51279 DISTORTIONS TO AGRICULTURAL INCENTIVES A GLOBAL PERSPECTIVE, 1955­2007 Edited by Kym Anderson DISTORTIONS TO AGRICULTURAL INCENTIVES DISTORTIONS TO AGRICULTURAL INCENTIVES A Global Perspective, 1955­2007 Kym Anderson Editor A COPUBLICATION OF PALGRAVE MACMILLAN AND THE WORLD BANK © 2009 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 12 11 10 09 A copublication of The World Bank and Palgrave Macmillan. PALGRAVE MACMILLAN Palgrave Macmillan in the United Kingdom is an imprint of Macmillan Publishers Limited, registered in England, company number 785998, of Houndmills, Basingstoke, Hampshire, RG21 6XS. Palgrave Macmillan in the United States is a division of St. Martin's Press LLC, 175 Fifth Avenue, New York, NY 10010. Palgrave Macmillan is the global academic imprint of the above companies and has companies and representa- tives throughout the world. Palgrave® and Macmillan® are registered trademarks in the United States, the United Kingdom, Europe and other countries. This volume is a product of the staff of the International Bank for Reconstruction and Development /The World Bank. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the 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 judgement on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without per- mission may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. ISBN: 978-0-8213-7665-2 (softcover) ISBN: 978-0-8213-7973-8 (hardcover) DOI: 10.1596/978-0-8213-7665-2 (softcover) DOI: 10.1596/978-0-8213-7973-8 (hardcover) eISBN: 978-0-8213-7666-9 Library of Congress Cataloging-in-Publication Data Distortions to agricultural incentives: a global perspective, 1955­2007 / edited by Kym Anderson. p. cm. Includes bibliographical references and index. ISBN 978-0-8213-7665-2 (pbk.)--ISBN 978-0-8213-7973-8 (hardback)--ISBN 978-0-8213-7666-9 (electronic) 1. Agricultural subsidies. 2. Agriculture and state. 3. International trade. I. Anderson, Kym. HD1415.D57 2009 338.1'8--dc22 2009011004 Cover design: Tomoko Hirata/World Bank. Cover photo: © Ray Witlin/World Bank Photo Library. Printed in the United States. Dedication To the memory of T. W. Schultz (1902­98), D. Gale Johnson (1916­2003), and Bruce L. Gardner (1942­2008), whose fine minds were sharpened at the University of Chicago and who contributed perhaps more than any other economists of the 20th century to our understanding of the need to reduce distortions to agricultural incentives around the world. OTHER TITLES IN THE SERIES Distortions to Agricultural Incentives in Africa, edited by Kym Anderson and William A. Masters, 2009. Distortions to Agricultural Incentives in Asia, edited by Kym Anderson and Will Martin, 2009. Distortions to Agricultural Incentives in Europe's Transition Economies, edited by Kym Anderson and Johan Swinnen, 2008. Distortions to Agricultural Incentives in Latin America, edited by Kym Anderson and Alberto Valdés, 2008. CONTENTS Foreword xxi Acknowledgments xxv Contributors xxix Abbreviations xxxiii Map: The 75 Focus Countries xxxvi PART I INTRODUCTION 1 1 Five Decades of Distortions to Agricultural Incentives 3 Kym Anderson PART II EVOLUTION OF DISTORTIONS IN ADVANCED ECONOMIES 65 2 Japan, Republic of Korea, and Taiwan, China 67 Masayoshi Honma and Yujiro Hayami 3 Western Europe 115 Tim Josling 4 United States and Canada 177 Bruce L. Gardner 5 Australia and New Zealand 221 Kym Anderson, Ralph Lattimore, Peter J. Lloyd, and Donald MacLaren PART III EVOLUTION OF DISTORTIONS IN EMERGING ECONOMIES 257 6 Eastern Europe and Central Asia 259 Kym Anderson and Johan Swinnen ix x Contents 7 Latin America and the Caribbean 289 Kym Anderson and Alberto Valdés 8 Sub-Saharan and North Africa 323 Kym Anderson and William A. Masters 9 China and Southeast Asia 359 Kym Anderson and Will Martin 10 India and Other South Asian Countries 389 Ashok Gulati and Garry Pursell PART IV GLOBAL MARKET AND WELFARE EFFECTS OF DISTORTIONS 417 11 Welfare-Based and Trade-Based Indicators of National Agricultural Distortions 419 Peter J. Lloyd, Johanna L. Croser, and Kym Anderson 12 Global Distortions to Key Commodity Markets 459 Kym Anderson, Johanna L. Croser, Signe Nelgen, and Ernesto Valenzuela 13 General Equilibrium Effects of Price Distortions on Global Markets, Farm Incomes, and Welfare 505 Ernesto Valenzuela, Dominique van der Mensbrugghe, and Kym Anderson Appendix A: Methodology for Measuring Distortions to Agricultural Incentives 565 Kym Anderson, Marianne Kurzweil, Will Martin, Damiano Sandri, and Ernesto Valenzuela Appendix B: Global Distortions Database, 1955­2007 595 Kym Anderson and Ernesto Valenzuela Index 619 Contents xi Figures 1.1 Gross Subsidy Equivalents of Assistance to Farmers, over Time and by Region, 1955­2007 17 1.2 NRAs to Agriculture, by Regions, 1980­84 and 2000­04 19 1.3 NRAs to Exportable, Import-Competing, and All Covered Agricultural Products, High-Income and Developing Countries, 1955­2004 22 1.4 NRAs, Key Covered Products, High-Income and Developing Countries, 1980­84 and 2000­04 27 1.5 NRAs to Agricultural and Nonagricultural Tradable Products and RRA, All Focus Countries, 1955­2004 31 1.6 RRAs to Tradables, Asia, Africa, and Latin America, 1965­2004 32 1.7 RRAs to Agriculture, High-Income Countries, 1955­2007 36 1.8 Cross-Country Dispersion of NRAs and RRAs, 2000­04 37 1.9 Relationships between Real GDP Per Capita and RRA, All Focus Countries, 1955­2007 39 1.10 Relationship between RRA and the TBI for Agriculture, Focus Regions, 1980­84 and 2000­04 40 1.11 Welfare Reduction Indexes for Covered Tradable Farm Products, by Region, 1960­2007 47 1.12 Regional Shares of Global and Country Shares of Asian Welfare Reduction from Agricultural Policies, 1981­2004 49 1.13 Trade and Welfare Reduction Indexes for 28 Major Agricultural Products, 2000­04 50 1.14 NRAs for Japan, Republic of Korea, and China, and Date of Accession to GATT or WTO, 1955­2005 56 2.1 RRA to Agricultural Versus Nonagricultural Tradables, Japan, Korea, and Taiwan, China, 1955­2007 86 2.2 NRA to Rice, Japan, Korea, and Taiwan, China, 1955­2007 94 2.3 RRA to Agriculture and Real GDP Per Capita, Japan, Korea, and Taiwan, China, 1955­2004 101 2.4 RRA to Agriculture and Relative GDP Per Agricultural Worker, Japan, Korea, and Taiwan, China, 1955­2004 102 3.1 NRAs to Agriculture, EU-6 and Western European Average, 1956­1964 128 3.2 NRAs to Agriculture, EU-9 and Western European Average, 1965­1974 133 3.3 NRAs to Agriculture, EU-12 and Western European Average, 1975­1984 137 3.4 NRAs to Agriculture, EU-15 and Western European Average, 1985­1994 140 3.5 NRAs to Agriculture, EU and Western European Average, 1995­2007 144 xii Contents 3.6 NRAs to Agriculture with and without Decoupled Payments, Western Europe, 1956­2007 145 3.7 NRAs to Exportable, Import-Competing, and All Agricultural Industries, EEC/EU and Other Western European Countries, 1956­2007 146 4.1 Farm Household Income as a Percent of National Household Income, United States, 1930­2005 178 4.2 Expenditure on Commodity Programs and Payments to Farmers, United States, 1955­2005 187 4.3 CCC Commodity Program Outlays, United States, Fiscal Years 1980­2007 189 4.4 NRAs to Exportable, Import-Competing, and All Covered Agricultural Products, United States and Canada, 1955­2007 193 4.5 NRAs to All Agriculture without and with Decoupled Support, United States and Canada, 1955­2007 203 4.6 NRAs to All Nonagricultural Tradables, All Agricultural Tradable Industries, and RRA, United States and Canada, 1955­2007 205 5.1 Real GDP Per Capita in Australia, New Zealand, and Other High- Income Countries Relative to the United States, 1870­2004 222 5.2 NRAs to Exportable, Import-Competing, and All Covered Products, Australia and New Zealand, 1946­2007 232 5.3 NRAs to Manufacturing, All Nonagricultural Tradables, All Agricultural Tradable Industries, and RRA, Australia and New Zealand, 1946­2007 236 5.4 Real Agricultural Total/Multifactor Productivity Growth, Australia and New Zealand, 1927­2004 249 5.5 Real Farmland Prices, New Zealand, 1978­2004 250 6.1 NRAs to Agriculture, Eastern European Countries, 1992­2007 268 6.2 NRAs to Agriculture, Individual Eastern European Focus Countries, 2000­03 271 6.3 RRAs to Agriculture, Eastern European Focus Countries, 1992­95 and 2000­03 280 7.1 NRAs to Agriculture, Individual Latin American Countries and Unweighted Regional Average, 1980­84 and 2000­04 301 7.2 NRAs, by Product, Latin American Countries, 1980­84 and 2000­04 302 7.3 NRAs to Exportable, Import-Competing, and All Agricultural Products, Latin American Region, 1965­2004 302 7.4 GSEs of Assistance to Farmers, Latin American Countries, 1980­84 and 2000­04 307 7.5 NRAs to Agricultural and Nonagricultural Tradable Products and RRA, Latin American Region, 1965­2004 311 Contents xiii 7.6 Relationship between RRA and the TBI for Agriculture, Latin American Focus Countries, 1980­84 and 2000­04 312 8.1 NRAs to Agriculture, Individual African Focus Countries and Unweighted Regional Average, 1975­79 and 2000­04 335 8.2 NRAs, Key Covered Products, African Focus Countries, 1975­79 and 2000­04 338 8.3 NRAs to Exportable, Import-Competing, and All Agricultural Products, African Region, 1955­2004 340 8.4 GSEs of Assistance to Farmers, African Focus Countries, 1975­79 and 2000­04 343 8.5 NRAs to Agricultural and Nonagricultural Tradable Products and RRA, Africa Region, 1960­2004 348 8.6 Relationship between RRA and the TBI for Agriculture, African Focus Countries, 1975­79 and 2000­04 352 9.1 NRAs to All Agriculture and to Rice, China and Southeast Asian Countries, 1980­84 and 2000­04 365 9.2 NRAs to Exportable, Import-Competing, and All Agricultural Products, China and Southeast Asia, 1970­2004 369 9.3 NRAs to Agricultural and Nonagricultural Tradable Products and RRAs, China and Southeast Asia, 1970­2004 372 9.4 Rice NRA and International Rice Price, Southeast Asia, 1970­2005 378 9.5 NRA for Rice, Malaysia, 1960­2004 379 9.6 Relationship between RRA and the TBI for Agriculture, China and Southeast Asian Countries, 1980­84 and 2000­04 381 9.7 RRAs and Log of Real Per Capita GDP, Select Asian Countries, 1955­2005 383 10.1 Real Effective Exchange Rate Index, India, 1964­2007 397 10.2 Unweighted Average Tariffs on Imports of Agricultural and Nonagricultural Goods, India, 2002­06 398 10.3 NRAs to all Agricultural Tradable Industries, All Nonagricultural Tradables, and RRAs, India, 1965­2004 402 10.4 NRAs to all Agricultural Tradable Industries, All Nonagricultural Tradables, and RRAs, Pakistan, 1973­2005 402 10.5 NRAs to All Agricultural Tradable Industries, All Nonagricultural Tradables, and RRAs, Bangladesh, 1974­2004 403 10.6 NRAs to all Agricultural Tradable Industries, All Nonagricultural Tradables, and RRAs, Sri Lanka, 1955­2004 403 10.7 RRAs, South Asia, 1965­2004 409 11.1 NRAs to Farmers in High-Income and Developing Countries, for All Covered Farm Products, 1960­2007 432 11.2 CTEs Affecting Covered Farm Products in High-Income and Developing Countries, 1960­2007 432 xiv Contents 11.3 NRAs Affecting Covered Farm Products in Africa, Asia, Latin America, and European Transition Economies, 1960­2007 433 11.4 CTEs Affecting Covered Farm Products in Africa, Asia, Latin America, and European Transition Economies, 1960­2007 434 11.5 WRIs for Covered Tradable Farm Products, by Region, 1960­2007 436 11.6 TRIs for Covered Tradable Farm Products, by Region, 1960­2007 439 11.7 WRIs and TRIs for Covered Tradable Farm Products, by Country, 2000­04 441 11.8 TRIs and WRIs for Covered Tradable Farm Products, by Region, 1985­89 and 2000­04 451 11.9 Country Contributions to the Global TRI and WRI, 2000­04 452 11.10 WRI and Real Per Capita GDP, All 75 Countries, 1961­2004 453 11.11 NRA, TRI, and WRI for Covered Tradable Farm Products, World, 1960­2004 453 12.1 GSEs of Assistance to Farmers Globally, by Product, 1980­84 and 2000­04 465 12.2 NRAs, Key Covered Products, High-Income and Developing Countries, 1980­84 and 2000­04 469 12.3 NRAs, Rice, Milk, and Sugar, by Country, 2000­04 471 12.4 NRAs, Beef, Pig Meat, and Poultry, by Country, 2000­04 473 12.5 NRAs, Wheat, Maize, and Soybeans, by Country, 2000­04 475 12.6 NRAs, Cotton, Cocoa, and Coffee, by Country, 2000­04 476 12.7 Rice NRA and International Rice Price, South Asia, 1970­2005 480 12.8 NRAs for Rice and Per capita Income, 1955­2007 481 12.9 TRI and WRI for 12 Key Covered Products, 2000­04 491 12.10 Global TRI and WRI for Covered Tradable Farm Products, by Commodity, 1985­89 and 2000­04 492 12.11 Country Share of the Global Commodity-Specific TRI for Rice, Sugar, Beef, Cotton, and Milk, 2000­04 493 12.12 Country Share of the Global Commodity-Specific WRI for Rice, Sugar, Milk, Beef, and Cotton, 2000­04 494 A.1 A Distorted Domestic Market for Foreign Currency 569 A.2 Distorted Domestic Markets for Farm Products 589 B.1 Number of Countries for Which NRA and CTE Estimates Are Provided for 30 Key Farm Products 596 B.2 Shares of Global Production of 30 Key Farm Products Covered in NRA and CTE Estimates 597 Contents xv Tables 1.1 Key Economic and Trade Indicators of Focus Countries, by Region, 2000­04 15 1.2 Poverty in Africa, Asia, Latin America, and Europe's Transition Economies, 1981­2005 16 1.3 Growth of Real GDP and Exports, Focus Countries, 1980­2004 17 1.4 NRAs to Agriculture, Focus Countries, 1955­2007 20 1.5 NRAs to Agricultural Exportables, Import-Competing Products, and the TBI, Focus Regions, 1955­2007 23 1.6 Dispersion of NRAs across Covered Agricultural Products, Focus Regions, 1965­2007 26 1.7 NRAs to Agricultural Products Relative to Nonagricultural Industries, 1955­2007 29 1.8 NRAs to Agricultural and Nonagricultural Tradables, and the RRA, by Region, 1955­2007 33 1.9 GSEs of Assistance to Farmers, Total and Per Farm Worker, by Region, 1965­2007 41 1.10 CTEs of Policies Assisting Producers of Covered Farm Products, Percent and Per Capita, by Region, 1965­2007 44 1.11 Intensity of Public Agricultural R&D Investment, High-Income and Developing Countries, 1971­2004 57 1.A.1 Export Orientation, Import Dependence, and Self-Sufficiency in Primary Agricultural Production, Focus Countries, 1961­2004 59 2.1 Economic Growth and Structural Transformation in Japan, Korea, and Taiwan, China, 1955­2004 72 2.2 Changes in Agricultural Structure in Japan, Korea, and Taiwan, China, 1955­2004 75 2.3 NRAs to Selected Agricultural Products in Japan, Korea, and Taiwan, China, 1955­2007 87 2.4 NRAs to Agricultural Relative to Nonagricultural Industries in Japan, Korea, and Taiwan, China, 1955­2007 90 2.5 CTEs for Selected Agricultural Products in Japan, Korea, and Taiwan, China, 1955­2007 96 2.6 Changes in Direct Tax Burdens and the Allocations of National Government Subsidies to Agricultural and Nonagricultural Sectors, Japan, 1878­1937 105 2.7 Farm-Nonfarm Income Disparity in Japan's Economic Development, 1885­2000 106 3.1 NRAs to Agricultural Industries Relative to Nonagricultural Industries, EU-15 and EFTA-3, 1956­2007 147 3.2 NRAs to Covered Farm Products, EU, 1956­2007 149 3.3 NRAs to Covered Farm Products, Non-EU Western European Countries, 1956­2007 151 xvi Contents 3.4 NRAs to All Agriculture, Individual Western European Countries, 1956­2007 154 3.5 Gross Subsidy Equivalents of Assistance to Farmers, Total, Per Farm Worker and by Product, Western European Countries, 1956­2007 155 3.6 CTEs of Policies Assisting Farmers, Covered Products, Total and Per Capita and by Product, Western European Countries, 1956­2007 158 4.1 Customs Receipts as a Percentage of Value of Imports, United States, 1821­2000 184 4.2 NRAs to Covered Farm Products, United States and Canada, 1955­2007 194 4.3 NRAs to Agricultural Relative to Nonagricultural Industries, United States and Canada, 1955­2007 201 4.4 GSEs of Assistance to Farmers, by Product, Per Farm Worker and Total, United States and Canada, 1955­2007 206 4.5 CTEs of Policies Assisting Farmers, Covered Products, Total and Per Capita and by Product, United States and Canada, 1960­2005 209 4.6 PAC Disbursements during the Election Cycle, United States, November 2004­October 2006 213 5.1 NRAs to Covered Farm Products, Australia, 1946­2007 229 5.2 NRAs to Covered Farm Products, New Zealand, 1955­2007 231 5.3 NRAs to Agricultural Products Relative to Nonagricultural Industries, Australia and New Zealand, 1946­2007 234 6.1 Key Economic and Trade Indicators, Eastern European and CIS Countries, 2000­04 261 6.2 NRAs to Agriculture, Eastern European and CIS Focus Countries, 1992­2007 269 6.3 NRAs, Key Covered Farm Products, Eastern European and CIS Focus Countries, 1992­2005 272 6.4 TBI, Eastern European and CIS Focus Countries, 1992­2007 273 6.5 Dispersion of NRAs across Covered Agricultural Products, Eastern European and CIS Focus Countries, 1992­2007 274 6.6 Components of NRAs to Agriculture, Eastern Europe and CIS, 1961­2007 275 6.7 GSEs of Assistance to Farmers, Total and Per Farm Worker, Eastern European and CIS Focus Countries, 1992­2007 277 6.8 Percentage CTE of Policies Assisting Producers of Covered Farm Products, Eastern European and CIS Focus Countries, 1992­2007 279 7.1 Key Economic and Trade Indicators, Latin American Countries, 2000­04 291 7.2 NRAs to Agriculture, Latin American Countries, 1965­2004 300 7.3 NRAs, Key Covered Farm Products, Latin American Focus Countries, 1965­2004 303 Contents xvii 7.4 Dispersion of NRAs across Covered Agricultural Products within Latin American Focus Countries, 1965­2004 304 7.5 NRAs to Agricultural Relative to Nonagricultural Industries, Latin American Region, 1965­2004 305 7.6 GSEs of Assistance to Farmers, Total and Per Farm Worker, Latin American Countries, 1965­2004 308 7.7 GSEs of Policies Affecting Farmers in Latin America, by Product and Subsector, 1965­2004 309 7.8 Percentage CTE of Policies Affecting Covered Farm Products, Latin American Countries, 1965­2003 314 7.9 Value of CTE of Policies Affecting Covered Farm Products, Latin American Countries, 1965­2003 316 8.1 Key Economic and Trade Indicators, African Focus Countries, 2000­04 325 8.2 NRAs to Agriculture, African Focus Countries, 1955­2004 334 8.3 Dispersion of NRAs across Covered Agricultural Products, African Focus Countries, 1955­2004 337 8.4 NRAs, Key Covered Farm Products, All African Focus Countries, 1955­2004 339 8.5 NRAs to Agricultural Relative to Nonagricultural Industries, African Region, 1955­2004 341 8.6 GSEs of Assistance to Farmers, Total and Per Farm Worker, African Focus Countries, 1955­2004 345 8.7 Percentage CTE of Policies Assisting Producers of Covered Farm Products, African Focus Countries, 1961­2004 349 9.1 Changes in Poverty in Asia, 1981­2005 360 9.2 NRAs to Agriculture, China and Southeast Asia, 1960­2004 366 9.3 NRAs, by Covered Product, China and Southeast Asia, 1970­2005 367 9.4 Annual GSEs of Assistance to Farmers, Total and Per Farm Worker, Asian Economies, 1955­2004 370 9.5 NRAs to Agricultural and Nonagricultural Tradable Industries and RRAs, China and Southeast Asia, 1960­2004 373 9.6 CTE of Policies Assisting Producers of Covered Farm Products, China and Southeast Asia, 1970­2004 376 10.1 Shares of Agriculture in GDP and Employment, South Asian Countries, 1965­2004 391 10.2 Distribution of Fertilizer and Electricity Subsidies and Subsidy Rates for Key Crops, India, 2004 395 10.3 Trade Status of Farm Commodities, South Asian Countries, 1965­2005 405 10.4 NRAs, TBIs and Dispersion of Covered Farm Products, South Asia, 1965­2004 406 xviii Contents 11.1 NRAs for Africa, Asia, Latin America, European Transition Economies, and High-Income Countries, All Farm Products, 1960­2007 429 11.2 CTEs for Africa, Asia, Latin America, European Transition Economies, and High-Income Countries, All Covered Farm Products, 1960­2007 431 11.3 WRIs for Asia, Africa, Latin America, European Transition Economies, and High-Income Countries, All Covered Tradable Farm Products, 1960­2007 437 11.4 TRIs for Asia, Africa, Latin America, European Transition Economies, and High-Income Countries, All Covered Tradable Farm Products, 1960­2007 440 11.5 PDIs for Asia, Africa, Latin America, European Transition Economies, and High-Income Countries, All Covered Farm Products, 1960­2007 442 11.6 CDIs for Asia, Africa, Latin America, European Transition Economies, and High-Income Countries, All Covered Farm Products, 1960­2007 443 11.7 WRIs, by Country and Region, All Covered Tradable Farm Products, 1960­2007 444 11.8 TRIs, by Country and Region, All Covered Tradable Farm Products, 1960­2007 447 12.1 Coverage of Gross Value of Agricultural Production at Undistorted Prices for 12 Key Covered Products, 2000­04 461 12.2 Share of Global Gross Value of Agricultural Production for Key Covered Products, by Region, 2000­04 463 12.3 Share of Regional Gross Value of Agricultural Production for Major Covered Products, by Region, 2000­04 464 12.4 GSEs of Assistance to Farm Industries, by Focus Country Group, 1965­2007 466 12.5 NRAs, 12 Key Covered Farm Products, All Focus Countries, 1965­2004 470 12.6 CTEs of Policies Assisting Producers of Covered Farm Products, All Focus Countries, 1965­2007 478 12.7 Shares of Production Exported and of Consumption Imported for Major Covered Products, by Region, 2000­03 482 12.8 Share of Global Production of Seven Mostly Nontraded Staple Crops, by Region, 1995­2004 484 12.9 Average of Focus Developing Countries' Self-Sufficiency Ratios for Seven Mostly Nontraded Staple Crops, by Region, 1961­2005 485 12.10 Additional Contribution of Seven Noncovered Staples to Values of Agricultural Production and to Aggregate NRAs in Focus Developing Countries, by Region, 1966­2004 486 Contents xix 12.11 Global TRIs, by Commodity, 1965­2004 489 12.12 Global WRIs, by Commodity, 1965­2004 490 12.A.1 Summary of NRA Estimates by Major Product, Africa, Asia, and Latin America, 2000­04 496 13.1 Structure of Price Distortions in Global Goods Markets, 1980­84 and 2004 509 13.2 Economic Welfare Impact of Going Back to 1980­84 Policies, by Country/Region 516 13.3 Impact of Going Back to 1980­84 Policies on Indexes of Real Export and Import Prices, by Region 519 13.4 Terms-of-Trade Contribution to Real Income Changes from Going Back to 1980­84 Policies, by Region 520 13.5 Impact of Going Back to 1980­84 Policies on Shares of Global Output Exported, and Developing Country Shares of Global Output and Exports, by Product 521 13.6 Impact of Going Back to 1980­84 Policies on Agricultural and Food Output and Trade, by Country/Region 523 13.7 Impact of Going Back to 1980­84 Policies on Self-Sufficiency in Agricultural and Other Products, by Product and Region 526 13.8 Impact of Going Back to 1980­84 Policies on Shares of Production Exported and of Consumption Imported by the World, High-Income and Developing Countries 528 13.9 Impact of Going Back to 1980­84 Policies on Shares of Agricultural and Food Production Exported, by Country/Region 529 13.10 Impact of Going Back to 1980­84 Policies on Real International Product Prices 530 13.11 Impact of Going Back to 1980­84 Policies on Real Factor Prices, by Region 531 13.12 Impact of Going Back to 1980­84 Policies on Sectoral Value Added, Agricultural and All-Sector Policy Changes 532 13.13 Impact on Real Income of Full Liberalization of Global Merchandise Trade, by Country/Region, 2004 536 13.14 Regional and Sectoral Sources of Welfare Gains from Full Liberalization of Global Merchandise Trade, 2004 539 13.15 Impact of Full Global Liberalization on Shares of Global Output Exported, and Developing Country Shares of Global Output and Exports, by Product, 2004 541 13.16 Impacts of Full Global Trade Liberalization on Agricultural and Food Output and Trade, by Country/Region, 2004 542 13.17 Impact of Global Liberalization on Share of Agricultural and Food Production Exported by Country/Region, 2004 545 xx Contents 13.18 Impact of Global Liberalization on Self-Sufficiency in Agricultural and Other Products, by Region, 2004 546 13.19 Share of Production Exported and of Consumption Imported by World, High-Income, and Developing Countries, before and after Full Global Liberalization of All Merchandise Trade, by Product, 2004 547 13.20 Impact of Full Global Liberalization of Agricultural and All Goods Markets on Real International Product Prices, 2004 548 13.21 Impacts of Full Global Merchandise Trade Liberalization on Real Factor Prices, 2004 549 13.22 Effects of Full Global Liberalization of Agricultural and All Merchandise Trade on Sectoral Value Added, by Country and Region, 2004 550 13.A.1 Protection Structure in GTAP Version 7 Prerelease and in the Distortion Rates Drawn from the World Bank Project, 2004 558 B.1 Summary of NRA Coverage Statistics, World Bank Agricultural Distortions Project 597 B.2 Coverage of Gross Value of Global Agricultural Production at Undistorted Prices, for 30 Key Products and Four Product Groups, 2000­04 599 B.3 Project's Coverage of National Agricultural Production in Focus Countries at Undistorted Prices, Regional Averages, 1980­2004 600 B.4 Shares of Global Agricultural Production for 30 Major Covered Products, by Region, 2000­04 601 B.5 Share of Regional Agricultural Production for 30 Major Covered Products, by Region, 2000­04 603 B.6 Share of Global Agricultural and Processed Food Exports for 30 Major Covered Products, by Region, 2000­03 605 B.7 Share of Global Agricultural and Processed Food Imports for 30 Major Products, by Region, 2000­03 607 B.8 Shares of Production Exported and Consumption Imported, by Farm Product and Region, 2000­03 609 B.9 Per Capita Income, Focus Countries, 2005 611 B.10 Variables in the Global Agricultural Distortions Database, 1955­2007 612 B.11 Variables in the Supplementary Global Agricultural Trade and Welfare Reduction Indexes Database, 1955­2007 615 FOREWORD In his seminal 1973 book on World Agriculture in Disarray, Professor D. Gale Johnson despaired at the persistence of high agricultural protection in Organisa- tion for Economic Co-operation and Development (OECD) countries, the anti- agricultural and antitrade policies of developing countries, and the tendency for both sets of countries to insulate their domestic food market from international price fluctuations, thereby exacerbating price volatility for the rest of the world. Since the vast majority of the world's poorest households depend on farming for their livelihoods, this disarray not only was highly inefficient but also contributed to global inequality and poverty. Yet the situation worsened over the next dozen years, with agricultural protection in Europe, North America, and Japan peaking and international food prices plummeting in 1986, thanks in large measure to an agricultural export subsidy war between the United States and the European Community. The World Bank's World Development Report 1986 was devoted to the issue of agricultural protection. It urged reform via the General Agreement on Tariffs and Trade's Uruguay Round, which was launched in September of that year and which had agricultural trade-related policies high on its agenda. Simultaneously, under the direction of its chief economist at the time, Anne Krueger, the Bank under- took a research project aimed at measuring the extent to which 18 developing- country governments were pursuing antiagricultural policies. That Krueger, Schiff, and Valdés project was summarized in an article in the World Bank Economic Review in 1988 and detailed in a series of five books in 1991 and 1992. It revealed that, during 1960­84, most developing countries were reducing farm incomes not only by heavily taxing agricultural exports, but even more so by protecting manufacturers from import competition and overvaluing the national currency. xxi xxii Foreword However, from the 1980s, many low-income and some high-income countries began to reform their agricultural price and trade policies. Sometimes this reform was undertaken unilaterally, but some was also undertaken in response to international pressures such as Uruguay Round stipulations, commitments required for accession to the World Trade Organization, and structural adjust- ment loan conditionality by international financial institutions. Meanwhile, reforms in some middle-income economies (most noticeably the Republic of Korea) had "overshot," going from taxing their farmers to protecting them from import competition, which raised concerns that other emerging economies may follow suit and pursue the same agricultural protection growth path of more- advanced economies. Though the OECD Secretariat began to monitor its members' agricultural policies beginning in the late 1980s, there has been no systematic comparable monitoring of policy developments in developing countries. The World Bank launched a major research project in 2006 aimed at filling this void. The papers emerging from that project (see http://www.worldbank.org/agdistortions) have since been edited into a series of four regional books, which are summarized in this volume along with comparable studies of high-income countries' policies since the mid-1950s. By including 75 countries that together account for more than 90 percent of global gross domestic product (GDP), agricultural output, and population, the study provides a representative study of global developments in policies affecting farmer incentives over the past half century. Moreover, the proj- ect generated a global panel dataset of annual estimates of distortion by product and country. By making this dataset freely available to the public, the expectation is that further economic analysis of these critical questions will be stimulated. The present volume concludes by reporting results from a global economy- wide model aimed at addressing the following questions, among others: How much have reforms since the early 1980s improved net incomes of farmers in developing countries? What more could be achieved by removing the remaining distortions to agricultural incentives? The authors find that the economic welfare cost to the world of global distortions to goods trade fell by 58 percent between the early 1980s and 2007, and the cost to developing countries fell by 46 percent. That is, the world has gone about halfway toward liberalizing goods markets globally during the past quarter century. Developing countries have gained dis- proportionately from those reforms, and their farmers have gained far more than nonfarmers in those countries. Moreover, developing countries would benefit 50 percent more than high-income countries from completely freeing global markets for agricultural and other goods, and again their farmers would be the major beneficiaries. Of the prospective overall gain to developing countries, half Foreword xxiii would be due to agricultural policy reforms; such is the extent of global distor- tions remaining in agriculture relative to other goods markets. In turn, this suggests that developing countries have a huge stake in whether the Doha Develop- ment Agenda, especially its agricultural negotiations, are brought to a successful conclusion. Justin Yifu Lin Senior Vice President and Chief Economist The World Bank ACKNOWLEDGMENTS This book provides an overview of the evolution of distortions to agricultural incen- tives caused by price, trade, and exchange rate policies in a large sample of countries spanning the world. Following the introduction and summary chapter, it includes studies of four sets of high-income economies and five sets of emerging economies that together account for more than 90 percent of agricultural production and 95 per- cent of global GDP. The chapters are followed by two appendixes: one provides the methodology used to measure the nominal and relative rates of assistance to farmers and the taxes and subsidies on food consumption, while the other provides statistical information on the coverage of annual estimates of those rates of assistance. The authors of the five emerging-economy chapters are indebted to the other authors of the country case studies underlying those regional summaries. The country studies are reported in four companion volumes, published by the World Bank in 2008 and early 2009, which cover Africa (coedited by Kym Anderson and William A. Masters), Asia (coedited by Kym Anderson and Will Martin), Latin America (coedited by Kym Anderson and Alberto Valdés), and European transi- tion economies (coedited by Kym Anderson and Johan Swinnen). Staff of the World Bank's regional departments provided generous and insight- ful advice and assistance throughout the project, including participating in Bankwide seminars on the draft studies of each region. The World Bank's country directors also offered advice on the studied countries when clearing the working paper versions of each chapter. The authors of this book's chapters benefited from feedback provided by many participants at workshops and conferences in which draft papers were presented over the past two years. We also appreciate the insightful comments and questions from the book's external reviewers. All the country authors are extremely grateful to Ernesto Valenzuela and the team of very able research assistants he managed, including PhD students Johanna L. xxv xxvi Acknowledgments Croser, Esteban Jara, Marianne Kurzweil, Signe Nelgen, Francesca de Nicola, and Damiano Sandri, all of whom helped compile the global mega-spreadsheet of dis- tortion estimates (accessible at http://www.worldbank.org/agdistortions, along with more than 70 working papers that have detailed appendixes not included in the published volumes). Our thanks extend to Johanna L. Croser and Marie Damania for assisting in the initial copyediting of many of the country chapters; to Janice Tuten, who handled the production and all the final copyediting of the manuscript; and to Stephen McGroarty, who supervised its publication. Both methodologically and in terms of previous estimates of distortions, we are indebted to the economists who have plowed this ground before us. They include the team of authors and editors who contributed to the seminal Krueger, Schiff, and Valdés volumes published in 1991­92; the team at the International Food Policy Research Institute (IFPRI) led by David Orden that generated a recent research report on producer support estimates and consumer support estimates (PSEs and CSEs) that covered four large Asian countries; and especially the team in the Trade and Agriculture Directorate of the Organisation for Economic Co- operation and Development (OECD), who have been generating PSEs and CSEs for high-income countries for more than two decades and for some transition and developing economies during the past four years. We are extremely grateful to OECD staff members for providing access to their files and for many useful com- ments on our work as it progressed. The direct contributions to the project by IFPRI (for the Ethiopian and Indian case studies and the South Asian overview), the OECD and the Food and Agricul- ture Organization of the United Nations (FAO, for the Ghana case study), and the United States Department of Agriculture (for the Russian case study) are also greatly appreciated. The OECD and FAO are jointly seeking funds from the Bill and Melinda Gates Foundation to continue the estimation of agricultural policy indicators for developing countries beyond the time series covered in the present study, beginning with a sample of African countries. Our thanks extend to the project's senior advisory board, whose members have provided sage advice and much encouragement throughout the planning and implementation stages of the project. The Board comprises Yujiro Hayami, Bernard Hoekman, Anne Krueger, John Nash, Johan Swinnen, Stefan Tangermann, Alberto Valdés, Alan Winters and, until his untimely death in March 2008, Bruce Gardner. For financial assistance, grateful thanks go to the Development Research Group of the World Bank and trust funds of the governments of the United Kingdom, The Netherlands. Japan, and Irelend. This combined support made it possible for the study to include countries from all regions of the world except the Acknowledgments xxvii Middle East (which accounts for less than 2 percent of global agricultural produc- tion). We also are extremely grateful to the Rockefeller Foundation for providing the opportunity for the authors of this book to gather to plan the project's final outputs at the Rockefeller Conference Center in beautiful Bellagio, Italy, Novem- ber 13­17, 2006. This book is dedicated to the memory of my mentors, T. W. (Ted) Schultz (1902­98) and D. Gale Johnson (1916­2003), and to Bruce Gardner (1942­2008), all of whose fine minds were sharpened at the University of Chicago's Department of Economics and who contributed directly and indirectly perhaps more than any other economists of the 20th century to our understanding of the need to reduce distortions to agricultural incentives around the world.1 Bruce Gardner, who was professionally active until shortly before his death, was a strong supporter of the present research project, both as a member of its senior advisory board and as author of the North America chapter in this volume. He was a gentleman as well as a scholar, and is sorely missed by friends and colleagues alike. Kym Anderson March 2009 1. T. W. Schultz's famous book, Transforming Traditional Agriculture (Yale University Press, 1964), helped the world understand that farmers, however poor, are efficient producers who respond to incentives and hence to government distortions to prices. The volume arising from a workshop he organized on resources, incentives, and agriculture for the American Academy of Arts and Sciences (Distortions of Agricultural Incentives, Indiana University Press, 1978), further clarified the wastefulness of government intervention in agricultural markets. The trade paper in that volume is by D. G. Johnson, who drew on his seminal book, World Agriculture in Disarray (Macmillan, 1973) which pointed to the inequity as well as inefficiency of farmers being taxed in poor countries and subsidized in rich countries. Both economists' contributions are celebrated in special features of journals: Schultz in the Review of Agricultural Econom- ics 28 (3), fall 2006; Johnson in Economic Development and Cultural Change 52 (3), April 2004. Bruce Gardner's seminal contributions include his paper "Causes of U.S. Farm Commodity Programs," in Journal of Political Economy 95 (2), April 1987, and his book American Agriculture in the Twentieth Century: How It Flourished and What It Cost (Harvard University Press, 2002). CONTRIBUTORS Kym Anderson is the George Gollin Professor of Economics at the University of Adelaide and a research fellow of the Centre for Economic Policy Research, London. During 2004­07, he was on an extended sabbatical as lead economist (trade policy) in the Development Research Group of the World Bank in Washington, DC. Johanna L. Croser has been a short-term consultant with this project and is a PhD and law student at the University of Adelaide, having completed her graduate eco- nomics coursework at the University of British Columbia in Vancouver, Canada. Bruce L. Gardner was a professor and the department chair of agricultural and resource economics at the University of Maryland in College Park, MD, until his untimely death in March 2008. He also served as an assistant secretary in the U.S. Department of Agriculture. Ashok Gulati is the Asian Director for the International Food Policy Research Institute in New Delhi, India. Prior to that, he headed the institute's Markets, Trade, and Institutions Division in Washington, DC. Yujiro Hayami is the chair of the graduate faculty of the Foundation for Advanced Studies on International Development and a visiting professor in the National Graduate Institute of Policy Studies, Tokyo. Masayoshi Honma is a professor of agricultural and resource economics at the University of Tokyo, where he focuses on farm policy issues. He is also a member of the board of trustees of the International Food Policy Research Institute in Washington, DC. xxix xxx Contributors Tim Josling is professor emeritus at the Freeman Spogli Institute for International Studies, and previously was a professor in the Food Research Institute at Stanford University. His research covers a wide range of agricultural trade policy areas, including protectionism. Marianne Kurzweil is a young professional at the African Development Bank in Tunis. During 2006­07, she was an extended-term consultant with this project in the Development Research Group at the World Bank in Washington, DC. Ralph Lattimore is a private consultant, but during much of the period of this project he was an economist with the Trade and Agriculture Directorate of the Organisation for Economic Co-operation and Development in Paris. Peter J. Lloyd is professor emeritus in and former dean of the Department of Economics at the University of Melbourne. Prior to that, he was a professorial fel- low in the Institute of Advanced Studies at the Australian National University, Canberra. Donald MacLaren is an associate professor in the Department of Economics at the University of Melbourne. His theoretical and empirical research focuses on agri- cultural trade policy issues, including the analysis of nontariff barriers to trade. Will Martin is research manager of the Rural Development Unit in the Develop- ment Research Group at the World Bank in Washington, DC. He specializes in trade and agricultural policy issues globally, but especially in Asia, and has written extensively on trade policies affecting developing countries. William A. Masters is a professor and the associate head of the Department of Agricultural Economics at Purdue University. He is currently coeditor of the jour- nal Agricultural Economics. Previously, he was a lecturer at the University of Zimbabwe (1988­90). Signe Nelgen was a short-term consultant with this project before becoming a PhD student at the University of Adelaide. Previously, she was associated with the Hamburg Institute of International Economics. Garry Pursell is a visiting fellow at the Australia South Asia Research Centre at Australian National University, after serving for many years in the South Asia Department of the World Bank in Washington, DC. Damiano Sandri is a PhD candidate in economics at the Johns Hopkins Univer- sity. During 2006­07, he was a short-term consultant with this project in the Development Research Group at the World Bank in Washington, DC. Contributors xxxi Johan Swinnen is a professor in the Department of Economics and director of LICOS Center for Institutions and Economic Performance at the Katholieke Uni- versiteit Leuven in Belgium, and a senior fellow of the Center for European Policy Studies in Brussels. Alberto Valdés is a research associate at Universidad Católica de Chile in Santiago. Previously, he was an adviser in the Agriculture Department of the World Bank and director of trade and food security at the International Food Policy Research Institute, both in Washington, DC. Ernesto Valenzuela is a lecturer and research fellow at the School of Economics and Centre for International Economic Studies at the University of Adelaide. During 2005­07, he was an extended-term consultant at the Development Research Group of the World Bank in Washington, DC. Dominique van der Mensbrugghe is the lead economist in the Development Prospects Group of the Development Economics Vice Presidency of the World Bank in Washington, DC, where he specializes in the global economywide modeling. ABBREVIATIONS AIDA Agricultural Income Disaster Assistance (of Canada) AMTA Agricultural Market Transition Act (of the United States) ASEAN Association of Southeast Asian Nations CAIS Canadian Agricultural Income Stabilization CAP Common Agricultural Policy (of the European Union) CCC Commodity Credit Corporation (of the United States) CDI consumer distortion index CEE Central and Eastern Europe CET common external tariff CFIP Canadian Farm Income Program CGE computable general equilibrium (model) cif cost, insurance, and freight CIS Commonwealth of Independent States (of the former Soviet Union) CMEA Council of Mutual Economic Assistance (of Eastern Europe and Central Asia) CMO common market organization CPI consumer price index CSE consumer support estimate (or earlier, consumer subsidy equivalent) CTE consumer tax equivalent CWB Canadian Wheat Board DDA Doha Development Agenda (of the WTO) EAEC Eurasian Economic Community EC European Community ECU European currency unit xxxiii xxxiv Abbreviations EEC European Economic Community EEP Export Enhancement Program (of the United States) EFTA European Free Trade Association ERA effective rate of assistance EU European Union FAO Food and Agriculture Organization FAS Foreign Agriculture Service (of the United States) fob free on board GATT General Agreement on Tariffs and Trade GDP gross domestic product GNP gross national product GRIP Gross Revenue Insurance Program (of Canada) GSE gross subsidy equivalent GTAP Global Trade Analysis Project IAC Industries Assistance Commission (of Australia) IFPRI International Food Policy Research Institute IMF International Monetary Fund ITC International Trade Commission Mercosur Mercado Común del Sur (Southern Common Market) MFP multifactor productivity NAFTA North American Free Trade Agreements NDP net domestic product NFF National Farmers Federation (of Australia) NISA Net Income Stabilization Account (of Canada) NPS non-product-specific NRA nominal rate of assistance NRP nominal rate of protection NSE net subsidy equivalent NTB nontariff barriers to trade NTM nontariff measure OECD Organisation for Economic Co-operation and Development PAC political action committee (in the United States) PDI producer distortion index PDS public distribution system PPP purchasing power parity PRSP Poverty Reduction Strategy Paper (of the World Bank) PSE producer support estimate (or earlier, producer subsidy equivalent) REER real effective exchange rate RRA relative rate of assistance Abbreviations xxxv SAL Structural adjustment loan (from the World Bank) SMAs statutory marketing authorities (in Australia) STE state trading enterprise TBI trade bias index TEC tax equivalent to consumers TFP total factor productivity TRI trade reduction index TRQ tariff rate quota UNCTAD United Nations Conference on Trade and Development URAA Uruguay Round Agreement on Agriculture USDA U.S. Department of Agriculture WDI World development indicators WRI Welfare reduction index WPI Wholesale price index WTO World Trade Organization Note: All dollar amounts are U.S. dollars (US$) unless otherwise indicated. The 75 Focus Countries (Shown in grey) IBRD 36062 September 2009 This map was produced by the Map Design Unit of the World Bank. The boundaries, colors, denominations, and any other information shown on this map do not imply, on the part of the World Bank Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. Part I Introduction 1 Five Decades of Distortions to Agricultural Incentives Kym Anderson "When you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind; it may be the beginning of knowledge, but you have scarcely, in your thoughts, advanced to the stage of science." -- Sir William Thompson (1889, pp. 73­74) Every decade or two, food becomes newsworthy globally. Mostly it is about a price spike, either upward (hurting consumers, as in 1973 and 2008) or downward (hurting farmers in open economies, as in 1986), and most such price spikes are a consequence of major policy shifts, since local weather-induced supply shocks in a many-country trading world tend to offset each other. In 1986, for example, it was the food export subsidy war between Western Europe and North America that drove real international food prices to their lowest level since 1930. The price hikes of 1973 and 2008, by contrast, were partly a consequence of a unilateral pol- icy decisions by a single large player. In 1973, the Soviet Union departed from its policy of self-reliance and entered the international grain market in a significant way to offset a domestic shortfall (Morgan 1979). In 2008, the United States and the European Union (EU) decided to subsidize biofuel production and mandate targets for its use domestically. On both occasions, other governments imposed export restrictions to help insulate their consumers from the price rise, which 3 4 Distortions to Agricultural Incentives: A Global Perspective pushed international prices even higher and drove more exporting countries to follow suit. Policy thus contributes to market volatility, an undesirable situation because volatility around the long-run trend terms of trade slows economic growth (Williamson 2008). Yet trade policy measures are very blunt instruments for dealing with volatility, especially in the modern era of myriad financial instru- ments for risk management, and their beggar-thy-neighbor feature diminishes the international public good contribution of trade openness. Less newsworthy to the mass media, but probably far more important in its effect on the long-run growth and distribution of global welfare, are gradual policy developments in individual countries and their combined effect on other countries via the trend terms of trade in international markets.1 This study is about one such set of trade-related policy developments that, over the past half century, has had dramatic effects--in some ways negative, in others positive--on distortions to agricultural incentives and thus also to consumer prices for food. Given the impor- tance of farm and food prices for the world's poor, those policy-imposed distor- tions affect not only economic growth but also income inequality and poverty. The benefits from specialization in production and exchange have been recog- nized for millennia, yet governments have chosen to restrict international trade, including in agricultural goods. Sometimes, these restrictions are carried out via export taxes, to raise government revenue or to lower the price of food for domes- tic consumers. An early example was the tax on wine exports from Greece in the first century BC and from France and Germany in the dark ages (Johnson 1989). More commonly, trade restrictions take the form of import duties or bans, often as part of a broader foreign policy, as with the United Kingdom's imports of wine from France versus Portugal and Spain in the 1700s and 1800s (Nye 2007). The practice was so striking that wine was used as the example of British imports in the first treatise on the theory of comparative advantage (Ricardo 1817). For advanced economies, the most common reason for farm trade restrictions in the past two centuries has been to protect domestic producers from import competition as they come under competitive pressure to shed labor in the course of economic development. But in the process, those protective measures hurt not only domestic consumers and exporters of other products but also foreign producers and traders of farm products, and they reduce national and global eco- nomic welfare. For decades, agricultural protection and subsidies in high-income (and some middle-income) countries have been depressing international prices of farm products, which lowers the earnings of farmers and associated rural busi- nesses in developing countries. The Haberler (1958) report to the General Agree- ment on Tariffs and Trade (GATT) contracting parties forewarned that such distortions might worsen, and indeed they did between the 1950s and the early 1980s (Anderson and Hayami 1986), thereby adding to global inequality and Five Decades of Distortions to Agricultural Incentives 5 poverty because three-quarters of the world's poorest people depend directly or indirectly on agriculture for their main income (World Bank 2007).2 In addition to this external policy influence on rural poverty, the governments of many developing countries have directly taxed their farmers over the past half century. A well-known example is the taxing of plantation crop exports in postcolonial Africa (Bates 1981). At the same time, many developing countries have chosen also to pursue an import-substituting industrialization strategy, predominantly by restricting imports of manufactures, and to overvalue their currency. Together, those measures indirectly taxed producers of other tradable products in developing economies, predominantly farmers (Krueger, Schiff, and Valdés 1988, 1991). Thus, the price incentives facing farmers in many developing countries have been depressed by agricultural price and international trade poli- cies in both their own and other countries. This disarray in world agriculture, as Johnson (1991) described it in the title of his seminal book, means there has been overproduction of farm products in high- income countries and underproduction in more-needy developing countries. It also means there has been less international trade in farm products than would be the case under free trade, thereby thinning markets for these weather-dependent products and thus making them more volatile. Using a stochastic model of world food markets, Tyers and Anderson (1992) find that instability of international food prices in the early 1980s was three times greater than it would have been under free trade in those products. During the past quarter century, however, numerous countries have begun to reform their agricultural price and trade policies, a process that has raised the extent to which farm products are traded internationally (see table 1.A.1 in the annex of this chapter), but not nearly as fast as globalization has proceeded in the nonfarm sectors of the world's economies.3 A key purpose of the present study is to examine empirically the extent to which those reforms have reversed the above-mentioned policy developments of the previous three decades. True, empirical indicators of farm sector support (called producer support estimates, or PSEs) have been provided in a consistent way for 20 years by the secretariat of the Organisation for Economic Co-operation and Development (OECD) for its 30 member countries (OECD 2008a). However, there are no comprehensive time series rates of assistance to producers of nonagricultural goods to compare with the PSEs, nor are there PSE figures for advanced economies in earlier decades, which are of more immediate relevance if how the two groups of countries' poli- cies evolved during similar stages of development is to be seen. As for developing countries, almost no comparable time series estimates have been generated since the Krueger, Schiff, and Valdés (1988) study, which covered the 1960­85 period for just 17 developing countries.4 An exception is a new set of estimates of nominal 6 Distortions to Agricultural Incentives: A Global Perspective rates of protection for key farm products in China, India, Indonesia, and Vietnam since 1985 (Orden et al. 2007). The OECD (2006) also has released PSEs for Brazil, China, and South Africa, and for several Eastern European countries. The present study complements and extends those two efforts and the seminal Krueger, Schiff, and Valdés (1988) study. It builds on them by providing similar estimates for other significant (including many low-income) developing economies, by developing and estimating new, more comprehensive policy indicators, and by employing the calculated price wedges in a global economy-wide model to estimate the effects of recent and prospective policy developments. These estimates can be helpful, as evidenced in subsequent chapters in this book, in addressing such questions as: Where is there still a policy bias against agricultural production? To what extent has there been overshooting in the sense that some developing-country food producers are now being protected from import competition along the lines of the examples of earlier-industrializing Europe and Japan? What are the political economy forces behind the more- successful reformers, and how do they compare with those in less-successful countries where major distortions in agricultural incentives remain? Over the past two decades, how important have domestic political forces been in bringing about reform relative to international forces (for example, loan conditionality, rounds of multilateral trade negotiations within the GATT, regional integration agreements, accession to the World Trade Organization [WTO], and the globalization of supermarkets and other firms along the value chain) and compared with forces operating in earlier decades? What explains the pattern of distortions across industries and the choice of support or tax instruments within the agricultural sector of each country? What policy lessons and market implications may be drawn from these differing experiences with a view to ensuring better growth- enhancing and poverty-reducing outcomes--including less overshooting, which results in protectionist regimes--in still-distorted economies during their reforms in the future? The study is timely for at least three reasons. First, the WTO is in the midst of the Doha Round of multilateral trade negotiations, and agricultural policy reform is one of the most contentious issues in those talks. Indeed economy-wide model- ing suggests that as much as two-thirds of the global welfare gains from removing all merchandise trade restrictions and agricultural subsidies would come from reform of agricultural policies, even though agriculture accounts for less than 8 percent of world gross domestic product (GDP) and exports (Anderson, Martin, and van der Mensbrugghe 2006). Second, poorer countries and their development partners are striving to achieve the Millennium Development Goals by 2015, in particular the goals related to alleviation of hunger and poverty. And third, world food prices spiked in 2007 and 2008 to extremely high levels. Governments in Five Decades of Distortions to Agricultural Incentives 7 some developing countries, in their haste to deal with the inevitable protests from consumers, reacted in far from optimal ways. Examining the policy responses to this and other episodes of food prices spikes (most notably in 1973­74) is impor- tant in determining which responses work better than others. The present study includes 75 countries that together account for between 90 and 96 percent of the world's population, farmers, agricultural GDP, and total GDP. The sample countries also account for more than 85 percent of farm produc- tion and employment in Africa, Asia, Latin America, and the transition economies of Europe and Central Asia. Per capita income in the sample countries covers a broad spectrum, from some of the poorest countries (Ethiopia and Zimbabwe) to the richest.5 Nominal rates of assistance and consumer tax equivalents (NRAs and CTEs) are estimated for more than 70 products, with an average of almost a dozen per country. In aggregate, the coverage represents around 70 percent of the gross value of agricultural production in the focus countries, and just under two-thirds of global farm production valued at undistorted prices over the period covered. Not all countries had data for the entire 1955­2007 period; the average number of years covered is 41 per country.6 Of the world's 30 most valuable agricultural products, the NRAs cover 77 percent of global output, ranging from two-thirds for livestock, three-quarters for oilseeds and tropical crops, and five-sixths for grains and tubers. Those products represent an even higher share (85 percent) of global agricultural exports (see appendix B of this volume for details). Having such a comprehensive coverage of countries, products and years offers the prospect of obtaining a reliable picture of both long-term trends in policies, and annual fluctuations around those trends, for individual countries and com- modities as well as for country groups, regions, and the world as a whole. The results can also serve as inputs into explanations of why government policies evolved as they did, and can thereby contribute to policy dialogues because, as Stigler (1975, p. ix) wrote,"Until we understand why our society adopts its policies, we will be poorly equipped to give useful advice on how to change those policies." This chapter begins with a brief summary of the long history of national distortions to agricultural markets. It then outlines the methodology used to gen- erate annual indicators of the extent of government interventions in markets, details of which are provided in Anderson et al. (2008a, 2008b) and appendix A of this volume. A description of the economies being examined and their economic growth and structural changes over recent decades is then briefly presented as a preface to the main section of the chapter, in which the NRA and CTE estimates are summarized across regions and over the decades since the 1950s. These esti- mates are discussed in far more detail in the regional studies that follow, chap- ters 2­10. A summary of an additional set of indicators of agricultural price dis- tortions, presented in chapter 11, is based on the trade restrictiveness index first 8 Distortions to Agricultural Incentives: A Global Perspective developed by Anderson and Neary (2005). In chapter 12, the focus shifts from countries to commodities, and various distortion indicators are used to provide a sense of how distorted each of the key farm commodity markets is globally. Chapter 13 uses the study's NRA and CTE estimates to provide a new set of results from a global economy-wide model. It quantifies the impacts of reforms under- taken since the early 1980s, and of the policies still in place as of 2004, on global markets, net farm incomes, and welfare. That chapter concludes by drawing on the lessons learned to speculate on the prospects for further reducing the disarray in world agricultural markets. National Distortions to Farmer Incentives: The Long History, Briefly While much government intervention in agricultural trade over the centuries has been aimed at stabilizing domestic food prices and supplies, there has been a gen- eral tendency for poor agrarian economies to tax agriculture relative to other sec- tors. As nations industrialize, their policy regimes typically gradually change from negatively to positively assisting farmers relative to other producers, and from subsidizing to taxing food consumers. Consider the United Kingdom, the first country to have an industrial revolu- tion. Prior to that period--from the late 1100s to the 1660s--England used export taxes and licenses to prevent domestic food prices from rising excessively. Then, from 1660­90, a series of laws gradually raised food import duties (making imports prohibitive under most circumstances) and reduced export restrictions on grain. Under the Corn Laws of 1815, these provisions were made even more protective of British farmers. The famous repeal of the Corn Laws in the mid- 1840s is often said to have heralded a period of relatively unrestricted food trade for the United Kingdom, although even then protection was retained for another generation for breweries and distilleries--and hence grain producers--via restric- tions on imports of wine and spirits. According to Nye (2007), it was only after the passage of the 1860 Anglo-French Treaty of Commerce that Britain moved closer to free trade than France, and that other European countries began to open up. But agricultural protection returned in the 1930s, and steadily increased over the next five decades. Indeed, the period of opening up in the 19th century was quite short for some countries in Europe, and agricultural protection levels in those countries throughout the 20th century were somewhat higher on average than in Britain. Kindleberger (1975) describes how the 19th century free trade movements in Europe reflect the national economic, political, and sociological conditions of the time. Agricultural trade reform was less difficult for countries, such as the United Five Decades of Distortions to Agricultural Incentives 9 Kingdom, with overseas territories capable of providing the metropole with a ready supply of farm products. The fall in the price of grain imports from the United States in the 1870s and 1880s provided a challenge for all, however. Denmark coped relatively well, by moving more into livestock production to take advantage of cheaper grain. Italy coped by sending many of its citizens to the New World. Farmers in France and Germany successfully sought protection from imports, however, and so began the post-industrial-revolution growth of agricul- tural protectionism in densely populated countries. Meanwhile, tariffs on West European imports of manufactures were progressively reduced after the GATT came into force in the late 1940s, thereby encouraging agricultural production rel- ative to manufacturing production (Lindert 1991; Anderson 1995). Japan provides an even more striking example of the tendency to switch from taxing to increasingly assisting agriculture relative to other industries. Its industri- alization began later than in Europe, after the opening up of the economy follow- ing the Meiji Restoration in 1868. By 1900, Japan had switched from being a small net exporter of food to becoming increasingly dependent on imports of rice (its main staple food, which was responsible for more than half the value of domestic food production). This was followed by calls from farmers and their supporters for rice import controls. Their calls were matched by equally vigorous calls from manufacturing and commercial groups for unrestricted food trade, since the price of rice at that time was a major determinant of real wages in the nonfarm sector. The heated debates were not unlike those that led to the repeal of the Corn Laws in the United Kingdom six decades earlier. In Japan, however, protectionist forces triumphed, and a tariff was imposed on rice imports from 1904. That tariff then gradually increased over time, raising the domestic price of rice to more than 30 percent above the import price during World War I. Even when there were food riots because of shortages and high rice prices just after the war, the Japanese gov- ernment's response was not to reduce protection but instead to extend it to its colonies and to shift from a national to an imperial rice self-sufficiency policy. That involved accelerated investments in agricultural development in the Korean and Taiwanese colonies behind an ever-higher external tariff wall that by the latter 1930s had driven imperial rice prices to more than 60 percent above those in international markets (Anderson and Tyers 1992). After Japan lost its colonies at the end of World War II, its agricultural protection growth resumed and spread from rice to an ever-wider range of farm products. Other high-income countries, which were settled by Europeans relatively recently and are far less densely populated, have had a strong comparative advan- tage in farm products for most of their history following Caucasian settlement. They have thus felt less need to protect their farmers than Europe or Northeast Asia. Indeed, until the present decade, Australia and New Zealand, like developing 10 Distortions to Agricultural Incentives: A Global Perspective countries, had in place policies that discriminated against their farmers (Anderson, Lloyd, and MacLaren 2007). In the Republic of Korea and in Taiwan, China in the 1950s, as in many newly independent developing countries, an import-substituting industrialization strat- egy was initially adopted, which harmed agriculture. But in those two economies, unlike in most other developing countries, that policy was replaced in the early 1960s with a more neutral trade policy that resulted in very rapid export-oriented industrialization. This development strategy in those densely populated economies imposed competitive pressure on the farm sector, which, just as in Japan in earlier decades, prompted farmers to lobby (successfully, as it happened) for ever-higher levels of protection from import protection (Anderson and Hayami 1986). Many less-advanced and less rapidly growing developing countries not only adopted import-substituting industrialization strategies in the late 1950s and early 1960s (Little, Scitovsky, and Scott 1970; Balassa and associates 1971) but also imposed direct taxes on their exports of farm products. It was common in the 1950s and 1960s--and in some cases all the way to the 1980s--to also use dual or multiple exchange rates so as to indirectly tax both exporters and importers (Bhagwati 1978; Krueger 1978). This added to the antitrade bias of developing countries' trade policies. Certainly within the agricultural sector of each country, import-competing industries tended to enjoy more government support than those that were more competitive internationally (Krueger, Schiff, and Valdés 1988; Herrmann et al. 1992; Thiele 2004). The Krueger, Schiff, and Valdés study also reveals that, at least up to the mid-1980s, direct disincentives for farmers, such as agricultural export taxes, were less important than indirect disincentives, such as import protection for the manufacturing sector or overvalued exchange rates, both of which pulled resources away from agricultural industries producing tradable products. In short, historically, countries have tended to gradually shift from taxing to subsidizing agriculture relative to other sectors in the course of their economic development, although less so, and at a later stage of development, the stronger a country's comparative advantage in agriculture (Anderson and Hayami 1986; Lindert 1991). Hence, at any point in time, farmers in poor countries have tended to face depressed terms of trade relative to product prices in international mar- kets, while the opposite was true for farmers in rich countries (Anderson 1995). The exceptions were rich countries with an extreme comparative advantage in agriculture--Australia and New Zealand. While the economic policy history of developing countries has been docu- mented extensively in previous surveys, less well known is the extent to which many emerging economies have belatedly followed the example of Korea and Taiwan, China in abandoning import substitution and opening their economies. Five Decades of Distortions to Agricultural Incentives 11 Some countries (for example, Chile) started in the 1970s while others (for exam- ple, India) did not do so in a sustained way until the 1990s. Other countries have adopted a very gradual pace of reform, with occasional reversals, while others have moved rapidly to open markets. Still others have adopted the rhetoric of reform but in practice have done little to open their economies. To get a clear sense of the overall impact of these reform attempts, there is no substitute for empirical analysis that quantifies the nature and degree of market intervention by governments over time. Methodology for Measuring Price Distortions7 The main focus of the present study's methodology is on the government- imposed distortions that create a gap between domestic prices and what they would be under free markets. Since it is not possible to understand the character- istics of agricultural development with a sectoral view alone, the study not only estimates the effects of direct agricultural policy measures (including distortions in the foreign exchange market) but also generates estimates of distortions in nonagricultural sectors for comparative evaluation. Specifically, the NRA for each farm product is computed as the percentage by which government policies have raised gross returns to farmers above what they would be without the government's intervention (or lowered them, if the NRA is less than zero). Product-specific input subsidies are included in those NRA esti- mates. A weighted average NRA for all covered products is derived using the value of production at undistorted prices as weights (in contrast to the PSEs and con- sumer support estimates, or CSEs, computed by OECD [2008a], which are expressed as a percentage of the distorted price). To that NRA for covered prod- ucts is added a "guesstimate" of the NRA for noncovered products (on average around 30 pecent of the total) and an estimate of the NRA from non-product- specific (NPS) forms of assistance or taxation. Since the 1980s, some high-income governments have also provided "decoupled" assistance to farmers, though because that support in principle does not distort resource allocation, its NRA has been computed separately and is not included for direct comparison with the NRAs for other sectors or for developing countries.8 Each farm industry is classified as import-competing, a producer of exportables, or a producer of nontradables (with its status sometimes changing over the years) in order to generate the weighted average NRAs for the two different groups of covered tradable farm products for each year. A production-weighted average NRA for nonagricultural tradables is also generated, for comparison with that for agricul- tural tradables, via the calculation of a percentage relative rate of assistance 12 Distortions to Agricultural Incentives: A Global Perspective (RRA), defined as: RRA 100*[(100 NRAagt) (100 NRAnonagt) 1], t t where NRAag and NRAnonag are the percentage NRAs for the tradable parts of the agricultural (including noncovered) and nonagricultural sectors, respectively.9 Since the NRA cannot be less than 100 percent if producers are to earn anything, neither can the RRA (since the weighted average NRAnonagt is non-negative in all our country case studies). If the agricultural and nonagricultural sectors are equally assisted, the RRA is zero. This measure is useful in that if it is below (above) zero, it provides an internationally comparable indication of the extent to which a country's sectoral policy regime has an antiagricultural (proagricultural) bias. This approach is not well suited to analysis of the policies of former socialist economies in Europe and Asia prior to their reform era, as prices played only an accounting function and currency exchange rates were enormously distorted. The price comparison approach, however, provides a valuable a set of indicators for the reform era in these countries, as for other market economies, of distortions to incentives for farm production, consumption, and trade, and of the income trans- fers associated with interventions.10 In addition to the mean NRA, a measure of the dispersion or variability of the NRA estimates across the covered farm products also is generated for each econ- omy. The cost of government policy distortions to incentives in terms of resource misallocation tends to grow as the degree of substitution in production increases. In the case of agriculture, which involves the use of farmland that is sector-specific but transferable among farm activities, the greater the variation of NRAs across industries within the sector, the higher the welfare cost of those market interven- tions. A simple indicator of dispersion is the standard deviation of the covered industries' NRAs. While most of the focus is on agricultural producers, this study also considers the extent to which consumers are taxed or subsidized. To do so, a CTE is calcu- lated by comparing the price that consumers pay for their food and the interna- tional price of each food product at the border. Differences between the NRA and the CTE arise from distortions in the domestic economy that are caused by trans- fer policies and taxes and subsidies that cause the prices paid by consumers (adjusted to the farmgate level) to differ from those received by producers. In the absence of any other information, the CTE for each tradable farm product is assumed to be the same as the NRA from border distortions and the CTE for non- tradable farm products is assumed to be zero. To obtain dollar values of farmer assistance and consumer taxation, the coun- try authors' NRA estimates are multiplied by the gross value of production at undistorted prices to obtain an estimate in U.S. dollars of the direct gross subsidy equivalent (GSE) of assistance to farmers. These GSE values are calculated in con- stant dollars and are also expressed on per-farm-worker basis. Likewise, a value of the consumer transfer is derived from the CTE, by assuming consumption value is Five Decades of Distortions to Agricultural Incentives 13 the gross value of production at undistorted prices divided by the self-sufficiency ratio for each product (production divided by consumption, derived from national volume data or the Food and Agriculture Organization's [FAO's] com- modity balance sheets). These transfer values can be summed across products for a country, and across countries for any or all products, to obtain regional aggre- gate transfer estimates for the studied economies. Needless to say, there are numerous challenges in applying the above method- ology, especially in developing economies with poor-quality data. Ways to deal with the standard challenges are detailed in the second section of appendix A, while country-specific issues are discussed in the relevant subsequent chapters. The next section summarizes estimates that have emerged from aggregating the NRA and related estimates provided by the project's country case studies,11 prefaced by a brief review of the relative size, economic growth, and structural changes that have taken place in key regions of the world over recent decades. After the national country studies were completed and the global database of NRA and CTE estimates was assembled (see Anderson and Valenzuela 2008), it was possible to estimate partial equilibrium country and global commodity indexes of the trade and welfare restrictiveness of agricultural policies. Since those estimates are not part of the country studies reported in Parts II and III of this volume, the methodology for them is not laid out until chapter 11, following which is a summary of the index estimates. Empirical Estimates of National Distortions to Farmer Incentives For the purposes of the present study, the world economy is divided into high- income areas (Western Europe, the United States and Canada, Japan, and Aus- tralia and New Zealand),12 three developing country regions (Africa, Asia, and Latin America), and Europe's transition economies in the 1990s plus Turkey. (Turkey is included in this last group because it is in the same geographic region and, like others in that region, has been seeking EU accession, which necessarily has influenced the evolution of its agricultural price and trade policies.) North America and Europe (including the newly acceded eastern members of the EU) each account for almost one-third of the global economy, and the remaining one-third is shared almost equally by developing countries and the other high-income countries. When the focus turns to just agriculture, however, developing countries are responsible for slightly over half the value added globally, with Asia accounting for two-thirds of that lion's share. The developing countries' majority becomes stronger still in terms of global population and even more so in terms of farmers, almost three-quarters of whom are in Asian develop- ing countries. Hence the vast range of per capita incomes and agricultural land 14 Distortions to Agricultural Incentives: A Global Perspective per capita, and thus agricultural comparative advantages, across the country groups in table 1.1 and the strong concentration of poor people in Asia. The num- ber of poor people in Asia has diminished dramatically over the past quarter cen- tury, though, and even more as a percentage of Asia's population (unlike in Africa), but 60 percent of the world's population living on less than US$1 a day still lived in Asia in 2005 (down from 87 percent in 1981 and 76 percent in 1993-- see table 1.2). The decline in Asian poverty was associated with much faster eco- nomic growth and export-led industrialization in the region than in the rest of the world: since 1980, Asia's per capita GDP has grown at four times, and exports at nearly two times, the global average (table 1.3). The share of Asia's GDP that is exported is now one-third above that for the rest of the world and for Latin America, as summarized from the World Bank (2007) by Sandri, Valenzuela, and Anderson (2007). In 2000­04, just 12 percent of Asia's GDP came from agriculture, on average. That stands in contrast to the situation in Africa, where the share for our focus countries ranges from 20 to 40 percent, and in even starker contrast to Latin America and Europe's transition economies, where it is down to 6 percent (and to just 2 percent on average in high-income countries). The share of employment in agriculture remains very high in Asia, however, at just under 60 percent--the same as in Africa and three times the share in Latin America and Eastern Europe, although more farmers work part-time on their farms in Asia than in other devel- oping countries. By contrast, less than 4 percent of workers in high-income countries are still engaged in agriculture. Hence the much greater impact of own- country and rest-of-world distortions to agricultural incentives to welfare, inequality, and poverty in developing countries than in high-income countries. NRAs and GSEs to agriculture Perhaps the simplest measure for capturing the aggregate extent of distortions to agricultural prices globally is to examine the trend in the GSE of assistance (posi- tive or negative) to farmers. Figure 1.1a shows the GSE for five-year periods since 1960 for the world as a whole, for developing countries, and for high-income countries plus Europe's transition economies. The dark line suggests that, apart from the dip during the period of high world food prices in 1973­74, the net global GSE has been steadily rising over the past half century, especially when the decoupled assistance to farmers in high-income countries is included.13 The decomposition of those global transfers by country group in figure 1.1a reveals two distinct trends, each kinked partway through the period. On one hand, aggregate support for farmers in high-income countries rose steadily throughout the period from the 1950s to the early 1990s, before declining slightly over the 15 subsequent years--somewhat more rapidly at the end of the period, when Table 1.1. Key Economic and Trade Indicators of Focus Countries, by Region, 2000­04 National relative to world (world 100) Share (%) of world Agricultural Agricultural RCA,a trade Agricultural Agricultural GDP land agriculture specialization Population Total GDP GDP workers per capita per capita and food indexb Africa 10 1 6 11 14 148 -- -- Asia 51 10 37 73 20 34 80 0.03 Latin America 8 5 8 3 64 171 -- -- Europe and Central Asia 7 4 6 3 48 178 -- -- Western Europe 6 29 16 1 454 46 106 0.03 United States and Canada 5 33 11 0.3 636 186 119 0.08 Australia and New Zealand 0.4 2 2 0.1 405 2,454 354 0.62 Japan 2 13 5 0.2 610 5 12 0.84 All focus countries 90 96 91 92 -- -- -- -- Other (nonfocus) developing and transition economies 10 4 9 8 -- -- -- -- Source: Sandri, Valenzuela, and Anderson (2007), compiled mainly from World Bank 2007. a. RCA, the revealed comparative advantage index, is the share of agriculture and processed food in national exports as a ratio of that sector's share of global exports. b. Primary agricultural trade specialization index is net exports as a ratio of the sum of exports and imports of agricultural and processed food products (world average 0.0). 15 16 Distortions to Agricultural Incentives: A Global Perspective Table 1.2. Poverty in Africa, Asia, Latin America, and Europe's Transition Economies, 1981­2005 % of people living on less than US$1/day who are rural, 1981 1993 2005 2002 People living on less than $1/day (millions): Sub-Saharan Africa 157 247 299 69 East Asia and Pacific 948 600 180 85 of which China 730 444 106 90 South Asia 387 341 350 75 of which India 296 280 267 74 Latin America and the Caribbean 27 34 28 34 European transition economies 3 10 16 50 World 1,528 1,237 879 74 Asia's share of world 87 76 60 n.a. People living on less than $1/day (% of population): Sub-Saharan Africa 40 44 39 East Asia and Pacific 69 36 10 of which China 74 38 8 South Asia 42 29 24 of which India 42 31 24 Latin America and the Caribbean 7 7 5 European transition economies 1 2 3 World 42 27 16 Source: Chen and Ravallion (2008) and, for rural share, Ravallion, Chen, and Sangraula (2007). Note: n.a. not applicable. world food prices shot up. On the other hand, farmers in developing countries were increasingly taxed by price and trade policies from the early 1960s to the late 1970s and early 1980s, a trend that gradually reversed and, by the mid-1990s, resulted in positive aggregate assistance to farmers. Thus, the GSE contributions of the two groups to the global trend are offsetting in the 1980s but additive from the early 1990s to 2004. Over 2005­07, when food prices in international markets rose steeply, transfers to farmers in high-income countries fell back considerably (simi- lar to what happened in 1973­74). Though there are not sufficient estimates to show the change during those recent years for developing countries, their governments also responded by reducing or suspending import tariffs and temporarily restricting the export of food, so developing countries may have con- tributed to, rather than offset, the downward GSE trend in high-income countries. Five Decades of Distortions to Agricultural Incentives 17 Table 1.3. Growth of Real GDP and Exports, Focus Countries, 1980­2004 (at constant 2000 prices, percent per year, trend-based) Total GDP per Export Agriculture Industry Services GDP capita volume Africa -- -- -- -- -- -- Asia 3.1 8.6 7.5 7.1 5.5 11.2 Latin America -- -- -- 5.4 3.6 7.2 Europe and Central Asia -- -- -- -- -- -- Western Europe 0.8 1.4 2.6 2.3 2.0 5.5 United States and Canada 2.7 2.6 3.4 3.2 2.0 6.7 Australia and New Zealand 2.8 2.5 3.6 3.3 2.0 6.5 Japan 1.7 2.0 3.0 2.5 2.1 4.0 All focus countries World 2.0 2.5 3.2 3.0 1.4 6.1 Source: Sandri, Valenzuela, and Anderson (2007), compiled from World Bank 2007. Note: -- not available. Figure 1.1. Gross Subsidy Equivalents of Assistance to Farmers, over Time and by Region,a 1955­2007 a. Over time 300 constant 2000 US$ billions 200 100 0 100 200 59 64 69 4 9 4 9 4 9 4 7 ­7 ­7 ­8 ­8 ­9 ­9 ­0 ­0 ­ ­ ­ 55 60 70 75 80 85 90 95 00 05 65 19 19 19 19 19 19 19 19 20 20 19 net, global (decoupled payments are included in the higher, gray line) developing countries (no averages for periods 1955­59 and 2005­07) high-income countries and Europe's transition economies 18 Distortions to Agricultural Incentives: A Global Perspective Figure 1.1. (continued ) b. By region 130 90 constant 2000 US$ billions 50 10 0 30 70 110 Africa Australia Latin Europe North Japan Western Asia and America and America Europe New and the Central Zealand Caribbean Asia 1980­84 2000­04 c. Among high-income countries 100 90 80 70 60 percent 50 40 30 20 10 0 56 59 62 65 68 71 74 77 80 83 86 89 92 95 98 01 04 07 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 20 Australia and New Zealand Canada United States Japan Western European non-EU countries EU countries Source: Anderson and Valenzuela (2008), based on estimates reported in the project's national country studies. Five Decades of Distortions to Agricultural Incentives 19 In figure 1.1b, GSE contributions are further subdivided into four high-income regions and four developing-country regions. It is clear that Japan and Western Europe are the biggest contributors among high-income countries and that Asia is the biggest contributor by far to the developing country trends--in both subperi- ods. As will become clear, the latter result is largely driven by China and India, although growth in protection in Korea added to the overall increase in assistance in the region. Figure 1.1c shows a further disaggregation of the contribution of the various high-income countries over the full time series. It reveals that while the EU (the EU-15, ignoring the new East European members), Japan, and the United States were equally large contributors in 2005­07, the original six members of the European Economic Community were responsible for more than half of the aggregate support in high-income countries during the first decade of the Com- mon Agricultural Policy (CAP) (1963­72) and Japan and the United States for less than one-fifth each. Though the GSE estimates account for inflation, by being expressed in constant dollar terms, they do not take into account the fact that the agricultural sector is growing and at different rates across countries. That shortcoming is avoided by using the NRA, which is the extent of policy-induced price support expressed as a percentage of the value of agricultural output at undistorted prices. Figure 1.2 and table 1.4, which show the same eight country groups as figure 1.1b, Figure 1.2. NRAs to Agriculture,a by Regions, 1980­84 and 2000­04a 140 120 100 80 percent 60 40 20 0 20 40 Africa Asia Latin Europe Australia North Western Japan America and and America Europe and the Central New Caribbean Asia Zealand 1980­84 2000­04 Source: Anderson and Valenzuela (2008), based on estimates reported in the project's national country studies. a. Includes noncovered products and NPS (but not decoupled) assistance. 20 Table 1.4. NRAs to Agriculture,a Focus Countries, 1955­2007c (percent) 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Africa 14 8 11 15 13 8 1 9 6 7 -- Asia 27 27 25 25 24 21 9 2 8 12 -- Latin America 11 8 7 21 18 13 11 4 6 5 -- Europe and Central Asiab -- -- -- -- -- -- -- 10 18 18 25 Western Europe 44 57 68 46 56 74 82 64 44 37 18 United States and Canada 13 11 11 7 7 13 19 16 11 17 11 Australia and New Zealand 6 7 10 8 8 11 9 4 3 1 2 Japan 39 46 50 47 67 72 119 116 120 120 81 Developing countries 26 23 22 24 22 18 8 2 6 9 -- High-income countries 22 29 35 25 32 41 53 46 35 32 17 All focus countries (weighted average) 3 5 6 0 2 5 17 18 17 18 -- Source: Anderson and Valenzuela (2008), based on estimates reported in the project's national country studies. Note: -- not available. a. Weighted average for each country, including NPS assistance and authors' guesstimates for noncovered farm products (but not decoupled assistance), with weights based on gross value of agricultural production at undistorted prices. Estimates for China pre-1981 and India pre-1965 are based on the assumption that the NRA to agriculture in those years was the same as the average NRA estimates for those countries for 1981­84 and 1965­69, respectively, and that the gross value of production in those missing years is that which gives the same average share of value of production in total world production in 1981­84 and 1965­69, respectively. Developing-country and world aggregates are computed accordingly. b. Countries in Europe and Central Asia are not included in the high-income or developing-country aggregates. Five Decades of Distortions to Agricultural Incentives 21 reveal much more similarity among the regions' NRAs than the GSEs for the developing-country groups. They also reveal that Japan is now proportionately far more protective of its farmers than Western Europe. The bottom row of table 1.4 provides a similar trend to the GSE trend line: the global average NRA, which was close to zero up to the early 1980s, increased as developing countries began to phase out export taxation, and has since hovered at about 18 percent. The GSE and NRA numbers are based not only on estimates of assistance to covered products but also on NPS assistance and on guesstimates of assistance to the roughly 30 percent of the value of farm products that have not been included in the study's explicit price comparison exercise. Figure 1.3 summa- rizes the most robust NRA estimates for covered farm products, categorized each year as either exportable, import-competing, or nontradable. The weighted average trends across those covered products are very similar to the GSE trends both for high-income countries and, in a slightly more muted fashion, for developing countries. What is more striking about figure 1.3 is the marked difference in the levels of support to import-competing versus exportable covered products. Exportables in high-income countries have received relatively little support other than during the export subsidy war of the mid-1980s, while in developing countries they were increasingly taxed from the late 1950s until the 1980s. Taxation has been gradually phased out over the past two decades, although a little remained in 2004, for exam- ple in Argentina, and considerably more was added by various developing coun- tries in 2008 in response to concerns about the spike in international food prices. Importables, by contrast, have been assisted throughout the past five decades, and the long-run fitted trend line has almost the same slope for both sets of countries (but a lower intercept for developing countries). Two lessons can be drawn from this: first, there is a strong antitrade bias for agricultural goods in high-income and developing countries that has not diminished much over the past half century (in fact, it increased substantially in the 1980s); and second, growth in agricultural import protection appears to have accompanied global economic growth through the 1990s, and has slowed only slightly since then. These lessons hold even when the eight regions examined here are disaggregated, as shown in table 1.5, which also reports estimates of a trade bias index (TBI) for each region. Though the index confirms the broad global conclusion, it indicates a sizable decline in the antitrade bias in developing countries since the 1980s, especially in Asia. That antitrade bias means that the rates of assistance are not uniform across commodities, which indicates that the resources being used within the farm sector are not being put to their best use. The extent of that inefficiency--over and above that caused by too many or too few resources in aggregate in the sector--is crudely indicated by the standard deviation of NRAs among covered products in each 22 Distortions to Agricultural Incentives: A Global Perspective Figure 1.3. NRAs to Exportable, Import-Competing, and All Covered Agricultural Products,a High-Income and Developing Countries, 1955­2004 a. Developing countries 90 70 50 30 percent 10 10 30 50 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 b. High-income countries plus Europe's transition economies 90 70 50 30 percent 10 10 30 50 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 import-competing total exportables Source: Anderson and Valenzuela (2008), based on estimates reported in the project's national country studies. a. Covered products only. The total also includes nontradables. focus country. This dispersion index, summarized for our eight regions in table 1.6, has fluctuated across time and varied between regions, but the global average has remained around 70 percent throughout the period, with no discernable trend. Table 1.5. NRAs to Agricultural Exportables, Import-Competing Products, and the TBI,a Focus Regions, 1955­2007 (percent) 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Africa NRA, agricultural exportables -- 30.1 38.4 42.6 42.6 35.0 36.7 35.8 26.1 24.6 -- NRA, agricultural import- competing products -- 18.6 11.8 1.9 14.5 13.2 58.3 5.2 9.8 1.6 -- TBI -- 0.41 0.45 0.44 0.50 0.43 0.60 0.39 0.33 0.26 -- Latin America NRA, agricultural Exportables -- 20.4 12.8 27.0 25.2 27.1 25.0 10.5 3.5 4.6 -- NRA, agricultural import- competing products -- 26.3 8.7 2.8 1.1 13.6 5.1 19.4 12.5 20.6 -- TBI -- 0.37 0.20 0.25 0.26 0.36 0.29 0.25 0.14 0.21 -- South Asiac NRA agricultural exportables -- 37.5 37.2 30.0 36.1 27.9 20.6 15.8 12.0 6.2 -- NRA agricultural import- competing products -- 39.2 41.2 39.4 45.1 37.9 63.3 25.1 14.5 26.5 -- TBI -- 0.55 0.56 0.50 0.56 0.48 0.51 0.33 0.23 0.26 -- China and Southeast Asia c NRA, agricultural exportables -- 55.5 55.1 51.8 50.1 50.0 41.0 20.8 2.2 0.1 -- NRA, agricultural import- competing products -- 10.3 8.9 9.4 2.6 0.5 15.1 3.3 13.3 12.3 -- TBI -- 0.50 0.51 0.47 0.49 0.50 0.49 0.23 0.14 0.11 -- (Table continues on the following pages.) 23 24 Table 1.5. NRAs to Agricultural Exportables, Import-Competing Products, and the TBI,a Focus Regions, 1955­2007 (continued) (percent) 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Japan, Korea, and Taiwan, China NRA, agricultural exportables 18.1 5.7 4.3 15.4 10.3 25.1 48.9 57.1 57.0 70.3 -- NRA, agricultural import- competing products 35.6 43.3 52.8 54.1 76.6 83.7 124.9 127.4 127.0 134.6 122.6 TBI 0.40 0.26 0.32 0.25 0.38 0.32 0.34 0.31 0.31 0.27 -- European transition economies NRA, agricultural exportables -- -- -- -- -- -- -- 3.2 1.0 1.0 15.2 NRA, agricultural import- competing products -- -- -- -- -- -- -- 32.5 35.4 35.7 32.3 TBI -- -- -- -- -- -- -- 0.27 0.27 0.27 0.13 Western Europe NRA, agricultural exportables 9.3 17.4 31.7 22.5 33.3 31.1 50.1 38.0 15.0 8.1 1.7 NRA, agricultural import- competing products 59.4 77.2 82.9 55.7 61.7 79.5 87.6 67.2 52.8 50.5 28.9 TBI 0.31 0.34 0.28 0.21 0.18 0.27 0.20 0.17 0.25 0.28 0.21 North America NRA, agricultural exportables 2.7 2.8 6.1 5.1 2.9 5.4 10.5 6.0 5.4 7.6 4.1 NRA, agricultural import- competing products 8.6 9.3 8.8 6.7 10.5 19.7 23.6 18.6 11.3 16.8 11.0 TBI 0.05 0.06 0.02 0.01 0.07 0.11 0.10 0.10 0.05 0.08 0.06 Australia and New Zealand NRA, agricultural exportables 3.8 4.7 6.6 5.8 5.5 7.6 6.5 3.6 2.2 0.2 0.2 NRA, agricultural import- competing products 7.9 8.3 9.3 11.7 8.7 8.4 6.5 3.8 2.0 2.0 1.5 TBI 0.04 0.03 0.02 0.05 0.03 0.01 0.00 0.00 0.00 0.02 0.01 Developing countriesc NRA, agricultural exportables -- 46.5 44.6 45.4 43.9 41.4 35.8 18.7 5.5 3.0 -- NRA, agricultural import- competing products -- 12.7 13.5 7.8 12.8 16.5 37.7 22.6 22.0 23.0 -- TBI -- 0.53 0.51 0.49 0.50 0.50 0.53 0.34 0.23 0.21 -- High-income countries NRA, agricultural exportables 4.2 7.4 13.5 10.3 11.3 12.1 22.3 15.9 8.1 6.9 2.9 NRA, agricultural import- competing products 31.2 45.9 50.2 36.5 47.4 58.1 71.4 62.4 53.9 50.7 30.8 TBI 0.21 0.26 0.24 0.19 0.24 0.29 0.29 0.29 0.30 0.29 0.21 World c NRA, agricultural exportables -- 23 20 23 25 24 17 7 1 0 -- NRA, agricultural import- competing products -- 35 37 27 34 38 57 43 38 36 -- TBI -- 0.43 0.42 0.39 0.44 0.45 0.47 0.35 0.28 0.26 -- Source: Anderson and Valenzuela (2008), based on estimates reported in the project's national country studies. Note: -- not available. a. NRAs for noncovered products are included here (unlike in figure 1.3). b. TBI (1 NRAagx 100)/(1 NRAagm/100) 1, where NRAagx and NRAagm are the weighted average percentage NRAs for the exportable and import-competing parts of the agricultural sector, with weights based on production valued at undistorted prices. TBIs shown here are calculated using the regional five-year averages of NRAagx and NRAagm. c. Estimates for China pre-1981 and India pre-1965 are based on the assumption that NRAs to agriculture in those years were the same as the average NRA estimates for those countries for 1981­84 and 1965­69, respectively, and that the gross value of production in those missing years is that which gives the same average share of value of production in total world production in 1981­84 and 1965­69, respectively. The developing-country and world averages are computed accordingly. 25 26 Table 1.6. Dispersion of NRAs across Covered Agricultural Products,a Focus Regions, 1965­2007 (percent) 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Africa 31 30 37 36 36 31 25 25 -- Asia 56 42 49 53 66 56 57 64 -- Latin America 49 44 52 52 44 42 32 40 -- Europe and Central Asia 34 33 41 26 39 56 39 45 44 Western Europe 119 85 112 98 122 86 69 74 64 United States and Canada 29 15 31 62 71 39 31 37 28 Australia and New Zealand 40 45 26 17 20 14 12 7 5 Japan 69 82 156 143 175 162 136 143 116 All focus countries (weighted average) 54 45 55 51 59 53 43 48 -- Product coverageb 68 70 71 73 73 72 71 68 70 Source: Anderson and Valenzuela (2008), based on estimates reported in the project's national country studies. Note: -- not available. a. Dispersion for each region is a simple average of the country-level annual standard deviations around a weighted mean of NRAs per country across covered products each year. b. Share of gross value of total agricultural production at undistorted prices accounted for by covered products. Five Decades of Distortions to Agricultural Incentives 27 The NRAs for different covered products are not random, because their weighted averages across developing or high-income countries cover a wide spectrum. Fig- ure 1.4 shows that rice, sugar, and milk (the rice pudding ingredients) are by far the most assisted farm industries in both sets of countries, with beef and poultry next. For high-income countries, cotton has the next-highest NRA after beef and poultry, yet it has the lowest (most negative) NRA in developing countries. When the country authors' best guesstimates of the NRAs for the various non- covered farm products are aggregated across countries, their weighted average tends to be less than half that for covered products. This finding is not surprising, since many of them are nontraded staples or fruits and vegetables that receive lit- tle or no government attention--part of the reason they were not covered by authors in the first place. Their inclusion in the weighted average across all prod- ucts therefore dampens the positive or negative NRA average for a region's covered products. Though NPS and decoupled assistance to farmers added only a Figure 1.4. NRAs, Key Covered Products, High-Income and Developing Countries, 1980­84 and 2000­04 a. Developing countries b. High-income countries sugar rice 387 milk sugar rice milk poultry beef wheat poultry maize cotton pig meat pig meat coffee soybeans soybeans maize beef wheat coconuts barley cotton rapeseed 100 50 0 50 100 150 50 0 50 150 250 percent percent 1980­84 2000­04 Source: Anderson and Valenzuela (2008), based on estimates reported in the project's national country studies. 28 Distortions to Agricultural Incentives: A Global Perspective little to the NRA for most developing countries, NPS added 2 or 3 percentage points until the early 1980s and has contributed about 5 points since then for high-income countries. If decoupled payments are included, the NRA for high- income farmers rises another 5 percentage points for the 1980s and early 1990s and as many as 10 points by during the present decade (table 1.7). Assistance to nonfarm sectors and RRAs The antiagricultural policy biases of the past were due to more than agricultural policies. Also important in developing countries, according to Krueger, Schiff, and Valdés (1988), has been border protection to the manufacturing sector--the dominant intervention in the tradables part of nonagricultural sectors. Unfortu- nately, it has not been possible with the resources available for this study to quan- tify the distortions to nonfarm tradable sectors as carefully as for agriculture. Authors typically have had to rely on applied trade taxes (for exports as well as imports) rather than being able to undertake price comparisons, and hence they usually do not capture quantitative restrictions on trade, which were important in earlier decades but decreasingly so in recent times. Nor do they capture distor- tions in the services sectors, some of which now produce tradables (or would in the absence of interventions preventing their emergence). As a result, the esti- mated NRAs for nonfarm importables are smaller and decline less rapidly than in fact was the case--and likewise for nonfarm exportables, except their NRAs in some cases would have been negative. Of those two elements of underestimation, the former bias certainly dominates, so the authors' estimate of the overall NRA for nonagricultural tradables should be considered a lower-bound estimate (more so in the past), so that its decline is less rapid than it should be.14 Despite these methodological limitations, the estimated NRAs for nonfarm tradables are very sizeable prior to the 1990s. For developing countries as a whole, the average NRA value has steadily declined throughout the past four to five decades, from around 45 percent in the 1960s to around 30 percent in the 1970s, 16 pecent in the 1980s, and less than 10 percent since the mid-1990s as policy reforms have spread (second last row of table 1.7). The decline in NRA has contributed to a decline in the estimated negative RRA for farmers: the weighted average RRA was worse than 50 percent up to the mid-1970s but improved to an average of 38 percent in the 1980s, 12 percent in the 1990s, and just above zero (1 percent) in 2000­04. The trend in RRAs and their two component NRAs for developing coun- tries is starkly illustrated in figure 1.5, where it is clear that the falling positive NRAs for nonfarm producers have contributed even more to the rise of the RRA than has the gradual disappearance of the negative NRAs for farmers. When trends are decomposed by region, it is clear that Asia has been the major contributor to this Table 1.7. NRAs to Agricultural Products Relative to Nonagricultural Industries, 1955­2007 (percent, weighted averages) a. Developing countriesd 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 1965­69 NRA, covered products 33.4 29.6 28.8 30.2 27.6 23.3 13.2 4.9 3.8 6.7 NRA, noncovered products 9.0 7.9 7.6 9.8 9.8 7.1 0.3 0.1 3.9 6.3 NRA, all agricultural products 27.1 24.0 23.1 24.9 23.1 19.1 9.8 3.5 3.9 6.6 Total agricultural NRA (including NPS)a 25.8 22.7 21.8 23.7 22.0 17.8 8.3 1.8 5.7 8.7 NRA, decoupled assistance 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.2 0.3 Total agricultural NRA (including NPS decoupled assistance) 25.8 22.7 21.8 23.7 22.0 17.8 8.3 1.6 5.9 9.0 NRA, all agricultural tradablesa 27.9 25.4 25.3 28.2 25.7 20.9 10.3 2.2 6.2 8.9 NRA, nonagricultural tradables 56.9 43.1 45.0 30.6 27.3 18.8 14.0 12.7 9.0 5.7 RRAc 54.1 47.7 48.4 44.9 41.6 32.9 21.2 13.1 2.5 3.1 (Table continues on the following page.) 29 30 Table 1.7. NRAs to Agricultural Products Relative to Nonagricultural Industries, 1955­2007 (continued) (percent, weighted averages) b. High-income countries (including Europe and Central Asia) 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 1965­69 1970­74 NRA, covered products 22.0 30.9 37.0 27.2 34.2 38.9 55.6 43.2 32.0 28.7 16.6 NRA, noncovered products 10.0 14.6 14.9 11.5 14.1 13.7 21.1 17.0 17.0 16.7 91.9 NRA, all agricultural products 18.5 26.2 30.7 22.8 28.6 31.8 45.6 35.6 27.5 25.0 12.6 Total agricultural NRA (including NPS)ba 22.3 28.8 32.4 23.7 30.2 36.3 49.6 40.0 31.3 29.4 16.5 NRA, decoupled assistance 0.0 0.0 0.0 0.0 0.6 4.7 5.8 5.5 8.1 9.6 9.9 Total agricultural NRA (including NPS decoupled assistance) 22.3 28.8 32.4 23.7 30.8 41.0 55.4 45.5 39.4 39.0 26.4 NRA, agricultural tradables 23.0 30.2 34.2 25.0 32.2 37.9 52.0 41.7 32.5 30.1 17.0 NRA, nonagricultural tradables 7.5 8.7 9.1 6.3 4.5 3.8 3.7 2.5 2.1 1.8 0.2 RRAc 14.3 19.7 23.0 17.6 26.5 32.8 46.6 38.2 29.8 27.8 17.2 Source: Anderson and Valenzuela (2008), based on estimates reported in the project's national country studies. a. NRAs including NPS assistance, that is, the assistance to all primary factors and intermediate inputs as a percentage of the total primary agricultural production valued at undistorted prices. b. TBI (1 NRAagx 100) (1 NRAagm 100) 1, where NRAagm and NRAagx are the average percentage NRAs for the import-competing and exportable parts of the agricul- tural sector. The regional average TBI is calculated from the regional averages of the NRAs for exportable and import-competing parts of the agricultural sector. c. RRA 100*[(100 NRAagt )/(100 NRAnonagt ) 1], where NRAagt and NRAnonagt are the percentage NRAs for the tradables parts of the agricultural and nonagricultural sectors, respectively. d. Estimates for the NRA and RRA for China pre-1981 and India pre-1965 are based on the assumption that the agricultural NRAs in those years were the same as the average NRA estimates for those countries for 1981­84 and 1965­69, respectively, and that the value of production in those missing years is that which gives the same average share of value of production in total world production in 1981­84 and 1965­69, respectively. Developing-country and world aggregates are computed accordingly. Five Decades of Distortions to Agricultural Incentives 31 Figure 1.5. NRAs to Agricultural and Nonagricultural Tradable Products and RRA,a All Focus Countries, 1955­2004 a. Developing countriesb 100 80 60 40 percent 20 0 20 40 60 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 b. High-income countries (not including those in Europe and Central Asia) 100 80 60 40 percent 20 0 20 40 60 9 4 9 4 9 4 9 4 9 4 ­5 ­6 ­6 ­7 ­7 ­8 ­8 ­9 ­9 ­0 55 60 65 70 75 80 85 90 95 00 19 19 19 19 19 19 19 19 19 20 NRA, nonagricultural NRA, agricultural RRA Source: Anderson and Valenzuela (2008), based on estimates reported in the project's national country studies. a. The RRA is defined as 100*[(100 NRAag t ) (100 NRAnonagt ) 1], where NRAagt and NRAnonagt are the percentage NRAs for the tradables parts of the agricultural and nonagricultural sectors, respectively. b. Estimates for China pre-1981 and India pre-1965 are based on the assumption that the NRA to agriculture in those years was the same as the average NRA estimates for those countries for 1981­84 and 1965­69, respectively, and that the gross value of production in those missing years is that which gives the same average share of value of production in total world production in 1981­84 and 1965­69, respectively. 32 Distortions to Agricultural Incentives: A Global Perspective Figure 1.6. RRAs to Tradables,a Asia, Africa, and Latin America, 1965­2004 20 10 0 10 percent 20 30 40 50 60 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 Africa Latin America and the Caribbean Asia Source: Anderson and Valenzuela (2008), based on estimates reported in the project's national country studies. a. Five-year weighted averages with value of production at undistorted prices as weights. In Asia, estimates for China pre-1981 are based on the assumption that the NRA to agriculture and nonagricultural tradables, and hence the RRA in those earlier years, were the same as the average NRA estimates for China in 1981­89. dramatic reform (figure 1.6 and table 1.8). Within Asia, China and India have con- tributed most to that outcome (Anderson and Martin 2009b). Although average changes in NRAs and RRAs for high-income countries look minor by comparison with those for developing countries, they hide major differences across the high-income countries. Even figure 1.7 doesn't do justice to the striking case of Australia and New Zealand, where the phasing out of high manufacturing tariffs provided nearly as large a boost to farmer incentives as in developing countries as a group. There are two differences between the trend in those two farm-exporting countries and the trend in developing countries, how- ever. One is that much of the cut in Australia's manufacturing protection occurred in the decade before 1955­59, when the RRA was far more negative (not shown in figure 1.7 but see Anderson et al. 2009); the other is that Australian and New Zealand farmers enjoyed considerable direct assistance for much of the time manufacturing was protected, and that farm assistance was cut at the same time as was manufacturing protection, which kept the RRA from rising greatly starting in the 1970s (table 1.8). Despite the very considerable policy reforms of the past quarter century, there is still a long way to go before differences in rates of assistance across countries are Table 1.8. NRAs to Agricultural and Nonagricultural Tradables, and the RRA,a by Region, 1955­2007 (percent) 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Africa NRA, agricultural -- 13.3 19.6 25.0 22.1 13.5 0.3 15.4 8.7 12.0 -- NRA, nonagricultural -- 3.7 2.7 1.5 5.7 1.6 9.2 2.7 2.0 7.3 -- RRA -- 15.2 21.4 26.0 25.9 13.1 8.3 17.1 10.4 18.0 -- Latin America NRA, agricultural -- 11.4 9.3 23.0 19.0 12.9 11.2 4.4 5.5 4.9 -- NRA, nonagricultural -- 26.9 31.3 27.8 23.3 18.5 16.8 7.3 6.6 5.4 -- RRA -- 30.2 30.9 39.8 34.2 26.6 24.0 2.7 1.0 0.5 -- South Asiab NRA, agricultural -- 4.1 4.4 9.7 7.7 1.8 47.1 0.2 2.4 12.7 -- NRA, nonagricultural -- 114.4 117.8 81.7 57.8 54.6 39.9 18.6 15.0 10.1 -- RRA -- 51.5 51.9 39.8 41.6 33.3 5.1 15.5 14.9 3.4 -- China and Southeast Asiab NRA, agricultural -- 43.6 42.6 40.1 35.7 34.5 27.8 12.0 4.9 7.1 -- NRA, nonagricultural -- 36.5 36.5 33.7 30.8 20.6 23.3 19.8 9.6 5.5 -- RRA -- 58.7 58.0 55.2 50.8 43.4 41.6 26.4 4.2 1.5 -- Japan, Korea, and Taiwan, China NRA, agricultural 30.1 39.9 48.8 51.3 75.5 78.8 124.3 129.9 130.5 138.1 126.1 NRA, nonagricultural 8.6 8.3 6.1 4.2 3.5 2.4 2.5 1.4 1.1 0.6 1.0 RRA 19.7 29.1 40.2 44.9 69.6 74.6 118.7 126.7 128.1 136.7 123.7 (Table continues on the following pages.) 33 34 Table 1.8. NRAs to Agricultural and Nonagricultural Tradables, and the RRA,a by Region, 1955­2007 (continued ) (percent) 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 European transition economies NRA, agricultural -- -- -- -- -- -- -- 10.0 18.3 16.1 17.0 NRA, nonagricultural -- -- -- -- -- -- -- 9.8 5.5 4.6 2.7 RRA -- -- -- -- -- -- -- 0.1 12.2 11.0 13.9 Western Europe NRA, agricultural 43.8 57.0 67.5 45.7 56.3 74.4 82.0 63.4 43.6 36.8 18.5 NRA, nonagricultural 8.0 7.2 5.7 3.8 2.5 1.5 1.7 1.3 1.5 1.4 1.2 RRA 33.1 46.5 58.6 40.4 52.6 71.9 79.0 61.3 41.5 34.9 17.1 North America NRA, agricultural 12.5 10.5 10.9 7.5 7.6 13.8 20.2 16.1 11.4 17.3 11.2 NRA, nonagricultural 6.1 7.4 7.4 5.5 4.1 3.8 3.7 3.3 2.1 1.5 1.3 RRA 6.0 2.9 3.3 1.8 3.4 9.7 15.8 12.4 9.1 15.5 9.7 Australia and New Zealand NRA, agricultural 5.5 6.6 8.3 7.9 7.3 10.6 8.7 4.3 2.9 1.0 0.6 NRA, nonagricultural 20.0 21.5 24.0 19.7 14.3 13.5 10.3 6.4 3.4 2.4 2.4 RRA 12.1 12.2 12.6 9.9 6.1 2.6 1.5 2.0 0.5 1.4 1.8 Developing countriesb NRA, agricultural -- 24.0 27.3 31.9 25.5 21.0 15.6 3.9 4.0 7.4 -- NRA, nonagricultural -- 58.3 60.0 45.8 37.3 34.6 27.0 16.7 9.8 6.3 -- RRA -- 52.0 54.5 53.3 45.8 41.3 33.6 17.6 5.3 1.1 -- High-income countries NRA, agricultural 23.0 30.9 36.8 26.5 34.7 43.0 55.5 48.2 36.6 33.9 18.3 NRA, nonagricultural 7.5 8.5 7.7 5.4 3.6 3.4 3.2 2.5 1.7 1.3 0.7 RRA 14.3 20.6 27.1 19.9 30.1 38.3 50.6 44.6 34.3 32.1 19.2 World b NRA, agricultural -- 5.6 7.6 0.8 2.6 5.7 18.7 19.7 18.4 18.6 -- NRA, nonagricultural -- 19.0 20.5 16.1 13.7 10.0 9.8 7.6 6.0 4.0 -- RRA -- 11.3 10.7 13.2 9.8 3.6 8.1 11.3 11.8 14.0 -- Source: Anderson and Valenzuela (2008), based on estimates reported in the project's national country studies. Note: -- not available. a. The RRA is defined as 100*[(100 NRAag t ) (100 NRAnonagt ) 1], where NRAagt and NRAnonagt are the percentage NRAs for the tradables parts of the agricultural and nonagricultural sectors, respectively. b. Estimates for the RRA for China pre-1981 and India pre-1965 are based on the assumption that the agricultural NRAs in those years were the same as the average NRA estimates for those countries for 1981­84 and 1965­69, respectively, and that the value of production in those missing years is that which gives the same average share of value of production in total world production in 1981­84 and 1965­69, respectively. Developing-world and country aggregates are computed accordingly. 35 36 Distortions to Agricultural Incentives: A Global Perspective Figure 1.7. RRAs to Agriculture,a High-Income Countries, 1955­2007 200 150 100 percent 50 0 50 9 4 9 4 9 4 9 4 9 4 7 ­5 ­6 -6 ­7 ­7 ­8 ­8 ­9 ­9 ­0 ­0 65 55 60 70 75 80 85 90 95 00 05 19 19 19 19 19 19 19 19 19 20 20 Non-EU Western Europe EU Japan/Rep. of Korea United states Canada Australia/New Zealand Source: Anderson and Valenzuela (2008), based on estimates reported in the project's national country studies. a. The RRA is defined as 100*[(100 NRAagt ) (100 NRAnonagt ) 1], where NRAagt and NRAnonagt are the percentage NRAs for the tradable parts of the agricultural and nonagricultural sectors, respectively. removed. The spread of NRAs and RRAs for all our focus countries for 2000­04 is shown in figure 1.8. As economic growth continues, there are fewer countries below the zero lines in those ladders, reflecting an ongoing tendency for political economy forces to transform countries from taxing to subsidizing their farmers relative to producers of nonfarm goods as per capita incomes grow. For the whole panel dataset, that tendency over the 50-year period examined appears to be almost exactly the same for developing as for high-income countries, the only dif- ference being the average RRA starting point (see the country fixed effects regres- sion lines in figure 1.9).15 A final way of summarizing two of the reform indicators used above--the RRA and the TBI, which allow assessment of changes in the antiagricultural or proagri- cultural production and trade biases in policy regimes--is to map them in two- dimensional space. Figure 1.10 shows agriculture's TBI on the horizontal axis and the RRA on the vertical axis. An economy with no antiagricultural or proagricul- tural bias (RRA 0) and no antitrade or protrade bias within the farm sector (TBI 0) would be located at the intersection of the two axes in figure 1.10. Africa, Asia, and Latin America (the last shown in the figure as Latin America and Five Decades of Distortions to Agricultural Incentives 37 Figure 1.8. Cross-Country Dispersion of NRAs and RRAs, 2000­04 a. NRAa Norway Switzerland Iceland Korea, Rep. of Japan Slovenia Taiwan, China Romania Latvia EU-15 Lithuania Hungary Slovak Republic Colombia Czech Republic Canada Turkey Philippines Vietnam Estonia India United States Russian Federation Mozambique Indonesia Mexico Ecuador Poland Sri Lanka Kenya China Chile Brazil Bulgaria Bangladesh Dominican Republic New Zealand Malaysia Pakistan Madagascar Australia Uganda Mali South Africa Chad Burkina Faso Cameroon Thailand Benin Togo Ghana Nicaragua Kazakhstan Nigeria Egypt, Arab Rep. of Senegal Ethiopia Ukraine Sudan Tanzania Argentina Côte d'Ivoire Zambia Zimbabwe 50 0 50 100 150 200 percent 38 Distortions to Agricultural Incentives: A Global Perspective Figure 1.8. (continued) b. RRA Korea, Rep. of Norway Switzerland Iceland Japan Taiwan, China Slovenia Romania Lithuania Latvia EU-15 Colombia Hungary Canada Estonia Czech Republic Turkey Philippines United States India Slovak Republic Poland Indonesia Mexico Russian Federation Mozambique Ecuador Kenya Chile China Malaysia Vietnam Brazil New Zealand Dominican Republic Australia South Africa Bulgaria Nigeria Thailand Ghana Nicaragua Sri Lanka Kazakhstan Madagascar Pakistan Cameroon Ukraine Bangladesh Sudan Uganda Argentina Senegal Egypt, Arab Rep. of Tanzania Ethiopia Côte d'Ivoire Zambia Zimbabwe 100 50 0 50 100 150 200 percent Source: Anderson and Valenzuela (2008), based on estimates reported in the project's national country studies. Note: NRA for all agriculture products, including NPS. Five Decades of Distortions to Agricultural Incentives 39 Figure 1.9. Relationships between Real GDP Per Capita and RRA,a All Focus Countries, 1955­2007 400 300 200 RRA (%) 100 0 100 1 0 1 2 3 ln real GDP per capita high-income country RRA observations high-income country fitted values developing-country RRA observations developing-country fitted values Coefficient Standard error R2 Developing countries 0.26 0.02 0.17 High-income countries 0.28 0.03 0.14 Source: Author's derivation with country fixed effects, using data in Anderson and Valenzuela (2008) that are based on RRA estimates reported in the project's national country studies. the Caribbean) were all well to the southwest of that neutral point as of 1980­84, but by 2000­04 all had moved much closer to the vertical axis (meaning they had reduced their antitrade bias in agriculture), and all but Africa had become closer to the horizontal axis--although Asia is now above rather than below that axis, which means the countries in that region are assisting farmers relatively more than producers of other tradable products. While that can lead to just as much waste of resources as the earlier, antiagricultural, policy bias, it is only in Korea and in Taiwan, China that the 2000­04 RRA is well above zero (it is just 1 percent for China and 4 percent for Southeast Asia). Producer assistance and consumer taxation in value terms In conclusion, this survey returns to the dollar values mentioned at the outset. Table 1.9 shows the full time series, in aggregate and by per person engaged in 40 Distortions to Agricultural Incentives: A Global Perspective Figure 1.10. Relationship between RRA and the TBI for Agriculture, Focus Regions, 1980­84 and 2000­04 150 Japan 100 Japan Western Europe RRA (%) 50 Western Europe North America Europe and Central Asia Asia 0 Africa Africa Latin America Australia and Asia and the Caribbean New Zealand 50 0.6 0.5 0.4 0.3 0.2 0.1 0 trade bias index 1980­84 2000­04 Source: Author's derivation using data in Anderson and Valenzuela (2008) that are based on NRA and RRA estimates reported in the project's national country studies. agriculture. Two points are worth stressing from the latter estimates. One is that for high-income countries, when decoupled payments are included, assistance per farmer has continued to rise over time, even when expressed in constant 2000 U.S. dollars. By 2000­04, the annual transfer was double what it was in 1975­79, which, in turn, is more than double the transfer in 1965­69. At US$13,400 per farmer, that is a very large transfer to a group whose household incomes are not significantly below those of nonfarm workers when adjusted for differences between urban and rural costs of living.16 The second point to note from table 1.9b is that for half this period, farmers in developing countries were effectively taxed between US$0.30 and US$0.50 per day by price-distorting policies, at a time when many hundreds of millions of them were struggling to survive on less than US$1 a day per house- hold member. Thankfully, most of the export taxation that drove those transfers was dismantled, although some was resurrected in response to high international food prices in 2007 and 2008. The move from negative to positive transfers to developing country farmers since the mid-1990s is not necessarily a good thing, however, even if society believes farmers should be compensated for the cost of adjusting to the structural changes that accompany rapid economic growth-- there are almost always more efficient ways to raise welfare of poor people than via price and trade policies. Table 1.9. GSEs of Assistance to Farmers, Total and Per Farm Worker, by Region, 1965­2007 a. Total (constant 2000 US$ billions per year) 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Africa 5 9 11 6 1 7 5 6 -- Asia 69 110 126 106 38 8 38 57 -- Latin America 2 13 13 14 10 4 5 4 -- Europe and Central Asia a 3 1 2 10 1 4 19 19 26 Western Europe 89 76 124 133 123 117 76 55 34 United States and Canada 19 16 19 31 36 30 20 30 32 Australia and New Zealand 1 1 1 1 1 1 0 0 0 Japan 16 18 37 33 53 59 53 42 22 All focus countries 46 21 30 64 165 201 208 202 -- Developing countries 79 133 152 136 48 6 58 75 -- High-income countries 125 112 182 199 214 207 150 127 80 High-income countries, including decoupledb 125 112 186 223 238 235 193 173 130 World (scaled)c 51 24 34 70 192 225 233 223 -- (Table continues on the following page.) 41 42 Table 1.9. GSEs of Assistance to Farmers, Total and Per Farm Worker, by Region, 1965­2007 (continued) b. Per person engaged in agriculture (constant 2000 US$ per year) 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Africa 77 137 142 81 10 69 44 46 -- Asia 114 166 174 136 45 9 41 60 -- Latin America 79 410 408 407 305 132 177 143 -- Europe and Central Asia 123 54 101 443 45 141 541 569 888 Western Europe 4,259 4,433 8,163 10,097 10,889 12,397 9,617 8,427 5,141 United States and Canada 3,733 3,436 4,189 6,730 8,220 7,478 5,582 9,182 8,763 Australia and New Zealand 3,504 3,452 3,492 4,744 3,226 1,613 1,221 442 1,061 Japan 1,301 1,845 4,830 5,432 10,234 13,957 16,234 16,933 12,469 All focus countries 59 27 34 66 164 185 181 172 -- Developing countries 116 177 185 154 51 6 53 67 -- High-income countries 3,261 3,491 6,483 8,166 9,972 11,325 9,745 9,871 7,180 High-income countries, including decoupleda 3,261 3,491 6,625 9,151 11,091 12,857 12,539 13,447 11,681 Source: Anderson and Valenzuela (2008), based on estimates reported in the project's national country studies. Note: -- not available. a. Data for Europe and Central Asia are only Turkey until 1991. Figures are from FAOSTAT, which may differ from national statistics. Estimates for the NRA and RRA for China pre-1981 and India pre-1965 are based on the assumption that the agricultural NRAs in those years were the same as the average NRA estimates for those countries for 1981­84 and 1965­69, respectively, and that the value of production in those missing years is that which gives the same average share of value of production in total world production in 1981­84 and 1965­69, respectively. Developing-country and world aggregates are computed accordingly. b. Decoupled payments to farmers are excluded from all rows except the final two. c. These figures assume that the NRA in nonfocus countries is the same as the average for the focus countries in each region (including decoupled payments in the case of other high-income countries), and that their share of the value of regional agricultural production at undistorted prices is the same as their average share of the region's agricultural GDP at distorted prices during 1990­2004. For the countries of North Africa (other than the Arab Republic of Egypt) and the Middle East, NRAs are assumed to be the same as Turkey's. Five Decades of Distortions to Agricultural Incentives 43 Since it is mostly border (trade) policies rather than domestic measures that are responsible for the nominal assistance provided to farmers, those border measures also confer a tax on consumers of farm products. The extent of those CTEs, shown in table 1.10 for all covered farm products, also differs from NRAs in that the consumption weights are not the same as production weights for tradable products. But in percentage terms, their order of magnitude is similar to the NRAs. Again, two points are noteworthy. One is that developing countries' price- distorting policies prior to the 1990s provided up to US$0.30 per capita per year to consumers, directly at the expense of farmers, who on average are far poorer than urban dwellers. The other is that the cost to consumers in high-income countries of farm support policies is miniscule relative to their incomes, amount- ing even at their peak in the most protected countries (other than Japan) to well under US$1 a day. It is thus not surprising that domestic opposition to these policies is weak. Additional Indicators of Agricultural Price Distortions The indicators used above to summarize trends in the extent of distortions within the agricultural sector of a country include the unweighted or weighted mean NRA of covered products, the standard deviation of covered product NRAs, the weighted mean NRA for exportable versus import-competing covered products, and the TBI of the agricultural sectors' covered plus noncovered trad- able products. The reason for reporting these various indicators of dispersion of NRAs--apart from their being informative in their own right--is that theory suggests the national economic welfare cost of government policy distortions to incentives in terms of resource misallocation tends to be greater, the greater the degree of substitution in production in response to changes in price. In the case of agriculture, which involves the use of farmland that is sector-specific but substitutable among farm activities, the greater the variation of NRAs across industries within the sector then the higher will be the welfare cost of those market interventions. While these various indicators of dispersion are useful, it would also be helpful to have a single indicator to capture the overall welfare or trade effect of each country's regime of agricultural price distortions in place at any time, and to trace its path over time and make cross-country comparisons. To that end, chapter 11 draws on a theoretical literature that has developed in recent years to provide such indicators for national price and trade policies that are well grounded in theory. They belong to the family of indexes first developed by Anderson and Neary (2005) under the catch-all name of trade restrictiveness indexes. 44 Table 1.10. CTEs of Policies Assisting Producers of Covered Farm Products, Percent and Per Capita, by Region, 1965­2007 a. Percent 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Africa 12 16 9 6 16 8 0 3 -- Asia 12 15 2 15 14 3 5 10 -- Latin America 7 18 13 12 10 13 6 8 -- Europe and Central Asiaa 17 6 7 28 1 2 9 17 12 Western Europe 74 49 59 70 65 49 37 32 17 United States and Canada 8 6 7 13 10 2 5 2 2 Australia and New Zealand 15 11 10 10 10 8 4 2 1 Japan 67 68 93 99 135 119 116 107 81 Developing countries 12 16 5 14 10 0 5 8 -- High-income countries 42 30 40 45 50 41 32 27 16 All focus countries 23 14 21 10 15 16 15 16 -- b. Per capita (constant 2000 US$ per year) 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Africa 15 27 16 12 21 10 0 3 -- Asia 6 13 2 22 18 4 6 10 -- Latin America 4 24 22 31 16 34 11 12 -- Europe and Central Asiaa 20 9 9 41 1 13 15 30 20 Western Europe 240 202 306 276 247 207 138 103 50 United States and Canada 40 35 43 67 42 8 18 9 11 Australia and New Zealand 112 87 80 68 56 49 22 12 8 Japan 130 157 287 258 435 498 443 344 221 Developing countries 7 15 5 22 13 1 6 8 -- High-income countries 153 136 207 195 199 176 124 94 59 All focus countries 30 18 39 23 27 29 26 23 -- Source: Anderson and Valenzuela (2008), based on estimates reported in the project's national country studies. Note: -- not available. a. Europe and Central Asia data are for Turkey only until 1991. 45 46 Distortions to Agricultural Incentives: A Global Perspective To capture distortions imposed by each country's border and domestic policies on its economic welfare and its trade volume, Lloyd, Croser, and Anderson (2009a) define a welfare reduction index (WRI) and a trade reduction index (TRI) and estimate them for the project's focus countries since 1960, taking into account that for some covered products the NRA and CTE differ (because there are domestic measures in place in addition to or instead of trade measures). As their names suggest, these two new indexes respectively capture in a single indica- tor the (partial equilibrium) welfare- or trade-reducing effects of distortions to consumer and producer prices of covered farm products from all agricultural and food policy measures in place (while ignoring noncovered farm products and indirect effects of sectoral and trade policy measures directed at nonagricultural sectors). The WRI measure reflects the welfare cost of agricultural price-distorting policies better than the NRA or CTE because it recognizes that the cost of a government-imposed price distortion is related to the square of the price wedge. It thus captures the disproportionately higher welfare costs of peak levels of assis- tance or taxation, is larger than the mean, and is positive regardless of whether the government's agricultural policy favors or hurts farmers. The WRI and TRI meas- ures also have the advantage of providing a theoretically sound indicator of the welfare (or trade) effect in a single sectoral measure that is comparable across time and place. In this way, the WRI and TRI come close to what a computable general equilibrium (CGE) can provide in the way of estimates of the trade and welfare (and other) effects of the price distortions captured by the product NRA and CTE estimates, while having the advantage of providing an annual time series. The time series derived for this project, available as Anderson and Croser (2009), is detailed in chapters 11 and 12. The WRI estimates in figure 1.11 indicate an increasing tendency for covered products' policies to reduce welfare between the 1960s and the mid-1980s, and then to improve welfare in the 1990s (when the WRI declines). This pattern is generated by different policy regimes in the different country groups, though: in high-income countries, covered products were assisted throughout the years reported, although less so after the 1980s, whereas covered products in developing countries were disprotected until the most recent years. That is, the WRI has the desirable property of correctly identifying the welfare consequences that result from both positive and negative assistance regimes, because it captures the disper- sion of NRAs among covered products: the larger the variance in assistance levels, the greater the potential for resources to be used in activities that do not maximize economic welfare. One consequence of these tendencies is that the WRI values are much higher than the NRAs for high-income countries. Another is that the WRI for Africa spikes in the mid-1980s, in contrast to the NRA, which moves close to zero. The Five Decades of Distortions to Agricultural Incentives 47 Figure 1.11. Welfare Reduction Indexes for Covered Tradable Farm Products, by Region, 1960­2007 a. Africa, Asia, and Latin America 80 60 percent 40 20 0 4 9 4 9 4 9 4 9 4 ­6 ­6 ­7 ­7 ­8 ­8 ­9 ­9 ­0 60 65 70 75 80 85 90 95 00 19 19 19 19 19 19 19 19 20 Africa Asia Latin America b. Developing countries, high-income countries, and Europe's transition economies 100 80 60 percent 40 20 0 4 9 4 9 4 9 4 9 4 7 ­6 ­6 ­7 ­7 ­8 ­8 ­9 ­9 ­0 ­0 60 65 70 75 80 85 90 95 00 05 19 19 19 19 19 19 19 19 20 20 high-income countries Europe's transition economies developing countries Source: Authors' calculations based on NRAs and CTEs in Anderson and Valenzuela (2008). 48 Distortions to Agricultural Incentives: A Global Perspective reason is that while Africa was still taxing exportables, it moved (temporarily) from low to very high positive levels of protection for import-competing farm products (table 1.4). At the aggregate level, African farmers received almost no government assistance then (NRA close to zero), but the welfare cost of its mix- ture of agricultural policies as a whole was at its highest according to the WRI. A third consequence is that for developing countries, the average WRI in the years 1995 to 2004 is around 20 percent, even though the average NRA for covered products in those years is close to zero (see figure 1.3), again reflecting the high dispersion across product NRAs--particularly between exportables and import- competing goods--in each country.17 It is possible to use the value of production (at undistorted prices) to estimate the contribution to global welfare reduction by the various regions' policies affect- ing covered agricultural products. Figure 1.12a shows that, since 1993, high- income countries (plus Korea) have been responsible for around 60 percent of that global welfare reduction, developing Asia for about 25 percent, and Africa for nearly 10 percent. Asia's share has been about 40 percent for the whole period since 1981, but as developing Asia's contribution declined over the 1980s, the con- tribution of Japan and Korea grew. Figure 1.12b reveals that China's contribution has fallen substantially over the 1990s. Recall, though, that the WRI picks up the variance for these countries' product NRAs within the sector. Thus, even though China had an aggregate agricultural NRA of only 6 percent in the period 1994­2005, it had negative NRAs on cotton and rice and positive NRAs for other products, so it still had a moderately large WRI. For developing countries as a group, the trade restrictiveness of agricultural policy was roughly constant until the early 1990s. Thereafter it declined, especially for Asia and Latin America, according to the TRI estimates. For high-income countries, the TRI time path was similar, although the decline began a few years later. Aggregate results for developing countries are driven by the exportables subsector, which is being taxed, and the import-competing subsector, which is being protected (albeit by less than in high-income countries). For high-income countries, policies have supported both exporting and import-competing agricul- tural products. Even though these policies heavily favor the latter, assistance to exporters has offset the antitrade bias from the protection of import-competing producers to some degree in terms of impacts on those countries' aggregate volume of trade in farm products. Thus, up to the early 1990s, the TRI for high-income countries was below that for developing countries. To use again the example of Africa, the TRI peaked in 1985­89, when the NRA was closest to zero, correctly identifying the trade-reducing effect of positive protection to the import-competing farmers and disprotection to producers of exportables (figure 11.6 and table 11.4 of Lloyd, Croser, and Anderson 2009a). Figure 1.12. Regional Shares of Global and Country Shares of Asian Welfare Reductiona from Agricultural Policies, 1981­2004 a. Regional share of global welfare reduction 100 90 80 70 60 percent 50 40 30 20 10 0 19 1 19 2 19 3 19 4 19 5 19 6 19 7 19 8 19 9 19 0 19 1 19 2 19 3 19 4 19 5 19 6 19 7 19 8 20 9 20 0 20 1 02 20 3 04 8 8 8 8 8 8 8 8 8 9 9 9 9 9 9 9 9 9 9 0 0 0 19 20 Latin America Africa other Asia Japan and Korea other high-income countries Western Europe b. National share of welfare reduction in developing Asia (excluding Rep. of Korea) 100 90 80 70 60 percent 50 40 30 20 10 0 19 1 19 2 19 3 19 4 19 5 19 6 19 7 19 8 19 9 19 0 19 1 19 2 19 3 19 4 19 5 19 6 19 7 19 8 20 9 20 0 20 1 20 2 20 3 04 8 8 8 8 8 8 8 8 8 9 9 9 9 9 9 9 9 9 9 0 0 0 0 19 Vietnam Thailand Taiwan, China Sri lanka Philippines Pakistan Malaysia Indonesia India China Bangladesh Source: Author's derivation using data from Anderson and Croser (2009), based on estimates in Anderson and Valenzuela (2008). a. The value of welfare reduction is computed as the WRI multiplied by the value of production at undistorted prices valued at constant 2000 U.S. dollars. European transition economies are not included, nor are countries not included in the Anderson and Valenzuela sample (together, these make up less than 5 percent of global GDP). 49 50 Distortions to Agricultural Incentives: A Global Perspective Chapter 12 differs from the preceding chapters in that it focuses on commodi- ties rather than countries, showing to what extent global markets for some farm commodities are distorted relative to others. It also shows that the basic food staples of the poor in low-income countries are reasonably well covered in our database for African countries where they matter most, and that their noncoverage in other regions does not bias the aggregate NRA estimates for covered products. Chapter 12 also reports new partial equilibrium estimates of global commodity TRIs and WRIs, to parallel the country-based ones reported in chapter 11. A sum- mary of both indicators for 28 key farm commodities is provided in figure 1.13 (taken from Lloyd, Croser, and Anderson 2009b). It shows that the most distorted of all commodities in 2000­04, in terms of both their global welfare cost and their trade restrictiveness, are rice, sugar, milk, and beef--although cocoa trade is just as restricted as beef, because export restrictions are still prevalent in major supplying countries such as Côte d'Ivoire. Figure 1.13. Trade and Welfare Reduction Indexes for 28 Major Agricultural Products, 2000­04 140 120 100 80 60 percent 40 20 0 20 40 su ice b lk ou ou n so dn ry h s se um co me ba oa m y ee let m t t ra g mtea se at so e ds yb gg w ans su m at pa ow e co cof oil co fee ru uts w va l m r gr p tto f lm er ss r oo co ee p oa ea ga ca bbe rg ut rle nf aiz i n lt pe e he e a c il r n sa e l pi sh welfare reduction index trade reduction index Source: Lloyd, Croser, and Anderson (2009b). Five Decades of Distortions to Agricultural Incentives 51 Economy-Wide Effects of Past Reforms and Remaining Policies It is clear that there has been a great deal of change over the past quarter century in policy distortions to agricultural incentives throughout the world: the antiagri- cultural and antitrade biases of policies of many developing countries have been reduced; export subsidies of high-income countries have been cut; and some rein- strumentation toward less inefficient and less trade-distorting forms of support, particularly in Western Europe, has begun. However, protection from agricultural import competition has continued an upward trend in both rich and poor coun- tries, notwithstanding the Uruguay Round Agreement on Agriculture (URAA), which aimed to bind and reduce farm tariffs. What, then, have been the net economic effects of agricultural price and trade policy changes around the world since the early 1980s? And how do those effects on global markets, farm incomes, and economic welfare compare with the effects of price distortions still in place as of 2004? The final chapter by Valenzuela, van der Mensbrugghe, and Anderson (2009) uses a global economy-wide model to provide a combined retrospective and prospective analysis that seeks to assess how far the world has come--and how far it still has to go--in harmonizing world agricultural policy. It quantifies the impacts of both past reforms and current policies by comparing the effects of the project's NRA and CTE distortion esti- mates for the period 1980­84 with those of 2004. Several key findings from that economy-wide modeling study are worth emphasizing. First, the policy reforms from the early 1980s to the mid-2000s improved global economic welfare by US$233 billion per year, and removing the distortions remaining as of 2004 would add another US$168 billion per year. This change suggests that in terms of global welfare, the world had moved three-fifths of the way toward global free trade in goods over a quarter century. Second, developing economies benefited proportionately more than high- income economies (1.0 percent compared with 0.7 percent of national income) from those past policy reforms, and would gain nearly twice as much as high- income countries by completing the reform process (an average increase of 0.9 percent compared with 0.5 percent for high-income countries). Of the prospective welfare gains from global liberalization, 70 percent would come from agriculture and food policy reform. This is a striking result given that the shares of agriculture and food in global GDP and global merchandise trade are only 3 and 6 percent, respectively. The contribution of farm and food policy reform to the prospective welfare gain for just developing countries is even greater, at 72 percent. Third, the share of global farm production exported (excluding intra-EU trade) in 2004 was slightly smaller as a result of those reforms since 1980­84, because there were less farm export subsidies. Agriculture's 8 percent share in 52 Distortions to Agricultural Incentives: A Global Perspective 2004 contrasts with the 31 percent share for other primary products and the 25 percent for all other goods--a "thinness" that is an important contributor to the volatility of international prices for weather-dependent farm products. If the poli- cies distorting goods trade in 2004 were removed, the share of global production of farm products that is exported would rise from 8 to 13 percent, thereby reduc- ing the instability of prices and quantities of products traded. Fourth, developing countries' share of the world's primary agricultural exports rose from 43 to 55 percent, and its farm output share from 58 to 62 percent, because of those reforms, with increases in both in nearly all agricultural indus- tries except rice and sugar. Removing the remaining goods market distortions would boost export and output shares to 64 and 65 percent, respectively. Fifth, the average real price in international markets for agricultural and food products would have been 13 percent lower had policies not changed over the past quarter century. Evidently, the impact of the decline in RRA in high-income countries (including the cuts in farm export subsidies) on international food prices more than offset the opposite impact of the increase in RRA (including the cuts in agricultural export taxes) in developing countries over that period, and the net effect was higher food prices. By contrast, removing the distortions remaining as of 2004 is projected to raise the international price of agricultural and food products by less than 1 percent on average. This is contrary to earlier modeling results based on the Global Trade Analysis Project (GTAP) protections database. For example, Anderson, Martin, and van der Mensbrugghe (2006) estimated prices would rise by 3.1 percent, or for just primary agriculture, by 5.5 percent. The lesser impact in the new results presented here is due to the fact that export taxes in developing countries based on the above NRA estimates (most notably for Argentina) are included in the new database: removing those taxes offsets the international price-raising effect of eliminating import protection and farm sub- sidies elsewhere. Sixth, for developing countries as a group, net farm income (value added in agriculture) is estimated to be 4.9 percent higher than it would have been without the reforms of the past quarter century, which is more than ten times the propor- tional gain for nonagriculture. If policies remaining in 2004 were removed, net farm incomes in developing countries would rise a further 5.6 percent, compared with just 1.9 percent for nonagricultural value added. In addition, returns to unskilled workers in developing countries--the majority of whom work on farms--would rise more than returns to other productive factors from that liber- alization. Together, these findings suggest both inequality and poverty could be alleviated by such reform, given that three-quarters of the world's poor are in farm households in developing countries (Chen and Ravallion 2007). Five Decades of Distortions to Agricultural Incentives 53 Finally, removal of agricultural price-supporting policies in high-income countries would undoubtedly lead to painful reductions in income and wealth for farmers in those countries if the farmers were not compensated. It should be kept in mind, though, that the majority of farm household income in high-income countries comes from off-farm sources (OECD 2008b), and that the gainers in the rest of their societies could readily afford to compensate the losers from the benefits of freeing trade. Prospects for Further Reform It is not obvious how future agricultural policies might develop. A quick glance at the indicators could lead one to view developments from the early 1960s to the mid-1980s as an aberrant period of welfare-reducing policy divergence (negative and declining RRAs in low-income countries alongside positive and rising RRAs in most high-income countries) that has given way to welfare-improving and poverty-reducing reforms during which the two country groups' RRAs are con- verging. But on inspection of the NRAs for exporting and import-competing sub- sectors of agriculture (figure 1.3), it is clear that the convergence of NRAs to near zero is mainly with respect to the exporting subsector, while NRAs for import- competing farmers are positive and trending upward over time at the same rate in both developing and high-income countries--notwithstanding the URAA, which aimed to bind and reduce farm tariffs. True, applied tariffs have been lowered or suspended as a way of dealing with the international food price spike in 2008, but this, and the food export taxes or quantitative restrictions imposed that year by numerous food-exporting developing countries, may be in place only until inter- national prices return to trend (as happened after the price hike of 1973­74 and the price dip of 1986­87). Indications are very mixed as to why some countries have reformed their price- distorting agricultural and trade policies more than others in recent decades. Some reforming countries have acted unilaterally, apparently convinced that it is in their national interest to do so. China is the most dramatic and significant example of a developing country adjusting its agricultural and trade policy over the past three decades according to its national interest, while among high-income countries only Australia and New Zealand (and to a lesser extent the EU) are in that category. Others countries have reformed their policies partly to secure bigger and better loans from international financial institutions and then, having taken that first step, have continued the process, even if somewhat intermittently. India is one example, but there are numerous others in Africa and Latin America. Few countries have gone backwards in terms of increasing their antiagricultural bias, but Zimbabwe and perhaps Argentina qualify during the present decade--and 54 Distortions to Agricultural Incentives: A Global Perspective numerous others joined them in 2008, at least temporarily, in response to the sud- den spike in international food prices. Still other countries have reduced their agricultural subsidies and import barriers at least partly in response to the GATT's multilateral URAA, the EU being the most important example (encouraged by its desire for preferential trade agreements, including its recent eastward expansion). EU reforms suggest growth in agricultural protection can be slowed and even reversed if accompanied by reinstrumentation away from price supports to decoupled measures or more direct forms of farm income support. The starker examples of Australia and New Zealand show that one-off buyouts can bring faster and even complete reform.18 But in developing countries, where levels of agricultural protection are generally lower than in high-income countries, there are fewer signs of a slowdown in the upward trend in agricultural protection from import competition over the time period studied. Indeed, there are numerous signs that developing-country governments want to retain the option to raise agricultural NRAs in the future, particularly via import restrictions. One indicator is the high tariff bindings developing coun- tries committed themselves to following the Uruguay Round of the WTO: as of 2001, actual applied tariffs on agricultural products averaged less than half the corresponding bound tariffs for developing countries of 48 percent, and less than one-sixth in the case of least developed countries (Anderson and Martin 2006). Another indicator of reluctance to undertake agricultural trade reform is the unwillingness of many developing countries to agree to major cuts in bound agri- cultural tariffs in the WTO's ongoing Doha Round of multilateral trade negotia- tions. Indeed, many developing countries believe that high-income countries should commit to reducing their remaining farm tariffs and subsidies before developing countries should offer further reform commitments of their own. Yet modeling results reported in Valenzuela, van der Mensbrugghe, and Anderson (2009) suggest that if high-income countries alone were to liberalize their agricul- tural markets, such subglobal reform would provide less than two-thirds of the potential gains to developing countries that could result from global agricultural policy reform. In addition, current negotiations have brought to the fore a new proposal for agricultural protectionism in developing countries, based on the notion that agri- cultural protection is helpful and needed for food security, livelihood security, and rural development. This view has succeeded in bringing "special products" and a "special safeguard mechanism" into the multilateral trading system's agricultural negotiations, despite the fact that such policies, which would raise domestic food prices in developing countries, may worsen poverty and jeopardize the food secu- rity of the poor (Ivanic and Martin 2008). Five Decades of Distortions to Agricultural Incentives 55 For developing countries to wait for high-income country reform before liber- alizing their own farm trade is unwise as a poverty alleviating strategy, not least because the past history revealed in the NRAs summarized above suggests such reform will be, at best, slow in coming. In the United States, for example, the most recent two five-year farm bills were a step backwards from the previous regime, which at least sought to reinstrument protection toward less trade-distorting measures (Gardner 2009). Nor have the world's large number of new regional integration agreements of recent years been very successful in reducing farm pro- tection. Furthermore, for developing countries to postpone their own reform would be to forego a major opportunity to boost their own and (given the size and growth in South-South trade of late) their neighbors' economies. It would be dou- bly wasteful if, by being willing to commit to reform in that way, they are able to convince high-income countries to reciprocate by signing on to a more ambitious Doha Round, the potential global benefits of which are considerable.19 Developing countries that continue to open their domestic markets and prac- tice good macroeconomic governance will keep growing. Typically, growth will be more rapid in manufacturing and services activities than in agriculture, especially in the more densely populated countries where agricultural comparative advan- tage is likely to decline. Whether such economies become more dependent on imports of farm products depends, however, on what happens to their RRA. The first wave of Asian industrializers (Japan, and then Korea and Taiwan, China) chose to slow the growth of food import dependence by raising their NRAs for agriculture even as they were reducing their NRAs for nonfarm tradables, such that their RRAs became increasingly above the neutral zero level. A key question is: will later industrializers follow suit, given the past close association of RRAs with rising per capita income and falling agricultural comparative advantage? Figure 1.9 suggests developing countries' RRA trends of the past three decades have been on the same trajectory as the high-income countries prior to the 1990s, so unless new forces affect their policies, the governments of later industrializing economies may well follow suit. One new force is constraints on farm subsidies and protection policies of WTO member countries following the Uruguay Round. Earlier industrializers were not bound under GATT to keep down their agricultural protection. Had there been strict discipline on farm trade measures at the time Japan and Korea joined GATT in 1955 and 1967, respectively, their NRAs may have been halted at less than 20 percent (figure 1.14). At the time of China's accession to WTO in December 2001, its NRA was less than 5 percent according to this present study, or 7.3 per- cent for just import-competing agriculture. Its average bound import tariff com- mitment was about twice that (16 percent in 2005), but what matters most is China's out-of-quota bindings on the items whose imports are restricted by tariff 56 Distortions to Agricultural Incentives: A Global Perspective Figure 1.14. NRAs for Japan, Republic of Korea, and China, and Date of Accession to GATT or WTO, 1955­2005 200 150 Japan (1955 16.6%) 100 Korea (1967 7.4%) NRA (%) China 50 (2001 4.5%) 0 50 100 9 4 9 4 9 4 9 4 9 5 ­5 ­6 ­6 ­7 ­7 ­8 ­8 ­9 ­9 ­0 55 60 65 70 75 80 85 90 95 00 19 19 19 19 19 19 19 19 19 20 Japan Korea China Source: Based on estimates in Anderson and Valenzuela (2008). rate quotas. The latter tariff bindings as of 2005 were 65 percent for grains, 50 per- cent for sugar and 40 percent for cotton (Anderson, Martin, and Valenzuela forth- coming). Clearly, the legal commitments even China made on acceding to the WTO are a long way from current levels of support for its farmers, and so are unlikely to constrain the government very much in the next decade or so. The legal constraints on developing countries that joined the WTO earlier are even less constraining. For India, Pakistan, and Bangladesh, for example, estimated NRAs for agricultural importables in 2000­04 are 34, 4, and 6 percent, respectively, whereas the average bound tariffs on their agricultural imports are 114, 96, and 189 percent, respectively (WTO, ITC, and UNCTAD 2007). And like other devel- oping countries, these countries have high bindings on product-specific domestic support (10 percent), and another 10 percent for NPS assistance, for a total of 20 more percentage points of NRA (17 percent in China's case) that legally could come from domestic support measures--compared to 10 percent in India and less than 3 percent in the rest of South Asia. It is hoped developing countries will choose not to make use of the legal wiggle room they have allowed themselves in their WTO agreements to follow Japan, Korea, and Taiwan, China into high agricultural protection. A much more effi- cient and equitable strategy would be to instead treat agriculture in the same way they have been treating nonfarm tradable sectors. That would involve opening the sector to international competition, relying on more efficient domestic policy Five Decades of Distortions to Agricultural Incentives 57 measures for raising government revenue (for example, income and consumption or value-added taxes), and assisting farm families--including younger members seeking off-farm employment--via public investment in rural education and health, rural infrastructure, and agricultural research (Otsuka and Yamano 2006; Otsuka, Estudillo, and Sawada 2009). Historically, developing countries' expendi- ture on public agricultural research has amounted to the equivalent of less than 1 percent of the gross value of farm production (table 1.11), so it would not be dif- ficult to double that level of investment with a diversion of just a small amount of the price support currently provided to farmers in those developing countries that provide farm import protection or input subsidies. As for high-income countries, the above distortion estimates show that they have all lowered price supports for their farmers since the 1980s. In some coun- tries, those supports have been partly replaced by assistance that is at least some- what decoupled from production. If that trend continues at the pace of the past quarter century, and if there is no growth of agricultural protection in developing countries, most of the disarray in world food markets may be removed before the middle of this century. However, if the WTO's Doha Development Agenda col- lapses, and governments thereby find it more difficult to ward off agricultural protection lobbies, it is more likely that developing countries will follow the same agricultural protection path this century as that which was taken by high-income countries during the last century. One way to encourage developing countries to follow a more liberal policy path could be to extend the Integrated Framework's Diagnostic Trade Integration Study process to a broader range of low-income countries. That process, which provides action plans for policy and institutional reform and lists investment and technical assistance needs, could be expanded to include the "aid for trade reform" proposal that has been discussed in the context of the Doha Round (Hoekman 2005)--regardless of the fate of that round. Table 1.11. Intensity of Public Agricultural R&D Investment, High-Income and Developing Countries, 1971­2004 (expenditure as percentage of gross value of agricultural production at undistorted prices, unweighted average across available countries) 1970s 1980s 1990s 2000­04 All high-income countries 3.6 2.8 1.9 1.5 All developing countries 1.0 0.8 1.0 1.0 Asia 0.3 0.6 0.7 1.3 Latin America 0.2 0.4 0.6 0.6 Sub-Saharan Africa (not including South Africa) 1.2 1.0 0.9 0.7 Source: Anderson and Valenzuela (2008), based on R&D data from the Consultative Group on International Agricultural Research (CGIAR) Agricultural Science and Technology Indicators Website at http://www.asti.cgiar.org. 58 Distortions to Agricultural Incentives: A Global Perspective Areas for Further Research The fact that indications are mixed as to why some countries appear to have reformed their agricultural and trade policies more than others, and that it is therefore unclear as to how policies might develop in the future, should not be surprising. After all, the long history of globalization is full of episodes of sen- sible policy reforms that for all sorts of reasons get reversed (Findlay and O'Rourke 2007; North, Wallis, and Weingast 2009). If one is to better understand what might happen in the context of continuing economic growth and terms-of- trade volatility, more in-depth analysis of the political economy of past policy behavior is warranted. That is now possible thanks to the new panel set of distor- tion estimates reported here, and some early findings from such analyses will appear in Anderson (forthcoming). That collection includes a broad range of the- oretical and econometric analyses aimed at better understanding the political economy forces that generated the evolving pattern of intersectoral and intra- sectoral distortions to farmer and food consumer incentives over recent decades. To contribute further to policy debate, a second area requires further research-- economy-wide modeling of the impacts on agricultural markets, national eco- nomic welfare and income distribution of alternative policies. Such research is now easier to do well following the recent development of microsimulation add- ons to such CGE models. How the reform of current policies--both in the country under consideration and in the rest of the world--affects the extent of poverty and inequality is explored in a series of new country case studies in Anderson, Cockburn, and Martin (forthcoming), using global and national economy-wide models that are enhanced with detailed earning and spending information of numerous types of urban and rural households. The impact of rest-of-world policy on each coun- try's terms of trade is informed by the project's agricultural distortions database and generated using the same model as in this book's final chapter (Valenzuela, van der Mensbrugghe, and Anderson 2009). It is hoped that other analysts will make use of the project's agricultural distortions database to explore other policy scenarios, including a continuation of past agricultural protection growth for import-compet- ing farmers as an alternative counterfactual to further freeing of trade. A third area for further research is growth diagnostics. For example, how much of the long-term divergence in per capita incomes between current high- income countries on the one hand and developing and transition economies on the other can be explained by the faster long-term growth of agricultural NRAs and RRAs up to the mid-1980s in the former? How much can be explained by the domestic market-insulating fluctuations in NRAs around those long-term assis- tance trends, which have contributed to the volatility of the terms of trade for agricultural-exporting countries? The recent work of Williamson (2008) offers one suggestion as to how such research might proceed. Finally, how much of Five Decades of Distortions to Agricultural Incentives 59 growth since the early 1980s in individual developing countries can be attributed to the reduction in domestic agricultural disincentives? Ravallion and Chen (2007) show that the decline in the antiagricultural bias in farm price policies has contributed significantly to China's poverty reduction. Rural growth has been shown to be a key contributor to the reduction in poverty in India, too (Ravallion and Datt 1996). These types of studies can now be revisited using the more com- prehensive set of measures of the extent of changes in distortions to agricultural incentives summarized above. Annex Table 1.A.1. Export Orientation, Import Dependence, and Self-Sufficiency in Primary Agricultural Production, Focus Countries,a 1961­2004 (percent at undistorted prices) a. Exports as share of production 1961­64 1970­74 1980­84 1990­94 2000­04 Africa 19 17 12 7 8 Asia 5 4 4 6 5 Latin America 24 27 16 16 27 Western Europe 13 16 27 37 43 United States and Canada 14 14 20 20 21 Australia and New Zealand 41 35 44 43 48 Japan 1 2 1 0 1 All focus countries 11 11 13 16 16 Developing countries 8 8 7 8 8 High-income countries 14 15 22 26 29 b. Imports as share of apparent consumption 1961­64 1970­74 1980­84 1990­94 2000­04 Africa 2 2 5 4 4 Asia 4 4 8 16 14 Latin America 2 4 7 10 17 Western Europe 32 28 34 41 46 United States and Canada 4 4 5 9 12 Australia and New Zealand 3 2 3 5 6 Japan 23 24 24 26 27 All focus countries 11 10 12 19 18 Developing countries 3 4 8 14 13 High-income countries 18 16 20 25 27 (Table continues on the following page.) 60 Distortions to Agricultural Incentives: A Global Perspective Table 1.A.1 Export Orientation, Import Dependence, and Self- Sufficiency in Primary Agricultural Production, Focus Countries,a 1961­2004 (continued) (percent at undistorted prices) c. Self-sufficiency ratio 1961­64 1970­74 1980­84 1990­94 2000­04 Africa 120 117 107 104 105 Asia 102 100 96 89 91 Latin America 129 132 110 107 114 Western Europe 78 85 90 94 94 United States and Canada 111 112 119 114 111 Australia and New Zealand 165 151 174 170 183 Japan 78 78 77 74 74 All focus countries 100 101 101 96 98 Developing countries 105 104 99 93 95 High-income countries 96 98 103 101 102 Source: Compiled using the project's estimates of total agricultural production valued at undistorted prices and the FAO's total agricultural trade value data, in Anderson and Valenzuela (2008). a. Includes intra-EU trade. Benin, Burkina Faso, Chad, Iceland, Mali, Togo, and the countries of Europe and Central Asia are not included. Notes 1. Some of the more transformational policy developments happened quite promptly, such as the end of colonization around 1960; the creation of the Common Agricultural Policy (CAP) in Europe in 1962; the floating of exchange rates and associated liberalization, deregualtion, privatization, and democratization in the mid-1980s; the opening of China's economy from 1979; and the demise of the Soviet Union in 1991. 2. According to the FAOSTAT (http://www.fao.org), fewer than 15 million relatively wealthy farm- ers in developed countries, with an average of almost 80 hectares per worker, currently are being helped, at the expense of not only consumers and taxpayers in those rich countries but also the major- ity of the 1.3 billion relatively impoverished farmers and their large families in developing countries who, on average, must earn a living from just 2.5 hectares per worker. 3. In the two decades to 2000-04, the value of global exports as a share of GDP rose from 19 to 26 percent, even though most of GDP is nontradable governmental and other services (World Bank 2007, as summarized in Sandri, Valenzuela, and Anderson 2007). 4. A nine-year update for the Latin American countries in the Krueger, Schiff, and Valdés sample by the same country authors, and a comparable study of seven central and eastern European countries, contain estimates at least of direct agricultural distortions (see Valdés 1996, 2000). The Krueger, Schiff, and Valdés (1991) chapters on Ghana and Sri Lanka have protection estimates back to 1955, as does the study by Anderson and Hayami (1986) for Korea and Taiwan, China (and Japan, and much earlier in the case of rice). 5. See appendix B. The only countries not well represented in the sample are the many small ones and those in the Middle East. In total, however, the omitted countries account for less than 5 percent of the global economy. 6. By way of comparison, the seminal multicountry study of agricultural pricing policy by Krueger, Schiff, and Valdés (1991) covered an average of 23 years to the mid-1980s for 18 focus countries, which collectively accounted for 5­6 percent of global agricultural output. The producer and consumer sup- port estimates of the OECD (2008a) cover 22 years for 30 countries that account for just over one- quarter of the world's agricultural output valued at undistorted prices. Anderson (2009) compares the Five Decades of Distortions to Agricultural Incentives 61 direct and total rates of protection estimated by Krueger, Schiff, and Valdés for their sample of coun- tries with the comparable NRAs and RRAs from the present study for the same sample of countries and same years, and finds them to be almost identical. However, the indicators of antiagricultural and antitrade bias are about two-thirds greater when the fuller sample of 41 developing countries in the present study is used. This suggests that, with a fuller sample, Krueger, Schiff, and Valdés could have stressed the key policy implications of their study even more forcefully. 7. Only a brief summary of the methodology is provided here. For details see Anderson et al. (2008a, 2008b) or appendix A. 8. The extent to which a payment is production-neutral (the degree of decoupling) differs depending on the way it is administered and the expectations of the recipient (Thompson, Dewbre, and Martini 2007). No attempt here is made to evaluate the extent of the distortion that might still be present as poli- cies are decoupled from output price and quantity produced. Rather, values of OECD estimates under certain categories of support are included as decoupled. For the years 1979­85, there was just one cate- gory, called "direct payments;" from 1986, those payments are specified to comprise the OECD's items C (payments based on area planted/animal numbers), D (payments based on historical entitlements), F (payments based on input constraints), and G (payments based on overall farming income); and for 2005­07, those items are replaced by the similar but newly defined items C to E. This categorization for economic purposes (and that for NPS assistance) should not be confused with the legal allocation of domestic support measures into the WTO colored "boxes" in the context of international commitments. 9. Farmers are affected not just by prices of their own products but also by the incentives nonagri- cultural producers face. That is, it is relative prices and hence relative rates of government assistance that affect producer incentives. More than seventy years ago, Lerner (1936) presented his Symmetry Theorem, which proved that in a two-sector economy, an import tax has the same effect as an export tax. This carries over to a model that also includes a third sector producing only nontradables. 10. Data availability also affects the year from which NRAs can be computed. For Europe's transi- tion economies, that starting date is 1992, for Vietnam it is 1986, and for China it is 1981. 11. While the next section summarizes the findings in subsequent chapters, readers should be aware that the emerging economy chapters in Part III of this book are themselves summaries of the findings in a large number of developing-country case studies that are detailed in one of four com- panion volumes covering Africa (Anderson and Masters 2009), East and South Asia (Anderson and Martin 2009a), Latin America and the Caribbean (Anderson and Valdés 2008) and European and Central Asian transition economies (Anderson and Swinnen 2008). 12. Korea and Taiwan, China are are categorized here as developing rather than high-income because at the beginning of the 50-year period under study they were among the poorest economies in the world. 13. The GSE estimates for 1955­59 are smaller for high-income countries but less negative for developing countries, so the aggregate effect is an empirical issue. Unfortunately, our sample of devel- oping countries for that period is too small to provide a reliable estimate of the net effect. 14. As a reality check, compare this project's regional NRAs for nonagricultural tradables for the 1960s with the spot-year national NRAs from manufacturing import protection (in brackets) for the eight countries reported in Little, Scitovsky, and Scott (1970) and Balassa and associates (1971): 120 percent in South Asia (96 percent in Pakistan); 40 percent in East Asia (8 percent in Malaysia, 29­46 percent in the Philippines, and 30 percent in Taiwan, China); and 30 percent in Latin America (141 percent in Argentina, 89-99 percent in Brazil, 89 percent in Chile, and 20­22 percent in Mexico). 15. The R2 values improve (to 0.16 and 0.26, respectively) and the slope of each line steepens (coef- ficients become 0.36 and 0.32, respectively) if the years since 1990 for high-income countries and before 1985 for developing countries are ignored. 16. The drop to US$6,000 per farmer in 2005­07 will likely prove to be temporary if international food prices return to trend levels soon. 17. National WRIs are aggregated across countries using as weights an average of the value of con- sumption and production at undistorted prices; TRIs use the absolute difference between the values of production and consumption at undistorted prices as weights. This is unlike NRAs and RRAs (or CTEs), which use as weights just the value at undistorted prices of production (or consumption). Like 62 Distortions to Agricultural Incentives: A Global Perspective NRAs, RRAs, and CTEs, national and regional WRIs and TRIs for the five-year periods are unweighted averages of the annual indexes. 18. For a detailed analysis of the buyout option versus the slower and less complete cashout option (moving to direct payments), as well as the uncompensated gradual squeeze-out or sudden cutout options, see Orden and Diaz-Bonilla (2006). 19. On the size of those potential net benefits compared with those from other opportunities that could address the world's most important challenges as conceived by the Copenhagen Consensus project (whose expert panel ranked trade reform as having the second highest payoff among those dozens of opportunities), see http://www.copenhagenconsensus.org, including the prepublication ver- sion of the trade paper by Anderson and Winters (2009). References Anderson, J. E., and J. P. Neary. 2005. Measuring the Restrictiveness of International Trade Policy. Cambridge, MA: MIT Press. Anderson, K. 1995. "Lobbying Incentives and the Pattern of Protection in Rich and Poor Countries." Economic Development and Cultural Change 43 (2): 401­23. ______. 2009. "Krueger/Schiff/Valdés Revisited: Agricultural Price and Trade Policy Reform in Devel- oping Countries Since the 1980s." School of Economics Working Paper 2009-21, University of Adelaide. Forthcoming in Applied Economic Perspectives and Policy. ______, ed. Forthcoming. The Political Economy of Agricultural Price Distortions. Cambridge and New York: Cambridge University Press. Working paper versions of the chapters are available at http://www.worldbank.org/agdistortions. Anderson, K., J. Cockburn, and W. Martin, eds. Forthcoming. Agricultural Price Distortions, Inequality and Poverty. Washington, DC: World Bank. Working paper versions of the chapters are available at http://www.worldbank.org/agdistortions. Anderson, K., and J. Croser. 2009. National and Global Agricultural Trade and Welfare Reduction Indexes, 1955 to 2007. Supplementary database at http://www.worldbank.org/agdistortions. Anderson, K., and Y. Hayami. 1986. The Political Economy of Agricultural Protection: East Asia in Inter- national Perspective. London: Allen and Unwin. Anderson, K., M. Kurzweil, W. Martin, D. Sandri, and E. Valenzuela. 2008a. "Methodology for Measur- ing Distortions to Agricultural Incentives." Agricultural Distortions Working Paper 02, World Bank, Washington, DC. ______. 2008b. "Measuring Distortions to Agricultural Incentives, Revisited." World Trade Review 7 (4): 1­30. Anderson, K., P. J. Lloyd, and D. MacLaren. 2007. "Distortions to Agricultural Incentives in Australia since World War II." The Economic Record 83 (263): 461­82. Anderson, K., R. Lattimore, P. J. Lloyd, and D. MacLaren. 2009. "Australia and New Zealand." Chapter 5 in this volume. Anderson, K., and W. Martin, eds. 2006. Agricultural Trade Reform and the Doha Development Agenda, London: Palgrave Macmillan and Washington, DC: World Bank. ______. 2009a. Distortions to Agricultural Incentives in Asia, Washington, DC: World Bank. ______. 2009b. "China and Southeast Asia." Chapter 9 in this volume. Anderson, K., W. Martin, and E. Valenzuela. Forthcoming. "Long Run Implications of WTO Accession for Agriculture in China." In China's Agricultural Trade: Issues and Prospects, ed. C. Carter and I. Sheldon. St. Paul, MN: International Agricultural Trade Research Consortium. Anderson, K., W. Martin, and D. van der Mensbrugghe. 2006. "Market and Welfare Implications of Doha Reform Scenarios." In Agricultural Trade Reform and the Doha Development Agenda, ed. K. Anderson and W. Martin. London: Palgrave Macmillan and Washington, DC: World Bank. Anderson, K., and W. Masters, eds. 2009. Distortions to Agricultural Incentives in Africa, Washington, DC: World Bank. Five Decades of Distortions to Agricultural Incentives 63 Anderson, K., and J. Swinnen, eds. 2008. Distortions to Agricultural Incentives in Europe's Transition Economies. Washington, DC: World Bank. Anderson, K., and R. Tyers. 1992. "Japanese Rice Policy in the Interwar Period: Some Consequences of Imperial Self Sufficiency." Japan and the World Economy 4 (2): 103­27. Anderson, K., and A. Valdés, eds. 2008. Distortions to Agricultural Incentives in Latin America. Washing- ton, DC: World Bank. Anderson, K., and E. Valenzuela. 2008. Global Estimates of Distortions to Agricultural Incentives, 1955 to 2007. Core database at http://www.worldbank.org/agdistortions. Anderson, K., and L. A. Winters. 2009. "The Challenge of Reducing International Trade and Migration Bar- riers." Policy Research Working Paper 4598, World Bank, Washington DC (also in Global Crises, Global Solutions (2nd edition), ed. B. Lomborg. Cambridge and New York: Cambridge University Press, 2009). Balassa, B., and associates. 1971. The Structure of Protection in Developing Countries. Baltimore: Johns Hopkins University Press. Bates, R. H. 1981. Markets and States in Tropical Africa: The Political Basis of Agricultural Policies. Berkeley: University of California Press. Bhagwati, J. N. 1978. Foreign Trade Regimes and Economic Development: Anatomy and Consequences of Exchange Control Regimes. Cambridge, MA: Ballinger. Chen, S., and M. Ravallion. 2007. "Absolute Poverty Measures for the Developing World, 1981­2004." Policy Research Working Paper 4211, World Bank, Washington, DC. ______. 2008. "The Developing World Is Poorer Than We Thought, But No Less Successful in the Fight Against Poverty." Policy Research Working Paper 4703, World Bank, Washington, DC. Findlay, R., and K. O'Rourke. 2007. Power and Plenty: Trade, War, and the World Economy in the Second Millennium. Princeton, NJ: Princeton University Press. Gardner, B. 2009. "United States and Canada." Chapter 4 in this volume. Haberler, G. 1958. Trends in International Trade: A Report by a Panel of Experts, Geneva: GATT. Herrmann, R., P. Schenck, R. Thiele, and M. Wiebelt. 1992. Discrimination Against Agriculture in Developing Countries? Tubingen: J. C. B. Mohr. Hoekman, B. 2005. "Making the WTO More Supportive of Development." Finance and Development 42: 14­18. Ivanic, M., and W. Martin. 2008. "Implications of Higher Global Food Prices for Poverty in Low- Income Countries." Agricultural Economics 39: 405­16. Johnson, D. G. 1991. World Agriculture in Disarray. London: St Martin's Press. Johnson, H. 1989. The Story of Wine. London: Mitchell Beasley. Kindleberger, C. P. 1975. "The Rise of Free Trade in Western Europe, 1820­1875." Journal of Economic History 35 (1): 20­55. Krueger, A. O. 1978. Foreign Trade Regimes and Economic Development: Liberalization Attempts and Consequences. Cambridge, MA: Ballinger. Krueger, A. O., M. Schiff, and A. Valdés. 1988. "Agricultural Incentives in Developing Countries: Measuring the Effect of Sectoral and Economy-wide Policies." World Bank Economic Review 2 (3): 255­72. ______. 1991. The Political Economy of Agricultural Pricing Policy, Volume 1: Latin America, Volume 2: Asia, and Volume 3: Africa and the Mediterranean. Baltimore: Johns Hopkins University Press for the World Bank. Lerner, A. 1936. "The Symmetry Between Import and Export Taxes." Economica 3 (11): 306­13. Lindert, P. 1991. "Historical Patterns of Agricultural Protection." In Agriculture and the State, ed. P. Timmer. Ithaca, NY: Cornell University Press. Little, I. M. D., T. Scitovsky, and M. Scott. 1970. Industry and Trade in Some Developing Countries: A Comparative Study. London: Oxford University Press. Lloyd, P. J., J. Croser, and K. Anderson. 2009a. "Welfare- and Trade-Based Indicators of National Dis- tortions." Chapter 11 in this volume. ______. 2009b. "Global Distortions to Key Commodity Markets." Chapter 12 in this volume. Morgan, D. 1979. Merchants of Grain. New York: Viking Press. North, D. C., J. J. Wallis, and B. R. Weingast. 2009. Violence and Social Orders: A Conceptual Framework for Interpreting Recorded Human History. Cambridge and New York: Cambridge University Press. 64 Distortions to Agricultural Incentives: A Global Perspective Nye, J. V. C. 2007. War, Wine, and Taxes: The Political Economy of Anglo-French Trade, 1689­1900. Princeton, NJ: Princeton University Press. Orden, D., F. Cheng, H. Nguyen, U. Grote, M. Thomas, K. Mullen, and D. Sun. 2007. "Agricultural Pro- ducer Support Estimates for Developing Countries: Measurement Issues and Evidence from India, Indonesia, China and Vietnam." IFPRI Research Report 152, International Food Policy Research Institute, Washington DC. Orden, D., and E. Diaz-Bonilla. 2006. "Holograms and Ghosts: New and Old Ideas for Reforming Agri- cultural Policies." In Agricultural Trade Reform and the Doha Development Agenda, ed. K. Anderson and W. Martin. London: Palgrave Macmillan and Washington, DC: World Bank. OECD (Organisation for Economic Co-operation and Development). 2006. Agricultural Policies, Mar- kets and Trade in the Central and Eastern European Countries and the New Independent States: Mon- itoring and Outlook. Paris: OECD. ______. 2008a. Producer and Consumer Support Estimates, OECD Database, 1986­2007. http://www.oecd.org (online database for 1986­2007 estimates; OECD files for estimates using an earlier methodology for 1979­85). ______. 2008b. The Role of Farm Households and the Agro-Food Sector in the Economy of Rural Areas: Evidence and Policy Implications, Paris: OECD. Otsuka, K., J. P. Estudillo, and Y. Sawada, eds. 2009. Rural Poverty and Income Dynamics in Asia and Africa. London and New York: Routledge. Otsuka, K., and T. Yamano. 2006. "Introduction to the Special Issue on the Role of Nonfarm Income in Poverty Reduction: Evidence from Asia and East Africa." Agricultural Economics 35 (supplement): 373­97. Ravallion, M., and S. Chen. 2007. "China's (Uneven) Progress Against Poverty." Journal of Development Economics 82: 1­42. Ravallion, M., S. Chen, and P. Sangraula. 2007. "New Evidence on the Urbanization of Poverty." Policy Research Working Paper 4199, World Bank, Washington, DC. Ravallion, M., and G. Datt. 1996. "How Important to India's Poor is the Sectoral Composition of Eco- nomic Growth?" World Bank Economic Review 10 (1): 1­26. Ricardo, D. 1817. On the Principles of Political Economy and Taxation. London: John Murray. Sandri, D., E. Valenzuela, and K. Anderson. 2007. "Economic and Trade Indicators, 1960 to 2004." Agricultural Distortions Working Paper 02, World Bank, Washington, DC. http://www .worldbank.org/agdistortions. Stigler, G. J. 1975. The Citizen and the State: Essays on Regulation. Chicago: University of Chicago Press. Thiele, R. 2004. "The Bias Against Agriculture in Sub-Saharan Africa: Has It Survived 20 Years of Struc- tural Adjustment Programs?" Quarterly Journal of International Agriculture 42 (1): 5­20. Thompson, Sir William (Baron Kelvin of Largs). 1889. Popular Lectures and Addresses, Volume I. London: Macmillan. Thompson, W., J. Dewbre, and R. Martini. 2007. "Is OECD Farm Support Less Decoupled Than It Use to Be?" Paper presented at the Annual Meetings of the American Agricultural Economics Associa- tion, Portland, OR, 29­30 July. Tyers, R., and K. Anderson. 1992. Disarray in World Food Markets: A Quantitative Assessment. Cam- bridge and New York: Cambridge University Press. Valdés, A. 1996. "Surveillance of Agricultural Price and Trade Policy in Latin America During Major Policy Reforms." World Bank Discussion Paper 349, World Bank, Washington, DC. Valdés, A., ed. 2000. "Agricultural Support Policies in Transition Economies." World Bank Technical Paper 470, World Bank, Washington, DC. Valenzuela, E., D. van der Mensbrugghe, and K. Anderson. 2009. "General Equilibrium Effects of Price Distortions on Global Markets, Farm Incomes, and Welfare." Chapter 13 in this volume. Williamson, J. 2008. "Globalization and the Great Divergence: Terms of Trade Booms and Volatility in the Poor Periphery, 1782 to 1913." NBER Working Paper 13841, Cambridge, MA. World Bank. 2007. World Development Indicators. Washington, DC: World Bank. WTO, ITC, and UNCTAD (World Trade Organization, International Trade Commission, and United Nations Conference on Trade and Development). 2007. Tariff Profiles 2006. Geneva: WTO. Part II EVOLUTION OF DISTORTIONS IN ADVANCED ECONOMIES 2 Japan, republic of Korea, and Taiwan, China Masayoshi Honma and Yujiro Hayami* The story of agricultural policy in Northeast Asia over the past 50 years illustrates the dramatic changes that can occur in distortions to agricultural incentives faced by producers and consumers at different stages of economic development. In this study of Japan, the Republic of Korea, and the island of Taiwan, China, the degree of distortions for key agricultural products and for the agricultural sector as a whole over a period when these economies transitioned from low- or middle- income to high-income status (the period from 1955 to 2004 plus, in the case of Japan, some pre-World War II experience) is estimated, a period that encompasses the so-called East Asian economic miracle of dramatic industrial development. Theodore Schultz (1978) established that as economies advance from low- to high-income status, agricultural policies tend to shift from taxation of to subsi- dization of agriculture. In this regard, Japan, Korea, and Taiwan, China are clear examples. This chapter compares the policy evolution in these economies and provides information on the effect of policies and underlying economic condi- tions on changes in agricultural distortions. Our findings shed light on how agri- cultural distortions may change over different stages of economic development in other countries. To begin, a succinct summary of core characteristics of the region in terms of the nature of the three economies--including their resource endowments, which * The authors are grateful for helpful comments from workshop participants, for invaluable help with data compila- tion by Johanna Croser, Esteban Jara, Marianne Kurzweil, Signe Nelgen, Francesca de Nicola, Damiano Sandri, and Ernesto Valenzuela. The working paper version of this chapter (Honma and Hayami 2008) contains additional back- ground material and appendix tables. This chapter draws on the introductory and country chapters in Anderson and Martin (2009), with data updated using Anderson and Valenzuela (2008). 67 68 Distortions to Agricultural Incentives: A Global Perspective to some degree determined the course of their modern economic growth and development--is given. The evolution of agricultural policies in the three economies is then reviewed, followed by a discussion of how to measure distor- tions to agricultural incentives using the methodology from Anderson et al. (2008a, 2008b), the focus of which is on nominal and relative rates of assistance (NRAs and RRAs). Implications of empirical findings of this study for policy reforms in the three economies are discussed in the final section, where lessons for other economies experiencing similar structural transformations in the course of their economic growth are also identified. The study finds that significant growth in agricultural protection began when Japan, Korea, and Taiwan, China entered the middle-income stage of economic development. Statistical observations are found to be consistent with the hypoth- esis that the success of rapid industrialization that advanced these economies beyond the middle-income stage resulted in a decline in agriculture's comparative advantage associated with the growing income disparity between farmers and employees in nonagricultural sectors. In several cases, demand from farmers for a reduction of farm-nonfarm income disparity materialized in the form of increased assistance to agriculture. This was manifest predominantly through rapid and sustained growth in border protection of agricultural products. Economic Development and Structural Change A government's choice of agricultural policies, particularly price-distorting poli- cies, is closely related to the process of economic development. As identified by Schultz (1978), there are two agricultural problems. One is the "food problem," which underlies policies commonly adopted in low-income countries that exploit or tax agriculture. Those policies contrast with the policies that protect or subsidize agriculture in many high-income countries seeking to solve "the farm problem." Schultz's hypothesis became an established paradigm among agricultural econo- mists, who supported it in several empirical studies (Anderson and Hayami 1986; Hayami 1988; Krueger, Schiff, and Valdés 1991). More recently, Hayami (2005) and Hayami and Godo (2004) have added the "disparity problem" as specific to mid- dle-income economies. They suggest it is important to study how distortions in agricultural incentives change as economies move through different stages of development. The most distinguishing characteristic of Japan, Korea, and Taiwan, China dur- ing the period examined here is their unusually rapid rates of economic growth and industrial development. In describing the so-called East Asian miracle, the World Bank (1993) depicted Japan as the front runner along the development spectrum, with Korea and Taiwan, China, together with Hong Kong, China and Japan, Republic of Korea, and Taiwan, China 69 Singapore, as the second group.1 Mainland China and the various countries of the Association of Southeast Asian Nations (ASEAN)2 are following behind, though with rapid rates of economic growth. Since the changing nature of distortions of agricultural incentives seems to be closely related to rate of economic growth and structural change, an overview is provided of the development of the three economies examined here and associated changes in economic and agricultural structures. Initial conditions and development strategies Weather and topographical conditions in Northeast Asia are characterized by reg- ular monsoon rain, together with mountainous, undulating land in which irriga- tion water is controlled relatively easily with efforts at the family and community levels. This makes the region well suited to rice production by small family farms organized into village communities. Indeed, the traditional agrarian structure involved smallholder farms of an average size of about one hectare that were pre- dominantly dependent on rice cultivation. It is important to recognize that, unlike in Southeast Asia, large agribusiness plantations dependent on hired labor were almost completely absent not only in Japan and Korea (temperate-zone coun- tries), but also in Taiwan, China, where tropical cash crops such as sugar and bananas comprised a significant share of agriculture. Prior to the land reforms that followed World War II, the rural community was stratified across landlords, land-owning cultivators, and landless tenants. Agricultural laborers subsisting on hired-labor wages were not a significant component of the rural population. Historically, there is a high degree of similarity in the agrarian structures of Japan, Korea, and Taiwan, China, due in part to the fact that Japan brought its institutions to its colonies--Taiwan, China starting in 1895 and Korea starting in 1910. In all of these places, the most fundamental land-related institution was the fee simple title granted to land owners through cadastral surveys, in return for their payment of land tax. Japanese efforts to develop the colonies concentrated on agriculture, and particularly on rice after Japan experienced a supply shortage after the Rice Riots in 1918. The promotion of rice production through agricul- tural research and extension systems, irrigation and drainage infrastructure, and import protection (see Anderson and Tyers 1992) was considered a major success from Japan's viewpoint in that rice imports from the two colonies increased from 5 to 20 percent of consumption in Japan between 1915 and 1935. The increased export of rice and other primary commodities and the correspon- ding inflow of manufactured commodities meant dependency of the Korean and Taiwanese economies on agriculture remained high. This tendency was especially pronounced in the southern part of Korea, as Japanese industrial development 70 Distortions to Agricultural Incentives: A Global Perspective efforts on the Korean Peninsula were concentrated in the north: the hydroelectric power of the Yalu River fed a complex of chemical industries larger than those that existed in Japan at the time. The heavy dependency on agriculture in the southern part of Korea was furthered by urban destruction during the Korean War (1950­ 53). Relative to what is now the Republic of Korea, Taiwan, China had more active commerce and industry, largely because the dominant cash-crop sector there required significant amounts of processing and marketing activities relative to subsistence crops such as rice and barley that were produced in Korea. In fact, the situation in Taiwan, China under colonial rule was somewhat akin to Japan between the mid-19th to the early 20th centuries, when commercial treaties imposed by Western countries deprived Japan of tariff autonomy. Consequently, Japan specialized in labor-intensive manufactured products based on farm- supplied materials such as silk reeling, tea processing, and cotton weaving. This scenario accorded with comparative advantage under virtual free trade, leading to a wide dispersion of small and medium industries in rural areas in Japan. The predominance of small-scale industry in Japan and Taiwan, China contrasts with the concentration of Korean industry in large-scale establishments in urban areas. The reasons Japan, Korea, and Taiwan, China came out as the forerunners in the East Asian miracle are numerous.3 Here, suffice it to say that their success was due to pertinent borrowing of technology from advanced economies. Gerschenkron (1962), for example, suggests that the later industrialization of an economy begins, the larger the scope for economic growth through technology borrowing. The question remains, however, as to why Japan, Korea, and Taiwan, China were particularly successful in technology borrowing among the many countries that began late starters. One commonly cited speculation is the fact that these three resource-poor countries were heavily endowed with cheap but relatively well-educated labor, a situation that made initial borrowing of labor- intensive technologies fairly efficient and smoothed the way for borrowing of capital- and knowledge-intensive technologies. Another reason is the great crises faced by these economies--Japan's defeat in World War II, the Korean War, and Taiwan, China's loss of the mainland to the communists--all of which compelled leaders to adopt policies that would lead to economic success for the sake of main- taining their legitimacy, instead of indulging in rent-seeking activities (Hayami and Godo 2005). Despite much similarity, there were also significant differences in the industri- alization strategy adopted by the three, especially between Korea and Taiwan, China. In Japan, despite policies aimed at promoting the development of capital- intensive industries after the recovery of tariff autonomy in 1911, small- and Japan, Republic of Korea, and Taiwan, China 71 medium-sized industries survived as a major component of the industrial sector, and many of them located in rural areas. In Taiwan, China, although the Nation- alist Party tightly controlled formal sectors, there was little government interven- tion in the activities of small- and medium-sized companies, which were able to grow through various marketing and financial linkages among themselves and with foreign firms. They became very internationally competitive (Ho 1979, 1982). In contrast, government control in Korea was stronger and more wide- ranging, especially under the military administration of Pak Chong-hui (1961­ 79). All formal credit issuance was channeled from nationalized banks to large industry, while foreign direct investment was tightly controlled (Cole and Park 1983; Amsden 1989). This strategy underlies the high concentration of industrial production in a small number of large enterprises in Korea. Economic growth and structural transformation Quantitative indicators of economic development in Japan, Korea, and Taiwan, China over the past five decades show similarities and differences, as summarized in table 2.1. The first three rows indicate real gross domestic product (GDP) per capita in constant 2000 prices at purchasing power parity (PPP), taken from Heston, Summers, and Aten (2006). In 1955, Japan's GDP per capita was more than US$3,000, whereas it was less than US$1,500 in Taiwan, China. Japan's strong economic growth thereafter pushed the figure to US$4,500 in 1960 and US$5,000 in 1961. In 1960, Korea and Taiwan, China had per capita GDP averages of around US$1,500; by 1978 in Taiwan, China, and 1983 in Korea, both countries were at US$5,000. Japan was the first, in 1970, to surpass US$10,000, a level that Taiwan, China and Korea reached in 1988 and 1991, respectively. Roughly speaking, in terms of growth of per capita GDP, Japan was ahead of Taiwan, China by about two decades and Taiwan, China was ahead of Korea by about half a decade, although these margins decreased over time. Though the criteria of classification are not universal, it is convenient to classify economic development in the three economies into four stages:4 · Low income (US$1,500 or less): pre-1950 for Japan and pre-1960 for Korea and Taiwan, China; · Lower middle income (US$1,500­5,000): 1950­60 for Japan and 1960­80 for Korea and Taiwan, China; · Upper middle income (US$5,000­10,000): 1960­70 for Japan and 1980­90 for Korea and Taiwan, China; and · High income (US$10,000 or more): post-1970 for Japan and post-1990 for Korea and Taiwan, China. 72 Distortions to Agricultural Incentives: A Global Perspective Table 2.1. Economic Growth and Structural Transformation in Japan, Korea, and Taiwan, China, 1955­2004 1955 1960 1970 1980 1990 2000 Real GDP per capita, Japan 3,128 4,509 11,391 15,520 21,703 23,971 constant 2000 $a Korea 1,429 1,458 2,552 4,497 9,593 15,702 Taiwan, China 1,241 1,444 2,846 5,963 11,248 19,184 Share of agriculture Japan 17.4 9.0 4.2 2.4 1.7 1.1 in GDP (%) Korea 46.9 39.1 29.2 16.2 8.9 4.9 Taiwan, China 28.9 28.2 15.3 7.5 4.0 2.0 Share of agricultural Japan 33.8 26.8 15.9 9.1 6.2 4.5 workers in Korea 79.7 60.2 49.1 37.1 18.1 10.0 economically active Taiwan, population (%)a China 53.6 50.2 36.7 19.5 12.8 8.9 Share of farm Japan 40.7 36.5 25.1 18.3 14 8.2 household Korea 61.9 58.2 44.7 28.4 15.5 8.6 population in total Taiwan, population (%) China 50.7 49.8 40.9 30.3 21.1 16.5 Agricultural GDP per Japan 51.5 33.6 26.4 26.4 27.4 24.4 worker/total GDP per Korea 58.8 65.0 59.5 43.7 49.2 49.0 worker (%) Taiwan, China 53.9 56.2 41.7 38.5 31.3 22.5 Sources: Heston, Summers, and Aten (2006); Japanese Ministry for Agriculture, Farming and Fisheries (JMAFF), Nogyo Hakusho Fuzoku Tokei-hyo (Statistical Appendix of Agricultural White Paper), various issues; government of Korea, Major Statistics of Agriculture, Forestry and Fisheries, various issues; government of Taiwan, China, Taiwan Agricultural Year Book, various issues. a. Shares of agriculture in GDP and labor force include forestry and fisheries. Changes in other indicators in table 2.1 are closely related to changes in per capita real GDP over the four stages. The GDP share of agriculture in Japan in 1955--the lower-middle-income stage--was 17 percent, substantially less than the share during the low-income stage for Korea and Taiwan, China (47 and 29 percent, respectively). By 1970 in Taiwan, China and 1980 in Korea, the shares had declined to similar levels to that of Japan at the lower-middle-income stage-- 15 percent and 16 percent, respectively. By the time Japan entered the high- income stage in 1970, its agricultural share of GDP was 4 percent, about the same as that of Korea and Taiwan, China in the 1990s, when they entered the high- income stage. Although Korea and Taiwan, China experienced similar changes in per capita real GDP over the four stages, significant differences can be observed in their economic structures. The GDP share of agriculture in 1955 in Korea was nearly 50 percent, whereas it was less than 30 percent in Taiwan, China. Similar differences Japan, Republic of Korea, and Taiwan, China 73 can be observed with respect to agriculture's share of the labor force, which was as high as 80 percent in Korea, versus less than 55 percent in Taiwan, China. In both countries, the share of agriculture in GDP declined significantly over time--to 3.8 percent in Korea and 1.7 percent in Taiwan, China by 2004--although the share in Korea was still double that of Taiwan, China at that point. Historical differences can be observed in agriculture's share of the labor force in Korea and Taiwan, China. Interestingly, however, the difference in the labor force share of agriculture disappeared by 2004, with about 7.7 percent of Korean employment in agriculture versus 7.5 percent in Taiwan, China. The relatively faster decline in the labor force share of agriculture in GDP in Korea reflects its concentration of industries in urban areas. In Taiwan, China, characterized by the wide dispersion of industries in rural areas, farmers increased their incomes from off-farm employment while continuing to be classified as farmers. In contrast, a much greater percentage of rural people in Korea had to quit farming and migrate to urban areas to obtain nonfarm employment. These differences are reflected in the much faster decrease in the farm household population as a percentage of the total population in Korea relative to Japan and Taiwan, China. The last rows in table 2.1 report the ratios of agricultural GDP per worker to total GDP per worker. This can be considered an indicator of the relative labor productivity of agriculture to total labor productivity of the whole economy. It may also be regarded as an indicator of the income gap between the agricultural sector and the whole economy. While the relative labor productivity of agricul- ture in nominal terms was not substantially different among the three economies in 1955, in Japan it declined sharply, from 52 percent in 1955 to nearly 25 percent in 1970 and thereafter. The ratio did not begin to decline in Korea and Taiwan, China until 1960. In Korea, it reached 44 percent in 1980 (from 65 percent in 1960) and stayed nearly constant thereafter. In Taiwan, China, the ratio contin- ued to decline, to 23 percent in 2000--a rate lower than that of Japan, reflecting the abundance of nonfarm employment opportunities for farmers in Taiwan, China. These measures should be interpreted with great care, however. At first glance, a faster decline in this ratio of agricultural labor productivity to total labor pro- ductivity in Taiwan, China relative to Korea appears to indicate more rapid growth in agricultural labor productivity in Korea. In fact, however, the faster decline ratio in Korea is caused by more rapid out-migration of farm labor to urban occupations in Korea relative to Taiwan, China. Thus, growth in the labor productivity of farmers engaging in agricultural activities relative to that of other workers would not have been slower and could have been even faster in Taiwan, China if the ratio was calculated using output per hour of labor instead of output per worker according to the official sectoral labor force classification. 74 Distortions to Agricultural Incentives: A Global Perspective Changes in the structure of agriculture How did the structure of agriculture in Japan, Korea, and Taiwan, China change in the course of their economic development? As of 2004, Japan's 2.9 million farm households accounted for 2.6 million workers engaged mainly in agricultural activities, an average of less than one person per household. Japan's Agricultural Census defines a farm household as one that either operates 0.1 hectare or more of farmland or produces annual sales of agricultural products of at least 150,000 yen (US$1,250 at the exchange rate of 120 yen/US$1). Thus, very small units of farm operation in which no full-time worker engages in farm production are classified as farm households. Indeed, full-time farm households having no family member engaged in agricultural employment accounted for 15 percent of total farm households in 2004. However, noncommercial farm households that operate less than 0.3 hectares of farmland or have annual sales of less than 500,000 yen accounted for 26 percent of total farm households. Moreover, part-time farm households whose income from nonagricultural sources exceeds agricultural income accounted for half of total farm households. Though the number of agricultural workers in Japan declined from 14 million in 1955 to 2.6 million in 2004, the number of farm households declined only from 6 million in 1955 to 2.9 million in 2004. Slow decreases in the number of farm households, together with decreases in agricultural land (from 6.1 to 4.7 million hectares between 1955 and 2004), resulted in a small increase in arable land per farm, from 1.01 hectares in 1955 to 1.61 hectares in 2004 (table 2.2). The average area of agricultural land per farm in Japan is very small by global standards-- Europe's are 20 to 45 times larger, and those of the United States are 125 times larger. The slow growth of small-sized operations has been a key constraint to the growth of agricultural productivity, resulting in a continual decline in the com- parative advantage of Japanese agriculture, particularly of land-intensive activi- ties, in the course of rapid industrial development.5 Table 2.2 shows that in 2004, 2.9 million farm households in Japan contained 9.4 million people, with an average family size of 3.2 persons. In the same year, 1.2 million farm households in Korea represented 3.4 million people and 1.8 mil- lion hectares of arable land, with average family and farm sizes of 2.8 persons and 1.5 hectares, respectively. In Taiwan, China, 0.72 million farm households con- tained 3.2 million people and 0.84 million hectares of arable land, while average family and farm sizes were 4.5 persons and 1.2 hectares, respectively. It is notable that the number of people in farm households in Korea declined at a much faster rate than in Japan and Taiwan, China as a result of faster decreases in both the number of farm households and the number of persons per household in Korea. These observations reflect the scarcity of nonfarm employment opportunities in Japan, Republic of Korea, and Taiwan, China 75 Table 2.2. Changes in Agricultural Structure in Japan, Korea, and Taiwan, China, 1955­2004 1955 1960 1970 1980 1990 2000 2004 Number of farm Japan 6,043 6,057 5,342 4,661 3,835 3,120 2,934 households Korea 2,218 2,350 2,483 2,155 1,768 1,383 1,240 (thousands) Taiwan, China 733 786 880 891 860 721 721 Population Japan 36,347 34,411 26,282 21,366 17,296 10,467 9,400 in farm Korea 13,300 14,559 14,422 10,827 6,661 4,031 3,415 households Taiwan, (thousands) China 4,603 5,373 5,997 5,389 4,289 3,669 3,225 Persons per farm Japan 6.01 5.68 4.92 4.58 4.51 3.35 3.20 household Korea 6.00 6.20 5.81 5.02 3.77 2.91 2.75 Taiwan, China 6.28 6.84 6.81 6.05 4.99 5.09 4.47 Arable land Japan 6,095 6,071 5,796 5,461 5,243 4,830 4,714 (thousands Korea 1,995 2,025 2,298 2,196 2,109 1,918 1,836 of hectares) Taiwan, China 873 869 905 907 890 852 836 Arable land Japan 1.01 1.00 1.08 1.17 1.37 1.55 1.61 per farm Korea 0.90 0.86 0.93 1.02 1.19 1.39 1.48 household Taiwan, (hectares) China 1.19 1.11 1.03 1.02 1.03 1.18 1.16 Share of Japan 70.7 49.5 31.9 17.0 13.8 13.1 14.3 agricultural Korea -- -- 75.8 65.2 56.8 47.2 39.3 income in total Taiwan, farm household China -- -- 48.7 24.8 20.1 17.6 22.0 income (%) Share of rice Japan -- 47.4 37.9 30.0 27.8 25.4 22.8 in value of Korea -- 59.3 37.3 34.1 36.9 32.9 27.6 agricultural Taiwan, production (%) China -- 36.5 25.7 19.8 12.1 9.6 7.1 Sources: JMAFF, Nogyo Hakusho Fuzoku Tokei-hyo (Statistical Appendix of Agricultural White Paper); government of Korea, Major Statistics of Agriculture, Forestry and Fisheries; government of Taiwan, China, Taiwan Agricultural Year Book; various issues. Note: -- not available. Korea's rural areas due to its urban-centered industrialization. Indeed, from 1970 to 2004, the share of agricultural income in the total income of farm households declined from 32 to 14 percent in Japan and 49 to 22 percent in Tai- wan, China, whereas in Korea it was 76 percent in 1970 and 39 percent in 2004. Japan has lost 23 percent of its arable land area over the past 50 years, falling from 6.1 million hectares in 1955 to 4.7 million hectares in 2004. The decrease in arable land under cultivation was thus a significant contributor to changes in farm 76 Distortions to Agricultural Incentives: A Global Perspective size in Japan in terms of arable land per farm household. Meanwhile, the arable land area in Korea deceased from 2 million hectares in 1955 to 1.8 million hectares in 2004, and arable land in Taiwan, China remained almost constant (0.87 million hectares in 1955 and 0.84 million hectares in 2004). Farm-size changes in Korea and Taiwan, China were almost exclusively the result of changes in the number of farm households. In Japan and Korea, average farm sizes increased slowly, from 1.0 and 0.9 hectares in 1955 to 1.6 and 1.5 hectares in 2004, respectively, whereas farm size in Taiwan, China remained almost constant during that period. The faster increase in farm size in Korea relative to Taiwan, China was the result, again, of faster out-migration of farm workers and their families to urban areas. The distinctly urban-centered industrialization of Korea is clearly reflected in its high share of agricultural income in total farm household income. In all three economies, the ratio decreased as off-farm employment for members of farm households increased. In Japan, the ratio decreased from 70 percent in 1955 to 32 percent in 1970 and further to 14 percent in 2004, corresponding to the shift from the lower-middle-income to upper-middle-income and finally high-income stage of development. In Taiwan, China, where the ratio of agricultural income to total farm income was already below 50 percent in 1970 when its economy was in the lower-middle-income stage, it dropped to 22 percent in 2004. In contrast, in Korea the ratio was 76 percent in 1970 and still nearly 40 percent in 2004, higher than in Taiwan, China but also higher than in Japan at comparable development stages. Major differences in the adjustment of agriculture to economic growth based on industrial development are also reflected in changes in the commod- ity mix of farm production. Traditionally, rice was the most important crop in all three economies, but its significance declined as per capita income increased. Changes in the relative importance of rice were different. From 1960 to 2004, the share of rice in the total value of agricultural production declined from 47 to 23 percent in Japan and from 59 to 28 percent in Korea. In contrast, the share of rice in Taiwan, China, which was less to begin with, at 37 percent, decreased rapidly, to 7 percent in 2004. The contrast reflects the fact that the agricultural sector of Taiwan, China traditionally depended less on rice because of both its greater opportunity to grow cash crops and its success in achieving greater agricultural diversification toward high-valued commodities such as vegetables, fruits, poultry, and pig meat more efficiently than Japan and Korea in response to the shift in domestic demand for more income-elastic commodities. Japan, Republic of Korea, and Taiwan, China 77 Evolution of Agricultural Policy after World War II Japan6 In the decade following World War II, Japan worked hard to recover from the devastation of the war. The primary emphasis of agricultural policy was on increasing domestic food production and delivering food equitably at low costs to consumers. To this end, the government invested heavily in agricultural research and extension and land infrastructure, while upholding many of the rigid controls on rice procurement from farmers and delivery to consumers established under the Food Control Laws enacted during the war. Immediately following World War II, land reform was carried out in accor- dance with the strong recommendations of occupying authorities. The urgent need to increase agricultural production through increased production incentives to cultivators was sufficiently strong to overcome the opposition of landlords to strengthening the rights of tenants through government control of rents and land prices. During the four years from 1947 to 1950, the government purchased 1.7 million hectares of farmland from landlords and transferred 1.9 million hectares, including state-owned land, to tenant farmers, which amounted to about 80 percent of the land under tenancy before the land reform. Although land reform resulted in a considerable change in the distribution of land ownership, the size distribution of operational holdings remained basically the same. As a result, the traditional agrarian structure of Japan, characterized by small-scale family farms with an average size of about l hectare, remained intact, despite the rise and the fall of landlordism (table 2.2). There is no doubt, how- ever, that the land reform promoted more equal asset and income distribution among farmers, and hence contributed to social stability in Japan's rural sector. But the fact that small-scale family farms continued to be the basic unit of agricultural production meant that land reform did not induce changes in the basic direction of technological developments. Though land reform contributed to an increase in standards of living and consumption levels, its contribution to capital formation and productivity growth in agriculture were not significant (Kawano 1969). As Japan set off on its "miraculous" economic growth path following the end of the Korean War in the 1950s, agriculture began to face serious adjustment prob- lems. The rate of growth in agricultural productivity, rapid by international stan- dards, was not rapid enough to keep up with growth in the industrial sector, and intersectoral terms of trade did not improve during the 1950s. This was partly because of the pressure of surplus agricultural commodities in the United States and other exporting countries, and partly because domestic demand for major 78 Distortions to Agricultural Incentives: A Global Perspective staple cereals (especially rice) approached saturation after the bumper crop of 1955. As a consequence, incomes and living standards of farm households lagged behind those of urban households during the 1950s. In 1961, real GDP per capita in Japan exceeded US$5,000 pushing Japan into the upper-middle-income stage of economic development. Correspondingly, the major goal of agricultural policy shifted from increased production of food sta- ples to reducing the rural-urban income gap. The need to assist farmers increased during the 1960s, as the rural-urban income gap progressively widened and the out-migration of agricultural labor accelerated. The difficulty of structural adjust- ment in agriculture as a result of the rapidly growing economy led to the enact- ment in 1961 of the Agricultural Basic Law, a national charter for agriculture. The law declared that it was the government's responsibility to raise agricultural pro- ductivity and thereby to close the gap in income and welfare between farm and nonfarm households. In order to raise agricultural productivity and improve farming efficiency, it was considered essential to increase the scale of farm operation by eliminating inefficient farm units and by promoting cooperative operations among remaining farms. Despite such efforts at structural adjustment, the rate of agricultural pro- ductivity growth was not increased sufficiently to prevent the rural-urban income gap from widening further. The Food Control System, which was originally designed to provide food security to consumers, thus became the chief instrument to protect farmers. Under the system, based on the 1942 Food Control Law, most food items were placed under direct government control. However, as the Japanese economy recovered from World War II, the number of items under control was reduced so that only rice remained under direct control after 1952. Initially, the whole marketing process of rice from producers to consumers was under direct control of the Japanese Food Agency and prices were regulated from the farmgate to the retail level, although the regulations were gradually relaxed. After Japan entered the upper-middle-income stage of development in the 1960s, the Food Control System became a powerful instrument for rice farmers, and they organized political lobbying to raise rice prices for government pur- chases. Their pressure resulted in a rice price determination formula in 1960 called the production cost and income compensation formula, established under the Food Control System. This formula was designed to reduce the gap between farm and nonfarm income and wages by raising rice prices. The goal appears to have been achieved: income per agricultural worker compared to income per worker in manufacturing improved after 1960 following a rapid rise in agricul- tural relative to manufacturing prices. The increase in the price of rice, which con- stituted about 40 percent of the total value of agricultural output before 1970, was a major factor in improving the domestic terms of trade for agriculture. The rise Japan, Republic of Korea, and Taiwan, China 79 in agricultural prices, together with increases in off-farm income, resulted in a marked reduction in the income per capita gap between agricultural and nonagri- cultural households. Protecting rice farmers through a price policy was possible in Japan because rice trade was completely controlled by the state-trading system. During the 1960s, the price of rice was raised not only far above the world price but also above the market equilibrium price under autarky. As an upper-middle-income country, Japan was able to let consumers and taxpayers shoulder the costs of agricultural protection. However, there was a limit on increasing agricultural protection through price policy. The high protected prices of rice resulted in an expansion of rice produc- tion in excess of consumption. The accumulated surplus of rice in government storage forced authorities to introduce controls on rice acreage in 1969 (they are still in place today). Further, the dramatic increase in income and wages of indus- trial workers after 1960 meant that the diet of Japanese people changed. Average consumption per industrial employee (deflated by the consumer price index) doubled over 1955­70, and again in the decade-and-a-half following that. Corre- spondingly, rice was no longer a major wage good for industrial workers. To cope with the increasing rice surplus, the Food Control System was revised. The direct control on rice distribution was relaxed by introducing nongovernment distribu- tion channels. Finally, in 1995, the Food Control Law was replaced by the Staple Food Law, whereby the role of government was limited to stock holding opera- tions for food security, although state trading of rice is maintained for interna- tional trade. Real GDP per capita in Japan exceeded US$10,000 in 1969, putting it over the high-income threshold. Demand for agricultural protection from the farm bloc increased. Japan's comparative advantage, though, continued to shift from agricul- ture to industry, while internal resistance to protectionism declined because the nonfarm population became affluent and, hence less resistant to shouldering the cost of agricultural protection in the form of high food prices or subsidies to farm producers. However, though internal resistance weakened, external pressure for liberalization of agricultural imports increased. Following the replacement of the General Agreement on Tariffs and Trade (GATT) with the World Trade Organization (WTO) in 1995, Japan was required to reform domestic agricultural policy. Under the WTO's Uruguay Round Agree- ment on Agriculture (URAA), in particular, Japan had to adjust agricultural poli- cies to be more consistent with the globalization of the economy. In 1995, Japan converted nontariff border measures to tariffs for 28 commodities. At the beginning of implementation, rice was exempted from tariffication in compensation of larger minimum access imports of rice, namely 4 percent of domestic consumption in 80 Distortions to Agricultural Incentives: A Global Perspective 1995, rising to 8 percent by 2000. In 1999, Japan adopted tarriffication for rice, leaving the minimum access imports at 7.2 percent of domestic consumption. The Basic Law on Food, Agriculture and Rural Areas, also enacted in 1999, was a replacement for the 1961 Agricultural Basic Law. Four years earlier, the Food Control Law had been abolished to liberalize the domestic rice market. The 1999 Basic Law obliged the government to draft a Basic Plan for Food, Agriculture and Rural Areas for the promotion of the comprehensive and systematic implementa- tion of policies on food, agriculture, and rural areas. The plan is intended to be redrafted every five years. Under the current Basic Plan, released in 2005, a key point of the new agricultural policy is to target government assistance to farmers who satisfy certain conditions, especially on minimum farm size. That is, it com- pels farmers who want to continue farming under government assistance to expand the size of their farm operation. Korea7 Before 1960, Korea was a low-income country, with per capita income below US$1,500. In the years just before 1960, the economy was struggling to recover from the Korean War. The agricultural policy adopted at this stage aimed to main- tain low domestic consumer prices for staple foods, notably rice and barley, as well as for fertilizer. The Grain Management Law, enacted in 1950, gave the govern- ment authority to regulate the price of staple foods. But government control was not very effective during the 1950s since the market share of government- controlled rice was less than 10 percent. The government, which was supposed to purchase grain directly from farmers, was unable to purchase sufficient amounts due to budgetary constraints and an inflation-driven spike in grain prices in the mid-1950s. As a creative solution, schemes to collect rice as an in-kind land tax and to barter fertilizer for rice were initiated. Though the former measure was successful, the latter was not because the implicit price of rice in the barter was lower than the market price. In the end, grain imports from the United States under Public Law 480, which accounted for 8 to 12 percent of total domestic grain during 1956­65, helped the Korean government hold down grain prices. In the 1960s, Korea launched wide-ranging policies to promote industrializa- tion under the development autocracy of Pak Chong-hui. Agricultural policies at this time were designed to keep the price of staple food crops low so as to uphold a low cost of living and stable wage rates for industrial workers, rather than main- taining adequate incomes for farmers. Government purchase prices for staple crops, which were considered necessary for the purpose of increasing industrial profits and capital formation, were below market prices. Over time, the Korean government's price intervention became more intense. The market share of Japan, Republic of Korea, and Taiwan, China 81 government-controlled rice was expanded to 20­25 percent during the 1960s; the increase was used mainly to maintain low domestic prices. These agriculture- taxing policies continued during the beginning of Korea's lower-middle-income stage of economic development. As Korea quickly advanced toward becoming an upper-middle-income coun- try, the direction of agricultural policy gradually moved toward supporting farm- ers. In the early 1970s, the buffer-stock operation for noncereal products was set in motion to counteract price declines. In addition to chemical fertilizers, pesticides and farm machineries were added to the list of subsidized inputs, alleviating the adverse impact of import protection to manufacturers of those inputs for farmers. Through the 1970s, the government's purchase prices for rice and barley were steadily raised with the aim of both increasing food production and reducing the urban-rural income gap. Although the government raised producer prices for sta- ple food grains, it did so without a comparable rise in the market prices of rice and barley in order to prevent the cost of living and the wage rate of industrial workers from rising. Likewise, it assisted livestock producers in part by using import quo- tas rather than tariffs to protect them from import competition, with the rent from those quotas being captured by the producer-managed meat import agency.8 The implementation of the two-price system, however, conflicted with the need to maintain financial and monetary stability. As the difference between the purchase and sale prices of rice and barley widened, the deficit of the grain man- agement fund increased. Since a large portion of this deficit was financed by long- term overdrafts from the Bank of Korea, the policy contributed heavily to infla- tionary pressure. Expansion of the government deficit due to the two-price policy became a serious constraint on the policy. Upon entering the upper-middle-income stage of development in the 1980s, the Korean government moved to reduce both tariff and nontariff protection for manufacturing industries. In contrast, agricultural policies toward protecting farmers were strengthened, and the producer prices of farm products were increased to levels far above border prices by means of quantitative import restric- tions on most agricultural commodities. Significant policy changes followed Korea's transition to a high-income coun- try in the early 1990s. Mostly, these changes were related to the URAA. According to the provisions, Korea's quantitative restrictions were converted to tariffs for all agricultural products except rice. But Korea retained the status of a developing country in the Uruguay Round negotiations, giving it special treatment in imple- menting commitments to reduce border protection. The agricultural products under tariffication were subject to a protection reduction commitment of 24 per- cent, on average, within 10 years, with a minimum cut of 10 percent. Tariff rates of Korean agricultural products were over 60 percent on average. Tariffs on products 82 Distortions to Agricultural Incentives: A Global Perspective considered particularly important in Korea were thus cut by the minimum rate of 10 percent. In addition, imports of many agricultural products were begun under the min- imum market access commitment. This commitment required that for all agricul- tural products, at least 3 percent of consumption must be purchased from over- seas in the first year and the import share must increase annually up to 5 percent of consumption within 10 years. Low tariff rates were applied to the in-quota vol- ume so as to guarantee easy market access from exporting countries. Key agricul- tural products such as rice, barley, oranges, red peppers, garlic, and onions began to be imported under the commitment. Rice, the most important agricultural product in Korea, was temporarily exempted from tariffication, as provided in annex 5.B of the URAA. Under the exception plan, rice was subject to an import quota, beginning with 1 percent of total consumption and gradually increasing to 4 percent in 2004, the final imple- mentation year. If Korean rice had not been exempted from tariffication, Korea would have complied with the standard market access commitment of 3­5 per- cent. The temporary exemption from tariffication expired in 2004, but Korea opted to continue invoking a rice exemption from tariffication for another 10 years, to 2014. Taiwan, China9 In the years after World War II, Taiwan, China suffered from high inflation rates, serious shortages of food and other necessities, and a heavy budget burden for its defense systems. The government gave the highest priority to economic stabiliza- tion, food production increases, and the repair of war damages. To alleviate the intense population pressure on limited land, it decided to grant incentives to farmers. Together with the land reform program implemented between 1949 and 1953, war-damaged irrigation and drainage facilities were repaired, fertilizers and other farm inputs were made available, and farmers' organizations were strengthened. During this recovery stage, the Sino-American Joint Commission on Rural Reconstruction (JCRR) was established in Nanking in 1948. That agency played an important role in the postwar rural reconstruction and development of Taiwan, China. From 1951 to 1965, the United States provided a total of US$1.5 billion in aid. Approximately one-third went to agriculture, which was used to build infrastructure and foster human resources for agriculture. Substantial imports of commodities financed by U.S. aid and increases in domestic production, especially of food, helped relieve demand pressures in Taiwan, China. Japan, Republic of Korea, and Taiwan, China 83 During its low-income stage of economic development (before 1960), agricul- tural policy in Taiwan, China was designed mainly to supply rice at low, stable prices to the nonfarm population. In those days, two important taxes were imposed on farmers: the farm land tax and the hidden rice tax. They were levied by means of compulsory rice purchases and the rice-fertilizer barter system. The compulsory purchase of paddy rice from landowners at official prices was another source of gov- ernment control over rice, such that all rice paddy land was subject to the paddy land tax plus the compulsory procurement of rice. The compulsory procurement was assessed on the basis of tax units determined by land productivity. The differ- ence between the government procurement prices and farmers' market prices con- stituted a hidden tax on paddy landowners, who were mostly farm operators after the implementation of land reform program. Though the hidden tax was gradually reduced as per capita income in Taiwan, China rose, it was not abolished until 1973. The government's rice collection by all of these methods during 1950­70 aver- aged 50 to 60 percent of the total amount of rice produced minus farmers' home consumption. By 1973, this share had declined to 20 percent. In subsequent years, it increased again because of the implementation of the guaranteed rice price policy. The total of this hidden rice tax was larger than the total income tax of Taiwan, China before 1963 and was more than twice the farm land tax before 1961, except in 1954. After 1961, when Taiwan, China moved into the lower-middle-income stage of economic development, the hidden rice tax decreased rapidly: the ratio of the hid- den rice tax to the total income tax was only 8.5 percent in 1971 (Kuo 1975). Agricultural policy geared to exploit agriculture for the sake of supporting industrial development (and military development) largely ended during the 1970s, when the shift to subsidizing agriculture began. This period was when Taiwan, China rapidly expanded its labor-intensive light industries in response to increases in export demand. Because several of these light industries, such as gar- ments and footwear, were located in rural areas, nonfarm incomes became increasingly important to farm households. Farmers in Taiwan, China were able to take advantage of employment in manufacturing without leaving home, and many of them engaged in nonfarm self-employed industrial activities during less- busy farm seasons. The need for farmers to rely on agricultural protection policies was therefore smaller than in Korea. It was 1978 when GDP per capita exceeded US$5,000 in Taiwan, China, push- ing it into the upper-middle-income stage of economic development. Still, to help equalize the income level of farm workers with that of the rapidly expanding industrial sector, the government offered loans and subsidies for promoting farm mechanization, which were designed to raise farmers' labor productivity. The expansion of rice production began to slow down in response to an increased emphasis on livestock and fishery products and high-value export crops. Increases 84 Distortions to Agricultural Incentives: A Global Perspective in industrial employment also drove up the cost of farm labor. Labor productivity in agriculture continued to lag behind that of the industrial sector, and the gap between farm and nonfarm per capita incomes was increased, especially for farm- ers who relied mainly on rice production. The problems faced by Taiwan, China agriculture were similar to those that many other industrial countries experienced at a comparable development stage, especially Japan in the early 1960s and Korea in the late 1970s. Per capita consumption of rice in Taiwan, China fell from 140 kilograms per year in 1968 to 74 kilograms in 1988. Correspondingly, an excessive stock of rice accu- mulated. In order to reduce production, farm extension workers encouraged farm- ers to plant other crops in rice fields, but their efforts were not successful because no economic incentive was provided. A six-year rice crop substitution plan inaugurated in 1984 gave direct subsidies of 1 metric ton of paddy rice per hectare to farmers who shifted their rice fields to corn or sorghum, or 1.5 metric tons of paddy rice per hectare to farmers who shifted to crops other than corn and sorghum. In addition, corn and sorghum were purchased by the government at guaranteed prices. Under the program, rice production declined to 1.84 million metric tons in 1988, 0.9 mil- lion metric tons less than the peak of 1976. In 1988, the paid-in-kind subsidy was changed to a cash payment to improve the efficiency of the program. Taiwan, China entered the high-income stage of development in 1988, when its real GDP per capita exceeded US$10,000. The most important change in agricul- tural policy in the years that followed was related to the economy's accession to the WTO on January 1, 2002. In line with its level of economic development, Taiwan, China agreed to bring its tariff rates to a level between those of Japan and Korea--as such, its average nominal tariff rate was reduced from 20 percent to 14 percent in the year following accession and gradually to 12.9 percent by 2004. The plan exempted 137 items covered under tariff rate quotas (TRQs). Of the 41 products that were under import quota restrictions before WTO accession, 18 were moved to tariffication after accession. Rice received a special exemption, and the remaining 22 items are now governed by the tariff rate quota regime. Similar to the arrangement in Korea, the special treatment of rice in Taiwan, China is based on the rules of annex 5 of the URAA. The quota of rice imports was set in 2002 at 8 percent of the average domestic consumption between 1990 and 1992 (144,720 tons of brown rice). By negotiation, this amount was divided into governmental and private import quotas. The government rice quota (65 percent of rice imports) was subject to the same treatment as rice purchased from local growers. The imported rice cannot be exported for food aid nor can it be used for animal feed. The remaining 35 percent was imported by private firms and allocated on first-come, first-served basis. For both private and government quo- tas, there is a ceiling on the price mark-up of NT$23.26 per kilogram for rice and Japan, Republic of Korea, and Taiwan, China 85 NT$25.59 for rice products when they are sold on the domestic market. If the sale of quota rice is slow, the price mark-up can be cut by NT$3 every two weeks. The mark-up reduction can be continued until all rice is sold. Measurement of Distortions to Agricultural Incentives The main focus of the study presented here is the gap between domestic prices and what they would be under free markets. Since it is not possible to understand the characteristics of agricultural development with a sectoral view alone, the project's methodology not only estimates the effects of direct agricultural policy measures (including any distortions in the foreign exchange market), but also generates estimates of distortions in nonagricultural sectors for comparative evaluation. Specifically, the study computes an NRA for farmers including an adjustment for direct interventions on inputs such as border protection on fertil- izers. It also generates an NRA for nonagricultural tradables, for comparison with that for agricultural tradables, via the calculation of an RRA (see Anderson et al. 2008a, 2008b). For Japan and Korea, an NRA is calculated for rice, wheat, barley, soybeans, beef, pig meat, poultry, eggs, and milk. For Taiwan, China, estimates are calculated for rice, wheat, beef, pig meat, poultry, and eggs. Domestic prices are converted to U.S. dollars using market foreign exchange rates except for 1955­64 in Korea and for 1955­61 in Taiwan, China, for which the shadow exchange rates estimated for Korea by Frank, Kim, and Westphal (1975) and for Taiwan, China by Scott (1979) are used in order to take into account the distortions to the foreign exchange market in early years. Aggregate NRAs on output for each county are calculated using weights based on domestic production of commodities valued at undistorted prices. In addition to the commodities above covered in this study, several other crops are included in the calculation of RRAs for Japan and Korea. These include apples, cabbage, cucumbers, grapes, mandarins, pears, spinach, strawberries, onions, and sugar for Japan and cabbage, red peppers, and garlic for Korea. The estimates for these products come from the Organisation for Economic Co-operation and Development's (OECD's) estimates of producer and consumer support estimates (PSEs and CSEs; see OECD 2008). Data for these crops are available only for years 1986 and later. Distortions of those crops prior to 1986 are assumed to be at the level of 20 percent in Japan and 90 percent in Korea of the NRAs for the available covered products. The percentage of agricultural output covered in this study is between 55 and 70 (valued at undistorted prices), though it is difficult to judge the levels of NRAs for the residual products. It is assumed to be made up of the following share trends (at distorted prices) between 1955 and the present: 50 to 80 percent for 86 Distortions to Agricultural Incentives: A Global Perspective import-competing products and 50 to 20 percent for nontradables in Japan and Korea. Distortions of the residual products are assumed to be zero for nontradables, and the same as that of the 11 (Japan) and 4 (Korea) OECD products for import- competing products. For Taiwan, China, distortions of all the noncovered residual products are assumed to be zero, as most of them are nontradable or exportable. To compute the RRA, the NRA for nonagricultural industries is first estimated. For the latter, weighted tariffs are available only for selected years for Japan, Korea, and Taiwan, China. Data are linearly interpolated for the years for which they are not available. For the early years, tariff rates are estimated as total tariff revenue divided by value of imports. Assuming the exportable industries receive no assis- tance, the weighted average tariff is multiplied by the share of import-competing industries in the value of all nonagricultural tradables. This procedure is likely to underestimate assistance to nonagricultural industries, especially in Korea, where subsidized credits targeting certain industries were the major form of assistance. The estimation results for nominal and relative rates of assistance (NRAs and RRAs) to selected commodities are summarized in five-year averages in tables 2.3 and 2.4 for Japan, Korea, and Taiwan, China.10 Annual movements of the RRA are shown in figure 2.1 to compare protection patterns of Japan, Korea, and Taiwan, China. Japan's RRA was 13 percent in 1955, when the country was still in the lower- middle-income stage of economic development. But it rose quickly as Japan Figure 2.1. RRA to Agricultural Versus Nonagricultural Tradables,a Japan, Korea, and Taiwan, China, 1955­2007 250 200 150 percent 100 50 0 50 100 1955 1959 1963 1967 1971 1975 1979 1983 1987 1991 1995 1999 2003 2007 Japan Taiwan, China Korea Source: Honma and Hayami (2008) a. RRA is defined as 100*[(100 NRAagt) (100 NRAnonagt) 1], where NRAagt and NRAnonagt are the percentage NRAs for the tradables parts of the agricultural and nonagricultural sectors, respectively. Table 2.3. NRAs to Selected Agricultural Products in Japan, Korea, and Taiwan, China, 1955­2007 (percent) a. Japan 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Import-competing products 53.7 66.5 79.9 77.8 110.8 111.8 153.1 149.3 147.3 146.5 124.3 Rice 72.5 91.0 122.9 164.9 210.8 267.2 591.6 656.2 535.4 607.0 362.8 Barley 35.7 38.7 16.5 10.1 63.1 88.6 203.6 141.7 129.0 121.5 197.6 Wheat 36.1 39.1 42.4 25.4 76.2 111.8 170.3 200.4 204.2 128.6 269.5 Beef 27.4 68.3 130.8 106.0 215.3 136.7 208.9 177.0 191.8 149.1 39.3 Pig meat 17.9 59.8 12.4 3.4 2.9 12.6 0.5 5.7 10.3 5.5 138.0 Poultry 33.1 42.8 33.5 36.7 31.1 16.8 17.6 25.1 41.4 74.0 11.7 Eggs 3.0 3.3 5.3 2.8 3.3 1.4 19.9 23.2 33.8 27.6 17.1 Milk 44.2 96.2 162.1 165.4 385.7 211.5 365.2 280.3 238.0 273.2 101.0 Apples -- -- -- -- -- -- 32.0 24.1 27.7 31.4 17.3 Cabbage -- -- -- -- -- -- 10.3 31.1 127.5 177.5 204.6 Cucumbers -- -- -- -- -- -- 57.1 17.4 29.8 43.2 31.1 Grapes -- -- -- -- -- -- 87.2 82.2 117.7 177.4 178.6 Mandarins -- -- -- -- -- -- 21.1 44.8 47.3 32.4 46.4 Pears -- -- -- -- -- -- 35.0 24.0 64.2 157.3 128.7 Spinach -- -- -- -- -- -- 89.2 138.0 236.7 134.4 32.3 Strawberries -- -- -- -- -- -- 11.0 25.1 26.5 16.8 7.2 Onions -- -- -- -- -- -- 55.3 80.8 144.4 284.2 294.9 Soybeans -- -- -- -- -- 410.7 259.4 21.3 42.2 67.2 68.5 Sugar -- -- -- -- -- 229.6 198.3 158.4 159.0 154.7 106.6 Exportables n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Total of covered products 53.7 66.5 79.9 77.8 110.8 111.8 153.1 149.3 147.3 146.5 107.4 from domestic measures 0.4 3.1 10.0 9.2 10.9 9.6 8.0 6.0 4.5 4.5 4.0 from border (import) measures 54.1 63.4 69.9 68.6 99.9 102.1 145.2 143.3 142.8 142.0 102.9 Dispersion of covered productsb 39.4 40.3 69.4 82.2 156.1 142.6 175.3 161.5 136.1 142.5 116.0 87 % coverage (at undistorted prices) 69 65 59 55 55 56 69 68 67 67 76 (Table continues on the following pages.) 88 Table 2.3. NRAs to Selected Agricultural Products in Japan, Korea, and Taiwan, China, 1955­2007 (continued) (percent) b. Republic of Korea 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Import-competing productsa 3.9 4.4 16.6 47.6 73.8 122.8 166.7 201.9 182.9 213.6 116.4 Rice 8.2 7.0 5.4 31.3 59.6 118.4 214.4 265.9 294.3 385.9 213.3 Barley 41.2 83.5 72.3 120.3 101.2 165.9 357.0 524.3 543.0 562.8 275.6 Wheat 43.0 26.7 11.2 0.4 26.5 92.2 144.4 216.0 122.8 135.4 -- Beef 38.8 34.4 64.9 73.9 162.6 163.2 126.2 200.8 159.9 167.8 182.3 Pig meat 15.2 21.7 158.7 204.1 202.9 169.1 124.7 149.3 116.2 134.4 103.1 Poultry 11.8 6.9 131.4 103.5 161.7 94.2 86.6 155.6 171.7 179.2 55.7 Eggs 27.1 24.7 23.0 0.1 7.5 14.9 19.4 28.0 26.6 54.3 31.6 Milk -- -- 173.3 108.8 189.0 179.8 185.2 203.7 140.7 149.8 137.0 Cabbage -- -- -- -- -- -- 30.0 30.0 29.1 27.6 27.0 Peppers -- -- -- -- -- -- 175.0 245.4 145.5 197.0 235.7 Soybeans 13.0 18.8 58.8 80.0 122.2 253.0 361.8 508.2 625.6 757.4 729.2 Garlic -- -- -- -- -- -- 250.3 288.8 213.3 122.6 128.1 Exportablesa n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Total of covered products 3.9 4.4 16.6 47.6 73.8 122.8 166.7 201.9 182.9 213.6 147.3 from domestic measures 0.2 0.4 0.9 4.2 7.1 5.3 5.5 5.9 6.1 5.2 4.4 from border (import) measures 3.7 4.7 15.7 43.4 66.7 117.5 161.2 196.0 176.9 208.5 143.0 Dispersion of covered productsb 34.1 40.5 85.0 82.5 89.0 80.1 114.8 164.2 200.1 225.4 206.0 % coverage (at undistorted prices) 48 57 67 65 65 61 60 57 51 46 55 c. Taiwan, China 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­02 2003­07 Exportablesa 23.5 7.5 5.7 20.7 13.4 35.9 89.5 161.4 167.6 203.1 -- Rice 29.6 6.6 17.9 9.4 7.6 32.5 103.3 161.4 167.6 203.1 -- Pig meatc 8.1 64.0 99.7 98.3 60.6 42.6 64.8 n.a. n.a. n.a. -- Import-competing productsa 33.0 5.3 21.7 26.7 32.5 49.1 55.4 93.6 126.3 160.0 -- Wheat 48.2 36.0 39.4 32.2 57.2 92.3 -- -- -- -- -- Beef 13.7 41.2 28.8 22.0 79.6 77.0 101.3 98.5 82.6 72.8 -- Pig meatc n.a. n.a. n.a. n.a. n.a. n.a. n.a. 107.1 131.3 173.2 -- Poultry 47.5 3.7 21.2 27.1 30.0 63.6 84.6 143.0 228.7 279.5 -- Eggsd n.a. n.a. n.a. n.a. n.a. 0.7 26.8 23.9 17.9 24.7 -- Nontradablesa 0.0 0.0 0.0 0.0 0.0 0.0 n.a. n.a. n.a. n.a. -- Eggsd 0.0 0.0 0.0 0.0 0.0 0.0 n.a. n.a. n.a. n.a. -- Total of covered productsa 23.2 7.2 6.2 20.0 14.0 35.1 76.1 109.5 134.0 167.8 -- Dispersion of covered productsb 33.4 35.3 47.5 40.5 40.5 34.5 56.9 66.1 86.9 106.4 -- % coverage (at undistorted prices) 53 49 49 48 50 42 35 34 35 36 -- Source: Anderson and Valenzuela (2008), based on authors' spreadsheet. Note: -- not available; n.a. not applicable. a. Weighted averages, with weights based on the unassisted value of production. b. Dispersion is a simple five-year average of the annual standard deviation around the weighted mean of NRAs of covered products. c. Pig meat changed trade status in 1989, from import-competing to exportable. The period average reported here corresponds to 1985­88 for the import-competing product, and 1989­94 for the exportable product. d. Eggs are assumed to be a nontradable product with zero distortions prior to 1983. 89 Table 2.4. NRAs to Agricultural Relative to Nonagricultural Industriese in Japan, Korea, and Taiwan, China, 90 1955­2007 (percent) a. Japan 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 NRA, covered products 53.7 66.5 79.9 77.8 110.8 111.8 153.1 149.3 147.3 146.5 107.4 NRA, noncovered products 5.3 6.5 7.7 8.0 12.1 13.0 23.7 26.7 42.0 50.5 18.6 NRA, all agricultural products (excluding NPS) 38.8 45.8 50.4 46.9 65.9 68.3 112.4 110.5 112.8 115.2 70.9 All importables 46.1 55.0 62.1 58.1 81.2 82.1 127.5 124.4 127.6 129.1 124.3 All exportables -- -- -- -- -- -- -- -- -- -- -- All nontradables 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 40.8 Trade bias index (TBI) 0.32 0.35 0.38 0.37 0.45 0.45 0.56 0.55 0.56 0.56 0.55 NRA, non-product-specific (NPS) assistance -- -- -- -- 4.8 4.0 6.4 5.8 6.8 5.1 3.3 Inputs -- -- -- -- 4.8 4.0 6.4 5.8 6.8 5.1 3.3 Other -- -- -- -- 0.0 0.0 0.0 0.0 0.0 0.0 0.0 NRA, all agricultural products (including NPS) 38.8 45.8 50.4 46.9 66.8 72.3 118.8 116.3 119.6 120.4 74.3 NRA, decoupled payments 0.0 0.0 0.0 0.0 3.6 15.1 7.1 2.5 2.7 4.6 5.0 NRA, all agricultural products (including NPS and decoupled assistance) 38.8 45.8 50.4 46.9 70.4 87.4 125.9 118.9 122.3 124.9 79.3 NRA, all agricultral tradables (including NPS) 46.1 55.0 62.1 58.1 87.6 86.1 133.8 130.2 134.4 134.2 127.7 NRA, all nonagricultural tradables 2.5 3.9 3.8 2.8 1.6 1.1 1.3 1.1 0.8 0.7 0.6 RRAd 42.5 49.1 56.2 53.7 84.6 84.0 130.9 127.6 132.4 132.7 126.4 b. Republic of Korea 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 NRA, covered products 3.9 4.4 16.6 47.6 73.8 122.8 166.7 201.9 182.9 213.6 147.3 NRA, noncovered products 1.7 0.2 7.0 15.3 25.3 37.4 64.3 88.0 74.6 71.7 49.3 NRA, all agricultural products (excluding NPS) 3.2 4.0 13.4 35.7 56.3 89.4 126.1 152.8 129.8 137.3 80.6 All importables 3.3 4.9 16.3 46.1 71.8 118.6 159.8 197.6 164.8 171.9 116.4 All exportables -- -- -- -- -- -- -- -- -- -- -- All nontradables 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 TBI 0.03 0.05 0.14 0.32 0.42 0.54 0.62 0.66 0.62 0.63 0.54 NRA, NPS assistance -- -- -- -- 0.6 0.7 2.2 7.1 7.3 4.4 3.9 Inputs -- -- -- -- 0.6 0.7 2.2 7.1 7.3 4.4 3.9 Other -- -- -- -- 0.0 0.0 0.0 0.0 0.0 0.0 0.0 NRA, all agricultural products (including NPS) 3.2 4.0 13.4 35.7 56.4 90.1 128.1 159.8 137.0 141.7 84.4 NRA, decoupled payments 0.0 0.0 0.0 0.0 0.1 0.4 0.5 5.3 2.7 7.2 12.3 NRA, all agricultural products (including NPS and decoupled assistance) 3.2 4.0 13.4 35.7 56.4 90.5 128.6 165.2 139.7 148.8 96.8 NRA, all agricultural tradables (including NPS) 3.3 4.9 16.3 46.1 71.9 119.3 161.7 204.7 171.9 176.3 120.3 NRA, all nonagricultural tradables 45.6 37.1 22.3 11.4 11.7 6.8 5.7 3.3 2.3 1.7 1.5 RRAd 32.6 21.4 4.8 30.5 54.0 105.4 147.8 195.0 165.8 171.6 117.0 (Table continues on the following page.) 91 Table 2.4. NRAs to Agricultural Relative to Nonagricultural Industriese in Japan, Korea, and Taiwan, China, 92 1955­2007 (continued) (percent) c. Taiwan, China 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­02 2003­07 NRA, covered productsa 23.2 7.2 6.2 20.0 14.0 35.1 76.1 109.5 134.0 167.8 -- NRA, noncovered products 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 -- NRA, all agricultural productsa 11.8 3.5 3.0 9.2 7.0 14.6 26.4 37.2 45.5 60.0 -- NRA, NPS assistance -- -- -- -- -- -- -- --. -- -- -- Total agricultural NRA (including NPS)b 11.8 3.5 3.0 9.2 7.0 14.6 26.4 37.2 45.5 60.0 -- TBIc 0.15 0.05 0.02 0.12 0.05 0.15 0.27 0.11 0.02 0.00 -- NRA, all agricultural tradables 15.8 4.7 3.9 12.0 8.9 18.5 32.7 45.0 53.6 69.2 -- NRA, all nonagricultural tradables 8.8 9.3 8.8 7.5 7.0 5.2 4.5 2.6 1.8 1.1 -- RRAd 22.5 4.2 4.5 4.2 1.7 12.7 27.0 41.3 51.0 67.3 -- Source: Anderson and Valenzuela (2008), based on authors' spreadsheet. Note: -- = not available. a. NRAs including product-specific input subsidies. b. NRAs including product-specific input subsidies and non-product-specific (NPS) assistance. Total of assistance to primary factors and intermediate inputs divided by total value of primary agriculture production at undistorted prices (percent). c. TBI (1 NRAagx 100) (1 NRAagm 100) 1, where NRAagm and NRAagx are the average percentage NRAs for the import-competing and exportable parts of the agricul- tural sector. d. RRA is defined as 100*[(100 NRAagt ) (100 NRAnonagt) 1], where NRAagt and NRAnonagt are the percentage NRAs for the tradables parts of the agricultural and nonagricultural sectors, respectively. Japan, Republic of Korea, and Taiwan, China 93 became an upper-middle-income country in the 1960s, to 30­40 percent. The RRAs of Korea and Taiwan, China, meanwhile, both in the low-income stage of development in the 1950s and the lower-middle-income stage in the 1960s, were at very low levels, involving negative rates for some years before the mid-1970s. Following Japan's entry into the high-income stage of development in the 1970s, it increased its RRA steadily, except during the global food crises in 1973­74. The RRA reached a peak in 1994, although this year followed a bad rice harvest (one- quarter below average). Japan's RRA was within the 100­150 percent range after the mid-1980s, except in 1994. In Korea, the rapid rise of agricultural assistance began in the late 1970s when the country moved from the lower- to the upper-middle-income stage of develop- ment. Taiwan, China followed Korea with an increase in RRA, though the differ- ence in the level of the RRA between the two economies continued to be significant during the upper-middle-income stage in both countries. It is interesting to see that Taiwan, China was behind Korea in terms of the RRA level until the mid- 1990s. After this time, however, Korea's RRA fell. After Korea and Taiwan, China entered the high-income stage of development in the 1990s, a relatively high RRA was maintained in both economies, albeit with some fluctuations. The wide fluctuations in RRA in the late 1990s were caused by the currency crises in East Asia that began in 1997. This resulted in a sharp decline in the RRA in Korea in 1997 and 1998. Sharp increases in the RRA in Taiwan, China in 1999 and 2000 were caused by shortages of livestock products because of the September 1999 earthquake and reduced production of pig meat resulting from the spread of foot-and-mouth disease in 1997. Although the paths of the RRAs were different during the middle- income stage of development, both Korea and Taiwan, China started at slightly negative protection levels in the low-income stage in the 1960s and reached very high RRAs (about 120 percent in Korea and 70 percent in Taiwan, China) by 2000. Movements of the NRAs for covered farm products were similar to those of the RRAs in all three economies until the late 1970s, when the growth of the NRAs for nonagricultural products was much faster, particularly in Taiwan, China. In con- trast to the path of the RRA, Taiwan, China kept pace with Korea in terms of the growth of the NRA for agriculture, albeit at the level of about 10 years behind that of Korea. Taiwan, China then caught up with Japan and Korea at a 150­180 per- cent NRA in the late 1990s. The three economies all maintained policies to protect covered agricultural products that were considered politically important and sen- sitive. However, the importance of these covered products declined over time with a smaller share of those commodities in the value of production. Thus, the growth in the RRA was less than that in the NRA because the RRA takes into account the nondistorted uncovered products whose share in value of production increased. Fluctuations in the RRA and NRA consist mainly of changes in the NRA of individual commodities and changes in the weight of each commodity. Because 94 Distortions to Agricultural Incentives: A Global Perspective Figure 2.2. NRA to Rice, Japan, Korea, and Taiwan, China, 1955­2007 900 700 500 percent 300 100 100 1955 1959 1963 1967 1971 1975 1979 1983 1987 1991 1995 1999 2003 2007 Japan Taiwan, China Korea Source: Honma and Hayami (2008) rice is the most important agricultural product in Japan, Korea, and Taiwan, China, its protection had a large influence on the RRA. A clear upward trend in the NRA of rice was present in all three economies (figure 2.2). In Japan, the NRA for rice was as high as nearly 100 percent in the 1960s, when Japan had already entered the upper-middle-income stage of development, whereas it was nearly zero in Korea and Taiwan, China in the lower-middle-income stage. From the late 1970s, when Korea and Taiwan, China approached the upper-middle-income stage, the NRA for rice began to rise sharply and continued to rise thereafter. The fastest increase in the NRA for rice occurred in Japan from the late 1970s. It peaked in the late-1980s. The rapid increase was caused, to a large extent, by a rapid appreciation of the Japanese yen relative to the U.S. dollar. Though the border price of rice declined sharply, there was no transmission to domestic market prices because of the control of rice imports by the government. In 1993, a temporary interruption in the peak of the rice NRA in Japan, caused by a bad harvest, resulted in a shortage of Japonica rice on world markets. This raised border prices, while domestic prices were kept relatively stable under the Food Control System. There- after, further increases in the NRA for rice were counteracted by yen depreciation and also by the acceptance of minimum access obligations in the URAA in 1995 and the later shift to tariffication in 1999. In Korea and Taiwan, China, the upward trend in the rice NRA continued after the 1970s. Such increases were a major factor underlying rapid increases in the Japan, Republic of Korea, and Taiwan, China 95 RRA in Korea during the upper-middle-income stage of development, because the weight of rice in agricultural production continued to be high. Korea's exemption from tariffication in the URAA allowed the NRA for rice to grow even under the implementation of WTO commitments. Similar to the NRA for covered products in figure 2.1, Taiwan, China followed Korea in the growth of the NRA for rice. Tai- wan, China mimicked the situation in Korea, though with a 5- to 8-year lag between the 1970s and the mid-1990s. In more recent years, the gap in NRAs for rice between Taiwan, China and Korea has increased, but the protection level of rice in the former appears to be maintaining a rising trend. Consumer tax equivalents on food Though support provided to farmers in Japan, Korea, and Taiwan, China has come mostly via food import restrictions, there have been additional schemes whereby crop producer prices were supported at levels above those charged to grain and soy- bean consumers (including feedmixers providing livestock producers with animal feedstuffs). Thus, the consumer tax equivalent (CTE) is below the NRA for some crop products. As a result, together with the different weights of various products in consumption as compared with production, the average NRA for covered products was approximately 50 percent above the CTE for both Japan and Korea in 2000­04 (compare tables 2.3 and 2.5). In this way, consumers were spared some of the implicit tax that otherwise would have been imposed on them had border measures alone been used to raise producer prices above international levels. Sources of Growth in Agricultural Protection Measures The experiences of Japan, Korea, and Taiwan, China are good examples of policy switching from exploitation to protection of agriculture when economies grow through industrial development. This shift is most clearly illustrated by the cases of Korea and Taiwan, China, whose agricultural protection levels were negative in the 1950s and the 1960s and began the rise sharply in the 1970s with the success of industrial development. The growth of agricultural protection in Japan, Korea, and Taiwan, China is empirically documented in Anderson, Hayami, and Honma (1986), which draws attention to three characteristics of the East Asian growth of agricultural protection, based on the nominal rates of protection for agricultural products, in comparison with other advanced economies: first, the rapid rise over time in pro- tection rates in the three economies in East Asia; second, the faster increase in agricultural protection in the three East Asian economies than in other industrial economies for the period of 1955 to 1980; and third, the fact that the highest level of agricultural protection reached by the three economies as of 1980 was rivaled 96 Table 2.5. CTEs for Selected Agricultural Products in Japan, Korea, and Taiwan, China, 1955­2007 (percent) a. Japan 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Apples -- -- -- -- -- -- 30.8 23.8 27.5 31.0 17.0 Barley 34.9 32.6 10.2 5.8 20.3 29.1 158.1 131.9 118.9 105.9 114.3 Beef 27.4 68.3 130.8 106.0 215.3 136.7 208.9 177.0 191.8 149.1 38.5 Cabbage -- -- -- -- -- -- 9.2 30.8 127.0 176.6 204.0 Cucumbers -- -- -- -- -- -- 56.4 17.2 29.7 42.9 30.9 Eggs 3.0 3.3 5.3 2.8 3.3 1.4 19.7 22.9 33.6 27.2 17.0 Grapes -- -- -- -- -- -- 85.3 81.6 117.2 176.3 177.8 Mandarins -- -- -- -- -- -- 20.0 44.5 47.0 32.0 46.0 Milk 44.2 96.2 162.1 165.4 385.7 211.5 365.2 280.3 238.0 273.2 93.5 Onions -- -- -- -- -- -- 54.0 80.4 143.9 282.9 294.0 Pears -- -- -- -- -- -- 35.0 24.0 64.2 157.3 103.4 Pig meat 17.9 59.8 12.4 3.4 2.9 12.6 0.5 5.7 10.3 5.5 138.0 Poultry 33.1 42.8 33.5 36.7 31.1 16.8 17.6 25.1 41.4 74.0 11.7 Rice 73.1 85.7 103.9 137.3 179.6 232.6 548.5 613.2 506.4 574.6 348.6 Soybeans -- -- -- -- -- 0.0 0.0 0.0 0.0 0.0 0.0 Spinach -- -- -- -- -- -- 89.2 138.0 236.7 134.4 32.3 Strawberries -- -- -- -- -- -- 10.0 24.8 26.2 16.4 7.0 Sugar -- -- -- -- -- 167.0 185.9 154.8 155.4 151.6 119.6 Wheat 35.9 33.0 27.4 14.1 26.7 37.2 136.1 108.4 73.4 68.7 73.6 All covered products 53.0 62.2 66.8 67.6 93.2 98.8 134.9 119.3 116.1 106.6 81.0 Import-competing 53.0 62.2 66.8 67.6 93.2 98.8 134.9 119.3 116.1 106.6 86.9 Exportables -- -- -- -- -- -- -- -- -- -- -- Dispersion, covered productsb 39.1 39.6 66.5 75.0 144.0 97.2 154.5 149.3 130.0 141.8 108.3 b. Republic of Korea 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Wheat 46.2 22.5 11.1 1.1 16.2 46.0 132.8 167.4 80.5 80.6 -- Barley 40.8 77.8 64.9 96.6 57.3 119.6 325.6 411.5 341.2 327.5 174.4 Rice 7.7 5.5 5.0 29.1 54.5 113.4 211.5 261.7 290.8 385.3 213.3 Beef 38.8 34.4 64.9 73.9 162.6 163.2 122.1 200.7 153.9 167.7 182.3 Pig meat 15.2 21.7 158.7 204.1 202.9 169.1 124.7 149.3 116.2 134.4 103.1 Poultry 11.8 6.9 131.4 103.5 161.7 94.2 86.6 155.6 171.7 179.2 55.7 Eggs 27.1 24.7 23.0 0.1 7.5 14.9 19.4 28.0 26.6 54.3 31.6 Milk -- -- 173.3 108.8 189.0 179.8 185.2 203.7 140.7 149.8 137.0 Cabbage -- -- -- -- -- -- 30.0 30.0 29.1 27.6 27.0 Peppers -- -- -- -- -- -- 175.0 245.4 145.5 197.0 235.7 Soybeans 19.8 8.2 51.6 63.2 95.2 245.4 112.2 75.5 63.6 66.8 91.9 Garlic -- -- -- -- -- -- 250.3 288.8 213.3 122.6 128.1 All covered products 5.0 5.4 14.5 39.7 63.9 114.3 148.5 176.4 144.9 154.1 135.1 Import-competing 5.0 5.4 14.5 39.7 63.9 114.3 148.5 176.4 144.9 154.1 115.7 Exportables -- -- -- -- -- -- -- -- -- -- -- Dispersion, covered productsb 34.7 37.6 85.3 81.1 92.3 82.3 95.1 118.1 107.3 116.2 81.7 (Table continues on the following pages.) 97 98 Table 2.5. CTEs for Selected Agricultural Products in Japan, Korea, and Taiwan, China, 1955­2007 (continued) (percent) c. Taiwan, China 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­02 2003­07 Rice 29.6 6.6 17.9 9.4 7.6 32.5 103.3 161.4 167.6 203.1 -- Wheat 38.3 16.4 29.6 14.6 1.6 0.3 -- -- -- -- -- Beef 13.7 41.2 28.8 22.0 79.6 77.0 101.3 98.5 82.6 72.8 -- Pig meat 8.1 64.0 99.7 98.3 60.6 42.6 76.5 103.9 131.3 173.2 -- Poultry 47.5 3.7 21.2 27.1 30.0 63.6 84.6 143.0 228.7 279.5 -- Eggs 0.0 0.0 0.0 0.0 0.0 0.7 26.8 23.9 17.9 24.7 -- All covered productsa 21.1 7.7 6.9 19.0 15.2 38.4 82.7 116.5 136.8 166.5 -- Import-competing 6.1 13.2 26.0 23.3 27.3 49.0 74.7 102.1 129.1 159.5 -- Exportables 23.7 7.0 5.2 19.1 13.7 36.2 89.6 161.4 167.6 203.1 -- Dispersion, covered productsb 33.3 35.9 47.2 40.6 40.2 32.2 34.1 56.2 87.1 106.0 -- Source: Anderson and Valenzuela (2008), based on authors' spreadsheet. Note: -- = data not available. a. Weighted averages, with weights based on the unassisted value of consumption, where consumption is derived using the value of production and self-sufficiency ratios (derived from the FAOSTAT Database) as production/consumption. b. Dispersion is a simple five-year average of the annual standard deviation around the weighted mean of CTEs of covered products. Japan, Republic of Korea, and Taiwan, China 99 only by Switzerland. Anderson, Hayami, and Honma also note that the growth of agricultural protection in these economies during the three decades to 1980 was exceptionally rapid compared with Western countries that started down the path of industrialization earlier. That is, East Asia was not exceptional in having increasing levels of agricultural protection, but it was exceptional in the speed at which it reached the world's highest level. Figures 2.1 and 2.2 indicate that the protection growth in terms of RRA and NRA continued at the same speed for about 20 years after the previous study period. The rapid growth of agricultural protection in industrializing economies can be explained largely by the shift in comparative advantage away from agriculture to industry as the result of successful industrialization. The decline in agriculture's comparative advantage increased the intersectoral resource adjustment costs that would have been shouldered by farmers if they were left to the competitive market mechanism. Instead, the increase in costs boosted farmers' demand for agricul- tural protection. This problem typically applies when industrial growth has been so rapid that intersectoral adjustments are not fast enough under free markets to prevent a widening rural-urban income disparity.11 The association between the rise in agricultural protection and the decline in agriculture's comparative advantage was tested in Honma and Hayami (1986) using multiple regression analysis and a pooled data set for 15 countries at six points in time ending in 1980. A strong correlation was found between the level of aggregate NRP12 and the index of agriculture's labor productivity relative to the total economy's labor productivity. Honma and Hayami conclude that the high level of agricultural protection in East Asia resulted not so much from factors unique to East Asia but mainly from factors common to all industrial countries. It should be noted, however, that there are differences in the process of intersectoral resource adjustment between Japan and the other two economies. In 1955, the first year of the analysis presented here, Japan was already in the middle-income stage of economic development and about to enter the so-called high-growth era character- ized by extremely rapid industrialization and a widening income gap between rural and urban households.13 Soon after shifting from the middle-income to the high- income stage of development, a process that took less than two decades, Japan increased its level of agricultural protection. Protection measures were raised to a level comparable with that of the European Economic Community during the 1960s. Meanwhile, Korea and Taiwan, China were still in the low-income stage of eco- nomic development in the 1950s. As both economies entered the middle-income stage in the 1960s, productivity growth in agriculture lagged behind that of non- agriculture as a result of successful industrialization. With delays in labor out- migration from farming, farmers' income levels tended to decline relative to those of nonfarmers. Nevertheless, because agriculture was such a large sector in terms 100 Distortions to Agricultural Incentives: A Global Perspective of both national income and labor force, it was impossible during the middle- income stage of development for governments to secure sufficient finance from nonagricultural sectors to raise support for farmers to the extent needed to close the income gap. Thus, despite the growing rural-urban income disparity, Korea and Taiwan, China retained low levels of agricultural protection until the late 1970s and early 1980s, respectively. The agricultural problem confronted by middle-income economies like Korea and Taiwan, China in the 1960s and 1970s has been called the "disparity problem" by Hayami (2005) and Hayami and Godo (2004), referring to the income dispar- ity between farm and nonfarm households. The problem is a lag in productivity growth in agriculture relative to nonagriculture, brought about by insufficient labor out-migration from farming in response to the successful industrialization that raised these economies to the middle-income stage. Farmers, who observe nonfarm workers' rapid escape from poverty, begin to realize how relatively poor they are, even if their income level did not decrease from the previous stage. The resulting dissatisfaction among farmers often becomes a significant source of social instability. Once an economy reaches the middle-income stage, that dissat- isfaction becomes a prime concern of policy makers, who might adopt agricul- tural protection measures to appease farmers and prevent the dissatisfaction from elevating into serious antigovernmental sentiment. Such protection may not be strong enough to close the income gap between farmers and urban workers until an economy graduates from the lower-middle- income stage, however. Because the share of agriculture in both national income and the labor force is still large, it is difficult to either raise sufficient revenue from the nonfarm sectors to close the growing farm-nonfarm income gap with direct support payments or to pass on the cost of agricultural protection to consumers by raising food import barriers, as increases in food prices erode real wages paid by the large number of small-scale enterprises that rely heavily on cheap labor. Faced with the disparity problem, policy makers in middle-income countries are forced to search for ways to protect farmers within the constraint of the food problem that is still binding because a large number of urban workers are still poor and thus still dedicate a large share of household expenditure to food. In the early 1990s, during which all three Northeast Asian economies were in the high-income stage, the decline in relative agricultural income (in terms of agricultural GDP per worker divided by total GDP per worker) stopped in Japan and Korea. In Taiwan, China, the relative agricultural income continued to decline until recently (table 2.1), despite the high level of agricultural protection. The rea- son relative agricultural income continued to decline in Taiwan, China was that the total economy's labor productivity increased more rapidly than agriculture's labor productivity even after 1990. Japan, Republic of Korea, and Taiwan, China 101 Figure 2.3. RRA to Agriculture and Real GDP Per Capita, Japan, Korea, and Taiwan, China, 1955­2004 250 200 150 RRA (%) 100 50 0 50 0 5,000 10,000 15,000 20,000 25,000 30,000 real GDP per capita at 2000 constant prices ($) Japan Korea Taiwan, China Source: Authors' computations. Agricultural protection in Korea rose faster and to a higher level during its upper-middle-income stage of development than in Taiwan, China or Japan during theirs. RRAs in Korea, for example, significantly exceeded those of Taiwan, China and Japan for the same levels of per capita income throughout their upper- middle-income stage (figure 2.3). The difference could reflect the different costs of intersectoral adjustment (corresponding to changes in comparative advantage) that farmers had to bear. In Korea, the shift of labor from agriculture to nonagricul- ture involved the migration of workers from rural to urban areas, whereas in Japan and Taiwan, China farmers increased their nonfarm activities while continuing to live in their home villages and towns and farming part-time. Correspondingly, both the pecuniary and psychological costs of intersectoral labor reallocation were much higher for farmers in Korea. In Japan, the decline in relative agricultural income ceased in the 1970s when the country reached the high-income mark. This was due to a deceleration in the growth of labor productivity in the total economy after reaching the high-income stage. The Korean experience after 1990, mean- while, is likely the result of fast increases in agricultural labor productivity resulting from the rapid out-migration of agricultural labor to urban activities (table 2.1). 102 Distortions to Agricultural Incentives: A Global Perspective Figure 2.4. RRA to Agriculture and Relative GDP Per Agricultural Worker, Japan, Korea, and Taiwan, China, 1955­2004 250 200 150 RRA (%) 100 50 0 50 0 10 20 30 40 50 60 70 ratio of agricultural GDP per worker to total GDP per worker (%) Japan Korea Taiwan, China Source: Authors' computations. The relationship between relative agricultural income and the RRA in Japan, Korea, and Taiwan, China for 1955­2004 is shown in figure 2.4.14 Except for Korea in 1990 and 2000, there is a negative correlation for all countries and all periods. The correlation is weak, however, when relative agricultural income is more than about 40 percent (a rate that corresponds to the low-income and lower-middle- income stages of economic development). Upon reaching the upper-middle-income stage of development in the 1980s, Korea and Taiwan, China strengthened their agricultural protection policies. This followed Japan's protection pattern in the 1960s, when the income gap was widening and protection measures were deemed necessary to close it. Under such circumstances, politicians were not able to resist pressure from the farm lobby to institute policies to prevent farmers' incomes from lagging behind those of non- farm workers. In the case of Korea, policy makers may have had a specific reason for strengthening agricultural protection, particularly at the farmgate level--the constant threat of communist aggression from the North. The imminent hostility prompted commercial and industrial interests to support farmers and thereby maintain political stability. If the income gap had been adequately dealt with during the middle-income stage of development, problems caused by agricultural protection in the upper- middle-income stage might have been avoided. Yet the disparity problem in the Japan, Republic of Korea, and Taiwan, China 103 middle-income stage has received relatively little attention in academic and policy debate, despite the fact that there are many economies attempting to reach the upper-middle-income and high-income stage through industrialization. The growing income disparity between farm and nonfarm populations could become a major source of social and political instability elsewhere in Asia, from ASEAN countries to China and eventually to South Asia, particularly India. In analysis by Honma and Hayami (1986), it was found that political power in the agricultural sector is maximized when the share of agriculture declines to 4 to 5 percent of GDP or 5 to 8 percent of the labor force. Japan has passed this range, and Korea recently entered this peak zone in terms of both GDP and labor force, as did Taiwan, China in terms of the labor share (having passed over this zone in 1990 in terms of GDP share). Political economy factors may well underlie the rise of agricultural protectionism in Korea at the high-income stage after 1990, as observed in terms of the NRA at the farmgate level despite no apparent further increase in its agricultural comparative disadvantage.15 Japan's Pre-World War II Experience The pattern of growth of agricultural protection in Japan, Korea, and Taiwan, China in the era of East Asian economic miracle, as outlined in the previous sec- tion, was consistent with the hypothesis that rapid growth in protection occurred when these economies were in the middle-income stage under the dictate of the "disparity problem" described earlier. Under the disparity problem, when farm- ers' income levels decline relative to those of nonfarmers, the economy is often characterized by a dual structure: a formal sector consisting of large, modern enterprises and government agencies, and an informal sector consisting of agri- culture and other small and medium enterprises. That was the case for Japan in the half-century before World War II, during which Japan advanced from the low- income to the middle-income stage of economic development. Japan set upon a modern economic growth path with the transformation in political structure that occurred under the Meiji restoration of 1868. In short, Japan went from a union of feudal fiefs under the hegemony of Tokugawa shogun (tycoon) to a modern nation state in the form of the constitutional monarchy under an emperor who was a symbol of national unification without actual ruling power. The immediate impetus for this political reorganization was the threat of colonization by Western powers that became obvious through the gunboat diplo- macy of the United States and the use of Admiral Perry's fleet. The national slogan of the Meiji state was to establish fukoku kyouhei (a wealthy nation and strong army) for the sake of preserving national independence. To achieve this goal, eco- nomic policies in Meiji Japan were aimed at the promotion of modern industries 104 Distortions to Agricultural Incentives: A Global Perspective to catch up to the economic power of Western nations. At the time, Japan did not have the freedom to set import and export duties above 5 percent ad valorem lev- els, as per commercial treaties signed by the Tokugawa tycoon with Western pow- ers in the mid-19th century. Thus, industrial promotion policy relied mainly on subsidies in areas such as the import of machines and factories and the purchase of their designs, the employment of engineers and skilled workers, and the collec- tion and dissemination of information on overseas technologies and markets. It was mainly through taxation of agriculture--through the newly established land tax system--that subsidies for industrial promotion and other modernization measures were financed. As the data in table 2.6 show, in the early Meiji period (before 1900), the agri- cultural sector shouldered about 90 percent of the total direct tax burden, which amounted to about 15 percent of agricultural GDP. Agriculture's share of the gov- ernment's subsidies at the time, however, amounted to less than one-quarter, or less than 5 percent of agricultural GDP. Evidently, at the beginning of its modern economic growth period, Japan promoted its modern sectors through the exploitation of the traditional sectors, a strategy commonly practiced by develop- ing economies upon becoming independent of colonial powers in the 1960s. Tax- ation and subsidization imbalanced between agriculture and nonagriculture was even greater than the data of table 2.6 reveal, as a disproportionately high share of the population educated at publicly financed schools were from nonfarm house- holds. Under strong promotion by the government, industrialization progressed rapidly in Japan, especially in labor-intensive manufacturing. Comparative advantage in this sector was unhampered owing to a virtual free trade scenario created by the absence of tariff autonomy in Japan. Changes in the position of agriculture in the total economy over the course of modern economic development under the Meiji restoration are summarized in table 2.7. The real GDP per capita series shows that the Japanese economy moved from the low-income to the middle-income stage by the beginning of the 20th cen- tury, with the share of agriculture in GDP at about 40 percent. This is roughly com- parable to the share of agriculture in GDP in Korea and Taiwan, China when they advanced to the middle-income stage. Meanwhile, the growth of labor productiv- ity in agriculture lagged behind that of industry, resulting in a continual decline in the ratio of labor productivity in agriculture to labor productivity in industry (table 2.7, column 6). This decline reflects successful industrial development, as in Korea and Taiwan, China during the era of East Asia's economic miracle. Neverthe- less, because the terms of trade did not improve for agriculture throughout the inter-war period (table 2.7, column 7), income per capita in farm households con- tinued to decline relative to that of nonfarm employees' households, parallel to the decline in agriculture/industry real productivity ratio (table 2.7, column 8). Japan, Republic of Korea, and Taiwan, China 105 Table 2.6. Changes in Direct Tax Burdens and the Allocations of National Government Subsidies to Agricultural and Nonagricultural Sectors, Japan, 1878­1937 a. Tax burdens Direct tax burdena Direct tax rateb Agriculture Nonagriculture Agriculture Nonagriculture (yen (% of (yen (% of millions) total) millions) total) (%) (%) 1878­82 63.6 91 6.3 9 -- -- 1888­92 58.5 86 9.8 14 14.9 2.0 1898­02 99.1 74 35.4 26 11.7 2.7 1908­12 153.4 54 132.2 46 11.2 5.5 1918­22 295.7 41 431.1 59 7.5 4.8 1928­32 205.5 33 421.3 67 8.1 3.8 1933­37 197.3 26 559.2 74 6.5 4.0 b. Subsidy allocations Subsidy receiptc Subsidy rated Agriculture Nonagriculture Agriculture Nonagriculture (yen (% of (yen (% of millions) total) millions) total) (%) (%) 1881 0 0 0.7 100 -- -- 1891 0 0 2.5 100 0 0.49 1901 0.4 2 18.7 98 0.05 1.41 1911 0.3 1 27.8 99 0.02 1.09 1921 0.6 1 51.8 99 0.02 0.55 1931 21.4 17 101.5 83 1.17 1.11 1934 28.3 28 71.0 72 1.14 0.58 Source: Tobata and Ohkawa (1956) for tax and subsidy data; Ohkawa and Shinohara (1979) for sectoral net domestic product (NDP) data. Note: -- not available. a. Includes both national tax and local rates. b. Direct tax burden divided by sectoral NDP. c. National government subsidies. d. Subsidy receipt divided by sectoral NDP. Growing dissatisfaction among farmers in Japan gave rise to strong political lobbying--organized by the politically powerful landlords--for reduced tax bur- dens and increased support to agriculture. The result was a significant reduction in the tax burden and a greater allocation of government subsidies to agriculture 106 Table 2.7. Farm-Nonfarm Income Disparity in Japan's Economic Development, 1885­2000 Agriculture/ GDP Share of Nominal Tariff Average Agriculture/ manufacturing Farm/nonfarm per capita agriculture rate of rate tariff rate, industry labor terms of household (2000 US$, in GDP protection on rice all products productivity trade Income ratio PPP)a (%)b for rice (%)c (%)d (%)e ratio (%)f (1885=100)g (%)h 1885 1,092 45 15 -- -- 75 100 76 1890 1,285 48 34 -- -- 67 115 87 1900 1,498 39 21 -- 3.7 49 102 52 1910 1,656 32 35 14 16.2 37 98 47 1920 2,154 30 16 10 10.7 50 99 48 1930 2,350 18 57 14 22.6 31 104 32 1935 2,693 18 134 41 23.8 24 136 38 1955 3,519 21 49 -- 3.5 55 163 77 1960 5,063 13 98 -- 6.5 39 169 70 1970 12,337 6 150 -- 6.9 25 303 94 1980 17,056 4 205 -- 2.5 25 342 116 1990 23,580 2 481 -- 2.7 26 379 115 2000 26,220 1 560 778 2.1 22 347 101 Source: See notes below. a. GDP per capita at PPP for 2000 from World Bank (2006), linked with the series from OECD (2003). b. The share of agriculture in nominal GDP for 1885­1935 and share in NNP for 1885­1935 are from Ohkawa and Shinohara (1979). Data for 1960­2000 are from World Bank (2006). c. Nominal rates of protection for rice for 1885­1960 are calculated by the difference between the domestic wholesale price of rice and the unit value of imported rice as percentage of the latter. For 1970­2000, it is calculated by the difference between the domestic wholesale price of rice and unit value of world rice imports multiplied by 1.18, expressed as a percentage of the latter. Data are from Kayo (1977) and Bank of Japan, Yearbook of Wholesale Price Indexes, various years for domestic wholesale price of rice, and Nihon Boeki Seiran, Toyo Keizai Shinposha, 1935, Yearbook of Japan Foreign Trade Statistics, Japan Tariff Association, and FAOSTAT, FAO for border prices. d. Tariffs for 1910, 1920, 1930 and 1935 are tariffs in 1908, 1918, 1928 and 1933, respectively, from Ohkawa, Shinohara and Umenura (1967). Tariff rate for 2000 is ad valorem tariff equivalent of specific duty, 341 yen per kilogram, which was reported to the WTO by the government of Japan. e. Tariffs for 1900, 1910, 1920, 1930, and 1935 are tariffs in 1898, 1908, 1918, 1928, and 1933, respectively, from Ohkawa, Shinohara, and Umenura (1967). Tariffs for 1955­ 2000 are average tariffs calculated by total tariff revenue as percentage of total import cif value in the Ministry of Finance, Monthly Report of Financial Statistics, various issues. f. The ratio of real GDP per worker in agriculture (including forestry and fishery) to real GDP per worker in industry (including mining). 1885­1970 from Hayami (1986). 1980­ 2000 values are extended from 1970 using real GDP and the numbers of employed persons from the Annual Reports of National Accounts. g. For 1985­1960, the ratio between the price index of agricultural products and the price index of manufacturing products in Ohkawa, Shinohara, and Umenura (1967). Values for 1970­80 are extended from 1960 using the Ministry of Agriculture, Forestry and Fishery's price index of agricultural products and the Bank of Japan's domestic corporate goods price index for manufacturing industry products. h. For 1885­1935, the ratio in household income per household member between farm and nonfarm households in Otsuki and Takamatsu (1982). Values for 1955­2000 are the ratio in per capita income between farm households and employees' households based on the Ministry of Agriculture's Farm Household Economy Survey and the Ministry of Internal Affairs' National Survey of Family Income and Expenditure. Farm households in 1990­2000 exclude noncommercial farm households. 107 108 Distortions to Agricultural Incentives: A Global Perspective in the first half of the 20th century. Before the beginning of the 20th century, land- lords were largely satisfied by the government's support to agricultural research and extension services and land infrastructure improvements such as irrigation and drainage systems, which proved to be highly effective in raising rice yields per hectare and thereby raising land prices and land rents for the benefits of landlords (Hayami and Yamada 1991). However, as rapid industrial development con- tributed to a continued decline in the comparative advantage in agriculture, landlords' demands began to shift toward border protection on agricultural com- modities, especially rice. Their strong lobbying achieved the implementation of a rice tariff at 15 percent ad valorem in the first year of the Russo-Japanese War (1904­05), a tariff that was justified as a way to raise government revenue to finance the war. Despite the initial intention to terminate the tariff at the end of the war, landed interests lobbied extensively to make it permanent in the form of a specific duty. Thereafter, the rice tariff became an issue of a major public contro- versy in Japan, similar to the controversy caused by the Corn Laws in the United Kingdom a century earlier and German grain tariffs half a century later. The imperial Agricultural Society, representing the landed interests, and the Tokyo Chamber of Commerce, representing the interests of manufacturers and traders of export commodities, lobbied strongly for opposite ends. The battle ended with a victory for the landed interests and the successful imposition of a specific duty on rice at 1 yen per 60 kilograms. This outcome contrasts with the victory of the bourgeoisie in the United Kingdom and the repeal of the Corn Laws in 1846, and resembles the situation in Germany in which tariff protection was installed on food grains (wheat and rye) in 1879 under Bismarck. But unlike the United Kingdom, which was able to estab- lish itself as the workshop of the world, the comparative advantage of industry was less certain in Germany, so that industrialists found it advantageous to seek protection on their products while approving some protection on agriculture. In addition, the rapid growth of the Social Democratic Party in Germany, a labor party initially based on orthodox Marxist doctrine, was considered a common menace by the Junkers (landed nobility in Prussia and Eastern Germany) and the industrialists in Western Germany. In fact, the simultaneous implementation of grain tariffs and iron and steel tariffs in Germany was the result of a united cam- paign of landlords in Eastern Germany and Western Germany (Gerschenkron 1943). This experience was repeated in other countries that came late to industrial- ization, such as France and Italy, in their attempts to match the United Kingdom in industrial strength (Kindleberger 1951). Japan's protectionist policies were similar: after tariff autonomy was recovered in 1911, Japanese industrialists actively lobbied for industrial protection, especially in heavy and chemical industries. They also Japan, Republic of Korea, and Taiwan, China 109 campaigned for reductions in tariffs on imports of industrial raw materials such as raw cotton and iron ore (Little, Scitovsky, and Scott 1970; Yamazawa 1984). As a result, Japan saw the emergence of tariff escalation, with lower rates applied to materials for industrial processing and higher rates applied to its imports of processed final products. Although agricultural production was raised by means of increases in food tariff rates, this was largely paralleled by increases in industrial tariff rates, which can be inferred from the movements in the average tariff rate for all products compared with the movements in just the rice tariff rate (table 2.7, columns 4 and 5). However, tariffs were largely exempt on the imports of raw materials for industrial production, so effective rates of industrial protec- tion were much higher than the nominal rates implied by the tariff rates. In par- ticular, a zero tariff on raw cotton was instrumental in making the cotton spinning industry the top foreign exchange earner in Japan and, at the same time, com- pletely eradicated domestic cotton farming. Although the rice tariff was raised successively from 14 percent ad valorem in 1910 to 41 percent in 1935, which with quantitative import restrictions increased the nominal rate of protection on rice from 21 percent in 1900 to 134 percent in 1935, improvement in the terms of trade for agriculture was slower than the decline in the agriculture/industry labor productivity ratio. This resulted in the continual decline of farmers' household income per capita relative to that of non- farmers' throughout the interwar period (table 2.7). Although agricultural protec- tion began to increase significantly during this period, it was evidently insufficient to counteract the loss of agriculture's comparative advantage owing to rapid industrial development. To undertake agricultural protection at a scale sufficient to close the widening income disparity, Japan waited until after World War II, when the Japanese economy advanced to the upper-middle-income and the high- income stages of development, so that nonagricultural sectors could bear the cost of agricultural protection. In the early 20th century, the share of food in house- hold consumption expenditure (the Engel coefficient) was higher than 60 percent. This implies that the elevation of food prices had a large effect on the cost of living and, hence, on the wage rate of workers, which caused serious damage to labor- intensive industries, which were then at the center of the Japanese economy. Indeed, as mentioned earlier, the fear that high food prices would damage indus- trial development was a major motive in the Japanese government's decision to launch rice development programs in Korea and Taiwan, China following the Rice Riots in 1918. These programs were launched despite opposition from landlords in Japan to policies fostering competition to Japanese rice farming. The importa- tion of Japonica rice free of tariffs from the two colonies became a major factor aggravating agricultural depression in Japan during the 1930s. 110 Distortions to Agricultural Incentives: A Global Perspective The situation changed dramatically after Japan advanced to the upper-middle- income stage of development in the 1960s. Although supports on agricultural product prices were raised rapidly, industrial wage rates were raised even faster, so that the Engel coefficient fell from 52 percent in 1955 to 31 percent in 1980 and further to 17 percent in 1995 (Hayami and Godo 2002). Meanwhile, the center of gravity in Japanese industry moved from labor-intensive manufacturing to capi- tal- and knowledge-intensive activities. In this environment, Japanese industrial- ists were able to tolerate increases in food prices so as to prevent farm-nonfarm income disparity from widening. Industrialists found it was to their advantage to support farmers and to keep them as allies against organized labor and left-wing activities under the Cold War regime (similar to the attitude of German industri- alists toward the grain tariff campaign a century earlier). The major surge of Japan's agricultural protectionism continued until it was counteracted by the seri- ous trade frictions with food-exporting countries, particularly the United States. What Have We Learned? This chapter's estimates of the NRAs for selected individual commodities and the RRA between agricultural and industry show that the growth of agricultural pro- tection in Northeast Asia, together with the decline of industrial protection rates, caused the RRA to rise in the three economies over the five postwar decades under investigation. The experience in Japan, Korea, and Taiwan, China can be explained by factors common to rapidly industrializing economies, especially the high cost of industrial adjustment shouldered by farm producers in the process of rapid industrial development. However, the agricultural protection level continued to grow even after 1980 in all the three economies despite apparently decreased need for agricultural support to prevent widening rural-urban income disparity. All three economies suffered a problem common in the high-income stage of economic development, notably a widening income gap between agricultural and nonagricultural sectors upon reaching middle-income status. If the income gap had been dealt with more appropriately at the middle-income stage, problems caused by agricultural protection in the high-income stage could have been signif- icantly less. All of this suggests that greater attention needs to be paid to the agricultural problem in the middle-income stage, the so-called "disparity problem." The chal- lenge at that stage of development is to find a compromise between the conflicting need, on one hand, to reduce the farm-nonfarm income gap, and on the other, to supply low-cost food to a large number of workers in urban areas when the gov- ernment's capacity to raise sufficient revenue from nonagricultural sectors is weak and when food import restrictions effectively tax net buyers of food. The Japan, Republic of Korea, and Taiwan, China 111 somewhat contrasting patterns of agricultural and industrial growth between Korea and Taiwan, China led to different solutions to that challenge, which may provide insights for some later-developing economies. Notes 1. On the dramatic transformation of Korea and Taiwan, China from slow, inward-looking eco- nomic growth to rapid, export-led economic growth, see, for example, Mason et al. (1980) and Tsiang (1980). 2. ASEAN countries include Brunei Darussalam, Cambodia, Indonesia, the Lao People's Demo- cratic Republic, Malaysia, Myanmar, the Philippines, Singapore, Thailand, and Vietnam. 3. For the perspective presented here on such a question, see Hayami (2005, section 8.4) and Aoki and Hayami (1998). 4. In terms of real GDP per capita in 2000 constant prices, for example, China exceeded US$1,500 in 1990 and US$5,000 in 2004, whereas Thailand passed the US$1,500 level in 1968 and the US$5,000 level in 1991. Among high-income countries, the United Kingdom and France exceeded US$10,000 per capita in 1960 and 1964, respectively, whereas the United States had sur- passed US$10,000 in 1950. 5. For an explanation as to why farm size has remained so small in Japan, see Godo (2007), who argues that farmers have far less interest in selling to neighboring farmers than in keeping their farm- land in the hope of making huge capital gains when the opportunity arises for their farmland to be converted to nonagricultural uses such as shopping centers. While in principle there are strict regula- tions on farmland use, in practice those laws can be manipulated to increase the probability of conver- sion, by pressuring local authorities and politicians. 6. This section draws heavily on Hayami (1988). 7. This section draws heavily on Moon and Kang (1989). 8. This drove a small wedge between the NRA for producers and the consumer tax equivalent for beef (Anderson 1986), similar to what occurred under the scheme operating in Japan in the 1970s. On why the government chose that scheme rather than a more efficient equally protective tariff plus a con- sumer subsidy funded by the tariff revenue, see Hayami (1979) and Anderson (1983). 9. This section draws heavily on Mao and Schive (1995). 10. The NRAs for commodities are different from those estimated by OECD. Major differences between the NRA study presented here and the OECD study for PSE are twofold: (1) domestic prices in the present study are wholesale prices, whereas the OECD uses farmgate prices for PSE and prices paid by consumers at the farmgate level for CSE; and (2) border prices in the calculations presented here are based on the study in Anderson and Hayami (1986), whereas the OECD uses a different set of reference prices. The fact that the producer price was often above the wholesale (consumer) price in the case of grains and soybeans in Japan and Korea is captured by setting the NRA equal to the measured CTE times the ratio NRA/CTE in Anderson, Hayami, et al. (1986) and Anderson (1989) for the period to 1985 and times the negative of the ratio PSE/CSE in OECD (2007) for the period from 1986. Most differences in NRA between the measures in the OECD and the present studies come from the differences in border prices. For example, the border price of rice is common for Japan and Korea as the world import unit value adjusted by a quality coefficient. But the OECD's border price of rice in Japan is based on the price of rice imports by Japan, while the price in Korea is China's export price of rice adjusted by transportation costs and, from 2001, average import prices of rice from China, Thailand, and the United States. The methodology in the present study makes the series of NRAs for rice more stable than that of the OECD for recent years. The discrepancy also explains the stability of the NRA for Korean rice presented here compared with the NRA in the OECD study. The border prices are also different for meat products. In the estimation of NRAs for beef, pork, and chicken, the OECD basically uses the meat data of Canada or the United States for border prices, while the study here uses Japan's import price for beef and unit values for pork and chicken (or Hong Kong, China import 112 Distortions to Agricultural Incentives: A Global Perspective prices for the 1950s). Our approach is preferred for estimating NRAs consistently for longer time peri- ods, particularly for the period when Korean imports were absent or negligible. Also, the approach here is preferred for comparing the NRAs of Korea and Taiwan, China for border prices. 11. For more on agricultural policies in the process of economic development, see Hayami (2005) and Hayami and Godo (2004). 12. Actual data used for the level of agricultural protection in the regression analysis are the nomi- nal protection coefficients (NPC 1 NRP 100). 13. Indeed, Japan's tariff protection for rice began in 1904; after 1918, it includes its colonies of Korea and Taiwan, China in what became an imperial rice self-sufficiency policy. An earlier set of esti- mates of the nominal rate of rice protection suggests it grew from 9 percent in 1903­07 to 21 percent in 1908­12, and to 27 percent in 1913­17. It then fell to an average of 13 percent in 1918­27 with the greater inflow of rice from the colonies before rising again to 26 percent in 1928­32, 45 percent in 1933-37, and 84 percent in 1938, according to Anderson, Hayami, and Honma (1986). See table 2.7 for new estimates for an even longer period. 14. The average RRA for 1955­59 is paired with agricultural GDP per worker relative to total GDP per worker in 1955 and so on. 15. The shares of Korean agriculture in GDP and the labor force were 3.8 percent and 7.7 percent, respectively, in 2004, while agriculture represented 1.7 percent of GDP and 7.5 percent of the labor force for the same year in Taiwan, China (table 2.1, rows 2 and 3). References Amsden, A. 1989. Asia's Next Giant: South Korea and Late Industrialization. New York: Oxford Univer- sity Press. Anderson, K. 1983. "The Peculiar Rationality of Beef Import Quotas in Japan." American Journal of Agricultural Economics 65 (1): 108­12, February. ------. 1986. "The Peculiar Rationality of Beef Import Quotas in Japan and Korea." In The Political Economy of Agricultural Protection: East Asia in International Perspective, ed. K. Anderson and Y. Hayami. Boston, London, and Sydney: Allen and Unwin. ------. 1989. "Korea: A Case of Agricultural Protection." In Food Price Policies in Asia, ed. T. Sicular. Ithaca: Cornell University Press. Anderson, K., Y. Hayami, et al. 1986. The Political Economy of Agricultural Protection: East Asia in Inter- national Perspective. Boston, London, and Sydney: Allen and Unwin. Anderson, K., Y. Hayami, and M. Honma. 1986. "The Growth of Agricultural Protection." In The Political Economy of Agricultural Protection: East Asia in International Perspective, ed. K. Anderson and Y. Hayami. Boston, London and Sydney: Allen and Unwin. Anderson, K., M. Kurzweil, W. Martin, D. Sandri, and E. Valenzuela. 2008a. "Methodology for Measur- ing Distortions to Agricultural Incentives." Agricultural Distortions Working Paper 02, World Bank, Washington, DC (and appendix A of this volume). ------. 2008b. "Measuring Distortions to Agricultural Incentives, Revisited." World Trade Review 7 (4): 1­30. Anderson, K., and W. Martin, eds. 2009. Distortions to Agricultural Incentives in Asia. World Bank, Washington, DC. Anderson, K., and R. Tyers. 1992. "Japanese Rice Policy in the Interwar Period: Some Consequences of Imperial Self Sufficiency." Japan and the World Economy 4 (2): 103­27. Anderson, K., and E. Valenzuela. 2008. Global Estimates of Distortions to Agricultural Incentives, 1955 to 2007. Core database at http://www.worldbank.org/agdistortions. Aoki, M., and Y. Hayami, eds. 1998. The Institutional Foundation of East Asian Economic Development. London: Macmillan. Japan, Republic of Korea, and Taiwan, China 113 Cole, D. C., and Y. C. Park. 1983. Financial Development in Korea, 1945­1978. Cambridge, MA: Harvard University Press. Frank, C. R., K. S. Kim, and K. Westphal. 1975. Foreign Trade Regimes and Economic Development: South Korea. New York: Columbia University Press. Gerschenkron, A. 1943. Bread and Democracy in Germany. Berkeley: University of California Press. ------. 1962. Economic Backwardness in Historical Perspective. Cambridge, MA: Harvard University Press. Godo, Y. 2007. "The Puzzle of Small Farming in Japan." Pacific Economic Papers 365, Australia-Japan Research Centre, Australian National University, Canberra. http://www.crawford.anu .edu.au/pdf/pep/pep-365.pdf Hayami, Y. 1979. "Trade Benefits to All: A Design of the Beef Import Liberalization in Japan." American Journal of Agricultural Economics 62 (2): 342­47. ------. 1986. Nogyo Keizairon [Agricultural Economics]. Tokyo: Iwanami. ------. 1988. Japanese Agriculture under Siege: The Political Economy of Agricultural Policies. London: Macmillan. ------. 2005. "Emerging Agricultural Problem in High-Performing Economies in Asia." Presidential Address at the 5th Conference of the Asian Society of Agricultural Economists, Zahedan, Iran, August 29­31. Hayami, Y., and Y. Godo. 2002. Nogyo Keizairon [Agricultural Economics] Tokyo: Iwanami. ------. 2004. "The Three Agricultural Problems in the Disequilibrium of World Agriculture," Journal of Agriculture and Development 1: 3­16. ------. 2005. Development Economics: From the Poverty to the Wealth of Nations. London: Oxford University Press. Hayami, Y., and S. Yamada, eds. 1991. The Agricultural Development of Japan: A Century's Perspective. Tokyo: University of Tokyo Press. Heston, A., R. Summers, and B. Aten. 2006. "Penn World Table Version 6.2." Center for International Comparisons of Production, Income and Prices, University of Pennsylvania Ho, S. P. S. 1979. "Decentralized Industrialization and Rural Development in South Korea and Tai- wan." Economic Development and Cultural Change 28: 77­96. ------. 1982. "Economic Development and Rural Industry in South Korea and Taiwan." World Development 10: 973­90. Honma, M., and Y. Hayami. 1986. "Determinants of Agricultural Protection Levels: An Econometric Approach." In The Political Economy of Agricultural Protection: East Asia in International Perspec- tive, ed. K. Anderson and Y. Hayami et al. Boston, London, and Sydney: Allen and Unwin. ------. 2008. "Distortions to Agricultural Incentives in Japan, Korea and Taiwan, China." Agricultural Distortions Working Paper 35, World Bank, Washington, DC. Kawano, S. 1969. "Effects of the Land Reform on Consumption and Investment of Farmers." In Agriculture and Economic Growth: Japan's Experience, ed. K. Ohkawa, B. F. Johnston, and H. Kaneda, 374­97. Tokyo: University of Tokyo Press. Kayo, N., ed. 1977. Basic Statistics for Japanese Agriculture. Tokyo: Norin Tokei Kyokai. Kindleberger, C. P. 1951. "Group Behavior and International Trade." Journal of Political Economy 59: 30­46. Krueger, A. O., M. Schiff, and A. Valdés, eds. 1991. Political Economy of Agricultural Pricing Policies. Baltimore: Johns Hopkins University Press. Kuo, S. W. Y. 1975. "Effects of Land Reform, Agricultural Pricing Policy, and Economic Growth on Multiple Crop Diversification in Taiwan." Philippine Economic Journal 14 (1 and 2). Little, I. M. D., T. Scitovsky, and M. Scott. 1970. Industry and Trade in Some Developing Countries. London: Oxford University Press. Mao, Y.-K., ,and C. Schive. 1995. "Agricultural and Industrial Development in Taiwan." In Agriculture on the Road to Industrialization, ed. J. Mellor. Baltimore: Johns Hopkins University Press. Mason, E. S., M. J. Kim, D. H. Perkins, K. S. Kim, and D. C. Cole. 1980. The Economic and Social Modernization of the Republic of Korea. Cambridge, MA: Harvard University Press. 114 Distortions to Agricultural Incentives: A Global Perspective Moon, P. Y., and B. S. Kang. 1989. "Trade, Exchange Rate, and Agricultural Pricing Policies in the Republic of Korea." World Bank Comparative Studies on the Political Economy of Agricultural Pricing Policy, World Bank, Washington, DC. OECD (Organisation for Economic Co-operation and Development). 2003. The World Economy: Historical Statistics. Paris: OECD Development Centre. ------. 2008. Producer and Consumer Support Estimates, OECD Database 1986­2007. http://www .oecd.org Ohkawa, K., and M. Shinohara. 1979. Patterns of Japanese Economic Development: A Quantitative Appraisal. New Haven: Yale University Press. Ohkawa, K., M. Shinohara, and M. Umenura, eds. 1967. Long-Term Economic Statistics in Japan since 1868. Tokyo: Toyo Keizai Shimposha. Otsuki, T., and N. Takamatsu. 1982. On the Measurement of Income Inequality in Prewar Japan. Tokyo: International Development Center of Japan. Schultz, T. W., ed. 1978. Distortions of Agricultural Incentives. Bloomington: Indiana University Press. Scott, M. 1979. "Foreign Trade." In Economic Growth and Structural Change in Taiwan: The Postwar Experience of the Republic of China, ed. J. Galenson. Ithaca: Cornell University Press. Tobata, S., and K. Ohkawa, eds. 1956. Nihon no Keizai to Nogyo [Economy and Agriculture in Japan]. Tokyo: Iwanami. Tsiang, S. C. 1980. "Foreign Trade and Investment as Boosters for Take-Off: The Experience of Taiwan, China." In Export-Oriented Development Strategies, ed. V. Corbo, A. O. Krueger, and F. Ossa. Boulder, CO: Westview Press. World Bank. 1993. The East Asian Miracle: Economic Growth and Public Policy. London: Oxford University Press. ------. 2006. World Development Indicators 2006. Washington, DC: World Bank. Yamazawa, I. 1984. Nihin no Keizai Hatten to Kokusai Bungyo [Economic Development and Interna- tional Division of Labor in Japan]. Tokyo: Iwanami. 3 Western Europe Tim Josling* European agricultural policy, in particular the Common Agricultural Policy (CAP) of the European Union (EU), has long been a matter of international interest. Over- seas producers view agricultural policy in Europe as a major impediment to the opening up of international trade in farm products. Internally, the policy has been no less controversial, with several member states seeking to reform the CAP and others regarding it as a foundation for economic integration. The overall percep- tion, both within and outside the European Union (EU), is of a highly protective policy that shelters a high-cost agricultural sector from the winds of competition blowing from the Americas, Australia, New Zealand, and North Africa. Supporters of the CAP claim social and political benefits from such protection, arguing that reduction of border protection and cutting of domestic support would lead to the depopulation of rural Europe and the destruction of social stability. Detractors see a policy that encourages overproduction, with surpluses dumped on world markets and misallocation of scarce resources away from more profitable uses--in short, they view the CAP as a poster child for agricultural distortion. Some facts are undisputed. Western Europe has transformed from a place of devastated infrastructure and productive capacity in the aftermath of World War II to the home of some of the world's most sophisticated economies. Economic integration, the development of a single market within the EU, and a common approach toward external trade are widely agreed to have contributed *The author is grateful for excellent research support by Uli Kleinwechter and Teresa Rojas Lara of Humboldt University; for invaluable help with data compilation by Johanna Croser, Esteban Jara, Marianne Kurzweil, Signe Nelgen, Francesca de Nicola, Damiano Sandri, and Ernesto Valenzuela; and for helpful comments from workshop participants and Stefan Tangermann. The working paper version of this chapter (Josling 2008a) contains additional background material and appendix tables. 115 116 Distortions to Agricultural Incentives: A Global Perspective to that success.1 Simultaneously, the agricultural sector has gone through a dra- matic transformation, from isolated national markets and traditional produc- tion methods to an integrated EU-wide market and technologically advanced farms. Transforming agriculture in this way within two generations is a remark- able feat. The transformation, though, has had positive and negative impacts. On one hand, it has encouraged investment in agriculture and sheltered producers from foreign competition. On the other hand, it has encouraged the production of high-cost commodities that have shrinking markets. This chapter focuses on agricultural policy developments in Western Europe and their consequent distortions to the economies within the region. It attempts to answer four related questions. · Has the level of protection, and the consequent distortion in resource use and consumer purchases, been increasing or decreasing over the five decades from 1955­2007? · How has the formation and subsequent enlargement of the EU influenced patterns and trends in agricultural protection? · What have been the main drivers of agricultural policy and to what extent have external influences impacted those forces? · What can one say about the future trends and protection levels in Western Europe? The chapter differs in some respects from the others in this book in several ways. First, Western Europe, besides encompassing a large number of countries, also contains different agricultural capacities, climatic conditions, economic structures, and political views. The differences among these countries are often as important as the similarities, and therefore generalizations are difficult to make. Inevitably, much of the focus is on the EU as an aggregate unit, rather than on each individual member country, though this misses some of the richness of the diversity of conditions. But from the viewpoint of the rest of the world, it is instructive to assess the aggregate impact of policies implemented in the EU and more broadly in Western Europe. Second, the process of economic integration in Western Europe has been more intense and comprehensive than in any other region, developed or developing. Though agriculture has been incorporated into the process of integration more fully than in most other regions, there are still differences in this regard within European countries. This means that the process of integration plays a much greater role in Western Europe than in other regions in explaining the pattern and trends in distortions. And as membership of the European Community (later the EU) grew, countries that previously had autonomous policies adopted the CAP. Western Europe 117 The number of agricultural policies in Western Europe has thus declined over the 50-year period considered by this study. Third, the external aspects of agricultural policy have played a more signifi- cant role in developing domestic policy in Western Europe than in most other regions. This is in spite of the fact that the CAP has often appeared resistant to pressures from abroad. The narrative of the past 50 years of agricultural policy in Western Europe is closely linked to the development of trade rules for agricul- tural products in the General Agreement on Tariffs and Trade (GATT) and the subsequent obligations undertaken as a result of the Uruguay Round and the resulting transformation of the GATT into the World Trade Organization (WTO). Other trade agreements, notably the obligations to former colonies through successive Lomé Agreements, have played a significant role in driving some commodity policies within the CAP. In the EU, the separation between domestic and foreign policy in the area of agriculture has always been blurred. Western European countries that are not members of the EU have been influ- enced by many of the same influences. Agriculture in Western Europe, 1955­2007: An Overview Agriculture in Western Europe enjoys a degree of diversity that reflects a wide vari- ety of soils and climatic conditions, the latter of which range from arid Mediter- ranean regions to the Arctic Circle. Superimposed on this natural diversity is the complexity of different social, economic, and political conditions in the 18 coun- tries that are the subject of this chapter.2 History has played a major part in creating this patchwork, particularly the different paths that countries took from feudal- ism to independent farming units and the inheritance laws that influenced the extent to which land ownership was transmitted from generation to generation. Average farm size varies considerably in the countries of Western Europe, in turn reflecting the relative political and social importance of landowners and small farmers. By the late 19th century, these various factors had determined the struc- ture of farming in the Western European region that is still visible today. The total utilized farm area in the 15 members of the EU in 2004 (EU-15) was 129 million hectares, distributed over 6.3 million holdings with an average size of 20.2 hectares.3 The sector employed 6.2 million persons, representing 3.8 percent of civilian employment in the EU (Eurostat 2006, table 2.0.1.2). Including the three countries not in the EU in 2004 (Iceland, Norway, and Switzerland), agriculture rep- resented 4.9 percent of the labor force, down from 30.0 percent in 1950. The value of output from these farms was 300 billion and gross value added was 155 billion (at distorted prices), or 1.6 percent of total gross domestic product (GDP) for those countries.4 118 Distortions to Agricultural Incentives: A Global Perspective Total Western European agricultural output has increased over the past 50 years by about 2.2 percent each year, a slower rate than for other sectors of the economy.5 As a result, the share of agriculture in GDP declined from 13.6 percent in 1955 to 2.9 percent in 2004. Only in Greece and Iceland is the share of agricul- tural output in GDP above 5 percent. Spain, Portugal, and Finland have agricul- tural sectors that contribute between 3 and 4 percent of GDP, while in Germany and the United Kingdom the share is now less than 1 percent. Participation in agricultural activities is highest in Greece and Portugal, with shares in excess of 12 percent of the active labor force. In Finland, Ireland, and Spain more than 5 percent of the labor force works in agriculture. By contrast, only about 1 percent of the United Kingdom's labor force participates in agricultural pursuits. In the immediate postwar period, 1949 to 1959, productivity growth in Western Europe's agricultural sector compared favorably with that in the manu- facturing, with output per worker in agriculture increasing by more than that in manufacturing in most countries. Productivity growth came from a combination of output increases as a result of mechanization and modernization, and the outflow of labor as other sectors absorbed rural workers. Though the productivity increase slowed somewhat in subsequent decades, it remained a key component in the development of the agricultural sector and the sector's role in postwar reconstruction. The strong farm productivity increase and accompanying farm labor force decrease is striking, and the decline in the share of agricultural workers in the labor force continues to the present day. The number of full-time equivalent "annual work units" employed in agriculture in the EU-15, as calculated by Eurostat, fell from 8.6 million to 5.9 million beteween 1991 and 2004 (Eurostat 2006). Structural change has also been rapid in European agricul- ture. The rate of consolidation of farms has risen over the past five decades, but the average size of farms still varies widely among countries, with the United Kingdom and Denmark having the largest farms, at 57 and 55 hectares per farm, respectively, and Italy and Greece having the smallest farms, at 7 and 5 hectares, respectively (Eurostat 2006). The countries of Western Europe also differ dramatically in terms of impor- tance of agriculture to international trade. In 2004, agricultural exports accounted for more than 10 percent of total merchandise exports in three countries: Denmark (18.7 percent), Greece (19.9 percent), and Ireland (11.6 percent). By contrast, they accounted for only 2.5 percent of German exports, 3.0 percent of Swedish exports, 2.8 percent of Finnish exports, and 4 percent of exports from the United Kingdom. Agricultural products, meanwhile, accounted for 6.1 percent of all imports to and 6.0 percent of all exports from the EU-15 as a whole. The EU-15 region ran a net deficit of 3.4 billion with respect to the rest of the world in food- stuffs and other agricultural products in 2004. Western Europe 119 In Western Europe, as elsewhere, agriculture has had to compete with nonfarm sectors for labor and capital. Growth rates of the manufacturing and service sec- tors therefore have been major influences on the economic health of the farming sector. In general, agriculture has provided an outflow of labor--both directly, as farmers and farmworkers become part of the industrial workforce (either by migration or by devoting more of their time each year to nonagricultural employ- ment), and indirectly, by offering less attractive employment to young people in rural areas. Though the outflow has been in progress for decades, the postwar period has been remarkable in the magnitude of the exodus. Obtaining capital within the agricultural sector, on the other hand, has been less problematic, as farmers typically have been able to raise funds in the financial markets, particu- larly through dedicated rural lenders and retained earnings. Though rates of return have not been high, nonpecuniary satisfaction and a lingering feeling on society's part that domestic agricultural activity enhances food security have encouraged rural investment. Sustained capital flows into the agricultural sector have resulted in significant transformation and modernization throughout Western Europe in the six decades since World War II. While the continent still has pockets of traditional farming, particularly in the south, in general, it now has a high level of technical expertise and moderate-sized farms. The transition has allowed Europe to become more internationally competitive in the second half of the postwar period compared to the first half. One major link between the agricultural and nonfarm sectors is through cur- rencies. Strong export performance by the industrial sector tends to appreciate the exchange rate and reduce the domestic cost of commodities, the prices of which are set in international markets. Thus, agriculture in countries with strong cur- rencies tends to be under pressure from reduced price levels as a result of export success in the nonfarm sector. For countries with weak currencies, exchange rate developments tend to raise the price levels of imports and exports, so the agricul- tural sector faces less competition from abroad--although governments may take action to lower domestic food prices. Foreign exchange market developments have had major impacts on the competitiveness of the agricultural sector in some countries, and have played a significant role in policy making in the EU. Agricultural Policy Prior to the Mid-1950s Current Western European agricultural policy reflects economic and social condi- tions of rural areas and political realities of the day, but many of the underlying factors are deeply rooted in experience and history. While some of this experience is shared among the countries of Western Europe, much of it is specific to the ways in which the countries concerned reacted to historical trends and events. 120 Distortions to Agricultural Incentives: A Global Perspective Many of these events were a product of the broad economic and political develop- ments in the 19th century, during which the pattern of land ownership was estab- lished and transportation and education systems extended into rural areas. But policy also influenced the reaction to these developments and led to significant disparities among neighboring countries. In addition to the social and political conditions that governed past agricul- tural policy in Western Europe, two other factors have been pervasive: the colonial experience of the countries concerned, and the reaction of those countries to the Industrial Revolution.6 The United Kingdom, with an extensive empire from which it could import both tropical and temperate agricultural products, was in a good position to benefit from trade. As the leader in both the technological revolution in agriculture and the industrialization of manufacturing processes, a policy of low-priced food played to the strengths of the economy. By contrast, Germany (and the numerous small states that preceded the creation of the unified Germany) had few overseas territories and lagged the United Kingdom in manu- facturing technology. As a result, German agriculture remained a protected sector. Political ideas reinforced these differences. In the United Kingdom, the free trade movement won widespread following by promising better living conditions for the urban workforce. Though landowners resisted, they lost ground over time to manufacturing interests. German intellectuals, however, pushed an infant indus- try strategy, which large landowners found to be in line with their own interests. By the end of the 19th century, significant differences between the United Kingdom and Germany had emerged in the prevailing economic paradigm and the agricultural policies that supported it. All of European agriculture was impacted by the growth of trade in temperate agricultural products from the New World in the 1870s, made possible by the opening up of new territory and falling rail and ocean transport costs. The intro- duction of refrigeration was important in that it began to make the transport of livestock products and grains from the Americas, Asia, and Australia profitable. In the case of the United Kingdom, the high tariffs that had been embodied in the Corn Laws had already been repealed in 1846. As a result, the political climate was such that manufacturing interests prevailed over agrarian pressures, and agricul- ture shrank in the face of overseas competition. Much of the cereals and meat came from colonies, and could be paid for in pounds, so import substitution was not a priority. In addition, the structure of farming in the United Kingdom was generally more able than in other countries to withstand the low prices.7 As a result, pressure for protection was less than in many other European countries. Other countries followed the lead of the United Kingdom and reluctantly accepted the benefits of cheap grain from the New World. Denmark stands out as the country that embraced the new relative price structure most completely, and Western Europe 121 the Netherlands reacted in a similar way. Livestock farming received a boost from the lower feed costs, in particular the rearing of cereal-fed livestock such as pigs and chickens. In addition, Denmark inherited an efficient farm structure from the early 19th century, and developed a cooperative system that fit well into the live- stock economy that flourished on the cheap grain of the 1890s.8 French agricul- tural markets, which were also relatively open to trade in the middle of the 19th century, increased protectionism sharply with the tariff of 1881, which imposed high duties on livestock imports (Tracy 1989). The level of protection peaked with the Meline tariff of 1892 and remained high until World War I. However, industrial tariffs were also increased over that time, masking the dis- tortive impact of agricultural protection. German farming benefited significantly from higher protection in the last quarter of the 19th century. As livestock from the United States and grain from Russia threatened to depress domestic prices, tariffs were introduced in 1879, ini- tially at a moderate level, though they increased over the next decade. Chancellor Otto von Bismarck was adamant that farm imports be controlled, and he presided over a bitter trade dispute with the United States over the sanitary con- ditions under which U.S. pork was produced for export to Germany (Snyder 1945). Upon Bismarck's ouster in 1890, Germany briefly returned to more open agricul- tural trade, albeit in the face of opposition by Prussian landowners (Tracy 1989). For the next 20 years, liberal and protectionist economic paradigms clashed, while disparate views on the desirability of industrialization kept the issue of agricultural polices at the political forefront. Such 19th-century differences between the "adjusters" and the "protectors" remained through the first half of the 20th century. The prime factors that played a role in the development of agri- culture during that period include the economic impact of World War I and the Great Depression. After World War I, the United Kingdom attempted to expand production by granting farmers subsidies to supplement their market earnings. These deficiency payments, introduced in 1917, were accompanied by a liberal import regime for farm products (except sugar). An attempt to introduce price guarantees in 1920 was repealed the following year. Though protection emerged in the 1930s as a result of depressed world prices, the effect was mitigated by imperial preferences that allowed agricultural products in from the Dominions (Australia, Canada, New Zealand, and South Africa) and the colonies. Domestic marketing became the focus of farm policy, and was institutionalized through the introduction of marketing boards in 1931, several of which remained in place until the 1970s. French agriculture was badly damaged during World War I, with both infra- structure and productive capacity destroyed. Agriculture did not get much assis- tance from trade protection in the 1920s, as protection in the nonfarm sector was 122 Distortions to Agricultural Incentives: A Global Perspective higher than that in agriculture. However, the reaction to the Great Depression was to introduce quotas on agricultural imports and intervene in the domestic market. State marketing was established through institutions such as the Office National Interprofessionel du Blé, founded in 1936. Germany, meanwhile, instituted the Ministry of Agriculture in the wake of the war, and the Weimar Republic attempted to take over responsibility for agricul- tural policy from the states. Rapid industrialization, however, reduced the signifi- cance of the farming sector, and low prices during the Depression took a heavy toll on farm incomes (Roesener 2000).9 The Third Reich attempted to capitalize on the decline of agriculture by promising state protection and higher social standing for the rural population, introducing policies designed to promote self-sufficiency and increase the control of the state over marketing and trade of agricultural products. Production, however, did not reach planned targets, and at the outbreak of World War II in 1939, the level of farm output was no higher than in 1935. Labor shortages and the need to keep urban prices down undermined attempts by the National Socialist German Workers' Party to return Germany to its rural past. Denmark, which remained neutral in World War I, expanded its agricultural sales to both the United Kingdom and to Germany in the postwar period (Tracy 1989), despite the increase in German tariffs in 1925. The Depression, however, hit Danish livestock production by 1931. Efforts to improve trade relations with the United Kingdom (to offset the preferences granted to competitors such as New Zealand) and with Germany were partly successful, and Denmark eventually had to compromise on its traditional liberal trade policy and introduce tariffs on grain. Later in the decade it began subsidizing producers of livestock products and restricting production by means of marketing quotas. Eventually, in 1938, grain imports were discouraged by compulsory mixing requirements for millers. Along with that of the Netherlands, Denmark's agricultural experiment of trading at world prices appeared to be at an end. The influence of all these political and economic trends in Western European agriculture in the early 20th century was disrupted by World War II. To an extent probably not experienced in any other region, World War II had a significant impact on the agricultural sector. Not only did the war itself cause havoc to infra- structure and destroy productive land, but the sector was drawn in to the war effort to provide food and industrial raw materials. The United Kingdom, with its vulnerability to blockades of imports, began to mobilize the civilian population to grow more food. In the postwar period, pro- duction rebounded rapidly to its pre-Depression levels, assisted by the introduc- tion of guaranteed prices in the 1947 Agriculture Act. The need for additional domestic production was premised in part on the chronic shortage of foreign exchange in the early postwar period, as exports failed to finance the imports Western Europe 123 needed for reconstruction and to service the debt that had been accumulated during the war. The fact that the pound was a global currency led policy makers to reject any efforts toward devaluation. Successive governments thus pursued a policy of high domestic prices as a way of discouraging agricultural imports. Occupied France, under the Vichy Regime, attempted to restore the country's agricultural "destiny."10 However, devasted infrastructure delayed the restoration of the sector in the immediate postwar period. France began a period of national planning, which included goals for the agricultural sector. The first agricultural plan (1948­52) called for an expansion of exports, while the second plan (1954­57) established a fund for market intervention. As economic integration became a reality, the opening up of markets in Europe to French farm products became an important goal. The salvation of rural France was to be in exporting products to the industrial heartland of Europe, assisted by some protection against overseas suppliers. Though Germany's agricultural sector maintained production levels through most of World War II, food shortages emerged in 1944. By the end of the war, nutritional deprivation was an acute problem and hunger relief became an inter- national issue. Wartime controls over trade were maintained and the new govern- ment in West Germany (established in 1949) encouraged domestic production by price incentives linked to production costs, with little regard to competitiveness. By the time Germany faced the prospect of opening its agricultural market to French and Dutch farm products, the predominance of inefficient small-scale German farms was a highly contentious political issue. Policy Distortions 1955­2004: A Chronology The current map of agricultural protection was drawn in the early postwar period, as the countries of Western Europe struggled to rebuild their economies and restore commercial and political relationships. The defining moment in the devel- opment of agricultural policy was undoubtedly the formation of the European Economic Community (EEC) in 1957. Indeed, no analysis of the distortions caused by agricultural policy in Western Europe can avoid a detailed examination of the development of the CAP and the enlargement of the EEC from the original six countries (Belgium, France, West Germany, Italy, Luxembourg, and the Netherlands) to the 15 that were members of the EU in 2004.11 The stages in the enlargement of the EEC/EU provide the backdrop to any timeline of the descrip- tion and tabulation of the changing pattern of direct and indirect distortions to farm incentives. Not only did each enlargement cause an examination of the EEC/EU policy toward agriculture, but it changed the reach of that policy by including more farms under its umbrella. Thus, the nature and magnitude of the 124 Distortions to Agricultural Incentives: A Global Perspective incentives faced by domestic producers and consumers of primary agricultural and processed food products in Western Europe were affected as much by the adoption of the CAP as through more traditional agricultural policy processes. The other side of this coin is that the number of countries in Western Europe that remain outside the EU has steadily declined.12 Three of these countries-- Norway, Iceland, and Switzerland--still have autonomous agricultural policy despite being members of the European Free Trade Association (EFTA) since 1960. Though these EFTA members represent a declining share of total agricul- ture in Western Europe, their policy decisions are of particular interest. While it would be too simplistic to say that they represent the "control group" in the exper- iment of designing and implementing a common policy for agriculture, they certainly offer important lessons. In this section, agricultural policy distortions are discussed in the context of each of the five decades from 1955 to 2004. The discussion relates this timeline to the evolution of the EU as it expanded membership and of the CAP as the dom- inant vehicle of support for Western European agriculture. Emphasis is placed on the levels of support and key policy prices in the prospective member countries relative to that of the EU as a whole. As each new group of countries gained access to the EU, so the geographical reach of the CAP changed. In turn, the CAP became the focus of much of the external pressure that faced the EU as it grew in significance in world trade. But the macroeconomic conditions, including infla- tion, exchange-rate fluctuations, and nonfarm growth in Western Europe, were a vital backdrop to the agricultural policy decisions and need to be considered in parallel with the more specifically agricultural aspects of the development of the CAP.13 1955­64: Agriculture in a period of rapid economic growth The period of postwar reconstruction was followed by a rapid expansion of economic activity during the 1950s and into the 1960s. This was the period of the German and Italian "miracles," as both these economies grew at rates far above those of the United Kingdom and France. In turn, this growth provided the demand for consumer goods that allowed neighboring countries to expand their own economies. Trade within the newly created EEC expanded rapidly as trade barriers in manufactured goods were removed over the period 1957­64. The formation of the EEC in 1957 set the scene for the integration of Western European agricultural markets.14 For this to happen, agricultural trade had to be included in the move toward the free flow of goods among the original six coun- tries.15 Though this was eventually incorporated, such an agreement on free inter- nal trade was possible only by erecting a protective border around the EEC to Western Europe 125 shelter agriculture from foreign competition. The development of a CAP was therefore the result of a compromise between those who wanted agriculture to be a full part of the free internal market and those who preferred a more interven- tionist system. A common market organization (CMO) was developed for each of the main commodities. Administered prices were related to a "target" price level for each commodity. Imports were allowed in at "threshold" price, calculated on the basis of the target price and transport costs. A "variable levy" was charged on the basis of the gap between offer prices on the world prices and the threshold price. Excess production could be taken off the market by national agencies at "intervention" prices--again, fixed relative to the target price.16 Prices were set each year by the Council of Ministers upon the recommendation of the European Commission. The period 1955­64 was one of policy initiatives for agricultural integration as well as the restoration of trade flows across Europe. National farm prices differed considerably across the six countries of the EEC and also between the six coun- tries and neighboring countries that had chosen not to join in the integration experiment. The path toward more integrated internal markets for agricultural goods proved rocky. Different farm structures, commodity balances, and histori- cal protection levels created a minefield for those advocating a free internal agricultural market and common prices. Eventually, in 1962, common rules for agricultural markets were agreed upon. A transition period was instituted until 1967, when all prices were supposed to have been harmonized. A look at selected Western European countries illustrates the wide differences in circumstances over this period, and illustrates the difficulties of achieving a common policy. West Germany, upon its creation in 1949, found itself with a structure of small farms because the most productive and largest farms were now in East Germany. Meanwhile, German industry was being encouraged to expand into other markets in the region and overseas, and its success added to the strains on agriculture. Tight controls over imports of cereals and the use of marketing boards to regulate the domestic market kept agricultural prices high, while pro- duction was encouraged from every small farm. The choice of a common price for wheat (and other grains) in 1961 was a major political issue in Germany: soft wheat prices were about US$110 per ton at that time, a level matched only by Italy among the six EEC members.17 In the end, the German government resisted calls for the common price to be set at a lower level, thus setting the stage for the devel- opment of surplus production in the EEC within the decade. Though France had farms larger than Germany's in the Paris Basin, it was hampered by the remnants of feudal strip farming in Normandy and Brittany and low productivity in the Massif Central and the Midi. Cereal farming in particular had recovered from the wartime disruption and by the 1960s had surpluses to 126 Distortions to Agricultural Incentives: A Global Perspective send to the deficit areas of Europe, including Germany. Soft wheat prices in France averaged only US$81 per ton at the start of the CAP transition period, and thus had to rise sharply to reach the agreed price levels in the young CAP. The Netherlands, along with Denmark, had an efficient farming structure based on milk, poultry, and eggs. While the Netherlands was already trading agricultural products with Germany, it sought increased market access. The fact that Denmark stayed out of the EEC, preferring to maintain its connection to the British market, gave Dutch farmers a welcome degree of preference. The Netherlands and Belgium, where the soft wheat price level in the early 1960s was between the high in Germany and the low in France, experienced some significant increases in the cost of animal feed in the movement to common prices. Italy shared with Germany some significant structural problems, with small farms dominating the southern part of the country and relatively high-cost cereal production in the center and the northern part. However, expansion of the live- stock sector in the north, based on imported grains, linked Italy's agricultural interests with those of the Netherlands. In fact, in the move to standardize prices, Italy was allowed to maintain imports of feedstuffs at a lower tariff than countries in Northern Europe. The Western European agricultural market was in effect split by the decision to push for a CAP for the six members of the EEC. The United Kingdom chose to stay outside the EEC, concerned about the element of "supranationality" intro- duced in the Treaty of Rome, and the CAP was negotiated without the input of the major Western European food import market. In any case, the EEC was not entirely in agreement over the prospect of the United Kingdom joining.18 The United Kingdom had its own troubles, with macroeconomic imbalances proving difficult to control. The balance of payments was chronically in deficit, and remained a problem until the devaluation of 1967. The 1957 Agriculture Act, which introduced deficiency payments as a way of maintaining high producer prices while keeping consumer prices close to world market levels, was an impor- tant aspect of U.K. foreign policy, because much of the imported food came from former colonies. Import patterns were so significant from a policy perspective that they contributed to the United Kingdom's decision not to join the EEC. However, in an attempt to show leadership among other countries that chose not to partici- pate in the EEC, the United Kingdom sponsored the EFTA in 1960. Seven coun- tries signed up to this "integration-lite" experiment, which differed from the EEC both in terms of its lack of a common tariff and supranational institutions and because it excluded agricultural (and fisheries) trade from its provisions. Facing the prospect of having to phase out preferential access from Dominions and former colonies, as well as the recurrent balance of payments problems and the fear of food-price-led inflation, led to significant debate on the costs of Western Europe 127 agricultural protection. Sparked by a paper by Nash (1955), there were several attempts to calculate the value of British farm output at world prices and compare the result with actual farm values. Several years later, Nash and Attwood (1961) repeated the same calculation using Danish prices, where distortions were notice- ably less, to value British production. McCrone (1962) elaborated these studies into a comparison of 13 countries in 1955­56. Howarth (1971) followed the same method and added an estimate for 1966. According to Howarth's estimates, agri- cultural protection levels increased markedly from 1956 to 1966, at a time when trade in nonagricultural goods was being liberalized.19 More evidence of the increase in protection was found in a study by Anderson and Hayami (1986). The level of protection (as measured by the nominal protec- tion coefficient) was calculated for eight Western European countries for the period 1955­1980. Switzerland stood out as having the highest level of protection, albeit a level that did not increase over the first decade of the period, 1955 to 1965. By contrast, estimated protection did increase in Italy, Sweden, Germany, and the Netherlands over this decade. In France and the United Kingdom, protection actually decreased in the decade up to 1965. The estimates made for the present study broadly confirm the conclusions of the Howarth and the Anderson and Hayami studies.20 Figure 3.1a shows the nom- inal rate of assistance (NRA), the percentage by which a product's domestic price exceeds the price at a country's border), including non-product-specific (NPS) support and aggregated over all commodities, for four of the original six members of the EEC.21 At the time when the Treaty of Rome was being discussed, assistance levels were modest (by later standards) and not too widely dispersed.22 The CAP was "launched" in 1962, with prices that reflected political compromise rather than economic foresight, and by the time the common price regime was in place European agriculture was operating on a price plateau for the major products that was well above world market levels. The corresponding calculations of the NRA for the EFTA countries (those that chose the path of less institutional integration and no free trade in agriculture) are shown in figure 3.1b. Austria, Denmark, and the United Kingdom had NRA levels in the same range as in the EEC. Portugal was in effect taxing its agricultural sec- tor by holding prices below their full market value--a stance that reflected both its own political structure and its level of development.23 Sweden increased its assis- tance to agriculture over the decade, and gave incentives to its agricultural sector to an even greater extent than the countries that formed the EEC. But the coun- tries with the highest level of support by far were Norway and Switzerland, where domestic prices were around twice as high as in other EFTA countries.24 The pattern for these countries is what one would expect from their experience in the prewar period. The United Kingdom, the Netherlands, and Denmark had 128 Distortions to Agricultural Incentives: A Global Perspective Figure 3.1. NRAs to Agriculture, EU-6 and Western European Average, 1956­64 a. EU-6 members 140 120 100 80 percent 60 40 20 0 20 1956 1957 1958 1959 1960 1961 1962 1963 1964 Germany Netherlands Western European average France Italy b. Other European countries 300 250 200 150 percent 100 50 0 50 1956 1957 1958 1959 1960 1961 1962 1963 1964 Switzerland Norway Sweden United Kingdom Denmark Austria Portugal Source: Anderson and Valenzuela (2008), based on authors' spreadsheet, which draws heavily on OECD (2008). Western Europe 129 lower levels of protection, reflecting their history of imports of grains, and France emerged in the decade leading up to World War II as an exporting country with lower production costs and less opportunities for protection at the border. Indeed, one of the main attractions of the EU for France was the opportunity it presented to sell French products abroad without having to bear the costs of sub- sidies. Switzerland, Norway, Sweden, and Germany have higher levels of protec- tion, as one might expect from predominantly importing countries attempting to generate acceptable incomes for farmers. The countries that made up the EEC increased their protection on average over the decade from 1956­64, explaining in large part the chorus of complaints from overseas suppliers about the protection- ist nature of the emerging CAP.25 If the United Kingdom stayed out of the EEC in part because the grouping was more protectionist in terms of agriculture than itself, the same could not be said of all the other EFTA members.26 Protection levels in Norway and Sweden, both of which joined the EFTA in 1965, were high, as they were in Finland (Gulbrandsen and Lindbeck 1973). In fact, the prospect of joining the EEC was seen as a threat to agriculture in these countries, as they would have had to reduce prices and face competition from the grains produced by France and the livestock produced by the Netherlands. Only Denmark looked favorably on the prospect of expanding its sales to the EEC countries, but in the end it chose to remain aligned with the United Kingdom, its traditional market for farm products. Another country that was among the "charter members" of the EFTA, reflecting its historic trade ties with the United Kingdom, was Portugal. The country had been under a dictatorship since 1928, when the military suspended democratic processes, and the economy was being run on a corporatist model with strong central control. Though Portugal's African colonies provided an income and Brazil remained a source of capital and a link with Latin America, Portugal was outside the mainstream of Western Europe and remained a relative backwater until its return to democracy in 1974 (Avillez, Finan, and Josling 1988; Corkhill 1995). Spain, also shunned by many European governments in the postwar years, was additionally hampered by a rigid corporatist economic policy. Innovation and social mobility were discouraged and central and southern landed interests domi- nated political life well into the 1960s. The Franco regime controlled wages, prices, and trade, and large state corporations were prefered over smaller urban enter- prises. Spain was excluded from the Marshall Plan in 1948, and it was not until 1958 that agricultural output regained the level of 1929, just prior to the Great Depression and before the Spanish Civil War. Farm structure, however, was better in Spain than in many parts of Western Europe, and by the 1970s investment in fruit and vegetable production had begun to increase output. The transition to democracy in starting in 1974 offered hope that accession to the EU was possible, 130 Distortions to Agricultural Incentives: A Global Perspective though economic growth stalled for much of the period from the death of Franco to accession to the EU (Lieberman 1995). For both Spain and Portugal, the oil price shocks of 1973 and 1975 had a major detrimental effect on their economies. Between the mid-1950s and 1964, Iceland was predominantly an agricultural and fisheries economy, with meager amouts of arable land and extensive imports of cereals. Dairy and sheep production was able to fulfill domestic and export needs of wool and meat. Though an agricultural development plan for 1951­60 tried to increase farm size (to ensure that every farm had at least ten hectares), it did not succeed (OECD 1967). A law passed in 1960 instituted fixed wholesale and retail prices for major food items and offered export subsidies to dispose of sur- pluses. Consumer subsidies were also used in this period. Under a revised agricul- tural plan of 1963­66, however, public investment in agriculture decreased as the burden of supporting high-cost production became apparent. In the 1950s, Greece was still a largely agricultural country, where cotton and tobacco represented a major portion of exports. Most temperate-zone foods were imported, and domestic production was hampered by low productivity and a fragmented farm structure that was in part the result of prewar land reform. Association with the EEC in 1961 promised to expand exports, and the govern- ment began to increase incentives for producers and institute minimum prices for domestic output (OECD 1967). But this program was carried out in the context of an economic strategy that supported manufacturing sector as the engine of growth (Pepelasis et al. 1980). 1965­74: Agriculture in a period of macroeconomic instability Price levels set under the CAP were at the high end of the range of existing prices in EEC members states, while strong upward convergence was evident over the period from 1962 to 1967. Price differences within the EEC reflected the gradient of prices from surplus to deficit areas, though this was modified by setting inter- vention prices in the surplus areas that did not always reflect transport costs. Achievement of a single support price in 1967, set in units of account with a value equal to the U.S. dollar, was a major political success, but the uniformity of the CAP was short-lived. Currency instability undermined the realization of an open internal market for agricultural products and a common price support scheme. The devaluation of the British pound sterling in 1967 heralded a period of financial instability in Europe. The revaluation of the Deutsche mark in 1968 and the devaluation of the French franc the same year caused havoc in the operation of the CAP, leading to the introduction of artificial "green" exchange rates for the conversion of prices decided in Brussels by the Council of Ministers into local Western Europe 131 currencies (Josling 1970; Josling and Harris 1976). The currency changes in 1968 also exposed the problems of administering regional price differentials. Earlier in the year, grain from France was flooded into the German market as a form of arbi- trage (the relative intervention prices were different from the relative values of the Deutsche mark and the French franc on money markets), causing serious storage problems. When the currency changes took place, compensation for what would have been price declines in Germany was granted by ad hoc tax relief for German farmers, and a system of border taxes and subsidies was introduced to offset the exchange rate changes. This process was extended to other EEC members by con- verting common prices through "green rates" that lagged the developments in the market rates. The border tax/subsidy regime, known as monetary compensatory amounts (MCAs), was imposed on both internal and international trade. The effect was to undermine the central concept of free trade within the EEC, to chal- lenge the notion of common price levels, and in essence give some control over price policy back to the member states (Heidhues et al. 1978). The level of protection in the United Kingdom remained low over the second half of the 1960s relative to major European countries (FAO 1973). Nevertheless, the protectionism debate heated up over the desirability of expansion of U.K. agriculture to help offset the balance of payments shortfall. Though the 1967 devaluation helped to correct a misalignment of the exchange rate, inflation reduced the competitive advantage of British exports. Agriculture was caught up in a wave of "buy British" sentiment in the late 1960s, and price supports were increased in the Annual Price Reviews. The debate on "import saving" exposed the economic costs of agricultural protection and led to a reevaluation of the role of the relatively small agricultural sector in the U.K. economy. The government moved to stabilize the costs of the "deficiency payment" program employed since 1947 by negotiating minimum import prices for cereals with exporting countries. Later, the policy of allowing relatively free import access was replaced by a system of variable levies, to generate some revenue from imports. But it was the prospect of entry into the EEC that dominated the agricultural policy debate in the United Kingdom in the late 1960s. The first enlargement of the EEC took place in 1973, with the accession of Denmark, Ireland, and the United Kingdom. The CAP, however, was a major prob- lem in the negotiations. On one hand, the United Kingdom was under pressure to accept the CAP as part of the "acquis communautaire," the accumulated regulations and directives of the existing EEC, and the enlargement process was supposed to focus merely on changing the language of regulations so as to reflect the new mem- bership. Any change in the policy itself was to be deferred until the new countries joined. On the other hand, at the political level, the United Kingdom was deter- mined to protect the preferential access of its former Dominions and colonies, 132 Distortions to Agricultural Incentives: A Global Perspective though this was unacceptable to the European suppliers for whom the prospect of free access to the United Kingdom, a large importer, was attractive. In the end, the compromise was to negotiate some assurances for traditional suppliers (New Zealand for specific quantities of meat and dairy goods, for example) and to incorporate the former colonies into the arrangements that had already been set up for those of Belgium, France, and the Netherlands. EEC negotiations with Ireland and Denmark were of a different nature. For those two countries, traditional exporters of livestock products to the United Kingdom, the opportunity to diversify to the European market was a welcome prospect. The higher farm support prices in the EEC were not a major hurdle, and tilted the balance of economic advantage in favor of accession. Norway also participated in the negotiations, and an agreement was reached among the governments. But when the question of Norwegian accession was put to a popular vote, the concerns of Norwegian farmers and fishermen prevailed. The vote went against membership, and the government withdrew from the tentative accession agreement. Internal debates about accession to the EU often have been focused on the impact of price changes, particularly for farmers and consumers, as EU prices for most farm products had been considerably higher during the 1960s than in the applicant countries (except Norway). But the aspiring members had agricultural support systems that were more in touch with the conditions on world markets. So the high world prices in 1973­75 for many commodities masked the full impact of the price increases expected from accession. As a result, the additional distortions due to the CAP were relatively small over the first two years of the EU-9. The countries that remained in the EFTA made hurried bilateral arrangements to preserve at least some of their access to the United Kingdom and Denmark. However, these agreements did not cover agricultural products, which had been excluded from EFTA trade liberalization. In the case of Portugal, some conces- sions were agreed: in 1972, a trade agreement was signed between the EEC and Portugal to allow for continued access of Portuguese exports into the bloc. But the domestic farm policies of Sweden, Switzerland, and Norway, along with Finland and Austria, continued to be determined largely on a national basis. These developments in rates of assistance are shown in figure 3.2. The NRA for the EU-6 (and the Western Europe average, heavily weighted by the EU countries) hovered at a little less than 80 percent over the decade 1965­74, plunging at the end of that period as world prices rose in 1973­74.27 The entry of the three new members did not cause this drop, even though it brought some new constraints to the development of the CAP. Meanwhile, the EFTA countries that chose to stay outside continued with their own policy trajectories (figure 3.2b). Switzerland and Norway kept domestic prices at more than twice the level of those on world Western Europe 133 Figure 3.2. NRAs to Agriculture, EU-9 and Western European Average, 1965­74 a. EU-9 members 100 90 80 70 60 percent 50 40 30 20 10 0 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 EU-6 Western European average Ireland United Kingdom Denmark b. Other Western European countries 300 250 200 150 percent 100 50 0 50 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 Switzerland Norway Sweden Finland Austria Portugal Source: Anderson and Valenzuela (2008), based on authors' spreadsheet, which draws heavily on OECD (2008). 134 Distortions to Agricultural Incentives: A Global Perspective markets, while Portugal continued to tax its farmers whatever the price levels on world markets. Rates of assistance to farmers in Sweden, Finland, and Austria were reduced somewhat in the early 1970s as the high world prices substituted for some of the protection given by policy interventions. 1975­84: Agriculture out of control The macroeconomic instability that followed the first global oil shock, in October 1973, had a significant impact on the level of policy prices set under the CAP. Farm input costs rose sharply and the real value of price supports declined. Politi- cians responded by increasing prices sharply to keep up with costs. In addition, the fear of worldwide food shortages made price restraint less attractive as a polit- ical argument. When commodity prices declined, the EU was left seriously uncompetitive in many temperate-zone agricultural products. The United Kingdom, which became a member of the EEC in 1973, was expected to assert a moderating influence on agricultural policy. Indeed, much of the opposition to the U.K. accession from countries such as France stemmed from the feeling that it would attempt to force its tradition of low market prices and high payments from the exchequer onto CAP policy. But these fears proved unfounded. The United Kingdom found a convenient way of keeping its own prices down through the medium of the "green money" system mentioned above, and turned its attention to limiting the budget cost of membership through a budget "rebate." The impact on agricultural distortions of the first enlargement of the EEC was therefore somewhat mixed. Protection levels went up in the United Kingdom and in Denmark and Ireland. But as a result of the transition arrangements, the United Kingdom was able for a time to avoid the impact of high EU prices. The accession period negiotiated for the United Kingdom (seven years) called for import subsi- dies paid by the EEC on farm products entering the United Kingdom. Coupled with the subsidies to offset the depreciation of the pound in the mid-1970s, the price of food rose by less that had been feared at the time of the entry debate. Denmark and Ireland made good use of their expanded opportunities for live- stock exports, selling to Europe and the United Kingdom. The green money system continued to add to the level of internal distortion in the late 1970s. In February 1979, a common price of 100 ECU28 by the Council of Ministers was equal to 110.8 ECU in Germany and 71.8 ECU in the United Kingdom. Thus, support prices were maintained in Germany at a level of 54 per- cent higher than that in the United Kingdom. Only Denmark eschewed the polit- ical convenience of masking exchange rates by the use of fictional green rates. The ECU system remained in place for another decade, undergoing modification as Western Europe 135 part of the single market of 1992, and was finally eliminated by the launch of the euro in 1999. Countries that remained outside the EFTA had the liberty of being able to run their agricultural policy without the need to comply with the CAP. EFTA had no direct impact on agriculture over this period. Switzerland maintained high prices in the late 1970s, with a support price for wheat of US$398 per metric ton, com- pared to US$200 in Germany and US$190 in the United Kingdom (Hallett 1981, based on numbers from the International Wheat Council). Any form of opening up of trade in grains between the EEC and Switzerland, therefore, would have been difficult to implement. Norway, which chose not to join the EEC in 1973, was able to pursue an autonomous policy based on the perceived need to retain population in the north- ern areas of the country. Protection levels were almost as high as in Switzerland, though Sweden resisted the temptation to farm the cold northern regions as a matter of national security. The experience of Finland fell somewhere in between that of its Scandinavian neighbors. Though it had an extensive area of high cost agriculture in the northern parts of the country, rural interests were not able to maintain such high levels of protection against imports as were granted in Norway. The EEC did welcome one new member over this period. Greece, long an asso- ciate member, was welcomed into full membership in 1981, following reestablish- ment of political freedoms. In agricultural terms, Greece posed no problems with respect to the temperate-zone products, as it was likely to increase its imports of them from the EEC as its own protection was withdrawn. But exports of Greek fruits and olives posed a problem for the region, and southern members argued (successfully) for extensive transition periods before opening up to Greek compe- tition. This set the scene for similar arrangements when Spain and Portugal fol- lowed Greece into the EEC five years later. In other regards, the experience of Spain and Portugal was different, as neither had negotiated associate membership status in the 1960s. Both countries had moved from dictatorships to democracy in the mid-1970s, but their agricultural policy was still focused on domestic concerns, including structural issues. Land reform, for example, was a serious concern in Portugal, following the 1974 revo- lution. The breakup of the large farms caused a significant drop in output of grains, a situation made worse by successive droughts. Prices, which were increased in an attempt to generate adequate income for the new class of small farmers created by land reform efforts, rose above those in Spain and the EEC. Dairy production was encouraged in the north and in the Azores. Though tomato processing became a significant export industry in the 1960s, other Mediterranean products were less advanced (Avillez, Finan, and Josling 1991). 136 Distortions to Agricultural Incentives: A Global Perspective By the mid-1980s, Portugal was not in a position to compete in Europe, and ended up needing a significant transition period to develop the institutional capacity to administer the CAP. In Spain, the period before accession also was one of economic stress and rela- tive stagnation. Cereal and livestock production lagged behind that of the rest of the continent, and marketing systems required modernization. Imports of feed grains and oilseeds, mostly from the United States, faced little in the way of trade barriers. Prices for other commodities--wheat, for example--did not differ greatly from those in the rest of Europe. But the fruit and vegetable sector in Spain, with support from the state, was undergoing a structural change--specifically, it had accelerated and wine and olive production. As a result, the prospect of substantial exports of these products loomed over the accession negotiations. The issue of the entry of Spanish farm products into the EEC took on particular political significance in Italy, France, and several countries in North Africa. Changes in the level of agricultural assistance in Western Europe over the period 1975­84 reflect these developments (figure 3.3). In the EU-9 as a whole, NRA levels increased steadily over the decade as a result of cost-based price deci- sions made under in an environment of rising inflation and lack of effective budget constraints. From a level of about 40 percent in 1975, the NRA increased to more than 80 percent in 1983. Overall support levels were very low for the applicant countries of Southern Europe, leading to political tensions over the adoption of the CAP by these countries. A transition period was needed both to cushion domestic consumers and to grant the producers in the EU-9 time to gear up for competition from Spain and Portugal. Of the EFTA countries, Switzerland, Norway, and Iceland continued to provide a high level of support to farmers, while Sweden, Finland, and Austria assisted their farmers at similar rates to those in the EEC. 1985­94: Agriculture as an international concern By the mid-1980s, the issue of domestic agricultural policy in Western Europe had become a central topic of concern in the multilateral trade system, for two rea- sons. First, high levels of border protection were retained to give a broad umbrella of protection against overseas competition under which the market orders could continue. Second, surpluses of cereals, meat, dairy products, and sugar, all of which other Organisation for Economic Co-operation and Development (OECD) countries produced for export, were growing. As a result, the CAP came under criticism abroad as a major cause of low world prices and at home for high sup- port costs and (at least in the United Kingdom) for high consumer prices (Tyers and Anderson 1992). Western Europe 137 Figure 3.3. NRAs to Agriculture, EU-12 and Western European Average, 1975­84 a. EU-12 members 120 100 80 60 percent 40 20 0 20 40 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 EU-9 Western European average Portugal Spain b. Other Western European countries 300 250 200 percent 150 100 50 0 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 Switzerland Norway Iceland Finland Sweden Austria Source: Anderson and Valenzuela (2008), based on authors' spreadsheet, which draws heavily on OECD (2008). 138 Distortions to Agricultural Incentives: A Global Perspective In 1984, the GATT began to discuss and the OECD to measure the extent of the distortions generated by domestic agricultural policy and by border instruments. The consensus was that for several markets, the impact of domestic policies spilled over to the world market in a way that caused a reaction by other countries, either to subsidize exports or to increase protection against imports. A common solution seemed to be the answer. If all domestic policies could be brought under control, and if the nature of border measures could be disciplined, then the situa- tion would be ameliorated. But the CAP was singled out as the policy that needed to change the most, and so the policy was firmly on the international agenda. Attempts to argue that U.S. support per farmer was significantly greater than in the EEC (as a result of larger farm size) garnered little sympathy. So long as the focus was on the impact on world markets, rather than the impact on domestic incomes, the CAP was subject to skepticism. Hence, the attitude of the European Union (EU)29 during the Uruguay Round was largely defensive. Eventually, in 1992, the MacSharry reforms forced the CAP to adjust from within and allowed the EU to agree to the negotiated strictures of the Uruguay Round Agreement on Agriculture (URAA) on export subsidies and domestic support payments. The 1992 CAP reforms marked a change in instrumentalities as well as price levels. In compensation for a price drop, farmers received a direct payment based on historical hectareage for cereals and oilseeds and a regional yield. The lower price for cereal-based animal feed led to an increase in the use of barley, wheat and corn, and helped to reduce cereal stocks. In addition, headage payments were introduced for beef and sheep, and the incentive to produce was reduced, as mar- ginal output increases were effectively sold at lower prices.30 The milk and sugar regimes, however, were not included in 1992 reforms. The adoption of the CAP reforms by the Southern countries (Portugal and Spain joined the EEC in 1986) posed few problems for the major crop and live- stock product producers, and did not raise the level of distortion overall. The main impact in those two countries was on the Mediterranean products--particularly olive oil, wine, citrus, and tobacco. For those products, the stimulus increased pressure on markets and hence calls for policy review. Several changes with respect to the position of the EFTA countries occurred between 1985 and 1994. They had been offered, and negotiated, an industrial free trade area (the European Economic Area, or EEA) with the EU. The EEA allowed free trade in manufactured goods and cooperation in regulatory issues: in effect, it extended previous bilateral agreements to include several aspects of trade that had been incorporated in the 1992 European single market. Though some quotas on agricultural goods were expanded, there was no progress toward the incorpora- tion of the rural sector in economic integration. The EFTA countries, which still did not have a say in EU decisions, were thus unable to influence regulations that Western Europe 139 would apply to them. Meanwhile, the politically neutral stance that previously prevented several EFTA countries from having close ties with the EEC became less important with the end of the Cold War. So, before the ink was dry on the EEA, Sweden made the decision to apply for full EU membership, and three other EFTA countries followed suit. Though the path toward EU accession was attractive to many countries of Western Europe, some ultimately did not make the journey. Norway, once again, chose to stay out. Switzerland also found that membership was problematic; even membership in the EEA was rejected in a referendum, particularly by the rural German-speaking Swiss cantons. In Austria and Finland, rural and agricultural concerns made negotiations complex, but both countries joined. A subsidy sys- tem, paid for in part by the new entrants themselves, was implemented to allow marginal farms in disadvantaged areas in both countries to continue to operate. Sweden faced a different dilemma: that it would have to reverse an agricultural policy reform that was generally in the direction that the EU wished to go. Until the late 1980s, Sweden maintained a policy much like that of other Nordic coun- tries. Suffering from overproduction, Sweden tried voluntary milk quotas and cereal set-asides, but to no avail. In June 1990, the parliament passed a bill that dramatically altered Swedish farm and food policy, abolishing agricultural sup- port and export refunds. In return, the government issued direct annual payments and an early retirement scheme to farmers. When export subsidies were reinstated upon Sweden's joining the EU, prices rose somewhat. One impact of joining the EU can be gauged by comparing cereal prices in the acceding countries and the EU. Finland, in particular, had to weigh the burden of reduced prices on the farm sector against the potential benefits of better manufac- turing-sector access to the rest of Europe. Nevertheless, the accession of Austria, Finland, and Sweden proceeded, a process smoothed by the fact that these coun- tries were able to pay for schemes that temporarily sheltered their farmers. These countries have not used their membership to press for higher agricultural prices at the EU level. The pattern of protectionism is illustrated in the calculated NRA for the EU and EFTA. Figure 3.4 shows the level of NRA for the EU-12 peaking at close to 100 percent in 1986, when world prices were at historical lows, and generally declining thereafter. By 1994, the NRA for the EU was down to 40 percent. Among the three countries that acceded in the 1985­94 period, Finland had the highest rate of assistance, and thus had the most adjustments to make upon accession. In the case of Austria, protection levels increased in the five years leading up to mem- bership, causing concern that the agricultural sector would be adversely effected after joining the EU. But the new EU members were not among the strongest sup- porters of agriculture in the EFTA countries, and so the tensions were manageable 140 Distortions to Agricultural Incentives: A Global Perspective Figure 3.4. NRAs to Agriculture, EU-15 and Western European Average, 1985­94 a. EU-15 members 180 160 140 120 percent 100 80 60 40 20 0 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 Sweden Finland Western European average EU-12 Austria b. Other Western European countries 500 450 400 350 300 percent 250 200 150 100 50 0 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 Iceland Switzerland Norway Source: Anderson and Valenzuela (2008), based on authors' spreadsheet, which draws heavily on OECD (2008). Western Europe 141 by means of transitional arrangements. As a result, there was very little impact on the EU from the enlargement from 12 to 15 members, in terms of either reducing or increasing the level of agricultural protection.31 The EFTA members that chose not to join the EU, meanwhile, maintained their high protection levels for agriculture. NRAs in Switzerland fell somewhat in 1989 but rebounded by 1990. Though there was a slight downward trend in Iceland and Norway over 1985­94, the rate of assistance remained well above the average for Western Europe.32 In any case, the bipolar nature of support in the region, as between nonmembers and members of the EU, was by that time firmly established. 1995­2007: Agriculture restrained The coming into force of the URAA in 1995 brought significant changes in the instrumentalities of the CAP. First, the agreement obliged the EU to convert its variable levy to a fixed tariff. Thus the threshold price from which the variable levy was calculated ceased to be the central determinant of protection levels. However, as a result of the careful choice of reference prices to calculate the tariff equivalent (known as "dirty tariffication") and the convenience of using an unweighted aver- age for tariff reductions, the impact on the price levels of sensitive goods such as sugar, beef, and dairy products was not great. For cereals, the creation of a special category of domestic support that was associated with output controls (the blue box) and hence deemed to be less distorting to trade meant that the new MacSharry payments were not required to be reduced. Moreover, a remnant of the "variable levy" was retained (at the insistence of the United States), because the tariff-inclusive import price was not allowed to rise above 150 percent of the intervention price. The time horizon of the URAA tariff and support reductions, 1995 through 2000, partially overlapped with that of the application of the MacSharry CAP reforms, which ran 1994­96. Though it became clear that though the 1992 CAP reform may have allowed the EU to agree to the URAA, the reform process would need to be continued if the WTO constraints were to be respected over the decade. Even more significant was the prospect of eastward expansion, where 10 countries already had an explicit promise of membership. While these countries were not large agricultural exporters, they did have significant farming populations. Farmers in the EU were focused on additional competition in certain commodi- ties, but the European Commission was concerned about the potential burden of paying subsidies to farmers in the prospective entrant countries. In either case, it was clear that the enlargement would have significant effects on the viability of the CAP. 142 Distortions to Agricultural Incentives: A Global Perspective In 1995, Franz Fischler took over as commisioneer for agriculture for the EU. He articulated a vision of a competitive, market-oriented (and simultaneously environmentally sustainable and socially acceptable) agricultural sector. The path chosen under his leadership was to continue down the path of the MacSharry reforms by reducing support prices and substituting direct payments tied by "cross-compliance" obligations to environmental goals. The Commission's propos- als for agricultural reform, included in the Agenda 2000 document, also dealt with budget and regional policy challenges facing the EU (Moyer and Josling 2002). The primary focus of Agenda 2000, though, was the enlargement of the EU to include the Central and Eastern European countries. Ten of these countries had already negotiated "Europe" agreements that would lead to bilateral free trade (including most of agriculture) over 10 years. As pressure mounted to fulfill the political pledge to allow for the reuniting of East and West Europe, concern about the impact on both agricultural markets in the EU and the financial cost of extending the CAP to the new members was also rising. Agenda 2000 was the Commission's suggestion for how the budget and agricultural challenges could be met. The agricultural reform component of Agenda 2000 was presented to the Council of (Agriculture) Ministers in 1997 and eventually endorsed with much of the Commission's ideas intact in 1999. Though the Agriculture Ministers' plan was modified by the European Council (a regular meeting of heads of state or government), the agreement that emerged was a major step in the evolution of the CAP.33 One important concession was won by those who favored even more reform of the CAP: there would be a midterm review of the effectiveness of the Agenda 2000 reforms in 2003 (the halfway point of the 2000­06 fiscal horizon). Fischler was able to convert what could have been a routine review of progress into a further step along the path set by the MacSharry and Agenda 2000 reforms by shepherd- ing through the Council of Ministers a proposal for consolidating the direct pay- ments associated with the compensation for price decreases in particular crops into the Single Farm Payment Scheme, a system that decreased links between farmers' production decisions and market prices on one hand and direct payments on the other (Swinnen 2008b). The 2003 reform package also began to tackle the difficult issue of reform of the dairy sector. Following on the heels of the 2003 reform, the European Commission pro- posed changes in the regimes for the Mediterranean crops--specifically, cotton, tobacco, rice, and hops. The thrust of reforms surrounding these products was to blend them with the Single Farm Payment Scheme while reducing distortions in the market. New proposals for fruits and vegetables were adopted in 2004, with subsidies for processing being converted into producer payments and those for Western Europe 143 the withdrawal of fresh produce from the market being chaneled through pro- ducer organizations. Wine and sugar reform proposals were introduced in 2005 and the sugar reform was agreed in 2006. This new regime aimed to eliminate the production quota system, paying compensation to the adversely impacted seg- ments of the wine market--producers, crushers and refiners, and overseas pro- ducers that rely on sales to the EU market--and cutting support prices. The wine reforms were adopted in 2008 with the aim of cutting capacity to meet shifting demand. The aggregated impact of these various policy changes is shown in figure 3.5. The EU-15 exhibited an NRA of around 40 percent following its absorbtion of three new members (and the introduction of the disciplines of the URAA) in 1995. By 2004, this had declined to approximately 30 percent. The reforms, combined with higher world prices, lowered the distortion to about 13 percent by 2007. The same trend is seen in the calculation of NRAs for the EFTA coun- tries that have retained their agricultural policy autonomy. For example, Norway and Switzerland have reduced distortions, as measured by the NRA, to about half the levels seen in the late 1990s. Iceland seems to be the EFTA coun- try where the tendency toward more moderate levels of assistance has been resisted most. Detailed Estimates of Nominal Assistance, 1955­2007 Having calculated the NRAs to inform the above description of the evolution of agricultural policies in Western Europe, it is possible to use the database to illus- trate several additional points. First, the NRA calculations need to be qualified by the change in the nature of the policy instruments adopted.34 In the latter 1980s, direct payments were limited to such products as olive oil and durum wheat, with market price support making up the bulk of assistance given to agriculture. From 1992, the total support (including direct payments and NPS support) and the market price support (limited to price support for individual commodities) began to diverge. By 2004, almost one-half of the 100 billion in assistance to agriculture came in the form of income payments and provision of services of an NPS nature (OECD 2006). There is no doubt that such payments benefit agriculture and keep resources that might otherwise leave in the sector, but the direct distortionary impact on commodity output, consumption, and trade is less, and arguably a much smaller source of the programs' economic cost.35 If the assistance from these somewhat decoupled measures had been included in the calculation of sup- port for Western European farmers, their NRAs would have declined much less after 1992 (see figure 3.6). 144 Distortions to Agricultural Incentives: A Global Perspective Figure 3.5. NRAs to Agriculture, EU and Western European Average, 1995­2007 a. EU-15 to 2004, EU-25 for 2005­06, and EU-27 for 2007 60 50 40 percent 30 20 10 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Western European average EU-15/25/27 b. Other Western European countries 250 200 150 percent 100 50 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Iceland Switzerland Norway EU-15 Source: Anderson and Valenzuela (2008), based on authors' spreadsheet, which draws heavily on OECD (2008). Western Europe 145 Figure 3.6. NRAs to Agriculture with and without Decoupled Payments, Western Europe, 1956­2007 120 100 80 percent 60 40 20 0 56 59 62 65 68 71 74 77 80 83 86 89 92 95 98 01 04 07 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 20 Western European average including Western European decoupled payments average Source: Anderson and Valenzuela (2008), based on authors' spreadsheet, which draws heavily on OECD (2008). Second, when products are classified as exportables, importables, and nontrad- ables, it becomes clear that exportables are assisted far less than import-competing farm industries in the case of the EU--though exporters are assisted only slightly less in the case of non-EU countries of the region (figure 3.7). Third, the rate of assistance to Western Europe's farmers dwarfs the small and declining NRAs for producers of nonagricultural tradable products, and so the relative rate of assistance (RRA) is almost the same as the NRA for agriculture (table 3.1). Fourth, the NRA varies greatly across the range of covered farm commodities (which account for more than 75 percent of the region's agricultural production when valued at undistorted prices). The extent of dispersion in NRAs for the EU, which increased up to the 1980s before starting to decline, has nevertheless contin- ued to increase in non-EU countries (see second-to-last row of tables 3.2 and 3.3). This suggests there is great scope to improve the efficiency of resource use within each country's farm sector. Fifth, leaving aside NPS and decoupled support, most of the assistance to farm- ers for covered products has come from border measures rather than domestic price supports (see near bottom of tables 3.2 and 3.3). This aspect of European Figure 3.7. NRAs to Exportable, Import-Competing, and Alla Agricultural Industries, EEC/EU and Other Western European Countries,a 1956­2007 a. EEC/EU 140 120 100 80 60 percent 40 20 0 20 40 60 56 59 62 65 68 71 74 77 80 83 86 89 92 95 98 01 04 07 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 20 import-competing total exportables b. Other Western European countries 300 250 200 percent 150 100 50 0 56 59 62 65 68 71 74 77 80 83 86 89 92 95 98 01 04 07 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 20 import-competing total exportables Source: Anderson and Valenzuela (2008), based on authors' spreadsheet, which draws heavily on OECD (2008). a. All NPS assistance is apportioned to tradables. The EU is the original 6 countries to 1972, 9 countries to 1985, 12 to 1994, 15 to 2004, 25 to 2006, and 27 thereafter. The other Western European countries in our study comprise the original 7 EFTA members (Austria, Denmark, Norway, Portugal, Sweden, Switzerland, and the United Kingdom) plus Finland, Iceland, Ireland, and Spain to 1970; then without Denmark, Ireland, and the United Kingdom from 1973 (when they joined the EEC-6); then without Portugal and Spain from 1986 (when they joined the EC); and without Austria, Finland, and Sweden from 1995 (when they joined the EU). 146 Table 3.1. NRAs to Agricultural Industries Relative to Nonagricultural Industries, EU-15 and EFTA-3, 1956­2007 a. EU-15 member countries (weighted average, percent) 1956­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 NRA, covered products 49.0 61.9 72.4 47.3 56.3 69.9 79.8 60.8 39.8 32.9 14.7 NRA, noncovered products 14.1 28.2 38.4 26.0 33.3 39.1 45.8 32.7 23.3 19.3 -- NRA, all agricultural products (excluding NPS) 39.8 53.6 64.4 42.5 51.3 62.7 72.6 54.6 36.2 30.0 11.2 All importables 56.0 74.3 80.2 52.9 58.8 76.5 84.0 63.8 50.4 48.2 28.6 All exportables 3.9 13.0 27.4 18.5 30.0 27.3 44.5 33.3 12.0 6.0 0.5 NRA, all agricultural products (including NPS) 39.8 53.6 64.4 42.5 53.3 71.0 76.8 59.2 40.4 34.2 15.4 NRA, decoupled payments 0.0 0.0 0.0 0.0 0.2 1.7 2.3 8.1 16.7 17.9 19.5 NRA, all agricultural products (including NPS and decoupled assistance) 39.8 53.6 64.4 42.5 53.5 72.6 79.1 67.3 57.0 52.1 35.0 NRA, all agricultural tradables (including NPS) 39.8 53.6 64.4 42.5 53.3 71.0 76.8 59.1 40.4 34.2 15.4 NRA, all nonagricultural tradables 8.2 7.4 5.7 3.8 2.5 1.4 1.7 1.3 1.4 1.4 1.1 RRAa 29.2 43.1 55.5 37.3 49.7 68.5 73.9 57.1 38.4 32.4 14.1 (Table continues on the following page.) 147 148 Table 3.1. NRAs to Agricultural Industries Relative to Nonagricultural Industries, EU-15 and EFTA-3, 1956­2007 (continued) b. Iceland, Norway, and Switzerland (weighted average, percent) 1956­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 NRA, covered products 256 257 255 252 256 232 339 295 204 171 99 NRA, noncovered products 169 172 173 179 166 169 150 111 78 64 141 NRA, all agricultural products (excluding NPS) 228 230 230 230 230 218 282 236 165 137 79 All importables 220 221 221 220 227 211 262 234 169 149 114 All exportables 247 249 249 255 239 240 334 242 159 119 70 NRA, all agricultural products (including NPS) 228 230 230 230 233 233 313 266 197 164 90 NRA, decoupled payments 0 0 0 0 11 56 31 32 48 68 68 NRA, all agricultural products (including NPS and decoupled assistance) 228 230 230 230 244 289 344 298 245 233 158 NRA, all agricultural tradables (including NPS) 228 230 230 230 233 233 313 266 197 164 90 NRA, all nonagricultural tradables 4.0 4.3 4.2 3.0 2.5 2.1 1.8 1.9 2.1 2.1 1.6 RRAa 216 216 217 220 225 226 306 259 191 159 87 Source: Anderson and Valenzuela (2008), drawing on authors' spreadsheet. Note: -- not available. a. The RRA is defined as 100*[(100 NRAagt ) (100 NRAnonagt ) 1], where NRAagt and NRAnonagt are the percentage NRAs for the tradables parts of the agricultural and nonagricultural sectors, respectively. Table 3.2. NRAs to Covered Farm Products, EU,a 1956­2007 (percent) 1956­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Exportables 17 12 4 7 17 11 8 10 6 2 1 Barley n.a. n.a. n.a. 35 45 14 n.a. n.a. n.a. 1 0 Rapeseeds n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 0 0 0 Rice 48 41 1 23 6 2 146 136 34 17 2 Tomatoes 9 16 n.a. n.a. n.a. n.a. n.a. 4 0 1 0 Wheat n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 16 4 0 Wine 16 13 4 4 4 11 1 11 4 2 1 Import-competing 60 90 103 68 70 93 93 70 56 55 28 Barley 39 59 66 n.a. n.a. n.a. 108 104 32 n.a. n.a. Beef 20 79 62 52 12 120 150 93 100 132 86 Eggs 28 18 20 1 23 17 22 10 9 1 0 Maize 9 24 55 37 56 43 90 89 31 25 19 Milk 259 301 314 269 431 335 291 122 87 61 19 Oats 3 34 46 20 45 0 37 47 47 23 8 Oilseeds 55 2 0 0 0 0 109 55 0 0 0 Pigmeat 28 75 139 113 88 111 24 13 25 28 17 Potatoes -- -- -- 43 90 48 16 16 13 10 10 Poultry 130 64 54 88 55 75 79 105 81 64 78 Sheepmeat 170 197 343 340 277 189 164 97 45 38 65 Soybeans 13 11 -- 0 0 0 121 61 0 0 0 Sugar 137 121 295 23 131 140 227 162 167 191 111 Wheat 11 52 67 5 2 38 84 75 n.a. n.a. n.a. 149 (Table continues on the following page.) 150 Table 3.2. NRAs to Covered Farm Products, EU,a 1956­2007 (continued ) (percent) 1956­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 NRA, all covered products 44 63 82 54 61 78 81 58 40 33 15 Domestic market support 0 0 0 0 0 0 3 2 0 0 0 Border market support 44 63 82 54 61 78 77 56 40 33 15 Dispersion of NRA of covered products 81 85 120 99 116 90 82 51 45 53 38 % Coverage at undistorted prices 74 77 78 80 79 78 78 78 79 79 74 Source: Anderson and Valenzuela (2008), drawing on authors' spreadsheet. Note: -- not available; n.a. not applicable. a. Weighted averages, with weights based on the unassisted value of production. Dispersion is the simple five-year average of the annual standard deviation around the weighted mean. The EU is the original 6 countries to 1972, 9 countries to 1985, 12 to 1994, 15 to 2004, 25 to 2006, and 27 thereafter. Table 3.3. NRAs to Covered Farm Products, Non-EU Western European Countries,a 1956­2007 (percent) 1956­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Barley 36 39 29 18 34 5 223 219 186 105 76 Beef 65 81 29 59 130 186 161 166 174 225 149 Eggs 64 69 71 80 321 328 145 148 296 161 186 Maize 91 53 47 34 48 34 51 89 124 89 64 Milk 259 275 251 262 301 223 227 240 242 181 73 Oats 23 48 71 58 132 34 103 115 154 113 83 Oilseeds 28 18 14 8 7 14 58 126 334 260 287 Pig meat 51 63 74 38 12 45 71 82 180 172 116 Potatoes 130 144 167 59 63 63 38 35 -- -- -- Poultry 56 52 25 101 188 207 104 74 370 437 383 Rice 18 25 14 29 21 12 52 -- -- -- -- Sheep meat 110 81 78 64 75 80 168 170 101 68 51 Soybeans -- -- -- 28 25 30 13 -- -- -- -- Sugar 265 253 446 47 126 121 250 198 233 255 173 Tomatoes 0 0 0 5 4 4 4 -- -- -- -- Wheat 29 42 43 7 14 46 126 170 195 91 61 Wine 0 0 0 0 0 0 0 0 -- -- -- Wool -- -- -- -- 124 36 78 207 187 167 207 (Table continues on the following page.) 151 152 Table 3.3. NRAs to Covered Farm Products, Non-EU Western European Countries,a 1956­2007 (continued ) (percent) 1956­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Exportables 37 43 55 37 35 30 98 133 236 172 70 Import-competing 82 81 66 43 57 61 169 191 194 171 114 All covered 68 69 63 41 50 51 119 152 204 171 99 Domestic market support 0 0 0 0 2 12 -20 -17 7 12 11 Border market support 68 69 63 41 46 33 138 169 198 159 88 Dispersion of NRA of covered products 90 91 125 67 109 103 81 73 84 105 112 % Coverage at undistorted prices 73 73 74 75 74 75 86 75 69 69 80 Source: Anderson and Valenzuela (2008), drawing on authors' spreadsheet. Note: -- not available. a. Weighted averages, with weights based on the unassisted value of production. Dispersion is the simple five-year average of the annual standard deviation around the weighted mean. Non-EU Western European countries in this study include the original seven EFTA members (Austria, Denmark, Norway, Portugal, Sweden, Switzerland, and the United Kingdom) plus Finland, Iceland, Ireland, and Spain to 1970; then without Denmark, Ireland, and the United Kingdom from 1973 (when they joined the EEC6); then without Portugal and Spain from 1986 (when they joined the EC); then without Austria, Finland, and Sweden from 1995 (when they joined the EU). Western Europe 153 policies is in sharp contrast to the situations in the United States, where domestic support plays a major role (Gardner 2009). Sixth, the average rate of assistance to farmers varies by country even among EU members (table 3.4). This is because the commodity composition of farm out- put varies across countries and over time, and hence so do the weights (based on value of production at undistorted prices) used to calculate the average. Sixth, the gross subsidy equivalent of the support provided to Western Europe's farmers varies hugely across countries (table 3.5a) and across commodities (table 3.5c). Milk, beef, and wheat are subsidized most in total. Expressed per per- son engaged in farming, subsidies are highest in Norway and Switzerland. Among EU members, Denmark, the Netherlands, and France have rates almost as high (table 3.5b). Note that the subsidy equivalent per EU farmer drops by two-thirds in 2005­07, much more than the 45 percent drop in the aggregate value over 2000­04, because the number of EU farmers has expanded dramatically with the latest eastern enlargement (even though, strictly speaking, the new member coun- tries are not in Western Europe). These figures are not actual cash transfers from the EU's budget in Brussels, but simply the value equivalent of the price support provided primarily via protection from non-EU imports. Finally, since most assistance to farmers comes from border measures, those same measures raise consumer prices of farm products. The extent of the con- sumer tax equivalent (CTE) is similar to that of the NRA for covered products in percentage terms (less so the further a country is from self-sufficiency), but vary considerably across countries when expressed on a per capita basis: after Iceland, Norway, and Switzerland, it is Ireland, Denmark, and the Netherlands where con- sumers are most harmed (table 3.6). Notice that the value of the implicit transfer from consumers is very small on a per capita basis, at about US$100 for EU coun- tries and US$300 for EFTA countries in 2000­04 (valued at 2000 US$). Since this is less than 0.5 percent of per capita income, it is not surprising that consumers see little benefit in joining together to lobby collectively against agricultural supports. Policy Trends and Turning Points since 1955 This section attempts an analytical explanation of the reasons behind the evolution of policy and policy-related distortions since the mid-1950s. It complements the historical/institutional narrative and NRA estimates of the two previous sections by emphasizing the political economy forces behind the long-run trends and turn- ing points, focusing on the CAP as the main policy for an increasing number of Western European countries (policy developments in the countries that opted to stay outside the EU show some similarity, however). Emphasizing the trends in policy allows for consideration of the changes in policy instruments over time and 154 Table 3.4. NRAs to All Agriculture,a Individual Western European Countries, 1956­2007 (percent) 1956­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Austria 25 33 36 12 17 17 49 71 45 41 19 Denmark 29 33 37 37 78 86 81 54 41 34 15 Finland 81 93 101 73 75 37 136 134 52 41 17 France 41 49 70 43 50 74 82 64 40 33 15 Germany 52 83 99 63 68 89 83 62 46 38 16 Iceland -- -- -- -- 229 252 346 289 173 144 153 Ireland 20 41 49 43 84 117 132 83 67 62 34 Italy 5 30 48 32 37 61 60 46 32 27 11 Netherlands 43 78 104 80 89 107 80 53 48 42 21 Norway 188 193 196 201 219 240 259 235 191 165 98 Portugal 6 4 7 9 24 15 41 37 28 26 15 Spain 13 14 12 2 4 1 53 44 31 26 14 Sweden 87 113 120 77 85 79 90 88 48 41 18 Switzerland 253 251 249 244 243 229 349 285 201 165 83 United Kingdom 65 60 43 32 68 83 95 70 47 40 20 Total Western Europe, weighted average 44 57 68 46 56 74 82 64 44 37 17 EU countriesb 40 54 64 43 53 71 77 59 40 34 15 Other Western Europeb 228 230 230 230 233 233 313 266 197 164 90 Source: Anderson and Valenzuela (2008) based on authors' spreadsheet. Note: -- not available. a. Weighted averages, with weights based on the unassisted value of production. Dispersion is the simple five-year average of the annual standard deviation around the weighted mean. b. The other Western European countries in our study comprise the original seven EFTA members (Austria, Denmark, Norway, Portugal, Sweden, Switzerland, and the United Kingdom) plus Finland, Iceland, Ireland, and Spain to 1970; then without Denmark, Ireland and the United Kingdom from 1973 (when they joined the EEC6); then without Portugal and Spain from 1986 (when they joined the EC); and without Austria, Finland, and Sweden from 1995 (when they joined the EU). Table 3.5. Gross Subsidy Equivalents of Assistance to Farmers, Total, Per Farm Worker and by Product, Western European Countries,a 1956­2007 a. Total (constant 2000 US$ millions) 1956­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Austria 422 566 651 192 462 440 8,184 3,108 1,715 1,338 541 Denmark 1,439 1,667 1,983 2,403 5,719 5660 4,133 3,804 2,565 1,920 888 Finland 1,219 1,416 1,820 1,733 2,185 932 3,644 3,685 1,000 730 287 France 9,866 14,437 22,418 17,393 26,423 31,314 26,608 25,399 15,707 11,012 5,225 Germany 8,735 18,158 25,895 21,893 32,312 33,297 24,905 21,113 13,869 10,125 3,975 Iceland -- -- -- -- 314 246 245 177 113 96 123 Ireland 254 572 710 1,235 3,744 4,026 3,553 3,421 2,646 2,035 837 Italy 651 6,596 11,772 8,903 13,826 17,067 13,585 13,061 8,728 6,257 2,646 Netherlands 1,812 3,778 5,634 6,153 9,735 10,432 7,083 6,695 5,228 3,593 1,483 Norway 1,432 1,497 1,710 1,924 2,898 2,876 3,061 3,437 2,566 1,836 1,208 Portugal 58 70 170 295 858 416 1,218 1,579 1,194 924 509 Spain 841 911 919 816 741 604 8,792 9,431 6,446 5,440 3,372 Sweden 2,573 3,323 3,832 3,050 4,229 3,363 3,180 3,096 1,499 1,047 413 Switzerland 3,148 3,486 3,981 5,130 5,599 5,817 6,491 6,541 4,438 3,014 1,945 United Kingdom 8,937 9,177 7,181 6,915 16,530 17,597 14,013 12,752 8,138 5,498 2,411 Total, Western Europe 41,271 65,653 88,676 76,403 124,095 132,880 128,695 117,300 75,850 54,866 31,173 EU membersa 21,065 42,969 65,718 54,342 108,415 119,393 103,890 97,256 68,734 49,920 27,618 Other Western Europea 20,206 22,685 22,957 22,061 15,680 13,487 24,805 20,044 7,116 4,946 3,555 (Table continues on the following pages.) 155 156 Table 3.5. Gross Subsidy Equivalents of Assistance to Farmers, Total, Per Farm Worker and by Product, Western European Countries,a 1956­2007 (continued ) b. Per person engaged in agriculture (constant 2000 US$) 1961­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Austria 805 1,161 409 1,257 1,361 28,429 12,023 7,947 7,661 3,468 Denmark 5,240 6,627 10,339 26,813 30,680 24,401 25,217 20,676 18,950 9,865 Finland 2,820 3,853 4,267 6,547 3,340 15,411 18,224 6,160 5,659 2,540 France 4,285 6,571 6,223 11,736 16,887 17,304 20,111 15,437 13,678 7,412 Germany 4,303 6,894 7,235 11,707 13,875 13,252 14,324 11,891 11,069 4,938 Iceland -- -- -- 24,186 19,696 17,016 11,597 8,195 7,529 10,223 Ireland 1,752 2,169 5,026 14,801 17,647 17,261 18,759 15,662 12,680 5,363 Italy 1,445 2,514 2,350 4,373 6,288 5,825 6,731 5,632 5,176 2,524 Netherlands 10,230 15,336 18,827 30,384 33,320 22,665 22,195 19,561 15,434 6,928 Norway 5,992 7,945 10,489 17,122 18,461 21,634 26,848 22,641 18,274 12,988 Portugal 39 143 276 765 357 839 1,925 1,668 1,492 891 Spain 182 230 255 259 218 2,973 5,303 4,364 4,418 3,024 Sweden 8,365 10,773 10,105 15,409 13,520 14,501 15,848 8,999 7,422 3,200 Switzerland 13,128 16,193 23,432 27,951 30,661 33,093 34,182 25,954 19,938 13,796 United Kingdom 10,358 8,954 10,507 23,461 25,825 22,013 21,525 14,852 10,863 4,971 Total, Western Europe 2,957 4,258 4,433 8,164 10,089 11,392 12,425 9,612 8,369 2,471 EU membersa 3,323 5,348 5,501 11,090 14,295 10,193 11,504 9,051 7,935 2,256 Other Western Europea 2,395 2,689 2,999 2,908 2,799 22,448 20,325 23,927 18,664 13,316 c. By product (constant 2000 US$ millions) 1956­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Barley 1,649 2,866 3,627 2,072 5,456 1,707 6,532 6,361 1,621 120 50 Beef 2,529 7,192 7,950 7,914 6,252 13,145 19,075 16,169 12,144 9,877 6,177 Eggs 604 817 534 893 1,808 1,436 1,673 1,037 571 130 123 Maize 57 557 1,323 1,396 2,715 2,204 3,589 3,297 1,394 1,036 1,020 Milk 20,793 23,917 28,101 31,074 45,150 41,790 37,896 32,811 22,919 15,418 2,822 Oats 405 1,229 1,611 1,081 2,027 229 812 849 335 144 70 Oilseeds 83 23 27 29 84 66 2,833 1,899 44 32 41 Pig meat 3,936 8,899 14,823 14,400 15,941 15,699 7,011 4,732 5,175 5,401 4,051 Potatoes 2,653 2,804 2,841 2,518 5,846 2,802 1,127 1,221 912 465 735 Poultry 1,863 1,973 1,987 3,229 3,717 4,060 4,337 5,687 4,492 3,424 4,042 Rice 246 267 62 578 126 104 529 598 220 84 12 Sheep meat 1,194 1,562 1,997 3,195 5,274 4,064 3,163 2,581 1,432 1,087 1,458 Soybeans 0 0 -- 2 2 1 571 337 0 0 0 Sugar 2,354 2,433 4,151 518 4,772 4,087 5,091 4,807 4,229 3,232 1,351 Tomatoes 44 387 710 1,046 1,539 1,345 887 356 44 75 0 Wheat 1,706 4,943 6,559 1,229 304 6,465 9,849 9,114 2,039 546 126 Wine 1,630 998 568 681 664 847 399 1,707 721 240 267 Wool -- -- -- -- 26 2 18 16 11 8 9 Total, Western Europe 36,784 56,703 75,678 65,415 101,451 99,844 105,391 93,581 58,303 41,319 21,916 EU membersa 19,158 36,948 55,749 47,291 88,543 88,805 84,016 77,079 53,218 37,773 19,033 Other Western Europea 17,626 19,755 19,929 18,124 12,908 11,038 21,375 16,502 5,086 3,546 2,883 Source: Anderson and Valenzuela (2008) based on authors' spreadsheet. Note: -- not available. a. The EU is the original 6 countries to 1972, 9 countries to 1985, 12 to 1994, 15 to 2004, and 27 thereafter. The EFTA countries are the original 7 to 1970; then 8 (with Iceland) to 1972; then 6 from 1973 (when the United Kingdom and Denmark joined the EEC) to 1985; then 5 from 1986 (when Portugal joined the EC) to 1994; then 3 after Austria, 157 Finland, and Sweden joined the EU in 1995. For 2005­07, the "EU members" and "total Western Europe" aggregates also include the most recent 12 countries that joined the EU in May 2004 and January 2007. 158 Table 3.6. CTEs of Policies Assisting Farmers, Covered Products, Total and Per Capita and by Product, Western European Countries,a 1956­2007 a. Aggregate CTE by country (percent) 1956­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Austria 74 88 77 22 24 22 63 88 39 35 17 Denmark 42 44 39 41 95 79 70 48 38 32 12 Finland 113 123 132 91 93 45 202 201 46 40 17 France 49 52 75 50 54 76 61 45 34 30 15 Germany 65 95 107 67 67 82 59 41 35 30 13 Iceland -- -- -- -- 42 72 272 210 117 95 104 Ireland 25 38 45 43 125 126 115 73 56 46 20 Italy 5 33 52 35 37 58 46 35 29 26 10 Netherlands 52 89 118 82 96 111 83 56 47 38 18 Norway 272 282 285 276 230 70 93 116 114 111 85 Portugal 1 11 18 11 34 21 37 33 25 24 16 Spain 19 20 19 4 0 5 36 30 24 21 13 Sweden 104 106 134 76 108 90 124 109 46 44 23 Switzerland 275 271 261 253 223 97 245 215 143 144 83 United Kingdom 69 65 46 34 71 89 77 54 44 40 22 Total, Western Europe 53 65 74 49 59 70 65 49 37 33 17 EU membersa 42 62 80 54 60 78 59 42 34 30 15 Other Western Europea 72 70 62 39 51 37 119 141 133 131 84 b. Total CTE per capita (constant 2000 US$) 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Austria 95 95 26 54 44 1026 354 145 108 46 Denmark 220 211 276 624 498 343 299 221 165 55 Finland 291 353 327 413 181 699 665 717 525 207 France 246 353 283 380 367 259 229 159 115 55 Germany 235 311 261 362 330 211 157 110 82 33 Iceland -- -- -- 1070 969 937 465 310 277 379 Ireland 94 114 252 792 696 495 435 299 221 65 Italy 125 218 165 226 243 178 162 117 87 33 Netherlands 228 322 290 439 408 271 245 188 129 48 Norway 319 367 446 593 333 374 447 326 241 226 Portugal 19 41 29 121 58 107 134 105 84 48 Spain 32 34 29 0 6 146 150 108 89 56 Sweden 256 346 254 457 311 303 284 138 119 52 Switzerland 708 704 843 790 627 820 792 486 354 266 United Kingdom 185 135 136 302 272 188 157 111 86 41 Total, Western Europe 162 205 198 306 276 226 207 138 103 46 EU membersa 161 232 240 339 319 209 178 128 96 41 Other Western Europea 165 156 119 193 128 293 488 422 310 254 Source: Anderson and Valenzuela (2008) based on authors' spreadsheet. Note: -- = not available. a. The EU is the original 6 countries to 1972, 9 countries to 1985, 12 to 1994, 15 to 2004, and 27 thereafter. The EFTA countries are the original 7 to 1970; then 8 (with Iceland) to 1972; then 6 from 1973 (when the United Kingdom and Denmark joined the EEC) to 1985; then 5 from 1986 (when Portugal joined the EC) to 1994; then 3 after Austria, Finland, and Sweden joined the EU in 1995. For 2005­07, the "EU members" and "total Western Europe" aggregates also include the most recent 12 countries that joined the 159 EU in May 2004 and January 2007. 160 Distortions to Agricultural Incentives: A Global Perspective the role of instrument changes in improving efficiency in meeting the various pol- icy objectives. Discontinuities in the trends can also come from the impact of changes in ideas among domestic political forces that impact farm policy in Western Europe, the introduction of international rules and institutions, and the impact of bilateral and other trade agreements. Long-run trends Several important trends can be identified as having shaped Western European agricultural policy in the postwar period: the production increases in European agriculture that have transformed the sector from a significant importer of temperate-zone agricultural products to a major exporter; the changing emphasis over time away from the production of undifferentiated farm commodities and toward quality and other attributes that allow for product differentiation; the increasingly globalized market that both provides raw materials and intermediate inputs for the European food system and offers expanding access to consumers in other countries; and the increasing concern with environmental and other objec- tives that are constantly changing the nature of the coalition of interests necessary for agricultural policy support. The reinstrumentation of agricultural policy has been a consequence of trends away from import protection that allowed the sector to have a sheltered environment for growth and toward direct payments that are supported as rewards for socially responsible farming practices. Intervention in the market and the dumping of surpluses abroad has largely given way to a regulatory approach that encourages quality foods that have a ready market and protection of locational and other indicators associated with quality attributes. The reinstrumentation itself has been largely responsible for the moderation of the levels of protection relative to nonagricultural sectors and for less consumer taxation mentioned above. Production increases Somewhat ironically, it is the strong growth of agricultural production in Western Europe that brought about the most intractable policy problems. Policies intro- duced in the EEC were premised on the fact that agriculture needed support because it could not compete with domestic industry for labor and capital or with production from overseas farms that were larger and perhaps less constrained by social and environmental factors. The instruments used were designed to insulate domestic producers from developments in world markets. World prices were seen as too volatile to act as a basis for stable domestic markets. When prices were propped up by imports entering over the protective wall, this system was at least manageable, despite its inefficiency. But when production outstripped demand, Western Europe 161 surpluses began to accumulate. New instruments had to be developed, while those that intended as occasional market supports undertook new functions. Pressures arising from production increases were first noticed in the cereals market. As a result of the initial decision on cereal prices, production of wheat, barley, and maize (corn) increased sharply. Use of grains for animal feed dropped rapidly, leading the search for alternative feeds to begin in earnest. One external factor played a significant role in the development of such alternatives: the level of protection of several nongrain feedstuffs was fixed at zero (or otherwise very low levels) in the Dillon Round of the GATT, which concluded in 1962. Feed com- pounders in Europe found that soybean meal (from imported soybeans) and cassava chips (from Thailand and Indonesia) made a cheaper feed for livestock than domestic corn and barley. The expanded imports that resulted exacerbated the domestic supply imbalances, adding to the growing stocks of cereals.36 Dairy prices had also been set high in the EEC, encouraging production and discouraging consumption. Structural change in the dairy sector added to the increase in production.37 Surpluses showed up early, and various mechanisms were suggested to get rid of the "butter mountains" that had resulted from inter- vention buying. Taxes on the use of nondairy products were suggested but not approved. Subsidies for the use of skimmed milk powder in animal feed were effective but did not discourage production. Modest price cuts (through corespon- sibility levies) also were too small to curb production. Finally, in 1981, quotas on milk production were introduced. They have been in place ever since, and have been an effective, if inefficient, way to restrain production. Beef production also expanded rapidly in the EEC, in part as a reflection of the increase in dairy production (most beef in Europe comes from the dairy herd) and in part a reaction to hill farming and other livestock subsidies. But in the 1970s, consumption of veal slumped as a result of health concerns related to the use of growth hormones, and in the 1990s beef consumption plummeted follow- ing the discovery of a probable link between Bovine Spongiform Encephalopathy (BSE) in cattle and a new variant of Creutzfeldt-Jakob disease (nvCJD) disease in humans. So the policy response had to focus on reassuring consumers rather than discouraging producers from overproduction. Another product where surpluses appeared was sugar, a crop that has a very long production history in the region (sugar beets were grown for 150 years as a rotation crop for Northern Europe and provided a valuable calorie source in times of interuption in trade).38 But both France and the United Kingdom had former colonies that depended on the European market to sell their sugar cane. When the United Kingdom acceded to the EEC in 1973, a guaranteed market was conceded for its overseas suppliers. So the EU has struggled with an import commitment alongside a growing export surplus. When other exporters challenged the legality 162 Distortions to Agricultural Incentives: A Global Perspective of the EU's policy of reexporting the imported sugar, a panel found that the EU had exceeded its export subsidy commitment under the WTO. The sugar policy has now been reformed, by reducing support prices and paying compensation to growers and beet crushers, so that it is less intrusive on world markets. In each of these cases, the main problem was that prices were rigid downwards, largely as a result of the decision-making system. Farm ministers were reluctant to return from the annual price negotiations in Brussels having agreed to price cuts. In some cases, financial control was also lacking, largely because the impact of price decisions on spending were not apparent to those (the farm ministers) mak- ing the decisions. As the budget cost of export subsidies was shared among all members (and the import levies pooled), the true (foreign exchange) cost to each country was the internal price. Under these circumstances, it was not surprising that price cuts were rare. Productivity growth put pressure on the costs of the pol- icy and on the external impact, but the lack of financial accountability and the dif- ficulty of getting the political support needed for a major policy change led to an environment of continual crisis and controversy. Comparisons with other high-income countries The predominant feature of agricultural policy in Western Europe over the past 50 years has been the high level of support (relative to other temperate-zone pro- ducers) given to most agricultural sectors. Though agricultural protection in East Asia rose from low levels in the 1950s to be well above that in Western Europe (Honma and Hayami 2009), protection in Australia, Canada, New Zealand, and the United States was generally at a much lower level (Gardner 2009; Anderson, Lattimore, Lloyd, and MacLaren 2009). Including agriculture in the experiment of economic integration was a bold move for the EEC. The economic rationale was similar to that in other sectors: the agricultural sector would gain from rationalization on a continent-wide basis and the price of agricultural products would be reduced. In turn, this created the hope that an integrated European agricultural system could compete with the Americas and the temperate-zone producers of the Southern Pacific. But this would have required earnest competition among EEC farmers and halting of support for the less efficient of them. In general, the level of protection in Western European agri- cultural policy over the past 50 years can be attributed to the reluctance of politi- cians (and to a large extent, of society in general) to expose the sector to true competition. Integration of agricultural markets has indeed taken place, but in an environment where inefficiencies were able to survive and production decisions did not reflect the realities of the market. In short, the distortions that have con- tinued in European agriculture have been the result of the political reluctance to encourage needed adjustments in the sector. The social costs of such adjustments Western Europe 163 played a dominant role in shaping crucial decisions to undertake agricultural support--support that absorbed resources that could have had greater benefits to society elsewhere. The process of outmigration from agriculture has undoubtedly been slowed by the incentive structure of the CAP. Thus the policy has been successful in its own terms, by mitigating the adjustment cost, but expensive in societal terms by delaying changes that would have been beneficial. The unwillingness of governments to subject their agricultural sectors to the rigors of the market was also evident in the decisions of those countries that chose not to join the EU. The need to lower support prices in Norway and Switzerland in the event of membership was a major factor in generating opposition to EU acces- sion in rural communities. These high-cost countries have been reluctant to expose their farmers to competition from the EU and to induce the structural changes and entrepreneural initiatives that offer the chance for competitiveness. Each has protected its agriculture behind high tariffs, thus blunting market signals and burdening the downstream sectors with higher costs. Environmental and quality objectives The five decades discussed above cover a period of a major transformation of agricultural policies not just in terms of the instruments used but in the objectives that lie behind government intervention in the agricultural sector. The focus of agricultural policy has over the period swung away from a narrow focus on the income level of farmers toward a raft of policy objectives ranging from the preser- vation of the coutryside and the provision of healthy food to rural development and animal welfare. Farm incomes have over time become more of a constraint on policy change than a rationale for the existence of policy. Those who benefit financially from the policy have been able to slow down the shift in emphasis (and budget allocation) but not stop it altogether. The need to include incentives in the improvement of farm structures was recognized in the 1960s, when the Mansholt Report emphasized the dangers of persuing all objectives through price support. The introduction of a structural program proved slow, as it was in competition with funds for market manage- ment. Similarly, the need for funds for rural development was realized in part as a way of helping agriculture adjust. Eventually, the "second pillar" of the CAP was constructed in the Agenda 2000 reform and supplied with enough funding to make it a significant aspect of policy. National funds are also now authorized, subject to constraints, to be shifted from price and income support to rural development programs (modulation) that meet other objectives. Second pillar spending currently accounts for about 19 percent of total CAP expenditure. Environmental objectives have also become important in shaping the CAP and policies in the EFTA countries. Originally, the thrust was to prevent the negative 164 Distortions to Agricultural Incentives: A Global Perspective impacts of intensive farming on the countryside (and adjacent urban areas) and on soil, air, and water pollution. Over time, this has evolved into an emphasis on the positive contribution of agriculture as the dominant activity in much of the Europe's more scenic rural areas. In the 1990s, environmental objectives became associated with the term "multifunctionality," reflecting the range of public goods provided by rural businesses. But other exporters pointed to the risk that this could become an excuse for a blank check to policy makers who were looking for reasons to continue income support in a more acceptable guise--thus perpetuat- ing distortions to resource allocation. The first evidence of major changes to the politics and political economy of the CAP came as a result of public concerns over food safety. Building on public reactions to reported instances of human exposure to hormones used in live- stock production, the EU banned the use of such substances in animal produc- tion. In addition to the international ramifications of this decision (U.S. and Canadian beef and beef products were banned), the impact on domestic policy was palpable, as it ushered in perhaps the most significant change in the percep- tion of the CAP, its link with food quality, and the production practices employed by farmers.39 Advocates of more emphasis on local foods and on more animal-friendly livestock practices joined with those who opposed the use of transgenic technology in foodstuffs to form a formidable new voice in agri- cultural policy. Alongside this trend, food safety and quality policies and regula- tions on animal welfare have become increasingly centralized. The policy has itself been redefined as one that emphasizes quality over quantity and the pro- motion of foods that reflect the cultural richness of rural Europe. Farmers spot- ting this trend have shifted their own production and marketing practices toward these new demands. While some argue that this market orientation reduces distortions, others might regard the policy-supported changes in demand as distortions for the consumers who cannot or do not wish to shift their eating habits in this way. Reinstrumentation The original policy instruments chosen by the EEC were intended to provide a stable price environment for the agricultural sector. The variable levy smoothed out the fluctations that would have otherwise been transmitted to the domestic market. The export subsidies (restitutions) allowed goods to be withdrawn from the domestic market and sold at lower world prices. Only dairy (after 1981) and sugar (strictly speaking, a "temporary" market order) were subject to supply con- trols. Some producer subsidies existed, including those for durum wheat and olive oil, but these were also backed up by tariff protection. Wine was regularly taken off the internal market for distillation, and producer groups could withdraw fruits Western Europe 165 and vegetables with assistance from the European Commission if prices fell to "crisis" levels. The changes that have come about in the instruments used in the CAP are as important in their impacts as the levels of protection themselves. Though these changes have not come easily, they have had a lasting impact on the shape of the CAP. The most significant modification of policy instruments came in 1992, with the MacSharry reforms.40 Direct payments in compensation for price declines added flexibility to the policy, as described earlier. But this transition also followed a pattern observed in other countries. The Unite States, for example, began to move from price supports to direct payments and to decouple those payments from production and prices during the 1980s (Moyer and Josling 1990). At the time of the Uruguay Round, it became apparent that one way to reduce overseas threats to the CAP was to make the instrumantalities more similar to those of trading partners. Since the mid-1980s, the gap between total producer support and that derived from market price support has widened, particularly after the 1992 reforms. Cur- rently, market price support accounts for just over one-half of total CAP support, compared to 90 percent 20 years ago. To the extent that the new instruments are significantly less trade distorting, they also are less distorting for the domestic economy. A similar trend is noticeable in Switzerland: the "decoupling" of pay- ments from output and price was introduced even though the absolute level of support is still greater in that country than in the EU. Though Norway has not increased its degree of decoupling, the absolute level was in any case lower as a result of subsidies aimed at keeping people in the northern regions of the country regardless of their levels of output. The decoupling of support from price and out- put has also allowed for a trend to attach payments to other policy objectives in these EFTA countries as it has in the EU. Turning points Though these trends have evolved since 1955, several turning points have been crucial in influencing the macroeconomic environment, which in turn affected the level of distortion in agricultural policies in Western Europe over the period. These include, for example, the increase in the price of oil in 1973 and changes in commodity markets and food prices. The oil price increase impacted the CAP in two respects. First, it increased costs to farmers in a way that was readily apparent and led to compensating price increases. Coming not long after the end of the transition period in 1967 (a time when policy makers were attempting to reduce price incentives) and the exchange rate changes of 1968 (when prices diverged again within the EU), 166 Distortions to Agricultural Incentives: A Global Perspective the pressure for price adjustment was difficut to resist. Moreover, the oil price increases coincided with the sharp rise in cereal and sugar prices on world mar- kets, thus fueling fears of longer-term shortages and masking the level of under- lying distortions that were present in the system. The opportunity to make use of the high price period to reduce the support prices was lost. Instead, support prices followed the world market upward, only to be left isolated when world prices declined again. The oil price hike also coincided with the accession of the United Kingdom to the EEC, a shift that was expected, based on the country's history of interna- tional openness, to bring a fresh look at European farm policy. Some countries, such as the Netherlands, welcomed this while others were prepared to resist (notably France but also Germany). However, the period of inflation and cur- rency movements that followed in the 1970s overwhelmed the possibility of pol- icy improvements and left the United Kingdom reluctantly accepting the CAP so long as the budget cost to the United Kingdom was constrained and it con- trolled the pace of adjustment of the "green" rate of exchange. This compromise lasted through the runup to monetary union, when policy prices in the EU became common again. Political changes also have had an impact on the CAP, and hence on the level of distortion in Western European agriculture.41 The fall of the Berlin Wall in 1989 and the subsequent unification of Germany offered an opportunity for the return of Central and Eastern European countries to the economic and political main- stream. Agreements were negotiated with Central and Eastern European countries that were deemed to be ready for accession, and agricultural products were included, albeit with some quantitative restrictions. All agricultural policy deci- sions since that time have had to be measured in light of their implications for the larger European market. This has acted as a disincentive to increase support levels in the existing EU. To an extent unmatched in most other developed regions, agricultural policy in Western Europe has been influenced by external pressures as well as by those of a more domestic nature. But a major turning point came in the mid-1980s as a direct result of this external pressure. Up to that point, decision making within the CAP always took place within the context of the state of world markets, which impacted budget costs in particular. But the Uruguay Round introduced, within the URAA, a legal framework that restricted the scope for domestic action. Through the negotiation and enforcement of tariff reductions and scheduled cuts in subsidies, the URAA also gave domestic policy makers a lever on the policies of other countries.42 Thus the CAP, along with the policies of other countries, was challenged after 1995, and policy changes that would have been mainly of internal interest became internationally significant.43 Western Europe 167 Prospects for Future Reform of Western European Policy The essence of the CAP is that it represents a bargain among member countries, as all countries in the EU maintain their own agricultural policy objectives even if they have ceded the ability to use many instruments to acheive those objec- tives. The CAP, then, is a result of the bargaining process within the EU as much as a coordinated and consistent European view of appropriate role for govern- ment in agriculture. Reform can come about as a result of changes in the terms of the bargain just as easily as a shift in the model of agriculture that underlies policy or the reactions to exogenous events. However, it is still possible to specu- late on the change in underlying attitudes to agriculture in the development of the CAP, and hence hazard a guess as to where those trends might be heading. In the light of the political economy constraints and incentives mentioned ear- lier, this section attempts to draw out the prospects for further policy reform in Western Europe. One issue is whether the consolidation of Western Europe in the EU, and by implication the extension of the CAP to all Western European agriculture, will be completed in the foreseeable future. This implies the entry of Norway and Switzerland to the EU. Thus far, Norway has resisted in part because its position as an oil and gas exporter has enabled it to have a high level of income as an "adjunct" to the EU market. Though investment could be stimulated by accession, it is not enough to convince the electorate to "share" its fish resources, subject its farmers to price decreases, and possibly force the modification of control and use of energy supplies. But the possible exhaustion of Norway's oil reserves and an autonomous reform of its agricultural policy (alongside a more acceptable fisheries policy in the EU) could change the situation. The level of agricultural protection, though, is still considerably higher in Norway than in the EU, and significant policy adjustments would be needed before Norway's agricultural sector could be assimilated into the CAP.44 In Switzerland, most aspects of policy, both industrial and agricultural, have been steadily converging with those of the EU. As a result, one would expect the economic aspects of accession to be less significant than the political aspects. Not only does Switzerland have a strong history of avoiding foreign entanglements, but its status as a participatory democracy poses problems for the transfer of sov- ereignty to the EU level. If a time comes when major sectors of the Swiss economy are harmed by exclusion from the EU, one might expect to see an application to join. But as in the case of Norway, the level of farm protection is still considerably higher than in the EU.45 An enlargement issue of a different kind is the prospect of the accession of Turkey to the EU, one that poses an economic and political dilemma for the EU. In this regard, the impact on the level of distortion in Western Europe may well be 168 Distortions to Agricultural Incentives: A Global Perspective significant. On the one hand, the geopolitical argument points to the benefits of offering membership to a strategically important country. On the other hand, Turkey has an income level of about one-third that of the average EU level, and its membership would add some 20 percent to the population of the bloc. The impact on the EU economy of Turkey's joining would thus be considerable (in a positive as well as a negative direction), with the challenge to agriculture being a major consideration. For these reasons, one might expect a long period of transition before Turkey is fully incorporated into the single EU market (see also Anderson and Swinnen 2008). More trade agreements are in the cards for the EU, as it consolidates its rela- tions with its neighbors in North Africa and the Middle East and extends its "neighborhood policy" to include Central Asia and the Caucasus. The current wave of trade agreements under negotiation also include arrangements between the EU and the countries of Latin America, through an EU-Mercosur trade pact. To avoid getting left behind relative to competitors in North America and East Asia, trade policy plans of the European Commission also include agreements with major Asian countries. Where they include improved market access for farm products, these agreements will tend to put downward pressure on agricultural distortions. The extent to which this pressure is reflected in real changes will depend on the political significance attributed to the free trade agreements them- selves. Domestic agricultural policy has proved resistant to many such trade policy developments, but not all. The Everything but Arms agreement that gave the least developed countries tariff- and quota-free access to the EU market included agricultural products, and led in turn to the reform of politically sensitive sugar policy in 2006. The CAP has always absorbed a major share of the EU budget, though that share, once above 75 percent, has now been reduced to about 45 percent. More countries are being included in the activities funded from the budget without cor- responding increases in the level of funding. While the prospect of substantially increased funds available for agricultural policy is slim, this in turn means that shifts in instruments toward the less distorting direct (decoupled) payments is increasingly difficult. More likely is the addition of funds for projects such as the development of biofuels (which could benefit agriculture in a more acceptable way, depending on how the perception of the environmental effects of such interventions evolve). The impact of the Uruguay Round on the CAP has been emphasized above. Negotiations surrounding the Doha Round, the WTO's follow-up to the Uruguay Round, have been underway since 2001. The Doha Round would cut tariffs, eliminate export subsidies, and reduce the scope for trade-distorting domestic support. Though the talks are presently on hold, there is still the possibility that Western Europe 169 they may be concluded by the end of 2009. If an agreement is eventually reached within the range of options on the table, the EU would phase out its export subsi- dies altogether by 2013 and significantly cut its ability to support prices. Though much of domestic policy is now sheltered in the green box (and thus not subject to reductions), the Doha Round would "lock in" those aspects of reform. Tariffs would be cut by about 60 percent, with some exceptions for the most sensitive products, such as dairy and beef. WTO dispute settlement panels have already found that the sugar and banana regimes were inconsistent with WTO obliga- tions: other countries will no doubt explore further challenges where they con- sider the CAP too distortive in world markets. One vital question with respect to the future of the CAP is whether the trend is toward a renationalization of the policy. There have been small moves in this direction, with flexibility built in to the direct payments. But the degree of decou- pling, for some commodities, is at the discretion of the member countries. How far this trend will go is a matter of political decisions well above that of the agri- cultural policy.46 So long as they do not influence competition, there is a logic to the renationalization of payments that are not related to farm output, and the possibility of an eventual division between the market mechanisms that would be controlled at the EU level and the payments for public goods that may revert to the national level. In sum, there could be a convergence of likely and desirable policy directions before 2020. A continuation of decoupling of support from price and output should over time significantly reduce distortions, though if such pay- ments are "recoupled" to other objectives there could be some economic cost. Trade agreements will reinforce this trend, as access to the EU market will be an important incentive for other countries to conclude such agreements. What, if any, are the policy lessons for developed, developing, and transition economies of the trends and turning points in Western Europe's agricultural poli- cies? In many ways, the EU is unique in having agreed at its founding on a system of internal free trade and a common policy for market management and external protection. In that regard, the lessons are unlikely to be directly applicable. But the development of the CAP does offer some general lessons. Among the most impor- tant of these has to do with the pitfalls of attempting to manage markets in such a way as to give farmers adequate incomes. Technical advances, demand shifts, and the natural responses by private actors to incentives offered by policy will under- mine the ability of policy administrators to achieve the required outcome. In general, the CAP is too inflexible to adapt to changes in the internal or external market, ministers of agriculture meeting periodically are unlikely to be able to adjust policy instruments and price levels in the face of such changes. So the pol- icy is always a step behind the market. In addition, the income streams generated by the policy ensure that there are always interests lined up against change. 170 Distortions to Agricultural Incentives: A Global Perspective The combination of inflexible decision making and a bias toward the status quo make it impossible for the CAP to keep its relevance over the decades.47 But external pressure through the GATT talks and the WTO, as well as a feeling that the CAP was no longer viable in its original form, have encouraged a series of changes that have shifted the nature of agricultural policy in Europe. The policy is still somewhat inflexible, and changes may be even harder now with 27 members, but the reduction of prices of most of the main products to near world market levels provides much more scope for creative policy making. The flexibility of direct payments, though it can be overused, gives hope for agricultural policy that is less distortive of world markets and less obtrusive of economic adjustments on the domestic front. The way in which this was accomplished should indeed be of interest to other countries. What Have We Learned? Three main conclusions emerge from this chapter. First, the level of distortion to the Western European economies resulting from direct and indirect policy inter- ventions in agriculture has declined in recent years. Protection of agriculture, highest in the 1960s, fell in the 1970s as world markets became tight and European agriculture appeared, temporarily, to be more competitive. But reaction to the period of high world prices and inflation at home led to a rapid increase in distor- tion between the mid-1970s and early 1980s. That the level of distortion in the EU remained high for several years attracted criticism from at home and abroad, prompting changes in policy that have led to steady decline and increased stability in the level since the mid-1980s. In the countries that remained outside the EU, distortions remain high and more variable, although in the last few years there have been signs of convergence with the EU. Second, the act of joining the EU has had a mixed impact on country levels of distortion, and hence on the total level of distortion in the region. In some cases the accession of new members has raised the protection level and in others it has decreased it. Protection levels in Western Europe increased as a result of the 1973 accession of Denmark, Ireland, and the United Kingdom. Agriculture in each of these countries expanded and their competitiveness with respect to world markets was eroded. The entry of Southern European countries in the 1980s also increased the overall level of protection by including countries such as Spain and Portugal, which had relatively low levels of assistance, in the CAP. But many of the countries that chose to stay outside the EEC in the 1960s were also those with high-cost farming sectors. The high levels of protection were not "imported"into the EU from countries such as Finland and Austria. Hence, the overall result of the incorporation of the many of the Nordic and Alpine countries has been to reduce Western Europe 171 overall protection in Western Europe somewhat when they eventually joined the EU. This is confirmed by the nature of the two major countries in Western Europe that have remained outside the EU (Switzerland and Norway), both of which have far higher levels of agricultural distortion than the EU. This suggests that their eventual accession could play some role in reducing distortions in these countries, even if not in the current EU countries. The accession of the 10 new EU members from Central and Eastern Europe posed--and continues to pose--a challenge for the CAP and the process of reform. Though producers in the EU expressed the fear that the competition from those countries would oversupply markets in the EU, particularly for products such as pig meat, the European Commission was more concerned with the implied com- mitment to distribute direct payments to the large number of farmers that were about to join the EU and come under the political umbrella of the CAP. As it hap- pened, prices rose steadily in the new entrants, but surplus production has not been a problem. Market disruption has been limited to isolated cases, and the east- ern half of Europe is rapidly becoming integrated with Western markets. The prospect of additional costs under the CAP proved a real stimulant for reform, as funds allocated for a EU containing 15 countries were to be spread over 27 coun- tries. So one can reasonably conclude that the latest EU enlargement has tightened the constraints already limting the CAP, even though it raised farm protection in Central and Eastern European countries (Anderson and Swinnen 2009). The third conclusion is that domestic policy operates in an international envi- ronment, even as domestic politicians proclaim their autonomy. The development of policy in Western Europe became an international issue in itself, and external pressures could not be ignored. This posed a major challenge for policy makers, who could have used the external pressure as a reason for positive change or as a negative force that calls for defenses that deflect its impact. For many years, the EU attempted the latter strategy, though since 1990 it has made significant use of bilateral and multilateral agreements to guide agriculture in a direction that makes it compatible with a more open trade system. So economic distortions have been reduced as compatibility issues with the international economic system have been resolved. The pace of such developments has been dictated by those who were gaining most from the unreformed policy, but in the end the political bene- fits that flowed from reforming the policy came to outweigh the costs of change. Though considerable distortion in Western European agriculture still exists, the path to reducing that distortion is opening up. Allowing farmers to respond to consumer demands not artificially stimulated by government regulation is the most desirable way to point agricultural policy. The rhetoric of the CAP has adopted this approach for the past five years, and the policy is slowly moving in that direction. 172 Distortions to Agricultural Incentives: A Global Perspective Notes 1. In a recent survey of European economic growth since 1950, Crafts and Toniolo (2008) conclude that incentive structures are a crucial explanation of comparative growth rates of the economies of Europe (east as well as west). 2. The countries of Western Europe, for the purposes of this chapter, include the 15 countries that were members of the EU (the EU-15) in early 2004 plus Iceland, Norway, and Switzerland. In terms of policy development, each country is considered to have abandoned its autonomous domestic farm policy when it joined the EU. Thus, 18 countries had independent policies at the start of the period, in 1955, but by 2004 the number of independent policies had dropped to four--the EU, Norway, Iceland, and Switzerland. The EU then expanded eastward to include eight Central European countries (plus Cyprus and Malta) in May 2004, and again in January 2007 to include also Bulgaria and Romania. Most of this chapter focuses on the period prior to the EU's expansion eastward, while the chapter by Anderson and Swinnen (2008) focuses on Eastern Europe and the former Soviet Union. 3. Comparable figures for the United States show a utilized agricultural area of 379 million hectares and 2.1 million farms, with an average size of 180 hectares. 4. The agricultural sector in the United States employs 0.7 percent of the civilian labor force and contributes 0.9 percent of GDP. 5. The economy of Western Europe as a whole expanded by 2.7 percent over 1955­2004. 6. For a discussion of the different reactions of the Western European countries to the mid- 19th-century period of relatively free trade, see Kindleberger (1975). Swinnen (2008a) emphasizes changes in social and political institutions (such as the expansion of voting rights, changes in landlord- tenant relationships, and the emergence of agribusiness organizations) that undoubtedly contributed to the policy environment. 7. The Enclosure Acts of the 18th century gave the United Kingdom a farm structure that allowed it to take advantage of the emerging technologies (mechanization) and farming practices of the period (Orwin 1949). 8. Output per hectare in Denmark more than doubled between 1880 and 1930, whereas the same measure stayed steady in the United Kingdom (Ingersent and Rayner 1999). Agriculture accounted for 45 percent of Denmark's GDP in 1880. 9. Rural incomes in Germany declined by almost 40 percent between 1929 and 1932. 10. Marshall Petain, in particular, had a vision of the agricultural destiny of France, reviving the notion of agricultural protection that had been promoted by Meline in the 19th century. 11. The enlargement of the EU to include 10 new members in May 2004 is discussed in the context of its influence on policy later in this chapter. Implications for the new members are dealt with in detail in Anderson and Swinnen (2009). 12. The abreviation EU is sometimes used even for the period before the transformation of the EEC into the EC and later into the EU. 13. See Josling (2008b) for a larger discussion on external influences on CAP reform. 14. The original six countries of the EEC were Belgium, France, Germany, Italy, Luxembourg, and the Netherlands. In economic terms, Belgium, Luxembourg, and the Netherlands had already agreed to an economic union (the Benelux Union) in 1948, and so were on the road to becoming a single unit. 15. The inclusion of agriculture in the free movement of goods within the EEC was at the insis- tence of the Dutch, who had struggled with the same issue in the formation of the Benelux Union. 16. This terminology was used for the cereal market regime. Some differences in instruments and nomenclature, however, were introduced in the other commodity-market organizations. 17. The world price over this period was about US$60 per ton. 18. The "non" from Charles de Gaulle in 1963 that ended the first set of talks about U.K. accession reflected a fear that the nature of the EEC, and in particular the CAP, would be compromised by the admission of a country that boasted of its "special relationship" with Washington. The United Kingdom was widely seen as a Trojan horse for U.S. policy interests, and its accession a sure recipe for continued pressure on the CAP. Western Europe 173 19. The United Kingdom appeared to be an exception to this trend, though Howarth's price sup- port measures did not reflect the full range of input subsidies that were introduced over those years. 20. The present study has a broader product coverage than the earlier attempts to measure protection. 21. Though independent data for the other two (much smaller) members, Belgium and Luxembourg, are not included in this study, it is likely that the distortions are similar in those two countries, as they have been in an economic union since 1922 (to which Luxembourg contributes just 4 percent of the union's population of less than 11 million). 22. Italy, however, had low rates of assistance through the 1950s. 23. Portugal was one of the countries studied in the study led by Kreuger, Schiff, and Valdés (1988). Its income level was below that of Brazil at the time, and the country had many of the features common to developing countries. 24. Price data for Switzerland and Norway for some products were not available prior to the latter 1970s so the NRAs for them were assumed to be similar to the earliest years for which data are avail- able. Hence the lesser degree of fluctuation in their lines in figures 3.2b and 3.3b. 25. The Kennedy Round of GATT talks (1963­68) took place over this time, and the main focus of the exporting countries was to constrain the protectionist tendencies of the CAP. In this they were largely unsuccessful. 26. The major political reason for the decision of Sweden and Norway not to join the EEC was the requirement for neutrality that each had enshrined in their constitution. Finland was even more con- strained, as it bordered on the Soviet Union, and was accordingly inhibited in foreign policy initiatives. Likewise, Austria was also constrained by the postwar treaties, and Switzerland by its determination not to join alliances or even multilateral organizations. Ireland stayed out both to keep in step with the United Kingdom, its major market, and to retain its neutrality. Ireland did not join EFTA in spite of the links with the United Kingdom, but it did retain some preferences into that market. 27. It may seem that the measure of assistance used in this study is unduly influenced by world price movements. But the NRA indicates the distortions at any particular level of world prices: if world prices change, then so do the distortions caused by policy. This phenomenon appeared again when food prices soared during 2005­08. 28. The European currency unit (ECU) was a basket of currencies used as a unit of account in the European Community prior to the launch of the euro in 1999. 29. The European Union was established in 1993 by the Maastricht Treaty. 30. Livestock protection was changed only incidentally, as tariffs and levies on pigs, poultry, and eggs had been tied to the levies on cereals. 31. Another member of EFTA, Iceland, did not participate in talks about joining the EU. The con- cern that that country would have to adhere to the Common Fisheries Policy has always been a major political hurdle. Lichtenstein, an independent country in a customs union with Switzerland, joined the EFTA-EU accord even though Switzerland did not. 32. Estimates for Iceland are included only from 1979, when comparable data became available from the OECD. 33. Among the other changes introduced in the Agenda 2000 reforms was the definition of a "second pillar" of the CAP to complement the market price and income support. This second pillar was to channel funds to rural development and be managed as a joint activity between the European Commission and the member states. An element of cofinancing was also introduced, as was the requirement to move some of the direct payments into this pillar (modulation). As spending on rural development has an ambiguous relationship with agricultural incentives (it is designed largely to offer attractive alternatives to farming), these second pillar funds are not be included here as distortionary payments. 34. The NRA was calculated with NPS payments and direct decoupled payments (both included and excluded) to help identify the importance of these policy changes. 35. The extent to which a payment is production-neutral (the degree of decoupling) differs depending on the way it is administered and the expectations of the recipient, so the categorization of 174 Distortions to Agricultural Incentives: A Global Perspective policies by the degree of production-neutrality is a hazardous endeavor. No attempt is made here to evaluate the extent of the distortion that might still be present from policies defined as decoupled from output price and quantity produced. Rather, the value the OECD estimates is included under certain categories of support as decoupled. For the years 1979­85, there was just one category, called "direct payments." From 1986, those payments are specified to comprise the OECD's items C (payments based on area planted/animal numbers), D (payments based on historical entitlements), F (payments based on input constraints), and G (payments based on overall farming income), all of which are defined in the OECD PSE-CSE Database. For 2005­07, those items were replaced by similar but newly defined items C to E. While the administration of the EU's direct payments policies varies among the member countries, data is insufficient to do more than assume the percentage points of NRA from decoupled payments is the same for all EU members. This categorization for economic purposes (and that for NPS assistance) should not be confused with the legal allocation of domestic support measures into the WTO's "colored boxes" in the context of international commitments. 36. In addition, the soybean oil produced along with the soybean meal was a major competitor for butter and for olive oil, increasing the surpluses of these crops. 37. Trade issues also played a role here, as the United Kingdom insisted on continued preferential access for New Zealand for butter as part of its accession arrangements. 38. Sugar beet production was developed and encouraged in the Napoleonic era as a food security strategy. 39. The emergence of this parallel justification for the CAP, as a guardian of a safe food supply, is described in Roderer-Rynning (2009). 40. A change in the oilseeds policy preceeded that in the cereals market, when a GATT panel ruled that payments to processors tied to the use domestic raw material were inconsistent with GATT obli- gations. This change allowed the EU to gain experience with a direct payment program. 41. The significance of EU enlargement in the Agenda 2000 and the 2003 Fischler reforms is explored in Swinnen (2008b). 42. Swinbank and Daugbjerg (2006) explore the links between the 2003 Fischler reforms and the WTO trade negotiations that began in 2001. 43. External influences on the CAP also included the many regional trade agreements concluded between the EU and other countries. In these agreements, domestic agriculture was usually sheltered from direct competition. But over time, the pressure to allow some market access became an additional stimulus to reform. 44. Much the same could be said for Iceland. 45. Due to the small economic size of Norway and Switzerland, however, the absolute value of the subsidies is much smaller than that of the EU. As a result, the international implications of such protection are less immediate. 46. In addition to the political nature of these decisions on funding the CAP, the institutional changes also have an impact. Thus, for instance, if the Lisbon Treaty were to be ratified, the European Parliament would have a right to codecision on major aspects of the CAP (though it would keep its consultative role for technical issues). See Roederer-Rynning (2009) for a fuller description of these changes. 47. The link between the CAP and the way in which decisions are made in the EU has been empha- sized by several writers. See Swinbank (1989) for an early discussion, and Pokrivcak, Crombez, and Swinnen (2006) for a more recent examination of the importance of decision-making institutions. References Anderson, K., and Y. Hayami. 1986. The Political Economy of Agricultural Protection. London and Sydney: Allen and Unwin. Anderson, K., R. Lattimore, P. Lloyd, and D. MacLaren. 2009. "Australia and New Zealand." Chapter 5 in this volume. Western Europe 175 Anderson, K., and J. Swinnen. 2009. "Eastern Europe and Central Asia." Chapter 6 in this volume. Anderson, K., and E. Valenzuela. 2008. Global Estimates of Distortions to Agricultural Incentives, 1955 to 2007. Core database at http://www.worldbank.org/agdistortions. Avillez, F., T. Finan, and T. Josling. 1988. Trade, Exchange Rates, and Agricultural Pricing Policy in Portugal: The Political Economy of Agricultural Pricing Policy, World Bank Comparative Studies, Washington, DC: World Bank. ______. 1991. "Portugal." In The Political Economy of Agricultural Pricing Policy: Volume 3 Africa and the Mediterranean, ed. A. Krueger, M. Schiff, and A. Valdés, Baltimore: Johns Hopkins University Press. Corkhill, D. 1995. The Portuguese Economy Since 1974. Edinburgh: Edinburgh University Press. Crafts, N., and G. Toniolo. 2008. "European Economic Growth, 1950­2005: An Overview." CEPR Discussion Paper 6863, Centre for Economic Policy Research, London. Eurostat Database. 2006. http://epp.eurostat.ec.europa.eu/portal/page/portal/eurostat/home/ (accessed June 2006). FAO (Food and Agriculture Organization). 1973. "Agricultural Protection: Domestic Policy and International Trade." FAO Conference Document C 73/LIM/9. Rome: FAO. Gardner, B. 2009. "The United States and Canada." Chapter 4 in this volume. Gulbrandsen, O., and A. Lindbeck. 1973. The Economics of the Agricultural Sector. Uppsala: Almquist and Wicksell. Hallett, G. 1981. The Economics of Agricultural Policy. Oxford: Blackwell Press. Heidhues, T., T. Josling, C. Ritson, and S. Tangermann. 1978. "Common Prices and Europe's Farm Policy," Thames Essay 14, Trade Policy Research Centre, London. Honma, M., and Y. Hayami. 2009. "Japan, Republic of Korea, and Taiwan China." Chapter 2 in this volume. Howarth, R. W. 1971. Agricultural Support in Western Europe. London: Institute of Economic Affairs. Ingersent, K., and A. J. Rayner. 1999. Agricultural Policy in Western Europe and the United States. Cheltenham: Edward Elgar. Josling, T. 1970. "Exchange Rate Flexibility and the Common Agricultural Policy of the European Economic Community." Review of World Economics 104 (1): 57­95. ______. 2008a. "Distortions to Agricultural Incentives in Western Europe." Agricultural Distortions Working Paper 61, World Bank, Washinton, DC. ______. 2008b. "External Influences on CAP Reform: An Historical Perspective." In The Perfect Storm: The Political Economy of the Fischler Reforms of the Common Agricultural Policy, ed. J. F. M. Swinnen. Brussels: Centre for European Policy Studies Publications. Josling, T., and S. Harris. 1976. "Europe's Green Money." The Three Banks Review 109: 57­72. Kindleberger, C. P. 1975. "The Rise of Free Trade in Western Europe, 1820­1875." Journal of Economic History 35 (1): 20­55. Lieberman, S. 1995. Growth and Crisis in the Spanish Economy, 1940­93. London: Routledge. McCrone, G. 1962. The Economics of Subsidizing Agriculture. London: Allen and Unwin. Moyer, W., and T. Josling. 1990. Agricultural Policy Reform: Politics and Processes in the EC and USA. Hemel Hempstead, U.K.: Harvester-Wheatsheaf Publishers. ______. 2002. Agricultural Policy Reform: Politics and Process in the EU and the US in the 1990s. Aldershot, U.K.: Ashgate. Nash, E. F. 1955. "The Competitive Position of British Agriculture," Journal of Agricultural Economics 11 (3): 222­41. Nash, E. F., and E. A. Attwood. 1961. The Agricultural Policies of Britain and Denmark: A Study in Reciprocal Trade. London: Land Books. OECD (Organisation for Economic Co-operation and Development). 1967. Agricultural Policies in 1966. Paris: OECD. ______. 2006. Agricultural statistics database. http://www.oecd.org (accessed April 2007). ______. 2008. Producer and Consumer Support Estimates, OECD Database 1986­2007. http://www .oecd.org (online database for 1986­2007 estimates; OECD files for estimates using an earlier methodology for 1979­85). 176 Distortions to Agricultural Incentives: A Global Perspective Orwin, C. S. 1949. A History of English Farming. London: Thomas Nelson and Sons. Pepelasis, A., G. Yannopoulos, A. Mitos, and others. 1980. "The Tenth Member--Economic Analysis." Sussex European Papers 7, University of Sussex. Pokrivcak, J., C. Crombez, and J. F. M. Swinnen. 2006. "The Status Quo Bias and Reform of the Common Agricultural Policy: Impact of Voting Rules, the European Commission and External Changes." European Review of Agricultural Economics 33 (4): 562­90. Roederer-Rynning, C. 2009. "The Common Agricultural Policy." In Policy-Making in the EU, ed. H. Wallace, M. Pollak, and A. Young. Oxford: Oxford University Press. Roesener, W. 2000. "The History of German Agriculture." In Agriculture in Germany, ed. S. Tangermann. Frankfurt: Verlag. Snyder, L. L. 1945. "The American-German Pork Dispute, 1879­1891." Journal of Modern History 17 (1): 16­28. Swinbank, A. 1989. "The Common Agricultural Policy and the Politics of European Decision Making." Journal of Common Market Studies 27 (4): 303­22. Swinbank, A., and C. Daugbjerg. 2006. "The 2003 CAP Reform: Accomodating WTO Pressures." Comparative European Politics 4 (1): 47­64. Swinnen, J. F. M. 2008a. "The Political Economy of Agricultural Protection: Europe in the 19th and 20th Century." Paper prepared for the Project on Political Economy of Agricultural Distortions, World Bank, Washington, DC. ______. 2008b. "The Political Economy of the 2003 Reform of the Common Agricultural Policy." In The Perfect Storm: The Political Economy of the Fischler Reforms of the Common Agricultural Policy, ed. J. F. M. Swinnen,. Brussels: Centre for European Policy Studies Publications. Tracy, M. 1989. Government and Agriculture in Western Europe, 1880­1988. Hemel Hempstead, U.K.: Harvester Wheatsheaf. Tyers, R., and K. Anderson. 1992. Disarray in World Food Markets: A Quantitative Assessment. Cambridge and New York: Cambridge University Press. 4 United States and Canada Bruce L. Gardner* There is much in common between the agricultural sectors of the United States and Canada. The number of farms in North America has been falling for decades, a well-known trend that has given the public a sense that agriculture is a hard-pressed industry in decline. What is less well known is that the decline has greatly slowed over the past 20 years. During the 1950s and 1960s, the number of farms in the United States fell at the rate of 3 percent annually, declining by half in 20 years. Between 1995 and 2005, the number declined by only 4 percent for the whole decade (figure 4.1).1 To fully understand the trend, it is necessary to consider farms of different sizes separately. The number of farms with annual sales of more than US$500,000, for example, quadrupled between 1978 and 2005, from 18,000 to 79,000. The number of farms sales between US$100,000 and US$500,000 also increased during those years. The number of farms with sales of less than US$25,000 per year, however, has held quite steady. Farms with sales between US$25,000 and US$100,000 represent the largest portion of the total decline. Along with the decline in the number of farms, U.S. land dedicated to farming has also declined, at an annual rate of 0.4 percent between 1950 and 2005. Over those years, this has resulted in the loss of about 230 million acres of farmland. *Professor Gardner died in March 2008. The editing of this chapter (which does not include a discussion of the latest U.S. Farm Bill, but see Orden, Blandford, and Josling [2009] for more on the topic) was done as faithfully as possible given his original manuscript. The author was grateful for helpful comments from workshop participants, and for invaluable help with data compilation by Johanna Croser, Esteban Jara, Marianne Kurzweil, Signe Nelgen, Francesca de Nicola, Damiano Sandri, and Ernesto Valenzuela. The working paper version of this chapter (Gardner 2008) con- tains additional background material and appendix tables. 177 178 Distortions to Agricultural Incentives: A Global Perspective Figure 4.1. Farm Household Income as a Percent of National Household Income, United States, 1930­2005 160 140 120 100 percent 80 60 40 20 0 30 35 40 45 50 55 60 65 70 75 80 85 90 95 00 05 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 Source: Gardner (2002, figure 3.12) and U.S. Department of Agriculture, Economic Research Service, Briefing Rooms (http://www.ers.usda.gov/Briefing/). Available cropland acreage, however, has remained relatively constant over this period, with 345 million acres harvested in 1950, versus 321 million acres in 2005. Considering the increase in irrigated and better-drained acreage, quality-adjusted cropland may even have increased slightly. Together with declining farm numbers, the acreage data imply an increase in the average size of farms. But as the decline in numbers has slowed significantly in recent decades, so has the growth in aver- age farm size. In 2005, the average U.S. farm had 444 acres (180 hectares); 30 years earlier, the average size was 431 acres (175 hectares). In Canada, the number of farms peaked in 1941. In the 40 years between 1961 and 2001, the number of farms declined by half, from 480,000 to 246,000. This is a rate of decline of 1.7 percent annually, and faster than the U.S. rate of 1.4 percent during the same period. Though Canada's land devoted to cultivated crops declined over the 1960­2000 period, it did so less rapidly than the number of farms: Canada had 111 hectares of cropland per farm in 1961, and 147 in 2001 (compared to about 75 hectares of cropland per U.S. farm in 2002). Yields and value of output per hectare of cropland were lower in Canada than in the United States, however, reflecting the generally cooler and drier climate of Canada. Total agricultural land declined much more slowly in Canada than in the United States--indeed, hardly at all, from 69.8 million hectares in 1961 to 67.5 million hectares in 2001. United States and Canada 179 Relative to the decline in the number of farms, the farm labor force in Canada and the United States continues to fall at a more rapid pace, mainly because there is less unpaid family labor. At the same time, the use of material inputs (fertilizers, fuels, and purchased feed additives) has doubled in the United States since 1950, leaving the aggregate input index calculated by the U.S. Department of Agriculture (USDA) remarkably constant, at 102 in both 1950 and 2000 (base 1996 100). But agricultural output has continued to grow at a steady clip, so total factor produc- tivity (TFP) growth has been an impressive 1.9 percent per year, with no evidence of slowdown in the trend in recent years despite energy price shocks, environmen- tal constraints, and concerns about exhaustion of TFP gains attributable to earlier breakthroughs in improved hybrid seeds and other innovations. So far, continued advances in genetics, livestock management, capital equipment, and economies of scale have kept the real cost of U.S. farm products on a pronounced downward path. These cost declines have been largely reflected in lower prices of farm prod- ucts and hence lower costs of raw materials for other foods. While the average price of farm products rose 40 percent between 1978 and 2005 (1.2 percent per year), the gross domestic product (GDP) deflator rose at an annual rate of 3.2 percent. Thus the real price of agricultural output fell by an average of 2 percent during this period--essentially the same as the rate of TFP growth. Overall farm size (in terms of total land and output per farm) has increased in Canada at a rate faster than in the United States since the 1950s. As of 2001, the average Canadian farm was 670 acres, compared to 440 acres for American farms. Larger farm size in Canada has not hindered output or productivity growth, though: real output per farm increased more rapidly than farm size, reflecting increased yields per hectare and a strong trend in total factor productivity growth, which is estimated to have increase 2.5 percent annually in recent decades (Furtan 2006). The fact that farm prices have fallen in real terms largely in parallel with cost decreases indicates that real incomes of farmers may not have benefited from farm productivity growth. Yet the incomes of people employed in agriculture have in fact grown, both in real dollar terms and relative to real incomes of the nonfarm population. From the 1930s to the 1960s (which includes the Great Depression and Dust Bowl years), farm households could be reasonably categorized as a low- income population. By the 1990s, farm and nonfarm household incomes had equalized. Since 2000, U.S. farm household incomes have been significantly higher than those of nonfarm households (figure 4.1). How has this income growth been achieved, given that the decline in real prices has equaled the decline in farmers' costs of production? The answer is the increas- ing importance of farm households' integration into the nonfarm economy, so that in recent years, off-farm income sources account for 85­90 percent of average 180 Distortions to Agricultural Incentives: A Global Perspective U.S. farm household incomes. These data suggest that income from farming itself may indeed be quite low. However, a full understanding of the farm income data requires consideration of differences between farms of different sizes. Numeri- cally, a majority of farms--87 percent in 2004--have less than US$10,000 annu- ally in sales. The costs of farming at this scale are such that these farms on average earned only US$1,020 from farming, and more than half are estimated to have losses from their farm enterprises. Nonetheless, the average household income of these farms is US$71,500, thanks to off-farm income. At the other end of the spec- trum, family farms classified by the USDA as commercial scale operations (those which have US$250,000 or more in sales) earned an average of US$145,300 from farming plus US$46,038 from off-farm sources, and these latter farms produce more than two-thirds of U.S. agricultural output.2 On a per-farm basis, net income from farming in Canada during 2001­05 was almost the same as in the United States, although net farm income in Canada, as in the United States, was highly variable from year to year.3 On average, Canadian farm households receive less off-farm income than in the United States. Agricultural Policies in the United States Legislative proposals in the United States to improve the economic situation of farmers through governmental intervention in commodity markets were first devel- oped conceptually as remedies for the precipitous fall in product prices in the after- math of World War I. After several failed attempts to enact farm support bills in the 1920s, the further decline in agriculture's situation with the onset of the Great Depression of 1929­32 led to legislative success in the landmark New Deal pro- grams beginning in 1933. These programs had the principal purposes of increasing the incomes of farm people (with an explicit goal stated in terms of "parity" of farm income with reference to a pre­World War I standard) and stabilizing farmers' rev- enues. The fateful choice was to attempt to achieve both these goals by means of supporting farmers' prices received for a subset of commodities ("basic" commodi- ties in U.S. law--initially, wheat, corn, cotton, rice, tobacco, pork, and milk). The four main feasible means of market intervention in pursuit of higher pro- ducer receipts had been implemented for key commodities and were already in use in the 1930s: production controls, government purchases of commodities for stockpiling at price support "loan rates" at which farmers could forfeit the basic commodities instead of making repayment of the loan plus interest, disposal of surplus stocks through distribution or subsidized sale for export or to domestic consumers, and direct payments to producers. The first direct payments, in 1933, were tied to farmers' idling of land or destruction of livestock, so from their incep- tion they were not classical production-inducing subsidies. United States and Canada 181 In 1936, the U.S. Supreme Court ruled that the federal government had no authority to administer land-idling acreage controls under New Deal farm legisla- tion, on the constitutional grounds "that powers not granted are prohibited. None to regulate agricultural production is given, and therefore legislation by Congress for that purpose is forbidden." Subsequently, the Court's alleged respect for precedent was not extended to this decision, and many later production control measures have passed constitutional muster.4 At the time, the result of the Court's decision was a merger of prior concerns about conservation with measures to remove acreage from commodity production. This was done in the Soil Conserva- tion and Domestic Allotment Act of 1936 principally by defining "soil-depleting" crops (the main basic commodities) and "soil-conserving" crops (grasses and legumes), and paying farmers to substitute the latter for the former. Trade policy was almost negligible in the 1930s' programs. The policy propos- als of the 1920s, in contrast, had given a central role to export promotion. The Smoot-Hawley Tariff Act, signed into law in June 1930 (despite a petition to Pres- ident Hoover signed by more than 1,000 economists asking that he veto the bill) protected manufacturing much more than agricultural products. It is notable, however, that the crossover Democratic votes needed to pass the bill in the Senate (the Republicans were then the main protectionists and the Democrats free traders) were lured by the desire to protect sugar (Louisiana), wool (Wyoming), and Florida fruit (Benedict 1953). By 1934, it was clear that even these elevated tariffs were insufficient to protect sugar, and the first Sugar Act added sugar to the list of basic commodities, authorizing import and production quotas. With respect to overall agricultural trade, by 1934 the combination of the Great Depres- sion and trade restrictions in the United States and elsewhere reduced both U.S. agricultural exports and imports to about US$600 million per year each, one- sixth of the levels of 1920. Experience during and after World War II made it apparent that export demand was capable of creating farm prosperity to an extent, and with far less cost and turmoil, than a decade of intensive effort by the federal government had been able to deliver in the 1930s. Under the Marshall Plan, U.S. exports of food- stuffs were 19 million tons annually in 1947­50, compared to 4 million tons in 1935­39. However, three aspects of the situation in the United States were obstacles to free-trade ideas in agriculture. First, the Marshall Plan and subsequent agricul- tural exports were in large part financed by subsidies rather than being bought abroad at world market prices. Second, U.S. commodity policy held some domes- tic commodity prices above world levels, so that import restrictions were vital to these policies (otherwise, the program would have to support the world price, not just the U.S. price). Third, the relatively few U.S. importable farm commodities-- notably sugar, dairy products, and some meats and fruits--had sufficient political 182 Distortions to Agricultural Incentives: A Global Perspective power to maintain protection via tariffs or quotas even when giving them up as part of a larger trade liberalization agreement would have been beneficial to the nation as a whole. Section 22 of the Agricultural Adjustment Act of 1933 required import quotas to be imposed if imports threatened the effectiveness of a price support program. This situation led the United States to join Europe in pressing for a waiver of agri- cultural products from General Agreement on Tariffs and Trade (GATT) members to reduce export subsidies or provide increased import access to their markets. Some experts argued vigorously for changing U.S. farm programs to make them compatible with liberal trade, notably in the "Brannan Plan" of 1948 and other proposals to replace production controls and price supports by payments to farmers and adjustment assistance (Johnson 1950). But U.S. policy did not turn to favor- ing the inclusion of agriculture in the GATT until the 1960s. By then, European farm policy had taken a decisively protectionist path that precluded significant agricultural trade liberalization. Beginning with the Agricultural Trade and Development Act of 1954 (P. L. 480), U.S. policy followed a path suggested by the Marshall Plan of using food aid to foreign countries as a mechanism for surplus disposal. During 1956­64, about one-fourth of U.S. agricultural exports were shipped under this program. Though P. L. 480 exports have had varying degrees of concessionary pricing, depending on the status of the importing country, overall there has been a substantial subsidy element in these exports. Since the 1970s, the program has shipped a fairly con- stant amount of just under US$1 billion in commodities annually. In addition, going back to 1935, government-provided export credit and guarantees of repay- ment to private sector lenders have been used to stimulate foreign demand for U.S. commodities. Exports have also been promoted through USDA grants to farm or commercial interests for the purpose of informational and sales efforts abroad. Revamped as the Marketing Assistance Program in 1990, spending under such programs still amounts to about US$100 million annually. More explicit export subsidies were paid through most of the post­World War II period, most notably in wheat, where their role was negotiated under the Inter- national Wheat Agreement starting in 1949.5 In the 1960s, more than 85 percent of U.S. wheat exports were assisted by subsidies. Then, during the worldwide commodities boom of the 1970s, it appeared that the era of export subsidies might be replaced by commodity scarcity, with trade policies restraining rather than subsidizing exports. The United States embargoed grain shipments to the Soviet Union and a few other countries for short periods during 1974­80. Com- modity scarcity proved temporary, however. The worldwide collapse in commod- ity prices of the 1980s provided the stimulus for export promotion programs. The European Community intensified its longstanding practice of export subsidies (Josling 2008). The United States began offering specific-destination export United States and Canada 183 subsidies in retaliation in the early 1980s, regularizing this approach in the Export Enhancement Program (EEP) established in 1983 under the Reagan Administra- tion. Where feasible, Canada met the subsidy competition through the pricing policies of the Canadian Wheat Board (CWB). The EEP, like the pre-1970s export subsidies, was first and foremost a wheat program. It began as low-price government sales of wheat stocks held by the Commodity Credit Corporation (CCC) as a result of the domestic price support programs to North Africa in 1983. The EEP was complicated, using a payment-in- kind approach. USDA would determine particular countries and commodities for which it believed export subsidies would be helpful in selling U.S. products. Exporters would then negotiate a deal with a foreign buyer at a price discounted from prevailing world trading prices. The exporter would then apply to USDA for a payment sufficient to make up the difference between the market price and the negotiated discount price. USDA, if it approved the sale, would give the exporter sufficient wheat from CCC stocks to cover the payment, called the export "bonus." By the late 1980s, the bonuses were totaling US$1 billion annually, with more than 80 percent of EEP commodities accounted for by wheat in 1985­89. The program was widened and generalized so that CCC wheat stocks could be used to subsidize exports of other commodities, and in 1990, when available CCC wheat stocks were exhausted, in-kind bonuses were replaced by cash. As the United States expanded its export subsidies in the 1980s, Europe and Canada, within their respective support structures, met the competition with increased export subsidies of their own. The Uruguay Round of GATT negotia- tions was a natural venue for mutual agreement to rein in this costly competition. After long and tortuous negotiations, the finalized Uruguay Round Agreement on Agriculture (URAA) contained disciplines that, together with strategic rethinking and changed grain market conditions, greatly reduced the role of export subsidies after the mid-1990s in both the United States and the European Union. Both have agreed such subsidies should be outlawed, which has allowed farm policy discus- sion in the Doha Round of the GATT's successor, the World Trade Organization (WTO), to focus primarily on market access (import protection via tariffs and quotas) and domestic price supports via subsidy payments to producers. Import protection measures A hundred years before the first domestic commodity support programs were enacted, international trade policy was already a hot political issue in the United States. Manufacturing interests of the North wanted protection and, after 1820, received it in the face of opposition from Southern agricultural interests, who bought imported manufactured producer and consumer goods and also linked their capacity to export some products, especially cotton, to the U.S. willingness to 184 Distortions to Agricultural Incentives: A Global Perspective import from Europe. The North, to succeed politically, needed an alliance with the West, which was in place until the Jackson presidency, when his vetoes of legisla- tion for Western improvements changed the balance of interests in favor of lower tariffs--ultimately, the "free trade tariff " legislation of 1846, which generated average tariff rates of 25 percent in 1850 (calculated from tariff receipts as a per- centage of the value of dutiable imports), as compared to 57 percent in 1830. Higher tariff protection resumed after the Civil War, especially on manufactured goods, with average rates of about 40 percent between 1870 and 1910. Through- out this period, there were tariffs on imported agricultural products such as wool and sugar, but rates averaged about five times as much for manufactured imports as for agricultural imports (Davis, Hughes, and McDougall 1965). Table 4.1 shows the evolution of U.S. protection as measured by customs duties as a percentage of the value of imports. This measure does not always provide a Table 4.1. Customs Receipts as a Percentage of Value of Imports, United States, 1821­2000 (percent) Manufactured Agricultural All products products merchandise 1821 47 47 43 1830 64 62 57 1840 21 00 18 1850 27 3 25 1860 19 2 18 1870 46 47 50 1880 39 6 29 1890 41 8 30 1900 40 15 28 1910 29 9 21 1920 7 2 6 1930 18 8 15 1940 15 7 13 1950 9 2 6 1960 -- -- 7 1970 -- -- 7 1980 -- -- 3 1990 -- -- 3 2000 -- -- 1.6 Sources: Davis, Hughes, and McDougall (1965) for manufactured and agricultural products; Carter (2006) for all merchandise. Note: -- not available. United States and Canada 185 good indicator of the trade effects of tariffs, however, notably because a tariff so high as to shut off all imports generates no customs duties and so counts the same as free trade in calculating the numerator of this measure. Irwin (2006) reviews this and other shortcomings of that table 4.1 indicator, and estimates a trade restrictiveness index that takes into account the effect of a tariff on imported quantities and the fact that the distortive effects of a tariff increase more than pro- portionately with the height of the tariff. His estimates indicate that for all mer- chandise trade, while customs duties as a percentage of all imports had fallen to 60 percent of the 1875 level in 1931, the TRI in 1931 remained at 97 pecent of its 1875 level--that is, the reduction in the crude measure greatly overstates the extent of liberalization. But between 1931 and 1960 the TRI fell faster than the crude measure. By 1960, the extent of overall trade liberalization between 1875 and 1960 was about the same for both measures. That is, by either measure, the restrictivenes of tariffs in the post­World War II period was about one-fourth the level of the late 19th century (Irwin 2006). Because the United States has always been a net exporter of agricultural goods, and imported manufactured goods are used directly and in the production of inputs used in farming, the tariff structure effectively taxed agriculture through- out the 19th century. The estimates of Irwin (2006) indicate that for the 1870­1900 period, an average 30 percent protection on imported goods (duties as a percentage of all import values including duty-free imports) generated a net subsidy to import-competing manufacturers of 15 percent and a net tax on agri- culture (and other exporters) of about 11 percent.6 This situation was reversed by increasing protection (through both import restrictions and domestic support) of agriculture after 1920. Import duties on wheat, maize, wool, sugar, and meat were raised sharply in "emergency" legislation of 1921 and the Tariff Act of 1922 when "the representatives of the agricultural states had committed themselves to a policy of high and even ruthless protection" (Taussig 1931, pp. 452­53). At the same time, imports of many manufactured products used in farming were made duty-free.7 After 1910, U.S. manufacturing became sufficiently export-oriented itself as not to be the strong political force for protection it once was, and after World War II protection of manufacturing steadily declined to the point that by 2000, import duties as a percentage of aggregate import value had declined to 1.6 percent (albeit with additional signficant protec- tion of politically sensitive sectors such as textiles through nontariff barriers). Tar- iffs on agricultural products were also reduced, though protection from imports increased due to quantitative restrictions. In short, between the broad periods of 1820­1900 and 1930­2000, there was a substantial turnaround in manufacturing as compared to agricultural import protection. However, it was already apparent in the 1920s that too much agricultural production was exported to make import 186 Distortions to Agricultural Incentives: A Global Perspective protection effective in alleviating farmers' losses in the post­World War I price plunge, and this changed the focus of farm policy in terms of other measures. Overall, protection of agricultural products relative to merchandise has increased over the past 50 years. Though rates of import protection of the agricul- tural and manufacturing sectors were equal in the late 1950s and 1960s, protec- tion of manufacturing has since fallen faster than assistance to agriculture. It should be noted, however, that comparing market-distorting agricultural protec- tion with merchandise tariff rates has become decreasingly relevant in capturing the main U.S. import protection elements in either agriculture or manufacturing, because in agriculture and in industries such as steel, automobiles, and textiles, the more important distortions of international trade have become quantitative restrictions, often in the form of "voluntary restraint" agreements between the United States and exporting countries. More economically relevant measures compare internal U.S. prices with international prices for the same goods, as dis- cussed below. Nonetheless, the picture remains essentially the same--that govern- mental action to assist agriculture is increasing in impact compared to action to protect manufacturing. Export subsidy measures An indicator of the role of export subsidies can be obtained from the value of sub- sidies paid per unit quantity of exports. In the United States, these have been most significant for wheat, as noted above. Subsidies were as high as US$1.3 billion for exported wheat in 1993 under the EEP. They were targeted to particular sales, however, and not available for all exports. This raised questions about their efficacy in actually increasing export quantities. By focusing subsidy funding on particular sales, the government's expenditures were made more effective at increasing those targeted sales. Subsidized wheat received payments as high as US$43 per ton in 1991. But at the same time, large sales to other importers, notably Japan, were not subsidized at all. Because of the targeted nature of the subsidies, the proper measurement of the subsidy, not to mention its effects on net exports, is difficult to nail down (as is the effects of the wheat export embargoes of the 1970s).8 The USDA budget provides information on the range of export promotion activities. In fiscal year9 2005, as in the several years prior, there were no export subsidies under the EEP or the more recent Dairy Export Enhancement Program. A total of US$2.2 billion was spent on USDA's Foreign Agriculture Service (FAS) programs, but US$1.7 billion of this was on food aid programs. Market development programs and export credit guaran- tees are export promotion activities that also have features making them similar to export subsidies. In fiscal 2005, FAS spent US$184 million on market development programs such as sending teams abroad to make the case for U.S. commodities United States and Canada 187 and informational campaigns in the United States and abroad. Export credit guar- antees covered US$2.6 billion in sales during fiscal 2005 (at a budgetary cost of US$137 million, which is a rough indicator of the subsidy element of the pro- gram). In response to a WTO dispute resolution panel in 2005, the United States has eliminated some high-risk countries from the guarantees and made other changes "intended to remove any long-term subsidy component of the program" (USDA 2006, p. 38). Unlike import restrictions, whose effects can be estimated by comparing internal and world reference prices because they are focused on a few commodities and create big effects, these export programs are spread so thinly across many commodities that it would not be credible to attribute to them any observed elevation in U.S. commodity prices relative to foreign prices for the same commodities. Rather, it makes more sense to treat such expenditures together with domestic programs as elements of overall non-product-specific (NPS) support, following Organisation for Economic Co-operation and Development (OECD) practice. Direct producer support through commodity programs Payments to producers have been a central element in U.S. agricultural policy since the first New Deal programs of 1933. Figure 4.2 shows payments in billions of U.S. Figure 4.2. Expenditure on Commodity Programs and Payments to Farmers, United States, 1955­2005 45 40 35 2000 US $ billions 30 25 20 15 10 5 0 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 spending level payments to farmers Source: Author's compilation from the Congressional Budget Office (CBO), Washington, DC. 188 Distortions to Agricultural Incentives: A Global Perspective dollars in real terms using the GDP deflator of 2000 100. For comparison, the figure also shows the total level of spending on "commodity stabilization and sup- port" as measured by the U.S. Office of Management and Budget. Spending levels are higher than payments prior to 1990 because much of the spending was for removing products from the market using stockpiling programs, purchases for sale to schools and other food assistance, or subsidized sales abroad. In recent years, spending has been overwhelmingly dominated by direct payments to producers.10 Summarizing the effects of U.S. direct support programs is difficult because of the variety of policy instruments used and their evolution toward being increas- ingly decoupled from production decisions over time. Throughout the 1950s, the price suppport loan rates at which farmers could forfeit basic commodities to the CCC remained a primary policy instrument, and were reinforced by payments for idling land. By the 1960s, lower loan rates and direct payments to reduce forfeitures at a supported price entered the policy arsenal. In the 1970s, direct payments were further institutionalized in the form of "target prices" and "deficiency payments" to farmers when the price of basic crops they produced fell below the target. By the 1980s, deficiency payments were being made on only 85 percent of a farmer's base acreage, and for a historical level of "program yields." Deficiency payments thus came to be somewhat decoupled from production decisions, although the specific basic crop for which payments were received had to be grown on the correspon- ding base acreage. Farmers also had to comply with annual acreage reduction requirements to be eligible for deficiency payments. Annual acreage reductions idled at an average of 46 million acres during 1983­88 as world agricultural prices collapsed in the mid-1980s, and then fell to much lower levels. Income support in the 1996 and 2002 Acts The most important change in support to farmers contained in the 1996 Farm Act was its Title 1, the Agricultural Market Transition Act (AMTA). With rising world prices for U.S. farm commodities in 1995 and 1996, AMTA replaced target prices, deficiency payments, and annual acreage set-asides for wheat, rice, feed grains, and cotton by a scheme of fixed "production flexibility" payments. The payments were based on amounts farmers received, or would have received if they had partici- pated, in the pre-1996 deficiency payment program. The amount of the payment was independent of prices and was fixed by each farm's production history. It could not be increased or decreased by changes in the farm's acreage or yield of the pro- gram crops, while the program allowed farmers to plant a wide range of crops on base acreage (the reasons for the "production flexibility" label). The production flexibility payments were in this sense further decoupled from both prices and pro- duction, although payments were not divorced from all production decisions in United States and Canada 189 that producers lost payments if they increased plantings of nonsupported fruits or vegetables or if they left farming completely. The aggregate of payments was sched- uled to decline from about US$6 billion annually in 1996 and 1997 to US$4 billion in 2002. AMTA payments constituted the bulk of projected commodity support in 1996, and the projections were for substantially less government spending on com- modity programs than had occurred before 1995, as figure 4.3 shows. The 2002 Farm Act Within a year after the 1996 Farm Act was introduced, commodity prices began to fall substantially. By 1998, as the Asian financial crisis and China's lack of an expected increase in imports weakened world demand and hence U.S. exports and prices, it became clear that prices were likely to remain at low levels for some time. These low prices triggered two policy responses, one automatically and one through Congressional response. The automatic response came through the "marketing loan" program. This program was introduced in 1990 to replace the former loan rate program of supported market prices through forfeitures of commodities at an established "loan rate" price. The forfeiture program created a market price floor at the loan-rate price, because farmers could always get that price through the loan program. To keep the program from resulting in the government accumulating Figure 4.3. CCC Commodity Program Outlays, United States, Fiscal Years 1980­2007 35 30 25 US$, millions 20 15 10 5 0 1980 1985 1990 1995 2000 2005 2010 CCC outlays 1996 projection 2006 revision 2002 projection Source: Author's compilation from the CBO, Washington, DC. 190 Distortions to Agricultural Incentives: A Global Perspective unwanted stocks of commodities, loan-rate prices were mostly kept below actual average prices after the 1960s. The marketing loan program precluded stock accu- mulation altogether by having the government not actually acquire grain, but instead by offering farmers a "loan deficiency payment" equal to the difference between the policy-determined loan-rate price (which varied from county to county) and the local county price. Because the loan-rate prices were set below average prices, hardly any loan deficiency payments were made during 1990­97. But by 1998, commodity prices fell sufficiently below loan-rate levels to trigger US$500 million in loan deficiency payments. Unlike the target-price-related defi- ciency payments paid on a historically fixed level of output, the loan deficiency payments are made on all of current output. In fiscal years 1999, 2000, and 2001, these payments rose to US$3.4 billion, US$6.4 billion, and US$5.3 billion, respec- tively, as all the grains, cotton, and rice experienced continuing historically low market price levels. The Congressional response to these low prices was to enact emergency "market loss assistance" programs in each of the years 1998­2001. These pro- grams provided an average of US$4.8 billion annually for those four years and added proportionally to each farmer's production flexibility payments (in most cases doubling them during 1999­2001). Together, the market loss assistance pro- gram and production flexibility payments were the main factors behind the huge expansion of commodity program spending during 1998­2001 as compared to the levels anticipated in 1996, shown in figure 4.3. Actual spending in those years averaged more than US$20 billion, rather than the US$6 billion that was expected. By 2001, it was clear that Congress was dissatisfied with the political wrangling involved with annual emergency legislation and that more permanent revision of the 1996 Farm Act would be legislated. The Farm Security and Rural Investment Act of 2002 provided for additional spending on farm programs of more than US$50 billion above the projected baseline spending if the 1996 Farm Act had been continued. This increase was possible because at that time the U.S. budget was--and was expected to continue to be--in surplus, and because farm com- modity producers had sufficient political power to defeat competitors for the funds available. This expanded commitment is the most significant--and to some the most shocking--aspect of the 2002 Farm Act.11 In addition to the issue of spending levels, the 2002 Act addressed structural issues in terms of the form and scope of farm subsidies. Some farm groups, mainly centered on wheat growing in the Great Plains, wanted to return to supply management, in order to reduce production and increase commodity prices. Environmental groups pushed to have a substantial part of the new spending allo- cated to conservation and environment-improving programs. As enacted, the 2002 legislation continued the production flexibility payments from 1996 United States and Canada 191 (renamed fixed direct payments). Though supply management was rejected, a new "countercyclical payment program" was added to provide payments that rise or fall inversely with market prices (although the quantity base for payments remained fixed, with an option for one-time updating to 2002, for each farmer). This amounted basically to a reinstitution of pre-1996 deficiency payments, but without annual acreage reduction set-aside requirements and with farmers retain- ing the planting flexibility introduced in 1996. In addition, earlier production quotas for peanuts were replaced with strengthened support programs similar to the programs for other crops, and a new program of direct support payments for milk was added. Conservation and environmental programs ended up with a sub- stantial share of the new spending, but not as much as the proponents of these programs had argued for. The Congressional Budget Office (CBO) estimated that the innovations of the 2002 Act would cost an average of US$8 billion over the 10 fiscal years 2002­11.12 Of this, US$4.5 billion was for direct payments to farmers under either the fixed payments or the mandated new countercyclical payment program. It was esti- mated that the Act would lead to 10-year spending increases of US$0.5 billion in marketing loans and loan deficiency payments, US$0.5 billion for the new peanut program, US$160 million for the new dairy program, and US$43 million for increasing support in the sugar program, partly offset by savings projected at US$26 million from tightening payment limitation slightly, for a total of a US$5.7 billion spending increase for all commodity programs. Meanwhile, new initiatives in the Conservation Reserve Program and the Wetlands Reserve Program (both long-term, paid land-idling programs that have enrolled a total of more than 30 million acres since 1990), as well as the Environmental Quality Improvement Program, the Farmland Protection Program, and a new Conservation Security Program were projected to cost US$1.3 billion annually over 10 years (FAPRI 2002). The expanded costs of these programs consisted mostly of payments made to farmers to encourage the use of soil con- servation or water quality improvement practices. Some of the practices reduce production, such as replacing cultivated crops near streams or lakes with "filter strips" of grass in the Conservation or Wetlands Reserve Program. Other "working lands" programs are roughly neutral with respect to production or even result in an increase in production, such as the Farmland Protection Program payments to farmers in exchange for their maintaining land in farming rather than selling it for commercial development. Total outlays in fiscal 2003 and 2004 were lower than projected because higher- than-expected commodity prices resulted in less spending than had been forecast on countercyclical payments and loan deficiency payments. But in fiscal 2005, these payments increased as commodity prices fell. 192 Distortions to Agricultural Incentives: A Global Perspective Overall summary of market-distorting support Table 4.2a shows the nominal rates of assistance (NRAs) for key U.S. agricultural products, which are partly based on the OECD's PSE-CSE Database for 1986­2007, and, for years back to 1955, on the author's use of the OECD's method. This measure of NRAs excludes the deficiency, production flexibility and fixed direct payments, and the market loss assistance and countercyclical pay- ments. These various support payments have been at the heart of U.S. farm poli- cies, as described earlier, but have become increasingly decoupled from produc- tion decisions and arguably can now be concluded to have minimal or at least small effects on production or trade.13 The main policies included in the NRAs in table 4.2a are import protection, most notably for sugar, but also for dairy prod- ucts and meats; export subsidies, most notably the EEP and earlier export subsi- dies for wheat, but also affecting rice and poultry in some years; and the various loan-related payments to producers that a producer can increase by increasing output, which are closest to a classical production subsidy. Plotting these data sep- arately for exported products and for imported products that compete with U.S. products shows somewhat higher rates of assistance for importables (figure 4.4a). Given the dominance of exportables in overall U.S. production, its line in figure 4.4a is close to the line for all covered products. NRAs vary greatly across U.S. agricultural commodities, as summarized in the dispersion measure shown in the second-to-last row of table 4.2a, and even more so if the other products that get negligible support are considered (the latter accounting in aggregate for one-third of the value of U.S. agricultural production at undistorted prices, see bottom row of table 4.2a). Gardner (2008, appendix table 4.4) describes support through various gov- ernment payments (including those excluded from table 4.2a, termed here as "decoupled") for the top 25 commodities, accounting for 91 percent of the U.S. farm value of production in 2004. Overall, commodities accounting for 42 per- cent of production received significant support, with payments amounting to 6.9 percent of the value of production in 2004. Yet 14 of the top 25 commodities, and a total of 58 percent of the value of production, received no significant support. Appendix table 4.5 of Gardner (2008) shows a broader picture of U.S. federal government activity in support of agriculture. In addition to the commodity programs, there are conservation programs, export programs, governmentally underwritten loan programs for farmers, crop insurance, research funding, and marketing and regulatory programs. These additional activities had a price tag of US$15 billion in 2005.14 The sum of US$34.1 billion for fiscal 2005 amounts to 14 percent of the market value of U.S. farm cash receipts for all crops and livestock. United States and Canada 193 Figure 4.4. NRAs to Exportable, Import-Competing, and All Covered Agricultural Products, United States and Canada,a 1955­2007 a. United States 70 60 50 40 percent 30 20 10 0 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 b. Canada 70 60 50 40 percent 30 20 10 0 1961 1966 1971 1976 1981 1986 1991 1996 2001 2006 import-competing products total exportables Source: Anderson and Valenzuela (2008), drawing on authors' spreadsheet. a. Includes the NRAs shown in table 4.2 and what is termed here as NPS assistance, all of which is attributed to tradables. NPS assistance for 1986­2007 includes payments classified by OECD as based on input use (E1) and miscellaneous payments (H). The total line does not include what here is termed "decoupled" assistance. Table 4.2. NRAs to Covered Farm Products, United States and Canada,a 1955­2007 (percent) 194 a. United States 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Exportables 1.0 1.1 6.7 5.7 2.1 5.0 10.5 5.5 5.1 8.1 4.8 Cotton 5.8 0.2 65.2 43.2 9.2 15.6 32.6 24.6 27.8 70.0 77.0 Eggs -- -- -- -- 4.1 5.8 4.6 8.1 2.2 0.0 0.0 Maize 0.0 1.4 2.4 3.4 0.4 2.6 10.4 0.6 4.8 7.0 6.6 Poultry 0.0 0.0 0.0 0.0 0.0 0.0 7.0 1.3 0.3 0.0 0.0 Rice 0.0 0.0 0.0 0.0 2.6 15.6 38.8 20.1 7.8 52.7 2.2 Sorghum -- -- -- -- -- 37.3 15.0 0.2 4.8 5.5 7.2 Soybeans 0.0 0.0 0.0 0.0 0.0 0.0 0.6 0.1 6.6 12.4 0.6 Wheat 0.0 3.6 17.8 14.4 9.6 11.2 29.1 25.5 5.4 4.3 0.2 Import-competing products 8.8 9.4 9.4 6.6 9.8 21.1 25.2 19.9 9.5 17.5 8.9 Barley -- -- -- -- -- -- 61.9 59.6 4.8 5.3 3.2 Beef 0.0 0.0 2.0 2.0 2.0 2.0 1.3 1.0 0.0 0.1 0.0 Milk 20.2 23.6 21.0 17.0 25.4 61.8 96.9 59.9 78.8 66.5 24.2 Pig meat 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.0 0.0 0.0 0.0 Sheep meat -- -- -- -- -- 7.1 2.4 1.2 3.4 13.8 9.9 Sugar 50.4 65.9 134.3 18.2 40.3 120.2 158.3 78.8 96.1 115.6 47.6 Wool -- -- -- -- -- -- 1.2 0.9 0.9 16.1 28.2 Nontradables 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Potatoes 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 All covered products 4.7 4.8 7.8 5.9 5.1 11.1 16.2 11.2 6.2 11.6 6.3 Domestic market support 0.0 0.0 0.0 0.0 0.1 0.9 2.3 0.4 2.2 3.6 1.0 Border market support 4.7 4.8 7.8 5.9 5.0 10.2 13.9 10.8 4.1 8.0 5.3 Dispersion of NRA of covered products 17.3 23.1 46.6 15.7 14.6 39.8 55.3 27.6 31.4 38.4 23.3 % Coverage at undistorted prices 66 66 66 69 70 69 66 67 65 66 68 b. Canada 1961­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Exportables 2.5 2.4 2.2 2.8 4.3 8.2 2.4 1.4 1.4 0.4 Barley 2.0 2.0 2.0 5.8 8.0 10.4 2.5 1.4 2.8 1.0 Beef 1.0 1.0 1.0 1.2 2.6 4.0 3.9 n.a. n.a. n.a. Peas 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Pig meat 2.0 2.0 2.0 2.0 2.0 6.3 5.1 3.8 2.0 0.4 Rapeseeds 0.0 0.0 0.0 0.0 0.0 9.6 0.5 0.1 0.1 0.0 Soybeans n.a. n.a. n.a. n.a. n.a. 2.8 0.4 0.5 1.2 0.3 Wheat 3.0 3.0 3.0 3.1 5.3 10.4 1.4 0.6 0.6 0.2 Import-competing products 15.7 12.0 12.2 32.3 47.2 53.1 44.0 41.3 49.1 46.9 Beef n.a. n.a. n.a. n.a. n.a. n.a. n.a. 2.0 1.8 0.7 Egg -- -- -- -- 41.1 30.1 33.1 27.7 10.6 64.7 Maize 4.9 4.7 3.5 2.3 2.0 12.8 2.5 3.3 8.7 2.9 Milk 34.4 34.4 34.4 162.8 307.2 314.2 182.6 109.4 125.9 94.4 Poultry 17.0 17.8 33.5 24.4 23.3 23.7 25.2 2.7 2.6 6.9 Soybeans 0.0 0.0 0.0 0.0 0.0 n.a. n.a. n.a. n.a. n.a. Sugar -- -- -- -- 32.5 79.5 31.0 17.7 -- -- Nontradables 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Potatoes 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 All covered products 8.7 7.2 7.0 16.0 22.5 28.6 20.8 12.3 13.5 11.4 Domestic market support 0.0 0.0 0.0 0.1 0.5 13.2 4.1 1.9 1.4 0.4 Border market support 8.7 7.2 7.0 15.9 22.1 15.4 16.6 10.4 12.1 11.0 Dispersion of NRA of covered products 11.1 11.0 13.8 47.5 84.5 87.2 50.1 31.1 35.9 32.5 % Coverage at undistorted prices 71 76 79 84 86 83 81 79 76 76 Source: Anderson and Valenzuela (2008), drawing on author's spreadsheet. Note: -- not available; n.a. not applicable. 195 a. The main U.S. policies included are import protection, export subsidies, and the various loan-rate-related price support and payments to producers. Excludes deficiency, production flexibility and fixed direct payments, and the market loss assistance and countercyclical payments (see text for discussion). Weighted averages use weights based on the unassisted value of production. Dispersion is the standard deviation shown as the simple five-year average of the annual standard deviation around the weighted mean. 196 Distortions to Agricultural Incentives: A Global Perspective Agricultural Policies in Canada Canada, which remained a British colony during the early period of trade policy examined here, did not introduce substantial tariff protection of manufacturing until the 1870s, and then at less than the U.S. levels (Fowke 1946). Agricultural protection became important earlier, notably with wheat import tariffs directly against U.S. exports (and U.S. duties levied on Canadian grain exports). The sig- nificance of particular policy steps has been debated, but a convincing case can be made that "The tariff (of 1843) was a true break in the old colonial system, and was brought about, not by capitalists seeking to establish new industries, but by pioneer farmers trying to exclude outsiders from the local markets" (Jones 1941, p. 537). Thus, while attempts to measure the relative protection of manufacturing and agriculture comparable to those cited above by Irwin for the United States for the 19th century are not available for Canada, it seems clear that agriculture was never implicitly taxed in Canada the way it was in the United States (or in Australia and New Zealand--see Anderson et al. 2009). With respect to the context of agricultural support, Canadian agriculture varies substantially from east to west in cropping patterns and size of farms, in ways roughly parallel to corresponding U.S. areas south of the border. In those parallel areas, there are many similarities, notably between the Prairie Provinces and the Northern Plains. However, the history of policy and current practices in the two countries are quite distinct in several respects. The Great Depression of the 1930s hit Canadian farmers as hard as it hit those in the United States. In both countries, the effect was particularly severe in the prairies. The Canadian government provided some debt relief through the Prairie Farm Rehabilitation Act of 1935, but Canada did not introduce an integrated set of programs or make substantial income transfers to farmers as the New Deal did in the United States. Indeed, it is arguable that at that time and still to the present day, while a series of ad hoc programs have been implemented, Canada has no compre- hensive farm policy comparable to the omnibus U.S. farm bills. This is attributable in part to Canada's comparably more decentralized political system, under which many policies are carried out at the provincial rather the federal level, notably in marketing, where provinces have set up marketing legislation for fruits and animal products going back to 1926 (see Schmitz, Furtan, and Baylis 2002). In Canada, the first major federal support similar to that of the United States was provided under the Agricultural Stabilization Act of 1958. This program guaranteed producers 80 percent of the average price over the previous 10 years, by means of payments from the federal government. Only small payments were made, and in 1975 the payment trigger was raised to 90 percent of a five-year moving average price and a mechanism was introduced for compensating farmers for rapid increases in cash costs of production, which "indicated the Canadian United States and Canada 197 Government's continuing commitment to ensuring short-term solvency in agri- culture" (OECD 1978, p. 29). Canada's agricultural policies for grains and oilseeds have followed a path sim- ilar in some respects to those of the United States, but with less reliance on crop supply management through acreage controls and more reliance on collective marketing, most notably through the CWB. The CWB has a legal monopoly on the sales of Canadian wheat and barley into foreign markets--all exports as well as sales of wheat to domestic millers of flour must be sold through the agency. Farmers then receive a pooled price depending on the receipts the CWB is able to earn from the exported and domestically sold commodities. The monopoly pow- ers of the CWB have been challenged in Canadian courts, much as the original U.S. programs of the 1930s were in U.S. courts, on constitutional grounds. In a case brought in 1994, a group of barley growers sued the CWB, arguing that it "breached the rights of individual farmers guaranteed under the Canadian Char- ter of Rights and Freedoms" (Schmitz and Furtan 2000, p. 145). The plaintiffs' plea was rejected, and following Canadian law, they had to pay court costs of the defendants--the government of Canada. The year 1995 was a watershed in that, under pressure because of federal budg- etary deficits and the URAA's disciplines on export subsidies, transportation sub- sidies under the Western Grain Transportation Act were ended. These had cost an average of about US$12 per ton of grain, and their elimination was estimated to have saved the federal government about US$400 million in 1995. With respect to income support for producers, Canada has undertaken far- reaching experimentation in its series of grain programs over the last 30 years. The Western Grains Stabilization Program (WGSP), initially enacted in 1976, made payments from a fund partly financed by growers when their aggregate cash receipts from grains fell below a five-year moving average. After accumulating large deficits without providing satisfactory income protection to producers, the WGSP was abandoned. The Farm Income Protection Act of 1991 marked an important turning point in Canada's approach to farm support in crop production, when it moved away from policies aimed at particular commodities and toward a whole-farm approach. The 1991 Act introduced the Gross Revenue Insurance Program (GRIP) and the Net Income Stabilization Account (NISA). These programs were tailored to each producer's situation, with the GRIP making crop-specific pay- ments to producers when their production multiplied by the marketwide average market price fell below that producer's established average yield multiplied by a "target" price. The program was packaged as insurance in that each producer paid a premium for this coverage (though approximately two-thirds of the premium cost was paid by a combination of provincial and federal funds). 198 Distortions to Agricultural Incentives: A Global Perspective The GRIP combined features of the U.S. deficiency payment and subsidized crop insurance programs, and its comprehensive approach has attractive features. But it proved to have too little political support from farmers to justify its budgetary costs in the belt-tightening environment of the mid-1990s, and the GRIP expired after 1995. NISA, a more broadly conceived (in that it avoided support of specific com- modities) and less costly program, continued for another decade. NISA is essentially a subsidized savings account into which producers may contribute 2 percent of the value of qualifying grain sales, to be matched by 1 percent each from provincial and federal governments. The producer may withdraw funds from the account if either annual farm operating income or family income falls below established triggers (see Huff [n.d.] and Gray and Smith [1997] for further discussion). In 1998, Canada introduced the Agricultural Income Disaster Assistance (AIDA) program. Under that program, funded 60 percent by federal and 40 per- cent by provincial governments, anyone who files income tax returns as a farmer qualifies for an indemnity payment if their gross returns fall below 70 percent of the similarly calculated returns over the average of the three preceding years (Schmitz, Furtan, and Baylis 2002). In 2001, the Canadian Farm Income Program (CFIP) replaced AIDA. The experimental nature of farm income support has been intensified since 2004 with the phasing out of NISA and AIDA, and replacement with new and ad hoc programs. The Canadian Agricultural Income Stabilization (CAIS) Program, introduced in fiscal 2003/04, combines insurance and income support features. It makes payments to producers when a farmer's "production margin" falls below the "reference margin" for the farm, calculated from previous years' experience. The reference margin is a measure of returns minus costs that counts fewer expense items than under previous programs, because "experience with previous farm programs such as the CFIP indicated that including a high number of allowable expenses often resulted in reference margins being low and in many cases negative. This often resulted in producers being ineligible for benefits" (Agriculture and Agri-Food Canada 2006). In response to the livestock losses resulting from the Bovine Spongiform Encephalopathy (BSE, "mad cow disease") crisis, which devastated the beef export business, and as a "bridge" to the CAIS Program, the Transitional Income Support Program was introduced in 2004 to assist both livestock and grain producers. This program, together with more liberal payouts under the CAIS, raised the costs of farm support in 2003­05 as compared to earlier years. In 2006, the government introduced payments under the Grains and Oilseeds Program that will further increase support. The 21st-century programs are organized and marketed under the Agricultural Policy Framework, which the government of Canada and the provinces and United States and Canada 199 territories agreed upon in late 2003 to coordinate agricultural policy in five areas: business risk management, environment, food safety and quality, innovation, and renewal (see Agriculture and Agri-Food Canada 2005). The latter two areas include what are traditionally described as rural development and research and extension programs. Business risk management covers CAIS and the more recent payment programs. All of the programs have notable differences in funding and delivery from province to province. Unlike for grains, Canada has maintained supply management programs with strong control measures for dairy, poultry, and eggs. This is in sharp contrast to the United States, where supply management in livestock is entirely absent.15 Canada's supply management history grew out of practices established by provincial mar- keting boards, which were well established by the 1960s (Schmitz, Furtan, and Baylis 2002). The Canadian Dairy Commission, created in 1966, introduced supply management ideas that culminated in market-share quotas, under which a farmer must have an established quota in order to sell milk. Import quotas (converted to tariff-rate quotas under the URAA) keep milk from entering Canada at the high established retail domestic price, while milk for processed dairy products is sold at lower prices so that it is competitive in world markets. Similar but less complex supply management programs exist for broilers (chickens), eggs, and turkeys. Overall summary of market-distorting support Table 4.2b shows the nominal rates of assistance (NRAs) for key Canadian agri- cultural products, which are partly based on the OECD's database of producer support estimates (PSEs) for 1979­2007, and, for years back to 1961, on the author's use of the OECD's method. The key products cover between 75 and 85 pecent of Canadian agricultural production. The average NRA for those covered products was around 8 percent up to the mid-1970s, rose to 28.6 pecent during the export price war period of 1985­89, and has since come back to around 12 percent. The dispersion in NRAs across the product range rose substantially up to the mid-1980s, though it has more than halved since then. As usual, the estimated NRA for importables is well above that of exportables according to the classification based on trade status presented here--and well above the average importables NRA for the United States. Agricultural Relative to Nonagricultural Support To create a more complete picture of the policy distortions to farmer incentives in North America, guesstimates of the NRAs for noncovered products (a weighted average across exportables, import-competing products, and nontradables) are 200 Distortions to Agricultural Incentives: A Global Perspective first provided. NPS support--that is, input subsidies and decoupled support, as measured by certain OECD PSE categories--is then considered over the past three decades. The NRA for tradable farm products (including NPS support but not decoupled payments) is also compared with the NRA for nonagricultural trad- ables by calculating a relative rate of assistance (RRA). These results are summa- rized in table 4.3. Input subsidies and other NPS assistance is of nontrivial importance to the overall NRA for agriculture in North America as compared with the rest of the world, adding between one-quarter and one-third to the sectoral NRA in the United States and only slightly less in Canada. Payments based on input use (OECD E1, as defined in the PSE-CSE Database) and miscellaneous payments (OECD H) are included in this category. Of even more significance than NPS assistance are decoupled payments, which are assumed to encourage production less than market price support policies. Following certain OECD categories, not only the various commodity income support payments described above as increasingly decoupled from prices and production over time, but also payments made for long-term acreage idling under programs such as the Conservation Reserve Program, are examined here, as well as subsidies for crop insurance and ad hoc annual disaster payments (see note to table 4.3 for the OECD categories included).16 As can be seen from table 4.3a and figure 4.5a, those decoupled payments nearly doubled the NRA for the United States in the mid-1980s and added about one-third in the 1990s and the present decade. For Canada, decoupled payments have been somewhat less important, although still nontrivial (table 4.3b and figure 4.5b). As pointed out, it is difficult to summarize the effects of U.S. support pro- grams because of the variety of policy instruments used. Although any estimate is conjectural, the set of most reasonable estimates indicate that in 1999­2005, the marketing loan program has increased U.S. output of grains and soybeans by about 2 percent, the direct payment program (including the 2002 Farm Act changes) by about 1 percent, and crop insurance subsidies by more than 1 per- cent, for a total effect of 4 to 5 percent more of these commodities being pro- duced than would have been in the absence of commodity support programs. The long-run consequences (on conservation, agricultural research, and technol- ogy adoption) of commodity support policies are quite a different matter not considered here. Looking at the levels of border protection and domestic support payments themselves, overall government support for farmers in the United States rose in the decade or so between the mid-1970s and mid-1980s, after falling over the two decades prior. Support has since fallen back slightly, even when decoupled pay- ments are included, although the level was nearly as high during the period of low Table 4.3. NRAs to Agricultural Relative to Nonagricultural Industries, United States and Canada,a 1955­2007 (percent) a. United States 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 NRA, covered products 4.7 4.8 7.8 5.9 5.1 11.1 16.2 11.2 6.2 11.6 6.3 NRA, noncovered products 8.0 7.7 6.0 5.0 4.7 6.6 10.9 8.4 8.5 9.6 5.4 NRA, all agriculture (excluding NPS) 5.8 5.8 7.2 5.6 5.0 9.7 14.4 10.2 7.0 10.9 5.2 All importables 8.6 8.9 8.5 6.1 8.4 16.9 20.9 16.4 9.1 15.0 8.7 All exportables 2.7 2.7 6.5 5.5 2.6 5.3 10.6 6.2 5.8 8.4 4.7 All nontradables 8.0 6.8 5.1 4.3 4.2 5.7 9.4 7.2 7.3 8.2 2.3 NPS 7.0 5.2 3.3 1.1 1.2 1.8 2.7 3.7 3.3 4.9 4.3 NRA, all agriculture (including NPS) 12.8 10.9 10.5 6.9 6.2 11.5 17.1 13.9 10.2 15.6 9.8 NRA, decoupled payments 0.0 0.0 0.0 0.0 0.1 5.2 9.3 5.9 6.3 8.4 5.6 NRA, all agriculture (including NPS and decoupled) 12.8 10.9 10.5 6.9 6.3 16.7 26.4 19.9 16.5 24.1 15.3 NRA, all agricultural tradables (including NPS) 12.5 10.8 10.8 7.0 6.2 12.0 18.0 14.4 10.4 16.5 9.9 NRA, all nonagricultural tradables 6.1 7.3 7.4 5.4 3.8 3.5 3.5 3.2 2.1 1.5 1.4 RRAb 6.0 3.2 3.2 1.5 2.2 8.2 13.9 10.8 8.1 14.7 8.4 (Table continues on the following page.) 201 Table 4.3. NRAs to Agricultural Relative to Nonagricultural Industries, United States and Canada,a 1955­2007 202 (continued ) (percent) b. Canada 1961­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Covered products 8.7 7.2 7.0 16.0 22.5 28.6 20.8 12.3 13.5 11.4 Noncovered products 6.6 5.4 4.9 9.8 12.9 14.9 10.6 6.2 7.1 6.7 All agriculture (excluding NPS) 8.1 6.8 6.5 15.0 21.2 26.3 18.9 11.0 12.0 9.1 All importables 15.1 11.7 12.0 31.1 45.2 50.9 41.8 36.2 42.0 45.0 All exportables 4.3 3.5 3.0 4.6 6.3 10.1 4.8 2.9 3.2 0.9 All nontradables 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 NPS -- 4.9 4.0 5.7 9.1 13.2 12.8 7.8 12.1 2.2 Inputs -- -- -- 0.5 0.8 6.9 5.6 2.4 1.8 2.1 Other -- 4.9 4.0 5.2 8.3 6.3 7.2 5.5 10.4 0.1 All agriculture (including NPS) 8.1 11.7 11.0 20.8 30.3 39.7 31.6 18.9 23.9 11.4 Decoupled payments 0.0 0.0 0.0 0.7 4.7 8.6 8.0 5.3 10.3 11.8 All agriculture (including NPS and decoupled) 8.1 11.7 11.0 21.5 35.0 48.3 39.6 24.3 34.2 23.2 All agricultural tradables (including NPS) 8.1 11.7 11.0 20.8 30.3 39.7 31.6 18.9 23.9 11.4 All nonag tradables 9.2 6.9 6.0 5.1 4.8 3.9 2.6 1.1 0.8 0.8 RRAb 1.0 4.5 4.7 14.9 24.3 34.5 28.3 17.6 22.9 10.5 Source: Anderson and Valenzuela (2008), drawing on authors' spreadsheet and OECD (2008 and earlier years). Note: -- not available. a. Decoupled support includes "direct payments" in the years 1979­85. From 1986 those payments are specified to comprise the OECD's items C (payments based on area planted/animal numbers), D (payments based on historical entitlements), F (payments based on input constraints), and G (payments based on overall farming income). And for 2005­07, those items replaced by similar but newly defined items C to E. The values of those payments are estimated by the OECD. See text for discussion of NPS support. b. RRA is defined as 100*[(100 + NRAagt)/(100 + NRAnonagt) ­ 1], where NRAagt and NRAnonagt are the percentage NRAs for the tradables parts of the agricultural and nonagricultural sectors, respectively. United States and Canada 203 Figure 4.5. NRAs to All Agriculture without and with Decoupled Support,a United States and Canada, 1955­2007 a. United States 70 60 50 40 percent 30 20 10 0 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97 00 03 06 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 20 b. Canada 70 60 50 40 percent 30 20 10 0 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97 00 03 06 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 20 decoupled support excluded decoupled support included Source: Anderson and Valenzuela (2008), drawing on authors' spreadsheet and OECD (2008). a. Decoupled support includes "direct payments" in the years 1979­85. From 1986 those payments are specified to comprise the OECD's items C (payments based on area planted/animal numbers), D (payments based on historical entitlements), F (payments based on input constraints), and G (payments based on overall farming income). For 2005­07, those items replaced by similar but newly defined items C to E. The values of those payments are estimated by the OECD. NPS support is also from OECD estimates as defined in the note to figure 4.4 (see text for discussion). 204 Distortions to Agricultural Incentives: A Global Perspective world agricultural prices during 2000­04 as it was when prices were low in the late 1980s (see table 4.3). It dropped substantially after 2005 as international food prices spiked, but this does not represent a change in policy so much as the coun- tercyclical design of some U.S. payments and a reduced need to support farmers because of higher prices. Support for farmers in Canada too rose steeply in the decade or so from the mid-1970s, but has since fallen back even more than in the United States, especially when decoupled payments are taken into account. As in the United States, the NRA for Canada dropped substantially after the early 2000s as international food prices rose. By contrast, assistance to producers of nonagricultural tradable goods has been lower than for farmers throughout this period, and has declined more than for the farm sector. Hence, even leaving decoupled payments aside, the rate of assistance to producers of agricultural products relative to those producing nonagricultural goods is now considerably higher than in the early 1960s in the United States, other than in the years of historically high international prices, namely 1995 and 2006­07 (table 4.3a and figure 4.6a). The 2008 Farm Bill left the existing support programs in place and created others that could raise expenditures even if agricul- tural prices remain higher than they generally were during 1990­2006.17 Assis- tance to Canada's producers of nonagricultural tradable goods, too, has been lower than for farmers since the mid-1960s, and has declined more than for Canada's farm sector and more than in the United States. Though the RRA to the farmers became considerably higher in the 1980s than in earlier decades, in the 1990s it returned to levels similar to those in the United States, again with the low- est levels in the years of historically high international commodity prices, namely 1995 and 2007 (table 4.3b and figure 4.6b). When expressed in real (constant 2000) U.S. dollars rather than percentage price wedges, the decine in border and domestic support recently is much less evi- dent, because over time the value of the farm sector has grown and the number of farmers has declined. The peak real value of support in total dollars was higher in the latter 1980s than this decade, but when expressed on a per farmer basis it was even higher in 2000­04 (before international food prices rose) than in 1985­89. When the market price support component is expressed by product, the lion's share in recent decades has gone to dairy and sugar, which receive more border protection than cotton and maize (table 4.4). Consumer Tax Equivalents (CTEs) While much of the support for farmers worldwide comes from border meas- ures, in the case of North America large shares also come from NPS meaures and from payments that are decoupled somewhat from production. Hence, United States and Canada 205 Figure 4.6. NRAs to All Nonagricultural Tradables, All Agricul- tural Tradable Industries, and RRA,a United States and Canada, 1955­2007 a. United States 50 40 30 percent 20 10 0 10 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 b. Canada 50 40 30 percent 20 10 0 10 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 agricultural tradables nonagricultural tradables RRA Source: Anderson and Valenzuela (2008), drawing on author's spreadsheet. a. The RRA is defined as 100*[(100 NRAag t ) (100 NRAnonagt ) 1]. consumers in the region are not taxed to the same extent as producers are assisted--although taxpayers bear additional costs. Because of the differing net trade status of each product, the weighted average of ad valorem CTEs across all covered farm products also differs because consumption rather than production Table 4.4. GSEs of Assistance to Farmers, by Product, Per Farm Worker and Total, United States and Canada, 1955­2007 206 a. By covered product (constant 2000 US$ millions, excluding NPS and decoupled payments) 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Barley -- 14 19 44 97 141 518 392 48 53 20 Beef 0 25 442 555 541 549 388 307 53 69 25 Cotton 458 19 1,573 1,541 593 792 1,170 1,030 1,094 1,692 2,442 Eggs -- -- -- -- 395 495 302 432 174 34 167 Maize 0 273 509 998 163 848 1,927 152 825 1,316 1,081 Milk 3,365 4,253 4,116 3,829 7,775 13,778 12,986 9,875 3,761 8,125 6,666 Peas -- 0 0 0 0 0 0 0 0 0 0 Pig meat 0 31 41 56 55 52 124 297 68 46 12 Potatoes -- 0 0 0 0 0 0 0 0 0 0 Poultry 0 105 127 248 222 197 791 363 68 38 90 Rapeseeds -- 0 0 0 0 0 68 4 1 2 0 Rice 0 0 0 0 54 212 261 219 76 344 33 Sheep meat -- -- -- -- 40 40 14 5 12 40 28 Sorghum -- -- -- -- 436 636 261 4 40 40 36 Soybeans 0 0 0 0 0 0 91 18 734 1,274 90 Sugar 515 594 995 476 711 1,015 1,173 800 759 824 552 Wheat 0 317 1,308 1,266 1,248 1,689 2,108 1,737 333 240 16 Wool -- -- -- -- -- -- 1 1 0 3 5 b. Per person engaged in agriculture (constant 2000 US$, including NPS but not decoupled payments) 1961­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Total, North America 3,115 3,739 3,445 4,181 6,720 8,222 7,370 5,580 9,165 7,364 United States 3,383 3,886 3,431 3,728 6,303 7,710 6,831 5,185 8,795 7,279 Canada 1,590 2,914 3,524 6,478 8,857 11,412 11,379 8,591 12,056 8,034 c. Total (constant 2000 US$ millions) 1961­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 North America, covered products 5,632 9,130 9,014 12,331 20,445 22,185 15,636 8,044 14,142 11,264 United States 4,708 8,115 7,731 9,029 16,177 18,088 12,781 6,195 12,227 9,109 Canada 924 1,015 1,283 3,302 4,268 4,096 2,856 1,850 1,915 2,156 North America, covered and noncovered products 9,886 12,539 12,345 15,387 25,132 28,745 20,788 13,195 19,816 12,985 United States 8,552 11,263 10,821 11,816 20,473 24,224 17,599 11,100 17,591 10,717 Canada 1,334 1,276 1,524 3,570 4,659 4,521 3,188 2,095 2,225 2,268 North America, NPS 7,658 6,192 3,399 4,077 5,822 6,863 8,580 6,999 10,228 10,166 United States 7,658 5,269 2,417 2,678 3,815 4,548 6,397 5,491 7,974 9,598 Canada 0 923 981 1,399 2,007 2,315 2,182 1,508 2,254 568 North America, including NPS 17,544 18,731 15,743 19,464 30,954 35,608 29,368 20,194 30,044 23,151 United States 16,210 16,533 13,238 14,494 24,288 28,772 23,997 16,590 25,564 20,315 Canada 1,334 2,199 2,505 4,970 6,666 6,836 5,371 3,603 4,480 2,836 North America, decoupled payments 0 0 0 529 11,199 16,979 11,526 11,012 15,598 14,795 United States 0 0 0 352 10,193 15,529 10,177 9,990 13,657 11,826 Canada 0 0 0 177 1,006 1,451 1,349 1,022 1,941 2,970 North America, NPS and decoupled payments 17,544 18,731 15,743 19,993 42,153 52,587 40,894 31,205 45,643 37,947 United States 16,210 16,533 13,238 14,846 34,481 44,301 34,174 26,580 39,222 32,141 Canada 1,334 2,199 2,505 5,147 7,672 8,286 6,720 4,625 6,421 5,806 207 Source: Anderson and Valenzuela (2008) based on author's spreadsheet. Note: -- not available. 208 Distortions to Agricultural Incentives: A Global Perspective weights are used. There are also direct consumer subsidies in the United States, notably through the food stamp program, and in some years for some products those direct subsidies more than offset the tax component of the trade measures used there to support producer prices, resulting in negative CTEs. The rises and falls in the degree of distortion on the consumption side of the market can be seen in table 4.5. Table 4.5b shows how trivial these transfers from or to food consumers in North America have been on a per capita basis. Even in Canada, they amount to only US$154 or less per capita per year. This, together with the free rider problem associated with collective action, helps explain why consumers in that region do not counterlobby farmers over farm support programs. It also means the trade and national economic welfare effects of U.S. programs are less than programs delivering the same NRA for farmers in other countries via trade measures that generate a CTE equal to the NRA. The Politics of United States and Canadian Policies Explaining the political forces behind United States and Canadian agricultural policies requires qualitative rather than quantitative analysis, but the contrasts between the political treatment of commodities within and between the two coun- tries suggests hypotheses in several areas: (a) historical legacies of commodity pro- ducer cooperation, (b) the importance of supply-chain participant cohesion and the link of that cohesion with technological change, (c) the role of budgetary pres- sures, and (d) the inherent weakness of opposition to agricultural support. There are economic and cultural factors behind each of these.18 The importance of a historical legacy of producer cooperation was emphasized by Olson (1985). He noted the traditional and continuing strength of the dairy industry and traced that back to the long-standing organization of producers into marketing cooperatives. Olson's underlying point is that lobbying is a voluntary collective activity of precisely the kind highly susceptible to free-rider problems, that this is in fact the chief hurdle to an interest group obtaining subsidies, and that cooperative organizations have already solved this problem sufficiently to permit effective lobbying. This hypothesis fits well with what would otherwise be perhaps the chief puzzle in Canadian as compared to U.S. agricultural support, namely why poultry has had a well entrenched support system via supply control and import protection in Canada but receives virtually nothing in the United States. As Schmitz, Furtan, and Baylis (2002) explain, supply control measures grew out of cooperative activity in those commodities by provincial producer organizations. But there were no such corresponding poultry organizations in the United States. Table 4.5. CTEs of Policies Assisting Farmers, Covered Products, Total and Per Capita and by Product, United States and Canada, 1960­2005 a. Aggregate CTE (percent) 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005 Total, North America 5.7 8.4 6.1 7.2 12.6 9.8 1.9 4.7 2.4 6.3 United States 5.4 8.4 5.9 5.8 11.0 7.5 0.4 6.8 4.2 8.5 Canada 10.1 8.2 7.8 19.1 27.3 31.0 24.6 14.1 15.6 14.1 b. CTE per capita (constant 2000 US$) 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005 Total, North America 27 41 36 44 68 42 8 18 9 26 United States 26 40 34 35 58 32 2 26 16 35 Canada 50 47 54 127 154 134 95 54 54 56 (Table continues on the following pages.) 209 210 Table 4.5. CTEs of Policies Assisting Farmers, Covered Products, Total and Per Capita and by Product, United States and Canada, 1960­2005 (continued ) c. CTE by covered product (constant 2000 US$ millions), United States, 1960­2005 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005 Barley -- -- -- -- -- 89 99 100 95 89 Beef 0 426 547 531 475 776 1,529 1,806 1,813 1,976 Cotton 16 1,489 1,284 482 657 1,028 865 965 1,320 1,955 Eggs -- -- -- 20 9 79 54 308 409 519 Maize 246 435 822 95 566 1,320 2,947 3,257 3,702 4,374 Milk 3,398 3,304 3,056 5,140 10,134 9,039 5,777 1,738 4,264 1,005 Pig meat 0 0 0 0 0 1,142 1,748 2,253 2,362 3,122 Potatoes 0 0 0 0 0 0 0 0 0 0 Poultry 0 0 0 0 0 134 1,076 1,263 1,328 1,657 Rice 0 0 0 25 100 46 138 201 199 357 Sheepmeat -- -- -- 29 5 6 6 13 53 57 Sorghum -- -- -- 0 0 143 189 185 163 161 Soybeans 0 0 0 0 0 199 337 365 376 483 Sugar 1,011 1,683 791 965 1,051 1,421 1,018 926 992 593 Wheat 147 591 572 449 546 88 126 917 859 1,293 Wool -- -- -- -- -- 2 1 1 0 0 United States 4,819 7,929 7,073 7,696 13,525 7,835 378 7,010 4,677 10,422 d. CTE by covered product (constant 2000 US$ millions), Canada, 1961­2005 1961­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005 Barley 11 15 29 68 86 17 16 0 0 0 Beef 26 41 52 54 108 70 54 6 5 0 Eggs -- -- -- 70 127 98 121 89 33 141 Maize 17 23 27 23 23 22 1 0 0 0 Milk 709 682 749 2,542 3,185 2,847 2,137 1,471 1,641 1,663 Peas 0 0 0 0 0 0 0 0 0 -- Pig meat 31 41 54 56 45 7 0 0 0 0 Potatoes 0 0 0 0 0 0 0 0 0 0 Poultry 105 127 245 227 203 271 361 30 31 30 Rapeseeds 0 0 0 0 0 22 1 1 1 0 Soybeans 0 0 0 0 0 8 1 2 5 6 Sugar -- -- -- 44 39 43 31 25 -- -- Wheat 28 25 37 38 55 152 42 3 6 0 Canada 927 954 1,194 3,122 3,871 3,557 2,767 1,622 1,710 1,841 Source: Anderson and Valenzuela (2008) based on author's spreadsheet. Note: -- = not available. 211 212 Distortions to Agricultural Incentives: A Global Perspective Of course the efficacy of lobbying depends not only on getting organized to make one's case, but also on the legislators' listening and acting favorably. It is almost axiomatic that legislators have an interest in listening to their electorate, who hire and fire them. But a legislator cannot act in accord with the requests of all his or her constituents. Why do agricultural commodity interests get heard well in so many cases? Traditionally, U.S. farmers were said to have a political advan- tage because of the structure of Congress, which has two Senators for every state, or one Senator for each 300,000 residents of the Dakotas, Wyoming, and Montana compared to one for each 20 million residents of California. However, recent efforts at reform (such as taking payments away from producers who have more than US$1 million in off-farm income) have been defeated in the House of Repre- sentatives (where each Representative equally has about 600,000 resident con- stituents) as soundly as in the Senate. The general principal of representation is that, whenever possible, a legislator provides constituents what they ask for. Complicating the efficacy of this system in the United States, however, is that constituents often request conflicting poli- cies. A strength of cotton growers is that they present a unified position of the supply chain, including growers, ginners, shippers, and millers, to the agriculture committees of Congress. This has led to policies such as the "step two" payments to cotton textile millers that compensate them for paying prices for cotton that exceed the low world prices that trigger payments to cotton growers. (This was stopped, however, to comply with a WTO ruling.) Similarly, a strength of corn growers is that they have powerful agribusiness allies who support ethanol subsi- dies, while a former weakness of the grain producers was their asking for acreage control measures that restricted raw material supplies for such agribusiness (for- mer because the 1996 Farm Act revoked the authority of the Secretary of Agricul- ture to administer annual acreage reduction programs that until the 1990s were a key feature of grains policy). Cross-commodity dispute is a related problem, aris- ing, for example, on the part of fruit and vegetable growers excluded from decou- pled payment programs. A political strength of sugar import restrictions is that high sugar prices have created a large market for corn-based sweeteners. Monetary contributions help in getting an interest group's case heard by politi- cal leaders. Though farmers are not notably profligate donors, agricultural political action committees (PACs) have been important in the United States. Table 4.6 summarizes PAC donations to politicians or candidates reported during the 2006 election cycle (November 2004­October 2006). These data are from the legally required reporting of PACs for donations of more than US$100,000 and do not capture all political spending. Though the US$6.7 million they donated may seem a large sum, there were 1,200 nonagricultural PACs that spent more than US$100,000 in the United States during this period, and their aggregate spending United States and Canada 213 Table 4.6. PAC Disbursements during the Election Cycle, United States, November 2004­October 2006 (US$ millions) Sugar Alliance 1.2 Farm Credit Council 1.0 Dairy 0.8 Texas Farm Bureau 0.5 Cotton Council 0.3 Sugar 0.3 Cattlemens Beef 0.3 Dairy 0.3 Broiler Chickens 0.3 FLA sugar 0.2 Indiana Farm Bureau 0.2 Sugar beets 0.2 Farmers' Group 0.2 Peanuts 0.2 Eggs 0.2 Michigan Farm Bureau 0.2 Farmer Coop 0.1 Sugar beets 0.1 Beef 0.1 Total 6.7 Source: Political Moneyline, http://www.tray.com. was US$800 million. The largest agricultural PAC, the Sugar Alliance, ranked 154th of the 1,200. Agriculture's share of the spending, US$6.7 million or just under 1 percent of the US$800 millon total, is about the same as agriculture's share of national GDP. Overall spending by PACs in the broad Standard Industrial Classifi- cation code for agriculture was US$15 million, but the majority of this was from agribusiness firms--notably Deere, Deans Foods, the International Dairy Foods Association, Cargill, Tysons, Archer-Daniel-Midlands, Conagra, and Heinz Foods. These companies have interests that are often the same as the commodity produc- ers' interests, but not always, and what the agribusiness firms focus on in their lob- bying is more typically their own specific issues of concern. In short, it does not appear likely that money is the source of farmers' exceptional political influence. Technical or other exogenous economic changes create diverse interests within a commodity group that can make policy making aimed at benefiting the group as a whole unattractive to legislators. This happened most notably in the case of a for- merly important commodity program that disappeared completely, the U.S. potato 214 Distortions to Agricultural Incentives: A Global Perspective price support program. During 1945­50, the potato program became highly con- tentious and eventually died because the Western potato growers were able to pro- duce profitably at prices lower than the prevailing support levels, and they saw their market potentially capped by restraints needed to control the costs of the support system (which at the time relied heavily on production control for the main sup- ported commodities). Similarly, the long-standing marketing order pricing sys- tems for California oranges and lemons was ended in the 1990s, mainly because of intraindustry disagreement about its operation, while differing interests of peanut and tobacco farmers with and without quotas to produce for the high-priced domestic market contributed to demise of these quota program earlier this decade. One might expect consumer interests to oppose farm legislation that would increase food prices, but this has only rarely surfaced as a significant political force (the notable case being U.S. grain export embargoes of the 1970s). This lack of opposition seems to be associated with longstanding positive feelings that the general public has about farmers and farming. A broad-based source of resistance to agricultural support that has been effective is budgetary pressure at times when fiscal discipline is perceived to be a high priority. In both the 1985 and 1990 farm legislation, reductions in payments were enacted so that farm bills could meet overall Congressional budget limits. The 1996 farm legislation was also designed to limit subsidy expenditures, but that budget discipline quickly broke down. As noted earlier, the Farm Security and Rural Investment Act of 2002 covering crops planted in 2002­07 increased spending well above baseline levels from con- tinuation of the legisltation it replaced. This Act was popular in Congress, having passed in the House of Representatives by a vote of 280 to 141 and in the Senate by 64 to 35. The Bush Administration did not raise serious objections, and the bill was signed into law with praise from farm-group representatives.19 However, small-farm and environmental advocacy groups were unhappy that amendments that would have imposed more stringent payment limits on large farms, redi- rected some commodity program payments to conservation and environmental programs, and imposed various regulatory restraints on agribusiness failed. Out- side the community of U.S. agricultural interests, the 2002 Act has been widely reviled, as mentioned earlier.20 Just after the Act was passed, three western provinces of Canada, along with a dozen Canadian farm groups, asked for Can$1.3 billion to offset the effects of the new U.S. payments. Since that time, Canada has raised its payments to producers too, as herein described. Prospects for Reform Both the United States and Canada have enacted agricultural support programs that have distorted their domestic commodity markets. Considering how similar the countries are in many respects, there are some notable differences between United States and Canada 215 the policy approaches in the two countries in the choice of policy instruments. But the similarities are more fundamental, and several of them bear on the pros- pects for reform: (a) strong political resilience of support for farmers, (b) weak political expression of consumers' interests, (c) trends toward less reliance on measures that seek to control prices in particular markets (via stockpiling, sup- ply management, and import tariffs), (d) trends toward more reliance on whole-farm income support (especially in Canada) and payments under com- modity programs that are delinked from current production (in the United States), (e) acceptance of multilateral liberalization of agricultural trade, but only with maintenance of protection for some producer interests, (f) modest increases in political influence of environmental protection, and (g) episodic political strength of taxpayer interests in cutting farm-support spending, under circumstances that give general budget reduction. As a result of these forces and trends, real spending on agricultural support has not diminished over time even as the share of the agricultural sector in the overall economy has diminished greatly, while at the same time both the United States and Canada have moved since the early 1990s in the direction of reduced directly market-distorting poli- cies. In the face of historically low commodity prices 1998­2002, the United States maintained policies that forestalled output reductions that these low prices would otherwise have induced. What options does the preceding summary suggest for reform? First, the best prospects remain international negotiations, most notably the Doha Round of the WTO, despite the failure of the negotiations to make much progress thus far. Nonetheless, it remains the case that the United States and Canada would sign on to such an agreement if it involved provisions that reform proponents could point to as offsetting gains in agricultural export markets. Second, there remain prospects that a combination of environmental and taxpayer interests could shift agricultural support spending toward public-good provision in ways that would be less market-distorting than current policies.21 Reforms in this direction were formulated and promoted by several organized coalitions of interests in preparation for the 2008 U.S. Farm Bill, mainly by environmentalists, farmland preservation- ists, and internationally oriented agribusiness groups. Third, while it is not in the cards now, it is possible that there will a resurrection of a general predisposition to economic liberalism, as occurred in Australia and New Zealand (Anderson et al. 2009). Not so long ago, this predisposition in North America served to limit the scope of market-distorting legislation considerably, and could again, as the recent legal debate in Canada on the CWB's authorities illustrates, despite its outcome. The ascent of free-market Republicans in the 1980s suggested this might be a real possibility for the United States. But the policy salience of this strain of opinion has been thwarted by that the party's embrace of muscular nationalism, military internationalism, and cultural conservatism. Similarly, once-prominent free-trade 216 Distortions to Agricultural Incentives: A Global Perspective advocacy by Democrats has been swamped by the wave of industrial protection- ism allied with antitrade prairie populism the party has undertaken in recent political campaigns. The prospects for each of the preceding openings to reform--a liberalizing WTO agreement, conservation and environmental pressures, and a resurgence of deregulatory policy dispositions--would be enhanced with the traditional econo- mist's recipe of compensating the losers from policy changes with nondistorting transfers. Though the direct payments under the U.S. 1996 and 2002 Farm Acts are a move in this direction, what is really required are one-time payments, or buyouts. U.S. policy has carried out three notable experiments in buyouts since 1980: in dairy, peanuts, and tobacco. The dairy buyout (Dairy Production Termi- nation Program) was implemented in 1986­87, at a time when milk surpluses were chronic, and was introduced following the calculation that having dairy farmers agree to sell their herds and leave dairying would reduce budgetary out- lays. Farmers made offers to the government stating the price per hundred pounds of milk producing capacity of their herds at which they would be willing to leave the business. Producers who participated had to sell all their female cattle and agree to remain out of dairy farming for five years, and attempts were made to ensure that cattle sold were slaughtered or exported and not sold to another dairy farm. The program is estimated to have reduced U.S. milk production by about 7 percent in the short term but to have had no long-term effect (Dixon, Susanto, and Berry 1991). The positive lesson of the program, however, is that buyouts can be successfully implemented by USDA as a means of getting farmers to participate in policy reform. Buyouts to permanently remove producers from certain commodity support programs have been undertaken for peanuts and tobacco since 2000. The 2002 Farm Act authorized payments to peanut producers to buy their production quo- tas, which had well-defined values under preexisting programs. Premiums over market values of quota had to be paid, but again the approach was proven feasible. Similarly, tobacco quotas and the entire tobacco price support program were ter- minated in 2005. However, whether buyouts would be feasible for commodities without production-limiting quotas or where program-created assets are not so precisely defined is questionable. An approximation to a buyout was, in the minds of some proponents of production flexibility, the contract payments in the 1996 Farm Act. But this legislation did not formally end the preexisting support struc- ture and, in fact, did not end the support programs. Rather, the approach enabled additional payments to farmers. While students of the prospects for this approach remain cautious (see Orden 2006 and Orden, Blandford, and Josling 2009), some- thing like it coupled with changes of approach as outlined above are the most likely feasible avenue to future reform. United States and Canada 217 Notes 1. This chapter's coverage of the United States draws to some extent on Gardner (2002). 2. There is one category of farms in the USDA classification, called "limited resource farms," that has low incomes both from farming and from off-farm sources. There were an estimated 199,000 such farms in 2004, or 10 percent of all farms; their average household income was US$7,700. Their average farm earnings was a loss of US$5,900. Apart from these farms, the other 90 percent of U.S. farm house- holds are doing well economically, either from farm or off-farm income sources. The data underlying this and subsequent information about farm household economics are developed by the Economic Research Service of the USDA, using the organization's annual Agricultural Resource Management Survey of about 10,000 farms. 3. For 2001­05, the annual net income per farm that yields the Can$11,000 average are, respec- tively: Can$11,000, Can$6,000, Can$11,100, Can$16,400, and Can$10,600 (Statistics Canada 2006). 4. The Court changed its view in decisions of Mulford v. Smith (1939) and Wickard v. Filburn (1942), which upheld, respectively, tobacco and wheat marketing controls enacted in legislation of 1938. The reasoning was that agricultural production affected by such programs influenced national markets and so could be legislated under constitutional powers to regulate interstate commerce. 5. An earlier wheat agreement, in 1933, created a schedule of quotas limiting the shipments of wheat by exporting countries, but this proto-OPEC broke down and the agreement was allowed to expire in 1935 (Gale and Zaglits 1949). Explicit export subsidies on Northwestern U.S. wheat were also implemented for a time early on in the New Deal (Nourse, Davis, and Black 1935). 6. Irwin's estimate of a net subsidy rate of 15 pecent is less than the tariff protection rate of 30 per- cent primarily because of the effect of the higher prices of import-competing goods on increasing the cost of production of nontraded goods. 7. It was such political successes of agriculture in tariff legislation that led H. L. Mencken to pillory farmers in one of his famous diatribes: "Has anyone heard of a farmer practising or advocating any political idea that was not absolutely self-seeking--that was not, in fact, deliberately designed to loot the rest of us to his gain? . . . There has never been a time, in good seasons or bad, when his hand were not itching for more . . . One might almost argue that the chief, and perhaps even only aim of legisla- tion in These States is to succor and secure the farmer" (Mencken 1958, pp. 158­60). 8. For estimates of effects of U.S. wheat export subsidies, see Gardner (1996). On the effects of grain export embargoes, see USDA (1986). 9. The U.S. government's fiscal year runs from October 1 of the previous calendar year to September 30 of the current calendar year. Thus, fiscal year 2005 represents the time period October 1, 2004 to September 30, 2005. 10. The spike in spending relative to payments in 2000 is an artifact of government budgetary spending being reported on a fiscal year basis while payments to farmers are reported on a calendar year basis. In 2000, an election year, Congress rushed to get payments to farmers that would normally have gone out after October 1 before October 1, resulting in payments that would normally have occurred in two fiscal years occurring in one calendar year. 11. A Washington Post editorial termed the bill "The Mother of All Pork." Business Week magazine opined: "It's a dreadful piece of legislation-bad for most farmers, bad for consumers, and horrendous for taxpayers" (May 7, 2002). The New York Times editorialized against the Act on several occasions, notably in a piece entitled "The Hypocrisy of Farm Subsidies" (December 1, 2002). 12. Ten-year projected spending was estimated in accordance with Congressional budgetary proce- dures, even though the Act only authorized programs for the six years 2002­07. The "baseline budget scoring" assumption is that those programs will be reauthorized to cover the 10-year period. 13. This argument is more difficult to make for the 1980s, when these payments were tied to signif- icant annual land-idling requirements. It also ignores wealth effects, insurance effects, and remaining production restrictions, as discussed in the appendix to Gardner (2008). Whether any of these effects are quantitatively important is an empirical question. While impossible to estimate with precision from the data available, the analysis to date suggests these effects are small. 218 Distortions to Agricultural Incentives: A Global Perspective 14. "Price tag" is a vague term and is used because the figures are not all derived from a consistent set of U.S. budgetary concepts. Most notably, the export credit guarantees are not the expenditures of the government on these guarantees; rather, they are the value of loans guaranteed. Unless there are defaults on these loans (funds borrowed by foreign importers to buy U.S. exports are not repaid to the U.S. lenders), the actual outlays on these programs is negligible. In fact, defaults are rare. In U.S. budg- etary parlance, the value of loans guaranteed is the "program level," as shown in the USDA budget summary. 15. In the original programs of the 1930s, there were U.S. supply control efforts in livestock. The only significant such program since 1955, however, is the Dairy Herd Buyout Program of the mid-1980s. 16. Neither the categories "NPS" nor "decoupled" presented here correspond directly to use of these terms in WTO domestic support notifications or as has occurred in dispute settlement argu- ments. To illustrate, the NPS for 1986-2007 herein includes diesel fuel tax exemptions not reported by the United States to the WTO under the URAA category of NPS, while it excludes crop and revenue insurance subsidies that are reported as WTO NPS support. The "decoupled" category herein includes not only the "decoupled income support" as defined in URAA annex 2 as WTO green box (which is how the United States reports its fixed direct payments); countercyclical payments that the United States reports as NPS are also included in this category. Subtleties in distinguishing between these cat- egories of payments are discussed in the appendix to Gardner (2008). Both classifications are subject to WTO dispute settlement litegation. The decoupled support herein also includes conservation pay- ments (both for long-term land-idling and working lands) that are reported to the WTO by the United States as green box and the crop and revenue insurance subsidies reported to the WTO as NPS sup- port. The use of these terms herein also differs from common classifications in U.S. farm bill discus- sions and budgets that, for example, often clearly separate commodity support from conservation expenditures. 17. See Orden, Blandford, and Josling (2009) for discussion of the 2008 U.S. Farm Act. 18. For more on the political economy of U.S. farm policy, see Gardner (2002) and Orden, Paarlberg, and Roe (1999). 19. As a reminder that U.S. presidents do not always accept what Congress delivers in support of agriculture, President Reagan in 1985 vetoed a farm bill on budgetary grounds, in the midst of the farm crisis of the 1980s. Congress could not muster the two-thirds majorities needed to override the veto, so the president's action was decisive. 20. It is also notable that the more market-oriented members of Congress, even if they represent agricultural constituencies, opposed the bill. Among the opponents were not only the House Republi- can leadership, but also members of the Agriculture Committee, such as Boehmer (Republican-Ohio) and Dooley (Democrat-California), who jointly wrote a Washington Post opinion piece entitled "This Terrible Farm Bill" (May 2, 2002). Similarly strong opposition was voiced in the Senate by Richard Lugar (Republican-Indiana), the senior Republican on the Senate's Agriculture Committee. 21. Even if these policies were more market-distorting, as they could possibly become, they would more likely be of the kind that would reduce agricultural output rather than increase it. References Agriculture and Agri-Food Canada. 2005. "Agricultural Policy Framework: Federal-Provincial-Territorial Programs." Program Planning, Integration, and Management Directorate, Ottawa. http://www .agr.gc.ca/progser/index_e.phtml. ______. 2006. "The Canadian Agricultural Income Stabilization (CAIS) Program." Program Planning, Integration, and Management Directorate, Ottawa. http://www.agr.gc.ca/caisprogram/factsheets/ faq_margins.html. Anderson, K., R. Lattimore, P. J. Lloyd, and D. MacLaren. 2009. "Australia and New Zealand." Chapter 5 in this volume. United States and Canada 219 Anderson, K., and E. Valenzuela. 2008. "Estimates of Global Distortions to Agricultural Incentives, 1955 to 2007." Core database at http://www.worldbank.org/agdistortions. Benedict, M. R. 1953. Farm Policies of the United States, 1790­1950. New York: Twentieth Century Fund. Carter, S. B., ed. 2006. Historical Statistics of the United States: Colonial Times to 1970, New York: Cambridge Uinversity Press. Davis, L., J. R. T. Hughes, and D. M. McDougall. 1965. American Economic History. Homewood, IL: Richard D. Irwin Inc. Dixon, B., D. Susanto, and C. R. Berry. 1991. "Supply Impact of the Milk Diversion and Dairy Termination Programs." American Journal of Agricultural Economics 73 (3): 633­40. FAPRI (Food and Agricultural Policy Research Institute). 2002. "Farm Security and Rural Investment Act of 2002: Preliminary FAPRI Analysis." Unpublished paper, FAPRI. Fowke, V. C. 1946. Canadian Agricultural Policy: The Historical Pattern. Toronto: University of Toronto Press. Furtan, W. H. 2006. "Canadian Agriculture and Food Policy." Saskatchewan Agrivision Corp., Saskatoon. Gale, E. G., and O. Zaglits. 1949. "Intergovernmental Agreements Approach to the Problem of Agricul- tural Surpluses." In Readings on Agricultural Policy, ed. O. Jesness, 306­27. Philadelphia: Blakiston. Gardner, B. L. 1996. "The Political Economy of U.S. Export Subsidies for Wheat." In The Political Economy of American Trade Policy, ed. A. Krueger, 291­331. Chicago: University of Chicago Press. ______. 2002. American Agriculture in the Twentieth Century: How It Flourished and What It Cost. Cambridge, MA: Harvard University Press. ______. 2008. "Distortions to Agricultural Incentives in the United States and Canada." Agricultural Distortions Working Paper 62, World Bank, Washington, DC. Gray, R., and V. Smith. 1997. "Harmonization and Convergence of Canadian and U.S. Grains and Oilseeds Policies." Policy Issues Paper 4, Montana State Univerity, Bozeman, Montana. Huff, H. B. n.d. "Changing Role of Public Policy in Canadian Agriculture." Unpublished paper. Irwin, D. A. 2006. "Historical Aspects of U.S. Trade Policy." NBER Reporter http://www.nber.org/ reporter/summer06/irwin.html. Johnson, D. G. 1950. Trade and Agriculture. New York: McGraw Hill. Jones, R. L. 1941. "The Canadian Agricultural Tariff of 1843." Canadian Journal of Economics and Political Science 7: 528­37. Josling, T. 2009. "Western Europe." Chapter 3 in this volume. Mencken, H. L. 1958. Prejudices: A Selection. New York: Vintage Books. Nourse, E., J. Davis, and J. Black. 1935. Three Years of the Agricultural Adjustment Administration. Wash- ington, DC: Brookings Institution. OECD (Organisation for Economic Co-operation and Development). 1978. Recent Developments in Canadian Agricultural Policy. Paris: OECD. OECD PSE-CSE Database (Producer and Consumer Support Estimates, OECD Database 1986­2007). Organisation for Economic Co-operation and Development. http://www.oecd.org/document/ 59/0,3343,en_2649_33727_39551355_1_1_1_1,00.html. Olson, M. 1985. "Space, Agriculture, and Organization." American Journal of Agricultural Economics 72 (3): 928­37. Orden, D. 2006. "Feasibility of Farm Program Buyouts." Presented at North American Agrifood Mar- ket Integration Workshop, Alberta, Canada, May 31­June 2. Orden, D., D. Blandford, and T. Josling. 2009. "Determinants of Farm Policies in the United States, 1996­2008." Agricultural Distortions Working Paper 89, World Bank, Washington, DC. http:// www.worldbank.org/agdistortions. Forthcoming in Political Economy of Distortions to Agricultural Incentives, ed. K. Anderson. Orden, D., R. Paarlberg, and T. Roe. 1999. Policy Reform in American Agriculture: Analysis and Progno- sis. Chicago: University of Chicago Press. 220 Distortions to Agricultural Incentives: A Global Perspective Schmitz, A., and H. Furtan. 2000. The Canadian Wheat Board. Regina, Saskatchewan: Canadian Plains Research Center. Schmitz, A., H. Furtan, and K. Baylis. 2002. Agricultural Policy, Agribusiness, and Rent-Seeking Behavior. Toronto: University of Toronto Press. Statistics Canada. 2006. CANSIM table. http://www40.statcan/101/cst01/prim10.htm. Taussig, F. W. 1931. The Tariff History of the United States. New York: G. P. Putnam's Sons. U.S. Department of Agriculture. 1986. "Embargoes, Surplus Disposal, and U.S. Agriculture." Economic Research Service, Agricultural Econonomics Report 564, U.S. Department of Agriculture, Washington, DC. ______. 2006. "Budget Summary and Annual Performance Plan." Office of Budget and Program Analysis, U.S. Department of Agriculture, Washington, DC. http://www.obpa.usda.gov/budsum/ 2007/fy07budsum.pdf. 5 Australia and New Zealand Kym Anderson, Ralph Lattimore, Peter J. Lloyd, and Donald MacLaren* Following its gold rush in the 1850s, Australia surpassed the United Kingdom in having the highest per capita income in the world, and New Zealand was not far behind. Both countries suffered from depressions in the 1890s before recovering their equal-first ranking just prior to World War I, though for the following seven decades their incomes lagged those of the United States and some other developed economies (figure 5.1).1 The middle of the 20th century brought a brief recovery in ranking thanks in part to the Korean War­induced boom in wool prices, when Australia and New Zealand were ranked equal second after the United States. But since then their rankings have continued to slide as the economies of Western Europe, Canada, and Japan grew faster than the United States, while Australia-- and especially New Zealand--grew slower (Anderson, Lloyd, and MacLaren 2007). By 2004, these two antipodean economies were ranked 25th and 37th respectively, according to the World Bank Atlas method of measuring gross national product (GNP) per capita (World Bank 2006). By that standard, at least, the long-term economic performance of both the Australian and New Zealand economies over the 100 years to the 1970s has to be described as relatively poor. During the past three decades, by contrast, these two small economies (with just 20 million and 4 million people, respectively) outperformed most other *The authors are grateful for research assistance from Johanna Croser, Ceren Erdogan, Esteban Jara, Marianne Kurzweil, Signe Nelgen, Francesca de Nicola, Damiano Sandri, and Ernesto Valenzuela, and for helpful comments from workshop participants and Rod Forbes, Robin Johnson, Jonathan Pincus, and Grant Scobie. The working paper version of this chapter (Anderson, Lloyd, and MacLaren 2007; Anderson et al. 2008) contains additional background material and appendix tables. 221 222 Distortions to Agricultural Incentives: A Global Perspective Figure 5.1. Real GDP Per Capita in Australia, New Zealand, and Other High-Income Countries Relative to the United States, 1870­2004 1.6 1.4 1 1.2 United States 1.0 0.8 0.6 0.4 18 5­ 4 18 0­ 9 18 5­ 4 18 0­ 9 19 5­ 4 19 0­ 9 19 5­ 4 19 0­ 9 19 5­ 4 19 0­ 9 19 5­ 4 19 0­ 9 19 5­ 4 19 0­ 9 19 5­ 4 19 0­ 9 19 5­ 4 19 0­ 9 19 5­ 4 19 0­ 9 19 5­ 4 19 0­ 9 19 5­ 4 19 0­ 9 20 5­ 4 00 99 4 7 7 8 7 8 8 9 8 9 9 0 9 0 0 1 0 1 1 2 1 2 2 3 2 3 3 4 3 4 4 5 4 5 5 6 5 6 6 7 6 7 7 8 7 8 8 9 8 9 9 ­0 18 0­ 7 18 Australia New Zealand other Western Europe Canada Nordic countries Source: Based on 1990 international Geary-Khamis dollars from Maddison (2003), shown relative to the United States, which is set as the numeraire at 1.00. "Nordics" include Denmark, Finland, Norway and Sweden; "other Western Europe" includes all countries with data from 1870, namely Belgium, France, Germany, Italy, Netherlands, Portugal, Spain, Switzerland, and the United Kingdom. high-income countries, with their per capita incomes growing half as fast again as the Organisation for Economic Co-operation and Development (OECD) average. The marked difference between recent and earlier relative performance in Australia and New Zealand is due to major economic policy reforms, in particular the belated opening of their economies, first to each other and then to the rest of the world. Having been more protectionist toward manufacturing than other OECD countries for most of the 20th century (Anderson and Garnaut 1987) and having not participated in the industrial trade policy reforms agreed to by other General Agreement on Tariffs and Trade (GATT) contracting parties in the first seven rounds of multilateral trade negotiations (1947 to 1979), Australia and New Zealand have since undergone a remarkable degree of opening up of their current and capital accounts. This liberalization reversed the downward trend in their trade as a share of gross domestic product (GDP), particularly for other OECD countries (including the United States, in which the share trebled over the past three decades) while accel- erating the downward trend in their average import tariff (Anderson et al. 2008). The fact that increased openness in Australia and New Zealand was accompanied Australia and New Zealand 223 by many domestic microeconomic and macroeconomic reforms, and that it coin- cided with a long period of rapid global economic growth that was stimulated by the information and communication technology revolution and by the opening up of other world economies (especially nearby resource-poor countries in East Asia) added to the scope for boosting gains from freeing their international trade and investment and floating their currencies. Another difference between the economies of Australia and New Zealand and most other OECD countries is that they are relatively well endowed with agricul- tural land per worker. Coupled with the development of valuable production and marketing technology, this endowment provides Australia and New Zealand with a significant comparative advantage in agricultural products. The advantage is especially strong for New Zealand, which, unlike Australia, does not contain an abundance of mineral and energy resources. Trade protectionism in these economies thus meant restrictions on imports of manufactured goods, making their trade policy regimes similar to those of developing countries. True, in some decades they also had periods of agricultural subsidies, but overall the trade policy regime in both countries has long involved an antiagricultural bias. The fact that those agricultural subsidies have been virtually eliminated over the past two decades also makes Australia and New Zealand an interesting political economy study, given the extreme difficulties other OECD countries have experienced in reforming their farm support programs (see Gardner 2009; Josling 2009; and Honma and Hayami 2009). This chapter examines the extent to which the antiagricultural bias in Australia and New Zealand has changed since World War II, identifying the forces that caused the policy evolution in each case. It begins by summarizing the structural changes that have accompanied economic growth in both countries since the 1950s. It then describes the emergence of manufacturing protection and some agricultural subsidies, and, after the 1970s, the dramatic dismantling of those inter- ventions. The analysis is conducted by compiling a new time series of nominal rates of protection for both agriculture and manufacturing, stretching back to the mid-1940s for Australia and to the mid-1950s for New Zealand. The reasons for these policy choices are also explored--first, the gradual growth in market inter- ventions to the 1970s and then their relatively rapid dismantling thereafter. The chapter concludes by discussing prospects for further policy reform and lessons for both other high-income countries and resource-rich developing economies. Growth and Structural Changes since 1950 The growth performance of Australia and New Zealand for most of the 20th cen- tury is in stark contrast with that since the late 1980s, when the two countries out- performed many other advanced economies in terms of GDP per capita growth 224 Distortions to Agricultural Incentives: A Global Perspective (World Bank 2006). The past two decades have also been a period of especially rapid total factor productivity (TFP) growth in Australia and New Zealand (Parham et al. 1999; Dowrick 2001; Statistics New Zealand 2006), mostly due to growth in employment and hours worked per worker (Card and Freeman 2004), in contrast to the United Kingdom, for example. Australia's annual TFP growth rate accelerated a full percentage point during the 1990s (Parham 2004), while New Zealand's rose four-fifths of a percentage point. Since that was not the expe- rience of other OECD countries, Parham asserts that domestic factors must explain a major part of the increase, an important one being the comparatively greater openness of the Australian economy to trade and investment.2 A more recent econometric study by Diewert and Lawrence (2006) demon- strates that productivity growth has been the dominant contributor to the growth in real welfare in Australia since 1960, with the terms of trade playing only a very minor role. Certainly prices in international markets for primary products relative to manufactures have been on a downward trend over the past century, but the decline has averaged less than 0.5 percent per year. But the difference between recent and earlier performance in Australia and New Zealand is also due very sub- stantially to economic policy reforms of the past three decades. The belated free- ing of markets in the two economies not only has arrested the decline in per capita income ranking in the two countries, but is having a remarkable influence on their patterns of production and trade. For these natural-resource-rich, relatively lightly populated economies,3 the most appropriate theory of comparative advantage is a blend of two core models developed in the 20th century: the Heckscher-Ohlin-Samuelson model, which assumes all factors of production are mobile between sectors, and the Ricardo- Viner model, which assumes some factors are sector-specific. Such a blend is pro- vided by Krueger (1977) and explored further by Deardorff (1984). They consider two tradable sectors each using intersectorally mobile labor plus one sector-specific factor (natural-resource capital or industrial capital). Assuming that labor exhibits diminishing marginal product in each sector, and that there are no serv- ices or nontradables and no policy distortions, then at a given set of international prices the real wage is determined by the aggregate per worker endowment of natural-resource and industrial capital. The commodity composition of a coun- try's trade--that is, the extent to which a country is a net exporter of primary or industrial products--is determined by its endowment of natural relative to indus- trial capital compared with that ratio for the rest of the world. Leamer (1987) develops this model further and relates it to paths of economic development. If the stock of natural resources is unchanged, rapid growth by one or more economies relative to others in the availability of industrial capital per worker would cause those economies to strengthen their comparative advantage Australia and New Zealand 225 in nonprimary products. However, a discovery of minerals or energy raw materials would strengthen that country's comparative advantage in mining and weaken its comparative advantage in farm and other goods, ceteris paribus. It would also boost national income and, hence, the demand for nontradables, which would cause mobile resources to move into the production of nontradables, further reducing farm and industrial production (Corden 1984). Domestic or foreign savings can be invested to enhance the stock and to improve the quality not only of industrial capital but also of labor or natural resources, and to provide capital to the nontradables sector. Any such increase in the net stock of produced capital per worker will put upward pressure on real wages. That will encourage, in all sectors, the use of more labor-saving techniques and the develop- ment and importation of new technologies that are less labor intensive. Which types of capital expand fastest in a free-market setting depends on their expected rates of return. The more densely populated, natural-resource-poor a country is, the greater the likelihood that the highest payoff would be in expand- ing capital stocks for nonprimary sectors. At early stages of development of such a country, when it has a relatively small stock of natural resources per worker, wages would be low and the country would have a comparative cost advantage in unskilled labor-intensive, standard-technology manufactures. As the stock of industrial capital grows, there would be a gradual move toward exporting more capital- and skill-intensive manufactures. Natural-resource-abundant economies such as Australia and New Zealand, however, would develop a comparative advan- tage in manufacturing at a late stage of development, and their industrial exports would be relatively capital intensive. In New Zealand, with lesser mineral and energy resources per worker and poorer climatic conditions for broadacre crop- ping than in Australia, agricultural comparative advantages would be stronger in aggregate but less focused on cereal and oilseed cropping than in Australia. The above theory of changing comparative advantages has been used successfully to explain the evolving pattern of exports of Australia and its Asian trading partners (Anderson and Garnaut 1980, 1987; Anderson and Smith 1981; Anderson 1998), and is also consistent with New Zealand's trade pattern. It can be used also to explain shocks to that evolutionary pattern, as with mining booms; and it is consistent with the larger shares of farm revenue from livestock and horticultural crops compared with grains, oilseeds, sugar, and cotton for New Zealand relative to Australia. But the evolving pattern of a country's production and trade specialization also depends on policy choices and their changes over time. In the case of Australia and New Zealand, their long history of industrial protectionism (reflected in the relatively high implicit tariffs) resulted in a smaller share of GDP traded than would be expected for economies of their size.4 Protectionism also ensured a big- ger manufacturing sector than would have emerged under free trade, which was 226 Distortions to Agricultural Incentives: A Global Perspective possible in their full-employment setting only at the expense of other sectors. In both countries, the proportion of GDP accounted for by manufacturing in the early 1960s were close to the OECD average of around 30 percent, even though Australia and New Zealand have always been lightly populated and so have weak comparative advantage in manufactures. The removal of the ban on key mineral raw material exports in the early 1960s and the tariff reforms of the 1970s and 1980s, however, corrected that distortion for Australia. Between 1960 and 2005, manufacturing's share of GDP fell much more rapidly for Australia than for the average OECD country, to just 11 percent, while the mining sector's share trebled (Anderson, Lloyd, and MacLaren 2007). The share of Australian exports accounted for by mining, meanwhile, more than trebled between the early 1960s and early 1980s, helped by the dramatic rises in energy raw material prices during the 1970s.5 Even though that trend lowered agriculture's relative contribution, the share of exports in the gross value of farm production increased considerably, from around 55 percent in the early 1960s to 75 percent after the mid-1970s (Anderson et al. 2008). Moreover, the growth in agricultural exports came from an increasing range of farm products as farmers diversified away from the traditional wheat and sheep enterprises to beef, cotton, sugar, dairy products, wine, and rapeseed. Natural-resource-based products were not the only exportables discouraged by protectionist measures in Australia and New Zealand, however. Export industries within the manufacturing sector, as well as services exports, were also affected. Together, those two sectors represented one-twelfth of Australia's exports in the early 1950s. By 1980, their contribution was barely above one-quarter, but by 1990 it had risen to one-third and by 2005 to 44 percent, or 22 percent for each sector, the first time either sector surpassed the 21 percent share of exports represented by agriculture. The transformation in New Zealand was similar, despite the absence of the mineral resources with which to create a mining boom. Furthermore, New Zealand was impacted much more than Australia by the coming into force from 1983 of the Australia New Zealand Closer Economic Relations Trade Agreement (ANZCERTA): its lower wages allowed it to rapidly expand exports of manufac- tures and services to the much bigger Australian economy under that preferential arrangement. Together with the virtual elimination of manufacturing protection, these forces brought to a halt the decline in agriculture's share of New Zealand's GDP, in fact raising it from a low of 7 percent in the late 1980s to 9 percent in the early 2000s--notwithstanding the abolition of nontrivial agricultural subsidies in the 1980s. Over that same period the share of food and agricultural products in New Zealand's exports have fallen somewhat, to the benefit of other manufactures and services. With the decade-long buildup and then removal in the mid-1980s of subsidies to agricultural exporting industries (see next section), agricultural and Australia and New Zealand 227 processed food as a percentage of the gross value of primary agricultural produc- tion expanded and contracted somewhat substantially, though it has since been stabilized despite the growth in exports of nonagricultural goods and services during the past two decades. Policy Evolution This section begins with a brief history of policies up to the early 1970s and then moves to a discussion of the changes in the next dozen years before reforms accel- erated in the mid-1980s. Since there were relatively few agricultural subsidies or farm import barriers (other than quarantine restrictions) through most of the past century, the story in both Australia and New Zealand is more about the indirect antiagricultural bias that resulted from the protection of manufacturing. Nonethe- less, policies that directly distorted various agricultural markets post­World War II are also examined.6 We know that it is relative prices, and hence relative rates of government assistance, that affect incentives. In a two-sector model an import tax has the same effect on the export sector as an export tax (the Lerner symmetry theorem); this carries over to a model that includes a third sector producing only nontradables (Vousden 1990). For that reason the average nominal rate of assistance (NRA) for the tradable parts of the agricultural sector is reported, based on NRA estimates for individual agricultural industries, along with the average NRA for the tradable parts of all nonagricultural sectors. The NRA is the equivalent of the percentage by which government policies have raised the producer price above what it would be without the government's intervention.7 With those two sectoral NRAs a relative rate of assistance (RRA) can be calculated. The RRA is defined as 100[(1 NRAagt 100) (1 NRAnonagt 100) 1], where NRAagt and NRAnonagt are the average percentage NRAs for the tradables parts of the agricultural and nona- gricultural sectors, respectively. Since the NRA must be greater than 100 percent if producers are to earn anything, so too must the RRA. The usefulness of this measure is that if it is below zero, it indicates the extent to which the policy regime has an anti- agricultural bias. The converse applies when RRA is positive. The cost of government policy distortions to incentives, in terms of resource misallocation, increases with the degree of substitution in production. In the case of agriculture, which involves the use of farmland that is sector-specific, the greater the variance of NRAs across industries within the sector, the higher the welfare cost of market interventions. A simple, if crude, index of that cost is the standard deviation of industry NRAs within agriculture. The weighted mean NRA for the sector (using the values of production at unassisted prices as weights), along with the standard deviation around that mean each year, is thus reported here. 228 Distortions to Agricultural Incentives: A Global Perspective Prior to the early 1970s The long history of industrial protectionism in Australia and New Zealand has its roots in the formation in 1901 of the Australian Federation, which New Zealand decided not to join, instead becoming separately independent of the United Kingdom. At that time, tariff revenue accounted for almost one-fifth of govern- ment revenue in both countries, a very high share for what at the time were the world's highest-income economies (the share of tariff revenue typically declines as per capita income rises) and twice that of the Nordic countries, for example, even though per capita incomes in those countries was barely half that in Australia and New Zealand (Anderson et al. 2008). Tariffs on manufactures rose steadily in the decades that followed (Lloyd 2008). They were supplemented by quantitative import restrictions first imposed in the late 1930s (as "wartime measures" in Australia and as industry protection in New Zealand), and then rose even further in the 1960s when they substituted for the import licences as the latter were removed (in 1960, with minor exceptions, in the case of Australia).8 This trend over the 1950s and 1960s contrasted strongly with what other high-income coun- tries were doing at that time, which was lowering tariffs on manufactures as part of multilateral trade negotiation under the GATT.9 By the early 1970s, the average manufacturing tariff in Australia and New Zealand exceeded that of any other OECD country (Anderson and Garnaut 1987). Meanwhile, the governments of Australia and New Zealand intervened in numerous markets for farm products, but the subsidies and protection they provided to agricultural industries was only a modest offset to the indirect disin- centives caused by manufacturing protection during this era. In the immediate post­World War II period, Australia's agricultural programs were directly taxing the farm sector. Most of that was removed by the end of the Korean War, however, at which time farmers enjoyed a boom in export prices that spurred the highest inflation in Australia since its gold-rush era of the 1850s. Farm assistance then rose gradually, such that by the end of the 1960s the NRA averaged 17 percent in Australia, whereas in New Zealand it averaged little more than 2 percent until the mid-1970s. The standard deviation of agricultural NRAs also rose during that period, indicating increasing misallocation of resources within the agricultural sectors of these two countries (tables 5.1 and 5.2). A striking feature of agricultural assistance in Australia and New Zealand at that time was that it applied to export industries as much as to import-competing ones (figure 5.2). Export industries in New Zealand, including wool, dairy prod- ucts, and meat, were assisted by so-called stabilization schemes (Sandry and Reynolds 1990). Similar schemes in Australia assisted wheat, manufactured dairy products, sugar, and dried vine fruit. The Australian schemes often also contained so-called home consumption price schemes, whereby domestic consumers were Table 5.1. NRAs to Covered Farm Products, Australia, 1946­2007 (percent, for fiscal years starting July 1) 1946­49 1950­54 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Exportables 7.5 0.9 6.4 7.0 10.0 7.6 3.6 4.6 5.6 4.8 3.0 0.0 0.0 Rice 3.2 1.1 11.4 15.0 14.8 22.0 20.4 15.2 10.6 2.5 2.3 1.7 1.9 Wheat 24.2 8.4 1.9 6.1 10.1 7.2 0.4 2.6 3.8 2.1 1.1 0.0 0.0 Barley 14.1 5.8 4.1 3.1 4.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Oats 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Grapes, total 10.5 4.5 5.6 9.7 18.7 39.7 19.2 21.3 18.3 13.3 4.9 0.0 0.0 Sugar 8.2 0.7 12.8 15.9 32.8 7.6 6.2 4.6 12.4 5.8 1.7 0.0 0.0 Cotton 0.8 2.0 26.7 52.1 73.9 53.4 17.6 4.4 2.0 0.0 0.0 0.0 0.0 Wool 0.0 0.0 0.0 0.0 0.0 6.0 1.4 1.0 1.0 5.4 0.7 0.0 0.0 Beef and veal 0.0 0.0 0.0 0.0 0.0 1.4 1.8 1.4 1.2 0.3 0.0 0.0 0.0 Mutton and lamb 0.0 0.0 0.0 0.0 0.4 1.6 1.8 1.4 1.8 0.9 0.0 0.0 0.0 Pig meat 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Milk 2.1 18.7 46.9 43.1 74.5 32.8 35.8 32.2 39.6 23.8 19.3 0.0 0.0 Apples -- -- 6.0 6.0 6.0 9.0 5.4 3.4 1.2 0.4 0.1 0.0 0.0 Sunflowers -- -- -- -- -- -- -- 5.6 1.4 0.0 0.0 0.0 0.0 Import- competing products 0.0 10.1 13.4 12.5 13.1 18.3 11.6 8.0 3.7 1.8 0.4 0.1 0.0 Maize 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Sorghum 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.6 2.0 0.0 0.0 0.0 0.0 Oilseeds 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Tobacco 0.0 34.2 51.0 46.9 51.3 250.0 122.2 56.4 37.6 48.5 19.8 0.0 0.0 229 Poultry 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 (Table continues on the following page.) 230 Table 5.1. NRAs to Covered Farm Products, Australia, 1946­2007 (continued ) (percent, for fiscal years starting July 1) 1946­49 1950­54 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Bananas -- -- 0.0 0.0 0.0 0.0 0.0 4.8 1.0 0.1 0.0 0.0 0.0 Olives -- -- 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Oranges -- -- 25.0 25.0 25.0 25.8 32.8 38.2 13.0 2.7 0.7 0.6 0.6 Soybeans -- -- 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.0 Nontradables 1.2 12.6 31.4 41.6 78.1 25.3 19.5 24.2 12.2 1.8 0.2 0.0 0.0 Eggs 1.7 14.7 43.7 61.8 141.2 35.0 26.0 35.8 18.4 3.4 0.4 0.0 0.0 Potatoes 0.0 8.0 8.0 8.0 8.0 7.2 7.2 8.0 3.2 0.0 0.0 0.0 0.0 Weighted average of covered products 7.0 1.8 7.8 8.5 12.3 8.8 4.6 5.4 5.7 4.4 2.6 0.0 0.0 Domestic market support -- -- 1.4 1.6 2.5 1.0 0.8 0.9 0.5 0.4 0.2 0.0 0.0 Border market support -- -- 6.5 7.0 9.8 7.8 3.9 4.6 5.2 4.0 2.5 0.0 0.0 Dispersion of covered products 7.4 12.2 17.3 21.3 36.1 54.0 27.9 16.9 12.2 10.9 5.8 0.4 0.4 % Coverage (at undistorted prices) 91 84 82 86 87 84 85 86 75 82 80 80 80 Source: Anderson and Valenzuela (2008), drawing on authors' spreadsheet. Note: -- not available. a. Weighted averages, with weights based on the unassisted value of production (actual back to 1966, and the average for 1966­69 for earlier years). Dispersion of NRAs is their standard deviation of the simple five-year average of the annual standard deviation around the weighted mean. Table 5.2. NRAs to Covered Farm Products, New Zealand, 1955­2007 (percent) 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 Exportables 0.1 0.1 0.2 2.8 13.1 18.8 11.8 1.2 0.8 0.9 0.6 Barley -- -- -- -- -- 10.6 1.8 0.0 0.0 0.0 0.0 Beef 0.1 0.1 0.3 5.0 10.0 15.6 11.0 1.4 1.0 1.0 0.3 Coarse grains 4.0 4.0 4.0 4.0 4.0 4.0 2.2 0.4 0.0 0.0 0.0 Other fruits and vegetables 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Maize -- -- -- -- -- 12.0 2.2 0.0 0.0 0.0 0.0 Milk 0.2 0.2 0.2 1.0 16.0 18.0 11.6 1.4 1.0 1.0 0.3 Oats -- -- -- -- -- 9.2 2.4 0.0 0.0 0.0 0.0 Other crops 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Sheep meat 0.1 0.1 0.3 5.0 19.0 32.9 27.8 1.8 1.0 1.0 0.3 Wool 0.0 0.0 0.0 5.0 11.0 19.0 10.2 1.4 1.0 1.0 0.3 Import-competing 28.3 28.3 28.8 32.0 27.0 28.1 41.1 22.5 19.6 15.8 8.9 Eggs 59.0 59.0 59.0 59.0 59.0 59.0 80.2 38.2 50.4 36.4 25.0 Pig meat 2.0 2.0 2.8 5.4 19.9 10.3 2.8 0.0 0.0 0.2 0.3 Poultry 31.0 31.0 31.0 31.0 31.0 31.0 61.6 57.0 40.8 34.6 28.3 Wheat 11.0 11.0 11.0 11.0 11.0 11.0 6.6 0.6 0.0 0.0 0.7 Mixed trade status Grapes 120.0 120.0 134.0 106.0 53.8 23.2 23.8 5.0 5.0 5.0 5.0 All covered 1.8 1.8 1.9 5.0 14.3 20.0 14.9 2.9 2.1 2.0 1.3 Domestic market support 0.0 0.0 0.0 0.0 0.0 0.2 2.0 0.0 0.0 0.0 0.0 Border market support 1.8 1.8 1.9 5.0 14.3 19.8 12.9 2.9 2.1 2.0 1.3 Dispersion of NRA of covered products 40.7 40.7 44.5 36.3 23.2 17.4 27.1 17.5 17.3 14.0 9.8 % Coverage at undistorted prices 100 100 100 100 100 100 100 100 100 100 80 Source: Anderson and Valenzuela (2008), drawing on authors' spreadsheet. 231 Note: -- not available. a. Weighted averages, with weights based on the unassisted value of production (actual back to 1966, and the average for 1966­69 for earlier years). Dispersion of NRAs is their standard deviation of the simple five-year average of the annual standard deviation around the weighted mean. 232 Distortions to Agricultural Incentives: A Global Perspective Figure 5.2. NRAs to Exportable, Import-Competing, and Alla Covered Products, Australia and New Zealand, 1946­2007 a. Australia 25 20 15 10 percent 5 0 5 10 15 20 06 01 46 51 56 61 66 71 76 81 86 91 96 20 20 19 19 19 19 19 19 19 19 19 19 19 b. New Zealand 45 40 35 30 percent 25 20 15 10 5 0 9 4 9 4 9 4 9 4 9 4 7 ­5 ­6 ­6 ­7 ­7 ­8 ­8 ­9 ­9 ­0 ­0 55 60 65 70 75 80 85 90 95 00 05 19 19 19 19 19 19 19 19 19 20 20 total import-competing products exportables Source: Anderson and Valenzuela (2008), drawing on authors' spreadsheet. a. The total NRA can be higher than both the exportable and importable averages because nontradables assistance is also included. forced to pay more than the export price (Sieper 1982; Edwards 2006). These schemes, which required the pooling of domestic and export returns, could only be implemented with the support of the Australian state governments.10 All other policy measures combined--fertilizer subsidies, income tax incentives, rural credit measures, subsidies to agricultural research and extension, and public Australia and New Zealand 233 investment in land and water development and rural infrastructure--added the equivalent of no more than 2 percent to Australian farmers' gross income as of the early 1970s (table 5.3). The net effect of farm and nonfarm policies on agricultural incentives is sum- marized in table 5.3 and illustrated in figure 5.3. For Australia, the negative effect on incentives from agricultural policies in the 1940s was trivial compared with that from nonagricultural ones, mostly import protection for manufacturers. Together, those policies effectively reduced farmers' gross returns by more than 20 percent. Price stabilization and other agricultural policies gradually provided more direct assistance to Australian farmers over the 1950s and 1960s, when manu- facturing protection remained steady, so that degree of overall taxation fell from 27 percent to just under 10 percent by 1965­69, as measured by the RRA. Of that decline, about two-thirds is due to changes in nonagricultural policies and only one-third to changes in direct assistance to farmers. New Zealand farmers, mean- while, were effectively taxed an average of more than 20 percent until 1972­73. The degree of antiagricultural bias in the two countries to the early 1970s was thus very similar to that of many developing countries in those decades. Meanwhile, by the late 1960s, home consumption price schemes imposed tax equivalents of over 100 percent on Australian consumers of butter, cheese, sugar, and eggs. These dis- tortions are unusual in that the imposition they imposed on consumers or users of exportables was greater than the price impact on producers--something that does not arise from standard trade barriers. Reforms of the early 1970s Disenchantment with interventionist trade and related economic policies gradually spread in the 1960s. At the 1968 National Development Conference in New Zealand, policy makers nearly agreed to simultaneously reduce import protection and tariff compensation to agriculture. But it was not until the 1970s that tariff reductions began in both countries. In Australia, a 25 percent across- the-board cut in July 1973, preceded by some minor cuts in 1970­71, started the tariff reform process. It was accelerated in the 1980s and continued through the 1990s. As a result, the average NRA to Australian manufacturing fell from 23 to 3 percent, and the effective rate from 36 to 5 percent over those three decades. In the 1990s alone, both the mean and the standard deviation of Australia's import tariffs on goods halved, bringing the average tariff for manufactures down to 4.2 percent in 1999.11 Currently, the only manufacturers with significant tariff protection are motor vehicles and parts, textiles, clothing, and footwear. Excluding those products, the average effective rate of assistance to Australian manufactur- ing is just 3 percent (Productivity Commission 2000a).12 Table 5.3. NRAs to Agricultural Products Relative to Nonagricultural Industries, Australia and New Zealand, 234 1946­2007 (percent, for fiscal years starting July 1) a. Australia 1946­ 1950­ 1955­ 1960­ 1965­ 1970­ 1975­ 1980­ 1985­ 1990­ 1995­ 2000­ 2005­ 49 54 59 64 69 74 79 84 89 94 99 04 07 NRA, covered products 7.0 1.8 7.8 8.5 12.3 8.8 4.6 5.4 5.7 4.4 2.6 0.0 0.0 NRA, noncovered products 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 NRA, all agriculture (excluding NPS) 6.4 1.5 6.3 7.4 10.7 7.6 3.9 4.6 4.4 3.7 2.1 0.0 0.0 All importables -- -- 3.8 4.5 4.9 7.4 4.9 3.5 1.2 0.8 0.2 0.0 0.1 All exportables -- -- 5.7 6.4 9.3 7.0 3.3 4.2 4.8 4.3 2.6 0.0 0.0 All nontradables -- -- 31.4 41.6 78.1 25.3 19.5 24.2 12.2 1.8 0.2 0.0 0.0 Trade bias index -- -- 0.02 0.02 0.04 0.00 0.02 0.01 0.04 0.03 0.02 0.00 0.00 NPS 0.6 1.2 1.8 2.1 2.0 2.1 1.4 1.1 0.9 0.7 0.8 0.5 2.2 Inputs -- -- -- -- -- -- 1.2 0.7 2.0 3.0 3.3 2.8 2.8 Other -- -- -- -- -- -- 1.4 1.1 1.1 2.3 2.5 2.3 0.6 NRA, all agriculture (including NPS) 5.7 2.7 8.1 9.5 12.7 9.4 5.3 5.8 5.3 4.4 2.9 0.5 2.2 NRA, decoupled assistance 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.9 1.4 1.2 1.2 2.3 2.6 NRA, all agricultural products (including NPS and decoupled assistance) 5.7 2.7 8.1 9.5 12.7 9.4 5.4 6.6 6.6 5.6 4.1 2.8 4.8 NRA, all agricultural tradables (including NPS) 5.9 2.3 7.3 8.4 10.9 8.9 4.9 5.2 5.1 4.5 3.0 0.5 0.2 NRA, all nonagricultural tradables 28.0 23.5 19.6 20.7 20.7 16.8 12.0 11.1 8.2 5.3 2.6 2.0 2.0 RRA 26.3 17.1 10.3 10.2 8.2 6.8 6.4 5.3 2.9 0.7 0.4 1.5 1.8 b. New Zealand 1955­59 1960­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 NRA, covered products 1.8 1.8 1.9 5.0 14.3 20.0 14.9 2.9 2.1 2.0 1.3 NRA, noncovered products n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. -- NRA, all agricultural products (excluding NPS) 1.8 1.8 1.9 5.0 14.3 20.0 14.9 2.9 2.1 2.0 1.2 All importables 28.3 28.3 28.8 32.0 27.0 29.6 42.1 24.3 20.7 17.6 9.8 All exportables 0.1 0.1 0.2 2.8 13.1 18.9 12.0 1.2 0.8 0.9 0.6 All nontradables -- -- -- -- -- -- -- -- -- -- -- Trade bias index 0.22 0.22 0.22 0.22 0.11 0.08 0.21 0.18 0.16 0.13 0.07 NRA, NPS -- -- -- -- -- -- 5.5 0.5 0.6 0.4 0.3 Inputs -- -- -- -- -- -- 4.6 0.5 0.6 0.4 0.3 Other -- -- -- -- -- 8.9 0.9 0.0 0.0 0.0 0.0 NRA, all agriculture (including NPS) 1.8 1.8 1.9 5.0 14.3 28.9 20.9 3.4 2.6 2.4 1.4 NRA, decoupled assistance 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.1 NRA, all agricultural products (including NPS and decoupled assistance) 1.8 1.8 1.9 5.0 14.3 28.9 20.9 3.5 2.6 2.4 1.5 NRA, all agricultral tradables (including NPS) 1.8 1.8 1.9 5.0 14.3 28.9 20.9 3.4 2.6 2.4 1.4 NRA, all nonagricultural tradables 21.3 24.0 34.3 30.0 21.7 20.3 16.6 10.8 6.5 3.8 3.3 RRA 16.1 17.8 24.1 19.0 6.0 7.1 3.5 6.7 3.6 1.3 1.8 Source: Anderson and Valenzuela (2008), drawing on authors' spreadsheet. Note: -- not available; n.a. not applicable. a. Before including Non-Product Specific (NPS) assistance. 235 b. Total of assistance to primary factors and intermediate inputs divided by the total value of primary agriculture production at undistorted prices (percent). c. RRA 100[(1 NRAagt 100) (1 NRAnonagt 100) 1], where NRAagt and NRAnonagt are the average percentage NRAs for the tradable parts of the agricultural and nonagricultural sectors, respectively. 236 Distortions to Agricultural Incentives: A Global Perspective Figure 5.3. NRAs to Manufacturing, All Nonagricultural Tradables, All Agricultural Tradable Industries, and RRA,a Australia and New Zealand, 1946­2007 a. Australia 50 40 30 20 10 percent 0 10 20 30 40 50 01 06 46 51 56 61 66 71 76 81 86 91 96 20 20 19 19 19 19 19 19 19 19 19 19 19 nonagricultural tradables agricultural tradables RRA b. New Zealand 60 50 40 30 percent 20 10 0 10 20 30 9 4 9 4 9 4 9 4 9 4 7 ­5 ­6 ­6 ­7 ­7 ­8 ­8 ­9 ­9 ­0 ­0 55 60 65 70 75 80 85 90 95 00 05 19 19 19 19 19 19 19 19 19 20 20 manufacturing nonagricultural tradables agricultural tradables RRA Source: Anderson and Valenzuela (2008), drawing on authors' spreadsheet. a. The RRA is defined as 100 [(100 NRAagt ) (100 NRAnonagt) 1]. Australia and New Zealand 237 Australia's agricultural subsidies and regulatory interventions also have been close to eliminated over those 35 years. The average NRA to the farm sector has fallen from 16 percent in the early 1970s to less than 2 percent at present, and the standard deviation has fallen from almost 60 percent to less than 1 percent. The process was piecemeal13 and gradual, often involving a series of partial steps, but it was persistent, beginning in 1972 with milk and then two to four years later with cotton and tobacco. It took another decade, though, for supports for rice and eggs to begin to be dismantled, and almost a further decade for assistance to the grape industry to be cut. The process also experienced a significant reversion--the introduction of the reserve price scheme for wool in 1973--that took until the early 1990s to unravel.14 There were just two farm groups still benefiting signifi- cantly from government programs in the latter 1990s: tobacco and milk produc- ers, each with an NRA of more than 20 percent in 1997. Deregulation of tobacco marketing arrangements began in 1995 and was completed in 2000, bringing effective assistance to tobacco growing down from 30 to 2 percent over that period. As from July 1, 2000, the remaining impediments to a free domestic mar- ket in fluid milk began to be dismantled, for which an untied, one-off compensa- tion grant (providing a total of around US$1.5 billion, or US$110,000 per dairy farm) has been paid, funded by a consumer levy at the retail level over 2000­08. Contrary to some pessimistic forecasts at the time, dairy output has continued to increase as production from those leaving the industry (17 percent over the first three years) has been more than compensated by output and productivity improvements on remaining farms (Harris 2005a, 2005b). In New Zealand, the antiagricultural bias declined in the 1970s as the result of a modest reduction in manufacturing protection and a big boost to beef, sheep, and dairy farmer assistance. In these two steps, the effective taxation of agriculture (the negative of the RRA) fell from 22 to 14 percent, and then to less than 7 per- cent. It fell further over the next dozen years as agricultural assistance increased slightly and manufacturing protection continued to fall slightly, then rose a little from the late 1980s as assistance to farmers fell faster than that to manufacturing, and finally fell to just 1 percent as the last of the interventions were removed from the late 1990s (figure 5.3). Thus, distortionary government assistance to both manufacturing and agricul- ture, and hence the overall antiagricultural bias, has now all but disappeared in both Australia and New Zealand after being in place for more than seven decades.15 Over the past 35 years, it was nonagricultural policy that overwhelm- ingly contributed to the improvement of farmers' incentives in Australia and New Zealand--in fact, the decline in manufacturing protection was sufficiently large enough to more than offset the decline in direct assistance to farmers from the late 1960s. Farmers also have benefited from the fact that service sectors too have not 238 Distortions to Agricultural Incentives: A Global Perspective been spared reform in these countries. Banking; post and telecommunications; ports; higher education; health; and rail, air, and sea transport have been opened up. In addition, there has been progressive outsourcing of many government serv- ices; and substantial reforms to competition policy and practice, including priva- tization and the corporatization and demonopolization of numerous government enterprises, are well advanced.16 Moreover, by 1985 both currencies were floating and foreign investment flows began to be freed up. That complemented financial sector reform and contributed to foreign direct investment, equity and foreign currency transactions growing at several times the pace of GDP. Even the previ- ously highly unionized labor markets have undergone considerable reform. Households have gained substantially from these widespread reforms, including consumers of food who for most of the past two decades have faced tax equiva- lents of well below 10 percent on their food purchases (compared with the OECD average of between 23 and 36 percent over that period--see OECD 2008). Reasons behind the Policy Evolution In the early postwar years, successive Australia and New Zealand governments sought to expand the output of the agricultural sector in order to improve the bal- ance of trade position in the context of a fixed exchange rate regime. There were numerous factors in common, but since some were country-specific, it is better to consider the two countries separately. Australia In the Great Depression years of the 1930s, and even more so in the two decades following World War II, agricultural policy instruments were introduced in profu- sion, including price stabilization schemes implemented through buffer funds, guaranteed prices and deficiency payments, home-consumption price schemes for wheat, dairy, sugar, and fruits and vegetables; input subsidies on fuel, fertilizers and interest rates; import subsidies; tax concessions; publicly funded research and development; publicly funded extension services; land development schemes; and the provision of rural infrastructure (Godden 1997; Mauldon 1990). For some commodities--for example, sugar and wheat--there were national- level statutory marketing authorities (SMAs) or, in the language of the GATT and, later, the World Trade Organization (WTO), state trading enterprises (STEs). This instrument was also used by state governments across a wide range of commodi- ties. It often took the form of a producer-controlled board, with compulsory monopsony or monopoly powers in order to effect "orderly marketing" and sometimes used in conjunction with marketing quotas. Frequently, the SMAs Australia and New Zealand 239 were linked with regulations that required the shipment and storage of grain to be handled by state-owned enterprises. This form of intervention by Commonwealth and state governments began in the 1920s and became central to marketing policy (Mauldon 1990).17, 18 In 1973, the Australian Labor Party won office and, as a political party with few roots in rural areas, set about establishing a review of the objectives and instru- ments of agricultural policy in order to define a set of principles that would underpin intervention. The result was the so-called Green Paper (Harris et al. 1974). The approach adopted by the authors was that of applied welfare economics: market failure provides a reason for government intervention to be potentially beneficial through improving economic efficiency. Some years later, Sieper (1982) argued that the public choice or private interest explanations fit better with reality, and particularly with the choice of policy instruments, than those provided by the public interest approach. Although the insights provided by that set of theories were not entirely new, Sieper's use of them provided fresh and powerful insights into the Byzantine world of Australian agricultural policy as it existed up to the late 1970s. In 1974, the Industries Assistance Commission (IAC) was established from the then-existing Tariff Board. The IAC was designed to act as a forum for public dis- cussion of economic policy by undertaking reviews and evaluation of economy- wide government intervention and regulation. During the 1970s and 1980s, it produced reports on wheat, sugar, dairy, dried vine fruits, rural adjustment, and income stabilization (Edwards 1987). This development signalled the govern- ment's intention that policies should be evaluated from the viewpoint of the public interest rather than, or perhaps as well as, the private interests of farm groups and regions. In other words, the year 1973 marked a watershed: applied welfare economic analysis would be used to analyse the effects of various policy instruments and the recommendations of that analysis would be treated objec- tively, even if not always implemented in full (or, in a few cases, at all). In 1982, the Balderstone Report set out the agricultural policy options for the 1980s (Balderstone 1982). The basic thrust was that of reduced protection, not just in agriculture but throughout the entire economy. The authors supported the position that the agricultural sector should receive compensation on the grounds of equity (but not efficiency as discussed in the 1970s; see Martin [1990]) for hav- ing to pay tariffs on imported inputs. In the report, the authors adopted an incre- mentalist approach that supported the continuation of many of the then-existing instruments--for example, input subsidies and price discrimination between domestic and export markets. They also suggested that the objective of price stabilization should be achieved through price underwriting linked to market conditions rather than through existing instruments such as buffer funds. 240 Distortions to Agricultural Incentives: A Global Perspective In 1988, the Commonwealth government issued a policy statement stating that the agricultural policy objectives were to be, among other things, enhancing pro- ductive capacity through balanced management of natural resources, development of human skills and improved research; developing a more productive industry structure through an economywide approach to policy reform with lower and more balanced assistance; responding to the international environment through better marketing of agricultural produce and reductions in international trade distortions; and offering positive assistance for structural adjustment (see Godden 1997). With this, it was obvious that the policy objectives of the earlier postwar years had been substantially altered. The Australian Wheat Board lost its monopoly in the domestic food wheat market in 1989, as had been recommended by the IAC in 1984, but it retained (and continues to retain) its export monopoly. In 2000, the single export desk for wheat was reviewed under the National Competition Policy and, in the absence of strong evidence one way or the other, the government accepted the recommenda- tion that the export monopoly be retained until 2010. Reviews under National Competition Policy of the single export desk for rice and for sugar also recom- mended their continuation, based on a national interest test. In 1991, a review of the statutory arrangements for the marketing of agricul- tural products (Industry Commission 1991) was published. The investigation focused on the economic effects of these arrangements, in particular the effects on efficient resource allocation. The Commission concluded that although many of the arrangements were beneficial to producers and to society generally, there were instances where this was not so. The SMAs did not operate in the public interest when they had powers to require producers to participate (that is, the compulsory acquisition of product), when potential producers were excluded from entry (for example, when there are nontradable quotas in place), or where prices were raised for users and consumers (for example, price discrimination between the domestic and export markets). In these instances, it was concluded that efficiency of resource use was not achieved.19 The outcome of this review has been the disap- pearance of most, but not all, SMAs. From its beginnings in the latter 1970s, the leadership of the National Farmers Federation (NFF, and before it the Grains Council) adopted an economy-wide approach to economic issues. The activities of the NFF also encouraged the devel- opment of a free-market faction in the Liberal Party in the late 1970s (Anderson and Garnaut 1987). That approach was to be the hallmark of much of the rigor- ous computable general equilibrium analysis conducted by the IAC over the decade to 1989 on matters both agricultural and more general. Early in that period, the government had convinced the farm sector of the need for rationaliza- tion of the sector. As a quid pro quo, the government would introduce measures Australia and New Zealand 241 to improve efficiency in the sectors supplying inputs to agriculture as well as the infrastructure required for distribution--for example, rail, storage, and the wharves. However, some commodity groups within the NFF complained that the leadership of the organization had turned its back on protecting its constituents from harsh economic conditions (for example, drought and the vagaries of inter- national markets) and from the deregulatory zeal of the government (including through one of its main sources of advice on agricultural matters, namely, the IAC/IC). The objectives and the instruments of agricultural policy that existed from the early postwar years until the late 1980s reflected not only the supply of assistance to producers through government from consumers (users) and taxpayers but also the demand for assistance from producer groups. As shown by Anderson (1978), the interaction between the two sides of the political market generated a wide dispersion of support from the late 1960s to the mid-1970s. His findings are con- sistent with the interest group theory of political economy, or public choice theory: the land-intensive subsectors such as cereals and livestock were found to be lightly supported, whereas the labor-intensive subsectors such as milk produc- tion and horticulture were heavily supported. Martin (1990) outlined several models in the literature on public choice theory and evaluated the extent to which each was relevant to the developments in agricultural policy that occurred in the 1980s. In assessing the role of the IAC, he concluded that "the establishment of the IAC also appears to have had a remarkably strong influence on the policy process, and to have been associated with and contributed to the change in ideas about industry assistance measures" (p. 198). MacLaren (1992) argues, in a comparison of the processes of agricultural policy reforms in Australia and the European Community (EC), that the IAC proved significant in the reform of Australian agricultural policy because it altered the political economy of policy making. No other industrial country has a body with the functions of the IAC, and the lack of such a body may help to explain why reductions in agricultural protectionism in other industrial countries, with the exception of New Zealand, have been very dif- ficult to achieve. Over the past 25 years, much of the intervention in both agriculture and manu- facturing in Australia has disappeared. In agriculture, the focus of government policy on incomes and prices has given way to concerns about the quality of the resources used by the agricultural sector and the efficiency with which they are used. During the same period, the principal farm lobby groups have acquiesced to the view that markets can perform better than governments in the provision of an economically healthy sector. Subsectors that traditionally were heavily supported have accepted the benefits of reduced market-based intervention. In 2000, even the dairy sector accepted total deregulation of the prices of both market and 242 Distortions to Agricultural Incentives: A Global Perspective manufacturing milk. The subsequent free interstate trade in market milk led to better exploitation of regional comparative advantages. The bribe paid to secure acceptance by all regions was an adjustment assistance scheme financed by con- sumers who will continue to pay a high price for milk during the transition period of eight years.20 This acceptance of reduced government involvement in the mar- kets for agricultural products contrasts markedly with the situation in most other OECD countries, and in many developing countries, as confirmed by the contin- uing difficulties in defining modalities for domestic support and market access in the Doha Round. New Zealand Well into the 1960s, New Zealand continued an isolationist economic policy (import selection through licensing) that had been introduced just after the Great Depression. Import selection effectively prohibited imports of competing goods by using a system of import licensing underpinned with high tariffs. Import pro- hibitions remained in place for selected food products, the most notable of which was for margarine, which could be purchased only with a doctor's certificate! Price supports, coupled with import licensing, increasingly protected the domes- tic wheat industry. At the other extreme, export prohibitions on coarse grains tended to lower the price of feed wheat, barley, and oats to poultry and pork pro- ducers and other users. For a short period around 1955, the import licensing regime was liberalized (Rayner and Lattimore 1991), but a balance of payment crisis in 1957 (falling dairy product prices) caused a policy reversal. In this strong import-substitution environment, exports were heavily concentrated in meat, wool, and dairy products--the industries in which New Zealand had its strongest comparative advantage and to which the government provided the least assistance (table 5.2). Until the mid-1950s, these "pastoral" exports represented nearly 95 percent of total merchandise exports from New Zealand. At that point, nonagricultural exports began to grow, so that by 1969 agriculture's share of merchandise exports fell to 84 percent, in part due to the advent of the highly selective New Zealand­Australia Free Trade Agreement that began in 1966. The New Zealand economy suffered a series of domestic and international eco- nomic shocks over the seven-year period starting in 1967 (Dalziel and Lattimore 2004). In 1967, export earnings were severely reduced by a fall in international wool prices, leading to the exchange rate being devalued. Furthermore, New Zealand was beginning to import inflation via its fixed exchange rate policy and the expanding U.S. fiscal deficit surrounding the Vietnam War. Domestically, an increase in the rate of inflation led to a breakdown in the national wage-setting Australia and New Zealand 243 mechanism, resulting over the next few years in large wage increases. The govern- ment then began increasing farm input subsidies to counteract the drop in income. Another terms-of-trade shock came in 1972­73, when meat prices rose to high levels as part of a general agricultural commodity price boom. Though New Zealand revalued its exchange rate to counteract the macroeconomic effects, the United Kingdom joined the European Economic Community (EEC) in 1973, reducing New Zealand's longstanding preferential access to this very profitable agricultural export market.21 The first oil crisis came close on the heels of these occurrences, causing an additional major adverse terms-of-trade shock. New Zealand's economic policies changed rather dramatically as a result. In an attempt to maintain consumption in the face of drastic falls in the terms of trade, a com- mand economy approach was adopted, under which industries, prices, interest rates, wages, and foreign exchange rates were more heavily regulated. High infla- tion and large fiscal and current account deficits ensued. Foreign borrowing increased significantly during this period, and was used in part to finance major energy projects in an attempt to reduce New Zealand's reliance on imported fuel. In order to finance this debt, agricultural output subsidies on exportables were used for the first time in an attempt to generate additional export earnings. Farm development subsidies also were increased substantially. The sheep and wool industries were the favored targets because they were the largest agricultural and export sectors, but export subsidies were also provided for some nontraditional export products. The result was a large increase in the NRA offered to the exportable sector and a closing of the historically large gap between rates of assistance to the import-competing and export sectors for about a decade (figure 5.2b). The NRA on outputs in exportable primary agricul- ture rose to 19 percent over 1980­84 (table 5.2). During the reform period from the mid-1980s, output assistance to exportable primary agriculture was abruptly removed and the rate of assistance fell to almost zero in less than a decade. The sharp rises within exportable primary agriculture in the early 1980s included 24 percent for beef in 1982 and 59 percent for mutton and lamb production in 1984. Output assistance was nearly as high for raw milk and wool production. No significant output subsidies applied to the other exportable subsectors, such as fruit and vegetables and deer production, nor to import-competing subsectors such as grains and oilseeds. Output assistance for other animal products appears to be the result of high nontariff phytosanitary barriers on pig meat (during the 1980s), poultry, and eggs. The NRAs for favored exportables (sheep, beef, wool, and dairy) all reached a peak in the 1975­84 period, although not in the same years because the deficiency payments were tailored to the different world price levels for each year. Most of 244 Distortions to Agricultural Incentives: A Global Perspective the importable NRAs were zero because New Zealand does not produce sugar, cotton, soybeans, or rice. The first agricultural adjustments occurred in 1981 in wheat, tobacco, and meat processing. Tobacco growers were offered a production quota buyout of NZ$7,000 per hectare. These quotas resulted from local-content regulations, which prescribed that half of tobacco leaf used for processing by the two firms (Wills and Rothmans) had to be grown in New Zealand. Around two-thirds of the producers accepted the grant and relinquished their production rights. The remaining producers continued in production. Wills ceased operations in New Zealand in 1987 but Rothmans continued until 1994, paying NZ$6.52 per kilo- gram for dried leaf grown in New Zealand even though the import price (and over-quota price) was around NZ$4 per kilogram. The NRA on tobacco at this time was, accordingly, 63 percent. In 1983, New Zealand entered into a new and much more comprehensive free trade agreement with Australia (ANZCERTA). This arrangement was the first vehicle, after 1955, used to begin the process of removing the 1938 import licens- ing system and reducing import tariffs. In the same year, under some pressure from the United States, it was announced that the agricultural output subsidies would be abolished. Wide-ranging macroeconomic and microeconomic reforms were introduced in 1984. They included monetary and fiscal policy reform and reductions in industry regulation and support. The process of removing almost all agricultural subsidies was begun by abolishing agricultural output subsidies and phasing out agricultural input subsidies--a process that was completed by 1990. Policies affecting nontradables also were important, as they promoted greater macroeco- nomic stability and resource mobility, both of which increased the effectiveness of changes in trade policies. The government did not remove the export monopoly rights of agricultural marketing boards over this period, and while it did signal in a general way that such action was on the agenda, it nonetheless introduced a new marketing board with monopoly export rights (for kiwifruit). Furthermore, wine- grape growers were given a subsidy to replace old varieties with newer ones. Import licenses for Australian goods were increasingly tendered and then abol- ished once tender premia had fallen to low levels. This was followed by the removal of all other import licenses; the last ones (on textiles, clothing, and footwear) were removed in the early 1990s. Tariffs were progressively reduced from 1986 until 2000. Further tariff reductions were then halted, as import tariffs had fallen to around the average of those in other high-income countries. Monetary policy in New Zealand was reformed in 1989, followed by liberaliza- tion of labor market legislation and fiscal policy constraints. These initiatives com- pleted the liberalization of the capital account begun earlier with the conversion to Australia and New Zealand 245 a floating exchange rate regime; full currency convertibility; and the removal of interest rate, price, and wage controls. In this environment, the New Zealand macroeconomy gradually stabilized over the period from 1984­91. The notable exception was unemployment, which continued to rise, to around 11 percent by 1991 (Evans et al. 1996). The unemployment rate did not fall to less than 4 per- cent, where it had been in 1984, until 2003. From 1990, the major change to agricultural policy was the removal of export monopoly rights administered by some agricultural marketing boards (for exam- ple dairy, meat, apples, and pears) and, in some cases such as dairy, the abolition of the marketing board itself. This process was complicated by the existence of bilateral import quotas--especially for exports to the European Union, United States, Canada, and Japan. In the case of dairy quotas, the import quota rights were given to the three export dairy companies. The largest of the three, Fonterra, subsequently bought those rights from the smaller ones. In the case of meat, a reg- ulatory body was retained within Meat New Zealand (the marketing and research organization that developed from the old New Zealand Meat Producers Board) to allocate bilateral quota use rights. Under pressure from trading partners in the Doha Round, Fonterra's quota rights are to be removed and allocated in an alter- native fashion that is yet to be decided. By 1990, government assistance to New Zealand agriculture was limited to biosecurity, adverse events relief, research and development subsidies, export market development assistance, and product quality control programs. The direct rate of assistance to exportable agriculture had fallen to around 1 percent. The consumer tax equivalent (CTE) for exportable food products was down to a similarly low level, and comprised mainly the import-export parity differential that can be gained by large export firms in dairy, meat, and forestry products. The CTE for importable food products is somewhat higher because there are still import tariffs on some of these import-competing items, although typically they are still within the range of 5­7 percent. The exceptions are poultry and eggs, imports of which are strictly controlled on animal health grounds. Accordingly, domestic prices of these items are significantly higher than world market prices. Because the OECD considers these phytosanitary restrictions nontariff barriers, they are analyzed here as elements of nominal assistance to outputs (even though part of that protection may be correcting for a market failure). By 2005, the rate of assistance to manufactured food products was in single digits. However, the traditional pattern of tariff escalation remains, in that tariffs on unrefined exportable products and noncompeting imported food components tend to be zero while higher value-added foods have tariffs in the 5­7 percent range. In the nonfood manufacturing sector, NRAs on output have fallen from 24 and 42 percent (on exportables and importables, respectively) in 1982 to 4 percent 246 Distortions to Agricultural Incentives: A Global Perspective in 2009. Following the removal of the last import licensing program in 1993, tar- iffs were reduced continuously until 2000, when the most favored nation (MFN) tariff reduction programs were halted. One reason for this policy change was to create bargaining chips in the pursuit of bilateral and plurilateral free trade agree- ments that New Zealand embarked on vigorously--most recently with China, the United States, and ASEAN (Association of Southeast Asian Nations) Plus Three. There were some larger, one-off tariff reductions over the period, though: early in the reforms the tariff on computers was reduced from 40 percent to zero, and in 1997 the last remaining MFN tariff on new cars (25 percent) was abolished. Prospects for Further Policy Reform Notwithstanding the huge amount of agricultural trade policy reform over the past two decades, plenty of policy issues remain on the table, predominantly in the natural resource and environmental areas. Three are illustrated here. The first is food and agricultural import restrictions for the protection of plants, animals, and human health, though the economic protection from import competition that this provides farmers has not been fully captured in the Produc- tivity Commission's NRA estimates for Australia, especially for horticultural products. While some of that protection may well be warranted on externality grounds, some (such as a complete ban on imports of certain fruits from all coun- tries) may be excessive from a national welfare viewpoint. The Australian govern- ment, mainly in response to pressure from other WTO members seeking greater market access, is slowly examining whether various measures are excessively restrictive. Typically, consumer costs are not included in such assessments, nor are any of the cheaper ways of reducing costs associated with the importation of disease (James and Anderson 1998). New Zealand, in particular, would be a beneficiary of a more liberal quarantine policy regime in Australia. Second, neither Australia nor New Zealand has so far allowed the commercial growing of genetically modified (GM) varieties of farm products, with the sole exceptions of cotton and carnations in Australia. While the Office of the Gene Technology Regulator in Australia has approved the commercial growing of GM canola, the state governments have placed moratoria on plantings. GM food can be sold only if strict labelling standards are adhered to (FSANZ 2007). These restric- tions may or may not be in economies' and consumers' interests, depending on their impact on market access abroad for Australian and New Zealand farm prod- ucts and on human health and the environment at home (Anderson and Jackson 2005), but emotion has played a larger role in formulating these policies than has sound technical and economic analysis, especially in the growing of GM crops. Third, environmental policy has become a very important issue in agriculture in Australia and New Zealand (Vitalis 2005). Water policy, for example, which was Australia and New Zealand 247 already becoming a major economic and political issue, was brought to a head in Australia during 2006 following the country's worst drought on record. Though there has been substantial reform in water policy reform in Australia in recent years, much remains to be done to make the most of this resource, particularly in rural areas, where the majority of water is used and proposals for reform and sev- eral national enquiries are underway (see, for example, Productivity Commission 2006). More efficient pricing of water may lead to substantial reallocation of resources within the agricultural sector, with possible declines in Australian pro- duction of cotton, rice, and milk as horticultural industries (and urban areas) bid away water from farmers.22 The remaining major frontier for policy reform that would boost farm incomes in Australia and New Zealand is the dismantling of agricultural subsidies and import protection abroad. A successful conclusion to the Doha Round of negotiations in the WTO provides the greatest promise for achieving that out- come. According to global Linkage Model results reported in Anderson, Martin, and van der Mensbrugghe (2006), removing all merchandise trade barriers and agricultural subsidies globally would raise agricultural value added or net farm incomes in Australia and New Zealand by 26 percent. Even using the Global Trade Analysis Project (GTAP) Model, in which supply response elasticities are lower than in Linkage, a similar study by Anderson and Valenzuela (2007) reports that agricul- tural value added in Australia and New Zealand would rise by 15 percent, compared with a rise of just 2 percent in value added by nonagricultural industries. Policy Lessons for Other Economies By way of conclusion, two lessons are worth emphasizing from the Australia and New Zealand experiences. For agricultural-subsidizing countries, these case stud- ies show that removing even the largest and longest-lasting farm subsidies is pos- sible. Even where that was done by providing generous adjustment assistance, support was time-bound. In the Australian case, adjustment assistance was able to be financed simply by delaying the rewards to domestic consumers and thereby creating a gap between the producer and consumer price, rather than through outlays from (and hence resistance by) the treasury.23 For developing countries still effectively taxing their agricultural sectors, these two case studies offer hope that good policy analysis and advisory institutions will be able to alter the political economy sufficiently to remove that taxation. More than that, the Australia and New Zealand cases illustrate the growth dividend that can come from reforming such distortionary policies. Having now dismantled virtually all their import protection and agricultural subsidy policy distortions, and having undertaken major domestic macroeconomic and microeconomic reforms over more than two decades, the outcomes of those undertakings are 248 Distortions to Agricultural Incentives: A Global Perspective beginning to be reaped. The impact on overall living standards was mentioned at the outset, but an indicator within the agricultural sector is the acceleration it has given to farm productivity growth. It needs to be borne in mind that farmers in Australia and New Zealand have not been immune from the standard problem facing small farms that requires them to "get big or get out" as the economy develops. It is true that farm sizes in Australia and New Zealand were large relative to those in most other market economies in the early post­World War II years, and that the "wool boom" of the early 1950s provided massive incomes for woolgrowers. Nonetheless, as wages grew in other sectors of the two economies, the need to adjust was felt strongly.24 Adjust- ment manifested itself in the same way as in other countries--that is, with farmers funding agricultural research and adapting and adopting the new technologies it generated as appropriate, and with the number of farms and farmers declining steadily to lower the labor intensity of the sector even as output expanded.25 Within that context, the removal of the antiagricultural policy bias over the past 30 years has substantially boosted, albeit with a not-unexpected delay, the rate of growth of farm productivity in Australia and New Zealand. Figure 5.4 shows that in New Zealand, TFP growth slowed during the dozen or so years of high agricultural subsidies, and only increased to its previous rate after those sub- sidies, along with manufacturing protection, were removed. This pattern has been reflected in farmland prices: after initially halving when the subsidy cuts were announced in the early 1980s, they more than recovered in real terms by the turn of the century as farmers profitably adjusted to the new deregulated, level-playing- field domestic economic environment (figure 5.5). In Australia, farm multifactor productivity (MFP) growth increased following the international price hikes in 1973­74, but then plateaued during the next decade until the reforms from the mid-1980s began to have an effect (figure 5.4). In the 1983­93 period, farm MFP grew at just 1.4 percent per year, low in com- parison to the 4.1 percent growth rate seen in 1993­2000 (Productivity Commis- sion 2005). Similar results are reported in Parham (2004): less than 1.5 percent during 1974­88, then 2.6 percent in 1988­93, and 4.3 percent in 1993­98.26 His estimates show that even that earlier rate of 1.5 percent compares favorably with that for the other sectors of Australia's economy, the MFP of which was well below 1 percent during 1973­93 (and only 1.8 percent in 1993­98). Clearly, farmers are capable not only of surviving without subsidies, but of becoming more productive after their removal--and without a noticeably faster rate of decline in the total number of farmers or farms than occurs with normal economic growth.27 The cases of Australia and New Zealand, in fact, contradict a popular view in the economic growth literature that natural resource abundance (including a comparative advantage in agriculture) is a curse rather than a blessing.28 Australia and New Zealand 249 Figure 5.4. Real Agricultural Total/Multifactor Productivity Growth, Australia and New Zealand, 1927­2004 a. Australia 140 120 100 100 index, 2002­03 80 60 40 20 0 5 7 9 1 3 5 7 9 1 3 5 7 9 1 3 ­7 ­7 ­7 ­8 ­8 ­8 ­8 ­8 ­9 ­9 ­9 ­9 ­9 ­0 ­0 74 76 78 80 82 84 86 88 90 92 94 96 98 00 02 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 hours worked capital input output multifactor productivity growth b. New Zealand 1,000 900 TFP: 4.4% 800 high subsidies 100 700 600 index, 1927 500 TFP: 2.7% 400 TFP: 3.4% TFP: 3.6% 300 200 100 0 27 31 35 39 43 47 51 55 59 63 67 71 75 79 83 87 91 95 99 03 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 Sources: Productivity Commission (2005) and Lattimore (2006). Three important aspects of the agricultural reform success in Australia and New Zealand need to be underlined. One is that it helps if assistance to nonagri- cultural sectors is cut at the same time. In Australia and New Zealand, those other microeconomic and macroeconomic reforms made it easier for farmers to adjust and raise their productivity. This is relevant for the many developing countries that still have industrial protection and behind-the-border restrictions on domestic 250 Distortions to Agricultural Incentives: A Global Perspective Figure 5.5. Real Farmland Prices, New Zealand, 1978­2004 18 1999 $NZ per hectare, thousands 16 14 12 10 8 6 4 2 0 19 8 19 9 19 0 19 1 19 2 19 3 19 4 19 5 19 6 19 7 19 8 19 9 19 0 19 1 19 2 19 3 19 4 19 5 19 6 19 7 19 8 20 9 20 0 20 1 20 2 20 03 04 7 7 8 8 8 8 8 8 8 8 8 8 9 9 9 9 9 9 9 9 9 9 0 0 0 19 arable dairy sheep Source: Lattimore (2006). market activities. For other high-income countries, where manufacturing protec- tion rates are already low and the macroeconomy is well managed, the reduction of high agricultural supports would be more painful unless coupled with adjust- ment assistance measures. Both sets of countries may benefit by following Aus- tralia's lead in the adoption of rural research and development corporations to manage research investments funded by a levy on farmers matched dollar for dol- lar by a grant from the federal government. Introduced in 1989, this innovative funding model, together with a similar model for the generic promotion of Australia's farm products,29 arguably has contributed significantly to the increased international competitiveness of Australian agriculture (see CIE 2003; Productiv- ity Commission 2007). Second, adjustment to agricultural assistance cuts is easier when heterogeneity within the agricultural sector is greater. This can take at least two forms. Hetero- geneity in rates of industry assistance within the sector, as captured by the standard deviation of farm industry NRAs shown near the bottom of tables 5.2 and 5.3, ensures that as high rates of assistance are lowered, resources will find it profitable to move to lightly assisted farm industries. Such transfers are far easier than shifting mobile resources to other sectors, and involve much less reduction in the value of sector-specific assets, particularly farmland. For Australia, the standard deviation was very high when the cuts in farm assistance began in the early 1970s, whereas for a country such as Norway, the support levels are very similar across farm indus- tries, allowing much less scope for adjustment within the sector. Australia and New Zealand 251 The other form of heterogeneity has to do with firm efficiency within each industry. If some farmers are much more efficient than others, then a cut in assis- tance to agriculture typically leads to the less efficient being bought out by the more efficient (who often are also more innovative and export-oriented). The productivity effects of that tendency could be substantial (Melitz 2003; Baldwin 2005; and Long, Raff, and Stahler 2007). In the case of the Australian dairy indus- try, it indeed seems to have been powerful (Harris 2005a). Productivity effects may also help explain the rapid rise in exports relative to domestic sales for a number of Australia's agricultural industries since the 1980s, including beef, canola, wine, and cotton. Four other indicators of the increasing globalization of agriculture in these two countries, two of which also indicate growth in intrasec- toral trade, are: an increasing share of consumption being imported even though the self-sufficiency indicator is rising, and a slightly declining ratio of net exports to exports plus imports of agricultural products (Anderson et al. 2008). The third important aspect of the agricultural reform success in Australia and New Zealand is the scope and quality of public institutions. Both countries had strong institutional arrangements that facilitated the development of policy advice, provided policy coordination, and administered the policy changes. The legal, accounting, and media systems ensured relatively transparent and informed policy debate, and assisted in resolving a host of firm-level complications that arose dur- ing policy implementation. The lesson for countries wishing to replicate the reform experience of Australia and New Zealand, especially in a developing-country con- text, is to not underestimate the importance of strong, pro-market, transparent, corruption-free institutions. Looking ahead in Australia and New Zealand, mean- while, getting domestic policies right in the food safety, natural resource (especially water), and environmental areas, along with securing tariff and subsidy cuts abroad from the WTO's Doha Development Agenda, have the potential to yield even further productivity growth for farmers. But those are stories for the future. Notes 1. Two recent analyses by economic historians point to successful exploitation of natural resources as the key explanation of Australia's relatively high per capita income prior to World War I (Broadberry and Irwin 2007; McLean 2007). 2. However, there is now evidence that the rate of growth of productivity in Australia has since slipped behind that of several OECD countries (Dolman, Parham, and Zheng 2007). 3. New Zealand has around five times the global average of both agricultural land per capita and arable land per worker, and Australia has around 25 times as much (Sandri, Valenzuela, and Anderson 2006). Of course the quality of farmland and associated water, rainfall, sunshine, etc. also matter, but even adjusting for these leave Australia and New Zealand are relatively very well endowed in agricul- tural resources per worker. 4. In terms of population, Australia is somewhat smaller than Argentina and Canada but similar in terms of arable land, while New Zealand is similar in population, arable land, and other agricultural attributes to the Nordic countries. But the antipodean location of Australia and New Zealand 252 Distortions to Agricultural Incentives: A Global Perspective compared with those other countries leads one to expect them to have traded less (and be specialized in more storable and less bulky exports) than these comparator countries, at least prior to East Asia's trade-led growth takeoff. 5. Mining was also an important export earner for Australia in the latter half of the 19th century, but due almost entirely to gold. Gold's share of total exports, 49 percent in 1861, fell to about one-sixth in the 1980s, returning to 28 percent in 1900. During 1961­90, wool and gold together accounted for almost three-quarters of all exports (Butlin 1962). 6. The portion of what follows that is focused on Australia draws on Anderson, Lloyd, and MacLaren (2007). 7. It is thus a generalization of the nominal rate of border protection due, for example, to an import tariff, which is the percentage by which the domestic price is raised above the import unit value. 8. The first major tariffs for the Australian Federation were imposed in 1907. According to the indexes constructed by Carmody (1952), by the 1920s the decade average of the general tariff on Australia's imports of items other than food beverages and tobacco was double that 1907 level, and by the 1930s it averaged 60 percent higher than in the 1920s. The Vernon Committee (Committee of Enquiry 1965) reports averages for tariffs above 12.5 percent for the period from 1938 to 1963: though they dipped somewhat in the late 1940s/early 1950s when import licences became the binding constraint, by the early 1960s they were back to the level of the late 1930s. The annual average level of protection since World War II is indicated by carefully constructed customs duty rates (Anderson, Lloyd, and MacLaren 2007). 9. The GATT came into effect in 1948. Even though Australia and New Zealand were founding signatories to that agreement, they both chose not to join the commitments to cut manufacturing tariffs--out of frustration with the unwillingness of other GATT contracting parties to commit to lowering their agricultural protection rates (Arndt 1965; Snape 1984; and Capling 2001). 10. That pooling was inefficient in at least two senses: it led to excessive volumes of production because producers received the average rather than the marginal price; and because there was little differentiation in terms of quality and variety, producers were discouraged from seeking out niche markets by differentiating their products. Additional stabilization schemes were implemented by indi- vidual states, such as for fresh milk and eggs, and these led to different incentives in the various states. These were possible because the states agreed not to trade across state borders, in contravention of Section 92 of the Constitution, which says there shall be no barriers to interstate trade. 11. This rate is still higher than that for other OECD countries (World Bank 2006). And WTO- bound tariffs average more than twice the applied rates. However, Australia uses nontariff import barriers less frequently than other OECD countries, apart perhaps from antidumping duties (Productivity Commission 2000a, 2000b, 2004). 12. Tariffs on motor vehicle imports fell from 40 to 15 percent over the 1990s and were cut again to 10 percent in 2005. For clothing, the decline over the 1990s was from 55 to 25 percent, and for footwear from 45 to 15 percent, with cuts to 17.5 and 10 percent in 2005, respectively. Further cuts, to as little as 5 percent, are scheduled for 2010 (Productivity Commission 2000a). 13. By contrast, the reductions in manufacturing protection were more systematic: the 1973 across-the-board tariff cut, a tariff review program begun in 1971 by the Tariff Board and subse- quently conducted by its successors (the Industries Assistance Commission, Industry Commission, and Productivity Commission), and the preannounced phased reductions in tariffs on textiles, cloth- ing, footwear, and motor vehicles and parts from 1988. 14. This stabilization scheme operated conservatively for 15 years until the government transferred the power to set the reserve price to growers in 1987. Growers promptly raised that reserve price, which operated on the world market, by 71 percent. Predictably, this encouraged growers to expand wool production and international buyers to reduce purchases (since the Australian Wool Corporation could then stockpile wool and thereby save the buyer the cost of storage). The scheme collapsed in 1991, following which the Australian Wool Corporation had to dispose of 4.75 million bales, at expense to the government and, even more, to woolgrowers (Richardson 2001). 15. Effective assistance to the mining sector is still slightly negative, as it was two decades ago (Industry Commission 1992; Productivity Commission 2004), although that will be less so when the government eases the current quantitative restrictions on exports of uranium and its derivatives. Australia and New Zealand 253 16. In addition, a comprehensive program of review of government regulations at all levels in Australia has been under way since the mid-1990s, with the aim of reducing/removing regulations that unjustifiably impede economic activities (Productivity Commission 2000b). For an early assessment of Australia's domestic microeconomic reforms, see Forsyth (1992, 2000). All Productivity Commis- sion reports on the myriad reforms are downloadable at http://www.pc.gov.au. Recent research on barriers to trade in a wide range of services in almost 40 countries found that services markets in Australia, relative to those in the other countries in the study, are now ranked as either very liberal (banking, distribution services, telecoms, engineering, and professional services) or moderately restrictive (other professional services and maritime services)--see Productivity Commission (2000c). 17. Even as recently as the late 1980s, there were some 10 Commonwealth and more than 50 state SMAs (Piggott 1990). 18. Agricultural SMAs were exempt from the regulation of competition under the Trade Practices Act 1974, although later on they were subject to assessment under the National Competition Policy. 19. In the course of its investigations, the Commission located more than 100 marketing arrange- ments at the state and territory level, though not all of which were SMAs. 20. The evidence suggests that the retail price of milk fell after deregulation as the supermarkets were able to exercise buying power with the milk processors, which, in turn, were able to exert down- ward pressure on the farmgate price that was no longer supported by state governments. 21. Unlike Australia, however, New Zealand continued to receive some nonreciprocal preferential access to the EEC market. 22. The impact of past underpricing of water for agriculture on farm returns has not been incor- porated in the NRA estimates reported in this paper. 23. For detailed descriptions of Australia's adjustment and adjustment assistance programs, see Harris (2005b). 24. In the first half of the 1950s, farm incomes averaged more than 20 percent above nonfarm self- employed incomes and more than twice those of male wage and salary earners. In the next dozen years, though, farm incomes fell to 5 percent below nonfarm self-employed incomes and to just 30 percent above those of male wage and salary earners (McKay 1967). The gap between farm and nonfarm incomes and spending power grew especially rapidly in between 1963 and 1973 (Glau 1971). 25. Australia's dairy industry saw several changes over the 25 years to 2004: the number of dairy farmers more than halved (from 22,000 to 10,000), the average herd size increased 2.5 times (from 85 to 210) average annual yield per cow nearly doubled (from 2,850 to 4,900 liters) and average annual milk production per farm quadrupled, from 0.25 to 1.05 million liters (Productivity Commission 2005; Harris 2005a). In addition, the dairy processing industry has become significantly more produc- tive (Balcombe, Doucouliagos, and Fraser 2007). Overall, the number of farmers in Australia has fallen by more than one-third over the past half-century (ABARE 2007). 26. See also the results since the mid-1970s in Fleming (2007), who shows that MFP growth in Australia's farms enabled producers to cope with the fact that, over the past three decades, the prices they paid for their inputs grew 1.6 percent more per year than the prices they received for their products. 27. Adjustment has been sharper within individual industries, of course, especially those that faced dramatic cuts in subsidies such as Australia's dairy industry (see previous footnote). 28. On this literature in general and its applicability to Latin America, see Lederman and Maloney (2006). See also Anderson (1998). 29. On the basic economics of these supply and demand shifting investments in the presence of spillovers, see, for example, Levin and Reiss (1988). References ABARE (Australian Bureau of Agricultural and Resource Economics). 2007. Australian Commodities Statistics 2006. http://www.abare.gov.au. Anderson, J. E. 1995. "Tariff Index Theory." Review of International Economics 3 (2): 156­73. 254 Distortions to Agricultural Incentives: A Global Perspective Anderson, K. 1978. "On Why Rates of Assistance Differ between Australia's Rural Industries." Australian Journal of Agricultural Economics 22 (2): 99­114. Anderson, K. 1998. "Are Resource-Abundant Economies Disadvantaged?" Australian Journal of Agricultural and Resource Economics 42 (1): 1­23. Anderson, K., and R. Garnaut. 1980. "ASEAN Export Specialisation and the Evolution of Comparative Advantage in the Western Pacific Region." In ASEAN in a Changing Pacific and World Economy, ed. R. Garnaut, 374­412. Canberra: ANU Press. ______. 1987. Australian Protectionism: Extent, Causes and Effects. Boston, London, and Sydney: Allen and Unwin. Anderson, K., and L. A. Jackson. 2005. "GM Crop Technology and Trade Restraints: Economic Impli- cations for Australia and New Zealand." Australian Journal of Agricultural and Resource Economics 49 (3): 263­81. Anderson, K., R. Lattimore, P. Lloyd, and D. MacLaren. 2008. "Distortions to Agricultural Incentives in Australia and New Zealand." Agricultural Distortions Working Paper 09, World Bank, Washington, DC. Anderson, K., P. J. Lloyd, and D. MacLaren. 2007. "Distortions to Agricultural Incentives in Australia Since World War II." The Economic Record 83 (263): 461­82. Anderson, K., W. Martin, and D. van der Mensbrugghe. 2006. "Distortions to World Trade: Impacts on Agricultural Markets and Incomes." Review of Agricultural Economics 28 (2): 168­94. Anderson, K., and B. Smith. 1981. "Changing Economic Relations between Asian ADCs and Resource- Exporting Developed Countries." In Trade and Growth in the Advanced Developing Countries, ed. W. Hong and L. Krause, 293­338. Seoul: Korea Development Institute Press. Anderson, K., and E. Valenzuela. 2007. "Do Global Trade Distortions Still Harm Developing Country Farmers?" Review of World Economics/Weltwirtschaftliches Archiv 143 (1): 108­39. ______. 2008. Global Estimates of Distortions to Agricultural Incentives, 1955 to 2007. Core database at http://www.worldbank.org/agdistortions. Arndt, H. W. 1965. "Australia ­ Developed, Developing or Midway?" The Economic Record 41 (95): 418­40. Balcombe, K., H. Doucouliagos, and I. Fraser. 2007. "Input Use, Output Mix and Industry Deregulation: An Analysis of the Australian Dairy Manufacturing Industry." Australian Journal of Agricultural and Resource Economics 51 (2): 121­35. Balderstone, J. S., et al. 1982. "Agricultural Policy: Issues and Options for the 1980s." Working Group Report to the Minister for Primary Industry, Australian Government Publishing Service, Canberra. Baldwin, R. E. 2005. "Heterogeneous Firms and Trade: Testable and Untestable Properties of the Melitz Model." NBER Working Paper 11471, National Bureau of Economic Research, Cambridge, MA. Broadberry, S., and D. A. Irwin. 2007. "Lost Exceptionalism? Comparative Income and Productivity in Australia and the UK, 1861­1948." The Economic Record 83 (262): 262­86. Butlin, N. G. 1962. Australian Domestic Product, Investment and Foreign Borrowing, 1861­1938/39. Cambridge, U.K.: Cambridge University Press. Capling, A. 2001. Australia and the Global Trade System: From Havana to Seattle. Cambridge, U.K.: Cambridge University Press. Card, D., and R. B. Freeman. 2004. "What Have Two Decades of British Economic Reform Delivered?" In Seeking a Premier Economy: The Economic Effects of British Economic Reforms, 1980­2000, ed. D. Card, R. Blundell, and R. B. Freeman, 9­20. Chicago: University of Chicago Press. Carmody, A. T. 1952. "The Level of the Australian Tariff: A Study in Method," Yorkshire Bulletin of Economic and Social Research 4: 51­65, January. CIE (Centre for International Economics). 2003. The Rural Research and Development Corporations: A Case Study of Innovation, Canberra: Centre for International Economics. Committee of Enquiry (The Vernon Committee). 1965. Report of the Committee of Enquiry. Melbourne: Wilke and Co. Corden, W. M. 1984. "Booming Sector and Dutch Disease Economics: Survey and Consolidation." Oxford Economic Papers 36 (3): 359­80. Australia and New Zealand 255 Dalziel, P., and R. Lattimore. 2004. The New Zealand Macroeconomy: Striving for Sustainable Growth with Equity. Melbourne: Oxford University Press. Deardorff, A. V. 1984. "An Exposition and Exploration of Krueger's Trade Model." Canadian Journal of Economics 5 (4): 731­46. Diewert, W. E., and D. Lawrence. 2006. "Measuring the Contributions of Productivity and Terms of Trade to Australia's Economic Welfare." Consultancy Report, Productivity Commission, Canberra. Dolman, B., D. Parham, and S. Zheng. 2007. "Can Australia Match US Productivity Performance?" Staff Working Paper, Productivity Commission, Canberra. Dowrick, S. 2001. "The Australian Productivity Miracle." In Creating an Environment for Australia's Growth, ed. J. Nieuwenhuysen, P. J. Lloyd and M. Mead, 33­49. Cambridge, U.K. and Sydney: Cambridge University Press. Edwards, G. W. 1987. "Agricultural Policy Debate: A Survey." The Economic Record 63: 129­43. ______. 2006. "Fifty Years of Agricultural Policy." Presented at the 50th Annual Conference of the Australian of the Australian Agriculture and Resource Economics Society, Sydney, 8­10 February. Evans, L., A. Grimes, B. Wilkinson, and D. Teece. 1996. "Economic Reform in New Zealand: The Pursuit of Efficiency." Journal of Economic Literature 34 (4): 1856­1902. Fleming, E. 2007. "Use of the Single Factorial Terms of Trade to Analyse Agricultural Production." Australian Journal of Agricultural and Resource Economics 51 (2): 113­19. FSANZ (Food Standards Australia New Zealand). 2007. "Assessing the Safety of New Foods and Technologies." FSANZ, http://www.foodstandards.gov.au/foodmatters/gmfoods/ assessingthesafetyof387.cfm. Forsyth, P., ed. 1992. Microeconomic Reform in Australia. London and Sydney: Allen and Unwin. ______. 2000. "Microeconomic Policies and Structural Change." In The Australian Economy in the 1990s, ed. D. Gruen and S. Shrestha. Sydney: Reserve Bank of Australia. Gardner, B. 2009. "The United States and Canada." Chapter 4 in this volume. Glau, T. E. 1971. "The Cost­Price Squeeze on Australia Farm Income." Australian Journal of Agricultural Economics 15 (1): 1­19. Godden, D. 1997. Agricultural and Resource Policy: Principles and Practice. Melbourne: Oxford University Press. Harris, D. 2005a. Industry Adjustment to Policy Reform: A Case Study of the Australian Dairy Industry, RIRDC Publication 05/110, Rural Industries Research and Development Corporation, Canberra. ______. 2005b. Rural Industry Adjustment to Trade Related Policy Reform, RIRDC Publication 05/173, Rural Industries Research and Development Corporation, Canberra. Harris, S. F., J. C. Crawford, F. H. Gruen, and N. Honan, 1974. The Principles of Rural Policy in Australia: A Discussion Paper, Canberra: Australian Government Publishing Service. Honma, M., and Y. Hayami. 2009. "Japan, Republic of Korea and Taiwan, China." Chapter 2 in this volume. Industry Commission. 1991. "Statutory Marketing Arrangements for Primary Products." Report 10, AGPS, Canberra. Industry Commission. 1992. Annual Report. Canberra: AGPS. James, S., and K. Anderson. 1998. "On the Need for More Economic Assessment of Quarantine/SPS Policies." Australian Journal of Agricultural and Resource Economics 42 (4): 525­44. Josling, T. 2009. "Western Europe." Chapter 3 in this volume. Krueger, A. O. 1977. "Growth, Distortions and Patterns of Trade Among Many Countries." Princeton Studies in International Finance Section 40: 1­45. Lattimore, R. 2006. "Farm Policy Reform Dividends." Paper presented to the North American Agrifood Market Integration Consortium Meetings, Calgary, 1­2 June. Leamer, E. E. 1987. "Paths of Development in the Three-Factor, n-Good General Equilibrium Model." Journal of Political Economy 95 (5): 961­99. Lederman, D., and W. Maloney, eds. 2006. Natural Resources: Neither Curse nor Destiny. Stanford, CA: Stanford University Press for the World Bank. Levin, R. C., and P. C. Reiss. 1988. "Cost-Reducing and Demand-Creating R&D With Spillovers." Rand Journal of Economics 19 (4): 538­56. 256 Distortions to Agricultural Incentives: A Global Perspective Lloyd, P. 2008. "100 Years of Tariff Protection in Australia." Australian Economic History Review 48 (2): 99­145. Long, N. V., H. Raff, and F. Stahler. 2007. "The Effects of Trade Liberalization on Productivity and Welfare: The Role of Firm Heterogeniety, R&D and Market Structure." Economics Discussion Paper 0710, University of Otago, Dunedin. MacLaren, D. 1992. "The Political Economy of Agricultural Policy Reform in the European Commu- nity and Australia." Journal of Agricultural Economics 43 (3): 424­39. Maddison, A. 2003. The World Economy: Historical Statistics. Paris: OECD Development Centre. Martin, W. 1990. "Public Choice Theory and Australian Agricultural Policy Reform." Australian Journal of Agricultural Economics 34 (2): 189­211. Mauldon, R. 1990. "Price Policy." In Agriculture in the Australian Economy, ed. D. B. Williams, 310­28. Sydney: Sydney University Press. McKay, D. H. 1967. "The Small-Farm Problem in Australia." Australian Journal of Agricultural Economics 11 (2): 33­52. McLean, I. 2007. "Why Was Australia So Rich?" Explorations in Economic History 44 (4): 635­56. Melitz, M. 2003. "The Impact of Trade on Intra-industry Reallocations and Aggregate Industry Productivity." Econometrica 71 (6): 1695­725. OECD. 2008. OECD PSE-CSE Database (Producer and Consumer Support Estimates, OECD Database 1986­2007). Organisation for Economic Co-operation and Development. http://www.oecd.org/ document/59/0,3343,en_2649_33727_39551355_1_1_1_1,00.html. Parham, D. 2004. "Sources of Australia's Productivity Revival." The Economic Record 80 (249): 239­57. Parham, D., T. Cobbold, R. Dolamore, and P. Roberts. 1999. Microeconomic Reforms and Australian Productivity: Exploring the Links. Canberra: Productivity Commission. Piggott, R. 1990. "Agricultural Marketing." In Agriculture in the Australian Economy, ed. D. B. Williams, 287­309. Sydney: Sydney University Press, 1990. Productivity Commission. 2000a. Review of Australia's General Tariff Arrangements. Canberra: Produc- tivity Commission. ______. 2000b. Regulation and its Review 1999­00. Canberra: Productivity Commission. ______. 2000c. Trade and Assistance Review 1999­2000. Canberra: Productivity Commission. ______. 2004. Trade and Assistance Review 2003­04. Canberra: Productivity Commission. ______. 2005. "Trends in Australian Agriculture." Research Paper, Productivity Commission, Canberra. ______. 2006. Rural Water Use and the Environment: The Role of Market Mechanisms. Canberra: Productivity Commission. ______. 2007. "Public Support for Science and Innovation." Research Report, Canberra. Rayner, A. and R. Lattimore. 1991. "New Zealand." In Liberalising Foreign Trade, ed. D. Papageorgiou, M. Michaely, and A. Choksi. Cambridge, MA: Basil Blackwell. Richardson, B. 2001. "The Politics and Economics of Wool Marketing, 1950­2000," Australian Journal of Agricultural and Resource Economics 45 (1): 95­115. Sandri, D., E. Valenzuela, and K. Anderson. 2006. "Compendium of National Economic and Trade Indicators by Region, 1960 to 2004." Agricultural Distortions Working Paper 01, World Bank, Washington, DC. http://www.worldbank.org/agdistortions Sandry, R. and R. Reynolds, eds. 1990. Farming without Subsidies: New Zealand's Recent Experience. Wellington: GP Books. Sieper, E. 1982. Rationalising Rustic Regulation. Sydney: Centre for Independent Studies. Statistics New Zealand. 2006. Productivity Statistics 1988­2005. Wellington: Statistics New Zealand. Snape, R. H. 1984. "Australia's Relations with GATT." The Economic Record 60 (168): 16­27. Vitalis, V. 2005. "The New Zealand Experience." In Subsidy Reform and Sustainable Development: Economic, Environmental and Social Aspects, ed. Lori Ridgeway and others. Paris: OECD. Vousden, N. 1990. The Economics of Trade Protection. Cambridge, U.K. and New York: Cambridge University Press. World Bank. 2006. World Development Indicators. Washington, DC: World Bank. Part III EVOLUTION OF DISTORTIONS IN EMERGING ECONOMIES 6 Eastern Europe and Central Asia Kym Anderson and Johan Swinnen* In a recent survey of European economic growth since 1950, Crafts and Toniolo (2008) conclude that incentive structures have been a crucial influence on compara- tive growth rates of the economies of Eastern and Western Europe. Predating that, a 2006 report on trade performance and policies in Eastern Europe and Central Asia included as one of its key recommendations the need to reduce the mean and vari- ance of the tariff equivalents of trade barriers, in particular the need to reduce uni- laterally governments' antiexport bias, especially in countries exporting primary products (Broadman 2006). To carry out such reform in Europe's transition economies efficiently and effectively--and to see how recent policies align with those of the European Union (EU)--requires better information on the extent of reform during the past two decades and of current policy influences on incentives within and between sectors. Immediately prior to their transition to market economies, policies in the region greatly distorted producer and consumer incen- tives, especially for agricultural products. Those distortions have been reduced substantially in several countries, but large variations remain across the region and distortions appear to be increasing again in some countries. Now is thus an oppor- tune time to examine how policies affecting agriculture are evolving in the region, particularly as part of the adjustment to EU accession for 10 of the countries. *The authors are grateful for data assistance and very helpful discussions with Organisation for Economic Co- operation and Development (OECD) Secretariat staff, especially Olga Melyukhina; for the distortions estimates provided by authors of the focus country case studies; and for assistance with spreadsheets by Johanna Croser, Esteban Jara, Marianne Kurzweil, Signe Nelgen, Francesca de Nicola, Damiano Sandri, and Ernesto Valenzuela. The working paper version of this chapter (Anderson and Swinnen 2009) contains additional background material and appendix tables. This chapter draws on the introductory and country chapters in Anderson and Swinnen (2008), with data updated using Anderson and Valenzuela (2008). 259 260 Distortions to Agricultural Incentives: A Global Perspective To assist that process, the study presented here assesses the changing landscape of agricultural protection or taxation patterns in Europe and Central Asia. It is based on a sample of 18 countries, including 11 Central and Eastern European (CEE) countries--the 10 new EU members (Bulgaria and Romania joined in January 2007 and the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, the Slovak Republic, and Slovenia joined in May 2004), plus Turkey--and seven Commonwealth of Independent States (CIS) countries (Kazakhstan, the Kyrgyz Republic, the Russian Federation, Tajikistan, Turkmenistan, Ukraine, and Uzbekistan). Together, these 18 countries accounted in 2000­04 for 89 percent of the region's agricultural value added, 91 percent of its population, and 95 percent of total gross domestic product (GDP). Some key characteristics of those economies are shown in table 6.1, drawn from the detailed compendium of indi- cators provided in the Sandri, Valenzuela, and Anderson (2007). Analyses of polit- ically feasible agricultural subsidy and trade policy reforms, and of policy options for coping with structural changes such as the recent boom in energy raw material prices that has intersectoral Dutch disease effects, need to be based on a clear understanding of the recent and current extent of policy interventions and the politico-economic forces behind their evolution. This study thus also seeks to bet- ter understand the political economy of distortions to agricultural incentives in Europe and Central Asia. With that better understanding, the study's third purpose is to explore prospects for further reducing distortions to agricultural incentives and their implications for agricultural competitiveness and trade of the different countries in the region, including those that have recently joined the EU. The great diversity within the group of European and Central Asian countries-- in terms of relative resource endowments and comparative advantages; stages of development and transition; agricultural and trade policy regimes; and member- ships of the EU, World Trade Organization (WTO), OECD, and regional trading agreements--make the set of countries chosen a rich sample for comparative study. Turkey and the CEE countries that are now EU members differ substantially from the rest of the former Union of Soviet Socialist Republics (USSR) that are now members of the CIS in that they have higher per capita income (three-quar- ters of the global average, compared with one-third for the CIS) and higher popu- lation density (half the global average land per worker and 70 percent of the global average agricultural land per capita, compared with 3.4 and 2.5 times, respectively, for the CIS). Growth and Structural Changes during Transition Before examining policy changes, it is helpful to review the economic growth and intersectoral changes that have taken place in Europe's various transition economies over the past two decades. The initial years of transition from central Table 6.1. Key Economic and Trade Indicators, Eastern European and CIS Countries, 2000­04 National relative to Percent of world: world (world 100) Agricultural RCA,a Trade Total Agricultural GDP land agriculture specialization Gini Population GDP GDP per capita per capita and food indexb Povertyc indexd Slovenia 0.03 0.07 0.04 216 32 52 0.68 0 -- Czech Republic 0.16 0.22 0.19 135 52 61 0.44 0 26 Hungary 0.16 0.20 0.14 122 72 90 0.40 0 27 Estonia 0.02 0.02 0.03 102 78 199 0.38 1 36 Poland 0.62 0.57 0.47 93 57 105 0.39 0 34 Slovak Republic 0.09 0.07 0.09 92 57 57 0.50 0 -- Lithuania 0.06 0.04 0.08 80 125 176 0.21 1 36 Latvia 0.04 0.03 0.03 76 132 364 0.51 0 38 Turkey 1.12 0.62 1.97 55 70 131 0.09 3 44 Romania 0.35 0.15 0.49 41 84 74 0.06 1 31 Bulgaria 0.13 0.05 0.15 39 86 143 0.37 0 29 CEE sample 2.75 2.05 3.67 74 70 98 0.09 1 37 Russian Federation 2.34 1.10 1.58 47 186 53 0.46 0 40 Kazakhstan 0.24 0.08 0.18 33 1737 76 -- 1 34 Ukraine 0.78 0.13 0.46 17 107 112 -- 0 28 Turkmenistan 0.07 0.01 0.06 18 881 92 -- 5 41 Uzbekistan 0.41 0.03 0.27 8 134 -- -- 0 37 Kyrgyz Republic 0.08 0.00 0.05 6 268 390 -- 0 30 Tajikistan 0.10 0.00 0.03 4 85 192 -- 7 33 CIS sample 4.02 1.37 2.62 34 270 -- 0.02 0 37 Other CEE/Central Asia 0.64 0.19 0.61 29 82 166 0.41 1 -- All CEE/Central Asia 7.43 3.60 6.90 48 179 -- 0.06 0 37 Source: Sandri, Valenzuela, and Anderson (2007), compiled from World Bank (2007). Note: -- not available. 261 a. RCA is the index of revealed comparative advantage share of agriculture and processed food in national exports as a ratio of that sector's share of global exports b. Primary agriculture trade specialization index (X M) (X M), 2000­02 (world average 0). c. Percentage of population living on US$1/day, from Chen and Ravallion (2007). d. Gini index for the most recent year during 2000­04, from Chen and Ravallion (2007). 262 Distortions to Agricultural Incentives: A Global Perspective planning to a more market-based economy saw production fall in the majority of sectors, before it recovered at varying rates from the mid-1990s. Real GDP for the Eastern Europe and Central Asia region as a whole fell by almost 6 percent per year during 1990­94. The decline for the CEE sample was only 0.6 percent, while for the CIS sample it was 11 percent and for the residual nonstudied countries of the CIS 12 percent. By contrast, annual GDP growth over 1995­2004 averaged 2.7 percent: the CIS sample was slowest (2.2 percent), the CEE sample somewhat higher at 3.2 percent, and the residual enjoyed 5.1 percent. Within those economies, agricultural value added measured at constant prices appears to have declined less rapidly than nonagricultural GDP in the early years of transition, but also to have grown less rapidly in the subsequent decade. The domestic terms of trade (the prices of their outputs relative to the prices of pur- chased inputs) apparently fell even more for farmers than for nonfarmers, how- ever, because agriculture's share of GDP measured in current prices declined even in the early transition period. Unlike in the central planning period, this did not allow faster industrialization but rather an expansion in the services sector, which increased from less than half the economy prior to 1993 to two-thirds by 2004. The halving of agriculture's share of GDP in Europe and Central Asia between 1992 and 2004 was accompanied by only a one-quarter decline in agriculture's share of employment, according to Food and Agriculture Organization (FAO) sta- tistics, which are not always consistent with national data because of definitional differences. In all three subgroups of countries, the share of the latter averaged three times the former by 2004, or five times in the case of the CEE-8 countries that joined the EU in 2004. This suggests much lower labor productivity on farms than in other employment. The share of farm and food products in total merchandise exports also has fallen, by as much as half in some countries in the region. When expressed as a ratio of that share for the world as a whole (an agricultural-revealed comparative advantage index), it becomes evident that most countries of the region have lost comparative advantage in farm products over the transition period. That index varies greatly across the region though, from a low of less than 0.5 for mineral-rich Russia and densely populated Slovenia to more than 3 for Latvia and the Kyrgyz Republic. The region as a whole, however, has become more open as a consequence of moving from a planned to a market economy, notwithstanding the continuation of numerous barriers to trade. A common indicator is the value of goods and services expressed as a percentage of GDP. For most countries, that percentage is now above the average for Western European countries (37 percent in 2004), with several countries approaching 60 percent by 2004.1 With this as background, the evolution of policy under Communism is briefly reviewed, followed by an examination of how sectoral and trade policies have Eastern Europe and Central Asia 263 changed in Europe and Central Asia in response to, or as contributors to, the macroeconomic and structural changes. Quantifying the Distortions to Agricultural Incentives The main focus of the present study's methodology is government-imposed dis- tortions that create a gap between domestic prices and what they would be under free markets. Since it is not possible to understand the characteristics of agricul- tural development with a sectoral view alone, the project's methodology not only estimates the effects of direct agricultural policy measures (including distortions in the foreign exchange market), but also generates estimates of distortions in nonagricultural sectors for comparative evaluation. Specifically, nominal rates of assistance (NRAs) for farmers are computed, including any input subsidies and non-product-specific (NPS) forms of assistance or taxation. It also generates a production-weighted average NRA for nonagricultural tradables, for comparison with that for agricultural tradables via the calculation of a relative rate of assis- tance (RRA; see Anderson et al. 2008a, 2008b). This approach is not well suited to analysis of policies of planned economies prior to their reform era, as prices then played only an accounting function and currency exchange rates were enormously distorted. During their reform era, however, starting in 1992, the price compari- son approach provides as valuable a set of indicators for European and Central Asian countries as for other market economies of distortions to incentives for farm production, consumption, and trade, and of the income transfers associated with interventions. While most of the focus is on agricultural producers, the extent to which consumers are taxed or subsidized is also considered. To do so, a consumer tax equivalent (CTE) is calculated by comparing the price consumers pay for their food and the international price of each food product at the border. Differences between the NRA and the CTE arise from distortions in the domestic economy that are caused by transfer policies and taxes or subsidies that cause the prices paid by consumers (adjusted to the farmgate level) to differ from those received by producers. To obtain dollar values of farmer assistance and consumer taxation, estimates of NRAs are multiplied by the gross value of production at undistorted prices to obtain an estimate in current U.S. dollars of the direct gross subsidy equivalent (GSE) of assistance to farmers. These GSE values are calculated in constant dollars, and are also expressed on a per-farm-worker basis. They (and their equivalent on the consumption side) can be added up across products for a country, and across countries for any or all products, to get regional aggregate transfer estimates for the studied economies. 264 Distortions to Agricultural Incentives: A Global Perspective To keep the task manageable, the sample of countries for which empirical esti- mates are provided below is limited to the 10 Central and Eastern European coun- tries that joined the EU in 2004 or 2007 plus Turkey and the three biggest CIS economies (Kazakhstan, Russia, and Ukraine).2 Nonquantitative policy assess- ments are undertaken for the other economies of Central Asia. Reliable price data are available only from 1992 to 2005 or 2007 or, in the case of Kazakhstan, for just 2000­04. The worst of the exchange rate distortions in the formerly planned economies were removed in the early 1990s, prior to the start of the period under study here. Since there were no reliable indicators of any remaining secondary market price for foreign currency, this study follows the practice of the OECD Secretariat in using official exchange rates in making price comparisons.3 Additionally, it is not possible to estimate the sectoral assistance equivalent of soft credits provided for some farms (and other enterprises). Throughout, there is no attempt to assess whether some interventions may have been warranted on national economic wel- fare grounds because of the presence of externalities, or failures in the markets for such things as land and credit, or policy failures such as underinvestment in public infrastructure in rural areas.4 The Communist era Incentives for agricultural producers and food consumers were massively dis- torted under Communist central planning, which was imposed from the 1920s until 1991 in the former Soviet Union and from the 1950s until the fall of the Berlin Wall in 1989 in Central and Eastern Europe. Distortions included collective farm property rights; centrally controlled organization of production allocation, processing, input provision, and marketing; price setting unrelated to demand- supply conditions (leading to rationing); and state-controlled trading and exchange rate systems. Land and farms were put under central planning and, in most countries (with the exception of Poland and former Yugoslavia), farming was forcibly organized in collective and state farms. In the former Soviet Union, this collectivization process and the associated forced migration (and worse) of many landowners and farmers contributed to massive incidence of hunger and death before World War II. From Lenin to Stalin and through most of the Khrushchev regime, agriculture was heavily taxed, and capital was drained from an impoverished countryside to finance urban industrial growth (Ellman 1988).5 All of this changed at the end of the Khrushchev regime, particularly under Brezhnev, when the leadership of the USSR decided to increase agricultural pro- duction, with a strong emphasis on livestock, a policy also followed by many of the Eastern European countries of the Soviet bloc (Liefert and Swinnen 2002). From Eastern Europe and Central Asia 265 the mid-1950s onward, and especially in the 1970s and 1980s, large amounts of support and investment were directed to agriculture. By 1980, almost 30 percent of total Soviet investment was going into agriculture (Gray 1990). At the same time, consumer prices were set low and producer prices high, with the gap covered by direct subsidies to processing and trading companies or by soft budget con- straints. Consequently, from 1970 to 1990, livestock herds and output in these countries grew by between 40 and 60 percent. The rise in feed requirements for the growing herds stimulated the crop sector. In the late 1980s, the average annual output of feed grain in Hungary and Poland increased by half and one-quarter, respectively, compared with output in the late 1960s. In the USSR, livestock feed requirements were so high that the country also became a substantial importer of feed commodities. By 1990, per capita consumption of livestock products and foodstuffs in gen- eral compared favorably with many OECD countries, even though per capita incomes in Central and Eastern Europe were much lower than the OECD average. This "achievement" came at a cost: large state subsidies, to both producers and consumers, were necessary to maintain the high levels of production and con- sumption. For example, by the end of the 1980s, direct budgetary subsidies to the agriculture and food economy were about 10 percent of GDP in the USSR and between 5 and 10 percent of GDP in most CEE countries. The bulk of these subsi- dies went to the livestock sector.6 Calculating the net transfers to farmers and to consumers under the Commu- nist regime is very difficult because of the large number of distortions caused by the state regulation of prices, production, consumption, exchange rates, and mar- keting organizations, along with the indirect nature of some of the subsidies. While it is generally true that producers of farm products were strongly subsidized by price setting toward the end of the Communist regime (in sharp contrast to the 1930s, when farmers were highly discriminated against), the complexity of the distortions led sometimes to offsetting effects. For example, while agricultural producers in the latter 1980s were supported by high output prices and low input prices, at the same time overvalued exchange rates effectively taxed agricultural (and other) exporters. Correcting for this overvaluation leads to significantly lower protection indicators. In addition, agriculture was not alone in being subsi- dized, as most (heavy) industry was also subsidized or at least protected from import competition. Available fragments of empirical evidence indicate that, in aggregate and real terms, there was substantial net subsidization of agriculture rel- ative to all other sectors as a group, although much more so for livestock produc- ers than for grain and oilseed farmers. This might suggest food consumers were taxed substantially, but under the central planning system wholesalers were told to sell their food to retailers below their production costs, for which they received 266 Distortions to Agricultural Incentives: A Global Perspective state subsidies. Overvalued exchange rates, which effectively taxed exports and subsidized imports, also lowered domestic consumer prices of tradable products. However, by restricting foreign imports and regulating trade, the Communist regime prevented its consumers from accessing higher-quality food products. Huffman and Johnson (2004) estimate that these welfare losses were equivalent to 50 percent to 75 percent of the direct subsidy benefits of consumers under the Communist regime. The reform era After 1989, the CEE-8 countries (those Central and Eastern European countries that joined the EU in 2004: the Czech Republic, Estonia, Hungary, Latvia, Lithua- nia, Poland, the Slovak Republic, and Slovenia) moved first and most rapidly toward market-based systems. Reforms in the Balkan countries, such as Romania and Bulgaria, were initially half-hearted and involved many inconsistencies during most of the 1990s, with government interventions continuing to heavily distort incentives. In the large CIS countries (Kazakhstan, Russia, and Ukraine), govern- ments continued important controls of the agricultural economy through inter- ventions such as regional trade controls, input supply controls, and the continuation of soft budget constraints. While the Kyrgyz Republic liberalized relatively quickly, other Central Asian countries moved more slowly, and some have undertaken far less reform and liberalization. Major controls are still in place in countries such as Turkmenistan and Uzbekistan. International trade was strongly regulated under the centrally planned system. Communist countries were integrated in the Council of Mutual Economic Assistance (CMEA) system, a planned, intercountry trading regime that traded mainly with other Communist countries. (One could think of the CMEA as the international version of a domestic centrally planned economy.) The CEE countries were less integrated than the Soviet republics, but still a large part of their trade volume went through the CMEA system. When the CMEA system collapsed in the early 1990s with the liberalization of the macroeconomy and of trade policies, impor- tant changes in trade and financial flows resulted. Trade liberalization reinforced the reallocation of production activities caused by the abolishment of central planning. Traditional international production allocations were no longer possi- ble when trade had to be financed by hard currencies and when inputs were accounted for at real costs. It also allowed the importation of high-quality Western agricultural products that previously had been restricted. At the same time, the liberalization of exchange rates removed discrimination against the sectors pro- ducing tradables. Trade liberalization led to a major international reorganization of production activities. Initially this had a very negative impact on the region's producers, as the Eastern Europe and Central Asia 267 traditional export markets dwindled due to a lack of hard currency and because Western countries remained closed to the region's agricultural exports. At the same time, the reduction of import constraints opened regional markets to imports from the West. Combined, these forces caused a worsening of the region's agricultural trade balance in the first half of the 1990s. Later on, however, agri- food trade intensified and growing exports (also to Western markets) contributed to the region's recovery. An important development was the shift from centrally imposed extreme specialization (for example, dairy production in the Baltics and cotton production in Central Asia) to more diversified production systems and less dependence on single commodities in those countries. Trade effects were only part of the international effects in the agri-food sys- tems. Possibly even more important was the massive inflow of foreign direct investment to food processing industries, which contributed to a major restruc- turing and to improvements in food quality and productivity enhancements and investments in agriculture (Dries and Swinnen 2004). Most recently, the wave of foreign investments in the retail sector caused further restructurings of the agri- food system, with important implications for both producers and consumers (Dries, Reardon, and Swinnen 2004). Progress in market reforms is not always correlated with the extent of distor- tions. On the one hand, Slovenia, which was a front runner in liberalization and developing a market economy, has a very high level of farm producer support that in 2004 was well above the average EU-15 rate. On the other hand, much slower reformers such as Bulgaria, Kazakhstan, and Ukraine have much lower--even negative--NRAs. Turkey, which has never been under Communist rule but nonetheless had a highly state-controlled food system (including price regulations and state processing companies), especially prior to the 1990s, maintained among one of the highest levels of support within Europe and Central Asia during 2004­ 05, despite the fact that there was a major policy reform after 2000, including a shift in assistance from market price support toward direct payments. NRAs to agriculture during transition As domestic markets and trade and currency exchange regimes were liberalized in the early 1990s in Europe and Central Asia, farm output declined dramatically. This was mainly the result of nominal input prices increasing much more strongly than output prices. Industrial output declined by a similar order of magnitude, while the services sector, which had been severely constrained under the Commu- nist system (at least as a stand-alone set of activities distinct from state-owned farm and industrial enterprises), grew rapidly after transition began. Beginning in the early 1990s, many trade and price distortions were removed throughout the region. Price, exchange rate, and trade policies were all liberalized; 268 Distortions to Agricultural Incentives: A Global Perspective Figure 6.1. NRAs to Agriculture, Eastern European Countries, 1992­2007 60 40 20 percent 0 20 40 60 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 Turkey CEE-10 Russian Federation Ukraine Source: Anderson and Valenzuela (2008), updated from estimates reported in Anderson and Swinnen (2009). subsidies were cut; hard budget constraints were introduced; property rights were privatized; and production decisions were shifted to companies and households. One consequence was that, on average, support to agriculture fell to very low lev- els in the early 1990s (as it did also for industrial production). Between 1992 and 1995, nominal assistance to agriculture averaged just 12 percent in the CEE-10 and was less than zero in Bulgaria and the three Baltic nations--as it was in Russia and Ukraine. By contrast, in Turkey, where nominal assistance averaged just 5 per- cent during 1986­89, the NRA rose to an average of 15 percent in 1992­95 and 25 percent in 1996­99 (figure 6.1 and table 6.2). Changes in policy in the region, and hence in rates of agricultural assistance, have not been smooth. Rather, they have been characterized by stop-go phases and sometimes even reversals of previous reforms, as is apparent from figure 6.1. Despite that heterogeneity of experiences, one can identify a couple of general phases in the policy changes. Following its initial collapse, support to agriculture increased during the mid- 1990s in some countries. In CEE, this was driven by the explicit introduction of new support policies, while in Russia it reflected primarily exchange rate develop- ments, which, in the presence of institutional constraints, constrained the pass- through of border prices to farmgate prices and pushed assistance rates up to high levels. Eastern Europe and Central Asia 269 Table 6.2. NRAs to Agriculture,a Eastern European and CIS Focus Countries, 1992­2007 (percent) 1992­95 1996­99 2000­03 2004­07d Bulgaria 19 11 0 7 Czech Republic 20 19 27 24 Estonia 14 20 20 23 Hungary 19 18 34 20 Latvia 15 30 36 28 Lithuania 19 29 32 29 Poland 10 24 7 32 Romania 22 29 53 50 Slovak Republic 28 26 30 21 Slovenia 64 79 80 31 CEE-10 12 22 24 31 Turkey 15 25 22 30 Russian Federation 8 25 13 19 Ukraine 21 1 11 2 Kazakhstan -- -- 0 5 All focus countries: Unweighted averageb 6 24 25 22 Weighted averagea 3 22 16 25 Dispersionc 26 21 26 14 Source: Anderson and Valenzuela (2008), updated from estimates reported in Anderson and Swinnen (2008). Note: -- not available. a. Weighted average for each country, including product-specific output and input distortions and NPS assistance as well as authors' guesstimates for noncovered farm products, with weights based on gross value of agricultural production at undistorted prices. b. The unweighted average is the simple average across the 14 countries of their national NRA (production-weighted) average NRAs. c. Dispersion is a simple four-year average of the annual standard deviation around a weighted mean of the national agricultural sector NRAs. d. Final column refers just to 2004­05 for Russia and Ukraine and 2004 for Kazakhstan; and the CEE values assume the NRA for each product is the same as for the EU-25 in 2004­06 and for EU-27 in 2007, such that the differences across CEE countries are due to differing national product weights. The increase in support started first in Central Europe, where, after the radical liberalization in the early 1990s, political pressures induced governments to rein- troduce a series of measures. NRAs increased from close to zero in 1992 to 20 to 30 percent (or more) in the second half of the 1990s, but then they stabilized in the lead-up to EU accession in 2004. Between 2000 and 2003, the average rate of assis- tance to agriculture in the CEE-10 was just under 25 percent (figure 6.1), slightly less than half the rate of assistance (including from programs somewhat decou- 270 Distortions to Agricultural Incentives: A Global Perspective pled from production) provided to farmers in the EU-15 at that time (see Josling 2008). Further East, two economic changes in the late 1990s had major impacts on agricultural incentives. First, the Russian financial crisis and the associated deval- uation of the ruble, in the presence of imperfect pass-through, caused a strong decline in the estimated rates of assistance to agriculture. This macroeconomic correction brought down estimated assistance rates dramatically. Second, in the 2000s, the hike in world energy and mineral prices and general economic growth improved many CIS governments' budgetary situations. The improvement in the budgetary situation, in turn, has increased the level of support to agriculture. In Russia, for example, the government announced that agriculture would be a pri- ority area for additional funding in 2005. Not all the additional funding is to go to subsidies, as some governments have plans to spend considerably on infrastruc- ture and quality upgrading in agriculture. Rural incomes have improved because of better (and timely) payments of farm workers' wages and pensions to farm and rural workers, and because of improved rural services. The combination of all these developments led to a somewhat lower weighted average NRA for agriculture in the region as a whole in the four-year period from 2000 to 2004 than in the period immediately before: 16 percent during 2000­03, compared with 22 percent in 1996­99 (table 6.2). In Russia, the level of average support fell even more (from 25 to 13 percent). However, supports rose again in 2004 and 2005, including in countries that have since joined the EU, before drop- ping again as international food prices rose in 2007. Meanwhile, the NRA moved closer to zero in Ukraine in 2005, but is probably still very negative in the rest of Central Asia. There is thus a very wide dispersion in average NRAs across coun- tries in the region, from very high levels in the highest-income country (Slovenia) to continuing negative levels in the poorest countries of Bulgaria, Kazakhstan, and Ukraine (figure 6.2). There are major differences in distortions across commodities, too. In the 1980s, virtually all commodities were supported, albeit some more than others. Now, some commodities are now taxed in the CIS (table 6.3). For example, by 2000­03, sugar, poultry, and milk were the most highly protected commodities in the CEE-10, while grains, beef, and pork were the least assisted. Meanwhile, in Russia and Ukraine, the range is even more extreme, from high positive assistance to livestock and sugar to high negative assistance to the production of the key feed inputs into livestock (coarse grains and oilseeds). Sunflower seeds, Russia's most commonly produced and traded oilseed, was in fact the only consistently exported commodity through the transition period. The case of Kazakhstan in 2004 was even starker, where import-competing producers were highly assisted while exporting industries had to endure negative assistance, such that even though the average NRA was close to zero, a strong antiagricultural trade bias prevailed. Eastern Europe and Central Asia 271 Figure 6.2. NRAs to Agriculture, Individual Eastern European Focus Countries, 2000­03 100 80 60 percent 40 20 0 20 ia ia Hu ia y a Re ic ic ey de a n Bu d ia an e ar ni Fe oni in tio n l bl en an tv ar ub rk st la ua ra ng pu La lg ra kh m Tu ov t Po Cz Rep Uk Es th Ro za Sl Li Ka ak h ec n ov ia Sl ss Ru Source: Anderson and Valenzuela (2008), updated from estimates reported in Anderson and Swinnen (2009). Government intervention and controls are especially important in a few key commodities within each country, often because of (real or imagined) food secu- rity concerns or the need to raise government revenue to meet other priorities. This is the case for grains and oilseeds in Bulgaria, Russia, and Ukraine, both those used for human consumption and those that support (via low feed input prices) the production of livestock products. It has also been the case for cotton in Tajik- istan, Turkmenistan, and Uzbekistan, where heavy taxation distorts incentives for producers--although open or porous borders make the taxing of cotton exports difficult when tax rates vary across countries in the subregion. The trade bias index (TBI) reported in table 6.4 is one way of capturing the diversity of assistance rates across farm commodities. The more negative that index, the greater the gap between assistance to import-competing farm indus- tries and assistance (or in some cases effective taxation) of export industries. Table 6.4 suggests that the antitrade bias has been a persistent feature of agricultural policies in the region throughout the transition period. Indeed, it has been worse in recent years than it was a decade earlier. A more comprehensive way to measure the extent of variance of rates across time is to calculate the standard deviation of NRAs for the covered products. These too have remained persistently high. On average, they have been higher in recent years than in the early stages of transition (table 6.5). 272 Distortions to Agricultural Incentives: A Global Perspective Table 6.3. NRAs, Key Covered Farm Products,a Eastern European and CIS Focus Countries, 1992­2005 (percent) 1992­95 1996­99 2000­03 2004­07 Wheat 6 13 2 9 Barley 1 16 5 6 Oats 11 7 27 4 Rye 0 14 10 2 Maize 16 3 16 21 Rapeseeds 8 18 1 0 Sunflower 13 13 13 4 Soybeans 45 0 9 4 Cotton 45 47 31 29 Sugar 23 80 73 91 Potatoes 25 25 60 57 Beef 16 20 36 53 Sheep meat 10 10 3 15 Pig meat 8 16 12 32 Poultry 26 43 52 75 Eggs 16 48 2 25 Milk 6 43 25 26 All covered products 2 19 13 22 Domestic market support 1 1 1 2 Border market support 4 15 11 19 Dispersion of product NRAs 21 29 29 31 Product coverageb 62 63 61 62 Source: Anderson and Valenzuela (2008), updated from estimates reported in Anderson and Swinnen (2009). a. Region's weighted average for each product and for all covered products, with weights based on gross value of agricultural production at undistorted prices. b. Dispersion is the standard deviation shown as the simple four-year average of the annual standard deviation around the weighted mean. The total amount of support is an imperfect indicator of distortions to incen- tives, since different trade, price, and subsidy instruments have different distortion effects. Most support to agriculture in the region was, and despite the reforms, still is provided via highly distortive and hence inefficient policy instruments. Under Communism, output price distortions were complemented by heavy distortions in input prices, in particular low fertilizer and energy prices and subsidized irriga- tion, while in the 1990s the majority of farm support in the CEE countries was provided by keeping output prices above border prices (see near the bottom of table 6.3). However, the share of support from those measures has declined over Eastern Europe and Central Asia 273 Table 6.4. TBI,a Eastern European and CIS Focus Countries, 1992­2007 (percent) 1992­95 1996­99 2000­03 2004­07b Bulgaria 0.02 0.17 0.18 0.30 Czech Republic 0.05 0.10 0.23 0.16 Estonia 0.21 0.16 0.01 0.24 Hungary 0.14 0.12 0.11 0.05 Latvia 0.35 0.18 0.15 0.22 Lithuania 0.50 0.32 0.19 0.07 Poland 0.19 0.19 0.24 0.26 Romania 0.19 0.28 0.40 0.23 Slovak Republic 0.03 0.09 0.05 0.06 Slovenia 0.26 0.40 0.38 0.18 CEE-10 0.15 0.16 0.23 0.02 Turkey 0.32 0.46 0.32 0.19 Russian Federation 0.11 0.31 0.34 0.24 Ukraine 0.12 0.25 0.21 0.42 Kazakhstan -- -- 0.01 0.32 All focus countries 0.15 0.16 0.23 0.02 Source: Anderson and Valenzuela (2008), updated from estimates reported in Anderson and Swinnen (2009). Note: -- not available. a. TBI [(1 NRAagx 100) (1 NRAagm 100) 1], where NRAagm and NRAagx are the average percentage NRAs for the import-competing and exportable parts of the agricultural sector. b. Final column refers just to 2004­05 for Russia and Ukraine and 2004 for Kazakhstan. the past decade, consistent with developments within the EU-15. These policy changes are reflected in the composition of the assistance that farms have received. Under the Communist system, price support and output subsidies were the main component of agricultural assistance in the CEE countries, accounting for more than 80 percent of their NRA. After the reforms in the early 1990s, the share of market support and output subsidies declined substantially, falling below half. Since then, it has grown again, to around half of the NRA. Other important com- ponents of the NRAs of CEE countries and Turkey were input subsidies, direct payments, and other NPS subsidies, plus some decoupled payments in the most recent years (table 6.6).7 In the CIS countries, those payments include soft loans and debt forgiveness, which continue to play an important role. While fiscal con- straints for most of the 1990s limited the government's ability to support farms by this means, the budgetary situation changed in the 2000s as earnings from mineral and energy exports grew and these earnings have become a more important source of government assistance to farmers. 274 Distortions to Agricultural Incentives: A Global Perspective Table 6.5. Dispersion of NRAs across Covered Agricultural Products,a Eastern European and CIS Focus Countries, 1992­2007 (percent) 1992­95 1996­99 2000­03 2004­07c Bulgaria 18 21 25 48 Czech Republic 27 28 23 53 Estonia 24 28 20 45 Hungary 34 41 62 49 Latvia 42 40 44 58 Lithuania 47 47 53 54 Poland 31 28 27 53 Romania 48 52 59 69 Slovak Republic 25 27 25 49 Slovenia 50 42 39 57 CEE-10 35 35 38 54 Turkey 62 65 53 69 Russian Federation 37 33 40 40 Ukraine 66 48 37 32 Kazakhstan -- -- 28 39 All focus countriesb 39 38 38 45 Source: Anderson and Valenzuela (2008), updated from estimates reported in Anderson and Swinnen (2009). Note: -- not available. a. Dispersion for each country is a simple four-year average of the annual standard deviation around a weighted mean of NRAs across covered products. b. Unweighted average, that is, the simple average across the 14 countries of their four-year simple average dispersion measures. c. Final column refers just to 2004­05 for Russia and Ukraine and 2004 for Kazakhstan. The GSE of the assistance to farmers, expressed in constant 2000 dollar terms, shows that Turkey has been the largest supporter between the early 1990s and mid-2000s, though Russia is rapidly catching up to Turkey. Romania and Poland are the next largest aggregate supporters. The region as a whole provides support of more than US$24 billion per year, compared with just US$3 billion in the early years of transition (table 6.7a). When expressed on a per farmer basis, the range is huge. In 2000­03, for example, it ranged from negative amounts (­US$300) in Ukraine and Kazakhstan to an average of US$980 in the CEE countries, US$430 in Russia, more than US$2,200 in Hungary and Romania, and a huge amount in high-income Slovenia (table 6.7b). This compares with US$8,400 per farmer in the EU-15 in 2000­04 (Josling 2008). Slovenia's support has come down signifi- cantly since its accession to the EU, however, to an average of just under Eastern Europe and Central Asia 275 Table 6.6. Components of NRAs to Agriculture, Eastern Europe and CIS, 1961­2007 (percent) a. CEE-10 1992­94 1995­99 2000­04 2005­07 NRA, covered products 10.4 16.4 25.5 17.7 NRA, noncovered products 10.6 17.3 26.9 24.7 NRA, all agriculture (excluding NPS) 10.5 16.7 26.1 21.3 All importables 19.1 32.1 45.2 31.7 All exportables 2.4 7.0 10.9 12.0 NRA, NPS assistance 1.9 2.5 2.5 4.4 NRA, all agriculture (including NPS assistance) 12.4 19.2 28.6 25.7 NRA, decoupled payments 0.6 0.8 2.9 12.1 NRA, all agriculture (including NPS and decoupled assistance) 13.0 20.1 31.5 37.8 NRA, all agricultural tradables (including NPS) 12.9 19.5 26.4 15.9 NRA, all nonagricultural tradables 5.7 4.9 4.4 4.6 RRA 6.7 14.1 21.1 10.7 b. Russian Federation and Ukraine 1992­94 1995­99 2000­04 2005 NRA, covered products 23.7 8.6 4.2 11.0 NRA, noncovered products 23.9 11.3 6.2 14.1 NRA, all agriculture (excluding NPS) 23.7 9.4 4.9 12.0 All importables 25.4 20.3 24.7 22.9 All exportables 21.5 13.4 15.3 2.0 NRA, NPS assistance 6.3 4.8 2.7 0.0 NRA, all agriculture (including NPS assistance) 17.4 14.2 7.5 12.0 NRA, decoupled payments 2.6 0.6 0.0 0.0 NRA, all agriculture (including NPS and decoupled assistance) 14.8 14.8 7.6 12.0 NRA, all agricultural tradables (including NPS) 17.4 14.2 5.9 6.3 NRA, all nonagricultural tradables 4.9 9.0 8.1 7.3 RRA 21.5 4.8 2.0 0.9 (Table continues on the following page.) 276 Table 6.6. Components of NRAs to Agriculture, Eastern Europe and CIS, 1961­2007 (continued ) (percent) c. Turkey 1961­64 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 2005­07 NRA, covered products 18.8 17.7 6.9 8.1 29.5 4.0 19.7 21.2 20.0 29.4 NRA, noncovered products 18.8 17.7 6.9 8.1 29.5 4.0 19.7 21.2 20.0 27.8 NRA, all agriculture (excluding NPS) 18.8 17.7 6.9 8.1 29.5 4.0 19.7 21.2 20.0 28.8 All importables 10.8 9.6 5.8 19.7 19.6 28.5 60.2 80.5 54.0 45.0 All exportables 29.9 28.4 18.0 23.3 35.5 8.1 2.5 1.9 3.5 21.9 NRA, NPS assistance 0.1 0.3 1.9 0.6 0.3 0.0 0.0 0.0 3.3 1.5 NRA, all agriculture (including NPS assistance) 18.9 17.9 5.0 7.5 29.2 4.0 19.7 21.2 23.2 30.3 NRA, decoupled payments 0.0 0.0 0.0 0.0 0.2 0.0 0.0 0.0 3.7 2.9 NRA, all agriculture (including NPS and decoupled assistance) 18.9 17.9 5.0 7.5 29.1 4.1 19.7 21.2 26.9 33.2 NRA, all agricultural tradables (including NPS) 18.9 17.9 5.0 7.5 29.2 4.0 19.7 21.2 20.4 23.5 NRA, all nonagricultural tradables 60.5 140.8 49.6 55.7 32.8 20.5 10.0 2.3 0.9 0.5 RRA 46.5 64.0 35.9 35.6 46.6 13.6 8.8 18.6 19.3 23.0 Source: Anderson and Valenzuela (2008), updated from estimates reported in Anderson and Swinnen (2009). Eastern Europe and Central Asia 277 Table 6.7. GSEs of Assistance to Farmers, Total and Per Farm Worker, Eastern European and CIS Focus Countries,a 1992­2007 a. Total (constant 2000 US$ millions per year) 1992­95 1996­99 2000­03 2004­07b Bulgaria 671 381 17 197 Czech Republic 784 632 711 689 Estonia 73 82 74 90 Hungary 856 768 1,205 920 Latvia 208 167 195 179 Lithuania 332 414 395 361 Poland 1,378 3,106 857 4,314 Romania 1,921 2,064 3,332 4,073 Slovak Republic 421 338 309 301 Slovenia 431 483 381 143 CEE-10 4,509 7,674 7,441 11,265 Turkey 4,671 8,033 6,070 10,525 Russian Federation 1,486 7,394 3394 3,100 Ukraine 4,461 70 1157 182 Kazakhstan -- -- 34 69 All focus countries 3,234 23,032 15,715 24,778 b. Per person engaged in agriculturec (constant 2000 US$ per year) 1992­95 1996­99 2000­03 2004­07 Bulgaria 1,429 1,075 65 1,010 Czech Republic 1,423 1,255 1,581 1,762 Estonia 678 898 931 1,267 Hungary 1,335 1,372 2,494 2,253 Latvia 1,038 993 1,333 1,393 Lithuania 1,113 1,693 1,932 2,123 Poland 283 683 204 1,118 Romania 879 1,135 2,202 3,311 Slovak Republic 1,393 1,199 1,197 1,281 Slovenia 10,781 18,225 22,105 14,254 CEE-10 466 893 977 1,682 Turkey 344 566 414 702 Russian Federation 152 842 431 439 Ukraine 956 17 333 60 Kazakhstan -- -- 27 59 All focus countries 86 647 451 752 Source: Anderson and Valenzuela (2008), based on NRA estimates updated from Anderson and Swinnen (2009) and data on the number of farmers from U.N. Food and Agriculture Organization Statistical Database (FAOSTAT). Note: -- not available. a. GSEs include assistance to nontradables and NPS assistance. The number of farmers in these countries is difficult to obtain on a consistent basis. The FAOSTAT numbers may be subject to error. For example, Slovenia's may be understated in FAOSTAT, in which case its GSE per farmer is overestimated. b. Final period refers to 2004­05 for Russia and Ukraine and just 2004 for Kazakhstan. 278 Distortions to Agricultural Incentives: A Global Perspective US$14,000 per farmer in 2005­07. For the EU-acceding countries, per farmer assistance over the next few years is likely to move closer to the EU average. Since most of the support for farmers previously came through price-support measures, most notably import restrictions, these have the effect of raising con- sumer prices by a similar degree when calculated at the farmgate level. That means that prior to the mid-1990s, policies in all but Turkey and Slovenia imposed the equivalent of low or negative taxes on food consumers (CTEs), but thereafter the CTEs have become positive. The region's weighted average CTE in 2000­03 was 17 percent (table 6.8), compared with nearly twice that in the EU-15. CTEs in Romania and Slovenia have been well above that EU average this decade, and so presumably will fall during those countries' transition to the EU's Common Agri- cultural Policy (CAP), especially given the EU's policy reinstrumentation toward more direct farm income supports that do not raise consumer prices of food. Assistance to agriculture relative to other tradable sectors Import tariffs on primary agricultural commodities in Europe and Central Asia are on average twice as high as average tariffs in industry, but only half as high as tariffs on processed food. This is true both for the CEE and CIS countries. It suggests that while the region's farmers receive more tariff protection from competition abroad than do nonagricultural producers, food processors may be effectively protected despite having to pay more than world prices for primary farm products. Import-competing producers are only part of each sector, however. When support for producers of exports in each sector is also taken into account, an overall NRA for all nonagricultural tradable industries can be used, together with the average NRA for agricultural tradable industries, to calculate the RRA. In so far as the NRAs for nonfarm industries are positive, the RRA is lower than the NRA for agriculture. But in most cases, the nonagricultural NRA is very low. Thus the overall NRA for tradable primary agriculture in the region during 2000­03 is estimated to have averaged more than three times higher than for producers of nonagricultural tradables (15 as com- pared with 5 percent), so the RRA averaged 10 percent. Only in three countries-- Bulgaria, Kazakhstan, and Ukraine--has agricultural production been assisted less than nonagricultural tradables (RRA 0) during the present decade. And in virtu- ally all countries for which there is a time series, the RRA is higher at the end of the sample period than in the first few years of transition (figure 6.3). Forces behind Transitional Policy Choices Several political economy stylized facts widely observed in market economies-- for reasons explained in, for example, Anderson and Hayami (1986), Anderson (1995), Swinnen (1994), and de Gorter and Swinnen (2002)--are also found in Eastern Europe and Central Asia 279 Table 6.8. Percentage CTE of Policies Assisting Producers of Covered Farm Products,a Eastern European and CIS Focus Countries, 1992­2007 (percent, at primary product level ) 1992­95 1996­99 2000­03 2004­07e Bulgaria 20 10 3 7 Czech Republic 23 19 22 20 Estonia 15 12 9 20 Hungary 18 15 22 16 Latvia 2 28 32 32 Lithuania 20 21 20 29 Poland 4 2 18 25 Romania 6 16 39 29 Slovak Republic 13 15 16 17 Slovenia 48 58 45 24 CEE-10 2 11 24 23 Turkey 10 20 16 9 Russian Federation 37 13 16 24 Ukraine 25 0 3 3 Kazakhstan -- -- 4 16 All focus countries: Unweighted average 1 16 19 20 Weighted averageb 13 13 16 14 Dispersion of national CTEsc 27 17 14 10 Dispersion of region's product CTEsd 35 37 47 53 Source: Anderson and Valenzuela (2008) updated from estimates reported in Anderson and Swinnen (2009). Note: -- = not available. a. Assumes the consumer tax equivalent (CTE) is the same as the NRA derived from trade measures (that is, not including any input taxes/subsidies or domestic producer price subsidies/taxes). b. Weights are consumption valued at undistorted prices, where consumption (from FAO) is production plus imports net of exports plus change in stocks of the covered products. c. Simple four-year average of the annual standard deviation around a weighted mean of the regional average CTE across the covered products. d. Simple four-year average of the annual standard deviation around a weighted mean of the national average CTE for covered products. e. Final column refers to 2004­05 for Russia and Ukraine and just 2004 for Kazakhstan; while CEE values assume the CTE for each product is the same as for the EU-25 in 2004­06 and for EU-27 in 2007, such that the differences across CEE countries is due to differing national consumption weights. Europe and Central Asia's transition economies. Specifically, for this region, as elsewhere, farmer assistance tends to be higher in higher-income countries, and in countries with weaker comparative advantage in agriculture. Hence, it is likely that similar politico-economic interactions and mechanisms are at work 280 Distortions to Agricultural Incentives: A Global Perspective Figure 6.3. RRAs to Agriculture,a Eastern European Focus Countries, 1992­95 and 2000­03 Slovenia Romania Latvia Lithuania Hungary Czech Republic Estonia Turkey Regional average Slovak Republic Poland Russian Federation Bulgaria Kazakhstan Ukraine 40 20 0 20 40 60 80 percent 1992­95 2000­03 Source: Anderson and Valenzuela (2008), updated from estimates reported in Anderson and Swinnen (2009). a. RRA is defined as 100*[(100 NRAagt ) (100 NRAnonagt ) 1], where NRAagt and NRAnonagt are the percentage NRAs for the tradables parts of the agricultural and nonagricultural sectors, respec- tively. For several countries no estimates are available for the 1992­95 period. in this region as in other parts of the world. However, those correlations are becoming weaker over time among the CEE countries. Taking on the EU's CAP is part of the explanation, but there are also other forces, both domestic and international, that underlie the political economy of agricultural policies in the CEE and CIS countries. Several of those key forces are discussed in the following subsections. Causes of rent extraction Traditionally, heavy negative government intervention in the form of depressed incentives tends to be concentrated on commodities that have the potential to provide export tax revenue for the government. This is especially the case in the cotton sectors of Tajikistan, Turkmenistan, and Uzbekistan. There, as in a number Eastern Europe and Central Asia 281 of African countries (see Anderson and Masters 2009), the government controls the cotton chain so as to extract rents, thereby depressing farmers' prices and pro- duction incentives. There is a clear division in Central Asia between the roughly neutral policy toward cotton in Kazakhstan and the Kyrgyz Republic (where cotton previously accounted for a relatively modest share of total exports) on the one hand, and on the other the extensive taxation and extraction of rents from cotton in Tajikistan, Turkmenistan, and Uzbekistan (countries where cotton traditionally was a very important export tax resource).8 In Turkmenistan and Uzbekistan, governments use state monopoly powers over marketing to transfer substantial resources out of agriculture. Most of the transfers in Uzbekistan appear to go to general govern- ment revenue, whereas in Turkmenistan much is wasted (for example, in ineffi- cient cotton mills, which have negative value added) or accrues to secret accounts under the president's personal control. Moreover, recently, potentially important reforms have been introduced in Uzbekistan to reduce some of the distortions to farm incentives, while almost none have taken place in Turkmenistan. In Tajikistan, the rent distribution is more opaque, but equally detrimental to farms, as a coalition between the government and a monopolistic private trading company has caused depressed prices and incentives for farmers. Not surprisingly, cotton farmers have responded sharply to these incentive distortions, both in area and output: with rapid growth in Kazakhstan and the Kyrgyz Republic, and with declines or stagnation in the other countries. The grain and oilseed export sectors of Bulgaria, Ukraine, and the grain sur- plus regions of Russia are similarly characterized by heavy government regulation and interventions. In traditional grain-exporting countries such as Bulgaria and Ukraine, for historic and psychological reasons, the grain sector has dispropor- tionate political significance. For example, in the mid-1990s in Bulgaria, ministers of agriculture were regularly forced to resign following reports of grain shortfalls or unregulated exports threatening the local grain supply. In Ukraine, ad hoc grain market interventions continued in recent years. Opportunities for rent seeking from distorted policies inhibit policy reform, as the few who benefit disproportionately from existing distortions lobby strongly for their continuation. This applies to various policies, such as cotton regulations in Central Asia, grain trade regulations in Bulgaria, Russia, and Ukraine and water policies in Central Asia. But it also applies to several policies in countries in which benefits go a specific group of farms. For example, the continuation of soft budget constraints in the large CIS countries, along with the failure of governments to enforce bankruptcies and strong land rights, disproportionately benefit large farming companies, while smaller family farms are often hurt by these policies. In Turkey, agricultural parastatal companies and marketing cooperatives benefit 282 Distortions to Agricultural Incentives: A Global Perspective from "farm support" and are major lobbyists in favor of market regulations and assistance packages. Sometimes, specific political, regional, or ethnic coalitions also play a role. In Kazakhstan, for example, many residents of the rich northern grain regions were Russian and German. After independence, power shifted to Kazakh nationals, lim- iting the Russian and German groups' influence in government and causing many to emigrate. Another recent example is Bulgaria, where the government's resist- ance to privatizing tobacco processing companies and its decision to allocate a dis- proportionate amount of subsidies to tobacco growers are due to the fact that the Turkish minority in Bulgaria is particularly active in the sector and traditionally has held key positions in the Ministry of Agriculture. Causes of increases in support during transition Increases in agricultural support in the CEE countries in the second half of the 1990s and more recently in CIS countries are the result of the interaction of domestic political forces with international events. The increase in farmer assis- tance in CEE countries was likely caused by "normal" domestic internal pressures that are brought to bear in a contestable political environment, which results in an increase in agricultural protectionism as per capita income increases and agricul- tural comparative advantage declines. Protectionism between the mid-1990s and mid-2000s was also a case of reversing somewhat the overshooting in reform dur- ing the first few years of transition. Overlaying these trends is the EU accession process, which encouraged CEE governments to target the levels of support expected in the EU by the end of the phase-in period of accession, so as to maximize the transfer of benefits from Brussels. However, it appears that in the years leading up to accession, the process had more impact on the introduction of new support instruments than on the overall level of support, probably because the full cost of that support had to be borne by the national economy prior to EU accession (Swinnen 2002). Another factor contributing to the increase in agricultural support in the mid- 1990s was the improvement in the government's budgetary situation, which allowed more subsidies to be given to farmers than was possible in the early years of transition. This factor has played a role throughout Europe and Central Asia, in particular in Russia and some of its neighbors, where recovery from the 1998 financial crisis has been aided by windfall gains from the dramatic rise in the export prices of energy raw materials. This factor has been stronger in countries where governments have more access to mineral resources, such as in Russia (oil and gas), Kazakhstan (oil), and Turkmenistan (gas). Eastern Europe and Central Asia 283 Crises and political change General political and economic crises have played an important role in inducing changes in agricultural distortions. The most obvious example is the fall of the Communist regime and the disintegration of the Soviet Union--and of the cen- tral directives coming from Moscow. However, even later there are several exam- ples in which general crises have triggered changes. Most often, the policy reforms came only after new elections induced a change in government, reflecting changed electoral preferences. For example, in Romania and Bulgaria, important progress in the removal of distortions and market reforms occurred only in the late 1990s, after electoral change brought reform-minded governments to power. In Bulgaria, significant progress was brought about by the financial crisis in 1996. Important reform progress was made in Ukraine in the years after the 1999 election in which the large farm lobby fell out with President Kuchma, who consequently introduced a series of important reforms that farmers had previously successfully opposed. Democratic political change, however, is not a sufficient condition in itself for better agricultural policies. For example, political revolutions in Ukraine and the Kyrgyz Republic (the "Orange Revolution" and the "Tulip Revolution," respec- tively) have not contributed to better agricultural policy. In fact, agricultural dis- tortions seem to have increased under the postrevolution Ukrainian government, while in the Kyrgyz Republic political change has mostly resulted in more market instability, while relatively little distortion remains in agriculture. Impact of international agreements EU accession, both prospective and actual, has had obvious and profound influ- ences on policy choices. The CEE countries that joined in May 2004 and January 2007 have raised domestic agricultural and food prices toward EU-15 levels (on average, since for some prices came down). An important part of the EU farm subsidies is now in the form of direct payments. Though CEE farms receive con- siderably less of these subsidies than do EU-15 farmers, they will gradually increase to reach EU-15 levels by 2010. Another important difference is that sub- sidies in the EU-15 will be given on a per farm basis (single farm payments) earlier than in CEE countries. CEE countries also have been induced also to undertake major regulatory improvements to stimulate their markets, including private investments in the food chain and public rural infrastructure investments. Likewise, their trade poli- cies have changed to allow free access for all products from other EU-27 member countries and, in most cases, also freer access for nonagricultural products from 284 Distortions to Agricultural Incentives: A Global Perspective non-EU countries (the latter because the common external tariff typically was lower than that previously applied in acceding countries). The EU accession process, however, has not caused a major increase in food prices in the CEE coun- tries. Two reasons behind this are the increased competition in consumer markets in the CEE countries with the full opening of agri-food markets to imports, and the massive inflow of foreign direct investment in the retail sector. Impacts of other international agreements (including WTO accessions) have varied. The Czech Republic, Hungary, Poland, Romania, the Slovak Republic, Slovenia, and Turkey have been members of the WTO since its creation in 1995. Albania, Armenia, Bulgaria, Estonia, Georgia, the Kyrgyz Republic, Latvia, Lithua- nia, and Ukraine joined the WTO later, while Russia and Kazakhstan are still negotiating their WTO accession. WTO accession has not strongly disciplined European and Central Asian coun- tries that were founding members in 1995 (Bacchetta and Drabek 2002). For those that had to negotiate their entry in the latter 1990s, constraints on introducing or maintaining distortions are more serious. And for large countries in the region still in the process of negotiating their accession, notably Kazakhstan and Russia, the WTO has been even tougher in its demands. Whether those demands will prove an agricultural trade-liberalizing force remains to be seen, but at least they will provide a ceiling on the extent to which agricultural protection and subsidies may be raised in the future. For the CEE countries, the most important WTO impact has been indirect: in anticipation of eastward enlargement, the EU was forced to introduce major changes to its CAP, which in turn has affected postaccession agricultural distor- tions in the CEE countries. A further and somewhat erratic influence has been regional trading arrange- ments among the European and Central Asian countries. These include the Eurasian Economic Community (EAEC), the Central European Free Trade Area, and the Baltic Free Trade Area. However, the impact of these agreements on reducing agricultural policy distortions has been generally limited, because the agreements include many exceptions for agricultural and food products, espe- cially for so-called sensitive products, which make up a substantial share of pro- duction. Moreover, Central Asian countries such as Kazakhstan and the Kyrgyz Republic have been reluctant to join the EAEC because the organization would impose Russia's trade and customs preferences on them. Influence of international institutions The role of other international institutions was very important at the start of tran- sition in Europe and Central Asia, as it provided policy reform guidance in all Eastern Europe and Central Asia 285 these countries. However, in more recent years, the influence of international institutions has declined. For countries joining the EU, policy advice from Brus- sels was perceived as more relevant. This is especially, but not only, the case for the EU accession countries. For countries aspiring to join the EU (such as most of the Balkan countries and Ukraine), or for those that view the accession countries as models for their own development strategies, policy advice from Brussels is taken seriously. Another reason is that in many of the countries of Southeast Europe and the CIS, improved macroeconomic situations have made them less beholden to international financial institutions requiring reforms as a condition for providing loans or financial assistance. Prospects for Further Reducing Distortions Clearly, there have been major reductions in distortions to agricultural incentives in Europe and Central Asia over the past two decades. In many countries in the region, average protection levels are now relatively low. However, there is still sub- stantial room for further reduction of distortions to agricultural incentives. This could be done through various means: overall reductions in support, shifting sup- port to less-distortive policy instruments, and focusing budgetary expenditures on public good investments (in infrastructure and in institutions that reduce trade costs) rather than on farm subsidies, shifting from a quantity-based to a quality-based policy paradigm, and so forth. In terms of further reductions in policy distortions, some of the most distortive cases concern taxation of agriculture, most notably the control and rent extrac- tion in the cotton sectors in some Central Asian countries. Removing those distor- tions would allow a substantial improvement in incentives to domestic producers. Though some progress has been made in recent years, much more can be done. Countries for which EU accession is unlikely to happen in the medium term, such as Turkey, Ukraine, and several of the Balkan countries, should focus their policy attention in the near term on efficiency improvements in both their policies and their agricultural economies. This is consistent with the objective of EU acces- sion, since the EU itself has moved in recent years to more decoupled farm sup- port and is demanding that member countries move in that direction and improve the efficiency of their farms and food companies.9 The same policy framework should be promoted in countries further east, including in those that are likely to spend more funds on agriculture in the com- ing years as their fiscal situation improves. Increased funding should be focused on upgrading infrastructure, on quality and efficiency of the agri-food system, and on the introduction or improvements of a variety of institutions necessary to support rural markets. In several of the poorer and larger CIS countries, institu- 286 Distortions to Agricultural Incentives: A Global Perspective tional and infrastructure problems, as well as corruption, remain major con- straints to trade and thereby distort farm incentives. Competition and antitrust policy are important related areas for policy atten- tion. In supply chains where farms have to sell their products to trading, process- ing, and retailing companies, the ability to choose freely between companies is of crucial importance in creating better conditions for farms. This applies across the region, where monopoly buyers (state-owned or private) push down prices and contract conditions, although the source of anticompetitive behavior and policy details are likely to differ, for example, between the increasing dominance of large retail chains in Central Europe versus some of the government-controlled cotton chains in Central Asia. Despite constraining political economy forces, there are prospects for further reducing distortions to agricultural incentives in the foreseeable future. Accession of the CEE countries to the EU has increased their levels of farm assistance, though they now face more competition within the enlarged EU. While reducing CEE farm assistance in the future will not happen without reductions in EU pro- tection levels, some reforms are currently underway in the EU (for example, the cut in EU sugar price support and the shift from per hectare payments to single farm payments). However, slow and intermittent progress in the WTO's Doha Round reduces the pressure for further reforms. Meanwhile, in the mineral- and energy-rich CIS countries, the rise in export earnings reduces budgetary con- straints on governments inclined to give assistance to farmers as national incomes grow. CIS regional trade policies that affect markets, largely ad hoc and nontrans- parent, are also important distortions. Eliminating these policy interventions would require fundamental reforms of Russia's political system, including a transforma- tion of attitudes and behaviors involving governance that Russian accession to the WTO is unlikely to alter in the medium term. Notes 1. This is a strong feature of Asia's economies in transition as well. For a comparison of the Asian and European transition experiences, see Swinnen and Rozelle (2006). 2. The only country from this region that was part of the Krueger, Schiff, and Valdés (1991) study was Turkey, for the period 1961 to 1983. However, a follow-up study subsequently undertaken for a few economies in transition has been made available in a World Bank technical report (Valdés 2000). 3. For a detailed discussion of how misalignments in the exchange rate affect the measurement and interpretation of NRAs in Russia during its transition period, see Liefert and Liefert (2008). 4. In addition to the present study's estimation of price wedges, the latter would require a sophisti- cated economic model of the national economy with some quantification of the difference between private and social marginal costs. Such welfare analysis is thus beyond the scope of the present study. 5. The dramatic implications-including millions of peasants dying of starvation-are documented in sobering detail in Conquest (1986). Eastern Europe and Central Asia 287 6. For an assessment of the support to farmers in the 1980s, see Cook, Liefert, and Koopman (1991). 7. Water price regulations and subsidies are important policy instruments in the irrigated regions of Central Asia, but it was not possible in this study to estimate their impact on NRAs. Energy policies are still used to assist various sectors, for example in Russia, but since they do not favor agriculture in particular, and are becoming less important, they too have been omitted from the NRA estimates. 8. Price and trade data were not sufficiently reliable to allow NRA calculations, but Pomfret (2008a, 2008b) and Christensen and Pomfret (2008) provide considerable informal information supporting the claims above. 9. From this perspective, it is important to point to the importance of other reforms, such as macroeconomic and regulatory reforms to stimulate food industry investment, labor market reforms to enhance off-farm employment opportunities, and credit reforms to stimulate access to rural credit. References Anderson, K. 1995. "Lobbying Incentives and the Pattern of Protection in Rich and Poor Countries." Economic Development and Cultural Change 43 (2): 401­23. Anderson, K., Y. Hayami, et al. 1986. The Political Economy of Agricultural Protection: East Asia in Inter- national Perspective. Boston, London, and Sydney: Allen and Unwin. Anderson, K., M. Kurzweil, W. Martin, D. Sandri, and E. Valenzuela. 2008a. "Methodology for Measur- ing Distortions to Agricultural Incentives." Agricultural Distortions Working Paper 02, World Bank, Washington, DC (and appendix A of this volume). ______. 2008b. "Measuring Distortions to Agricultural Incentives, Revisited." World Trade Review 7 (4): 1­30. Anderson, K., and W. Masters, eds. 2009. Distortions to Agricultural Incentives in Africa. Washington, DC: World Bank. Anderson, K., and J. Swinnen, eds. 2008. Distortions to Agricultural Incentives in Europe's Transition Economies. Washington, DC: World Bank. ______. 2009. "Distortions to Agricultural Incentives in Eastern Europe and Central Asia." Agricul- tural Distortions Working Paper 63, World Bank, Washington, DC. Anderson, K., and E. Valenzuela. 2008. "Estimates of Global Distortions to Agricultural Incentives, 1955 to 2007." World Bank, Washington, DC. http://www.worldbank.org/agdistortions. Bacchetta, M., and Z. Drabek. 2002. "Effects of WTO Accession on Policy-Making in Sovereign States: Preliminary Lessons from the Recent Experience of Transition Countries." Staff Working Paper DERD-2002-02, World Trade Organization, Geneva. Broadman, H. G., ed. 2006. From Disintegration to Reintegration: Europe and Central Asia in Interna- tional Trade. Washington, DC: World Bank. Chen, S., and M. Ravallion. 2007. "Absolute Poverty Measures for the Developing World, 1981­2004." Policy Research Working Paper 4211, World Bank, Washington, DC. Christensen, G., and R. Pomfret. 2008. "The Kyrgyz Republic." In Distortions to Agricultural Incentives in Europe's Transition Economies, ed. K. Anderson and J. Swinnen. Washington, DC: World Bank. Conquest, R. 1986. The Harvest of Sorrow. London: Oxford University Press. Cook, E. C., W. M. Liefert, and R. B. Koopman. 1991. "Government Intervention in Soviet Agriculture: Estimates of Consumer and Producer Subsidy Equivalents." Staff Report AGES 9146, Economic Research Service, U.S. Department of Agriculture, Washington, DC. Crafts, N., and G. Toniolo. 2008. "European Economic Growth, 1950­2005: An Overview." CEPR Discussion Paper 6863, Centre for Economic Policy Research, London. de Gorter, H., and J. Swinnen. 2002. "Political Economy of Agricultural Policies." In volume 2 of Hand- book of Agricultural Economics, ed. B. Gardner and G. Rausser. Amsterdam: Elsevier Science. Dries, L., and J. Swinnen. 2004. "Foreign Direct Investment, Vertical Integration and Local Suppliers: Evidence from the Polish Dairy Sector." World Development 32 (9): 1525­44. 288 Distortions to Agricultural Incentives: A Global Perspective Dries, L., T. Reardon, and J. Swinnen. 2004. "The Rapid Rise of Supermarkets in Central and Eastern Europe: Implications for the Agrifood Sector and Rural Development." Development Policy Review 22 (5): 525­56. Ellman, M. 1988. "Contract Brigades and Normless Teams in Soviet Agriculture." In Socialist Agricul- ture in Transition: Organizational Response to Failing Performance, ed. J. C. Brada and K. E. Wädekin, 23­33. Boulder, CO: Westview Press. Gray, K. R., ed. 1990. Soviet Agriculture: Comparative Perspectives. Ames, IA: Iowa State University Press. Huffman, S. K., and S. R. Johnson. 2004. "Impacts of Economic Reform in Poland: Incidence and Wel- fare Changes within a Consistent Framework." Review of Economics and Statistics 86 (2): 626­36. Josling, T. 2008. "Distortions to Agricultural Incentives in the Western Europe." Agricultural Distor- tions Working Paper 61, World Bank, Washington, DC. http://www.worldbank.org/agdistortions. Krueger, A. O., M. Schiff, and A. Valdés. 1991. The Political Economy of Agricultural Pricing Policy, Volume 3: Africa and the Mediterranean. Baltimore: Johns Hopkins University Press for the World Bank. Liefert, W., and O. Liefert. 2008. "The Russian Federation." In Distortions to Agricultural Incentives in Europe's Transition Economies, ed. K. Anderson and J. Swinnen. Washington, DC: World Bank. Liefert, W., and J. Swinnen. 2002. "Changes in Agricultural Markets in Transition Economies." Agricul- tural Economic Report AER806, Economic Research Service, U.S. Department of Agriculture, Washington, DC. Pomfret, R. 2008a. "Kazakhstan." In Distortions to Agricultural Incentives in Europe's Transition Economies, ed. K. Anderson and J. Swinnen. Washington, DC: World Bank. ______. 2008b. "Tajikistan, Turkmenistan and Uzbekistan." In Distortions to Agricultural Incentives in Europe's Transition Economies, ed. K. Anderson and J. Swinnen. Washington, DC: World Bank. Sandri, D., E. Valenzuela, and K. Anderson. 2007. "Economic and Trade Indicators for the Europe's Transition Economies, 1990 to 2004." Agricultural Distortions Working Paper 22, World Bank, Washington, DC. http://www.worldbank.org/agdistortions. Swinnen, J. 1994. "A Positive Theory of Agricultural Protection." American Journal of Agricultural Eco- nomics 76 (1): 1­14. ______. 2002. "Transition and Integration in Europe: Implications for Agricultural and Food Markets, Policy and Trade Agreements." The World Economy 25 (4): 481­501. Swinnen, J., and S. Rozelle. 2006. From Marx and Mao to the Market: The Economics and Politics of Agri- cultural Transition. New York: Oxford University Press. Valdés, A., ed. 2000. "Agricultural Support Policies in Transition Economies." World Bank Technical Paper 470, World Bank, Washington, DC. World Bank. 2007. World Development Indicators Database. World Bank. http://go.worldbank.org/ B53SONGPA0 (accessed June 2008). 7 Latin America and the Caribbean Kym Anderson and Alberto Valdés* While the vast majority of the world's poorest households depend on farming for their livelihoods, poverty tends to be less focused on rural areas in Latin America than in Africa or South Asia, as Latin American countries generally have higher levels of development, larger nonfarm sectors in relation to their overall economies, more extensive urbanization, and a greater concentration of land ownership. Nonetheless, poverty is sufficiently prevalent in numerous parts of Latin America and the Caribbean that it continues to be a concern. In the past, farm earnings in the region have often been depressed by the pro-urban, antiagri- cultural bias of government policies. True, progress has been made over the past two decades in reducing the policy bias, but many trade-reducing price distor- tions remain between sectors, as well as within the agricultural sectors of most Latin American countries. This study of Latin America is based on a sample of eight countries, including the big four economies of Argentina, Brazil, Chile, and Mexico; Colombia and Ecuador, two of the poorest tropical South American countries; the Dominican Republic, the largest Caribbean economy; and Nicaragua, the poorest country in Central America. Together, in 2000­04, these countries accounted for 78 percent of the region's population, 80 percent of the region's agricultural value added, and 84 percent of the total gross domestic product (GDP) of Latin America. *The authors acknowledge invaluable help with data management from Ernesto Valenzuela; data compilation by Johanna Croser, Esteban Jara, Marianne Kurzweil, Signe Nelgen, Francesca de Nicola, and Damiano Sandri; and help- ful comments from workshop participants. The working paper version of this chapter (Anderson and Valdés 2008b) contains additional background material and appendix tables. This chapter draws on the introductory and country chapters in Anderson and Valdés (2008a), with data updated using Anderson and Valenzuela (2008). 289 290 Distortions to Agricultural Incentives: A Global Perspective The key characteristics of these economies--which account for only 4.5 per- cent of worldwide GDP, but 7.7 percent of agricultural value added and more than 10 percent of agricultural and food exports--are shown in table 7.1. The table reveals the considerable diversity within the region in terms of stages of develop- ment, relative resource endowments, comparative advantages and, hence, trade specialization and the incidence of poverty and income inequality. This means that these countries represent a rich sample for comparative study. By compari- son, Nicaragua's per capita income is only one-seventh the global average, while the incomes of Colombia and Ecuador are one-third of this average. By contrast, the per capita income in Argentina and Chile averages just one-eighth below and in Mexico one-eighth above the global average. Only Argentina, Brazil, and Nicaragua are well above the global average in endowments of agricultural land per capita; the Dominican Republic and Ecuador are well below this average; and Chile, Colombia, and Mexico are a little less than one-third above the average. Income inequality is high throughout Latin America and the Caribbean compared with the rest of the world--the Gini coefficient is near or above 0.5 and averages 0.52. This is well above the Gini coefficient for Africa and Asia. Likewise, the Gini coefficient for land distribution is high in Latin America: 0.58 for Chile and more than 0.7 in Argentina, Brazil, Ecuador, and Nicaragua, compared with an average of less than 0.5 in Asia (World Bank 2007). Even so, there is comparatively little absolute poverty except in the poorest tropical parts of the region. Though it relies on nearly twice as much agricultural land per capita as the rest of the world, Latin American agriculture is characterized by concentrated land ownership and a structure of production whereby medium and large commercial farms contribute the bulk of agricultural output. It is also a region with a high degree of urbanization. These features are important in understanding the forces behind agricultural policies. So, too, is the fact that, until a few years ago, most countries in the region were experiencing a high degree of macroeconomic insta- bility and high inflation. The manipulation of food prices for urban consumers in an attempt to reduce inflation was (and, in Argentina, still is) a dominant feature driving farm pricing policy. Most Latin American countries have gone undergone major economy-wide policy reforms that began, for some countries, in the 1980s (or the 1970s for Chile) and, for others, in the mid-1990s. Reforms centered on macroeconomic stabilization, trade liberalization, deregulation, and some privatization of state agencies. There was a considerable reassessment of the role of government in guiding economic development. Agricultural policies were an integral part of this reform process, although not the principle motivation of the reforms. This chapter begins with a brief summary of economic growth and structural changes in the region since the 1960s and of agricultural and other economic Table 7.1. Key Economic and Trade Indicators, Latin American Countries, 2000­04 National relative to Share (%) of world: world (world 100) Gini Agricultural index Agricultural RCA,a trade for per Total Agricultural GDP land agriculture specialization Poverty capita Population GDP GDP per capita per capita and food indexb incidencec incomed Latin American focus countries 6.49 4.49 7.73 69 178 219 0.42 7 52 Argentina 0.61 0.54 1.04 89 426 541 0.85 5 51 Brazil 2.88 1.54 3.38 54 184 355 0.66 8 57 Chile 0.25 0.22 0.24 86 120 386 0.63 2 55 Colombia 0.70 0.24 0.77 35 132 264 0.25 7 59 Dominican Republic 0.14 0.06 0.18 41 54 474 0.29 3 52 Ecuador 0.20 0.07 0.16 33 80 487 0.59 16 44 Mexico 1.62 1.82 1.89 112 133 64 0.17 7 46 Nicaragua 0.08 0.01 0.06 14 169 952 0.26 44 43 Other Latin American countries 1.84 0.84 2.05 46 148 -- -- -- -- Caribbean 0.20 0.07 0.13 36 23 -- -- -- -- Central America 0.52 0.21 0.78 41 55 504 0.26 -- -- South America 1.12 0.56 1.13 50 213 157 0.16 13 -- All Latin American countries 8.33 5.33 9.78 64 171 -- -- -- -- Source: Sandri, Valenzuela and Anderson (2007), compiled mainly from World Bank (2008). Note: -- not available. a. RCA revealed comparative advantage. The RCA index is the share of agriculture and processed food in national exports as a ratio of that sector's share of global exports. b. Primary agricultural trade specialization index is net exports as a ratio of the sum of exports and imports of agricultural and processed food products (world average 0.0). 291 c. Percentage of the population living on less than US$1 per day. d. The poverty incidence and Gini index are for the most recent year available between 2000 and 2004, except for Ecuador where they refer to 1998. The weighted averages for the focus countries use population as the basis for weights. 292 Distortions to Agricultural Incentives: A Global Perspective policies as they affected agriculture before and after the reforms of the mid-1980s to mid-1990s. It then summarizes estimates of the nominal rate of assistance (NRA) and the relative rate of assistance (RRA) to farmers delivered by national farm and nonfarm policies over the past several decades, as well as the impact of these policies on the consumer prices of farm products. Both farmer assistance and consumer taxation tend to be negative in periods when there is an antiagri- cultural, pro-urban consumer bias in a country's policy regime. The final sections describe lessons learned and draw key policy implications for the region. Growth and Structural Changes1 Since 1980, real GDP in Latin America has grown at an average annual rate of 5.4 percent, or 3.6 percent per capita. These rates are somewhat above the averages of other developing countries of 4.1 percent total and 2.3 percent per capita, but somewhat below Asia's averages of 7.1 percent total and 5.5 percent per capita. The region's comparative growth performance was much less rosy in the 1960s and 1970s, however, before it moved away from an import-substitution industri- alization regime. Among the focus countries in our study, Chile and Mexico have been the star performers since 1980, while Ecuador and Nicaragua have been the slowest growers. Nicaragua's civil conflict set the country back in the 1980s, but its economy grew twice as fast as Ecuador's in the 1990s. Though the industrial sector has grown much more slowly than overall GDP over the past 25 years, agriculture has grown even more slowly, at barely half the rate of the rest of the economy, while the service sector has taken the lead. Among our sample countries, the economies of Chile and Mexico have been among the most rapidly growing, and Argentina's and Ecuador's the most slowly growing, apart from Nicaragua. As a result of strong growth in service activities during the past two decades, the share of services in GDP has risen from barely one-half to two-thirds, while agriculture's share fell from 9 to 6 percent, on average, in our sample economies. The relative decline of agriculture has been slowest in Argentina, Brazil, and Nicaragua and the most rapid in oil-exporting Ecuador and Mexico, but also in Chile. In 2000­04, agriculture's share of GDP ranged from 4 percent in Chile and Mexico to twice that in Brazil and Ecuador, three times that in Colombia and the Dominican Republic, and more than four times that in Nicaragua. The share of overall employment accounted for by farming activities in Latin America and Caribbean countries has fallen somewhat more slowly than agricul- ture's GDP share, according to statistics in the FAOSTAT Database of the Food and Agriculture Organization of the United Nations (which, because of defini- tional differences, is not always consistent with databases produced by countries). Latin America and the Caribbean 293 In general, the share remains at much higher level than the GDP share, implying relatively low and slow-growing labor productivity on farms. The most rapid decline has occurred in Brazil, where the employment share in agriculture has fallen from one-half to less than one-sixth during the past 40 years. Agriculture's average share in exports has also declined, by about one-third each decade since the late 1960s. The only exception is Chile, where the share has risen dramatically, from one-eighth to one-third. Chile contrasts markedly with the other rapidly growing economy in our sample, Mexico, where the share of farm products in all goods exports has fallen from 58 percent to only 6 percent. Additionally, the decline in relative importance of farm exports has been more rapid in Latin America than in the rest of the world: the index of the revealed comparative advantage of Latin America in these products (defined as the share of agriculture and processed food in national exports as a ratio of the share of such products in worldwide merchandise exports) has fallen by about one-third since the 1960s, as has the region's index of trade specialization (defined as net exports as a ratio of the sum of the imports and exports of agricultural and processed food products). There has been a marked upturn in these two indexes during the past decade, however, not only in Chile but in several other reforming Latin American countries, including Argentina and Brazil. The indexes are now at high levels in all countries in the sample apart from Mexico, which is also the only country in the sample with a revealed comparative disadvantage in agriculture. Finally, before examining the region's policy reforms, it is important to note the increases in export orientation. A common indicator is the value of goods and services expressed as a percentage of GDP. Since the early 1990s, this indicator has roughly doubled in the three largest economies (Argentina, Brazil, and Mexico) but changed little in the other countries in our sample, apart from Chile, where it rose a few years earlier. Another indicator is the share of primary agricultural pro- duction that is exported. This share has jumped dramatically in the past 20 years, including in Mexico, where it is now more than 30 percent as a result of sharply increased specialization within the sector following the agricultural and trade policy reforms begun in anticipation of the North American Free Trade Agreement (NAFTA), which came into effect in 1994. Note, however, that import dependence has also grown as a consequence of trade specialization. Indeed, 17 of the region's 21 countries on which data are available are net food importers (de Ferranti et al. 2005). Only Argentina was a net exporter of cereals during 2003­05, even though all eight countries in our sample (except Mexico) are more than 100 percent self- sufficient in agricultural products in aggregate and even though the share of these countries in global exports of agriculture and food jumped from 6.8 to 9.6 percent between 1990­94 and 2000­04. 294 Distortions to Agricultural Incentives: A Global Perspective The Evolution of Agricultural and Trade Policies Like most other regions, Latin American countries encompass a diverse range of policies, political structures, and institutions, though there has been, to some extent, a common evolution in the ideology motivating economic policies, begin- ning in the 1960s. Prior to the reforms of the mid-1980s and early 1990s Until approximately the mid-1980s, agricultural price interventions in the region were largely a by-product of a development strategy based on a claim that the best way to grow the economy was to adopt a protectionist policy to encourage import-substitution industrialization. This policy also raised budgetary resources in the form of import tax revenue, which was supplemented in some countries (such as Argentina) through agricultural export taxes. Both of these approaches harmed the region's most competitive farmers and were offset only slightly by farm credit and fertilizer subsidies. Between the 1950s and the 1980s, there were concerns about high rates of infla- tion, especially where urban populations had strong political influence. Policy makers were under pressure to avoid large increases in food prices, which would potentially impact wage rates and thereby (according to theory prevailing at the time) accelerate inflation through the so-called cost-push effect. In addition to fiscal and inflation objectives that made farm export taxes attractive, there was, in the 1950s and 1960s, a widespread belief among the region's policy makers and followers of the structuralist school associated with Prebisch (1950, 1959, 1964)--notwithstanding the seminal book by Schultz (1964)--that the efficiency losses generated through the extraction of rents in agriculture were low and that the main impact would be to reduce land rents and land values. Argentina is a prime example of a case in which the view persisted that farmers in Latin America were unresponsive to price incentives. While the belief in this unresponsiveness has now largely disappeared, a few countries-- Argentina is one--still tax agricultural exports to generate fiscal revenues and lower consumer food prices. An empirical study of agricultural pricing policies led by Krueger, Schiff, and Valdés (1991) included five Latin American countries for the period 1960­84. Its main findings are fourfold. First, over the period examined and for the farm prod- ucts selected, the direct interventions affecting importables were positive, on aver- age, while the direct interventions on exportables were negative. Second, aggregating over all selected products, one sees that the net effect was negative, indicating that the direct tax on exportables dominated the protection on importables. Third, the Latin America and the Caribbean 295 rate of indirect taxation on agriculture (because of industrial protection policies and the overvaluation of the real exchange rate) was large and dominated the rate of direct taxation. And fourth, direct price policies stabilized agricultural prices relative to world prices, while indirect policies contributed little, if at all, to food price stability. The study found that direct protection for agricultural importables averaged 13 percent, while for exportables, it amounted to ­6 percent. The indi- rect taxation rate in the region averaged 21 percent so that the total taxation rate (direct and indirect) averaged 28 percent. The highest direct taxation was found in Argentina and the Dominican Republic (about 18 percent). As a percent of agri- cultural GDP, net income transfers out of agriculture (direct and indirect) reached 84 percent in Argentina, 56 percent in Chile, 43 percent in the Dominican Republic, and 42 percent in Colombia. Economic reforms from the mid-1980s to early 1990s By the 1980s, there was disillusionment with the results of the import-substitution strategy and wider acceptance of theoretical developments regarding the causes of inflation and macroeconomic instability in general. During the 1980s and early 1990s, a macroeconomic framework designed for open economies gradually dis- placed the closed economy approach in most Latin American countries. Govern- ments introduced economy-wide reforms with special emphasis on macroeco- nomic stabilization, deregulation, unilateral trade liberalization, and privatization. The goal of the reformers was to create a better climate for productivity and private investment in all economic sectors, including agriculture. In most Latin American countries, the major change in trade policy was the partial or total removal of most quantitative restrictions on imports and exports, the elimination of export taxes, and a program of gradual reduction in the levels of import tariffs. This yielded incentives to move resources from import-competing to export- oriented sectors, including in agriculture, which enhanced competitiveness and led to greater integration with the world economy. By the mid-1990s, the exchange rate was recognized as the most important "price" affecting the agricultural economy. At the outset of the reforms, it was expected that trade liberalization and the reduction of the fiscal deficit would lead to a depreciation of the real exchange rate (Krueger, Schiff, and Valdés 1988). Yet the reforms were followed by a significant appreciation of the currency that was associated with the opening of the capital account, greater inward foreign invest- ment, and a major increase in domestic real interest rates. Reforms in the service sector also played a critical role. Deregulation and privatization had a major impact on the availability in the marketplace of the more-reliable and lower-cost services used in agriculture, such as ports, airlines, and shipping transport. 296 Distortions to Agricultural Incentives: A Global Perspective The timing of reforms differed somewhat across countries, however. Colombia, for example, became a more open economy through export promotion beginning in 1967; it adopted a more ambitious liberalization of trade in 1990 and then went into a policy reform reversal beginning in 1992. In Chile, the controlled markets of 1950 to 1974 were followed by radical economic reforms toward trade liberaliza- tion, deregulation, and privatization between 1978 and 1982, before a second phase of reforms beginning in 1984. Mexico, meanwhile, introduced strong policy changes starting in the mid-1980s, before the signing of NAFTA. The changes involved more openness, deregulation, and privatization; a reduction in credit subsidies; and major changes in the role of government in the marketing of farm products. A wide variety of policy instruments have been applied to influence agricul- tural prices, even during the postreform period. Colombia, for example, has had minimum support prices, in addition to import tariffs, price compensation schemes, procurement agreements, a monopoly on grain imports by a govern- ment agency, export licenses and subsidies, and safeguards on imports. Moreover, until 1990, all imports of inputs were subject to prior import licenses. Then, in 1995, tariffs and tariff surcharges associated with price bands on more than 100 products were introduced. Mexico is another leader in interventions, includ- ing in the transition from highly government-controlled markets before the mid- 1980s to more market-oriented policies. Its policies include price support pro- grams (before the mid-1980s and in conjunction with state trading), credit and input subsidies, and direct income payments to farmers (ProCampo). Argentina, however, has undertaken simpler interventions. Agricultural exportables that are also wage goods have been subjected to export taxes, complemented by export bans in some years. The return to sizeable export taxes in late 2001 and their sub- sequent rises has been controversial, with the most recent rises leading to pro- longed protests by farmers in urban areas in mid-2008. Estimates of Latin American Policy Indicators The net effect of these various interventions on farmer and consumer incentives are quantified using the common methodology (Anderson et al. 2008a, 2008b) that has been adopted by the authors of this volume and the four preceding regional volumes. After a brief word on methodology, a summary of results follows.2 Methodology The NRA is defined as the percentage by which government policies have raised gross returns to producers above what they would be without the government's Latin America and the Caribbean 297 intervention (or lowered them, if the NRA is below zero). If a trade measure is the sole source of government intervention, the measured NRA will also be the con- sumer tax equivalent (CTE) rate at that same point in the value chain. The NRAs are based on estimates of assistance to individual industries at farmgate. The tar- geted degree of coverage of the products for which agricultural NRA estimates are generated is 70 percent of the gross value of farm production at undistorted prices. The authors of the country case studies also provide guesstimates of the NRAs for noncovered farm products. For countries with non-product-specific (NPS) agricultural subsidies or taxes, such net subsidies are then added to product-specific assistance to obtain NRAs for total agriculture and also for trad- able agriculture for use in generating an RRA. Farmers are affected not only by the prices of their own outputs, but also (albeit indirectly, because of the changes to factor market prices and the exchange rate) by the incentives nonagricultural producers face. In other words, both absolute and relative prices--and hence, relative rates of government assistance-- affect producer incentives. The direction of the economy-wide effect of distor- tions to agricultural incentives may be captured by the extent to which the trad- able parts of agricultural production are assisted or taxed relative to producers of other tradables. By generating estimates of the average NRA for nonagricultural tradables, it is then possible to calculate an RRA, which is defined in percentage terms as: RRA 100[(1 NRAagt 100) (1 NRAnonagt 100) 1], where t t NRAag and NRAnonag are the weighted average percentage NRAs for the trad- able parts of the agricultural and nonagricultural sectors, respectively. Since the NRA cannot be less than 100 percent if producers are to earn anything, neither can the RRA. If both sectors are assisted equally, the RRA is zero. Although this measure cannot fully capture the ultimate impacts on resource allocations to various sectors including nontradables (a computable general equilibrium model is needed for that), it is nonetheless useful in comparing policy biases across time and countries. If the RRA is below (above) zero, it indicates that a country's trade policy regime has an antiagricultural or proagricultural bias. In calculating the NRA for producers of agricultural and nonagricultural trad- ables, the methodology seeks to include distortions generated by dual or multiple exchange rates. Such direct interventions in the market for foreign currency were common in Latin America in the 1970s and 1980s, but not since the reforms. However, some authors of the Latin American country studies have had difficulty finding an appropriate estimate of the extent of this distortion, so the impact on NRAs has been included only for the Dominican Republic, Ecuador, and Nicaragua. Its exclusion for the other five countries means the estimated (typi- cally) positive NRAs for importables and (typically) negative NRAs for exporta- bles are smaller than they should be for these countries. In cases where the NRA 298 Distortions to Agricultural Incentives: A Global Perspective for importables dominates that for exportables, this omission would lead to an underestimate of the average (positive) NRA for such tradables sectors. This applies to nonagricultural sectors for all the countries studied in this chapter. In the most common cases in earlier decades, where the estimated NRA for importa- bles is dominated by a negative NRA for exportables in the farm sector, the esti- mate of the sectoral average NRA for agriculture would be less negative than it should be, and, hence, so would the RRA estimate.3 To obtain the values for farmer assistance and consumer taxation, the NRA estimates of the country authors have been multiplied by the gross value of pro- duction at undistorted prices to obtain an estimate in constant U.S. dollars of the direct gross subsidy equivalent (GSE) of assistance to farmers. This GSE value is then added up across products for each country and then across countries for any or all products to obtain regional aggregate transfer estimates for the countries being examined. An aggregate estimate for the rest of the region is obtained by assuming that the weighted average NRA for the countries not within the study is the same as the weighted average NRA for the countries within the study and that the share of each country in the region's gross value of farm production at undistorted prices is the same each year as the share of the country in the region's agricultural GDP measured at distorted prices. These GSE values are also expressed on a per farm worker basis. To obtain comparable value estimates of the consumer transfer, the CTE esti- mate at the point at which a product is first traded is multiplied by consumption (obtained from the FAO SUA-FBS Database), valued at undistorted prices, to obtain an estimate in constant U.S. dollars of the tax equivalent to consumers of primary farm products. These, too, are summed across products for a country and across countries for any or all products to obtain regional aggregate transfer esti- mates for the countries under study. Estimates of NRAs in agriculture On average (whether simple or weighted), agricultural price and trade policies in Latin America reduced farmer earnings throughout the postwar period right through to the 1980s. The extent of these (when expressed as a nominal tax equiv- alent) peaked at more than 20 percent in the 1970s but still averaged close to 10 percent in the late 1980s. The only countries in our sample that received posi- tive assistance from farm policies during that period were Chile (from the late 1970s, but only to a minor extent), Colombia, and Mexico. Argentina, Brazil, the Dominican Republic, and Ecuador each had negative rates of assistance that aver- aged well above 20 percent for at least one five-year subperiod, and, apart from the Dominican Republic, each had a negative average NRA even in the 1990s, as did Latin America and the Caribbean 299 Nicaragua. However, by the mid-1990s, Brazil and the Dominican Republic joined Chile and Colombia in that they had positive average NRAs. Meanwhile, Mexico had raised its assistance considerably before engaging in reform following negoti- ations to join the World Trade Organization and the NAFTA, while Argentina had all but eliminated its discrimination against its exporters in the 1990s, only to reinstate explicit export taxes again in late 2001 when it abandoned its fixed exchange rate with the U.S. dollar and nominally devalued by two-thirds. The NRAag for the region in the 1990s and the first half of the present decade averaged slightly under 5 percent (table 7.2). Its switch from negative to positive occurred in 1992. The effect of the policy reforms on NRAs over the past two decades is illus- trated in figure 7.1. For all countries except Chile, the national average NRA was less negative or more positive in 2000­04 than in 1980­84. This is true, too, for the majority of the commodity NRAs for the region, although assistance for several commodities (such as milk and poultry) was cut. This pattern may be seen in figure 7.2 and table 7.3, which also illustrate the diversity of the region's average rates across commodities. There is also a great deal of diversity across commodities within each country's farm sector. The extent of this diversity (as measured by the standard deviation) diminished, on average, by only about one-quarter during 1990­2004 compared with the prereform period of 1965­89. This is evident in table 7.4. The table reports the standard deviation of NRAs for covered products, which account for more than two-thirds of the value of agricultural production. This means there is still a great deal that may be gained in terms of improved resource reallocation within the agricultural sector if differences in rates of assistance for different industries are reduced. One striking feature of the pattern of farm price distortions in Latin America and the Caribbean as a whole is the strong antitrade bias. This is shown for agri- culture's import-competing and export subsectors in the region in figure 7.3 and for each country in table 7.5 (along with a trade bias index, or TBI). These esti- mates reveal that there has been little decrease in the bias over the past four decades, except in Brazil. Indeed, the average NRA for exportable farm products has been negative throughout virtually the whole period analyzed in all countries other than Chile (plus Brazil in the 1990s and Colombia since 2000), while the regional average NRA for import-competing farm industries has increased from near zero in the 1970s to 20 percent or more in the period since 1990 (with Chile again an exception with its NRA for import-competing industries falling to near zero). That is, despite the lower taxation of farm export industries, the region's antitrade bias has persisted because the average NRA for import-competing farm products has been rising recently in several of the countries being examined. 300 Table 7.2. NRAs to Agriculture,a Latin American Countries, 1965­2004 (percent) 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 Argentina 22.7 22.9 20.4 19.3 15.8 7.0 4.0 14.9 Brazilc 6.1 27.3 23.3 25.7 21.1 11.3 8.0 4.1 Chile 16.2 12.0 4.5 7.2 13.0 7.9 8.2 5.8 Colombia 4.7 14.8 13.0 5.0 0.2 8.2 13.2 25.9 Dominican Republic 5.0 17.5 21.2 30.7 36.4 1.0 9.2 2.5 Ecuadorc 9.6 22.4 15.0 5.9 1.0 5.3 2.0 10.1 Mexico -- -- -- 2.9 3.0 30.8 4.2 11.6 Nicaraguac -- -- -- -- -- 3.2 11.3 4.2 Latin American focus countries Unweighted averageb 2.8 15.5 14.5 7.7 8.3 2.3 3.2 4.9 Weighted averagea 7.2 21.0 18.0 12.5 10.9 4.2 5.5 4.8 Dispersion of individual country average NRAsd 13.8 15.4 10.8 17.4 17.1 13.5 8.6 11.9 Source: Anderson and Valenzuela (2008), based on estimates reported in Anderson and Valdés (2008a). Note: -- not available. a. Weighted average for each country, including product-specific input distortions and NPS assistance as well as authors' guesstimates for noncovered farm products, with weights based on gross value of agricultural production at undistorted prices. b. The unweighted average is the simple average across the eight countries of their national NRA (weighted) averages. c. Ecuador and Brazil in the 1965­69 column refer to 1966­69 data; and Nicaragua 1990­94 column to 1991­94 data. d. Dispersion of average NRAs across countries is a simple five-year average of the annual standard deviation around a weighted mean of the national agricultural sector NRA each year. Latin America and the Caribbean 301 Figure 7.1. NRAs to Agriculture, Individual Latin American Countriesa and Unweighted Regional Average, 1980­84 and 2000­04 40 20 percent 0 20 40 a o r ile av a il bl n ua a do az bi d ric in ic pu ica Ch ag e) N c ex nt m Br ua te e i in gh Am ar lo ge M Ec om ic Co Ar ei in Re D nw at (u L 1980­84 2000­04 Source: Anderson and Valenzuela (2008), based on estimates reported in Anderson and Valdés (2008a). a. There are no estimates for Nicaragua for 1980­84. The contributions to the overall NRA for agriculture for the region as a whole provided by covered products, noncovered products, and NPS assistance are sum- marized in table 7.5. NPS assistance has added only one or two percentage points during the past four decades. Input price distortions have also contributed little, on average, to the overall regional NRA in agriculture, reducing the negative value slightly in the 1980s and adding slightly to the positive value during the past decade or so. In Chile, input distortions have reduced the positive NRA in the farm sector because of protectionist policies that have raised the price of imported or import-competing farm inputs. This has also been the case in Argentina since the early 1990s and, to a smaller extent, in Colombia since the 1960s. There is little in the way of domestic producer subsidies or taxes, on average, in the region; the main exception is positive support measures in Mexico and slightly negative support measures in Argentina. The dollar value of the positive or negative assistance to farmers arising from agricultural price and trade policies has been nontrivial. The antiagricultural bias Figure 7.2. NRAs, by Product, Latin American Countries, 1980­84 and 2000­04 milk rice sugar poultry cotton pig meat coffee wheat beef maize cocoa soybeans eggs 80 60 40 20 0 20 40 60 80 100 120 percent, weighteda average across countries 1980­84 2000­04 Source: Anderson and Valenzuela (2008), based on estimates reported in Anderson and Valdés (2008a). a. Weights based on gross value of agricultural production at undistorted prices--that is, each NRA (by country, by product) is weighted by the country's value of production of that commodity in a given year. Products with less than 1 percent of the gross value of regional production are excluded. These include: apples, cassava, cocoa, garlic, onions, palm oil, peanuts, and sesame. Figure 7.3. NRAs to Exportable, Import-Competing, and Alla Agricultural Products, Latin American Region, 1965­2004 30 percent, weighted averages 20 across eight countries 10 0 10 20 30 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 import-competing total exportables Source: Anderson and Valenzuela (2008), based on estimates reported in Anderson and Valdés (2008a). Table 7.3. NRAs, Key Covered Farm Products, Latin American Focus Countries,a 1965­2004 (percent) 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 Rice 27 5 10 5 8 12 26 34 Wheat 2 15 11 6 6 18 3 2 Maize 10 8 15 5 13 0 4 3 Other grains 4 3 4 6 2 0 14 11 Soybeans 3 5 15 11 21 10 4 10 Other oilseeds 4 3 15 21 23 11 16 21 Sugar 17 61 46 54 43 20 7 27 Cotton 7 2 14 16 23 12 6 11 Coffee 27 26 32 42 29 1 9 3 Cocoa 6 16 13 4 14 16 12 7 Fruits and vegetables 12 22 31 5 33 16 24 20 Beef 23 21 11 10 4 2 5 1 Pig meat 6 14 13 19 20 6 3 4 Poultry 110 144 108 33 23 23 8 19 Eggs -- -- -- 0 6 2 16 16 Milk 2 7 19 104 70 45 29 45 All covered products 13.0 25.1 19.6 14.6 14.3 0.9 0.8 2.7 Sources: Anderson and Valenzuela (2008) based on estimates reported in Anderson and Valdés (2008a). Note: -- not available. 303 304 Table 7.4. Dispersion of NRAs across Covered Agricultural Productsa within Latin American Focus Countries, 1965­2004b (percent) 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 Argentina 18.5 17.8 19.9 15.7 12.1 7.1 9.4 12.6 Brazil 28.1 37.2 41.0 35.9 25.5 27.4 8.5 7.6 Chile 33.0 37.2 30.4 17.0 26.1 16.5 14.7 13.3 Colombia 34.8 21.2 29.9 42.5 34.1 27.2 31.0 46.0 Dominican Republic 86.5 64.0 89.3 83.0 102.3 137.1 92.6 132.8 Ecuador 99.0 88.6 104.8 106.2 48.5 18.8 27.9 29.6 Mexico -- -- -- 71.9 60.1 57.7 30.6 41.1 Nicaragua -- -- -- -- -- 40.1 35.7 27.7 Latin American focus countries Unweighted averagec 50.0 44.3 52.5 53.2 44.1 41.5 31.3 38.8 Product coveraged 54 65 68 71 68 65 69 70 Source: Anderson and Valenzuela (2008), based on estimates reported in Anderson and Valdés (2008a). Note: -- not available. a. Dispersion for each country is a simple five-year average of the annual standard deviation around a weighted mean of NRAs across covered products each year. b. The 1965­69 column for Ecuador and Brazil refers to 1966­69 data; the 1990­94 Nicaragua column to 1991­94 data. c. The unweighted average is the simple average across the eight countries of their five-year simple average dispersion measures. d. Share of gross value of total agricultural production at undistorted prices accounted for by covered products in the region. Table 7.5. NRAs to Agricultural Relative to Nonagricultural Industries, Latin American Region, 1965­2004 a. Unweighted averages for eight focus countries (percent) 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 a NRA, covered products 9.1 21.8 17.0 8.8 8.9 1.0 1.1 4.4 NRA, noncovered products 0.5 9.2 10.0 6.5 7.5 1.4 0.9 0.4 NRA, all agricultural productsa 5.4 17.0 15.0 8.3 9.3 0.4 0.7 2.7 Total agricultural NRA (including NPS)b 2.8 15.5 14.5 7.7 8.3 2.3 3.2 4.9 TBIc 0.22 0.18 0.31 0.41 0.33 0.26 0.25 0.26 NRA, all agricultural tradablesb 6.0 19.0 16.4 7.2 8.2 2.6 3.5 5.7 NRA, all nonagricultural tradables 16.8 20.6 15.6 14.3 13.4 7.7 7.3 6.5 RRAd 19.5 32.9 27.7 18.8 19.1 4.8 3.5 0.8 b. Weighted averages for eight focus countries (percent) 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 NRA, covered productsa 13.0 25.1 19.6 14.6 14.3 0.9 0.8 2.7 NRA, noncovered products 3.3 15.5 15.0 10.9 13.1 0.7 3.8 2.1 NRA, all agricultural productsa 8.6 21.7 18.1 13.6 14.0 0.8 1.7 2.5 Total agricultural NRA (including NPS)b 7.2 21.0 18.0 12.5 10.9 4.2 5.5 4.8 TBIc 0.20 0.25 0.26 0.36 0.29 0.25 0.14 0.21 NRA, all agricultural tradablesb 9.3 23.0 19.0 12.9 11.2 4.4 5.5 4.9 NRA, all nonagricultural tradables 15.9 27.8 23.3 18.5 16.8 7.3 6.6 5.5 RRAd 21.4 39.8 34.2 26.6 24.0 2.7 1.0 0.6 Source: Anderson and Valenzuela (2008), based on estimates reported in Anderson and Valdés (2008a). a. NRAs including product-specific input subsidies. b. NRAs including NPS assistance, that is, the assistance to all primary factors and intermediate inputs as a percentage of the total primary agricultural production valued at undistorted prices. c. TBI (1 NRAagx 100)/(1 NRAagm 100) 1, where NRAagm and NRAagx are the average percentage NRAs for the import-competing and exportable parts of the agricul- 305 tural sector. The regional average TBI is calculated from the regional averages of the NRAs for exportable and import-competing parts of the agricultural sector. d. RRA is defined as 100*[(100 NRAagt) (100 NRAnonagt) 1], where NRAagt and NRAnonagt are the percentage NRAs for the tradable parts of the agricultural and nonagricultural sectors, respectively. 306 Distortions to Agricultural Incentives: A Global Perspective peaked for the region in the 1975­84 period at nearly US$17 billion per year in constant 2000 dollar terms, assuming that the Latin American countries not under study had the same NRAs as the countries under study, keeping aside the case of Mexico (see the bottom row of table 7.6a). This is equivalent to a gross tax of almost US$400 for each person engaged in agriculture. Around 60 percent of this US$17 billion arose because of policies in Brazil. Thanks to the reforms of the past two decades, this taxation has gradually disappeared in all the countries under study except Argentina and Nicaragua. However, the reform does not mean that there is now no intervention. Rather, the old policy has been replaced by pos- itive assistance to farmers in the remaining six countries. This assistance has aver- aged US$6 billion per year, or around US$140 per farmworker, over the 1995­ 2004 period. The US$140 is small compared with per capita income for the region (about 4 percent), but it ranges from more than US$450 in Colombia (one-quar- ter of that country's per capita GDP in 2000­04) to US$1,700 in Argentina (negative one-third of that country's per capita GDP). The extent of this dramatic transformation in the region as a whole over the past two decades is illustrated in figure 7.4 for the individual countries and for key products. Table 7.7 reveals that, as in most other regions of the world, the lion's share of assistance goes to milk, sugar, and rice. Assistance to nonfarm sectors and RRAs The antiagricultural policy bias of the past was caused not merely by agricultural policies. The significant reduction in border protection for the manufacturing sector and the indirect impact of this on the drop in the price of nontradables after the initiation of the reforms, together with the deregulation and privatization of services, have also been important in the changes in the incentives affecting intersectorally mobile resources. The reduction in assistance to nonfarm tradable sectors has been as responsible for the expansion in agricultural exports since the early 1990s as the reduction in direct taxation on these agricultural exports. Quantifying this distortion in nonfarm tradable sectors as accurately as the quantification of the distortion in agriculture has not been possible. Our authors have had to rely on applied trade taxes (for exports, as well as imports) rather than undertaking price comparisons for nonfarm goods, and, hence, they have not cap- tured the quantitative restrictions on trade that were important in earlier decades but that have been less important recently.4 Nor have they captured distortions in the services sectors; many of these sectors now produce tradables (or would do so in the absence of interventions preventing the emergence of this production). As a result, the NRAs for nonfarm importables are underestimated, and the decline indicated is less rapid than the decline that actually occurred; the situation is Latin America and the Caribbean 307 Figure 7.4. GSEs of Assistance to Farmers, Latin American Countries, 1980­84 and 2000­04 a. Total per country 4,000 2,000 constant 2000 US$ millions 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 a ua bl n ile r zil a o do in bi ic pu ica a Ch ag ic ex nt m Br ua Re in ar lo ge M Ec om ic Co Ar N D b. Total per product milk poultry sugar rice pig meat cotton coffee wheat cocoa other grains other oilseeds beef eggs maize fruits and vegetables soybeans 0 0 0 00 0 0 00 0 0 00 00 00 00 00 00 00 0 0 0 1, 2, 3, 6, 5, 4, 3, 2, 1, constant 2000 US$ millions 1980­84 2000­04 Source: Anderson and Valenzuela (2008), based on estimates reported in Anderson and Valdés (2008a). Table 7.6. GSEs of Assistance to Farmers, Total and Per Farm Worker, Latin American Countries,a 1965­2004 308 a. Total (constant 2000 US$ millions) 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 Argentina 1,699 2,630 2,466 2,850 1,533 738 595 2,473 Brazil 790 7,905 8,141 12,724 9,142 3,578 3,101 1,509 Chile 482 378 167 267 394 380 465 294 Colombia 358 1,555 1,719 583 5 905 1,562 1,835 Dominican Republic 61 457 603 694 561 22 150 39 Ecuador 192 477 453 121 23 132 68 324 Mexico -- -- 389 1,581 762 7,426 984 2,805 Nicaragua -- -- -- -- -- 32 140 54 Latin American focus countries 2,496 12,647 13,604 13,716 10,098 4,210 5,459 4,279 All Latin American countriesa 3,082 15,613 16,794 16,933 12,467 5,197 6,740 5,283 b. Per person engaged in agriculture (constant 2000 US$) 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 Argentina 1,094 1,776 1,727 2,030 1,054 498 404 1,693 Brazil 51 482 475 736 561 240 224 118 Chile 650 515 216 324 442 401 478 299 Colombia 119 483 483 153 1 244 419 496 Dominican Republic 88 641 859 1,003 803 33 238 66 Ecuador 199 475 446 114 20 108 54 260 Mexico -- -- 51 194 90 867 115 329 Nicaragua -- -- -- -- -- 81 351 137 Latin American focus countries 85 411 417 408 305 132 177 144 All Latin American countriesa 81 390 396 386 283 119 156 124 Source: Anderson and Valenzuela (2008), based on estimates reported in Anderson and Valdés (2008a). Note: -- not available. a. Assumes the rate of assistance in nonfocus countries is the same as the average for the focus Latin American countries excluding Mexico, and that their share of the value of Latin American and Caribbean (excluding Mexican) agricultural production at undistorted prices is the same as their average share of the region's agricultural GDP at distorted prices during 1990­2004, which was 23 percent. Farmer numbers are from FAOSTAT and may differ from national statistics. Table 7.7. GSEs of Policies Affecting Farmers in Latin America, by Product and Subsector, 1965­2004 a. By product (at undistorted farmgate prices, US$ millions) Rice Wheat Maize Other grains Soybean Other oilseeds Sugar Cotton 1965­69 24 17 92 0 1 0 8 19 1970­74 40 216 162 1 55 0 1,829 8 1975­79 230 91 475 56 436 81 1,619 159 1980­84 55 116 396 53 428 110 3,260 156 1985­89 55 65 707 10 1,533 151 1,980 380 1990­94 201 395 17 5 386 92 988 158 1995­99 569 79 373 151 279 256 233 36 2000­04 614 30 307 113 1,371 241 970 78 Fruits and All covered Cocoa Coffee vegetables Beef Pig meat Poultry Eggs Milk products 1965­69 1 127 19 289 1 10 -- 2 516 1970­74 8 169 41 440 4 15 -- 29 2,987 1975­79 32 815 163 404 53 116 51 236 4,131 1980­84 8 3,014 165 1,027 565 423 14 1,603 7,003 1985­89 17 1,738 623 327 504 344 66 944 6,716 1990­94 14 30 610 188 93 533 19 1,471 661 1995­99 10 536 977 704 110 378 225 1,393 476 2000­04 7 76 750 264 111 1,048 285 1,915 1,504 (Table continues on the following page.) 309 310 Table 7.7. GSEs of Policies Affecting Farmers in Latin America, by Product and Subsector, 1965­2004 (continued) b. By subsector (at undistorted farmgate prices, US$ billions) Total GSE, all direct assistance to farmersa GSE for just covered GSE for just non- Import- farm productsb covered farm products Total Exportables competing Nontradables 1965­69 0.5 0.1 0.6 0.7 0.1 0.0 1970­74 3.0 1.1 4.0 3.9 0.2 0.0 1975­79 4.0 1.5 5.5 5.5 0.0 0.0 1980­84 7.0 2.2 8.5 12.1 2.9 0.0 1985­89 6.7 3.1 7.5 10.7 0.9 0.0 1990­94 0.7 0.4 3.8 4.6 5.7 0.0 1995­99 0.5 1.2 5.3 2.3 3.9 0.0 2000­04 1.5 0.6 4.3 3.3 5.4 0.0 Source: Anderson and Valenzuela (2008), based on estimates reported in Anderson and Valdés (2008a). Note: -- not available. a. GSEs including assistance to nontradables and NPS assistance. b. GSEs including product-specific input subsidies. Latin America and the Caribbean 311 similar for nonfarm exportables, except that the actual NRAs would have been negative in most cases. Of these two elements of underestimation, the former bias probably dominated. Thus, the author estimations of the overall NRA for nonagricultural tradables should be considered a lower-bound estimate; this is especially true as one looks back in time, so that the decline indicated in the NRA estimates is less rapid than it actually would have been.5 Despite these methodological limitations, the estimated NRAs for nonfarm tradables prior to the 1990s are sizeable. For Latin America as a whole, the average value of the NRAs for nonfarm tradables has steadily declined throughout the past four decades as policy reforms have spread. This has therefore contributed to a decline in the estimated RRA among farmers. Thus, the RRA has fallen from below 30 percent in the 1970s to an average of 1 percent in 2000­04 (see table 7.5), and this appears (in figure 7.5) to have been caused as much by falling positive NRAs among nonfarm producers as by falling negative NRAs among farmers. The extent of the change in RRAs among individual countries over the past two decades is striking, particularly in the case of Brazil and the Dominican Republic (where negative RRAs virtually disappeared) and of Colombia (where RRAs switched from negative to positive). In figure 7.6, this is depicted by Figure 7.5. NRAs to Agricultural and Nonagricultural Tradable Products and RRA,a Latin American Region, 1965­2004 40 30 percent, weighted averages across eight countries 20 10 0 10 20 30 40 50 1965­69 1970­74 1975­79 1980­84 1985­89 1990­94 1995­99 2000­04 NRA, nonagricultural tradables NRA, agricultural tradables RRA Source: Anderson and Valenzuela (2008) based on estimates reported in Anderson and Valdés (2008a). a. The RRA is defined as 100*[(100 NRAagt ) (100 NRAnonagt) 1], where NRAagt and NRAnonagt are the percentage NRAs for the tradable parts of the agricultural and nonagricultural sectors, respectively. 312 Distortions to Agricultural Incentives: A Global Perspective Figure 7.6. Relationship between RRA and the TBI for Agriculture, Latin American Focus Countries, 1980­84a and 2000­04 a. 1980­84 0.3 0.2 0.1 Ecuador Chile 0 Mexico RRA Nicaragua 0.1 Colombia 0.2 Argentina 0.3 Dominican Republic 0.4 Brazil 0.5 0.6 0.5 0.4 0.3 0.2 0.1 0 TBI b. 2000­04 0.3 Colombia 0.2 0.1 Mexico Chile Ecuador 0 Dominican Republic Brazil RRA Nicaragua 0.1