the consumer between the and 31.4 cents per kilogram outside quota4); primary product (cotton) and the final product Uzbekistan (10 percent); and Zimbabwe (15 per- (cloth). Second, consumers of clothing (as opposed cent duty plus 5 percent import tax). to consumers of, say, beverages) must pay attention The remainder of this section analyzes the struc- to a host of factors before they make their purchas- ture and degree of interventions in the United ing decision. The decision involves brand, color, States, European Union, and China. It also looks at style, size, type of cotton (typically identified by its Uzbekistan, a country that taxes its cotton sector. country of origin), content (for example, 80 per- cent cotton, 20 percent polyester), and care instruc- United States tions. Adding to that already congested list infor- mation on whether the cotton is of organic origin is The main channels of support in the United States rather difficult. Note that this decision-making are decoupled payments (formerly known as Cotton: Market Setting, Trade Policies, and Issues 265 TABLE 14.2 Direct Government Assistance to Cotton Producers, 1997­98 to 2002­03 (US$ millions) Country 1997­98 1998­99 1999­2000 2000­2001 2001­02 2002­03 United States 1,163 1,946 3,432 2,148 3,964 2,620 China 2,013 2,648 1,534 1,900 1,196 750 Greece 659 660 596 537 735 718 Spain 211 204 199 179 245 239 Turkey -- 220 199 106 59 57 Brazil 29 52 44 44 10 -- Mexico 13 15 28 23 18 7 Egypt, Arab Rep. 290 -- 20 14 23 33 -- Not available. Data for 2001­02 are preliminary. Source: ICAC 2002 and 2003, U.S. Department of Agriculture, and European Union. TABLE 14.3 Government Assistance to U.S. Cotton Producers, 1995­96 to 2002­03 (US$ millions) Policy Instrument 1995­96 1996­97 1997­98 1998­99 1999­2000 2000­01 2001­02 2002­03 Coupled payments 3 0 28 535 1,613 563 2,507 248 PFC/DP 0 599 597 637 614 575 474 914 Emergency/CCP 0 0 0 316 613 613 524 1,264 Insurance 180 157 148 151 170 162 236 194 Step-2 34 3 390 308 422 236 196 -- Total 217 759 1,163 1,946 3,432 2,148 3,964 2,620 -- Not available. Note: PFC denotes production flexibility contracts, DP denotes direct payments, CCP denotes countercyclical payments. Source: U.S. Department of Agriculture (assistance); International Cotton Advisory Committee (production); and author's calculations. production flexibility contracts), loan deficiency former. Export subsidies, or Step-2 market pay- payments (through the loan-rate mechanism), ments, are made to eligible cotton exporters and insurance, subsidies to domestic mills (the so- domestic end users of cotton when domestic U.S. called Step-2 mechanism, also referred to as prices exceed c.i.f. (cost, insurance, and freight) export subsidy), and emergency payments (intro- prices in northern Europe by a certain level and duced in 1998 to compensate for the loss of the world price is within a certain level of the base income caused by low commodity prices but loan rate. The objective of the Step-2 payment is made "permanent" under the 2002 Farm Bill) to bridge the gap between higher U.S. domestic (table 14.3). Direct payments, predetermined prices and world prices so that U.S. exporters and annual payments based on historically enrolled textile mills maintain their competitiveness. areas of cotton, were introduced with the 1996 In 2002 the U.S. Congress passed a farm bill that Farm Bill to compensate farmers for "losses" is expected to be in place for the next six years. The stemming from elimination of earlier loan defi- 2002 Farm Bill retained the earlier support through ciency payments. Market price payments are various loans, flexibility contracts, and insurance, designed to compensate cotton growers for the as well as the Step-2 payment, while legitimizing difference between the world price and the loan emergency assistance under the term "countercycli- rate (the target price) when the latter exceeds the cal payments." If cotton prices remain at their 266 Global Agricultural Trade and Developing Countries 2001­02 levels, then U.S. support to its cotton sec- level, implying that when this restriction is con- tor is expected to be in the order of $3.5 billion to verted to individual basis, it creates not only $4.0 billion for the next six years, implying the U.S. administrative complexities but also leads to misal- cotton producers will be receiving close to twice the location of resources (see chapter 5 on decoupling world market price. for more on this issue). Karagiannis and Pantzios (2002) found that the current system failed as a sur- plus containment mechanism and also resulted in European Union farm income losses. In the 1960s there were three cotton producers in Between 1995­96 and 1999­2000 the budgetary Europe. Greece and Spain produced an average of expenditure on cotton aid ranged between 740 85,000 tons each; Bulgaria produced 25,000 tons. million and 903 million, implying that, on aver- Throughout the 1970s Bulgaria's output declined, age, EU cotton producers received more than twice while that of Greece and Spain stayed at the levels the world price of cotton. Note that even in periods seen during the 1960s. Cotton production by the of high prices, EU cotton producers would receive three countries taken together declined by 0.4 per- support since the amount allocated to the cotton cent a year between 1960 and 1982. With the Euro- sector had to be disbursed. In addition to output pean Union's expansion and the subsequent acces- subsidies, EU cotton producers receive subsidies on sion of Greece and Spain, cotton production grew inputs such as credit for machinery purchase, by an annual average of 7.3 percent during the insurance, and publicly financed irrigation. On 1990s, averaging 325,000 and 78,000 tons in Greece September 23, 2003, the EU Commission proposed and Spain, respectively. to reform its cotton, sugar, and tobacco sectors Under the EU Common Agricultural Policy (European Commission 2003). Under the cotton (CAP), support is given to cotton growers based on reform proposal, EU support to the cotton sector the difference between the market price and a will consist of the following parts: a single farm guide (support) price. Advance payments are made payment scheme; a production aid scheme, granted to ginners based on estimates of seed cotton pro- as an area payment; and development measures. duction. They pass the subsidy on to growers by paying higher prices. The policy also influences the China quantity of cotton produced by setting a maxi- mum guaranteed quantity of seed cotton for which China is currently the largest producer, consumer, assistance is provided--782,000 tons of seed cot- and stockholder of cotton. China's cotton sector ton for Greece, and 249,000 for Spain, approxi- became fully government-controlled in 1953 after mately equivalent to 255,000 and 82,000 tons of the introduction of the first Five-Year Plan (Zhong cotton lint. and Fang 2003). The central planning policies The European Union reformed its cotton pro- adopted then were similar to those of the Soviet gram in 1999 (European Commission 2000). While Union and remained in place for the next 35 years. the guide price level and the maximum guaranteed The central government set production targets and quantity of seed cotton for which assistance is pro- procurement quotas. This monopoly was easily vided have been maintained, "penalties" (that is, exercised because all ginning facilities were owned reductions in subsidy) for excess production over by the cooperatives. A step to boost cotton produc- the maximum guaranteed quantity increased. tion was taken in 1978 by increasing the price of Under the reformed policy, for each 1 percent of cotton as well as supplying more fertilizer. A second excess production, the level of subsidy is lowered by boost came in 1980 with the partial abolition of the 0.6 percent of the guide price as opposed to 0.5 per- communal production system under the House- cent prior to 1999. As production increases, the hold Responsibility System, which gave land use penalty becomes stiffer, effectively, putting an rights to individual farmers. upper limit on the budgetary outlays to the cotton Evidence suggests that the government of China sector. It is important to note that this quantitative protects its cotton sector through support prices, restriction (the so-called maximum quantity guar- import tariffs, export subsidies, and public stock- anteed) applies at the aggregate (that is, country) holding. The government sets a reference price for Cotton: Market Setting, Trade Policies, and Issues 267 cotton, typically above world prices. China also cotton was either consumed by mills in Russia maintains tariffs on imports that bridge the gap (then considered domestic trade) or shipped to between domestic and world prices. Following Eastern European countries under barter arrange- its WTO (World Trade Organization) accession ments. Following the collapse of the Soviet Union, arrangements the tariffs will be reduced to 15 per- Uzbekistan began exporting its cotton to Western cent, but at the same time a tariff-related quota countries in exchange for foreign currency (until system will be implemented to manage imports. 1996 some cotton still went to Russia in barter The International Cotton Advisory Committee trade terms). found that support to the cotton sector in the six Although 12 years have passed since the change seasons beginning in 1997­98 ranged from $0.8 bil- in the trade regime, most aspects of production, lion to $2.6 billion. Huang, Rozelle, and Chang marketing, and trade of the sector closely resemble (2004) estimated that during 2001 the nominal pre-1991 arrangements. Numerous entities are rate of protection for cotton averaged 17 percent. involved in all postproduction activities of cotton. Fang and Beghin (2003), however, estimated that The three most important ones are the state between 1997 and 2000, the nominal protection company handling ginning; the state trading coefficient for cotton has averaged 0.80, implying organizations handling exports; and the Ministry that China taxes its cotton sector by 20 percent. The of Foreign and Economic Relations, handling different views on the nature and degree of inter- financial transactions. vention, however, should not be surprising given All pre- and post-ginning operations of cotton the complexities of China's agricultural policies as are handled by UzKhlopkoprom/UzPakhtasanoitish well as the unreliability of the data. (UKP), a state company that used to be a ministry. In September 1999 the government of China UKP is responsible for collecting, storing, ginning, announced reform measures which included the and classifying cotton, making payments to growers, creation of a cotton exchange to facilitate domestic and providing inputs. UKP owns considerable spot trading; the reduction of prices paid to pro- assets, including all ginning and storage facilities as ducers; and a reduction in stocks. In some sense well as handling machinery and equipment. the reforms have worked: China's stocks declined The second important entities are the three state from 4.1 million tons in 1998­99 to 2.3 million trading organizations (STOs) in charge of handling tons in 2000­01. In September 2001 further all aspects of cotton exports. The main responsibil- reforms were announced and are currently under ities of these organizations include contracting cot- way (Zhong and Fang 2003). First, the internal cot- ton merchants for the sale of cotton, organizing the ton market would be open to cross-regional trade. availability and shipment of cotton, receiving pay- Second, various enterprises would be allowed to ments and converting them into local currency, and buy cotton directly from producers with approval paying UKP. Although these organizations have a granted by the provincial government. Third, number of other responsibilities (such as purchas- ginning operations would be separated from ing machinery and equipment on behalf of the gov- marketing cooperatives, in effect making them ernment), exporting cotton is their core activity. commercial enterprises. Because each organization has been allocated a quota of cotton to be exported, there is no competi- tion involved in the export process. Uzbekistan The third important entity is the Ministry of Uzbekistan, the world's fifth largest cotton pro- Foreign Economic Relations, which reports directly ducer and second largest cotton exporter, produces to the government. Its main function is to manage more than 1 million tons of cotton annually, most cotton export operations, including setting prices, of which is exported. During 1998­99 cotton selecting buyers, and monitoring dollar receipts. exports accounted for one-third of total merchan- Several other entities are involved in the sector dise exports, while the sector contributed an aver- including the state company responsible for age of 6.4 percent to the country's GDP. Before domestic and international transportation of cot- 1991 all aspects of Uzbekistan's cotton sector were ton, the organization responsible for quality moni- under state control (of the Soviet Union). Most toring, and the customs agency. 268 Global Agricultural Trade and Developing Countries It appears that cotton growers are heavily taxed Quirke (2002) estimated that removal of pro- both directly through the lower price received by duction and export subsidies by the United States UKP (which, in turn, receives a fixed price from the and the European Union are likely to induce a STOs, as dictated by the Ministry) and indirectly 20 percent reduction in U.S. cotton production and through the exchange rate regime. A recent study a 50 percent reduction in U.S. cotton exports, with (Uzbekistan 2003) found that at an ex-ginnery much higher figures for the European Union. He price of $1.03 a kilogram, the STOs receive the also estimated that if support was not in place, equivalent of $0.63 a kilogram (these calculations world cotton prices would be 10.7 percent higher were based on a Cotlook A Index (price) of $1.24 a compared to their 2001­02 levels. kilogram). With respect to the difference between Based on a partial equilibrium model, Tokarick $1.03 and $0.63 a kilogram, the study concluded: (2003) found that multilateral trade liberalization "It is not clear exactly where this profitability figure in all agricultural markets (including cotton) is is allocated. It is alleged that, after a marketing fee is expected to induce a 2.8 percent increase in the deducted, the balance is paid to the Ministry of world prices of cotton, with 0.8 percent coming Finance as an export duty." The declared price to be from the removal of market price support and the paid to farmers by UKP is 126,000 Sum per ton of remaining 2 percent coming from the removal of seed cotton, which, at an exchange rate of 960 Sum production subsidies (removal of market price sup- per U.S. dollar and a 32 percent ginning out-turn port most likely applies to the United States Step-2 ratio, implies a price of $0.41 a kilogram, about payment). Tokarick also calculated that global one-third of the A Index. reforms would lead to $95 million in total change Perhaps, it is not unreasonable to conclude that in welfare a year. even though cotton exports from Uzbekistan FAPRI (2002) found that under global liberal- moved from a barter to a commercially oriented ization (that is, removal of trade barriers and structure, the sector is still tightly controlled by the domestic support of all commodity sectors), the government. Moreover, growers are taxed heavily, world cotton price would increase over the baseline receiving only about one-third of the export price scenario by an average of 12.7 percent over the of cotton. 10-year period (table 14.4). The largest gains in trade would go to Africa, which would increase its exports by an average of 12.6 percent. Exports from Impact of Distortions and Uzbekistan and Australia would increase by 6.0 and Prospects for Reform 2.7 percent, respectively, while exports from the The ICAC (2003a) concluded that in the absence of United States would decline by 3.5 percent. The direct subsidies, average cotton prices during the most dramatic impact is on the production side 2000­01 and 2001­02 seasons would have been 17 where the European Union's cotton output would and 31 cents a pound higher, respectively. If the decline by more than 70 percent. The latter out- United States alone removed its subsidies during come should not be a complete surprise, consider- these two seasons, world cotton prices would have ing that the European Union's cotton output been 6 and 11 cents higher, respectively. These fig- during the late 1990s was, on average, three times as ures imply cotton prices 30 and 71 percent higher much as it was before CAP took effect on the cotton than the actual averages of 57.2 and 41.8 cents a sector. pound. The study, which is based on a short-run Prospects for policy reforms by major producers partial equilibrium analysis, does acknowledge that subsidizing the sector are mixed. Support for cot- while removal of subsidies would result in lower ton in the European Union is unlikely to increase production in the countries that receive them (and for two reasons. First the countries expected to join hence higher prices in the short term), such impact the EU are not cotton producers and hence there would be partially offset by shifting production to will be no budgetary pressure. Second, the current nonsubsidizing countries in the medium to longer support scheme is subject to an upper spending cap terms; similarly higher prices are likely to reduce that appears to be a binding constraint; both the growth of cotton consumption, making the Greece and Spain, being among the world's long-run impact less striking. highest-cost cotton producers, are unlikely to Cotton: Market Setting, Trade Policies, and Issues 269 TABLE 14.4 Estimated Effect of Removal of Distortions (percentage changes over baseline) Variable 2003­04 2005­06 2007­08 2009­10 2011­12 Averagea World price 15.6 13.7 13.0 12.2 11.7 12.7 Exports Africa 12.1 15.1 14.0 13.1 12.3 12.6 Australia 3.9 3.0 2.7 2.3 2.1 2.7 United States -8.4 -6.6 -4.0 -1.5 0.9 -3.5 Uzbekistan 5.4 6.9 6.7 6.4 6.2 6.0 World 3.9 5.6 6.2 6.7 7.3 5.8 Production Africa 4.5 7.5 7.1 6.7 6.3 6.0 European Union -77.4 -77.7 -78.3 -78.8 -79.0 -70.5 United States -18.3 -7.9 -5.9 -4.1 -2.3 -6.7 Uzbekistan 3.1 4.7 4.6 4.4 4.2 4.0 a. Average is taken over the 10-year period 2001­02 to 2010­11. Source: FAPRI 2002. increase production given the reduced support they The United States took a step in the right direc- would receive if they exceed the current output tion with the replacement of the deficiency pay- levels. At the same time, support is not expected to ment system by decoupled payments in 1996, but be eliminated because it supposedly goes to low- all progress was eliminated with the 2002 Farm Bill, income areas and hence it is regarded as a poverty which effectively legitimized emergency payments reduction program. introduced in 1998­99 following the sharp decline The nature of support is shifting away from in prices; renamed them as countercyclical pay- direct price support toward partially decoupled pay- ments; increased target prices; and made it more ments. Beginning in 2006, the EU cotton sector will convenient for larger farmers to increase the sup- go through another reform. Under the Luxembourg port they receive. Historically, U.S. farm bills either Council's decisions of April 22, 2004 (which was give what they promise or give more than they based on the September 2003 proposal), an esti- promise (as the recent experience showed). Hence, mated 700 million will fund two support meas- if history is any guide, it is reasonable to expect that ures, with 65 percent of the total coming in the form U.S. cotton farmers will be receiving generous sup- of a single farm payment decoupled from current port for the next six years, unless the support production decisions and the remaining 35 percent exceeds WTO commitments, in which case the U.S. in the form of an area payment. Eligibility for the secretary of agriculture has the discretion to inter- decoupled payment will be limited to growers who vene and reduce it. produced cotton during the three-year period from A number of factors may induce some early 1999 to 2001. The area payment will be given for a reforms, however. First, the substantial increase of maximum area of 380,000 hectares in Greece, the support to the U.S. cotton sector along with 85,000 hectares in Spain, and 360 hectares in Portu- 30-year record low prices and the fact that 10 per- gal and will be proportionately reduced if claims cent of U.S. cotton growers receive 90 percent of the exceed the maximum area allocated to each country. support (hence falsifying the claim that support To receive decoupled payments, cotton growers preserves the small farm), is likely to put pressure must keep the land in good agricultural use. To for altering the nature of policy sooner. Second, receive area payments they must plant (not neces- Brazil's request for consultations at the WTO re- sarily produce) cotton. Karagiannis (2004) esti- garding U.S. cotton subsidies may create some pres- mated that the reformed regime is likely to reduce sure to lower subsidies (WTO 2002). Third, four EU cotton production between 10 and 25 percent West African cotton-producing countries (Benin, (depending on the assumed elasticity of supply). Burkina Faso, Chad, and Mali) pressed for removal 270 Global Agricultural Trade and Developing Countries of support to cotton sector through the WTO. In an Reforms in East African cotton-producing unusual move, the president of Burkina Faso countries were in response to the inefficiencies addressed the WTO on June 10, 2003, asking for faced by the parastatals that used to handle most financial compensation for cotton-producing low- (and in some occasions all) aspects of marketing income countries to offset the injury caused by and trade. For the most part, policy reforms meant support. This compensation, according to the elimination of the monopoly powers of the paras- request, should be in place for as long as subsidies tatals. Although the outcome of these reforms are in place. appears to have been mixed, if one considers that China appears to be the most promising case of the countries that undertook reforms also faced the reform. The reforms undertaken in 1999 and more most difficulties, one may argue that reforms have recently in 2001 indicate that its cotton sector will been successful. For example, during the eight-year be soon exposed to internal and external competi- period staring in 1995­96, cotton output in Uganda tion. China is also in the process of establishing a has averaged 17,000 tons, an almost three-fold cotton futures exchange, indicating that market increase compared with the eight seasons before forces within the sector are likely to play a more sig- 1995­96. The corresponding world price average nificant role in the future (Shuhua 2003). before 1995­96 was $1.56 a kilogram; after it was On the international side, while the phase-out of $1.40 a kilogram. The farmers' share in world prices the Agreements on Textiles and Clothing (ATC) is rose from less than 50 percent to 70 percent after supposed to end the distortions imposed on the the reforms, while a number of new traders and location of the textile industries, it is uncertain exporters entered the sector. This success came whether the expected benefits will be fully realized. despite the failure of most credit mechanisms that were launched after the reforms. · First, ATC is back-loaded, with most of the In Zimbabwe reforms appear to have been suc- reforms expected to take place in the last year, cessful. First, cotton production is up substantially. thus increasing the risk of noncompliance. During the eight seasons since 1995­96, cotton · Second, a number of (mainly European Union) output has averaged 115,000 tons, 50 percent countries have repeatedly sought to impose higher than the eight-year period average before antidumping duties on textile imports from Asia 1995­96. Some 30 percent of the 1997­98 cotton in recent years. harvest was marketed entirely by private entities. · Third, a number of provisions under the ATC Private companies now transport most of the allow for the imposition of temporary duties in cotton. Competition has pushed the price farmers the case that current domestic textiles suffer receive to close to 80 percent of international "significant damage" following the phaseout. prices, and producers are being paid faster. Zimbabwe has also retained the premium for qual- ity it used to receive in the world market. Reform Initiatives in Africa The outcome of cotton reforms in Tanzania has During the 1990s, a number of African cotton- been mixed. On the positive side, the share of pro- producing countries undertook substantial reforms. ducer prices increased to 51 percent (from 41 per- The reform process and its outcome have been stud- cent before the reforms). Furthermore, cotton ied extensively. See, for example, Kähkönen and growers receive payments quickly, a major achieve- Leathers (1997) for Zambia and Tanzania; Sabune ment compared with the delays encountered before (1996) and Lundbæk (2002) for Uganda; Larsen the reforms. Contrary to what many reports show, (2002) for Zimbabwe; Baffes (2001) for Uganda, quality of cotton appears not to have suffered con- Zimbabwe, and Tanzania; Baffes (2000), Badiane siderably. At the same time, cotton production and others (2002), and Goreux and Macrae (2003) since 1995­96 has averaged less than before reforms for Francophone Africa; Baffes (2004) and Gibbon (55,000 after, compared with 61,000 tons before). (1999) for Tanzania. Poulton and others (2004) On the policy side, the Cotton Board along with the looked at the cotton sectors of six African countries, two line ministries (Agriculture and Food Security, while Shepherd and Farolfi (1999) reviewed export and Cooperatives) still play a major role in the sec- commodity sectors for a number of sub-Saharan tor that goes far beyond the regulatory role they are African countries. supposed to play. Collection and dissemination of Cotton: Market Setting, Trade Policies, and Issues 271 data (as well as accuracy of statistics) are poor even Major subsidizers are the United States, $3.96 bil- by the government's own admission. lion in 2001­02 and the European Union--Greece Reforms are also under way in West Africa. The and Spain--$0.98 billion (compare this to $20 bil- World Bank has argued that the discipline and lion, the value of world's cotton production, evalu- responsibility that a free-entry competitive system ated at 2001 prices and quantities). This level of imposes on market participants would make for a support implies that prices received by U.S. and EU more resilient, flexible, self-reliant, and innovative cotton producers are 87 and 160 percent above national cotton sector. Improved competition world prices. China reportedly has been supporting through market reforms offers important oppor- its cotton sector during the last few seasons by an tunities for regional trade and cooperation, the estimated $1.5 billion annually. Many cotton- latter in areas such as research, phytosanitary regu- producing countries have reacted to low prices by lations, and seed development and certification. introducing offsetting support. Support in Turkey, Most important, improved sector performance Brazil, Mexico, Egypt, and India, totaled $0.6 billion would contribute to alleviating poverty by raising during 2001­02. Further, Brazil initiated a WTO cotton prices to levels enjoyed by farmers else- consultation process claiming losses to its cotton where in the world. exports due to subsidies by the United States. WTO Significant developments have taken place dur- determined in its interim ruling that indeed the ing the last few years, which indicate the future U.S. cotton program has violated the Agreement on direction of institutional changes in the region's Agriculture. Not only is this decision an important cotton sector. Three countries, Benin, Côte d'Ivoire, victory for Brazil, but it may also trigger similar and Togo, have now opened their sector to private cases, especially in view of the expiration of the ginners. Benin and Côte d'Ivoire have eliminated Peace Clause in the Uruguay Round Agreement on the monopoly power of their national companies Agriculture. More recently, four West African and transferred key responsibilities to the private cotton-producing countries (Benin, Burkina Faso, sector. Chad, and Mali) pressed for removal of support to the cotton sector through the WTO (the so-called "cotton initiative"). This compensation, according Summary and Conclusions to the request, should be in place for as long as sub- Cotton is very important to a number of low- sidies are in place. income African and Central Asian countries, in Given the highly distorted nature of the cotton some cases contributing as much as 40 percent to market and the fact that millions of rural poor merchandise exports and between 5 and 10 percent households in developing countries depend on this to GDP. Considering that in most countries cotton commodity, what are the alternatives? As discussed is a smallholder crop, the implications of price earlier, a number of developing countries, espe- changes (either induced by market forces or policy cially in Sub-Saharan Africa have undertaken pol- interventions) as well as changes in market share icy reforms during the 1990s. Setting aside the are enormous. For example, a 40 percent reduction lively debate on the motives of the reforms, in many in price (the equivalent of the price decline that respects the reforms have been successful. For took place from December 2000 to May 2002) example, in the few cases reviewed here, cotton implies a 7 percent reduction in rural income in growers received a higher share of f.o.b. prices, they Benin--a typical cotton-producing country in also received payments more promptly, and there West Africa. Cotton also faces intense competition was considerable supply response. In an environ- from chemical fibers, especially following techno- ment of declining commodity prices, these are not logical improvements in the early 1970s that trivial achievements. In a number of cases, how- brought their prices down to cotton price levels. ever, the reform process has either not been com- Since 1975, polyester and cotton have been traded pleted (Tanzania), has been reversed (Zimbabwe), at roughly the same price levels. Currently, the has been slow (West Africa), or has not even started share of cotton in total fiber consumption is 40 per- (Uzbekistan). In these cases further reforms are the cent (down from 68 percent in 1960). only feasible alternative. Although cotton faces minimal border restric- A second issue that should receive attention is tions, there is considerable domestic support. the enabling policy environment regarding the use 272 Global Agricultural Trade and Developing Countries of genetically modified cotton. In China for exam- wool and silk). Likewise, man-made fibers can be further ple, where genetically modified cotton is used divided into inorganic and organic fibers. Inorganic fibers are materials such as ceramic, glass, and carbon (typically not used extensively by smallholders, the costs of producing in garments.) Organic man-made fibers, on the other hand, are cotton declined by 20­25 percent. This cost reduc- mostly used in garment production either as substitutes or as tion meant doubling the net income for cotton complements to natural fibers. Organic fibers are further subdi- vided into natural and synthetic polymers. Natural polymers growers. One should also note that genetically (often called cellulosic) are made from wood. The most com- modified cotton has not been subject to negative mon natural polymer is viscose, also known as rayon. The syn- consumer reaction as has been the case with genet- thetic polymers are made from crude oil. The most common synthetic polymers are polyester, acrylic, and polyamide (also ically modified food products. known as nylon). Per capita chemical fiber consumption in 1960 A third issue (and one closely related to geneti- and 2000 was 1.75 and 4.52 kilograms, respectively. China is the cally modified cotton) is organic cotton. Producers world's dominant producer of chemical fibers, accounting for of organic products typically command significant 6.7 million tons each year. 3. China's 2003 tariff rate quota of 856,250 tons was premiums. However, organic cotton production exhausted. has not been as profitable as other organic crops 4. The U.S. tariff rate quota for 2002 was 73,207 tons, while (such as coffee and tea). The main reason is weak cotton imports totaled 6,295 tons. demand, which appears to be a reflection of the "distance" between the farm product--cotton-- References and the final product--cloth. It is because of this Badiane, O., D. Ghura, L. Goreux, and P. Masson. 2002. "Cotton distance that genetically modified cotton has not Sector Strategies in West and Central Africa." Policy Research Paper 2867. World Bank, Washington, D.C. faced resistance by the consumers, which further Baffes, John. 2000. "Cotton Reforms in West and Central Africa, reinforces the conclusion that genetically modified and the World Bank." Cotton Outlook, Special Issue: Cotton in cotton is something that developing countries the Franc Zone. Liverpool: Cotlook Limited. should consider seriously. ------. 2001. "Policy Reforms in the Cotton Market." In T. Akiyama, J. Baffes, D. Larson, and P.Varangis, eds., Commod- The price prospects (and consequently the ity Market Reforms: Lessons from Two Decades. Regional and export shares of low-cost producers, including Sectoral Studies. Washington, D.C.: World Bank. many African countries) can be improved consid- ------. 2004. "Tanzania's Cotton Sector: Constraints and Chal- lenges in a Global Environment." Development Policy Review erably if support by developed countries is reduced 22 (January): 75­96. substantially or eliminated altogether. Given the European Commission. 2000."Commission Analysis Paper: The low probability of eliminating support, however, Cotton Sector in the European Union." Brussels. ------. 2003. "Agricultural Reform Continued: Commission a second-best alternative would be for support to Proposes Sustainable Agricultural Model for Europe's be given in a nondistortionary manner. A type of Tobacco, Olive Oil and Cotton Sectors." Brussels. support with minimal distortionary effects--the Fang, C., and J. C. Beghin. 2003. "Protection and Comparative so-called decoupled support mechanisms--has Advantage of Chinese Agriculture: Implications for Regional and National Specialization." In D. Sumner and S. Rozelle, regained popularity recently. Income transfers eds., Agricultural Trade and Policy in China: Issues, Analysis under decoupled mechanisms are based on past and Implications. Aldershot, U.K.: Ashgate Press. production and prices and thus have no effect on FAPRI (Food and Agricultural Policy Research Institute). 2002. "The Doha Round of the World Trade Organization: Liber- current production decisions. What makes decou- alization of Agricultural Markets and its Impact on Develop- pled support in the cotton sector an interesting ing Economies." Paper presented at the IATRC Winter (and potentially applicable) alternative is that Meetings, San Diego, CA. Gibbon, P. 1999. "Free Competition without Sustainable Devel- almost all support comes in the form of domestic opment? Tanzanian Cotton Sector Liberalization, 1994­95 measures. Therefore, changing the nature of sup- to 1997­98." Journal of Development Studies 36:128­50. port does not require changing the sources of fund- Goreux, L., and J. Macrae. 2003. "Reforming the Cotton Sector ing, as it would in the case of border measures. in Sub-Saharan Africa." Africa Region Working Paper Series 47. World Bank, Washington, D.C. http://www.worldbank. org/afr/wps. Notes Huang, J., S. Rozelle, and M. Chang. 2004. "The Nature of Dis- tortions to Agricultural Incentives in China and Implica- 1. A more detailed version of this chapter is presented in tions of WTO Accession." In D. Bhattasali, S. Li, and W. Annex 14 of the attached CD-ROM. Martin, eds. China and the WTO: Accession, Policy Reform, 2. Fibers include a wide variety of products that can be and Poverty Reduction Strategies. Washington, D.C.:World divided into two broad categories: natural and man-made. Nat- Bank, 2004, pp. 81­97. ural fibers can be further divided into fibers of plant origin ICAC (International Cotton Advisory Committee). 2001. Survey (such as cotton and linen) and fibers of animal origin (such as of the Costs of Production of Raw Cotton. Washington D.C. Cotton: Market Setting, Trade Policies, and Issues 273 ------. 2002. Production and Trade Policies Affecting the Myers, D., and S. Stolton. 1999. Organic Cotton: From Field to Cotton Industry. Washington, D.C. Final Product, London: ITDG Publishing . ------. 2003a. Production and Trade Policies Affecting the Poulton, C., and others. 2004. "Competition and Coordination Cotton Industry. Washington, D.C. in Liberalized African Cotton Market Systems." World Devel- ------. 2003b. World Textile Demand. Washington, D.C. opment 32 (3): 519­36. ------. Various Issues. Cotton: Review of the World Situation. Pray, C., D. Ma, J. Huang, and F. Qiao. 2001. "Impact of Bt Washington, D.C. Cotton in China." World Development 29: 813­25. Kähkönen, S., and H. Leathers. 1997."Is There Life after Liberal- Quirke, D. 2002. Trade Distortions and Cotton Markets: Implica- ization?: Transaction Costs Analysis of Maize and Cotton tions for Global Cotton Producers. Canberra: Cotton Research Marketing in Zambia and Tanzania." Prepared for the U.S. and Development Corporation, Centre for International Agency of International Development. Center for Interna- Economics. tional Reform and the Informal Sector, University of Sabune, J. 1996."Experiences of Uganda with Privatization of its Maryland, College Park. Cotton Industry." Paper presented at the 55th Plenary Meet- Karagiannis, G. 2004. "The EU Cotton Policy Regime and the ing of the International Cotton Advisory Committee, Implications of the Proposed Changes for Producer Wel- Tashkent, Uzbekistan. fare." Report prepared for the Food and Agriculture Organi- Shepherd, A. W., and S. Farolfi. 1999. Export Crop Liberalization in zation. University of Macedonia, Greece. Africa: A Review. Rome: Food and Agriculture Organization. Karagiannis, G., and C. Pantzios. 2002. "To Comply or Not to Shuhua, G. 2003. "Introduction to China's Cotton Futures Trad- Comply with Policy Regulation--the Case of Greek Cotton ing." Cotton Outlook, Special Feature--China: The Future. Farmers: A Note." Journal of Agricultural Economics, 53: Liverpool: Cotlook Limited. 345­51. Tokarick, S. 2003."Measuring the Impact of Distortions in Agri- Larsen, M. N. 2002. "Is Oligopoly a Condition of Successful Pri- cultural Trade in Partial and General Equilibrium." IMF vatization? The Case of Cotton in Zimbabwe." Journal of Working Paper, WP/03/110. International Monetary Fund, Agrarian Change 2:185­205. Washington, D.C. Lundbæk, J. 2002. Privatization of Cotton Sub-Sector in Uganda: Townsend, T., and A. Guitchounts. 1994. "A Survey of Cotton Market Mechanisms and Institutional Mechanisms to Over- Income and Price Support Programs." Proceedings of the come These. Masters thesis, The Royal Veterinary and Agri- Beltwide Cotton Conferences. National Cotton Council of cultural University, Department of Economics and Natural America, Washington, D.C. Resources, Copenhagen, Denmark. Uzbekistan. 2003. "Uzbekistan: Cotton Policy Note." Govern- MacDonald, S.A. 1997. "Forecasting World Cotton Prices." In mental publication. Processed. Debra E. Gerald, ed., Federal Forecasters Conference--1997, WTO (World Trade Organization). 2002. United States-- Papers and Proceedings. National Center for Education Subsidies on Upland Cotton: Request for Consultations by Statistics. http://nces.ed.gov/pubs98/98134.pdf (accessed Brazil. WT/DS267/1. Geneva, October 3. December 9, 2002). Zhong, F., and C. Fang. 2003."China's Cotton Policy." Center for Minot, N., and L. Daniels. 2002. "Impact of Global Cotton Agricultural and Rural Development, Iowa State University, Markets on Rural Poverty in Benin." International Food Pol- Ames, IA. Processed. icy Research Institute, Washington, D.C. 15 SEAFOOD: TRADE LIBERALIZATION AND IMPACTs ON SUSTAINABILITY Cathy A. Roheim In many ways, fish as a commodity is treated differ- grams per year in China and to 25.8 kilograms per ently from agricultural products. For one thing, it is year in Southeast Asia, whereas it will remain con- not part of the agricultural negotiations of the stant or decline in developed countries (Delgado World Trade Organization (WTO). That it contin- and others 2003). ues to be treated as an industrial product in negoti- The goal of this chapter is to present the struc- ations may be a mixed blessing--leading to lower ture and important features of the global seafood trade protection but less discipline on domestic market, including illustrations of the complexities subsidies. Yet fish is the most important source of of the market, followed by a discussion of the protein for many around the globe. Seafood consti- impacts of trade liberalization, with a particular tutes the biggest category of food and agriculture focus on developing countries. Developing coun- exports from developing countries, at an annual tries play a very important role in international average of $33 billion (2000­01), or 18 percent of seafood trade (FAO 2002b). Many rely on seafood exports--more than combined exports of coffee, for export earnings--among them the Maldives, cocoa, tea, spices and nuts, cotton, and sugar. Glob- Mozambique, Peru, Senegal, and Sierra Leone. ally, per capita consumption of fish is estimated at Fisheries production, both caught and farmed, has 14.3 kilograms per year (Delgado and others 2003). doubled in the last 30 years, and most of that Per capita consumption in 1997 was led by Japan, increase has come from developing countries. Over with 62.6 kilograms per year, and China, at 26.5 half of global fish exports by value come from Latin kilograms per year (up from 8.1 in 1985). The America and the Caribbean and the developing European Union (EU) consumes 23.6 kilograms nations of Asia and Africa, and the majority of that per year per capita, and Southeast Asia 23 kilo- production goes to developed nations. grams, up from 19.8 in 1985. By 2020 per capita With rapid growth in production and trade have consumption of fish is expected to rise to 35.9 kilo- come the overexploitation of fish stocks and a rapid The author would like to thank Mirvat Sewadeh for her contributions to the trade policies and trade flows and Baris Sivri for his help with the data analysis. The CD-ROM accompanying this volume contains a longer version of this chapter with additional figures, tables, and boxes and an expanded narrative on fish and aquaculture production, value, trade, and policy matters. 275 276 Global Agricultural Trade and Developing Countries expansion of aquaculture. Both have had severe Thus, with the combination of capture fisheries impacts on the environment (FAO 2002c). Thus, and aquaculture, the volume of world production the issue of trade liberalization in seafood markets has doubled in the last 30 years. Most of the growth relates directly to sustainability of fish production in aquaculture is occurring in developing countries, and, by implication, the sustainability of interna- especially China, where it is destined predominantly tional trade in fish products. for domestic consumption. Marine aquaculture has grown very slowly in developed countries, largely because of limited available shoreline. Production China is the world's largest producer of captured Production of fish (finfish, mollusks, and crus- fish, marine and inland, at 17 million tons (fig- taceans) takes two forms, aquacultured (or farm ure 15.1). Peru and Chile follow, primarily captur- raised) and captured. The vast majority of captured ing anchoveta, largely used to produce meal and oil fish (by volume) are marine, while the majority of for industrial use. U.S. fleets catch large volumes of aquacultured fish are freshwater species. The fishing low-value pollock off Alaska. Most of the catch goes sector has expanded considerably in the past into surimi, a refined, stabilized fish protein concen- 50 years, with capture fisheries landing 19 million trate used in making imitation crab meat and (metric) tons in 1950 to 98 million tons in 2000 processed fish such as breaded fish sticks and patties. (FAO 2002b). During this time the importance China is the leading producer of carp. The of developed countries in the fishing sector has majority of that harvest is retained for domestic declined relative to the developing nations because consumption. Norway, Chile, Scotland, Ireland, of overfishing of waters contiguous to developed and Canada are the leading producers of farmed countries and an increase in fishing in the develop- salmon and trout, and most of that production is ing world. Aquaculture has further expanded traded on the international market. the seafood industry, increasing production from China and Thailand produce almost 50 percent 2.5 million tons in 1970 to more than 35 million tons of the world's supply of shrimp (figure 15.2), with in 2000,with most of the increase occurring in devel- other developing countries supplying most of the oping nations (FAO 2002b). Production of carp and rest. Shrimp and prawns account for just 6.4 per- mollusks dominated aquaculture production during cent of the volume of the world fish trade but about the 1990s, but shrimp have the highest value. 20 percent of its value (OECD 2003a). The global FIGURE 15.1 Fish Catches by Leading Countries, 1991­2000 (millions of metric tons) Millions of metric tons 45 40 35 30 25 20 15 10 5 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 United States Chile Japan Peru China Source: FAO. Fishstat Database. Seafood: Trade Liberalization and Impacts on Sustainability 277 FIGURE 15.2 World Aquaculture Production generally agreed, however, that this rapid expansion of Shrimp, by Volume, 2000 has had considerable environmental costs. The area Brazil 2% Other 9% under shrimp culture tripled in 10 years, from the mid-1980s to the mid-1990s, covering 130,000 hec- Mexico 3% tares by 1999 (UNEP 1999a). In the process, man- Philippines 4% groves have been removed and replaced by coastal Ecuador 5% ponds. The ponds have increased the salinity of adja- Thailand 28% India 5% cent land, jeopardizing its future productivity. The costs of restoration would likely be very Bangladesh 5% high. Disappearing mangroves have deprived the marine ecosystem of valuable habitat and nursery Vietnam 6% areas for fish reproduction. In addition, sustainable Indonesia 13% China 20% shrimp farming is threatened by its reliance on the Source: FAO. Fishstat Database. collection of wild shrimp fry, which are then "grown out" to appropriate sizes for export, a prac- tice that threatens the sustainability of wild shrimp stocks as well. Disease sometimes breaks out in shrimp trade is valued at more than $10 billion shrimp ponds and may spread to the wild shrimp annually. population. Finally, the feed for cultured shrimp is It is estimated that more than 60 percent of based on fish meal, which is produced from fully Asia's mangroves have already been converted to used, if not overused, stocks of anchovies, herring, aquaculture farms, primarily for the production of menhaden, and sardines. shrimp (ESCAP and ADB 2000), degrading habi- Recognizing the negative externalities caused by tats and land. Because shrimp is such an impor- shrimp culture in Bangladesh, the U.N. Environ- tant export earner for Southeast Asia and South ment Programme (UNEP) recommended that ef- America and has such a marked negative effect on fective environmental policies with proper enforce- their environments, it is worthwhile to discuss its ment should be implemented to ensure that trade production in some detail. liberalization did not lead to externalities that re- duced overall welfare (UNEP 1999a). Sustainable Shrimp Aquaculture in Bangladesh Thailand is the world's largest producer of and Thailand shrimp, with approximately 23,413 farms covering Subsistence fishermen have caught shrimp in an area of 72,663 hectares (in 1996) (FAO 2000). By Bangladesh for hundreds of years. But since the 2000, Thailand was exporting 249,638 metric tons mid-1980s the cultivation of shrimp for export has of shrimp, valued at some $2.7 billion, to the world grown significantly. In 1972­73 exports of captured market. Shrimp production in India, Indonesia, shrimp were valued at $2.9 million. By 1985 ex- and Vietnam combined equal what Thailand pro- ports had growth to $90 million, primarily from duces in export value. The same environmental aquaculture. In 2000 the figure was $330 million issues highlighted for Bangladesh apply to (FAO 1999, 2002b). Thailand--satisfying the huge export market for Some of the credit for this rise goes to a struc- cultivated shrimp has led to significant environ- tural adjustment program in which Bangladesh mental damage. received a World Bank loan of $1.76 billion over the period 1979­96 (UNEP 1999b). Under the pro- The Shrimp Industry in Madagascar gram, policies that had limited trade were replaced with new policies that encouraged exports. The Madagascar's shrimp industry is the country's lead- changes created an environment in which private ing foreign exchange earner. Exports grew from $20 investments in shrimp culture, shrimp processing, million in 1980 to $102 million in 1999 and now and shrimp exports flourished. account for 7 percent of gross domestic product Shrimp now accounts for almost 91 percent of (GDP). Approximately one-half of the shrimp pro- fish exports from Bangladesh (FAO 2002b). It is duced are from capture fisheries, the other half 278 Global Agricultural Trade and Developing Countries from aquaculture. The industry provides direct Pacific, and other species such as menhaden and employment for approximately 53,000 people and herring found around the globe. In processing the indirectly for another 30,000 people (World Bank fish are cooked, pressed, dried, and milled. The 2003). dry remainder is fishmeal; oil is extracted from In the shrimp capture industry, there are three pressing. types of fisheries: traditional, artisanal, and indus- Fishmeal and fish oil, used in animal feeds (for trial. The bulk of employment occurs in traditional both terrestrial livestock and aquacultured fish) but fisheries, in which fishers have no motorized equip- not for human consumption, are industrial prod- ment. Entry into the fishery is open; no license is ucts. Demand for fish meal from the farmed fish required. Most of the catch of traditional fishermen industry has increased dramatically in the last is consumed domestically. Production was about 20 years. Growing poultry and pig industries in 3,400 tons in 2000 (World Bank 2003). China and Southeast Asia also create strong de- The cost of the license required to ply the arti- mand for fish meal. sanal fisheries depends on the power of the fishing The primary producer of fish meal has long boat's motor. Most artisanal boats belong to a com- been Latin America, with a total of 2.8 million tons pany rather than being individually owned. Indus- produced in 1997 and an annual growth rate of trial trawlers that fish in Madagascar's waters are 1.7 percent between 1985­97 (Delgado and others mostly foreign owned and have processing facilities 2003). Much of that production, from Chile and on board. In 2000, approximately 8,200 tons of Peru, is susceptible to the vagaries of El Niño. The shrimp were captured by artisanal and industrial most heavily exploited fish is the Peruvian fisheries, which directly employed some 10,500 anchoveta (figure 15.3). World production in 1997 people. Virtually all of the shrimp captured in these was 6.1 million tons, with the balance after Latin two fisheries are exported, with France and Japan America made up primarily by China, Southeast being the primary markets (World Bank 2003). Asia, Japan, and the European Union. If world markets and production do not change substantially over the next 17 years, fish meal prices Industrial Products are projected to rise by 18 percent (Delgado and Developing countries are important exporters and others 2003). Conversely, if aquaculture expands by importers of fish meal. Fishmeal and oil are derived 50 percent, then the price of fish meal will increase from small, wild-caught pelagic fish such as capelin by 42 percent. Greater efficiency in the use of fish from the North Atlantic, anchovies from the South meal in animal feed could push prices down. In the FIGURE 15.3 World Aquaculture Production by Value, 1991­2000 US$ billions 40 35 30 25 20 15 10 5 0 1991 1995 2000 Carps Salmon Trout Shrimp Tilapia Mollusk Source: FAO. Fishstat Database. Seafood: Trade Liberalization and Impacts on Sustainability 279 worst-case scenario, in which the world experiences Fish and fish products have not always been an ecological collapse in fisheries yielding fishmeal major internationally traded products. Several and oil, the price will rise by 134 percent (Delgado influences led to the rapid expansion in interna- and others 2003). Any of these potential price tional trade beginning in 1975. Certainly the pass- changes would dramatically affect livestock pro- ing of the International Law of the Sea and the duction in developing countries. institution of the 200-mile EEZ in 1977 had a large impact. The establishment of the EEZs effectively created importers out of countries, such as Japan, International Markets with very large distant water fleets, and created ex- Fish is one of the most traded food commodities in porters out of those countries, such as the United the world. The value of world imports of fish prod- States, that had large marine resources and rela- ucts was $60 billion in 2000, greater than interna- tively low domestic demand. tional trade in many agricultural products (fig- The most important trade commodities in order ure 15.4). The most valuable component of seafood of their value in 2000 are shrimp ($10.8 billion), trade is shrimp, with world trade in 2000 valued at salmon and trout ($5.2 billion), tuna ($4.8 billion), more than $10 billion (FAO 2002b). groundfish ($4.4 billion), crabs and lobsters A myriad of issues underlies fisheries and aqua- ($3.8 billion), mollusks ($2.8 billion), cephalopods culture production. Capture fisheries still supply ($2.7 billion), fish meal ($2.1 billion), small pelag- the majority of fish production, but fully 60 percent ics ($1.6 billion), large pelagics ($1.1 billion) and of the world's fisheries are already being fished at or flatfish ($1.1 billion) (Anderson 2003). over capacity (Grainger and Garcia 1996). Even Thailand is the world's top exporter of food fish with the establishment of 200-mile exclusive eco- in the world, followed by China, Norway, and the nomic zones (EEZs) in 1977, which brought one- United States (see figure 15.4). Seventy-four per- third of the world's oceans under the jurisdiction of cent of Africa's exports are destined for the coastal states, most fisheries management plans European Union, while exports from Central and have not achieved their stated goal of maintaining South America go primarily to the United States, sustainable fisheries. Many countries, mostly devel- Canada, and the European Union. oping, do not have management policies or lack The major importing nations are the European resources to enforce them. Union ($19.5 billion), Japan ($15.5 billion), and the FIGURE 15.4 Food Fish Exports by Top Countries, 2000 US$ billion 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 China Chile Spain Thailand Norway States CanadaDenmark (China) Indonesia United Taiwan Source: FAO 2002b. 280 Global Agricultural Trade and Developing Countries FIGURE 15.5 World Food Fish Exports by Institutional Influences on International Value of Major Commodity Trade in Fishery Products Group, 2000 Even though most caught, farmed, and traded fish Crustaceans and Fish, fresh, chilled or are clearly food products, no fish is included in the mollusks, canned 7% frozen 45% WTO Uruguay Round Agreement on Agriculture Fish, canned 10% (URAA). The concern among some nations is that fishing as an industry involves not only market access, but also resource access on a scale unprece- dented in other areas of agriculture. Therefore, negotiations regarding trade liberalization for fish have proceeded far differently from those on agri- cultural commodities. Crustaceans and mollusks Fish, dried, salted or Tariffs on fish products, in contrast to those on 33% smoked 5% agricultural products, have been reduced with Source: FAO 2002b. every successive trade round. And international agreements on sanitary and phytosanitary meas- United States ($10.4 billion). Within the European ures, technical barriers to trade, antidumping, rules Union, imports go to Spain ($3.35 billion), France of origin, import licensing, and safeguards have ($3.0 billion), Italy ($2.5 billion), Germany ($2.3 all been applied to trade in fish. Subsidies in the billion), the United Kingdom ($2.2 billion), and fishing industry fall under the GATT (General Denmark ($1.8 billion) (FAO 2002b). Agreement on Tariffs and Trade) Agreement on Crustaceans account for 19 percent of the Subsidies, whereas in agriculture they fall under the weight of exports but 33 percent of the value. Fin- URAA. fish, by contrast, contribute 63 percent of volume This section of the chapter discussed the domes- but only 45 percent of value (figure 15.5). The most tic and international policies and institutions most widely traded processed seafood products are items relevant to global trade in fish and fish products. such as canned tuna, canned crab and lobster The domestic policy interventions are fisheries meats, canned herring and sardines, roe (such as management policies, fishing subsidies, and trade caviar), shelled and deveined shrimp, and dried or barriers, including tariffs, technical barriers to salted finfish. trade, sanitary and phytosanitary measures, and Significant reexporting occurs in the world antidumping and countervailing measures. seafood markets. Thailand, for example, imports a significant amount of the world's tuna catches, Fisheries Management Policies processes it into cans, and then exports it. Similarly, China is a major reprocessing market for U.S. and To fully understand the impact of trade liberaliza- Norwegian seafood. tion on fishery products, one first must understand For low-income, food-deficit countries, exports the factors influencing supply. The impacts of trade are far larger in value than imports. When fish meal liberalization will differ depending on several fac- and oil are excluded from export values, the picture tors, including production method (capture or changes only slightly, since their value is not high aquaculture) and domestic fisheries management and many of these countries do not participate in policies. fish meal or fish oil production. Among the devel- Fish in capture fisheries belong to a common oping countries that rely on exports of seafood as pool. Before 1977 jurisdiction of most nations over a primary source for export earnings are the fishing grounds extended only 12 nautical miles Maldives, Mozambique, Peru, Senegal, and Sierra from shore. Expansion to 200-mile EEZs was dis- Leone (FAO 2002b). Thus, reductions in the stocks cussed and agreed to in 1977 by nations at the of fish in developing countries because of poor Third Law of the Sea Convention (UNCLOS-III, management have the potential to jeopardize the 1973­1982) (Hannesson 1996). EEZs cover 40 per- food supply while reducing household incomes and cent of the world's oceans and 90 percent of its liv- export earnings. ing marine resources (Deere 2000). Seafood: Trade Liberalization and Impacts on Sustainability 281 UNCLOS assigns the exclusive right to coastal Output controls are designed to limit directly states to manage and exploit marine living resources the volume of the catch from a given fishery. The and to regulate fisheries resources through a com- critical necessity for this form of management is prehensive management system. There is consider- the ability to monitor the catch. In some fisheries, able debate over the effectiveness with which coastal managers may have personnel at the dockside to nations have managed their EEZs with respect to count the number (or weight) of fish caught as they the sustainability of production. Creating an EEZ are landed. In other cases, on-board observers may does not remove the common-pool property of the monitor the catch. In either case, once the total resource; it simply redistributes the use of the allowable catch (TAC) is reached, the fishery is gen- resource to new (domestic) market entrants. Many erally closed for the season. nations, especially developed nations, encouraged Management by TAC has at least three short- expansion of the domestic fleet to increase the comings. First, it induces fishers to compete to catch national capacity to catch fish that foreign nations as much as possible before the TAC is reached. Sec- would have caught in the past. Catches quickly grew ond, as fishers become more intensively capitalized, as the number and size of fishing boats increased. In the TAC is reached in a shorter time, leading to a ensuing years, however, supplies in many fisheries backlog of fish for processors that pushes down decreased drastically as fish stocks were reduced fishermen's prices and reduces product quality. beyond the sustainable limit and the remaining fish Third, idle vessels may move to fish in another fish- became harder to find. ery, leading to overcapitalization in yet additional An often-quoted statistic is that fully 60 percent fisheries (Conrad 1999). of the world's major fisheries resources are already The management systems of fisheries have being exploited at or above capacity (Grainger and caused vexing trade and environment issues, and Garcia 1996). Fish stocks in OECD (Organisation several cases have landed before dispute panels dis- for Economic Co-operation and Development) cussions with the WTO and format GATT. The dis- countries, in particular, have been subject to large putes below (drawn from Robb 2001) were directly fishing pressure over the years and are mostly over- related to fisheries management policies. fished (OECD 2003b). The problem derives not from a lack of regulations per se, but rather from a · Canada v. U.S.--Prohibition of Imports of Tuna lack of effective regulations. and Tuna Products from Canada, 1982. In a fishery where there is no restriction on · Canada v U.S.--Measures Affecting Exports of entry into fishing, the management system (or lack Unprocessed Herring and Salmon, 1988. of it) is referred to as open access. It is well known · Mexico v. U.S.--Restrictions on Imports of Tuna from economic theory and experience that open (Tuna/Dolphin I), 1991. access will lead to overexploitation of the fish stock, · European Economic Community and Netherlands as individual fishermen have little incentive to v. U.S.--Restrictions on Imports of Tuna (Tuna/ restrain their fishing efforts to promote a sustain- Dolphin II), 1994. able fishery, because the fish forgone by one fisher- · India, Malaysia, Pakistan, and Thailand v. U.S.-- man will simply be captured by someone else Import Prohibitions of Certain Shrimp and (box 15.1). Most fisheries in the United States and Shrimp Products--1998. European Union operate under some form of lim- ited access or limited harvest. Because of poor management and other factors, Management policies can be categorized as the status of fish stocks worldwide is alarming. The being either an input or output control. Input con- implications of trade liberalization for capture fish- trols, the oldest type of fishery management tool, eries are many, but the most obvious implication is are designed to limit either the number of people that the current level of catches from capture fish- fishing or the efficiency of fishing (National eries is unsustainable. Should trade liberalization Research Council 1999). Input controls generally provide incentives to fishermen to catch even more lead to inefficient outcomes. They raise the cost of fish, it would simply speed up the overfishing and fishing but generally do not reduce effort or depletion process, leading to unsustainable interna- capacity. tional markets as well. This is not to say that further 282 Global Agricultural Trade and Developing Countries BOX 15.1 Impacts of Trade Liberalization in Uganda's Fishing Industry Economywide liberalization and reforms in ing practices are on the rise, as the catch of Uganda's trade regimes have made the fisheries native fish has declined. For example, exotic industry one of the country's most important in species are being introduced to lakes and rivers. terms of employment and export earnings. The In another example, poisons are being used, ille- fisheries sector is Uganda's second-largest gally, to stun the fish, bringing them to the sur- national export producer, with export values face, making them easy to scoop up in nets. The growing from $1.4 million in 1990 to $78 mil- poisoning has led the European Union to impose lion by 2001 (UNEP 2002a). More than 1 million a ban on fish exports from Uganda due to food workers are directly engaged in harvesting, safety concerns. transporting, processing, distributing, and mar- Other issues related to food safety include a keting fish (UNEP 1999b). lack of refrigeration facilities to preserve fish after With success have come problems common to harvest. Transportation to processing facilities is fishing industries elsewhere. Uganda's fish come made difficult and slow by poor road conditions, from the country's many lakes and rivers. Current further degrading the quality and safety of the legislation allows open access to lake fishing. fish prior to processing. There are relatively few restrictions on who may Other environmental concerns include efflu- fish, and few technical measures to control fishing ent pollution from fish-processing industries. mortality. Poor data make it difficult to determine Raw, untreated waste is dumped directly into the amount of fish that can be taken without the very rivers and lakes from which the fish are depleting the stocks beyond a sustainable level, being pulled, contaminating the environments particularly in Lake Victoria, which borders not for tomorrow's catch. only Uganda but also Kenya and Tanzania. Thus it Social problems also threaten the fishing has been difficult to establish harvest limits. The industry, as most of the products are destined U.N. Environment Programme recommends that for export markets, where they fetch higher Uganda should determine the level of fish stocks prices. Much of the local population can afford it currently has, establish a total allowable catch only fish rejected by processors for the export that is in line with sustainable harvests in each of market. Food security concerns have been the major water bodies, and implement an indi- raised, as well, as Nile perch feed heavily on vidual transferable quota system. freshwater shrimp that are also caught and used Overfishing is not the only problem in as animal feed. Uganda. According to UNEP, unsustainable fish- Source: UNEP 1999b. liberalization should not occur, but rather that trade-distorting subsidies" in the fisheries sector overfishing and other externalities must be consid- (WTO 1999, 2000, 2001). In 2001, at the Fourth ered in free-trade discussions. Ministerial Conference in Doha, Qatar, the WTO explicitly included fisheries subsidies in the negoti- ating agenda to improve current discipline on Fishing Subsidies subsidies--this as a result of discussions in the The most sensitive issue related to capture fisheries CTE. The Doha Declaration states that the need to before the WTO Committee on Trade and the Envi- "clarify and improve WTO disciplines on fisheries ronment (CTE) is fishing subsidies. Subsidies exist subsidies, taking into the account the impor- in the fishing sector globally and have come to be tance of this sector to developing countries" (WTO recognized as having a significant impact on the 2003a: 28). quantities of fish traded, largely because they lead to In an excellent review of fisheries subsidies, unsustainable fishing practices. At the WTO High- Schrank (2003: 49) cites three implications. Level Symposium on Trade and Environment in March 1999, five WTO member nations (Australia, Three implications are noted: (1) countries that Iceland, New Zealand, the Philippines, and the do not subsidize and that restrain total catch to United States) submitted a joint statement on the maintain the resource lose the extra catch to need to eliminate "environmentally damaging and countries that subsidize and do not restrain total Seafood: Trade Liberalization and Impacts on Sustainability 283 catch; (2) competition from subsidized distant nations for a total value of more than 400 million. water fleets can make it economically unviable The countries with the largest negotiated fees in for developing countries to develop their own 2000 were Morocco (114 million), Mauritania fisheries and therefore to realize the benefits of (54 million, box 15.2), Argentina (16 million), their own 200-mile zones of fishery jurisdiction; Angola (13 million), and Senegal (12 million) (3) subsidies can contribute to stock depletion, (OECD 2003a). The primary beneficiaries of the with negative economic, trade, and environ- access agreements are Spain and France. Portugal, mental effects for other countries that have an Italy, and Greece have also benefited. interest in the stock. The agreements are very controversial. Fishing access agreements have been seen as a way to reduce The greatest contrast to agricultural subsidies is capacity in the European Union while securing the effect noted in Schrank's first point. Fishing employment and supplies of fish for the European subsidies create not only a trade distortion in the market (Institute for European Environmental Pol- markets, but also, in the case of straddling or icy 2002b). On the environmental side, catch limits migratory fish stocks, a negative externality on the are either not imposed on the foreign fleets or the nation competing to capture the fish. limits are not enforced, and so the sustainability of The relationship between fisheries subsidies and stocks of fish in accessed waters is in doubt in many their environmental and social impacts is obviously nations. complex. According to Hussein Abaza, who heads UNEP's Economics and Trade Branch, "It is Trade Barriers: Tariffs becoming clear that developing countries stand to gain a great deal from trade in fisheries products, Tariffs in OECD member countries are important but only if trade and fisheries policies are reformed barriers to the developing nations that export to to support sustainable management of these them. But a good deal of South-South trade is also resources" (UNEP 2002a). The policy recommen- affected by tariffs.1 dation is simple--eliminate trade and domestic Tariffs on seafood in developing countries are distortions while adopting environmental policies generally higher and more transparent than those that address overfishing. But the implementation in OECD countries. The structures of the tariff of sound environmental management is the real regimes, however, differ considerably among devel- policy challenge. oping countries. Among developing countries, Thailand has the highest tariff levels on seafood products (60 percent across all product forms), fol- Fishing Access lowed by India, whereas Chile and Malaysia gener- In a form of fishing subsidies, the European Union ally apply the lowest duty rates. Yet all developing signed its first fishing access agreement with countries for which detailed tariff schedules are Senegal in 1979, shortly after nations exercised available implement transparent tariff structures their rights to the 200-mile EEZ. Since many devel- with all product lines subject only to ad valorem oping nations with EEZs did not have the capacity duties. to make use of their resources, they opted to sell After the Uruguay Round, average weighted access to these resources to third parties. The Euro- import tariffs on fish products in developed pean Union has been predominant in negotiating countries were reduced to around 4.5 percent (Lem these agreements on behalf of its member countries 2003). This average hides a number of tariff issues, and has been paying the access fees. however, including some tariff escalation and tariffs Most fishing access agreements have been on specific items (such as canned tuna in the United reached between the European Union and African States). The European Union and the Republic of countries and a few other nations. In these access Korea have the highest tariffs (ranging from 4 per- agreements, an amount is negotiated to guarantee cent to 33 percent), whereas the United States and access to foreign waters by portions of the EU Canada have the lowest (0­5 percent) (figure 15.6). industrial fishing fleet. During 1999­2000, the But despite their relatively high tariffs,both Korea European Union had agreements with 20 different and the European Union have very transparent tariff 284 Global Agricultural Trade and Developing Countries BOX 15.2 Foreign Fishing Access Agreements Involving Mauritania The fishing sector in Mauritania accounts for In response to critics, the European Union has more than 40 percent of exports and about begun to increase the value of the access pay- 6 percent of gross domestic product. The only ments (for Mauritania up 61 percent over the major export items are squid and octopus, with previous agreement) and to work toward agree- an export value of $68 million in 2000 (FAO ments that promote sustainable development of 2002b). Only $639,000 were fish products the fisheries in the target nations. To that end, exported in processed form and that was for the agreements, by design, allow the Mauritan- dried, salted, or smoked products. Total fish ian authorities to inspect and control fishing product exports were $74 million. activities--requiring a daily log of catches by the The primary source of earnings from the fish- foreign vessels and setting up a system of ery sector in Mauritania is not from exports but observers on board vessels. These opportunities from access fees. The European Union pays for for Mauritania are not fully taken advantage of. its fleet to fish in Mauritanian waters. In a sense Restricted fishing zones have increased in size. one might say that Mauritania exports its fish re- There remain no catch limits. sources, while they are still in their habitat, di- Determining economic benefits for either rectly to the European Union fishing fleet. Eighty party to the agreement is uncertain, as there is percent of fish in Mauritania, or 450,000 tons, little information on catch statistics. However, were landed by industrial vessels in 2001 (WWF based on the previous agreement between the 2003). A new agreement on fishing access by European Union and Mauritania, for each euro The European Union was enacted in 2001 and paid to Mauritania in 1996, the value of the is effective until August 2006. The European catch was two times greater. In 1997, the value Union is paying 430 million, creating access of the catch was three times greater than the to Mauritanian water for 248 vessels, target- cost of access. Little of the access money ing hake, squid, crawfish, and tuna. The EU appears to be utilized to build within Mauritania vessels are predominately from Spain and a domestic infrastructure to nationalize its France, but also from Italy, Portugal, Greece, the resources rather than selling foreign access. In Netherlands, Germany and, to a minor extent, addition, reports from nongovernmental organi- Ireland. zations, such as the World Wildlife Fund, indicate In addition to the access fees, vessel owners that the agreements have negative effects are required to pay 29 per ton of catch taken on local communities and on sustainable by freezer tuna seiners, and 19 per ton for development. catches from pelagic fish trawlers. A license fee is Source: Institute for European Environmental also payable, based on tonnage per year in some Policy (2002a and b); cases and a flat annual fee for tuna vessels. www.integratedframework.org. structures.All tariffs applied on seafood products are The European Union offers free access to all ad valorem duties. In comparison, Japan and the seafood products from the least-developed coun- United States implement more complex tariff struc- tries and partial tariff exemption to most of tures. In Japan about 20 percent of the tariff lines on seafood exports from Africa-Caribbean-Pacific intermediate seafood products are either per-unit- (ACP) countries and other developing countries. specific or compound duties. Similarly, 38 percent of The United States grants free access for all develop- U.S.tariff lines on intermediate seafood products are ing countries for all seafood products. Japan also per-unit-specific or compound. The U.S. tariffs do grants free access to some seafood imports from the not seem to be aimed at concealing protection, since least-developed countries and maintains only one their average ad valorem equivalent is only a little seafood tariff line for other developing countries more than 2 percent. At the same time, the products (table 15.1). that receive tariff protection in the United States, Table 15.1 shows that the trade-weighted tariff such as canned tuna, are protected only through averages across the OECD countries exhibit some high ad valorem tariffs. trade escalation for imports from developing coun- Most industrial countries offer preferential tries and all other countries, but not for the least- access to developing countries' seafood exports. developed countries. However, in the context of Seafood: Trade Liberalization and Impacts on Sustainability 285 FIGURE 15.6 Tariff Structure by Level of Processing, (1998­2001) Average of all Tariffs 20 15 10 5 0 Canada United States Japan EU Korea, Rep. of Raw Inter Final Source: WTO IDB database. TABLE 15.1 Trade-Weighted Tariff Averages for Developing Countries' Fish Product Exports to OECD Countries, by Processing State (percent) Level of Processing Least-Developed Developing All Other All 2.5 2.9 3.2 Unprocessed 2.5 2.5 2.5 Fillets 2.8 2.5 2.0 Semi-processed 0.5 1.9 1.4 Processed 1.7 4.3 8.0 Total value (US$ millions) 437 10,689 21,992 Source: OECD 2003a. tariffs on agricultural goods, tariffs in seafood A great deal of regulatory activity concerning products are lower and the level of tariff escalation country-of-origin labels is occurring in the United is very moderate. States and European Union. Ecolabeling and organic labeling are voluntary programs, but the WTO is interested in whether such labels constitute Trade Barriers: Technical Barriers a nontariff trade barrier. Currently, these labels are In recent years, there has been a large increase in not considered to be trade barriers as long as they policies that could potentially come under the are nondiscriminatory (WTO 2003b). heading of technical barriers to trade. Among them are labeling programs and the resultant tracing Trade Barriers: Sanitary and Phytosanitary capability they require. The programs are typically Measures found in developed countries but can have poten- tially large impacts on developing countries. Import regulations based on hazard-analysis, Among the labeling programs are ecolabeling, critical-control-point (HACCP) principles, adopted country-of-origin labeling, and other labeling by many of the major importing nations, are related to the production process, such as "organic." regarded as nontariff barriers by many developing 286 Global Agricultural Trade and Developing Countries countries, as the investment required to bring pro- Kenya has adapted to the new realities by cessing plants up to code can be substantial (Filhol restricting the number of facilities handling fish to 2000). During 1997­98, the European Union im- be exported. Only five fishing villages (out of nearly posed bans on the import of seafood from India, 300) are authorized to handle fish landings. This Bangladesh, Kenya, Madagascar, Mozambique, causes fishermen from elsewhere to incur higher Tanzania, and Uganda, citing food safety concerns transportation costs to land their catch. both in processing and in possible contamination The costs of exporting to nations with strict prior to catch in both capture fisheries and aquacul- quality controls are not trivial, but Kenya has had to ture (Filhol 2000). incur those costs to remain in the international For example, most of the fish caught in Kenya market. As long as Nile perch continues to be in are from Lake Victoria; the majority of that catch is demand in the world market, it is likely that Kenyan Nile perch (FAO 2002b).2 Nile perch are also the producers can more than cover their costs. Should main export from Kenya, earning about $50 mil- the prices rise too far, however, other white-fleshed lion annually. Of the 18 fish processing and export- fish will become competitive substitutes. The inter- ing firms now in Kenya, 10 specialize in Nile perch national seafood market in white-fleshed fish is and 8 in marine products such as shrimp, other very competitive, particularly now that farmed crustaceans, and tuna. In 1997 the European Union tilapia and catfish are available in large quantities. became concerned about the safety of fish from Kenya when Spain and Italy both banned fish Trade Barriers: Antidumping and Countervailing imports because of the presence of salmonella. Measures Some other members of the European Union con- tinued to import from Kenya, but exports declined As tariff barriers have been relaxed and the aquacul- by 34 percent between 1996 and 1997. In 1998 the ture industry has boomed globally, more and more European Union banned imports of fish from fishing industries in the United States have found Kenya because of a cholera outbreak, causing a themselves competing with lower-priced imports. 66 percent drop in fish exports to the European Thus, the United States in particular has been quite Union. In 1999 the European Union banned fish active in pursuing antidumping and countervailing from Lake Victoria yet again, this time because of duty suits against foreign competitors. The United the presence of pesticides, causing another 68 per- States brought antidumping and countervailing cent decline in fish exports. In 1997 Kenyan exports charges against imports of Norwegian farmed were $52 million, in 1998 $39 million, and in 1999 salmon in 1990, Chilean farmed salmon in 1997, $32 million. In 2000 they were back up to $39 mil- crawfish from China in 1997, and farmed catfish lion (FAO 2002a). from Vietnam in 2003. A petition was filed with the In response to the requirement for a HACCP U.S. International Trade Commission (USITC) in program to export to many nations, Kenya has December 2003 against six exporters of farmed instituted stringent quality control procedures. The shrimp. Details on several of these cases follow. Fisheries Department controls quality though pro- visions of the Kenya Fisheries Act and the Fish Crawfish from China. The imported product was Quality Assurance Regulation of 2000. However, defined as freshwater crawfish tail meat in all its fish quality comes at a cost. There are strict regula- forms, grades, and sizes. China supplied 62 percent tions on production, handling, processing, packag- of all imports by the United States in 1997 and ing, and transportation of fishery products. In 92 percent in 2001. U.S. production of crawfish in addition, strict regulations govern construction of 1996 was 12.5 million pounds; in 1997, 23 million buildings, equipment, purification tanks, and stor- pounds (U.S. Department of Commerce 1997). age facilities. Costs were incurred to train workers Meanwhile imports of crawfish from China were in hygiene related to fish handling. There is also the 2.6 million pounds in 1996 and 5.8 million pounds additional cost of electricity to maintain strict tem- in 1998.3 The average value per pound of imports perature controls. Finally, the cost to fishermen is from China was $1.85 in 1997, compared with significant. They must invest in newer boats that $5.82 per pound for the domestically produced have chillers to maintain the quality of caught fish. product. As a result, antidumping duties of Seafood: Trade Liberalization and Impacts on Sustainability 287 223.01 percent were imposed (USITC 2003b). (SSA), a group of shrimp harvesters and processors However, Chinese crawfish continue to dominate in the United States, filed antidumping petitions imports to the U.S. market, with sales of 8 million with the USITC, alleging that Brazil, China, pounds in 2001 and 7.5 million pounds in 2002, Ecuador, India, Thailand, and Vietnam were dump- worth a total of $38.7 million in 2001 and $22.2 ing shrimp (primarily farmed) with an approximate million in 2002. The ruling was reviewed in 2003, annual value of $2.4 billion into the U.S. market. and it was determined that the antidumping duties The SSA is petitioning for tariffs on imports of should remain in place. shrimp from these countries ranging from 30 per- cent to 267 percent. It argued that "a variety of Catfish from Vietnam. In 2002, independent financial incentives provided by national govern- processors and the Catfish Farmers of America, a ments and international institutions over a number trade association of U.S. catfish farmers and proces- of years have overstimulated the infrastructure and sors, brought a petition to the USITC regarding production of farm-raised shrimp in these coun- dumping of frozen catfish fillets into the U.S. mar- tries" (emphasis added) (McGovern 2003). Thus it ket by Vietnam. Catfish farming is the largest aqua- seems that the investment by organizations such as culture industry in the United States. Production in the World Bank and others in helping build an 2000 was 150.6 million pounds (USITC 2003a). export industry in some of these countries is per- The primary producing states are Mississippi, ceived to have created unfair subsidies for these Arkansas, Louisiana, and Alabama. Prior to 1999 shrimp-exporting nations. imports were largely absent from the U.S. market. In 1999, Vietnam exported fewer than 2 million Impacts of Trade and Domestic pounds of what the Vietnamese call catfish into the Policy Reforms U.S. market. By 2001 that number had increased to 15.9 million pounds. Although the Vietnamese The previous section makes clear that the primary product was successfully labeled and marketed as trade barriers for capture and aquaculture fisheries catfish, the Latin names of the imported species are tariffs, countervailing and antidumping meas- were Pangasius bocurti, Pangasius pangasius, and ures, and the discriminatory potential of ecolabel- Pangasius micronemus. American catfish are from ing, country-of-origin labeling, and sanitary and the Ictaluridae family. phytosanitary measures for seafood safety. A problem in world markets for fish is that once To analyze the impacts of trade liberalization on fish is processed, it is very difficult to determine its trade in seafood, particularly on seafood from species. Some fish marketed as red snapper are not, developing countries, one must distinguish between in fact, red snapper, a high-value fish. Due to many the impacts of trade liberalization on seafood de- cases of intentional and unintentional fraud in rived from capture fisheries and on seafood from seafood markets, in which consumers were falsely aquaculture. This is because of their distinct attrib- led to believe that they were buying a certain prod- utes. Capture fisheries are generally ill-managed. As uct or confused by the same product being mar- such, changes in trade policies may create changes keted under different names, the U.S. Food and in welfare that differ between the short and long run Drug Administration has become more rigorous in because of the sustainability of fish stocks. With its regulations of appropriate names for fish. respect to the effects of trade liberalization, aquacul- Vietnam is now required to label its fish not as ture is more similar to agriculture. However, to the catfish, but instead as basa and tra.1 However, extent that aquaculture is dependent on feed de- Vietnamese basa and tra are still considered similar rived from capture fisheries or seed stock from wild enough to American catfish to be subject to fisheries, trade liberalization may have a different antidumping measures. Producers have had to pay effect on aquaculture than on agriculture. antidumping duties of between 36.84 percent and Whereas the research literature on markets for 63.88 percent. fish is extensive (Wessells and Anderson 1992; Kinnucan and Wessells 1997), there has been little Shrimp from some developing countries. As of empirical analysis of the impacts of trade liberaliza- December 2003, the Southern Shrimp Alliance tion through tariff reductions related to fish and 288 Global Agricultural Trade and Developing Countries fish products. This is partly because of the complex As expected the results show that there would be nature of the global seafood market, partly because significant increases in export volumes (and prices) of a lack of data, and partly because of a govern- under the Bogor Declaration and the ESVL relative mental and academic focus diverted away from to the baseline. If only the developed countries seafood markets toward the economics of manage- removed their tariffs, the simulation shows that ment of capture fisheries. In addition, although there would be little difference from the baseline nongovernmental organizations and international because developed-country tariffs are generally development agencies have produced many studies small. The greatest change would occur under the on trade liberalization and its impacts on the agri- EVSL scenario, at least initially. By 2020 the effects cultural sector in developing countries, there is of the Bogor and EVSL agreements would be the a spectacular lack of quantitative information on same (tables 15.2 and 15.3). the impacts of trade liberalization for developing Significant benefits from import tariff reduc- countries with respect to fish. tions accrue to the "Other APEC" countries of The study by Cox, Stubbs, and Davies (2000) is Brunei, Indonesia, Malaysia, Mexico, New Zealand, the notable exception. This section begins by dis- Papua New Guinea, the Philippines, the Russian cussing its findings on trade liberalization in Asia Federation, Singapore, and Vietnam. Pacific Economic Cooperation (APEC) countries. Removing Subsidies in Capture Fisheries Trade Liberalization in APEC Countries The previous section discussed the types of subsi- Cox, Stubbs, and Davies (2000) investigate the dies found in the fishing sector and the concern of short-run effects of trade liberalization on seafood the WTO CTE about fishing subsidies as a potential products in the APEC countries. These countries distorter of trade and contributor to unsustainabil- maintain a tariff and other trade barriers against ity of fish stocks around the globe. To analyze the fish and fish products. With the conclusion of the trade impacts of these subsidies, a logical place to Uruguay Round, WTO member nations agreed to begin may be to calculate producer subsidy equiva- lower tariff rates. However, the APEC agenda was lents (PSEs). According to the OECD,"the PSE is an more ambitious. Under the 1994 Bogor Declara- indicator of the value of the transfers from domes- tion, APEC made a commitment to fully liberalize tic consumers and taxpayers to the producers all markets by 2020, with 2010 as the deadline for resulting from a given set of agricultural policies at developed countries. This was followed by "early, a point in time" (FAO 2003). voluntary sectoral-liberalization" (EVSL) proposals The PSE also may be useful in assessing the in which nine sectors, including fisheries, would advantages of producer subsidies in the fisheries accelerate tariff removals beyond the Bogor Decla- sector. The complicating factor is management. In ration. Rather than having 2010 and 2020 as dead- agriculture, it is assumed that subsidies are com- lines for developed and developing countries, pared to a subsidy-free world characterized by eco- respectively, the timeline was moved to December nomically efficient allocation of goods at various 31, 2005. prices. However, if the fishery is managed under an Cox, Stubbs, and Davies (2000) developed a open-access system, for example, then the subsidy- simulation model to evaluate the impact of seafood free world is not economically efficient, because tariff removals under the Bogor Declaration, EVSL, that system does not lead to efficient allocation. To and another scenario wherein only the developed be truly efficient, the subsidies would not exist and countries in APEC would remove their tariffs while there would be perfect management of fish stocks those of developing countries remained the same. so that all negative externalities were incorporated The model included all the APEC countries and the into the price of each fish. PSEs for fisheries prod- rest of the world as sources and destinations. ucts have not been calculated because fish are Seafood products were generally grouped together highly heterogeneous and reference prices to except for a focus on species particularly important measure market-price support are hard to pin to Australia such as tuna, lobsters, and shellfish. down. Seafood: Trade Liberalization and Impacts on Sustainability 289 TABLE 15.2 Simulated Changes in the Real Value of World Exports in 1995 Prices (percent) Developed APEC Annual Growth EVSL Bogor Countries in Base Type of Export (1995­2020) 2010 2020 2010 2020 2010 2020 Unprocessed Tuna 1.3 5.0 4.8 1.5 4.8 1.6 1.5 Other fish 8.2 32.4 28.0 0.4 28.0 -1.9 -1.9 Rock lobster 1.4 1.5 1.5 0.7 1.5 0.7 0.7 Prawns 2.6 11.3 15.5 0.4 15.5 0.4 0.4 Other crustaceans 7.1 38.5 51.2 1.3 51.2 1.3 0.7 Abalone 3.4 20.2 21.9 2.6 21.9 2.6 2.0 Scallops 6.2 15.5 23.5 0.5 23.5 0.6 0.1 Other mollusks 3.9 22.5 23.7 1.3 23.7 1.3 1.4 Processed Tuna 2.4 0.4 0.8 0.3 0.8 0.3 0.3 Other fish 4.4 11.7 17.0 2.6 17.0 2.6 1.5 Rock lobster 0.4 0.0 0.0 0.1 0.0 0.1 0.0 Prawns 1.2 0.6 0.7 0.3 0.7 0.3 0.3 Other crustaceans 1.6 0.4 0.4 0.4 0.4 0.4 0.4 Mollusks 3.0 8.4 9.4 4.0 9.4 4.0 3.9 Total 5.1 20.4 24.1 1.0 24.1 0.2 -0.4 Source: Cox, Stubbs, and Davies 2000. TABLE 15.3 Simulated Benefits from Tariff Reductions, by Country (percent) Growth Developed EVSL and Developed in Base EVSL Bogor APEC Countries Bogor APEC Countries Country (1995­2020) 2010 2010 2010 2010 2010 Australia 2.7 0.0 0.1 0.1 0.7 0.1 Canada 2.4 1.9 0.1 0.0 5.9 0.1 Chile/Peru 0.2 0.0 0.0 0.0 0.4 0.0 China 9.1 0.0 0.0 0.0 -0.3 0.0 Hong Kong (China) 4.8 -0.3 0.0 0.0 -1.4 0.1 Japan 1.9 -0.3 -0.2 -0.2 0.0 -0.3 Korea, Rep. of 3.7 1.1 0.1 0.0 2.5 0.0 United States 1.7 0.6 0.0 0.0 2.9 -0.1 Other APEC 4.6 1.0 0.0 -0.1 0.9 -0.2 Total APEC 5.4 0.2 0.0 0.0 0.2 -0.1 Non-APEC 1.8 0.2 0.0 0.0 0.7 0.0 World 4.0 0.2 0.0 0.0 0.4 0.0 Note: These benefits represent changes in the sum of consumer surplus, producer surplus, and import tariff revenue. Source: Cox, Stubbs, and Davies 2000. 290 Global Agricultural Trade and Developing Countries The impact of removing subsidies may be ana- Hannesson (OECD 2003a) has investigated the lyzed according to the type of subsidy. For subsidies effects of liberalizing trade in fish, fishing services, that lower the costs of production (such as govern- and investments in fishing vessels. Three styles of ment-paid fishing access fees, low-cost vessel con- fisheries management are defined: open access, struction loans, and tax exemptions), removal will catch control, and efficient management. As we saw increase costs of production. A large portion of earlier, under open access fishermen are free to the world's subsidized fishing fleet is from the respond to prices by increasing or decreasing their European Union, Japan, Russia, China, and other catch. Increased prices will invite entry into the nations that subsidize (Milazzo 1998). A reduction fishery by more participants, so that in the long run in these subsidies would almost certainly benefit the fishery will be overfished. Under catch control fish stocks--as well as decreasing trade. and efficient management, total supplies are fixed Milazzo (1998) provides an excellent summary and will not change with changes in prices. This is of the benefits to developing nations of removing because a TAC will have been set to guide the fish- subsidies. ing effort and guarantee a sustainable fishery. The difference between catch control and efficient man- · Subsidies that pay for access arrangements agement is that the TAC catch control imposes no support continued operations primarily by constraint on each fisherman, who retains the European and East Asian distant-water fleets incentive to catch as much as he can, as fast as he off Africa and in the Western Pacific. These can, before the TAC is reached and the fishery subsidized operations reduce the fishing oppor- closed. Catch control alone is economically in- tunities available to local fishermen. In most efficient because it allows too many fishermen in cases, the payments probably do not compensate the fishery, and the capitalization and effort are adequately for the full economic value of the too high. resources. If trade barriers are removed--that is, if fish- · There is scattered evidence that subsidized importing countries lift their barriers--prices access arrangements are beginning to compro- decline in the importing country and rise in the mise local food needs. exporting country to a global equilibrium (account- · The combination of developed countries' subsi- ing for transportation costs). What are the impacts dies to their distant-water fleets and to their of such a development, assuming adequate manage- domestic (coastal) fleets minimizes to some ment measures? Table 15.4 shows the expected extent trade opportunities that should be avail- outcome. able to developing countries. The "double dividend" refers to the gain in the · Fishing subsidies are highly nontransparent in importing country from getting fish at a lower the sense that more than three-quarters of the price and redirecting resources from the domestic subsidies are not budgeted, and a good share of fishing industry to higher-value uses. Although budgeted subsidies are controlled by govern- there is no reason to assume that both the import- mental agencies other than those responsible for ing and exporting countries share the same type of fisheries. management regime, if both have an effective · Environmentally harmful subsidies outweigh regime then the results will be very similar to the the effect of subsidies that are environmentally classic outcome of agricultural trade liberalization. benign or positive. Milazzo's estimates show that With open access and catch control, under possibly no more than 5 percent of all subsidies which a change in prices induces increased effort in support conservation. the exporting country, it is conceivable that a coun- try could end up worse off with trade liberaliza- tion (Brander and Taylor 1997a, 1997b, 1998; Influence of Management Regime on Effects of Hannesson 2000). This is because the total quantity Trade Liberalization in Capture Fisheries caught in the open-access fishery will increase at An alternative means of looking at impacts of trade first but then decline as the fishery becomes over- liberalization is to assess their implications un- fished. With a decline in prices resulting from the der different management programs. Rögnvaldur elimination of import barriers, however, the effort Seafood: Trade Liberalization and Impacts on Sustainability 291 TABLE 15.4 Effects of Relaxing Trade Barriers Fish-Exporting Country Fish-Importing Country Catch Efficient Catch Efficient Regime Open Access Control Management Open Access Control Management Short-term Increased Increased No change in Lower effort, Lower effort, No change in effects effort, larger effort, no effort unless smaller no change in effort unless catches, more change in higher allowed catches, catch, lower smaller allowed trade gains catch, higher catch, gains more trade, profits, gains catch, gains from from trade profit gains from trade, gains from from trade trade, lower from trade higher market trade market value of value of quotas quotas and and licenses licenses Long-term Fish stocks Increased Same as above Fish stocks Reduction of Same as above effects decline, catch investment in recover, fishing fleets, may decline, fishing boats, catch may no change in possibly loss no change in increase, catch, double from trade catch, small double dividend from gains from dividend trade trade from trade Source: OECD 2003a: 170. in the importing country would be likely to decline, Impact of Trade Liberalization in Aquaculture giving fish stocks a chance to recover. This is not The implications of trade liberalization in aquacul- necessarily the predicted outcome. Indeed, in many ture would likely be very similar to those in agricul- cases, as the price of fish has decreased, fishermen ture, because aquaculture shares many of the have actually increased their effort to maintain total resource constraints and externalities of agriculture revenue, at least in the short run. (tables 15.5 and 15.6). Certainly, if tariffs in the The results above were premised on two sepa- European Union, the Republic of Korea, and Japan rate stocks of fish--one in the importing country were reduced, the quantity of aquacultured prod- and one in the exporting country. The discussion ucts sold to those countries would grow. can be made much more complicated by assuming The concern among many is that increased trade that several countries share the resource. in cultivated shrimp has had a large and negative Consider the European Union and Uganda as effect on the environment and that the effect rises trading partners. Much of Uganda's fisheries prod- with production and exports. The same is true for ucts come from Lake Victoria and are exported to salmon farming. Chile, Norway, Scotland, Canada, the European Union. Uganda has an open-access and Ireland are the largest producers of farmed management regime on Lake Victoria and shares salmon, with Chile and Norway being by far the lake with Kenya and Tanzania. If trade were to the largest. Environmental groups are concerned be liberalized, the amount traded would increase. not only about pollution but also about effects on Fishing pressure on Lake Victoria and its stock of the genetic diversity of wild fish from escaped Nile perch would increase, putting further pressure farmed fish that may not be indigenous to the area on the fish stock from both Uganda and Kenya. The (Porter 2003). price of the fish would rise as fewer and fewer fish Both salmon and shrimp production rely on fish were found. Food security would decline as the meal for feed. Any increase in aquaculture produc- local community found it increasingly difficult to tion of either species will have an impact on afford Nile perch. Unless some type of enforcement demand for fish meal. I have already discussed the management regime were set up to limit total catch various issues associated with fish meal production, from the lake, this source of export earnings might including the growing concern that the stocks of be short-lived. Holding all else constant, trade lib- fish from which fish meal is produced (herring, eralization would deplete stocks in Lake Victoria anchovies, capelin, menhaden) are themselves more quickly than if trade were not liberalized. 292 Global Agricultural Trade and Developing Countries TABLE 15.5 Effects of a Rise in the Price of Cultivated Fish on Aquaculture Output and Fisheries Catch If Feed Is Held Constant Management Regime Effect on Output Effect on Output in Capture in Capture Fisheries in Aquaculture Fisheries for Consumption Fish Open access Output rise for sufficiently low Lower stocks of feed fish lead to prices, but as the price of feed less growth of consumption fish. fish increases, the stocks will Higher price of consumption fish ultimately be exploited beyond leads to less supply as stocks are MSYa, supply of feed falls, and pushed beyond MSY aquaculture output falls Capture fisheries Output rise and flattens out Output of consumption fish for feed fish and as supply of feed cannot be falls as the price exceeds consumption fish further augmented a certain level managed separately All capture fisheries As above, but aquaculture As above, but output of capture managed as a whole is initiated at a higher price fisheries continues to rise with price longer before starting to fall. a. Maximum Sustainable Yield. Source: OECD 2003a: 204. TABLE 15.6 Effects on Price and Quantities of Market Liberalization: Relaxing Border Measures in the Importing Country (two-country situation) Exporter Importer Fishery managed by TAC Increase price, no change Decrease price, no change set without reference to in quantity in quantity economic factors Open access a) Stock above MSY Increase price and quantity Decrease price and quantity b) Stock at MSY Increase price, decrease quantity Decrease price and quantity c) Stock less than MSY Increase price, decrease quantity Decrease price, increase quantity Aquaculture a) Feed available without Increase price and quantity Decrease price and quantity significant price rise b) Managed fishery for Increase price and increase Decrease price and decrease captured feed fish or leave unchanged quantity or leave unchanged quantity c) Open access fishery for Same as open access above Same as open access above captured feed fish Source: OECD 2003a, page 200. overfished. Unless effective management of the tion consists of herbivores such as carp. Carp con- total catch in those fisheries is instituted, the sus- tribute significantly to food security in China and tainability of aquaculture may not be possible until other nations, particularly as they tend not to be an alternative to fish meal is developed. found on the export market, so trade liberalization This section has focused so far on carnivorous is likely to have little impact. fish (salmon and shrimp), and the impact of trade The other face of aquaculture is farmed shell- liberalization on the source of feed. However, a fish, which makes up a good proportion of aqua- large portion of the world's aquaculture produc- culture production worldwide. In Thailand Seafood: Trade Liberalization and Impacts on Sustainability 293 production of green mussels, blood cockle, oysters, As stocks in developed countries have declined, and other shellfish doubled from 73,976 million their fleets have gone elsewhere to capture fish. The tons in 1988 to 138,202 million tons in 2000 valued governments of the European Union, for example, at approximately $47 million (Chalermwat, Szuster, have paid several developing countries for access to and Flaherty 2003). Because the primary concern their fishing territory. While the developing nations with these products is the placement of the farms in gain access fees, enforcement of fish-management unpolluted areas, the WTO Agreement on Sanitary policies to limit the catches of the foreign fleets are and Phytosanitary Measures is likely to have the minimal, resulting in an overfishing of these fish largest effect on this sector. Table 15.6 summarizes stocks. Thus, developing countries derive a short- the discussion in this section. term gain by allowing foreign fleets to fish in their waters; that value disappears in the long run. Removing foreign access from developing coun- Conclusions tries' waters may not be the complete answer, even Global seafood markets are truly international. though foreign access is usually subsidized by the Production, consumption, imports, and exports foreign fleets' governments. Developing countries cover the globe, just as several species of fish have fisheries resources within their exclusive eco- migrate around the globe. Because the global mar- nomic zones. Removing foreign fleets from those ket for fish and fish products has specific dynamics waters is good for the fish stocks, but if the country and issues separate from global agriculture, under- itself has no means to capture the value of the standing the impacts of trade liberalization on resource, it gains little else. Two options present seafood and fishery products requires an under- themselves under such circumstances: first, to standing of the differentiated markets for the vari- negotiate better access agreements to ensure that ous products. the true value of the resource is being paid to the Key aspects of trade liberalization on global developing country, and, second, to invest in the seafood, fish meal, and fish oil markets have developing country's fishing capacity so that it can emerged from the discussion. Impacts of trade lib- take advantage of its rightful resource. It should go eralization on the welfare of countries depends without saying that in either case an effective man- critically on the fisheries management systems of agement system must be put in place to prevent the producing countries, since negative externali- overfishing. ties in global seafood markets are much larger and Tariffs in global seafood markets have come more detrimental than those specific to agriculture. down significantly and may no longer be a prime Open access, the management regime in many trade barrier, except perhaps in South-South trade. developing countries, invariably leads to overfish- In the United States, as the markets for certain ing. Any event that raises prices for fish from seafood species has become more competitive, exporting developing countries creates incentives industries in the United States have increasingly to fish even more, exacerbating overfishing and turned to antidumping and countervailing duty leading quickly to collapses in stocks. Even trade measures to protect themselves from competition liberalization in the aquaculture industries is not from developing countries. immune from the effects of fisheries management The WTO has the opportunity to use its purview regimes to the extent that the feed for that produc- over subsidies through the Agreements on Subsi- tion is derived from a poorly managed capture dies and Countervailing Measures to encourage fishery. members to drop fishing subsidies and thus to cure Increased trade in aquacultured products, inde- the trade distortions caused by the subsidies while pendent of issues with feed, can lead to increased encouraging sustainability of fish stocks globally. In environmental degradation from conversion of addition, from the developing countries' perspec- land from benign agricultural use to less benign tive, an important focus in WTO negotiations must aquacultural use.Little has so far been done to inter- be the Agreement on Technical Barriers to Trade nalize the negative externalities caused by excess and the Agreement on Sanitary and Phytosanitary fishing, unintended trapping of other marine life, or Measures. The processes by which developed coun- water pollution from aquaculture operations. tries impose technical barriers to trade must be 294 Global Agricultural Trade and Developing Countries transparent and demonstrably nonarbitrary. Devel- Conrad, J. M. 1999. Resource Economics. Cambridge: Cambridge oping countries need resources to assist them to University Press. Cox, Anthony, Matthew Stubbs, and Luke Davies. 2000. Southern meet current sanitary and phytosanitary measures Bluefin Tuna and CITES: An Economic Perspective. Report by building infrastructure that permits them to prepared for the Fisheries Resource Research Fund and meet the requirements and training workers to Environment Australia, ABARE Research Report 99.2, Canberra. maintain the proper measures. Deere, C. 2000. "Net Gains: Linking Fisheries Management, Finally, international trade in fish and fish prod- International Trade and Sustainable Development." Interna- ucts also has an impact on food security. Often the tional Union for Conservation of Nature and Natural Resources (IUCN), Washington, D.C. domestic market in exporting developing countries Delgado, C. L., N. Wada, M. W. Rosengrant, S. Meijer, and retains only the inferior fish, while the better, more M. Ahmed. 2003. Fish to 2020: Supply and Demand in valuable fish are sold abroad. A collapse in the stock Changing Global Markets. Washington, D.C.: International of the fish consumed domestically may lead to sig- Food Policy Research Institute. ESCAP and ADB. 2000. State of the Environment in Asia and nificant food security problems. Similarly, if fish Pacific 2000. Economic and Social Commission for Asia and meal prices were to rise for any reason, the increase the Pacific and Asian Development Bank. New York: United would have an impact on the ability of some Nations. www.usepscap.org/enrd/environ/soe.htm. FAO (U.N. Food and Agriculture Organization). 1999."Country nations to feed terrestrial livestock. Profile: The People's Republic of Bangladesh." Rome. Everyone has an interest in ensuring that fish- August. eries and aquaculture are managed in a sustainable ------. 2000. "Country Profile: The Kingdom of Thailand." Rome. May. way. As externalities are internalized into the pro- ------. 2002a. FAO Yearbook--Fisheries Statistics, Aquaculture duction process and their value incorporated into Production. volume 90/2. Rome. the prices of fish products, then it is likely that trade ------. 2002b. FAO Yearbook--Fisheries Statistics, Commodities. liberalization will bring about a net benefit to trad- vol. 91. Rome. ------. 2002c. The State of World Fisheries and Aquaculture. ing partners. The distribution of benefits across Rome. countries, producers, and consumers can best be ------. 2003. "Report of the Expert Consultation on Identify- judged after effective management measures are in ing, Assessing and Reporting on Subsidies in the Fishing Industry." FAO Fisheries Report 698. Rome. place. That distribution is not easily judged today. Filhol, Agnes. 2000."Effect of World Trade Organization's Regula- tion on World Fish Trade."GLOBEFISH 65 (February). Rome. Notes Grainger, R. J. R., and S. M. Garcia. 1996. "Chronicles of Marine Fishery Landings (1950­94): Trend Analysis and Fisheries 1. The primary source for the material in this section is Abila Potential." FAO Fisheries Technical Paper 359. Rome. 2003. Hannesson, Rögnvaldur. 1996. Fisheries Mismanagement: The 2. www.st.nmfs.gov/webpls. Case of the North Atlantic Cod. Oxford: Fishing News Books. 3. The Farm Security and Rural Investment Act of 2002 ------. 2000."Renewable Resources and the Gains from Trade." (Farm Bill) states that for the purposes of the Federal Food, Canadian Journal of Economics 33: 122­32. Drug, and Cosmetic Act, "the term `catfish' may only be consid- Institute for European Environmental Policy. 2002a. "Subsidies ered to be a common or usual name (or part thereof) for fish to the European Union Fisheries Sector." London. October. classified within the family Ictaluridae" (USITC 2003a). ------. 2002b. "Fisheries Agreements with Third Countries-- Is the EU Moving Towards Sustainable Development?" References London. November. Kinnucan, H., and C. R. Wessells. 1997. "Marketing Research Abila, Richard. 2003. "Food Safety in Food Security and Food Paradigms for Aquaculture." Aquaculture Economics and Trade, Case Study of Kenyan Exports." International Food Management 1: 73­86. Policy Research Institute, Washington, D.C. September. Lem, Audun. 2003. The WTO Doha Round and Fisheries: What's Anderson, J. L. 2003. The International Seafood Trade. Oxford: at Stake. FAO Fact Sheet for WTO Ministerial Conference in Woodhead. Cancún, Mexico. July. Brander, J. A., and M. S. Taylor. 1997a. "International Trade and McGovern, Dan. 2003. "SSA Files 6 Shrimp Antidumping Open-Access Renewable Resources: The Small Open Econ- Petitions; Alleges Dumping Margins of Over 200 Percent." omy Case." Canadian Journal of Economics 30: 526­52. The Wave News Network, December 31. ------. 1997b. "International Trade between Consumer and Milazzo, Matteo. 1998."Subsidies in World Fisheries: A Reexam- Conservationist Countries." Resource and Energy Economics ination." World Bank Technical Paper. Washington, D.C. 19: 267­98. National Research Council. 1999. Sustaining Marine Fisheries. ------. 1998. "Open Access Renewable Resources Trade and Washington, D.C.: National Academy Press. Trade Policy in a Two-country Model." Journal of Interna- OECD (Organisation for Economic Co-operation and Develop- tional Economics 44: 181­210. ment). 2003a. Liberalising Fisheries Markets: Scope and Chalermwat, K., B. W. Szuster, and M. Flaherty. 2003. "Shellfish Effects. Paris. Aquaculture in Thailand." Aquaculture Economics and Man- ------. 2003b. Review of Fisheries in OECD Countries: Country agement 7: 249­61. Statistics 1999­2001. Paris. Seafood: Trade Liberalization and Impacts on Sustainability 295 Porter, G. 2003. "Protecting Wild Atlantic Salmon from Imports No. 731-TA-1012 (Final). Publication 3617. August. of Salmon Aquaculture: A Country-by-Country Progress Washington, D.C. Report." www.worldwildlife.org. May. ------. 2003b. "Crawfish Tail Meat from China." Investigation Robb, Cairo. 2001. International Environmental Law Reports, 731-TA-752 (Review). Publication 3614. July. Washington, vol. 2, Trade and the Environment. Cambridge: Cambridge D.C. University Press. Wessells, C. R., and J. Anderson. 1992."Innovations and Progress Schrank, W. 2003. "Introducing Fisheries Subsidies." FAO in Seafood Demand and Market Analysis." Marine Resource Fisheries Technical Paper 437. Rome. Economics 7: 209­28. UNEP (United Nations Environment Programme). 1999a. Envi- World Bank. 2003. "Madagascar Diagnostic Trade Integration ronmental Impacts of Trade Liberalization and Policies for Study." Prepared for the Integrated Framework for Trade- the Sustainable Management of Natural Resources: A Case Related Technical Assistance to Least Developed Countries, Study of Bangladesh's Shrimp Farming Industry. UNEP: Washington, D.C. New York. WTO (World Trade Organization). 1999. "Benefits of Eliminat- ------. 1999b. "Environmental Impacts of Trade Liberaliza- ing Trade Distorting and Environmentally Damaging tion and Policies for the Sustainable Management of Nat- Subsidies in the Fisheries Sector." WT/CTE/W/121. Geneva. ural Resources: A Case Study of Uganda's Fisheries Sector." ------. 2000. "Environmental Benefits of Removing Trade New York. Restrictions and Distortions: The Fisheries Sector." ------. 2002a. "Africa Environment Outlook: Past, Present and WT/CTE/W/167. Geneva. Future Perspectives." www.unep.org/aeo. ------. 2001. "Environmental Benefits of Removing Trade ------. 2002b. "Well Managed Fisheries Vital for Environmen- Restrictions and Distortions: The Fisheries Sector-- tally Friendly Development in Poor Parts of the Globe." Press Addendum." WT/CTE/W/167/Add.1. Geneva. release. New York. March. ------. 2003a. "Possible Approaches to Improved Disciplines U.S. Department of Commerce. 1997. "Fisheries of the on Fisheries Subsidies." TN/RL/W/77. Geneva. United States, 1997." National Marine Fisheries Service, ------. 2003b. "Understanding the WTO." Third Edition. Sep- Washington, D.C. tember. www.wto.org. WWF (World Wildlife Fund). 2003. USITC (United States International Trade Commission). 2003a. "WWF Fact Sheet: West African Marine Ecoregion." Dakar, "Certain Frozen Fish Fillets from Vietnam." Investigation Senegal. 16 COFFEE: MARKET SETTING AND POLICIES John Baffes, Bryan Lewin, and Panos Varangis All coffee is produced in the tropics, primarily by There are two types of coffee. Arabica, grown at smallholders. Most is consumed in high-income high altitudes in Latin America (including Brazil) countries. Latin America accounts for 60 percent of and northeastern Africa, accounts for two-thirds of global output, followed by Asia (24 percent), and total world output. It has a strong aroma and low Africa (16 percent). More than half of global coffee level of caffeine. Robusta, with a much stronger output is accounted for by the three dominant pro- taste than arabica, is grown in humid areas at low ducers: Brazil (33 percent), Colombia (10 percent), altitudes in Asia, western and southern Africa, and and Vietnam (10 percent). Some other African and Brazil. During the last decade, production of Latin American countries, however, are heavily robusta, which is particularly suitable for instant dependent on their exports of coffee, despite their coffee, has increased (table 16.1).1 low share in global output. For example, coffee During the last decade, the coffee market has accounts for more than half of total merchandise gone through a number of structural changes. On exports in Burundi, Rwanda, and Ethiopia and the supply side, Brazil's production capacity more than 20 percent in Guatemala, Honduras, and expanded enormously, with new plantations in the Nicaragua. More than 80 percent of coffee produc- north that are less affected by frosts and, because tion is traded internationally. Historically, coffee is of irrigation, not affected by droughts. Vietnam the second most traded primary commodity after entered the coffee market in a major way in the crude oil, generating more than $15 billion in 1980s--it currently supplies more than 12 million export revenue (evaluated at 1997­98 average prices bags, making it the world's second-largest coffee and volumes). Overall, consumption volumes have exporter. On the demand side, consumption of spe- stagnated in the mature markets, in which the cialty coffees has expanded, currently accounting United States accounts for about 18 percent, fol- for an estimated 6­8 percent of total consumption. lowed by Brazil (10 percent), Germany (9 percent), Demand for low-quality coffee beans has also Japan (6 percent), and France and Italy (5 percent increased, primarily reflecting new technologies each). However, consumption has been increasing that enable roasters to remove the harsh taste of in some new (especially transition) markets. robustas for normal coffee while continuing to The authors thank Ataman Aksoy and Harry de Gorter for helpful comments and suggestions on earlier drafts. 297 298 Global Agricultural Trade and Developing Countries TABLE 16.1 The Changing Structure of the Coffee Market (thousands of 60-kg bags) Arabica Year Colombian Milds Other Milds Naturals Subtotal Robusta Total 1992 16,959 25,122 23,317 65,398 27,291 92,689 1993 13,256 23,398 28,555 65,209 26,989 92,198 1994 15,059 24,582 29,300 68,941 27,901 96,842 1995 15,503 27,525 18,545 61,573 27,193 88,766 1996 12,489 27,040 27,126 66,655 37,033 103,688 1997 13,498 27,965 23,436 64,899 32,753 97,652 1998 12,509 27,380 35,024 74,913 33,506 108,419 1999 11,821 31,698 30,178 73,697 39,706 113,403 2000 12,026 28,480 30,717 71,223 45,638 116,861 2001 13,229 26,123 28,540 67,892 42,834 110,726 2002 13,179 25,585 43,667 82,431 41,720 124,151 2003 13,352 26,318 26,217 65,887 39,945 105,232 Market share (percent) 1992 18 27 25 71 29 100 1993 14 25 31 71 29 100 1994 16 25 30 71 29 100 1995 17 31 21 69 31 100 1996 12 26 26 64 36 100 1997 14 29 24 66 34 100 1998 12 25 32 69 31 100 1999 10 28 27 65 35 100 2000 10 24 26 61 39 100 2001 12 24 26 61 39 100 2002 11 21 25 66 34 100 2003 13 25 25 63 37 100 Source: U.S. Department of Agriculture. meet the increasing demand for instant and fla- (ICAs) under the auspices of the ICO, the last of vored coffees, which primarily use robusta coffees. which collapsed in 1989. Government intervention For most of the 20th century the coffee market in domestic markets was also prevalent in many has been subject to various supply-control schemes. countries through parastatals that controlled mar- The most important were the price-stabilization keting and trade in the coffee industry. Following schemes implemented by Brazil at the beginning of the collapse of the last ICA, most parastatals were the century, the Inter-American Coffee Agreements either dismantled or their roles diminished. Cur- implemented during and after the Second World rently, the global coffee market is, to a large extent, a War, the agreements administered by the Interna- distortion-free market. On the trade side, import tional Coffee Organization (ICO) from 1962 to restrictions are nonexistent, except some tariff esca- 1989, and more recent attempts by the Association lation in coffee products (such as instant coffee). of Coffee Producing Countries (ACPC). Although the stated objective of these arrangements was to Global Balance and Price Trends stabilize prices, prices often ended up being higher than they would have been in the absence of Brazil, by far the largest coffee producer and the arrangements. The most influential of these exporter and the second-largest consumer, schemes were the International Coffee Agreements accounts for one-third of global output and Coffee: Market Setting and Policies 299 TABLE 16.2 Coffee Production, Selected Years (thousands of 60-kg bags) Country 1960 1970 1980 1990 2000 2002 2004 Brazil 29,800 11,000 21,500 31,000 34,100 51,600 42,400 Vietnam 53 39 77 1,200 15,333 11,167 12,000 Colombia 7,260 8,000 13,500 14,500 10,500 11,712 11,600 Indonesia 1,327 2,327 5,365 7,480 6,495 6,140 5,750 India 1,225 1,914 1,977 2,970 5,020 4,588 4,835 Mexico 2,100 3,200 3,862 4,550 4,800 4,350 4,500 Ethiopia 1,687 2,589 3,264 3,500 3,683 3,693 4,000 Guatemala 1,500 1,965 2,702 3,282 4,564 3,802 3,671 Uganda 1,925 2,667 2,133 2,700 3,205 2,910 3,200 Peru 598 1,114 1,170 1,170 2,824 2,760 2,980 Honduras 291 545 1,265 1,685 2,821 2,661 2,753 Costa Rica 951 1,295 2,140 2,565 2,502 2,207 2,050 Nicaragua 437 641 971 460 1,610 997 1,500 Côte d'Ivoire 0 4,414 3,973 4,734 5,700 3,568 1,444 El Salvador 1,452 2,054 2,940 2,603 1,624 1,351 1,285 Papua New Guinea 61 426 889 969 1,051 1,118 1,210 Cameroon 855 1,180 1,860 1,450 1,113 801 1,100 Kenya 566 999 1,568 1,455 864 926 1,085 Thailand 1 19 201 785 1,692 757 950 Ecuador 594 1,255 1,517 1,830 1,005 790 750 Total 64,999 58,838 85,738 99,911 116,861 124,151 117,650 Source: U.S. Department of Agriculture. produces both arabica and robusta coffee. It is fol- approximately 5.5 kilograms each. U.S. per capita lowed by Colombia (arabica) and Vietnam consumption fluctuates between 4 and 5.5 kilo- (robusta), each accounting for about 10 percent of grams; in the United Kingdom and Japan it is global output. Other significant producers are between 2.5 and 3 kilograms. Only five coffee pro- Indonesia and Mexico (6 percent each) and India ducers consume a substantial portion of their (4 percent) (table 16.2). output: Brazil and Ethiopia (30 percent each), The technology of coffee production has Indonesia (23 percent), Mexico (19 percent), and changed significantly in the past 30 years, but not Colombia (11 percent), which together account for all countries have shared equally in the changes. about 20 percent of global output; the remaining Average yields in Asia are double those in Sub- 80 percent is internationally traded. Saharan Africa, and yields in Latin America are Vietnam's emergence as a major robusta pro- 60 percent higher than in Africa. Annual yield ducer altered the landscape of the global coffee growth during the 1990s was 2.6 percent in Asia, market in a permanent way. In 1980 Vietnam pro- 1.7 percent in Latin America, and 1.1 percent in duced 77,000 bags--less than 0.1 percent of world Sub-Saharan Africa, according to data from the production. In 2000, it exceeded 15 million bags-- U.N. Food and Agriculture Organization (FAO). more than 13 percent of world production. On the demand side, the United States con- Vietnam entered the coffee market in response to a sumes about 18 percent of global output, followed series of policy reforms in the early 1990s that by Brazil (10 percent), Germany (9 percent), Japan changed the balance of incentives toward export (6 percent), France and Italy (5 percent each). On a crops. These reforms facilitated land ownership per capita basis, Scandinavian countries consume and liberalized input and output markets. Follow- about 10 kilograms a year, followed by Germany ing the reforms, for example, fertilizer prices (8 kilograms), and France, Italy, and Spain with declined by almost 50 percent. Other reforms 300 Global Agricultural Trade and Developing Countries (known as Doi Moi) encouraged internal migra- Numerous studies have identified several factors tion to the Central Highlands because of easy access that are likely to further influence coffee process- to new land (eventually to be used for coffee pro- ing and consumption patterns (see, for example, duction). These reforms, combined with the 1994 IADB/USAID/World Bank 2002, and Lewin, coffee price spike, made Vietnam an important Giovannucci, and Varangis 2004). First, roasters are player in the coffee market. It is worth noting that able to work with a lower level of stocks. Second, Vietnam's coffee expansion took place without new technology enables them to remove the harsh assistance from either national or multilateral taste of robustas, achieving the same level of quality funding. However, some help came from the Soviet with lower-quality beans. Third, roasters have been Union and Eastern European countries in the form more flexible in their ability to make short-term of technical assistance during the early 1980s. switches between coffee types, implying that the Because neither Vietnam nor these countries were premiums commanded by certain types of coffee ICO members, and hence not bound by any quota cannot be retained for long. Finally, a small seg- obligations, they could expand coffee production ment of the market has emerged that focuses on and trade without any restriction. The expansion product differentiation, such as organic, gourmet, was also aided by the desire of the Soviet Union and and shade coffee. The implication of all this is that Eastern European countries to have access to coffee the demand outlook is likely to be different for dif- without paying hard currency. ferent coffee producers. Specifically, if any expan- Brazil has been able to maintain unprecedented sion in coffee demand takes place, it is likely to be at output levels, averaging more than 35 million bags the two ends of the spectrum: lower-quality beans during the last four seasons. Extensive mechaniza- (reflecting improved technology and increased tion of coffee harvesting, along with the develop- demand for soluble coffee) and specialty coffees ment of high-yielding varieties, has reduced costs (reflecting expansion to niche markets).3 Efforts to of production, while shifting production to the increase coffee consumption may also come at the north, away from the frost-prone areas of the expense of tea consumption, a commodity pro- south, has reduced the likelihood of weather- duced mainly by low-income (and often coffee- related supply disruptions. Extensive use of irriga- producing) countries. tion in areas such as Bahia and the Cerrado has Coffee prices are highly volatile. (figures 16.1a stabilized and sustained yields. Another signifi- through d). During the 1990s arabica prices ranged cant development in Brazil is emergence of semi- from $1.17 a kilogram in August 1992 to $5.89 a washed arabicas; a process that makes better coffee. kilogram in May 1997. Robusta prices ranged from About 3 million bags of semi-washed arabicas $0.82 a kilogram in June 1992 to $4.03 a kilogram compete directly with higher-quality coffee from in September 1994. The price volatility stems in Central America. part from weather conditions in Brazil, where frost Given that both Vietnam and Brazil are low-cost affects crops every five to six years and severe producers, they are unlikely to reduce coffee pro- droughts also occur periodically. While short- duction. Consequently production cutbacks to selling and buying by hedge funds are sometimes restore the balance of supply and demand are now cited as a reason for the high volatility of coffee coming from the higher-cost African and Central prices, this activity probably contributes only to American producers. In Central America, for exam- short-term volatility.4 ple, production of the lower-altitude, lower-quality Coffee prices have declined considerably since coffees that can be easily replaced in commercial 1998 (figure 16.1). In January 2002 robusta dropped blends by Brazilian arabicas have fallen sharply.2 to $0.50 a kilogram (the lowest nominal level since While Latin America and Asia have increased the $0.49 a kilogram price of May 1965 and 86 per- their shares in global coffee output, Africa's share cent below its high four years earlier), while in has declined from 33 percent in 1970 to 18 percent October 2001 arabica averaged $1.24 a kilogram, a in 2000. Africa's coffee output has never surpassed nine-year low and 76 percent below its high four its peak in 1972. After remaining almost constant at years earlier. The combination of increased avail- 20 million bags for two decades following that ability from Vietnam and Brazil, as well as domestic peak, it has been in slow decline since then. policies in many producing countries that retard Coffee: Market Setting and Policies 301 FIGURE 16.1(a) Nominal Coffee Prices, 1990­2003 (US$ per kilogram) 6.0 5.0 4.0 3.0 2.0 1.0 0.0 Jan-90Jul-90 Jan-91Jul-91 Jan-92Jul-92 Jan-93Jul-93 Jan-94Jul-94 Jan-95Jul-95 Jan-96Jul-96 Jan-97Jul-97 Jan-98Jul-98 Jan-99Jul-99 Jan-00Jul-00 Jan-01Jul-01 Jan-02Jul-02 Jan-03Jul-03 Robusta Arabica Source: World Bank. (b) Real Coffee Prices, 1960­2003 (US$ per kilogram) 10.0 8.0 6.0 4.0 2.0 0.0 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 Robusta Arabica Source: World Bank. exit from the market of uncompetitive producers, poverty implications of coffee in these countries led to these historically low prices, which, in the are enormous--in seven coffee-dependent African absence of any international supply control mecha- countries, per capita gross domestic product nism, gave rise to the so-called coffee crisis and (GDP) ranged between $112 and $336 (table 16.3). probably prolonged its length. Areas with relatively high labor costs and large Exports from small coffee producers are a farms that are heavily dependent on seasonal labor, minuscule proportion of global trade in coffee but especially in Central America, can feel the effects can loom large in the exporters' economies. For of changing prices in a significant way (Lewin, example, three African countries (Burundi, Giovannucci, and Varangis 2004). For example, the Rwanda, and Ethiopia) derive more than half of rural labor employed in the coffee sectors of five their total merchandise exports from coffee. The Central American countries represented, on 302 Global Agricultural Trade and Developing Countries (c) Nominal Price Indexes for Coffee and Other Commodities, 1990­2002 (August 2002 = 1.0) 5.0 4.0 3.0 2.0 1.0 0.0 Jan-90Jul-90 Jan-91Jul-91 Jan-92Jul-92 Jan-93Jul-93 Jan-94Jul-94 Jan-95Jul-95 Jan-96Jul-96 Jan-97Jul-97 Jan-98Jul-98 Jan-99Jul-99 Jan-00Jul-00 Jan-01Jul-01 Jan-02Jul-02 Jan-03Jul-03 Coffee composite Agriculture Source: World Bank. (d) Real Price Indexes for Coffee and Other Commodities, 1960­2003 (2001 = 1.0) 8.0 6.0 4.0 2.0 0.0 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 Coffee composite Agriculture Source: World Bank. average, 28 percent of the labor force in those coun- The International Environment tries (Nicaragua, 42 percent; Guatemala, 31 per- Regulation of coffee supplies at the international cent; Costa Rica, 28 percent; Honduras, 26 percent; level has a long history (box 16.1). Calls for supply El Salvador, 17 percent). controls were made as early as 1902 following price declines due to Brazil's oversupply (Hutchinson The Policy Environment 1909). At least three successful stabilization schemes The coffee market has been subject to considerable took place in Brazil between 1905 and 1921. intervention at the international and national lev- However, the coffee market became depressed els. Those interventions are the subject of the next following the crash of 1929. Attempts by Brazil to two subsections. convince other coffee producers to coordinate Coffee: Market Setting and Policies 303 BOX 16.1 Coffee Supply Controls in the Twentieth Century Calls for supply controls in the coffee market of free entry and competition. However, came as early as 1902 at an (unsuccessful) Inter- 30 years of controls had taken a toll: Brazil's national Coffee Conference held in New York share in the export market of coffee had fallen following price declines due to Brazil's oversup- from 87 percent in 1905 to 55 percent in 1940. ply (Hutchinson 1909). At the time, Brazil Brazil's coffee problem had become Latin accounted for more than 85 percent of world's America's coffee problem. coffee output of 18.2 million bags. Chronic Weak demand from Europe during the Second oversupplies prompted the state of São Paulo, World War, coupled with the desire of the United which accounted for three-quarters of Brazil's States to keep Latin America on the side of the coffee, to initiate a price stabilization scheme in Alliance, led to the formation of the Inter- 1905--called valorization--and to prohibit new American Coffee Agreement. Its membership plantings. The stockholding mechanism that consisted of the two dominant coffee producers regulated sales of coffee was financed by the (Brazil and Colombia) and several smaller produc- federal government and several foreign banks. ers in Central America (Wickizer 1943). A second The scheme not only stabilized prices but also agreement was negotiated after the war. The two kept them at levels higher than demand and agreements had the same outcome. Supply supply would have supported. A second val- restrictions and investment activity by European orization was undertaken in 1917 (following the countries in their colonies brought African pro- disruption of coffee consumption in Europe dur- ducers into the market. Latin America's coffee ing World War I), and a third in 1921. These problem became the Western world's coffee schemes were very profitable for their promot- problem. ers. However, greater stability in the coffee mar- In 1962 coffee-producing countries account- ket arising from the supply controls encouraged ing for 90 percent of global output and almost the rapid extension of new plantings in Brazil all developed consuming countries formed the and led to new calls for even more state inter- International Coffee Organization and signed vention (Wickizer 1943: 143). Thus, São Paulo's the International Coffee Agreements (ICAs). The coffee problem became Brazil's coffee problem. objective of the ICAs was to stabilize coffee Following the success of the valorization prices through mandatory export quotas. The schemes, a permanent supply-control scheme United States enthusiastically backed the agree- was envisaged for the newly created São Paulo ment, considering it a means of increasing the Coffee Institute. The institute began buying cof- income of Central American coffee producers, fee after the 1927­28 bumper crop and con- hoping that this would contain the spread of vinced other states that had become important communism. Consumer-country support for suppliers to join in. It withdrew from coffee pur- export quotas was also encouraged by large chasing after the crash of October 1929, when importers, who benefited from export-tax coffee consumption plummeted and financing rebates offered by the Instituto Brazilieno de dried up. The Brazilian government then Café in return for high-volume purchase com- attempted to convince other Latin American cof- mitments. Western European countries viewed fee producers (which had increased their market the ICAs with sympathy, believing that high cof- shares considerably) to find ways to regulate fee prices were a good way to aid their former exports. Two Pan-American Coffee conferences colonies (Bates 1997). To satisfy their quota (in Bogotá in 1936 and Havana in 1937) ended obligations, governments of coffee producers with no agreement. In the meantime, several bought stocks using part of their coffee tax coffee destruction schemes were undertaken by revenues. The export-quota system, first imp- Brazil using public funds. During 1931­38, a lemented in 1963, continued intermittently total of 68.7 million bags were destroyed--twice until 1989. the world's annual coffee output. Following After Vietnam entered the world market in inaction by other coffee producers (not surpris- the 1980s, with assistance from the Soviet ingly, since they were enjoying the benefit of Union, the West's coffee problem became the controls), Brazil abandoned restrictions in favor world's coffee problem. 304 Global Agricultural Trade and Developing Countries TABLE 16.3 Coffee's Importance to Developing Countries, 1997­2000 Averages Percent of Merchandise Exports Per Capita GDP Country Merchandise Exports (millions of current US$) (constant 1995 US$) Burundi 72.2 64 143 Rwanda 58.1 65 227 Ethiopia 51.5 520 112 Uganda 40.1 509 336 Sierra Leone 29.7 11 161 Nicaragua 25.4 613 435 Honduras 21.7 1,398 715 Guatemala 20.2 2,505 1,535 El Salvador 18.2 2,580 1,737 Tanzania 13.4 637 185 Madagascar 11.1 616 241 Source: International Coffee Organization. supply-containing mechanisms failed. Brazil then the price increase already under way, but the agree- introduced a number of coffee destruction schemes. ments were overtaken by the price rises that Between 1931 and 1938, a total of 68.7 million bags followed the Brazilian frosts in 1994. During 2000 were destroyed--twice the world's annual global and 2001 the ACPC worked to persuade coffee- coffee output. Following years of weak demand producing countries to retain part of their exports from Europe during the Second World War, Brazil so as to staunch the decline in coffee prices that had negotiated two agreements with other produc- started in 1998 and accelerated in 2000. Following ing countries in Latin America. Those agreements some initial enthusiasm, ACPC's efforts failed, and were largely unsuccessful. The countries that agreed the association was dissolved in February 2002, one to restrict their exports in return for Brazil's month after robusta prices reached their historic coffee stock destruction did not respect their low. The ACPC failed for several reasons, but a commitments. principal one was that in a liberalized market, the In the early 1960s most coffee-producing coun- institutional structure necessary to ensure compli- tries (accounting for 90 percent of global output) ance in the member countries--a single-desk mar- and almost all developed coffee-consuming coun- keting agency--had been dismantled. There was tries formed the ICO, which attempted to stabilize also the problem of free-riding by nonmembers. coffee prices through mandatory export quotas The ICO attempted once again in 2002 to reduce under the International Coffee Agreements. The coffee availability in a new agreement under which export quota system, first implemented in 1963, coffee producers are to remove low-quality beans was temporarily suspended in 1972 as coffee prices from the market. Regulation 407 of the ICO states soared. Quotas were restored in 1980 and sus- minimum specifications for export qualities, but it pended again in 1986 due to soaring prices. They will depend entirely on voluntary compliance for were reintroduced in 1987 and suspended indefi- its success. nitely in 1989. These agreements kept coffee prices A final consequence of the ICAs was that they higher than they otherwise would have been gave rise to rent-seeking behavior by governments (Gilbert 1995). and marketing boards. The extent of this problem Following the collapse of the last International was revealed in the late 1980s, shortly before the Coffee Agreement, several coffee producers-- collapse of the last ICA, when the Instituto including Brazil and Colombia but not Vietnam Brazilieno de Café decided to auction 10 percent of and Mexico--formed the Association of Coffee its export quotas. The very high prices exporters Producing Countries in September 1993. In the paid for the quotas revealed to the entire domestic following year, export restrictions did contribute to coffee industry the extent of the rents being Coffee: Market Setting and Policies 305 extracted. A common consequence in many coun- domestic prices adjusted faster after the reforms tries of the end of the ICA was an end to opportu- than they did before the reforms. In addition to nities for rent seeking--this led to significant shifts higher prices, considerable private investment in in domestic support policies in several producing the marketing, processing, and transportation sec- countries. tors took place. Increased supply response also took place in most occasions. At the same time, the gap created by the with- Domestic Policies drawal of the state has not been filled in all cases-- Since the collapse of the ICAs, domestic policies of the quality of public-sector services has deterio- coffee producers have focused on the reform and rated. It has been often argued that the quality of liberalization of marketing systems, and more coffee declined after the reforms, but this cannot be recently on helping producers survive periods of substantiated from the data. Quality may have low prices, sometimes through state intervention. declined after the collapse of the ICAs, since during Akiyama (2001) reported that only 15 of the the coffee agreements, quality improvements were world's 51 coffee-producing countries had private the only means of increasing revenue.5 marketing systems in 1985. Twenty-five countries Uganda undertook sweeping reforms in 1990 sold coffee through state-owned enterprises,includ- (Akiyama 2001). An overvalued exchange rate, the ing marketing boards and stabilization funds, and inefficiencies of the country's Coffee Board, politi- another 11 countries had mixed state and private- cal instability, and the price decline of 1989 made sector marketing bodies. Most aspects of coffee reform the only viable alternative. Under the marketing and trade, especially in Sub-Saharan reforms, producer prices rose from 40­50 percent Africa, were handled by government-controlled of export prices to 70­80 percent. The supply agencies, which typically resulted in heavy taxation response has been considerable, and many entre- of the sector. Although the reasons behind the tax preneurs have entered the market. Regulation, qual- policies varied, the main ones were low price elastic- ity control, and promotion issues were assigned to ity of short-run supply, implying minimal impact of the newly established Uganda Coffee Development taxation on supply; less social and political resist- Authority. In addition to increased output, Uganda ance to taxation for cash crops than for food crops; regained its reputation as a reliable robusta pro- the relative simplicity of tax collection, facilitated by ducer, commanding a premium for its exports. the single marketing channel; and support for the Reforms in neighboring Tanzania have been less government budget and balance of payments successful. Before 1990 the Tanzania Marketing through foreign exchange earnings. Board and the cooperative unions handled all mar- Many coffee-producing countries undertook keting (including input provision, transportation, reforms during the 1990s by removing or redefin- and processing) and trade aspects of the sector. The ing the role of the parastatals. A combination of cooperatives were also responsible for managing falling prices and rent-seeking activities by some the large estates nationalized in the early 1970s. of the marketing boards led several countries to Some reforms were introduced in 1990, but they reform their coffee sectors altogether. The outcome affected only inputs, price announcements, and of these reforms has been mixed and mirrors the retention of export earnings. More comprehensive outcome of similar efforts in other export crop sec- reforms begun in 1994 allowed private traders to tors (Akiyama and others 2003; Shepherd and purchase coffee directly from growers and process it Farolfi 1999). Bohman, Jarvis, and Barichello in their own factories for the first time in more than (1996) showed that in many cases, prices paid to 30 years. The outcome of these reforms has been growers were lower under the ICAs compared with mixed. Growers receive a higher share of f.o.b. (free what they would have been under a free market. on board) prices, they are paid promptly, and entre- Krueger (1990) showed that this was the case for preneurial activity has increased enormously. But other commodities as well. Krivonos (2003), who the Tanzanian coffee sector is still plagued by over- evaluated the impact of reforms undertaken in regulation (including mandatory auction), high 14 coffee-producing countries during the late 1980s taxation, and ad hoc decisions by the Tanzanian and early 1990s, concluded that in most cases Coffee Board (Baffes forthcoming). 306 Global Agricultural Trade and Developing Countries Domestic policies in producing countries nels such as supermarkets, foodservice, and remain sensitive to international developments as other institutions and they compete strongly on well as to local pressures, and consequently distor- the basis of price. Differentiated coffees are often tionary domestic policies appear in many coun- distinguished by a closer and sometimes direct tries. Although coffee prices have been in long-term relationship with a roaster or buyer rather than decline, the volatility discussed above can make it being traded in bulk or via the commodity difficult to determine whether price changes are markets. temporary or a genuine shift in market fundamen- Differentiated coffees can help the cof- tals. Under such circumstances it is equally difficult fee industry compete with other beverages by to determine the correct policy response. An addi- leveraging unique characteristics that in- tional complication is that the shift to lower-cost clude: (1) geographic indications of origin producers has been paralleled by the fact that those (appellations); (2) gourmet and specialty; countries with greater market power have lower (3) organic; (4) fair trade; (5) eco-friendly dependency on coffee export volumes. This makes or shade-grown; (6) private or corporate defensive policymaking difficult for countries with standards." greater dependency. The approaches taken to recent domestic policy- The expansion of differentiated coffees has two, making are varied. Among the larger producing often overlapping, dimensions. The first is social. countries, Brazil has long had a policy of preferen- Rising consumption of fair-traded, eco-friendly, tial credit access but more recently has been auc- shade-grown, or bird-friendly coffees is driven by tioning put options to farmers at well below fair social concerns. Consumers wish to ensure that value; these options are exercisable as sales of coffee coffee growers receive higher prices (fair trade) to the government. In Central America govern- or to improve the effects of coffee growing on ments have bailed out the banks that had lent heav- the environment (shade-grown or bird-friendly ily to the coffee sector, but because most loans had coffee). The second dimension relates to taste or been made to larger, more creditworthy farmers, preference. Here, increasing consumption derives the bailout failed to have much impact on the poor- from geographic indications of origin as well as est, except by maintaining employment in larger gourmet and specialty coffees (such as Kona coffee estates. or Kilimanjaro coffee). Consumers are willing to pay a premium for these coffees because of their superior characteristics. Quite often these two Niche Markets and Changing dimensions overlap in the sense that consumers Patterns of Consumption may demand specialty coffee that also satisfies cer- The last decade has witnessed the emergence of tain social criteria. nontraditional channels of production, marketing, Certification of nontraditional coffees is compli- and consumption of "new coffees"--gourmet (or cated and often contentious. Currently, no govern- specialty6), organic, fair trade, eco-friendly (shade- ment agency or international organization has the grown or bird-friendly), and other certified coffees. official mandate to certify nontraditional coffees. Lewin, Giovannucci, and Varangis (2004: 99) make With the exception of organic coffee, all certifica- the following distinction between differentiated tion comes from nongovernmental organizations-- and mainstream coffees: hence some of the value of the certified coffee rests with the reputation of the certifying organizat- "Differentiated coffees are those that can be ion. Organic coffee carrying a legally protected clearly distinguished because of distinct origin, trademark is regulated in the European Union, defined processes, or exceptional characteristics Japan, and the United States. such as superior taste or zero defects. In con- The rise of self-certification by large supermar- trast, mainstream coffees are nearly always pre- ket chains, often with standards below those set by ground blends that are often unidentified in the independent certification agencies, raises the terms of origin. These are usually, though not issue of credibility and thus of the further expan- always, distributed through mainstream chan- sion of niche markets. Parallel to the question of Coffee: Market Setting and Policies 307 self-certification is the emergence of institutional Colombia. At the same time, numerous niche mar- buyers that require producers to meet certain sus- kets have emerged. Currently 6 to 8 percent of cof- tainability criteria but do not offer a price premium fee output is traded outside traditional marketing for doing so. channels. On the other hand, new technologies Firm estimates of the market share of differenti- have enabled roasters to be more flexible in their ated coffee do not exist, but the figure is probably ability to make short-term switches among coffee between 6 and 8 percent of global coffee consump- types, implying that premiums for certain types of tion. Organic consumption in major consuming coffee cannot be retained for long. countries reached 700,000 bags in 2002­03 (or about Given the inability of the various supply-control 0.6 percent of global coffee consumption). In terms measures to arrest the decline in coffee prices, and of market share the highest rates of consump- in the absence of any new international initiative or tion were in Denmark (2.8 percent), Switzerland distorting domestic policies by dominant produc- (2.3 percent), Austria (2.0 percent), and Germany ers, the outlook for the coffee market rests entirely (1.2 percent), followed by the United States and on supply and demand. Neither the supply nor the Canada (1.1 percent each). Japan's share was 0.5 per- demand outlook favors a reversal of the events that cent. In the fair-trade coffee market about 240,000 shaped the coffee market during the last decade. bags were traded in 2001, 43 percent of which were Per capita coffee consumption in high-income consumed by Germany and the Netherlands. countries, where more than three-quarters of coffee To summarize, several characteristics of these is consumed,has remained virtually unchanged over "new" markets must be highlighted. First, the phe- the past decade, implying a near-zero income elas- nomenal growth of these markets reflects, in part, ticity for coffee. According to recent International a low base--implying that as a share of global Coffee Organization calculations, per capita coffee output, niche markets are small. Second, supply consumption in Western Europe declined from and demand conditions will soon saturate these 5.8 kilograms a year in 1993 to 5.5 kilograms in 1999 markets--there is increasing evidence of falling and in the United States from 4.5 kilograms a year to premiums for these coffees in some markets. Third, 4.2 kilograms (table 16.4). That is the same as the the benefits usually accrue to producers with some 1910­20 average.Annual per capita coffee consump- organizational structure, who are usually not the tion in the United States peaked at about 8 kilograms poorest. after World War II and declined to 6.5 kilograms during the 1960s, before returning to its 1910­20 average (Pan-American Coffee Bureau 1970). Synthesis Like tea, coffee faces strong competition from The coffee market may have been subject to supply the soft drink industry. In 1970 annual per capita controls longer than any other important com- consumption of soft drinks in the United States was modity. Apart from stabilizing (and perhaps rais- 86 liters; in 1999 it exceeded 200 liters, according ing) prices in the short term, these agreements to the U.S. Department of Agriculture. With the brought new entrants into the coffee market. With exception of a few coffee producers, low-income few exceptions, the trade and marketing regimes countries that have high income growth potential of coffee-producing countries are largely free of and high income elasticities for food do not con- domestic support or taxation measures. At the sume much coffee. Efforts to penetrate new mar- international level, there are no tariffs or quantita- kets (China and Russia, for example) have only tive restrictions, with the exception of some tariff recently begun. Even if such efforts succeed, two escalation on coffee products (such as soluble cof- points must be made. First, success is likely to come fee). This escalation is very small compared to at the expense of tea consumption, which is often other commodities, however. produced by the same countries that produce coffee During the 1990s Brazil expanded its coffee out- (the tea industry has also engaged in efforts to put to less frost-prone areas, thus reducing the increase consumption). Second, any increase in probability of weather-induced supply disruptions. coffee consumption by developing countries is Vietnam emerged as the dominant supplier of likely to come in the form of soluble coffee, which, robusta coffee; it now produces as much coffee as as mentioned earlier, requires lower quality beans. 308 Global Agricultural Trade and Developing Countries TABLE 16.4 Per Capita Coffee Consumption of Major Consumers, 1993­99 (kilograms per year) Country 1993 1994 1995 1996 1997 1998 1999 European Union Germany 7.93 7.53 7.37 7.16 7.22 7.01 7.46 France 5.73 5.30 5.48 5.69 5.68 5.39 5.52 Italy 5.18 5.00 4.86 4.95 5.08 5.16 5.16 Spain 4.19 4.28 4.21 4.49 4.63 4.68 5.15 United Kingdom 2.61 2.71 2.25 2.43 2.46 2.62 2.30 EU average 5.76 5.57 5.33 5.57 5.56 5.51 5.52 Japan 2.83 2.92 2.98 2.83 2.90 2.91 3.01 United States 4.50 4.01 3.98 4.10 4.00 4.14 4.24 Average 4.88 4.64 4.51 4.64 4.59 4.62 4.69 Source: International Coffee Organization. With the aggressive production prospects of RIAS (2002) found no evidence of collusion or a cartel, it is also major Asian producers, especially Vietnam; with the case that the coffee industry wishes to sell the volume that maximizes profits, which appears not to be the highest possible Brazil's expansion, considerable efficiency gains, volume. and reduced likelihood of frosts; and with weak 4. Highly liquid coffee futures contracts, where the hedge demand prospects due to low income elasticity and fund activity takes place, are traded at the New York Board of Trade for arabica and at the London International Financial strong competition from soft drinks, the outlook Futures and Options Exchange for robusta. Less liquid coffee for the coffee market is poor. While prices are contracts are traded at the São Paulo Commodity Exchange, expected to recover from their current lows when Singapore Commodity Exchange, Bangalore Commodity Exchange, and Tokyo Grains Exchange. the downward adjustment of supply takes place, 5. Quality deterioration has been presented as a negative prices are unlikely to reach the highs experienced consequence of policy reforms. However, the two studies that during the boom years of the late 1970s or the have looked at the issue in some detail, albeit for different com- mid-1990s. modities found little or no evidence of lower quality of cocoa in Cameroon (Gilbert and Tollens 2003) and cotton in Tanzania (Baffes 2004) after the reforms. Notes 6. UNCTAD (2002: 65) describes specialty coffees: "It is fair to say that `specialty coffee' has become a generic label covering a 1. Arabica typically commands a highly volatile premium range of different coffees, which either command a premium over robusta. However, a bivariate time series error-correction price over other coffees or are perceived by consumers as being model that examined the comovement of arabica and robusta different from widely available mainstream brands of coffee. prices using monthly data from January 1983 to September 2001 The term has become so broad that there is no universally found extremely low comovement. In the 1990s, for example, accepted definition of what constitutes `specialty coffee', and it the price differential fluctuated between 13 percent in October frequently means different things to different people. Given this 1995 and 156 percent in August 1997. lack of precision in definition it is extremely difficult to describe 2. The concentration of coffee production has increased the market in a global way." (from 0.11 in 1970 to 0.14 in 2000), mainly reflecting the increased shares of Brazil and Vietnam. The concentration index, also known as the Herfindahl index, is defined as the squared sum of production shares of all countries. A value of References unity indicates that a single country accounts for the entire pro- duction. Values close to zero indicate that a large number of Akiyama, T. 2001. "Coffee Market Liberalization since 1990." In countries have equal shares. Takamasa Akiyama, John Baffes, Donald Larson, and Panos 3. There has been some concern that the increasing concen- Varangis, eds., Commodity Market Reforms: Lessons from Two tration of the coffee industry has allowed for rent-seeking by the Decades. Regional and Sectoral Studies. Washington, D.C.: coffee industry. Evidence cited includes the very high profits World Bank. made by the coffee industry in times of low prices and the "stick- Akiyama, T., J. Baffes, D. Larson, and P. Varangis. 2003. "Com- iness" of retail prices, which do not fall as fast as world green cof- modity Market Reform in Africa: Some Recent Experience." fee prices. It is claimed that this reduces final demand because of Economic Systems 27: 83­115. higher-than-necessary retail prices, thus holding down world Baffes, J. (forthcoming) "Tanzania's Coffee Sector: Constraints demand for any given level of supply. Although recent work by and Challenges." Journal of International Development. Coffee: Market Setting and Policies 309 Baffes, J. 2004. "Tanzania's Cotton Sector: Reforms, Constraints Kreuger A. O. 1990. The Political Economy of Controls. Public and Challenges." Development Policy Review 22: 75­96. Policy and Development: Essays in honor of Ian Little. Oxford: Bates, R. H. 1997. Open-Economy Politics: The Political Economy Oxford University Press. of the World Coffee Trade. Princeton, N.J.: Princeton Univer- Krivonos, Ekaterina. 2003. "The Impact of Coffee Market sity Press. Reforms on Producers'Prices and Price Transmission."Devel- Bohman M., Lovell Jarvis, and R. Barichello. 1996."Rent Seeking opment Prospects Group, World Bank, Washington, D.C. and International Commodity Agreements: The Case of Lewin, B., D. Giovannucci, and P. Varangis. 2004. "Coffee Coffee." Economic Development and Cultural Change 44(2): Markets: New Paradigms in Global Supply and Demand." 379­404. Agriculture and Rural Development Internal Report, World Gilbert, C. L. 1995. "International Commodity Control: Retro- Bank, Washington, D.C. spect and Prospect." Policy Research Working Paper 1545. Pan-American Coffee Bureau. 1970. Annual Coffee Statistics. No. World Bank, International Economics Department, Com- 34. New York. modity and Policy Analysis Unit, Washington, D.C. RIAS (Rabo International Advisory Services). 2002. "Raising the Gilbert C. L., and E. F. Tollens. 2003. "Does Market Liberalisation Income of Coffee Growers." Paper prepared for the Ministry Jeopardise Export Quality? Cameroonian Cocoa, 1988­2000," of Foreign Affairs, The Netherlands. Journal of African Economies 12: 303­342. Shepherd, A. W., and S. Farolfi. 1999. Export Crop Liberalization Hutchinson, L. 1909. "Coffee `Valorization' in Brazil." Quarterly in Africa: A Review. Rome: United Nations Food and Agri- Journal of Economics 23(3): 528­35. culture Organization. IADB/USAID/WB (Inter-American Development Bank, U.S. UNCTAD (U.N. Conference on Trade and Development). 2002. Agency for International Development, World Bank). 2002. Product and Market Development--Coffee: An Exporter's "Managing the Competitive Transition of the Coffee Sector in Guide. International Trade Center, Geneva. Central America." Paper presented at the Regional Workshop Wickizer, V. D. 1943. The World Coffee Economy with Special "The Coffee Crisis and its Impact in Central America: Situa- Reference to Control Schemes. Palo Alto, Calif.: Stanford tion and Lines of Action,"Antigua, Guatemala, April 3­5. University Press. Index Page numbers followed by b, f, or t indicate box, figure, or table. ACP (African, Caribbean, and Pacific). See Cotonou in developing countries, 42­43 Agreement and trade preferences; Lomé escalation of tariffs, 48­49, 48t3.8, 49t3.9 Convention evolution of, 39­42 ACPC. See Association of Coffee Producing Countries export subsidies. See export subsidies acreage set-asides, 80 fruits and vegetables tariffs and import restrictions, ad valorem and specific duty rates, 45, 46t3.4 246­248, 246t13.7, 247t13.8, 256 aflatoxin problems, 94, 107, 227 developing countries, 248, 250t13.10 See also food safety standards groundnuts, 223­228, 232, 233 Africa African exporters, 226 See also African Growth and Opportunity Act trade liberalization scenarios, 228­234, 229t12.13, (AGOA); Southern Africa Customs Union 230t12.14 (SACU); specific countries and regions in OECD countries, 39­42 agricultural exports, 31, 108 See also OECD countries coffee production, 297 reform implications for, 50­52, 51t3.10­3.11, 52­53, cotton 118­120 exports, 260, 262 See also reform; trade liberalization production, 260, 261 seafood, 280, 283­287, 287­288 reform, effect of, 268, 269­271 tariff levels, 42, 46­48, 47t3.6, 48t3.7 dried fruit exports, 107­108 tariff rate quotas, 49, 50f3.6 reform, effect on, 120 fruits and vegetables, 251 rice and trade liberalization, 190 sugar, 148, 149, 156 seafood exports, 279 tariff transparency, 44­46, 45f3.5 trade preferences, 65­67 ad valorem and specific duty rates, 45, 46t3.4 African Growth and Opportunity Act (AGOA) non-ad valorem lines as share of total, clothing exports, 57 44­45, 45f3.5 U.S. preferences under, 59, 59t4.2, 60t4.3 non-ad valorem lines by degree of processing, cumulation permitted, 70 46, 46t3.5 fruits and vegetables, 249, 251­252 Uruguay Round Agreement and, 38­39 Agenda 2000 (EU), 82 wheat tariff rates and quotas, 200­203, 200t11.3, Agreement on the Application of Sanitary and 202t11.4 Phytosanitary Measures. See Sanitary and agricultural trade Phytosanitary Agreement evolution of. See changes in agricultural trade flows Agreements on Textiles and Clothing (ATC), 270 importance of, 1 Agricultural Act of 1949 (U.S.), 86 policies. See agricultural protection; specific agricultural productivity and reform, 128­131 commodities aggregate welfare, 130, 130f 7.4 reform. See reform; trade liberalization high-productivity assumption, 130­131, agricultural trade shares, 23­26, 23t2.6, 25t2.9 131t7.12­7.13 excluding intra-EU and intra-NAFTA trade, 25, 26t2.10 output impact, 129­130, 129t7.11 alcoholic beverages trade impact, 128­129, 129t7.10 growth in trade of, 28 agricultural protection, 5­6, 37­53 tariff lines, 45 See also specific commodities Algeria aquaculture, 291 dried fruit exports, 108 coffee, 298 wheat imports, 206 cotton tariffs, 264 Andean Group and dairy product trade, 168 current structure of, 43­50, 44t3.3 Andean Trade Preference Act, 249 decoupling. See decoupling Angola fishing access, 283 311 312 Index animal diseases, 103, 104 shrimp production, 276 antidumping duties wheat exports, 211­212 HFCS, 149 Asia Pacific Economic Cooperation (APEC) orange juice, 249b13.2 milk and dairy products, 168 seafood, 286­287 asparagus exports, 254b13.3 APEC countries and seafood trade liberalization, 288 Association of Coffee Producing Countries (ACPC), aquaculture 298, 304 See also seafood Australia carp, 292 agricultural protection levels, 43­44 environmental effects, 277, 291, 293 cotton expansion of, 276, 278­279 exports, 260 fish meal, effect on, 291­292, 294 genetically modified, 263 production, 276, 277f15.2, 278, 278f15.3 production, 259, 260 salmon farming, 291­292 reform, effect of, 268 shellfish, 292­293 exporting of agricultural products, 26 shrimp, 276, 277, 277f15.2 milk and dairy products See also shrimp production and manufacturing, 161, 162 Thailand and Bangladesh, 276 reform, effect of, 171, 174 trade liberalization, impact of, 291­293, trade and domestic policy, 165­166 292t15.5­15.6 trade trends, 164 Arabica coffee, 297 rice exporters, effect of liberalization on, 189 Argentina sugar cotton tariffs, 264 exports, 146 fishing access, 283 reform, effect of, 151 food safety complaints, 103 trade liberalization, impact of, 124, 125 fruits and vegetables exports, 244 wage changes for unskilled labor, 125 groundnuts wheat exports, 108 export credits for, 203 global trade in, 218 as major exporter, 205 policies, 215, 225­226, 233, 234 production and yields, 197 prices, 220 Austrian organic coffee, 307 HFCS production, 146 avian flu, 103 as major exporter, 28 milk production, 162, 163 BAAC (Bank for Agriculture and Agricultural reform, impact of, 28, 124 Cooperatives), 182 high-productivity assumption, 130 bagasse, 141­142 rice exporters, effect of liberalization on, 190 Bangladesh trade liberalization, impact of, 125 food safety, costs of compliance in shrimp industry, wage changes for unskilled labor, 125 97, 98t6.1 wheat, 201 rice, 182 domestic support, 205 seafood bans for food safety reasons, 286 as major exporter, 205, 206 shrimp Arkansas Global Rice Model (AGRM) to study domestic food safety, costs of compliance in, 97 rice policies, 190­191 sustainable aquaculture in, 277 artificial sweeteners, 148 Bank for Agriculture and Agricultural Cooperatives Asia (BAAC), 182 See also specific countries and regions Benin and cotton reform initiatives, 270 coffee production and yields, 297, 299, 300 beverages poverty, 18 growth in trade of, 28, 30 protectionism of, 3 soft drink competition with coffee, 307 rice tariff lines, 45 production and consumption, 177 bilateral trade agreements trade liberalization, effect of, 188­189 fruits and vegetables, 249, 250 seafood groundnut trade liberalization, 228­233 consumption, 275 milk and dairy products, 168 fishmeal production, 278 trade liberalization and, 4 Index 313 Bogor Declaration, 288 groundnuts Brannan Plan of 1949 (U.S.), 78 market for, 217 Brazil policies, 227­228 coffee HFCS production, 146 association of producers, 304 milk and dairy products consumption, 297, 298, 299, 307 production trends, 163 domestic policies, 305 reform, effect of, 170, 171, 173, 174 output, 297 trade and domestic policy, 165­166 price and price stabilization, 298, 300, organic coffee, 307 303b16.1 rice trade liberalization, effect of, 190 production, 297, 298­299, 300, 307, 308 salmon farming, 291­292 rent-seeking behavior by governments and seafood marketing boards, 304­305 exports, 279 supply controls, 302­304 tariffs, 283 types grown, 297 trade liberalization, impact of, 124, 125, 233 cotton wheat government assistance, 264 export credits for, 203 production, 260 as major exporter, 205 reform, effect of, 269 Cancún Ministerial Meetings (2003), 37, 50, 175 tariffs, 264 capital market, change due to reform, 125­126, 127t7.8 WTO process initiated by, 269, 271 Caribbean Basin Economic Recovery Act, 249 food safety complaints, 103 Caribbean Basin Initiative, 68, 156 fruits and vegetables tariffs, 248 Caribbean countries and sugar reform, 152 groundnut exports, 108, 218 carp production, 276, 292 as major exporter, 28 catfish, 287 milk and dairy products Catfish Farmers of America, 287 production trends, 163 cattle headage payments (EU), 83 trade trends, 164­165 Central America. See Latin America orange juice trade dispute, 249b13.2 cephalopods, 279 reform, impact of, 28, 124 changes in agricultural trade flows, 4­5, 17­35 high-productivity assumption, 130 disaggregated export and import performance, 28, rice trade liberalization, effect of, 190 29t2.12 sugar export shares by product and region, 31, 33t2.16 employment, 145 growth in trade, 22­23, 22t2.5 ethanol production, 143, 147 distribution of, 26­28, 27t2.11 policies, 147­148, 150 output growth rates, 23, 23t2.8 production and trade, 145 poverty and, 17­20 reform, effect of, 151, 158 price decline, 23, 23t2.7 wheat, 201 processing, degree of, 30­31, 31t2.14, 32t2.15 imports, 206, 208 ratio of farm household income British Retail Consortium Technical Food to all household income for high-income countries, Standard, 95 18, 19f2.2 Bt cotton, 263­264, 272 to nonfarm household income for developing bulk-commodity approach to export expansion, 4 countries, 18, 19f2.1 Burundi coffee exports, 297, 301 ratio of farm income to total income of farm households for high-income countries, 20, 21f2.3 Cahill, Sean A., 77 role of demand and changes in market share, Cairns Group and dairy products, 170 23, 24b2.1 Canada rural income and, 17­21, 18t2.1­2.2 See also Quad countries income distribution, 20­21, 21t2.4 agricultural protection levels, 44 structure of income sources, 20, 20t2.3 tariff lines, 45 trade shares, 23­26, 23t2.6, 25t2.9 Crow Rate program for grain shippers, 79, 86­87 excluding intra-EU and intra-NAFTA trade, fruits and vegetables 25, 26t2.10 domestic support, 245 trade structure changes, 28­30, 30t2.13 tariffs, 246 trends, 21­31 314 Index Chile exports, 212 food safety complaints, 103 imports, 206, 209 free trade agreement with U.S., 249 production and yields, 196 fruits and vegetables exports, 244 Clean Air Act amendments of 1990 (U.S.), 143 milk and dairy products, 168 Codex Alimentarius Commission, 93 salmon farming, 291­292 coffee, 13, 297­309 seafood certification of nontraditional, 306­307 antidumping and countervailing measures, changing structure of coffee market, 297, 298t16.1 286­287 consumption of, 297­298, 299, 300, 306­308, 308t16.4 fishmeal production, 278 decline in trade, 28 production, 276 developing countries, importance to, 297, 301, tariffs, 283 304t16.3 China differentiated, 306­307 carp production, 276 as distortion-free market, 298 coffee, 307 domestic policy, 305­306 cotton global balance and price trends, 298­302, exports, 262 301­302f16.1 genetically modified, 263­264 international policy, 302­305 government assistance, 264, 266­267 low-quality beans, demand for, 297­298, 307 production and consumption, 259, 260, 263 market outlook, 307­308 reform, effect of, 270 new markets for, 307 tariffs, 264 niche marketing and changing patterns of trade in, 260­261 consumption, 297­298, 306­307 food deficit trends, 118 price-stabilization schemes, 298, 303b16.1 fruits and vegetables production, 297­298, 299, 299t16.2, 300 exports, 238, 244 rent-seeking behavior by governments and marketing unit values, 241 boards, 304­305 groundnuts soft drink competition, 307, 308 exports, 108, 217 supply controls, 298, 302, 303b16.1, 304 global trade in, 218 tariffs, 298 policies, 226, 228, 234 types of, 297 prices, 220 Colombian coffee production, 216­217 association of producers, 304 trade liberalization scenarios, 229, 232, 233­234 consumption, 299 milk and dairy products, 168­169 output, 297, 299 production trends, 163 production, 307 reform, impact of, 3, 118, 124, 125, 127 commodities high-productivity assumption, 130 See also specific commodities rice, 179­180 change of product mix in global trade, 28­30, 30t2.13, exporters, effect of liberalization on, 189 34, 69 seafood Conservation Reserve Program (U.S.), 205­206 antidumping and countervailing measures, corn and aflatoxin problems, 107 286­287 corn starch, 142 consumption, 275 Costa Rican coffee prices, 302 crawfish, 286­287 Côte D'Ivoire and cotton reform initiatives, 270 exports, 279 Cotonou Agreement and trade preferences, 61, 63, fishmeal production and use, 278 63t4.5­4.6, 68 production, 276 rules of origin and, 70 reprocessing market for U.S. and Norway, 280 cotton, 12, 259­273 subsidies, 290 See also specific countries and regions sugar, 145, 147 border protection and, 3 policies, 148, 150 developing countries reform, effect of, 151 economic importance of, 262­263, 262t14.1, 271 trade surplus of, 28 production, 259 wage changes for unskilled labor, 125 genetically modified, 263­264, 272 wheat government assistance, 264, 265t14.2 Index 315 historical background, 259 New Zealand exit grant to farmers, 79, 87 importing, 261 U.S. peanut quota, 79, 86, 157 market distortions, 264­268 politics of, 77­78 reform of, 268­270 recommendations for, 88­89 market share vs. polyester, 259, 260f14.1, 271 U.S. cotton payments, 264­265, 269 nonconventional production, 263­264 "Decoupling: The Concept and Its Future in Canada," 76 organic, 263­264, 272 Denmark phase-out of Agreements on Textiles and Clothing farm household incomes, 18 (ATC), 270 organic coffee, 307 pricing, 261­262, 271, 272 seafood imports, 280 production and consumption, 259­261 developing countries stockholding, 260 See also specific countries and regions subsidies, effect of, 3, 50, 271 agricultural protection policies of, 37­38, 42­43, tariffs, 264 43t3.2, 52 trade liberalization, effect of, 268­270, 269t14.4, 271 most-favored-nation tariffs for, 42, 43f3.4 country of origin agricultural trade flow of, 25, 26­28, 27t2.11 labeling of seafood, 285 agricultural trade shares of, 23­26, 23t2.6, 31 rules of origin and trade preferences, 57, 69­70 cotton crab and lobster, 279, 280 economic importance of, 262­263, 262t14.1, 271 crawfish, 287 production, 259 Crow Rate program for grain shippers (Canada), food safety issues, 97­99, 101, 103, 111­112 79, 86­87 disputes and complaints through WTO, 103­105 crustaceans, 280 fruits and vegetables exports, 237, 244, 244t13.4, Cuban sugar 245t13.5, 256 exports, 146 growth in trade, 22, 28, 29, 34 reform, effect of, 151 export shares by product and region, 31, 33t2.16 Cyclospora, 109 milk and dairy products Czech Republic and dairy products, 166, 168 market growth prospects for, 175­176 production trends, 164 dairy products, 9­10, 161­176 reform, effect of, 30, 123­124 See also specific countries agricultural productivity, 128­130 background, 162­163 supply response, 132­134, 133t7.14 developing economies, prospects for, 175­176 rice trade and policies, 179 milk components as factor in sector growth, 175 trade liberalization, effect of, 190 milk production and dairy product manufacturing, rural household incomes, 20, 20t2.3 162­163 sugar production, 157­158 subsidies, effect of, 3 tariff rates, 42, 47 tariff lines, 45 trade preferences and least-developed countries, trade liberalization and, 3, 4, 161­162, 169­174 55­73 full liberalization scenario, 170­174, 172t9.3 See also trade preferences other scenarios, 174 wheat prospects for, 175 food aid programs, 195 trends in trade, 164­165 importing of, 198 decoupling, 6­7, 75­89 development strategy based on agricultural exports, 3 assessment of, 87­89 dextrose, 142 definition of, 75­77 disaggregated export and import performance, economics of, 78 28, 29t2.12 EU cotton payments, 269 diversification and trade preferences, 68­69 experience with broad attempts, 78­86 Doha Declaration EU, 81­83, 82t5.3 fishing subsidies, 282 Mexico, 83­85, 84t5.4 Doha Round, 57, 232 Turkey, 85­86, 85t5.5 Doi Moi, 300 U.S., 79­81, 80t5.2 Domestic Food Aid Plan (Czech Republic), 166 experience with one-time buyouts, 86­87 domestic subsidies, effect of, 3 Canadian Crow Rate program for grain shippers, See also specific commodities 79, 86­87 decoupling. See decoupling 316 Index East Asia aquaculture tariffs, 291 See also specific countries cattle headage payments, 83 cotton Common Agricultural Policy (CAP), 81­83 consumption, 260 cotton, 266 importing, 261 decoupled payments and, 77, 78 growth in importing, 28 sugar, 152 milk and dairy products, 168­169 wheat, 201, 205 trade and domestic policy, 168­169 cotton reform, impact of, 124, 125 decoupled payments, 269 Eastern Europe government assistance, 264, 266 See also specific countries reform, effect of, 268 agricultural protection reductions, 42 decoupling agricultural support, 77, 81­83, 82t5.3, 88 milk and dairy products Economic Partnership Agreements (EPAs), 154 production trends, 163 Everything But Arms. See Everything But Arms trade and domestic policy, 166­168 initiative (EBA) trade trends, 164 export trade subsidies, 50 wheat imports, 206 farm subsidies, 41, 126 EBA. See Everything But Arms initiative final products, shares of, 31, 34 eco-friendly coffee, 306­307 fish bans, 105­106 eco-labeling of seafood, 285 food safety standards, 94 Economic Partnership Agreements (EPAs) and sugar aflatoxins, 107 policies, 154 complaints from developing countries, 104 Economic Research Service of USDA on Japanese sugar EURECAP Fruit and Vegetable Standard, 95­96 policies, 155 fish bans, 105­106 EEZ. See exclusive economic zones harmonization of measures, 105 Egypt product rejection data, 100­103, 102t6.2 cotton recognition of developing country standards, 96 government assistance, 264 fruits and vegetables tariffs, 264 domestic support, 245 groundnut exports, 108 entry price scheme, 247b13.1 rice exporters, effect of liberalization on, 189 as largest market and supplier, 241­242, 243 wheat imports, 206, 207­208 preference system, 60, 249­250, 250f13.7, 252, El Salvador 253f13.10 coffee prices, 302 production growth, 238 environmental effects tariffs, 246, 248 aquaculture, 277, 293 unit values, 241 shrimp aquaculture, 277, 291 groundnuts Uganda's fishing industry, 282b15.1 market for, 217 EPAs (Economic Partnership Agreements) and sugar policies of, 227­228 policies, 154 prices, 221 Essential Commodities Act of 1955 (India), 148 rejection for food safety reasons, 101 ethanol, 143, 147, 148 technical standards for, 220, 227 Ethiopia trade liberalization scenarios, 229 coffee Mac Sharry reform, 81 consumption, 299 milk and dairy products exports, 297, 301 reform, effect of, 170, 171, 173, 174, 175 income sources, 20 tariff rates, 47 trade preferences, 67 trade and domestic policy, 165­166 EURECAP Fruit and Vegetable Standard, 95­96 organic coffee, 306 European Union (EU) rice, 184 See also Quad countries domestic trade and policy reforms, 191 Agenda 2000, 82 importers, effect of liberalization on, 190 agricultural protection levels, 40­41, 44 policy distortions, 185 average tariffs, 47 seafood tariff lines, 45, 46 bans, 282b15.1, 286 agricultural trade shares of, 24­25, 26 consumption, 275 Index 317 country-of-origin labels, 285 peanut buyout, 79, 86 fishing access, 283, 284, 293 rice, 184 fishing sector in Mauritania, 283, 284b15.2 wheat, 206 fishmeal production, 278 FDA. See Food and Drug Administration (FDA, U.S.) imports, 279­280 Federal Agricultural Improvement and Reform (FAIR) management styles, 290 Act of 1996 (U.S.), 80­81 subsidies, 290 feed, use of wheat for, 198 tariffs, 283­284 Fiji self-sufficiency of, 34 sugar sugar industries employment, 145 enlargement issues, 154 reform, effect of, 152 HFCS production, 146 trade preferences, 68 as major global producer and exporter, 145, 146 fish. See seafood protection, 141, 142 Florida and sugar production, 157 reform, effect of, 151, 152­154, 158 food aid programs, effect on wheat, 195, 203, 203f11.5 tariff transparency, 45 Food and Agricultural Policy Research Institute on trade liberalization, impact of, 124, 125, 233 liberalization's effect on wheat, 212 trade preferences under GSP and Cotonou Food and Agriculture Organization (FAO) Agreement, 58, 60­63, 61t4.4, 63t4.5­4.6 on cotton production, 263 entry price of tomatoes, 62b4.1 historical trends of growth, 118 proportion of trade covered by, 65, 66t4.11 Food and Drug Administration (FDA, U.S.) rules of origin and, 70 Center for Food Safety and Applied Nutrition, 94 value of, 68, 68f4.2, 71 Guatemalan fruits, standards for, 109, 110 wage changes for unskilled labor, 125 increased inspections by, 100 wheat Food Corporation of India and rice importation, 180 domestic support of, 195, 204 food safety standards, 7­8, 91­114 export subsidies, 200, 203 See also specific commodities as major exporter, 205 asparagus exports, 254b13.3 processed wheat products trade, 201 capacity to comply with minimum standards, 96­97 production and yields, 196 costs of compliance, 97­99 Everything But Arms initiative (EBA), 57, 61, 82, effect of, 91­93 153­154, 249, 250 exports of high-value agro-food products and, rules of origin and, 70 99­105 exchange rates, 42 border detentions and rejections of agricultural exclusive economic zones (EEZ), 279, 280­281, 283 and food products, 100­103, 102t6.2 export diversification and trade preferences, 68­69 WTO disputes and complaints, 103­105, 104t6.3 Export Enhancement Program (U.S.), 203 fruits and vegetables, 109­111 export shares by product and region, 31, 33t2.16 Guatemalan raspberries, 109­110 export subsidies, 49­50 horticultural product standards, proliferation of, See also specific commodities 109­111 export taxes, 37 Kenyan fresh produce, 110­111 Argentina's groundnut policy, 225 mycotoxins and aflatoxins, 107­109 as political issues, 77 FAIR Act. See Federal Agricultural Improvement and Sanitary and Phytosanitary Agreement, 93­96, 293 Reform (FAIR) Act of 1996 seafood, 95, 102, 105­107, 282b15.1, 285­286, fair trade coffee, 306­307 293, 294 FAO. See Food and Agriculture Organization viewed as barriers, 4 Farm Bill of 1985 (U.S.), 78 wheat, 201 Farm Bill of 1996 (U.S.) Food Security Act of 1985, 80 cotton, 265 food security and wheat, 203­204, 213 decoupling and, 78­79, 80, 88 foot and mouth disease controls, 103, 104 Farm Bill of 2002 (U.S.) Former Soviet Union (FSU) cotton, 265 See also Russia dairy products, 175 cotton decoupling, effect of, 81 economic importance of, 262­263 groundnuts, 215, 217, 222­223, 224, 233 production, 260 318 Index Former Soviet Union (continued) policies, 226­227 milk and dairy products prices, 220 production trends, 163 trade liberalization scenarios, 229, 232 trade trends, 164 GAP (Good Agricultural Practice), 254b13.3 wheat GATT (General Agreement on Tariffs and Trade) exports, 209­211, 210f11.9 disputes. See World Trade Organization (WTO) feed use, 198 Enabling Clause and trade preferences, 56 imports, 206 fishing subsidies, 280 production and yields, 197 Standards Code, 93 tariff protection, 201 sugar tariff rate quotas for U.S., 156 France Generalized System of Preferences (GSP), 56 See also European Union (EU) EU preferences under, 60­63, 63t4.5­4.6 coffee consumption, 297, 299 fruits and vegetables, 249 farm household incomes, 18 harmonizing preference regimes as goal of, 57 seafood Japanese preferences under, 63­64, 64t4.8­4.9 fishing access, 283 rules of origin and, 70 imports, 280 sugar duties and, 156 Free Trade Area of the Americas (FTAA), 168 U.S. preferences under, 59­60, 59t4.2, 60t4.3 fructose, 142 genetically modified cotton, 263, 272 See also high-fructose corn syrup (HFCS) Germany fruits and vegetables, 11­12, 237­257 See also European Union (EU) See also specific countries and regions coffee developing countries as exporters, 237, 244, 244t13.4, consumption, 297, 299 245t13.5, 256 fair trade, 307 domestic support and export subsidies, 245­246, organic, 307 245t13.6 seafood imports, 280 dried fruit exports from Africa, 107­108 global merchandise trade reform, 118­120, EU entry price scheme, 247b13.1 119t7.3 EURECAP Fruit and Vegetable Standard, 95­96 caveats about, 119­120 exports, 241­244, 242t13.3 tariffs' role, 120, 121t7.4 determinants of success, 252­256 Global Trade Analysis Project (GTAP) food safety standards, 109­111 data problems, 119 global trade patterns, 241­244 sugar policies and, 151 Guatemala programs, 109­111 trade elasticities, 134 imports, 238, 239t13.1, 240f13.3­13.4, 241t13.2, Good Agricultural Practice (GAP), 254b13.3 242t13.3, 256 Goodwin, Barry K. and Ashok K. Mishra, 77 preferential market access, 248­352 Greece EU hierarchy of preferences, 249­250, 250f13.7 See also European Union (EU) pricing, 240­241, 241f13.5 cotton production, 259, 260, 261 processed, 239­240, 240f 13.4 fishing access, 283 preferential access, 255 Green Revolution, 196­197, 211 tariffs, 248, 248t13.9 groundnuts, 11, 215­235 production and trade growth, 28, 30, 237­241, See also specific countries and regions 238f13.2, 239f 13.3­13.4 African exports, 108, 215, 226­227, 227t12.11 tariff rate quotas, 251 confectionery, 227, 234 tariffs and import restrictions, 246­248, 246t13.7, domestic policies of major countries in world 247t13.8, 256 markets, 221­228 developing countries, 248, 250t13.10 African exporters, 226­227, 227t12.11 tomatoes and EU system of entry prices, 62b4.1 Argentina, China, and India, 224­226, trends in production and trade, 238­241, 225t12.10, 233 238f13.1 high-income importers, policies of, 227­228, urbanization, effect of, 244 228t12.12 FSU. See Former Soviet Union U.S. policies, 222­224, 222t12.5, 223t12.6, FTAA (Free Trade Area of the Americas), 168 224t12.8­12.9, 233 global trade in, 217­220 Gambian groundnuts consumption, exports, and market shares, global trade in, 108, 217, 218­220 218, 219f12.1 Index 319 share of groundnut products in total merchandise ICAC. See International Cotton Advisory Committee exports, 218, 220t12.4 ICAs. See International Coffee Agreements value of net exports by product, 217, 218t12.3 Iceland agricultural protection levels, 43 high-income importers, policies of, 227­228 ICO. See International Coffee Organization international prices of, 220­221, 231­233 India Rotterdam prices, 220, 221f12.3 cotton unit price of edible raw groundnuts in Gambia, production and consumption, 259, 260, 263 Senegal, and South Africa, 220, 221f12.2 tariffs, 264 oil markets, 221 trade in, 261 patterns of trade, 229 food safety complaints, 103 policy reforms, impact of, 228­233, 229t12.13, fruits and vegetables 230t12.14 exports, 238 production, 215­217, 216t12.1, 217t12.2 tariffs, 248 trade liberalization scenarios, 3, 215, 228­234, groundnuts, 3 229t12.13, 230t12.14 policies, 226, 228, 233 U.S., 222­224 production, 216 welfare gains, 229, 231 trade liberalization scenarios, 229, 232, 233­234 Group of Three and dairy product trade, 168 milk and dairy products, 168­169 growth in trade, 22­23, 22t2.5 production trends, 163 See also specific commodities rice, 180­181 distribution of, 26­28, 27t2.11 exporters, effect of liberalization on, 189 GSP. See Generalized System of Preferences seafood GTAP. See Global Trade Analysis Project food safety issues, 99, 286 Guatemala tariffs, 283 coffee sugar exports, 297 ethanol production, 148 prices, 302 as largest global consumer, 146 fruits and vegetables exports, 244 as major global producer, 145, 147 raspberries and food safety standards, 109­110 policies, 148, 150 rice importers, effect of liberalization on, 190 trade liberalization and, 3 trade surplus of, 28 Haiti and trade preferences, 68 wheat Harbison, Stuart, 50­52 exports, 211 hazard-analysis, critical-control-point (HACCP) production and yields, 196­197, 197f11.1 principles, 97, 98, 285­286 Indonesia seafood imports, 285­286 coffee output and consumption, 299 Hennessy, David A., 77 fruits and vegetables exports, 244 herring, 280 milk and dairy products, 168­169 high-fructose corn syrup (HFCS), 142­143, 146, 149 rice, 181­182, 189 EU quotas, 153 sugar, 151, 158 Japan, 146, 155 tariff rates, 47 U.S., 142­143, 143f 8.2, 146, 156 wheat imports, 201, 207 high-income countries. See industrial countries; OECD industrial countries countries See also OECD countries; specific countries and regions Honduras coffee agricultural protection policies, 37 exports, 297 average tariffs, 47 prices, 302 Harbinson reform proposal, 50­52 Hong Kong agricultural support, effect of, 21 liberal trade regime of, 28 demand and changes in market share, 24b2.1 rice trade liberalization, effect of, 190 farm income, distribution of, 20, 31 horticultural trade import decline, 22, 28 EU, 243 reform, effect on, 121­122, 123­124 export subsidies, 245, 245t13.6 agricultural productivity, 128­130 opportunities for reform and, 4, 256 Integrated Framework for Trade-Related Technical pesticide residues, 95, 109 Assistance, 72 product standards, proliferation of, 109­111 Integrated Program for Agricultural and Environmental Household Responsibility System (China), 266 Protection (Guatemala), 110 320 Index Inter-American Coffee Agreements, 298 rice, 183 International Coffee Agreements (ICAs), 298, 303b16.1, policy distortions, 38, 185 304­305 seafood International Coffee Conference (N.Y. 1902), 303 consumption, 275 International Coffee Organization (ICO), 298, fishmeal production, 278 303b16.1, 304 imports, 279 International Cotton Advisory Committee (ICAC), 261, subsidies, 290 264, 267, 268 tariffs, 284 International Food Policy Research Institute, 105 trade preferences, 68 International Law of the Sea, 279 sugar, 154­155, 154f 8.4 International Maize and Wheat Improvement HFCS production and consumption, 146, 155 Center, 196 protection, 141, 142 International Office of Epizooties, 93, 103 reform, effect of, 151, 158 International Plant Protection Convention, 93 trade liberalization, impact of, 125 International Sugar Organization on China's sugar tariff trade preferences under GSP, 58, 63­64, rate quota, 148 64t4.7­4.9 inulin, 153 proportion of trade covered by, 65­67, 66t4.12 Iran rules of origin and, 70 groundnut exports, 108 value of, 68, 69f4.3 wheat imports, 207 wage changes for unskilled labor, 125 Ireland wheat milk production and dairy product manufacturing, 162 imports, 201, 206, 208­209 salmon farming, 291­292 support to domestic producers, 205 isoglucose. See high-fructose corn syrup (HFCS) Israel Kazakhstan cotton production, 261 See also Former Soviet Union (FSU) free trade agreement, 249 wheat exports, 209­210 minimum import commitments for sheep and dairy Kenya products, 38 fruits and vegetables exports, 244 Italy food safety standards, 110­111 coffee consumption, 297, 299 seafood exports and food safety standards, 98, 99, seafood 106­107, 286 bans for food safety reasons, 286 sugar industries employment, 145 fishing access, 283 wheat imports, 201 imports, 280 Korea, Republic of agricultural protection levels, 42, 44 Japan average tariffs, 47 See also Quad countries aquaculture tariffs, 291 agricultural protection levels, 44 fruits and vegetables tariffs, 246, 248 average tariffs, 39 groundnut policies of, 228 tariff lines, 45 HFCS production, 146 aquaculture tariffs, 291 as major importer, 28 coffee milk and dairy products, 168­169 consumption, 297, 299 production trends, 163 organic, 306, 307 rice, 183 farm household incomes, 19 importers, effect of liberalization on, 188 fruits and vegetables policy distortions, 38, 185 domestic support, 245 seafood tariffs, 283­284 imports, 243 tariff rates, 47 tariffs, 248 taxation to finance decoupling, 78 unit values, 241 trade liberalization, impact of, 125 groundnuts wheat imports, 201, 207 global trade in, 217 policies of, 228 labeling of seafood, 285 importing, 26, 28, 213 names, regulation of, 287 milk and dairy products, 168­169 land contribution to agricultural income and impact of reform, effect of, 171 reform, 126, 128t7.9 Index 321 Latin America Malaysia See also specific countries milk and dairy products, 168­169 agricultural protection, evolution of, 43 rice trade liberalization, effect of, 190 agricultural surplus trends, 118 seafood tariffs, 283 coffee tariff rates, 47 domestic policies, 305 wheat imports, 201 output, 297, 300 Maldives seafood exports, 275, 280 prices, 301­302 Mansholt Plan of 1968 (Europe), 78, 81 production, 300 market share regulation of exports, 303b16.1 cotton vs. polyester, 259, 260f14.1, 271 type grown in, 297 groundnuts, 218, 219f12.1 yields, 299 role of demand and changes in, 23, 24b2.1 cotton production, 260 Sub-Saharan Africa, 31 fishmeal production, 278 Mauritania fishing access, 283, 284b15.2 fruits and vegetables exports, 238, 244 Mauritius milk and dairy products sugar industries employment, 145 production trends, 163 trade preferences, 68, 71 reform, effect of, 173 Mercosur countries trade and domestic policy, 168 EU free trade negotiations with, 82 trade trends, 164 milk and dairy product trade, 168 reform, impact of, 124­125 reform, effect of, 173 rice rice trade, 192 policy distortions, 185 wheat trade, 201, 208 trade liberalization, effect of, 190 Mexico seafood exports, 279 See also NAFTA countries shrimp production, 278 agricultural protection levels, 42 sugar, effect of trade reform, 151 coffee output and consumption, 299 wage changes for unskilled labor, 125 Conasupo programs, 83 least-developed countries and trade preferences, cotton and government assistance, 264 55­73 decoupling agricultural support, 83­85, 84t5.4, 88 See also trade preferences fruits and vegetables exports, 109­110, 244 liberalization. See reform; trade liberalization milk and dairy products licensing production and manufacturing, 163 elimination in developing countries, 42 trade trends, 164 wheat importers, 202 National Program for Direct Assistance to Rural LINKAGE model and simulations, 115­118 Areas, 78 agriculture and food trends, 117­118, 117t7.1­7.2 Procampo program, 83­84 baseline simulation, 116 rice importers, effect of liberalization on, 190 LMC International on sugar production costs, 144, rural household incomes, 20 144t8.1 sugar Lomé Convention HFCS production, 146 fruits and vegetables, 249, 250, 251, 251f13.8 policies, 148­149, 150, 156 sugar protocol, 153 reform, effect of, 151 low-income countries. See developing countries trade liberalization, impact of, 134, 233 wheat imports, 201 Mac Sharry reform (EU), 81 microanalysis, importance of, 2­3 Madagascar seafood Middle East bans for food safety reasons, 286 See also specific countries shrimp industry, 277­278 milk and dairy products Malawi trade and domestic policy, 169 groundnuts, 227 trade trends, 164 exports, 217 reform, effect on, 120 global trade in, 218, 220 rice importers, effect of liberalization on, 190 production, 216 milk. See dairy products trade liberalization scenarios, 229, 232 Milk Income Loss Contract program (U.S.), 175 income sources, 20 molasses, 142 sugar and trade preferences, 68 mollusks, 276, 279 322 Index MONASH model and trade elasticities, 134 Nigerian groundnuts Morocco global trade in, 217, 218 fruits and vegetables policies, 227 preferential agreement with EU, 252 prices, 220 tariffs, 248 production, 216 tariff rates, 47 trade liberalization scenarios, 229, 231, 232, 234 most-favored-nation tariffs Nile perch and food safety standards, 106­107, 286 developing countries, 42, 43f3.4 non-ad valorem lines fruits and vegetables, 246, 248 by degree of processing, 46, 46t3.5 Mozambique as share of total, 44­45, 45f3.5 seafood nontariff barriers bans for food safety reasons, 286 See also specific types (e.g., licensing) exports, 275, 280 removal of, effect of, 38 sugar and trade preferences, 68 wheat and, 201­202 trade preferences, 67 North Africa mycotoxins and food safety standards, 101, 107­109, 201 See also specific countries milk and dairy products NAFTA countries trade and domestic policy, 169 agricultural trade shares of, 24­25, 26 trade trends, 164 final products, shares of, 31, 34 North American Free Trade Agreement. See NAFTA fruits and vegetables intraregional trade, 242, 249 countries groundnut tariffs, 223­224 Norway peanuts, 86 agricultural protection levels, 43 rice trade, 192 tariff lines, 45 sugar trade, 156 salmon farming, 291­292 side-letter agreement between U.S. and Mexico, seafood 149, 156 antidumping and countervailing measures, wheat 286­287 processed wheat products trade, 201 exports, 279 tariffs, 200­201 wheat support to domestic producers, 205 Namibia, fruits and vegetables tariff rate quotas, 251 Nash, Eric F., 76 Oceania. See Australia; New Zealand National Program for Direct Assistance to Rural Areas OECD countries (Mexico), 78 See also Quad countries; specific countries Netherlands agricultural protection fair trade coffee, 307 evolution of, 39­42, 40f3.1, 41f3.2, 42f3.3, 52 farm household incomes, 18 market price support, 39 New Deal farm programs, 79 percentage attributable to border protection and New Zealand direct subsidies, 40­41, 41t3.1 agricultural protection levels, 43­44 product coverage of tariffs, 47­48 exit grant to farmers, 79, 87 seafood, 283, 284­285 exporting of agricultural products, 26 structure of, 43­44, 44t3.3 milk and dairy products agricultural trade shares of, 24 production and manufacturing, 161, 162 domestic subsidies in, 3 reform, effect of, 171, 174 export trade subsidies, 50 trade and domestic policy, 165­166 farm household incomes, 18 trade trends, 164 fruits and vegetables trade liberalization, impact of, 124, 125 domestic subsidies, 245 wage changes for unskilled labor, 125 tariffs, 246, 247t13.8, 248, 248t13.9 wheat production and yields, 197 sugar policies, 150­157, 158 Nicaragua tariff rates, 39­40, 40f3.1, 47­48 coffee quotas, 49 exports, 297 trade liberalization and, 233 prices, 302 wheat food safety, costs of compliance in shrimp industry, domestic support, 204­205, 204t11.5, 205t11.7 97, 98t6.1, 99 exports, 195 rice importers, effect of liberalization on, 190 OPEC countries and wheat imports, 206 Index 323 orange juice trade dispute, 248, 249b13.2 processing, degree of, 30­31, 31t2.14, 32t2.15, 34 organic coffee, 306­307 non-ad valorem lines by, 46, 46t3.5 organic cotton, 263­264, 272 tariff escalation and, 49 output growth rates, 23, 23t2.8 wheat products, 201 producer subsidy equivalent (PSEs) and seafood, 288 Pakistan protection. See agricultural protection cotton production and consumption, 259, 260, 263 Quad countries trade in, 261 agricultural protection levels, 44, 45 wheat exports, 212 reform, effect on, 120 peanuts underestimation of tariffs, 46 mycotoxins. See mycotoxins and food safety standards fruits and vegetables tariffs, 246, 246t13.7 U.S. quota buyout, 79, 86, 157 milk and dairy products pelagics, 276, 279 production trends, 163 Peru trade and domestic policy, 161, 164, 165­166 asparagus exports, 254b13.3 quotas. See agricultural protection milk and dairy products, 168 seafood Railway Act of 1925 (Canada), 86 exports, 275, 280 raspberries and food safety standards, 109­110 fishmeal production, 278 real income gains from reform, 120­123, 121t7.5, production, 276 122f 7.1 Peruvian Commission for Export Promotion, red snapper, regulation of names, 287 254b13.3 Refined Sugar Reexport Program (U.S.), 156 pesticide residues. See food safety standards; reform, 8­9, 115­137 horticultural trade agricultural productivity and, 128­131 Philippines aggregate welfare, 130, 130f 7.4 food safety complaints, 103 high-productivity assumption, 130­131, fruits and vegetables exports, 244 131t7.12­7.13 milk and dairy products, 168­169 output impact, 129­130, 129t7.11 rice trade impact, 128­129, 129t7.10 importers, effect of liberalization on, 188­189 of agricultural protection, implications for, 50­52, minimum import commitments, 38 51t3.10­3.11, 52­53, 118­120 sugar reform, effect of, 151 global merchandise trade reform, 118­120, phytosanitary measures. See Sanitary and Phytosanitary 119t7.3 Agreement caveats about, 119­120 Polish milk and dairy products, 167, 168 tariffs' role, 120, 121t7.4 politics and decoupling, 77­78 impacts of, 118­127 polyester, 259, 260f14.1, 271 LINKAGE model and simulations, 115­118 Portugal See also LINKAGE model agricultural protection levels, 42 milk and dairy products, 169­174 fishing access, 283 multicommodity approach to, 3­4 poverty problems in pursuing, 2­3 agriculture and, 1, 3 real income gains, 120­123, 121t7.5, 122f 7.1 changes in agricultural trade flows and, 17­20 structural implications, 123­127, 123t7.6 trade preferences and, 57 capital market change, 125­126, 127t7.8 preferences. See trade preferences change in rural value, 124­125, 124f 7.2, 125f 7.3 pricing land contribution to agricultural income, 126, See also specific commodities 128t7.9 decline, 23, 23t2.7 wage changes for unskilled labor, 125, 126t7.7 Procampo program (Mexico), 83­84 sugar, 150­152 processed food transitional impact, 132, 132f 7.5 fruits and vegetables, 239­240, 240f13.4 supply response in low-income countries, 132­134, preferential access, 255 133t7.14 tariffs, 49, 248, 248t13.9 trade elasticities, 134­135, 135f 7.6­7.7 rejection for food safety reasons, 101 of trade preferences, to benefit developing countries, trends in baseline simulation to assess impact of 56 reform, 117­118, 117t7.2 transitional impact, 131­132 324 Index Revised Sugar Price Adjustment Law of 2000 (Japan), 155 salmon farming, 291­292 rice, 10, 177­193 Sanitary and Phytosanitary Agreement, 93­96, 293 See also specific countries and regions increased complexity of the standards environment, concentration of production and consumption in 95­96 Asia, 177 private and public oversight and monitoring, 94 definitions of rice trade flows, 178b10.1 separating legitimate from illegitimate differentiation, development of new types of, 178 93­94 domestic rice trade and policy reforms, 190­191 sardines, 280 policy distortions, 38, 185­186, 185t10.4, 191 Scandinavian countries pricing volatility, factors contributing to, 177­178, 191 See also specific countries self-sufficiency of countries in, 177, 191 coffee consumption, 299 as share of calories by region and income level, 177, Scottish salmon farming, 291­292 178t10.1 seafood, 12­13, 275­295 trade and policies in major producing and consuming aquaculture. See aquaculture nations, 179­184, 179f10.1, 180t10.2 bans for food safety reasons, 105­107, 286 net rice trade, 179, 181t10.3 border protection and, 3 trade liberalization and, 3, 4, 188­192 captured, 276 Africa, impact on, 190 catch control, 290­291 Asian exporters, impact on, 189 common-pool property, 280­281 Asian importers, impact on, 188­189 consumption, 275 developing countries, impact on, 190 eco-labeling, 285 flow and price impact, 186, 187t10.5, 188 efficient management, 290 Latin America, impact on, 190 environmental concerns, 282b15.1 North America, impact on, 190 fish catches in leading countries, 276, 276f15.1 other exporters, impact on, 189 fisheries management policies, 280­282 other importers, impact on, 190 fishing access, 283 price impact, 188 fishing subsidies, 280, 282­283, 288, 290, 293 type of rice, impact by, 188, 192 fish meal, 278­279, 291­292, 294 welfare impact, 188­190 food fish RICEFLOW as model to study rice trade liberalization, exports by top countries, 279, 279f15.4 186, 187t10.5, 190, 191 exports by value of major commodity group, 280, robusta coffee, 297 280f15.5 roe, 280 foreign fishing access agreements, 283, 284, rules of origin and trade preferences, 57, 69­70 284b15.2, 293 rural income global aquaculture production by value, 278, 278f15.3 changes in agricultural trade flows and, 17­21, growth in trade of, 28 18t2.1­2.2 hazard-analysis, critical-control-point (HACCP) income distribution, 20­21, 21t2.4 principles, 285­286 structure of income sources, 20, 20t2.3 institutional influences on international trade in trade liberalization, impact of, 125 fishery products, 280­287 rural value, change in from reform, 124­125, 124f 7.2, international markets, 279­280 125f 7.3 labeling, 285 Russia limited access, 281 See also Former Soviet Union (FSU) names, regulation of, 287 average tariffs, 47 open access, 281, 290­291 coffee, 307 overexploitation and, 279, 281 cotton importing, 261 overexploitation of fish stocks, 275­276 global milk production trends, 163 prices, 278­279 seafood subsidies, 290 production, 275, 276­280 sugar imports, 145­146, 149 industrial products, 275, 278­279 wheat shrimp. See shrimp exports, 209­211 sustainability, 276, 294 imports, 206 tariffs, 280, 283­285, 285f15.6, 285t15.1, 287­288, 293 Rwandan coffee exports, 297, 301 total allowable catch (TAC), 281, 290 trade barriers saccharin, 148 antidumping and countervailing measures, 286­287 SACU. See Southern Africa Customs Union sanitary and phytosanitary measures. See food salmonella. See food safety standards safety standards Index 325 technical barriers, 285, 293­294 fruits and vegetables trade liberalization, impact of, 4, 280, 287­293, 293 dried fruit exports, 108 in APEC countries, 288, 289t15.2­15.3 preferential agreement with EU, 252 aquaculture, 291­293, 292t15.5, 292t15.6 tariff rate quotas, 251 fisheries management policies, 280, 281­282 groundnuts management regime in capture fisheries, influence global trade in, 108, 218 of, 290­291, 291t15.4 policies, 227 removing subsidies in capture fisheries, 288, 290 production, 216 Uganda's fishing industry, 282b15.1 trade liberalization scenarios, 229, 232 200-mile exclusive economic zones (EEZ), 279, rice trade liberalization, effect of, 190 280­281, 283 sugar industries employment, 145 self-certification of coffee, 306­307 South America. See Latin America self-sufficiency, 30 South Asia EU, 34 See also specific countries rice, 177, 191 milk and dairy products, 168­169 wheat, 195 production trends, 163 Senegal trade and domestic policy, 168­169 groundnuts trade trends, 164 global trade in, 108, 217, 218­220 supply response and reform, 133, 133t7.14 policies, 226­227 Southeast Asia prices, 220­221 fishmeal production and use, 278 production, 215 seafood consumption, 275 trade liberalization scenarios, 229, 232, 234 shrimp production, 278 seafood Southern Africa Customs Union (SACU) exports, 275, 280 trade liberalization, impact of, 125 fishing access, 283 supply response, 133, 134 set-asides, 80, 82­83 Southern Shrimp Alliance (SSA), 287 shrimp Spaak Report of 1956 (EU), 81 antidumping and countervailing measures, Spain 286, 287 See also European Union (EU) Bangladesh seafood food safety in, 97 bans for food safety reasons, 286 sustainable aquaculture in, 277 fishing access, 283 environmental effects, 291 imports, 280 exports, 280 Spriggs, John, and Dale Sigurdson, 76­77 food safety, 97 state trading of wheat, 202 as important trade commodity, 279 Sub-Saharan Africa Madagascar, 277­278 See also African Growth and Opportunity Act Nicaragua, food safety in, 97 (AGOA); specific countries production, 276­2778 agricultural surplus trends, 118 Thailand coffee production, 276 domestic policies, 305 sustainable aquaculture in, 277 yields, 299 trade liberalization, impact of, 277 food safety issues, 101 value of trade in, 276­277 groundnuts world aquaculture production, by volume, global trade in, 217 277f15.2 policies, 233 Sierra Leone seafood exports, 275, 280 prices, 220 Singapore production, 215 liberal trade regime of, 28 trade liberalization, 233 milk and dairy products, 168­169 market share of, 31 rice trade liberalization, effect of, 190 milk and dairy products Slovak Republic milk and dairy products, 168 production trends, 163 Slovenia milk and dairy products, 167, 168 trade trends, 164 soft drink consumption, 307, 308 reform, effect on, 120 South Africa supply response, 133, 133t7.14, 134 average tariffs, 47 wage changes for unskilled labor, 125 food safety complaints, 103 sucrose, 142 326 Index sugar, 9, 141­159 Tanzania See also specific countries and regions coffee, domestic policies, 305 background, 141­147 cotton, reform initiatives, 270 bagasse as production by-product, 141­142 seafood bans for food safety reasons, 286 Brazil policies, 150 tariffication, 38 China policies, 150 tariffs. See agricultural protection; decoupling decline in trade, 28 taxation employment in sugar industries, 145, 145t8.2 of African groundnut farmers, 226 ethanol as by-product of. See ethanol export taxes. See export taxes EU policies, 152­154 replacing tariffs, economic effect of, 78 See also European Union (EU) VAT in China, 226 exports and net imports, 141, 142f8.1 tea consumption, 307 global market, 145­147, 146t8.3 Thailand government support to producers, 150, 150t8.4 corn and aflatoxin problems, 107 HFCS. See high-fructose corn syrup food safety complaints, 103 India policies, 150 fruits and vegetables exports, 244 Japanese policies, 154­155, 154f 8.4 as major exporter, 28 See also Japan milk and dairy products, 168­169 Lomé Convention preference sugar, 153 poultry safety issues, 103 Mexico policies, 150 reform, impact of, 28 molasses as production by-product, 142 high-productivity assumption, 130 NAFTA side-letter agreement, 149 rice, 182­183 OECD countries' policies, 150­157 exporters, effect of liberalization on, 189 prices, 146­147, 152, 155 seafood production costs, 143­145, 144t8.1 exports, 279 protection, 146­147, 158 production, 276 history of, 141 tariffs, 283 reform, effect of, 132, 150­152 shellfish, 292­293 transitional impact, 132, 132f 7.5 shrimp Russia policies, 150 production, 276 tariff lines, 45 sustainable aquaculture in, 277 Thailand policies, 149­150 sugar U.S. policies, 155­157, 157f 8.5 exports, 146 See also United States policies, 149­150 Sugar Price Stabilization Law of 1965 reform, effect of, 151 (Japan), 155 tuna exports, 280 surplus. See specific countries Third Law of the Sea Convention (UNCLOS-III, sustainability 1973­1982), 280­281 aquaculture, 294 Togo and cotton reform initiatives, 270 fish products, international trade in, 276 tomatoes and EU system of entry prices, 62b4.1 Swerling, Boris, 76 trade elasticities and impact of reform, 134­135, Switzerland 135f 7.6­7.7 agricultural protection levels, 43 trade liberalization, 3­4 horticultural export subsidies, 245 See also reform organic coffee, 307 cotton, effect of, 268­270, 269t14.4, 271 wheat support to domestic producers, 205 developing countries and, 28 synthetic fibers, 259 environmental effects of, 277 Syria groundnuts, 215, 228­233, 229t12.13, 230t12.14 cotton production, 261 milk and dairy products, 3, 4, 161­162, 169­174 fruits and vegetables exports, 244 full liberalization scenario, 170­174, 172t9.3 other scenarios, 174 Taiwan prospects for, 175 as major importer, 28 rice, 3, 4, 188­192 rice seafood, 280, 282b15.1, 287­293 importers, effect of liberalization on, 188 shrimp, 277 policy distortions, 185 sugar, 150­152 trade liberalization, impact of, 125 wheat, 212 Index 327 trade preferences, 6, 55­73 Ukraine benefits of, 56­57 See also Former Soviet Union (FSU) direct transport requirement, 70 wheat exports, 211 erosion by multilateral tariff reductions, 70­71 UNCLOS (Third Law of the Sea Convention), 280­281 EU preferences under GSP and Cotonou Agreement, United Kingdom 60­63, 61t4.4, 63t4.5­4.6 British Retail Consortium Technical Food entry price of tomatoes, 62b4.1 Standard, 95 proportion of trade covered by, 65, 66t4.11 coffee consumption, 299 export diversification and, 68­69 fruits and vegetables standards, 110­111 factors determining impact of, 58 seafood imports, 280 failure to reach potential of, 57 United Nations Conference on Trade and Development Generalized System of Preferences. See Generalized (UNCTAD) System of Preferences (GSP) food safety studies, 105 Japan's preferences under GSP, 63­64, 64t4.7­4.9 Generalized System of Preferences. See Generalized proportion of trade covered by, 65­67, 66t4.12 System of Preferences (GSP) out-of-quota quantities and, 58 United Nations Environment Programme in principle and practice, 56­58 (UNEP), 105 problems with, 55­56, 71­72 United States reform needed to benefit developing countries, 56 See also Quad countries rules of origin and, 57, 69­70 coffee scope of, 58 consumption, 297, 299, 307 undesirable effects of, 57­58 ICAs and, 303b16.1 U.S. preferences under GSP and AGOA, 59­60, organic, 306, 307 59t4.1­4.2 soft drink competition, 307 proportion of trade covered by, 65, 65t4.10 cotton value of, 67­68 decoupled payments, 264­265, 269 EU, 68, 68f4.2 exports, 260 Japan, 68, 69f4.3 export subsidies, 50, 265, 268 U.S., 67­68, 67f4.1 genetically modified, 263 trade shares, 23­26, 23t2.6, 25t2.9 government assistance, 264, 265t14.3 excluding intra-EU and intra-NAFTA trade, 25, 26t2.10 market distortions, 264­266 trade structure changes, 28­30, 30t2.13 production and consumption, 260 transitional impact of reform, 131­132 reform, effect of, 269 sugar, 132, 132f 7.5 tariffs, 264 tuna, 279, 280, 284 decoupling agricultural support, 79­81, Tunisia and dried fruit exports, 108 80t5.2, 88 Turkey farm household incomes, 19 agricultural protection levels, 42 farm subsidies, 41 cotton food safety standards, 94, 100­103, 102t6.2 government assistance, 264 fruits and vegetables production and consumption, 259, 260 imports, 243, 243f13.6 trade in, 261 production growth, 238 decoupling and reform efforts, 79, 85­86, 85t5.5, 88 tariffs, 246, 248 fruits and vegetables groundnuts exports, 244 confectionery, 234 preferential agreement with EU, 252 exports, 108, 217 groundnut exports, 108, 109 policies, 215, 222­224, 228, 233 rice importers, effect of liberalization on, 190 prices, 220 tariff rates, 47 production, 216­217 wheat, 201 trade liberalization scenarios, 229, 231­234 horticultural export subsidies, 245­246 Uganda milk and dairy products coffee, domestic policies and, 305 production trends, 163 seafood reform, effect of, 170, 171 bans for food safety reasons, 286 trade and domestic policy, 165­166 management styles, 290 orange juice protection, 248, 248b13.2 trade liberalization, 282b15.1 peanut quota buyout, 79, 86, 157 328 Index United States (continued) U.S. Commodity Credit Corporation, 261 rice, 184 U.S. Department of Agriculture domestic trade and policy reforms, 191 Economic Research Service of USDA on Japanese exporters, effect of liberalization on, 189, 190 sugar policies, 155 seafood food safety issues and exports, 100, 105 antidumping and countervailing measures, wheat price study, 212 286­287, 293 U.S. Food and Drug Administration. See Food and country-of-origin labels, 285 Drug Administration (FDA, U.S.) exports, 279 U.S. General Accounting Office study on sugar imports, 279­280 policies, 151 tariffs, 283­284 U.S. International Trace Commission and antidumping sugar policies, 145 measures, 286­287 ethanol production and consumption, 143 Uzbekistan high-fructose corn syrup (HFCS), 142­143, See also Former Soviet Union (FSU) 143f8.2, 146 cotton prices and loan rates, 156­157 export, 260 protection, 141, 142 government assistance, 267­268 reform, effect of, 151, 152, 155­157, 158 production, 260 tariffs reform, effect of, 268 average tariffs, 39, 47 tariffs, 264 transparency, 45 trade liberalization, impact of, 124, 127, 233 valorization, 303b16.1 trade preferences under GSP and AGOA, 58, 59­60, value-added tax (VAT) in China, 226 59t4.1­4.2 Van Donkersgoed, Elbert, 76 proportion of trade covered by, 65, 65t4.10 vegetables. See fruits and vegetables value of, 67­68, 67f4.1 Vietnam wheat coffee domestic support of, 195 output, 297, 299 embargoes, 203­204, 213 prices, 300 export subsidies, 200, 203 production, 299­300, 307, 308 food aid and surplus production, 203, 203f11.5 world issues and, 303b16.1 as major exporter, 205 income sources, 20 pricing, 199, 199f11.4 poultry safety issues, 103 production and yields, 196­197, 197f11.1 rice, 182 support to domestic producers, 204 exporters, effect of liberalization on, 189 University of Wisconsin-Madison World Dairy seafood Model, 170 antidumping and countervailing measures, 287 unskilled labor and wage changes, 125, 126t7.7 catfish, 287 urbanization, effect on fruits and vegetables trade liberalization, impact of, 125 trade, 244 wage changes for unskilled labor, 125 Uruguay Round Agreement on Agriculture (URAA), 38­39, 200 wage changes for unskilled labor, 125, 126t7.7 Amber Box, 38­39, 166, 190 wheat, 10­11, 195­214 Blue Box, 39, 82, 170, 183 See also specific countries and regions Green Box, 39, 80, 166, 169 characteristics and trends, 196­200 groundnut tariffs, 223­224 domestic support, 204­205 milk and dairy products, 163, 164, 165, 166t9.1, OECD countries, 204­205, 204t11.5, 205t11.7 169­170 export OECD countries' level of support and, 42, 44 emerging exporters, 209­212, 210f11.8­11.9 Peace Clause expiration, 271 major exporters, 205­206, 205t11.6, 206t11.8 protection reduction efforts of, 22, 46, 58 promotion, 203 rice, 180, 183, 184, 192 feed use of, 198 seafood, 280 food consumption levels, 198, 198f11.2, 200 tariffs, 283 food security and, 203­204, 213 sugar, 153, 156 importers, 206­209, 207f11.6­11.7, 213 wheat, 200, 203 market access, 200­203 Index 329 policy environment, 200­205 Committee on Trade and the Environment (CTE) pricing, 199f11.3­11.4 fishing subsidies, 282, 288 production and yields, 196­197, 196f11.1, 196t11.1, cotton subsidies, disputes and complaints, 269, 271 199­200 Doha Round, prospects for, 192 self-sufficiency policies, effect of, 195 EU sugar regime, disputes and complaints, 154 stocks of, 198­199, 199f11.3 fisheries management, disputes and complaints, 281 surplus disposal programs, 195, 203, 203f11.5 food safety standards, disputes and complaints, tariff rates and quotas, 200­203, 200t11.3, 103­105, 104t6.3 202t11.4 High Level Symposium on Trade and Environment, trade liberalization, effect of, 212 282 trade policies, 197­198, 197t11.2, 212­213 orange juice, disputes and complaints, 249b13.2 World Bank food safety studies, 105 sectoral or product approach, shortcomings of World Trade Organization (WTO) negotiations, 2 agreements Thailand and sugar tariff rate quota, 149 groundnuts, 215 sanitary and phytosanitary measures. See Sanitary Zimbabwe and Phytosanitary Agreement cotton on Subsidies and Countervailing Measures, 293 reform initiatives, 270 on Technical Barriers to Trade, 293 tariffs, 264 China entry and sugar tariff rate quota, 148 foot and mouth disease controls, 103 T R A D E A N D D E V E L O P M E N T S E R I E S G lobal Agricultural Trade and Developing Countries explores the outstanding issues in global agricultural trade policy and evolving world production and trade patterns. This book presents research findings based on a series of commodity studies of significant economic importance to developing countries. Setting the stage with background chapters and investigations of cross-cutting issues, the authors describe trade and domestic policy regimes affecting agricultural and food markets and analyze product standards and compliance costs and their effects on agricultural and food trade. They then examine the impact and effectiveness of preferences and review the evidence on attempts to decouple agricultural support from agricultural output. Finally, they assess the potential gains from global liberalization in agricultural and food markets, and their sensitivity to various assumptions. Within this broad context of global agricultural policies and reforms, the authors then present detailed studies of commodity markets that feature distorted policy regimes among industrial and developing countries or that are important contributors to exports of developing countries. The commodities analyzed are sugar, dairy, rice, wheat, groundnuts, fruits and vegetables, cotton, seafood, and coffee. These commodity studies analyze current policy regimes in key producing and consuming countries, document the magnitude of these distortions, and estimate the distributional impacts--winners and losers--of trade and domestic policy reforms as well as their impact on trade flows and production location. GlobalAgricultural Trade and Developing Countries will aid policymakers and researchers in approaching global negotiations and in evaluating domestic policies on agriculture. This book complements the findings of Agriculture and the WTO: Creating a Trading System for Development. THE WORLD BANK TMxHSKIMBy358634zv":&:>:<:$ 0-8213-5863-4