34772 Poverty& THE WTO IMPACTS OF THE DOHA DEVELOPMENT AGENDA Editors Thomas W. Hertel · L. Alan Winters POVERTY AND THE WTO POVERTY AND THE WTO Impacts of the Doha Development Agenda Edited by Thomas W. Hertel and L. Alan Winters A copublication of Palgrave Macmillan and the World Bank ©2006 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 09 08 07 06 A copublication of The World Bank and Palgrave Macmillan. Palgrave Macmillan Houndmills, Basingstoke, Hampshire RG21 6XS and 175 Fifth Avenue, New York, NY 10010 Companies and representatives throughout the world Palgrave Macmillan is the global academic imprint of the Palgrave Macmillan division of St. Martin's Press, LLC and of Pal- grave Macmillan Ltd. Macmillan® is a registered trademark in the United States, United Kingdom and other countries. Palgrave is a registered trademark in the European Union 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 permis- sion 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-10: 0-8213-6314-X (softcover) ISBN-10: 0-8213-6370-0 (hardcover) ISBN-13: 978-0-8213-6314-0 eISBN-10: 0-8213-6315-8 eISBN-13: 978-0-8213-6315-7 DOI: 10.1596/978-0-8213-6314-0 Library of Congress Cataloging-in-Publications Data Poverty and the WTO : Impacts of the Doha Development Agenda / edited by Thomas W. Hertel, L. Alan Winters. p. cm. ­ (Trade and development series) Includes bibliographical references and index. ISBN-13: 978-0-8213-6314-0 ISBN-10: 0-8213-6314-X 1. World Trade Organization--Developing countries. 2. International trade. 3. Poverty--Developing coun- tries. 4. Developing countries--Economic conditions. 5. Developing countries--Economic policy. I. Hertel, Thomas W. (Thomas Warren), 1953- II. Winters, L. Alan. III. Series. HF1385.P68 2005 339.4'6091724--dc22 2005044583 Contents Acknowledgments xv Contributors xvii Abbreviations and Acronyms xix Part I EVALUATION OF THE DOHA DEVELOPMENT AGENDA 1 Poverty Impacts of a WTO Agreement: Synthesis and Overview 3 Thomas W. Hertel and L. Alan Winters 2 Scenarios for Global Trade Reform 31 Kym Anderson and Will Martin 3 Assessing the World Market Impacts of Multilateral Trade Reforms 57 Thomas W. Hertel and Maros Ivanic Part II PRICE LINKAGES 105 4 Multilateral Trade Liberalization and Mexican Households: The Effect of the Doha Development Agenda 107 Alessandro Nicita 5 The Doha Trade Round and Mozambique 129 Channing Arndt v vi Contents Part III HOUSEHOLD IMPACTS OF PRICE CHANGES 6 The WTO Doha Round, Cotton Sector Dynamics, and Poverty Trends in Zambia 155 Jorge F. Balat and Guido G. Porto 7 The Doha Round, Poverty, and Regional Inequality in Brazil 183 Joaquim Bento de Souza Ferreira Filho and Mark Horridge 8 Growing Together or Growing Apart? A Village-Level Study of the Impact of the Doha Round on Rural China 219 Marijke Kuiper and Frank van Tongeren Part IV A FOCUS ON LABOR MARKETS 9 Structural Change and Poverty Reduction in Brazil: The Impact of the Doha Round 249 Maurizio Bussolo, Jann Lay, and Dominique van der Mensbrugghe 10 Impacts of the DDA on China: The Role of Labor Markets and Complementary Education Reforms 285 Fan Zhai and Thomas W. Hertel 11 The Social Impact of a WTO Agreement in Indonesia 319 Anne-Sophie Robilliard and Sherman Robinson Part V FISCAL REPLACEMENT OF LOST TARIFF REVENUE 341 12 The Poverty Impacts of the Doha Round in Cameroon: The Role of Tax Policy 343 Christian Arnault Emini, John Cockburn, and Bernard Decaluwé 13 Doha Scenarios, Trade Reforms, and Poverty in the Philippines: A CGE Analysis 375 Caesar B. Cororaton, John Cockburn, and Erwin Corong Part VI CROSS-COUNTRY ANALYSIS 403 14 The Effects of a Prospective Multilateral Trade Reform on Poverty in Developing Countries 405 Maros Ivanic Contents vii Part VIITHE DOHA DEVELOPMENT AGENDA, GROWTH, AND POVERTY 427 15 Implications of WTO Agreements and Unilateral Trade Policy Reforms for Poverty in Bangladesh: Short- versus Long-Run Impacts 429 Nabil Annabi, Bazlul Khondker, Selim Raihan, John Cockburn, and Bernard Decaluwé 16 The Impact on Russia of WTO Accession and the DDA: The Importance of Liberalization of Barriers against FDI in Services for Growth and Poverty Reduction 467 Thomas Rutherford, David Tarr, and Oleksandr Shepotylo 17 Global Impacts of the Doha Scenarios on Poverty 497 Kym Anderson, Will Martin, and Dominique van der Mensbrugghe INDEX 529 Box 1.1. Elements of the DDA Scenario Based on the July Framework Agreement 10 Figures 2.1 Converting the Harbinson Formula into a Tiered Formula 45 2.2 Tiered Tariff-Cutting Formula without Discontinuities 45 3.1 Validating the GTAP-AGR Model: Predicted versus Observed Standard Deviations for Wheat 62 3.2 Transmission of Global Results to a National Model 91 3A.1 Demand and Supply for a Single Export in Type A Model 94 3A.2 Demand and Supply for a Single Export in Type B Model 95 3A.3 Demand and Supply for a Single Export in a GTAP Simulation 96 3A.4 Demand and Supply for a Single Import 99 4.1 Household Income 112 4.2 Household Consumption 113 5.1 Distribution of Changes in Household Welfare 147 6.1 Dynamics of Cotton Income Shares 164 6.2 Dynamics of Cotton Income Shares, Central Province 165 6.3 Dynamics of Cotton Income Shares, Eastern Province 165 6.4 Dynamics of Cotton Income Shares, Southern Province 166 6.5 Cotton Prices and Household Income 169 viii Contents 6.6 Cotton Prices and Household Income, Regional Analysis 169 6.7 Relative Probability of Cotton Production 175 6.8 Relative Probability of Cotton Production, Regional Analysis 175 7.1 Brazilian States, Shaded According to Proportion in Poverty 193 8.1 Household Supply Response with Price Bands 224 8.2 Village Production Response under Alternative Liberalization Scenarios 239 9.1 Growth Incidence Curves, BaU Scenario: All, Agricultural, and Nonagricultural Households 264 9.2 Decomposition of Poverty Changes, BaU Scenario, All Households 267 9.3 Growth Incidence Curves for the BaU and Trade Scenarios, Poorest 30 Percent of All Households 275 10.1 Change in Sector Output, Full-Lib 306 10.2 Change in Sector Output, Doha 307 10.3a Impacts on Households, Full-Lib 311 10.3b Impacts on Households, Doha 311 10.4a Incremental Impacts on Households, Edu-Lib 314 10.4b Cumulative Impacts on Households, Edu-Lib 314 13.1 Income Distribution and Poverty: The Philippines (1985­2000) 382 14.1 Description of the Method 411 14.2 Decomposition of Poverty Effects of the Trade Reforms 424 15.1 Aggregate Welfare Effects 442 16.1 Production and Allocation of Output 475 16.2 Distributions of Estimated Welfare Gains from Russian WTO Accession for the Entire Sample, the Poorest Decile, and the Richest Decile 490 16.3 Distributions of Estimated Welfare Gains for Russian Households from ROW Free Trade for the Entire Sample, the Poorest Decide, and the Richest Decile 491 16.4 Distributions of Estimated Welfare Gains for Russian Households from Doha for the Entire Sample, the Poorest Decile, and the Richest Decile 492 16.5 Distributions of Estimated Welfare Gains: WTO Accession, Doha, and ROW Free Trade Model Results Compared 493 Tables 1.1 Poverty Impacts of a Prospective DDA 27 2.1 Sectoral Contributions to Comparative Static Estimates of Economic Welfare Gains from Completely Removing Merchandise Trade Barriers Globally, Post-Uruguay Round 33 2.2 Change in Exports Valued at 2001 Base Prices from Completely Removing Merchandise Trade Barriers Globally, Post- Uruguay Round 34 Contents ix 2.3 Sectoral and Regional Contributions to Comparative Static Estimates of Economic Welfare Gains from Completely Removing Merchandise and Services Trade Barriers Globally, Post- Uruguay Round 35 2.4 Key Features of Applied Agricultural Tariffs, by Country and Region, 2001 41 2.5 Bound and Applied Agricultural Tariff Rates, by Country and Region, 2001 42 2.6 Cuts in Domestic Support under a Tiered Formula with 75 Percent Cuts in High-Supporting Countries 47 2.7a Agricultural and Food Tariffs, Percent 49 2.7b Aggregate Merchandise Trade Tariffs, Percent 50 2.8a Agricultural and Food Tariffs Faced by Exporters, Percent 51 2.8b Aggregate Merchandise Trade Tariffs Faced by Exporters, Percent 52 3.1 Tariff Rates on Imports to and from the Focus Regions 64 3.2 Full Liberalization: Import Prices for All Regions 68 3.3 Full Liberalization: Export Prices for All Regions 72 3.4 Full Liberalization: Import Quantities for All Regions 74 3.5 Full Liberalization: Export Quantities for All Regions 76 3.6 Full Liberalization: Trade Balance 79 3.7 Doha: Import Prices for All Regions 82 3.8 Doha: Export Prices for All Regions 84 3.9 Doha: Import Quantities for All Regions 86 3.10 Doha: Export Quantities for All Regions 88 3A.1 Interaction of the GTAP Model with the Brazil Model 100 3A.2 Interaction of the GTAP Model with the China Model 101 4.1 Poverty in Mexico 110 4.2 Scenarios--Doha Implementations and Full Trade Liberalization 115 4.3 Pass-Through 118 4.4 Change in Real Income (Doha) 121 4.5 Change in Real Income ("Doha Plus") 122 4.6 Change in Real Income ("Doha Plus-Plus") 123 4.7 Improved Scenario: Variance of the Gains 124 4.8 Change in Real Income ("Doha Plus-Plus") 125 5.1 Components of GDP 136 5.2a Sectoral Shares in Value Added, Exports, and Imports 137 5.2b Sectoral Shares in Value Added, Exports, and Imports 138 5.3 Indications of Import Competition 139 5.4 Share of Value of Home Consumption in Total Consumption 140 5.5 Consumption by Quintiles 141 5.6a Export and Import Price Changes and Tariff Cuts for Simulations 142 5.6b Export and Import Price Changes and Tariff Cuts for Simulations 143 x Poverty and the WTO 5.7 Simulations 144 5.8 Macroeconomic Indicators 145 5.9 Equivalent Variation for Households 145 5.10 Microsimulation, Percentage Changes in Welfare by Quintile 146 6.1 Poverty in Zambia 161 6.2 Rural Poverty Trends: 1991, 1996, and 1998 161 6.3 Sources of Income in Rural Areas, 1998 162 6.4 Income Shares from Agricultural Activities, Rural Zambia, 1998 163 6.5 Income Gains from Cotton Production 173 6.6 Child Nutrition in Rural Areas (0­60 months old) 177 6.7 Effects on Child Nutrition and Education from Market Agriculture, Cotton versus Subsistence 179 7.1 Poverty and Income Inequality in Brazil, 2001 191 7.2 Share of Occupations in Each Activity's Labor Bill 194 7.3 Share of Each Activity in Total Labor Bill, by Occupation 197 7.4 Wage Bill Distribution According to Occupational Wages and Household Income Classes 200 7.5 Brazilian External Trade Structure 202 7.6 Shocks to the CGE Model 205 7.7 Selected Macroeconomic Results 208 7.8 Activity Level Variation by Industry: Percentage Change 209 7.9 Regional Results in 27 Regions 211 7.10 Average Household Income, CPI by Household Income Class and Gini Index Percentage Change 212 7.11 Percentage Changes in the Proportion of Poor Households (FGT0) and in the Poverty Gap Ratio (FGT1) by Household Income Groups 213 7.12 Percentage Changes in Number of Poor Households by Region and Total Number Change 214 8.1 Shocks Administered to Village Model 228 8.2 Activities and Income by Household Group 230 8.3 Substitution Elasticities for Cropping Activities 233 8.4 Household Production and Marketed Surplus with Full-Liberalization Scenario A 235 8.5 Household Production and Marketed Surplus with Full-Liberalization Scenario B 237 8.6 Household Production and Marketed Surplus with the Full-Liberalization Scenario 238 8.7 Equivalent Variation per Adult Equivalent by Household Group (in yuan) and as Percentage of Base Adult Equivalent Income 241 8.8 Contribution of Price Changes and Employment to Income Gains 242 Contents xi 9.1 Medium-Term Labor Market Structural Adjustments 261 9.2 The BaU Scenario's Output and Trade Sectoral Growth Rates and Employment Intensities 262 9.3 Poverty and Inequality in the BaU Scenario, by Sectors 263 9.4 Poverty and Inequality in a Distributionally Neutral Scenario 266 9.5 Trade Shock: Tariff Reductions and International Price Changes 269 9.6 Initial (Year 2001) Structure of the Brazilian Economy 271 9.7 Brazil's Structural Adjustment, Percent Changes in the Final Year between BaU Scenario and Trade Shocks 272 9.8 Factor Market Effects 274 9.9a Poverty and Inequality in the Doha Scenario, by Sector 277 9.9b Poverty and Inequality in the Full Liberalization, by Sector 278 9.10a Poverty Impact of Trade, by Migration Choices 279 9.10b Poverty Impact of Trade, Nonagricultural Stayers 279 9.10c Poverty Impact of Trade, Sectoral Movers 279 10.1 Elasticity Parameters for Trade and Sectoral Structure of Production and Trade 291 10.2 Substitution Elasticities in Production 294 10.3 Inputs from the Global Model 299 10.4 Aggregated Results 303 10.5 Effects on Inequality and Poverty 305 10.6 Sector Volume Impacts of Trade Liberalization: Percentage Deviation from Baseline 308 11.1 Trade Structure of Indonesia 322 11.2 A Macroeconomic SAM for Indonesia 327 11.3 Macroeconomic Results of Trade Liberalization Reference Scenarios 330 11.4 Social Impact of Trade Liberalization Reference Scenarios 331 11.5 Macroeconomic Results of Alternative Scenarios 334 11.6 Social Impact of Alternative Scenarios 335 11.7 Monte Carlo Simulations on the Social Impact of Alternative Labor Market Closures 338 12.1 Key Elasticities and Parameter Values in the Model 349 12.2 Changes in Selected Macro Variables 351 12.3 Sources of Household Factor Incomes 353 12.4 Poverty and Inequality Indexes before and after Simulations 354 12.5 Doha Scenario with VAT as the Replacement Tax: Sectoral Effects 356 12.6 ROW versus Domestic Liberalization: Sectoral Effects on Prices and Volumes of Goods and Services 359 13.1 Elasticities and Key Parameters, 1994 384 13.2 GTAP-Simulated World Prices and Demand Variations 386 xii Poverty and the WTO 13.3 Macro Effects 388 13.4 Effects on Prices and Volumes by Major Sector 390 13.5. Effects on Factor Remunerations 391 13.6 Sources of Household Income at the Base 392 13.7 Changes in Household Income and Sources 393 13.8 Poverty Indexes: Initial Value and Percent Change from Base under Alternative Scenarios 394 14.1 Regional Aggregation Used in the Model 407 14.2 Producer and Consumer Commodity Aggregation Used in the Model 408 14.3 Price and Income Changes (in percent) under Doha-SDT Deflated by the Domestic CPI at the Poverty Level of Consumption 414 14.4 Price and Income Changes (in percent) under Doha-All Deflated by the Domestic CPI at the Poverty Level of Consumption 415 14.5 Price and Income Changes (in percent) under Full Liberalization Deflated by the Domestic CPI at the Poverty Level of Consumption 416 14.6 Percent Change in Poverty under Three Scenarios 417 14.7 Change in Poverty Headcount (thousands) under Three Scenarios 419 14.8 Differences in the Probabilities of Poverty Growth and Poverty Reduction in a Sequential Application of the Reforms between Doha-SDT and Full Trade Liberalization 422 15.1 Base-Run Statistics 437 15.2 Household Income Composition 438 15.3 Macro Results 441 15.4 Sectoral Trade and Price Effects and Export Demand Shocks 443 15.5 Sectoral Volume Effects 446 15.6 Income and Welfare Effects 450 15.7 BaU Poverty Levels and Poverty Effects 453 16.1 Structure of Value Added, Factor Shares, Imports, and Exports in Russia 472 16.2 Tariff Rates, Export Tax Rates, Estimated Ad Valorem Equivalence of Barriers to FDI in Services Sectors and Estimated Improved Market Access from WTO Accession 474 16.3 Factor Income Shares and Their Standard Deviations by Consumption Decile 478 16.4 Changes in Export and Import Prices Facing Russia on World Markets as a Result of the Doha Round or Rest of the World Free Trade 482 Contents xiii 16.5 Impact of Russian WTO Accession and the DDA on Economywide Variables in Russia 483 16.6 The Mean and Standard Deviation of the Welfare Impacts of WTO Accession on Russian Households, from Poorest to Richest 487 17.1 Gains from Global Merchandise Trade Reform 504 17.2 Regional and Sectoral Source of Gains from Global Trade Reform 507 17.3 Impacts of Global Merchandise Trade Reform with and without Productivity Changes 509 17.4 Impact of Labor Productivity from Full Merchandise Trade Reform 510 17.5 Real Income Gains from Doha Scenarios, 2015 514 17.6 Real Income Gains from Doha Scenarios as Percent Change from Baseline, 2015 516 17.7a Macro Poverty Impacts from Global Reform and Alternative Doha Scenarios, US$1 per Day 518 17.7b Macro Poverty Impacts from Global Reform and Alternative Doha Scenarios, US$2 per Day 519 17A.1 Sectoral Concordance for the LINKAGE Model 523 17A.2 Sectoral Concordance for the LINKAGE Model 525 ACKNOWLEdGMENTS The editors are grateful to all the authors in this volume for their valuable contri- butions and for the high priority that they gave this project over the past year. We also thank the discussants who provided valuable input at the conference held in The Hague, December 2­4, 2004, where these papers were first presented. In addi- tion to authors listed elsewhere in this volume, these discussants included: Rashad Cassim, Alexander Keck, Hernan Lacunza, Sam Laird, Patrick Osakwe, and George Rapsomanikis. This research also benefited greatly from the input of Car- los Primo Braga, who participated in the conference and helped to guide subse- quent dissemination activities. Finally, William Cline provided a comprehensive and incisive set of referee comments on the entire manuscript. This entire project would not have been possible without the generous support of the Bank-Netherlands Partnership Program, which had the foresight both to commission this research work, as well as to ensure that sufficient funds were available for its widespread dissemination. We also express our thanks for the excellent logistical support provided by Rebecca Martin of the World Bank, and by the Dutch Agricultural Economics Research Institute (LEI), which hosted the December 2004 conference. Maria Lourdes Kasilag of the World Bank did an out- standing job formatting the manuscripts. These acknowledgments would not be complete without thanking our spouses, Adriela Fernandez and Zhen Kun Wang. Their understanding and encouragement is greatly appreciated. In closing, it should be noted that the material in this volume represents the authors' own views and not necessarily those of their employers or of the World Bank Group, its Board of Executive Directors, or the governments the Directors represent. xv Contributors Kym Anderson, World Bank Nabil Annabi, Université Laval, Québec City, Québec, Canada Channing Arndt, Ministry of Planning and Development, Maputo, Mozambique, and Purdue University, West Lafayette, Indiana Jorge Balat, World Bank Maurizio Bussolo, World Bank John Cockburn, Université Laval, Québec City, Québec, Canada Erwin Corong, De La Salle University, Manila, Philippines Caesar B. Cororaton, International Food Policy Research Institute, Washington, DC Bernard Decaluwé, Université Laval, Québec City, Québec, Canada Christian Arnault Emini, Centre d'Etudes et Recherches en Economie et Gestion (CEREG), University of Yaoundé II, Yaoundé, Cameroon, and Université Laval, Québec City, Québec, Canada Joaquim Bento de Souza Ferreira Filho, Universidade de São Paulo, Brazil Thomas W. Hertel, Purdue University, West Lafayette, Indiana, and World Bank J. Mark Horridge, Monash University, Centre of Policy Studies, Melbourne, Australia Maros Ivanic, World Bank Bazlul Khondker, University of Dhaka, Dhaka, Bangladesh xvii xviii Contributors Marijke Kuiper, Agricultural Economics Research Institute, Wageningen University and Research Centre, Wageningen, the Netherlands Jann Lay, Kiel Institute for World Economics, Kiel, Germany Will Martin, World Bank Dominique van der Mensbrugghe, World Bank Alessandro Nicita, World Bank Guido Porto, World Bank Selim Raihan, University of Dhaka, Dhaka, Bangladesh, and University of Manchester, Manchester, United Kingdom Anne-Sophie Robilliard, Institut de Recherche pour le Développement (IRD), Paris Sherman Robinson, University of Sussex, Brighton, United Kingdom Thomas Rutherford, University of Colorado, Boulder, Colorado Oleksandr Shepotylo, University of Maryland, College Park, Maryland David Tarr, World Bank Frank van Tongeren, Agricultural Economics Research Institute, Wageningen University and Research Centre, Wageningen, the Netherlands L. Alan Winters, World Bank Fan Zhai, Asian Development Bank, Manila, Philippines Abbreviations and Acronyms ACP African, Caribbean and Pacific Group of States AGE applied general equilibrium AMS aggregate measure of support ASEAN Association of Southeast Asian Nations ATC Agreement on Textiles and Clothing BaU business as usual BOP balance of payments CEPAL Comisión Económica par América Latina y el Caribe CEPII Centre d'Etudes et Prospectives d'Informations Internationales CES constant elasticity of substitution CET constant elasticity of transformation CGE computable general equilibrium CHNS China Health and Nutrition Survey CIF cost, insurance, and freight CNIS Cameroon National Institute of Statistics CPI consumer price index CRS constant returns to scale CV coefficient of variation DDA Doha Development Agenda DWP Doha Work Program EFTA European Free Trade Association ENIGH Encuesta de Ingresos y Gastos de los Hogares (Survey of Household Income and Expenditures) xix xx Abbreviations and Acronyms EPA Economic Partnership Agreement EPR effective protection rate EU European Union EV equivalent variation FAO UN Food and Agriculture Organization FDI foreign direct investment FGT Foster-Greer-Thorbecke FIES Family Income and Expenditure Survey FOB free on board FTAA Free Trade Area of the Americas GATT General Agreement on Tariffs and Trade GDP gross domestic product GEP Global Economic Prospects GTAP Global Trade Analysis Project haz height-for-age HBS household budget survey HES household expenditure survey HIPC heavily indebted poor country IBGE Instituto Brasileiro de Geografia e Estatística (Brazilian Institute of Geography and Statistics) INE Instituto Nacional de Estatística (National Institute of Statistics) ITC International Trade Centre LCMS Living Conditions Monitoring Surveys LDC least developed country LES linear expenditure system Mercosur Mercado Común del Sur MFN most favored nation MH multiple households MPS market price support MS microsimulation NAFTA North American Free Trade Agreement OECD Organisation for Economic Co-operation and Development PNAD Pesquisa Nacional por Amostragem de Domicílios (National Household Survey) POF Pesquisa de Orçamentos Familiares (Household Expenditure Survey) QR quantitative restrictions RH representative household ROW rest of the world Abbreviations and Acronyms xxi SAM social accounting matrix SDT special and differential treatment SITC Standard International Trade Classification SKTIR Special Survey on Saving and Household Investment TOT terms of trade TRP trade reform program TRQ tariff rate quota UN United Nations VAT value added tax waz weight for age whz weight for height WTO World Trade Organization Part I evaluation of the doha development agenda 1 Poverty Impacts of a WTO Agreement: Synthesis and Overview Thomas W. Hertel and L. Alan Winters Summary This chapter reports on the findings from a major international research project investigating the poverty impacts of a potential Doha Development Agenda (DDA). It combines in a novel way the results from several strands of research. First, it draws on an intensive analysis of the DDA Framework Agreement, with particularly close attention paid to potential reforms in agriculture. The scenarios are built up using newly available tariff line data, and their implications for world markets are established using a global modeling framework. These world trade impacts form the basis for 12 country case studies of the national poverty impacts of these DDA scenarios. The focus countries are Bangladesh, Brazil (2 studies), Cameroon, China (2 studies), Indonesia, Mexico, Mozambique, the Philippines, the Russian Federation, and Zambia. Although the diversity of approaches taken in these studies limits the ability to draw broader conclusions, an additional study that provides a 15-country cross-section analysis is aimed at this objective. Finally, a global analysis provides estimates for the world as a whole. Some of the main findings are: · The liberalization targets under the DDA have to be quite ambitious if the round is to have a measurable impact on world markets and hence poverty. · Assuming an ambitious DDA, the near-term poverty impacts are found to be mixed; some countries experience small poverty rises and others more 3 4 Poverty and the WTO: Impacts of the Doha Development Agenda substantial poverty declines. On balance, poverty is reduced under this DDA, and this reduction is more pronounced in the longer run. · Allowing minimal tariff cuts for just a small percentage of special and sensitive products virtually eliminates the global poverty reduction due to the DDA. · Deeper cuts in developing country tariffs would make the DDA more poverty- friendly. · Key determinants of the national poverty impacts include the incomplete transmission of world prices to rural households, barriers to the mobility of workers between sectors of the economy, and the incidence of national tax instruments used to replace lost tariff revenue. · To generate significant poverty reductions in the near term, complementary domestic reforms are required to enable households to take advantage of new market opportunities made available through the DDA. · Sustained long-term poverty reductions depend on stimulating economic growth. Here, the impact of the DDA (and trade policy more generally) on productivity is critical. To fully realize their growth potential, trade reforms need to be far reaching, addressing barriers to services trade and investment in addition to merchandise tariffs. Introduction and Motivation International trade is arguably the most direct economic means by which rich countries influence poor countries. Exports of manufactures by developing coun- tries have increased rapidly over the last 30 years, due in part to falling tariffs in the Organisation for Economic Co-operation and Development (OECD) countries as well as in developing countries, declining transport costs, increased specialization, and sustained economic growth. Manufactures accounted for just 25 percent of developing country exports in 1965, and this share tripled to nearly 75 percent over the next three decades, while agriculture's share of developing country exports has fallen from 50 percent to under 10 percent (Hertel and Martin 2000). Increased manufactures trade has benefited many developing countries, helping them make the transition out of agriculture and lifting many out of poverty. Some of the poorest developing countries, however, have gained relatively little from increased manufactures trade. Market access for their most competitive manufactured export (apparel) remains highly restricted, as it does for their key source of employment and exports, farming, and the problem with agricultural exports is exacerbated by the massive government subsidies provided to farmers in OECD countries. When poverty within the poorest countries is considered, developed countries' agricultural policies become even more central. A majority Poverty Impacts of a WTO Agreement 5 of the poor are concentrated in rural areas, where agriculture is usually the main source of economic activity (World Bank 2004), and in the poorest developing countries, large shares of households (including most of the very poorest) depend on self-employment in agriculture for virtually all of their income (Hertel and others 2004). Together, these facts highlight the potential influence that multilat- eral trade policies can have on poverty in developing countries. The DDA negotiations, sponsored by the World Trade Organization (WTO), experienced a blow in Cancún, Mexico, because of the question of rich countries' agricultural support and its potential impacts on poverty in developing countries. The Doha negotiations are now emphasizing the need to better understand the linkages between trade policies--particularly in rich countries--and poverty in the developing world. Moreover, poverty reduction is now widely accepted as a central focus for development efforts and has become the main mission of the World Bank and other development institutions. For example, the Millennium Development Goals commit the international community to halve poverty in developing countries by 2015 and identify several key means to achieving this goal with international trade. With this high level of policy interest, it is hardly surprising that the issue of trade and developing-country poverty has become a focus of much research activ- ity over the last several years. This book contributes to this literature by offering the first comprehensive analysis of the national poverty impacts of specific policy reforms proposed under the auspices of the WTO. To do so, it combines the results from several strands of research in a novel way. First, it draws on an inten- sive analysis of the July 2004 DDA Framework Agreement, particularly of poten- tial reforms in agriculture, which, as will be shown, have special significance to the poor. The scenarios analyzed below are built up from newly available tariff line data on bound and applied tariff rates. Similarly detailed analysis is undertaken in the case of domestic support for agriculture and export subsidies, as well as for nonagricultural market access. Second, the research assesses the implications of these alternative Doha scenarios for world markets. These are established using a state-of-the-art, global modeling framework that incorporates the most recent econometric evi- dence on supply and demand elasticities, with particularly close attention paid to food and agriculture markets that prove crucial in assessing the poverty impacts of the DDA. The outputs of this part of the project include export and import price changes for each region of the world, along with changes in export volumes. Third, these world trade impacts form the basis for analyzing the poverty impacts of the DDA on 10 individual countries by way of a dozen case studies. 6 Poverty and the WTO: Impacts of the Doha Development Agenda These case studies use a variety of innovative techniques to establish the potential impacts of the DDA on different household groups and, in some cases, different regions within the country. The focus countries are Bangladesh, Brazil (two stud- ies), Cameroon, China (two studies), Indonesia, Mexico, Mozambique, the Philip- pines, Russia, and Zambia. Some case studies also examine other poverty policies in addition to trade reforms--for example, education reform or agricultural extension services. Sometimes these are complementary to the Doha Round in the sense of enhanc- ing its effect, but more often they are independent. They are explored here as yard- sticks against which trade reform can be measured and as suggestions concerning how governments can seek to overcome any adverse poverty effects from the Doha Round. However, the authors do not subscribe to the view that such "com- plementary policies" are necessary for the Doha Round to be beneficial. Choice of Methodologies Organization of the research underpinning this volume had two contrasting objec- tives. On the one hand, the studies had to be consistent with one another to ensure an accurate global assessment of the DDA, as well as comparability across studies. On the other hand, research into the poverty impacts of trade reform is new, and almost the only consensus it has reached is that countries differ. From this perspec- tive, it was important to both encourage a variety of approaches at the country level and exploit the specific skills and knowledge of the case studies' authors to gear their country models most closely to local characteristics and issues.1 The project, therefore, is a composite in which the global analysis--the methodology for deriving the global findings and passing them over to the national case studies--is unique and consistent with current standards in the field of quantitative trade policy analysis, the country case studies display a wide range of methodological innovations and topical design features. This variety has been fruitful, with different country studies emphasizing alternative links between trade and poverty and providing a diversity of insights. Nevertheless, as a check and to draw some broader conclusions, two more uniform exercises are included: a 15-country cross-section analysis, in which a common, fully integrated trade- poverty analysis is provided for a range of developing countries, and, a global analysis of aggregate poverty impacts derived by applying simple poverty elastici- ties to the predicted outcomes for developing countries in a global simulation of a prospective Doha agreement. In most of this book, the methodology known as computable general equilib- rium (CGE) analysis is used. This is the dominant methodology for the ex ante Poverty Impacts of a WTO Agreement 7 analysis of the economic consequences of comprehensive trade agreements whether multilateral or bilateral in nature (Francois and Shiells, 1994). This is the dominant methodology because no other approach offers the same flexibility for looking at prospective changes in trade policy while respecting the fundamental economy- wide consistency requirements such as balance of payments equilibrium and labor and capital market constraints that are so important in determining the conse- quences of comprehensive trade reforms. The CGE approach has come under sub- stantial criticism (for example, from Jorgenson [1984], McKitrick [1998], and Kehoe [2005]) for having insufficient econometric underpinnings and for not being ade- quately validated. Accordingly, this volume offers a number of econometric-based analyses that focus on key dimensions of the trade and poverty question, including price transmission from the border to households, cropping choices made by farm households, labor market participation decisions, and the intersectoral movement of labor. In addition, when the global market impacts are assessed, a CGE model is used that is based on the most recent econometric evidence on supply and demand elasticities and for which some (modest) validation has been undertaken. The majority of the studies reported in this volume are based on comparative static analysis. The authors abstract from the impact of trade reform on invest- ment and productivity and therefore economic growth. There are two reasons for this emphasis. First, most of the issues that arise in the popular debate over the poverty impacts of trade policy are fundamentally comparative static in nature. Concerns about the urban poor being adversely affected by higher food prices, the potential loss of jobs by women in the apparel sector, or the poverty impacts on low-income farmers in developing countries are all questions about the redistrib- utive impact of trade policy reform. Answering them requires a disaggregated, comparative static framework. Of course, there is also a keen interest in the poten- tial for economic growth to alleviate poverty, and five of the studies use a dynamic framework that accounts for the growth effects of changes in investment deriving from trade policy reform. However, quantifying the impact of trade reform on growth and poverty through channels such as the effect on productivity or the benefits of increasing the range of available goods remains a lively topic for cur- rent research on which consensus has yet to emerge. Hence, the second reason for using the comparative static approach is to avoid any appearance of overstating the poverty-alleviating benefits of liberalization. In the end, it must be said that this project has proven to be a very ambitious undertaking--attempting to bridge micro-based research focusing on the choices and opportunities facing individual households in developing countries with macro-based research on the global impacts of multilateral trade policy reform. The payoff to this exercise must be judged by the insights offered. 8 Poverty and the WTO: Impacts of the Doha Development Agenda The Global Impact of the Doha Development Agenda Chapter 2 of this book, by Kym Anderson and Will Martin, takes as its starting point the July, 2004, WTO Framework Agreement for the Doha Agenda. It explores the issues emerging from of this document--in particular, the annexes dealing with export subsidies, domestic support and market access in agriculture, and market access for nonagricultural goods. Chapter 2 examines seven different Doha scenarios, one of which is adopted as the core scenario for this book. In con- structing this scenario, the authors have taken considerable care to distinguish those trade reforms that are actually being negotiated under the DDA from those that have been agreed to previously. This distinction is complicated because nearly all of the policy databases used predate completion of the Uruguay Round Agree- ment. In fact, the starting point for all of the analysis in this book is 2001, the most recent year for which comprehensive data are available for tariffs, domestic sup- port, and export interventions. Therefore, before constructing the Doha scenario, a "pre-experiment" is undertaken to account for the major developments in trade policy since 2001. These include tariff reforms undertaken by newly acceding WTO members (most notably China), the phase-in of remaining Uruguay Round commitments by developing countries, European Union (EU) enlargement to 25 countries, and the abolition of export quotas on textiles and apparel under the Agreement on Textiles and Clothing. Thus, even though the full impact of some of these reforms is yet to be felt, the analysis in this book looks beyond these reforms, envisioning a global economy in which they have been fully implemented and focusing on the further impacts of trade liberalization undertaken in the context of the Doha negotiations. The most important finding from chapter 2 is that, unless the DDA is consid- erably more ambitious than the Uruguay Round in terms of depth of cuts in bound tariffs and domestic support, it will achieve little development stimulus. The main problem on the market access side is binding overhang. For example, in agriculture--one of the key areas of the DDA with respect to trade and poverty-- bound tariffs in developing countries average 48 percent, but applied tariffs aver- age only 21 percent. In the case of the least developed countries (LDCs), the respective figures are 78 percent and 13 percent. Even in the EU (21 percent bind- ing versus 12 percent applied) and the United States (6 percent binding versus 3 percent applied), there is substantial binding overhang in agriculture. So, for many countries and products, bound tariffs can be cut deeply with no impact on applied protection and hence international trade. In the central Doha scenario featured in this book, agricultural tariffs are cut using a tiered formula, with marginal cuts changing at 15 and 90 percent bound Poverty Impacts of a WTO Agreement 9 tariff rates. The marginal cuts are 45 percent for the lowest agricultural tariffs, 70 percent for tariffs in the middle range, and 75 percent marginal cuts for the high- est tariffs.2 For developing countries, the inflection points are placed at 20, 60, and 120 percent bound tariff levels in agriculture, with marginal cuts of 35, 40, 50, and 60 percent, respectively. The LDCs are not required to cut tariffs under this central scenario. Because of a lack of specificity in the July Framework Agreement, nona- gricultural tariffs are simply cut by 50 percent across the board (33 percent in developing countries and 0 percent in LDCs). Box 1.1 summarizes the central Doha scenario. There is much more to the DDA than just agriculture and nonagricultural market access--for example, trade facilitation, services liberalization, and rules on antidumping and regionalism. This book focuses on the former issues partly because they are quantifiable and provide a large agenda in themselves. Mainly, however, they are likely to be the major issues in terms of both effects and nego- tiators' need for detailed quantitative advice. Moreover, the other issues are basi- cally additive to the analysis of market access for goods, so that as their outcomes and consequences become clear, they may be added to these results to get an over- all picture. As a consequence of the relatively ambitious tariff cuts analyzed here, average worldwide tariffs for all merchandise trade drop from 4.7 percent in the baseline to 3.2 percent. This masks rather different cuts for countries at different income levels. High-income countries' tariffs fall from 2.9 percent to 1.6 percent, middle- income countries' tariffs from 7.2 percent to 6.3 percent, and low-income coun- tries' tariffs (including LDCs, which do not cut tariffs at all) fall from 15.6 percent to 14.6 percent. (Anderson and Martin report these cuts in detail in chapter 2.) In the case of domestic support, there is also a problem of bound versus applied protection, with bindings generally much higher than applied aggregate measure of support (AMS). But even more severe is the definition of the AMS itself--particu- larly its reliance on administered prices as a benchmark. This feature makes it pos- sible for administrators in some countries to bring programs into WTO compli- ance with the stroke of a pen, simply by abolishing the administered price. The core Doha scenario assumes that industrial countries with domestic support in excess of 20 percent of production cut their bound AMS commitments by 75 percent, and others cut by 60 percent. Developing countries are assumed to cut their AMS by 40 percent. Even with these ambitious reductions, only six WTO members would be required to reduce actual support, based on 2001 notifications: Australia, the EU, Iceland, Norway, Thailand, and the United States. Export subsidies are the one area where bold cuts (full elimination) are on the table, but these have diminished in importance over time. At present, they remain a significant factor only in the case of the EU (and in the United States for 10 Poverty and the WTO: Impacts of the Doha Development Agenda Box 1.1. Elements of the DDA Scenario Based on the July Framework Agreement · Agriculture: ­ Market access--use nonlinear (tiered) formula (as with progressive income tax): · For developed countries, marginal rates (45, 70, and 75 percent) change at 10 and 90 percent tariffs · For developing countries, marginal rates (35, 40, 50, and 60 percent) change at 20, 60, and 120 percent tariffs · For LDCs, no cuts to tariffs ­ AMS: apply tiered formula: · For developed countries, marginal rates of 60 percent (AMS less than 20 percent) and 75 percent · For developing countries, marginal rate of 40 percent · For LDCs, no cuts to domestic subsidies ­ Export subsidies abolished · Nonagriculture market access: 50 percent cuts in tariffs (33 percent devel- oping countries, 0 percent LDCs). dairy products), and the abolition of export subsidies has been made conditional on equivalent treatment of food aid and state trading. Preliminary estimates suggest that reform of the latter two items will have little impact, but the linking of these fea- tures to the WTO negotiations makes the whole process much more complex. The central Doha scenario in this book assumes that export subsidies are abolished. In addition to this central Doha scenario, this book also considers an impor- tant variant in which developing countries fully reciprocate the tariff cuts made by developed countries, thereby eliminating one of the historical pillars of special and differential treatment. The rationale for considering this alternative, labeled Doha-all, becomes clear with the discussion of the results of the global poverty analyses later in this chapter. Under Doha-all, average merchandise tariffs in the middle- and low-income countries drop further, to 5.6 and 13.4 percent, respec- tively. In the case of the low-income countries, this represents a larger incremental cut in average tariffs than was achieved in the central Doha scenario itself. Assuming that negotiators honor their initial vision as set forth in the DDA and make significant cuts in agricultural and nonagricultural protection, what impact might this have on poverty? Will they really put development squarely into the DDA? Answering these questions is the primary goal of this book. Poverty Impacts of a WTO Agreement 11 The impact of the Doha reforms on world market prices is the subject of chapter 3. Here, Thomas Hertel and Maros Ivanic use a global CGE model to assess the potential impact on world market prices and trade volumes. As estab- lished in chapter 2, agricultural protection is central to any assessment of global trade reform, and the analysis in chapter 3 bears this out. The trade reform sce- narios invariably have the biggest impact on prices and trade volumes for farm and food products, followed by textiles and apparel. Given the predominance of the poor in rural areas and their heavy reliance on unskilled wages elsewhere, these are the key industries in any poverty assessment. The strongest world price increases are for the heavily subsidized farm products: rice and other grains, cot- ton, dairy products, and beef. The ranking of the price increases arises from the composition of cuts, both across the three sets of agricultural distortions and across countries. The other important point made in chapter 3 is that, given the increasingly differentiated nature of traded products, there is no one "world price," and careful attention must be paid to bilateral patterns of trade and coun- try-specific price changes. Finally, chapter 3 outlines the methodology for transmitting the price and vol- ume changes to the national case studies. This represents an important innovation in the linking of global economic outcomes with national impacts. Price Transmission The analysis of the country case studies is structured around the conceptual framework laid out by Winters (2000) and Winters, McCulloch, and McKay (2004). This begins with the question of price transmission: How much of the world price shock is transmitted to producers and consumers? With a majority of the poor in most countries located in rural areas--often poorly served by transportation and communication infrastructure--it is impor- tant to ask whether developments in global markets will really have an impact on these households. Of course, this is an empirical question, subject to econometric investigation, and this is precisely what Alessandro Nicita does in chapter 4 for the case of Mexico. He shows that, indeed, world prices are differentially transmitted to the regions of the country, depending on their distance from the border and the nature of the commodity in question. He begins his analysis by examining the extent of "pass-through" from international prices to domestic prices at the bor- der. Here, he finds that for manufactured goods, about two-thirds of the interna- tional price change passes through to the domestic market, whereas the compara- ble figure for agriculture is just one-quarter. Nicita's econometric estimates also show that the transmission of world mar- ket price changes diminishes with distance from the border. In addition, urban 12 Poverty and the WTO: Impacts of the Doha Development Agenda areas are more sensitive to border prices changes, when compared to rural areas. Therefore, he concludes that in the more remote, rural regions of Mexico, very lit- tle of the international price changes will be felt, particularly in the case of agri- cultural products. As a consequence, the impact of the Doha scenarios--which have only modest impacts on world prices, anyway--are negligible in rural Mex- ico, except in the north, near the U.S. border, where rural households see some small gains. Urban consumers face higher food prices and a small decline in unskilled wages as the privileged Mexican position in the U.S. market is eroded by most-favored-nation tariff cuts. Thus, the urban poor experience small losses. Nicita also explores the impact of complementary domestic reforms that might permit rural producers to respond to improved world market conditions without incurring additional costs (for example, a productivity gain or the employment of surplus labor). This enhances the welfare outcome for rural households in all regions except the south. Rural households in the south benefit from Doha only when the reforms are accompanied by enhanced price transmission--for exam- ple, through improved transport and market infrastructure. Thus, there is an important interaction between price transmission and the distribution of gains from global trade reforms. One of the poorest countries in the world, which also has very poor infrastruc- ture and is plagued by high domestic marketing costs, is Mozambique. In fact, work by Arndt and others (2000) estimates producer-consumer margins as high as 300 percent (for cassava). The biggest margins reported in their study are for food products, which tend to dominate both the consumption and production bundles of the poor. So the existence and behavior of these margins is critically important for any poverty study. Chapter 5, by Channing Arndt, explores this issue in the context of the Doha Round scenarios for Mozambique. As with the Mexico study, the combination of these marketing margins with modest world price changes means that the impact on household welfare in Mozambique is quite small. Indeed, about one-third of rural households are unaffected by the Doha scenario. The largest rural losses are about 1 percent of income, with some households experiencing modest gains. The dispersion among urban households is larger because of the presence of smaller marketing margins. Overall, the impact of multilateral trade reform on Mozambique is adverse as preferences are eroded and prices of imports rise. The Disaggregated Impact on Households Moving beyond the question of price transmission, the studies in this book move on to the issue of household-level impacts of--and household responses to--the price changes ensuing from trade reforms. The simplest way of exploring this link Poverty Impacts of a WTO Agreement 13 is to focus on a single commodity. This is the approach taken by Jorge Balat and Guido Porto in chapter 6 on the impact of trade reform on cotton producers in Zambia. They note that the critical factor in this case is the share of household income generated by cotton production. To a first-order approximation, the real income impact of a change in the price of cotton may be obtained by multiplying this income share by the percentage change in cotton price. This leads Balat and Porto to focus on the evolution of cotton income shares among the poor in Zam- bia. Because cotton is grown in significant quantities in only three provinces, this is where they focus attention. One of the striking things about world cotton markets in the late 1990s was the collapse in world prices. Between 1996 and 1998, cotton prices in Zambia fell by 20 percent. Therefore, it is surprising that cotton's share in income among the poor rose sharply in the eastern and southern provinces over this same period. Among the poorest households in the eastern province, the increase was nearly fivefold, even as the income share fell for wealthier households. Although there are many factors that may bear on this change, the authors argue that the most likely reason was the reform of the cotton marketing board system and the implementation of an out-grower scheme that proved effective in getting seed and fertilizer into the hands of credit-constrained, small-scale producers. This increase in the cotton share boosts the potential benefits from multilateral agri- cultural reforms, because one of the main consequences of such reform would be to raise cotton prices. Despite the increase in cotton income shares over this period, the income impact on the poor of higher cotton prices--the authors assume a 12 percent price rise, based on several independent studies of world cotton markets--is still relatively modest (on the order of 1 percent of real income, on average) because the average income share is about 8 percent. This brings the authors to a discus- sion of complementary domestic reforms. In particular, they cite evidence from other research they have conducted in Zambia, which finds that access to exten- sion services can boost productivity by more than 8 percent, resulting in an aggre- gate gain of more than 9 percent when combined with higher cotton prices. But the largest poverty reduction benefits appear to arise when subsistence households switch to cotton production in the wake of increased demand for exports. Here, a careful matching of subsistence and cotton-producing house- holds shows that, all else constant, subsistence producers could boost their incomes by nearly 20 percent if they switched to cotton production. Such a switch would be greatly facilitated by continued improvement of the out-grower schemes and strong demand for cotton exports. When combined with improved extension services and higher cotton prices, the switch from subsistence production to cot- ton could boost incomes of some of the poorest households in Zambia by nearly 14 Poverty and the WTO: Impacts of the Doha Development Agenda one-third. In sum, Balat and Porto conclude that trade reform alone is not suffi- cient to raise a large number of poor out of poverty in Zambia, but when the mar- ket opportunities presented by trade reforms are combined with complementary domestic reforms, significant headway in the fight against poverty is possible. Of course, global trade reforms do not simply alter one single commodity price: rather they potentially affect all prices in the economy, including the prices of nontradeable commodities and services as well as wages and returns to land and capital. So, the next study seeks to account for the full range of price impacts at a highly disaggregated level. The unusual thing about Joaquim Ferriera-Filho and Mark Horridge's chapter 7 in this volume is the very large number of individ- uals considered in their analysis--264,000 adults who are members of 112,000 households spread across the 27 regions of Brazil. The authors argue that the regional dimension of their study is critical, given the tremendous disparities in income and poverty incidence across regions. The proportion of poor households ranges from about 14 percent in parts of the southeast, to nearly 60 percent in the north (Amapá). When combined with large variations in industrial composition across regions, there is a recipe for great differences in poverty impacts due to trade reform. Ferriera-Filho and Horridge find that the Doha scenarios benefit agriculture at the expense of industry. This is no surprise, because virtually all previous studies of global agricultural trade reform have concluded that Brazil would be a substan- tial beneficiary from such a development. However, the real question is, which households within Brazil will benefit? Many believe that all of the benefits will go to large farmers, thereby worsening the income distribution in Brazil. The research reported in chapter 7 argues that, when one takes account of the addi- tional employment generated by the expansion of agriculture and related indus- tries in many of the poorer states of Brazil, the largest gainers are actually the households that are most heavily reliant on low-skill labor. As a consequence, the income distribution in Brazil improves under the Doha scenario. This is a very important finding. It is a point that has been previously emphasized in more highly aggregate research on trade and poverty reported by Harrison and others (2003). As a percentage of initial poverty, the estimated national decline in chapter 7 is modest (less than 1 percent), but it still amounts to a large number of persons: Under the Doha scenario, poverty falls by about 236,000, and it declines by about twice that amount in the case of the full-liberalization scenario. The declines in poverty are fueled by the growth in agricultural activity--Brazilian farm and food exports expand strongly in the wake of trade reform--and the subsequent increase in demand for the lowest skill workers, 41 percent of whom still work in the farm sector. Poverty Impacts of a WTO Agreement 15 Of course, these wage gains hinge on the existence of an operational labor mar- ket. Such a market may not exist in some cases, and the potential consequences of factor market failure are explored in considerable depth in chapter 8 by Marijke Kuiper and Frank van Tongeren. These authors approach this problem by employ- ing a village-level model of a community in Jiangxi province in China. They cap- ture the heterogeneity of household types by grouping them according to their factor endowments. In particular, they distinguish whether or not households have access to draft power and whether or not they have family members involved in temporary migration outside the province. After a detailed analysis of circum- stances in this village, they conclude that the markets for labor, land, and capital are imperfect, thereby preventing households from simply taking wages and rental payments as given when making decisions about consumption and production. This "nonseparability" complicates the household's decision-making process and can result in some striking results in the wake of trade reforms. In the case of Doha reforms, the real income gains for the village are quite modest--about 1.2 percent of income--and relatively evenly spread across the different household groups. However, in the case of full liberalization, the aggre- gate gains are four times as large and much more unevenly spread across house- holds, with the gains to households with draft power nearly twice as large as those for the other household groups. This reflects the intensification of production in agriculture engendered by higher prices for rice and other farm products. Labor Markets The main resource with which the poor are endowed is their own labor. Whether they are self-employed farmers, providers of services, or wage earners, their income is closely tied to conditions in the labor market. This point surfaces clearly in the Brazil and China studies discussed above, both of which emphasize the importance of labor markets as a mechanism for transmitting favorable develop- ments in the world marketplace, as well as elsewhere in the domestic economy, to impoverished households. The next set of studies focuses primarily on the labor markets in Brazil and China, as well as on a third country, Indonesia. The first of these is chapter 9, by Maurizio Bussolo, Jann Lay, and Dominique van der Mens- brugghe, on Brazil. Their focus is specifically on the link between the farm and nonfarm labor forces. They model the decision to move out of agriculture based on an econometric model that predicts the likelihood of a given individual chang- ing sectors, based on the historical evidence in Brazil. The other important feature of this chapter is the authors' analysis in the context of a 2001­15 baseline for the Brazilian economy. This permits viewing the impacts of trade reform in the con- text of ongoing changes in the economy, labor markets, and poverty. 16 Poverty and the WTO: Impacts of the Doha Development Agenda In their baseline projection, Bussolo, Lay, and van der Mensbrugghe find that the poverty headcount falls by almost 14 percent. The majority of this decline is due to poverty reduction in agriculture, a sector that grows considerably faster than the nonfarm economy under their business-as-usual forecast. The majority of this poverty reduction is due to factor price changes (for example, higher wages), but a significant portion is attributable to the exit of labor from the rela- tively low-wage agricultural sector to higher-wage, nonfarm jobs. This intersec- toral movement is particularly important to the poorest farm households. Having established this baseline scenario, Bussolo, Lay, and van der Mensbrug- ghe analyze the implications of alternative trade reforms for poverty--in particular for the different labor force groups: the "movers" who move from agriculture to nonagriculture over the course of the baseline, the "stayers" who remain in agricul- ture, and the "stayers" in nonagriculture. The largest percentage point reduction in poverty over the baseline is for the "movers," who experience a 22.4 percentage point reduction in their headcount (down from 53.4 percent to 31 percent). This is the poorest of the three groups, and it is also the group that experiences the great- est incremental poverty reduction, above and beyond the baseline, as a result of the Doha trade reforms. Overall, the authors find quite modest poverty gains from the Doha scenarios (just 3 percent of the baseline change over the 2001­15 period). Full liberalization generates estimates of national poverty reduction that are three times as large as the Doha reductions, but still modest in the context of projected baseline changes. This underscores the fact that trade reforms taken alone are a rel- atively small piece of the overall poverty reduction puzzle. Chapter 10 by Fan Zhai and Thomas Hertel takes a deeper look at the Doha reforms through the lens of a labor-focused CGE model of China and the scope for enhancing these outcomes through complementary education reforms. Like chap- ter 9, this chapter emphasizes the farm-nonfarm labor market linkage, which Zhai and Hertel argue is partly a function of educational attainment and therefore sus- ceptible to change through educational policy. They also emphasize the link between rural and urban labor markets in China through the temporary migration of workers. (Permanent migration is still restricted in that country.) In their analy- sis of multilateral trade reforms, the authors find that poverty falls across all of their household categories: by 1.3 percent in the case of Doha and 2.7 percent in the case of full liberalization. Inequality also declines slightly under these scenarios. Zhai and Hertel cite econometric evidence that suggests that an additional year of education boosts an individual's chances of obtaining an off-farm job in China by 14 percent. Educational attainment is also important for workers seeking to meet the needs of an increasingly integrated global marketplace, yet education expenditures per pupil in the rural areas lag significantly behind their urban coun- terparts in China. The authors explore the implications of accompanying trade Poverty Impacts of a WTO Agreement 17 reform with additional educational investments in rural areas to enhance rural labor mobility, productivity, and income. In particular, they boost expenditures per pupil enrolled in mandatory education by 16 percent to reach the comparable urban level. This increment is assumed to be financed in part by public funds, raised through additional taxation, and in part through increased private contribu- tions taken out of rural households' disposable income. This combination of edu- cational and trade reforms has a much stronger impact on poverty alleviation, with the number of poor (living on less than US$2 per day) falling by 13.4 percent. This scenario also has a favorable impact on rural-urban income inequality. The final chapter focusing on labor markets, chapter 11, is a case study of Indonesia by Anne-Sophie Robilliard and Sherman Robinson. Instead of focusing on the farm-nonfarm or rural-urban movement of labor, these authors draw a sharp distinction between the formal and informal labor markets. The formal sec- tor offers high wages, but few opportunities for employment. The informal sector, by contrast, has a flexible wage that is assumed to clear the market. Robilliard and Robinson explicitly model each individual's decision to participate in one or the other of these labor markets. In this way, they are able to predict which types of individuals will lose their job when formal sector employment contracts and which will be hired when employment expands. These changes in employment represent an important determinant of the welfare impacts on households of any change in a country's pattern of trade, production, and employment. Robilliard and Robinson explore the poverty impacts of multilateral trade reform under three alternative labor market closures: fixed aggregate employment and flexible wages; fixed, sector-specific labor (no change in employment by sec- tor); and fixed real wages and variable aggregate employment (that is, changes in unemployment are permitted). They focus on the full-liberalization scenario for this sensitivity analysis and find that the largest reduction in poverty comes from the fixed employment scenario: about 1.4 million people are lifted out of poverty. The proportional reduction is slightly higher in the rural areas and more favorable to the poorest of the poor as well, so that the national Gini index falls in this clo- sure. When labor is not permitted to move across sectors, the poverty reduction is much smaller--only 900,000--because the economy is not permitted to fully adjust to the new world prices, efficiency gains are blunted, and the national rise in per capita income is muted. The third case, in which wages are fixed and the unemployment rate is permit- ted to fall in the wake of increasing labor demand, presents a particularly interest- ing contrast in chapter 11. With increasing aggregate employment, national per capita income rises more than in the first case with fixed employment and flexible wages. The authors point out that the poverty outcome depends critically on who gets the new jobs. If the new jobs go to individuals from nonpoor households 18 Poverty and the WTO: Impacts of the Doha Development Agenda (that is, families with other wage earners or other sources of income), the unem- ployment specification could worsen income inequality because the pool of unemployed workers prevents unskilled wages from rising and, without the bene- fit of higher wages, the poverty reduction would be muted. To quantify this out- come, the authors have estimated the likelihood that each type of unemployed individual will obtain one of the newly available jobs. There is a considerable uncertainty associated with these estimates, and Robilliard and Robinson reflect this by reporting their results in terms of the mean and standard deviation of a Monte Carlo simulation for each closure or scenario. Although the mean poverty reduction under the unemployment closure is larger than that under the standard labor market specification, the standard deviations suggest that the two are not significantly different in a statistical sense. Interactions with Tax Policies An important theme in many of the chapters in this volume is the potential for interactions between the Doha scenarios and domestic policies to alter the poverty outcomes obtained from multilateral trade reform. Does multilateral trade liberalization lessen the distortions introduced by domestic commodity and factor market policies, or does it exacerbate them? To what extent can comple- mentary reforms of domestic policies enhance the degree of poverty reduction? When trade liberalization results in reduced tax revenues, how will this shortfall be made up? Two of the chapters in this volume focus squarely on the question of tax replacement.3 Chapter 12, by Christian Arnault Emini, John Cockburn, and Bernard Decaluwé, focuses on the case of Cameroon. They examine the poverty impacts of the central Doha scenario, paying particular attention to the structure of the domestic tax system and the different options available for replacement of the lost tariff revenue. They view the value added tax (VAT) as the most likely tax replace- ment tool in Cameroon. This tax has a very heterogeneous impact on sectors, with effective rates ranging from 0 percent in the case of agriculture to 13 percent in the case of petroleum refining. When the authors combine this tax replacement tool with the Doha scenario, they find that poverty falls slightly, by about 22,000 peo- ple, in Cameroon, as does inequality. Of course, with relatively small tariff cuts under the Doha scenario, tax replacement is not all that central in this scenario. In the case of full liberalization, tax replacement becomes much more impor- tant, and the authors consider three alternative tax scenarios with these tariff cuts. In every case, poverty rises, but the size of the poverty increase, as well as its causes, vary with the choice of replacement tax. When Arnault Emini, Cockburn, and Decaluwé use a nondistorting production tax, 106,000 people are estimated Poverty Impacts of a WTO Agreement 19 to be lifted out of poverty, but 193,000 formerly nonpoor fall into poverty, result- ing in a net poverty increase of 87,000 people. This occurs despite an increase in aggregate welfare in Cameroon, so it is clearly a consequence of the pattern of imports and exports in that country. When trade reform is coupled with an increase in consumption taxes, the poverty rise is much larger--nearly half a mil- lion people. This impact is lessened somewhat (a 300,000-person increase) by the use of the VAT to replace the forgone tariff revenue. Clearly, in the case of Cameroon, the choice of tax instrument used to replace the lost tariff revenue can be as important as the type of trade liberalization (full liberalization versus Doha reforms only). Chapter 13, by Caesar Cororaton, John Cockburn, and Erwin Corong, is a study focusing on the issue of tax replacement in the Philippines. This is an inter- esting case because the agriculture sector has evolved from net exporter to net importer over the past three decades. Because the country is a relatively recent net food importer, there is widespread concern in the Philippines that trade reforms will jeopardize food security. However, in their analysis of the Doha scenarios, the authors find that the national poverty headcount is barely affected. There is a small rise in poverty among the self-employed households, particularly those in rural areas, but poverty among salaried urban workers falls. In contrast to many of the focus economies studied in this volume, the Doha reforms are not favorable to Philippine agriculture, and this effect is more pronounced under full liberaliza- tion. Because Philippine agriculture currently receives relatively high protection, full liberalization results in a contraction of the agricultural sector and an increase in rural poverty. This is offset by a reduction in poverty among the urban popula- tion, where wages rise. As a consequence, there is a small decline in the national poverty headcount. However, when the authors switch from the VAT to a uniform income tax for purposes of tariff replacement, poverty rises in the full-liberaliza- tion case. Once again, the pattern of exemptions in the indirect tax system favors the poor, and its use for purposes of tax replacement is a critical piece of the poverty puzzle. Cross-Country Comparisons With their differences in factor market closures, elasticities of substitution, methodologies for grouping households and modeling labor markets, and so forth, the country case studies discussed up to this point have not been comparable. This makes it difficult to generalize on the basis of cross-country comparisons. There- fore, chapter 14, by Maros Ivanic, features a cross-country comparison for 15 coun- tries, each of which has been treated symmetrically. Although this approach is somewhat stylized, and therefore less definitive for any given country, each of the 20 Poverty and the WTO: Impacts of the Doha Development Agenda focus country databases has been built up from the same types of individual household surveys as the single-country case studies. Another virtue of this chapter is that is offers a fully integrated, global-national-micro modeling approach. In particular, Ivanic has augmented the Global Trade Analysis Project (GTAP) global CGE model with reconciled data on 140 disaggregated household groups for each of the 15 focus countries. His grouping is based on income specialization--for example, agriculture-specialized households rely almost entirely on agricultural self-employment for their income, and similarly for a wage-specialized stratum, and so forth. Because Ivanic uses a global framework, he can simulate all of the trade reform scenarios directly in his model, which also facilitates further decom- position of the elements of trade reform and their poverty impacts. Ivanic's findings with respect to the poverty impacts of the DDA are particu- larly interesting. Specifically, he finds that the Doha trade reform scenarios are not as poverty friendly as the global liberalization scenario. If Doha represented the same mix of policy reforms as full liberalization, both simulations would be expected to have the same pattern of poverty reduction, but with larger cuts under full liberalization because of its deeper cuts in protection (for example, 100 per- cent versus 33 percent). However, this is not the case, and, in a decomposition analysis, Ivanic shows why. The DDA as outlined in chapter 2 has a variety of different elements, and these have conflicting impacts on poverty. The removal of export subsidies in the EU and the United States tends to raise poverty in most of the developing countries in Ivanic's sample, even while reducing poverty among the agricultural households in these poorer countries. This is hardly surprising in light of earlier studies high- lighting the vulnerability of low-income, net-food-importing countries to higher world prices for these products (see, for example, Valdes and McCalla [2004]).4 Because these export subsidies are fully removed under the Doha scenario, this impact is fully realized under that partial reform. However, Ivanic finds that cuts in developing-country tariffs as a group have a very favorable impact on national poverty in the focus countries.5 Yet there is very little reform of developing-country tariffs under Doha--first as a result of limited reciprocity (part of special and dif- ferential treatment), and second as a result of the extensive binding overhang in developing countries. Thus, although developing-country tariff cuts are among the most poverty-friendly elements of global trade reform, very little of the beneficial impact of these reforms is felt under the Doha scenario. When combined, these facts explain why Doha is less poverty friendly than the comprehensive reform sce- nario. It accentuates those aspects of reform that adversely affect poverty (export subsidies), while largely omitting those aspects that benefit the poor. This suggests that deeper cuts in developing-country tariffs under the Doha sce- nario might have a beneficial impact on the poverty outcome. This is explored Poverty Impacts of a WTO Agreement 21 under the alternative scenario, Doha-all, in which developing countries fully recip- rocate the developed-country reductions in tariff bindings. Ivanic shows that Doha- all does have a more favorable poverty outcome than the base Doha scenario. An additional finding from Ivanic's cross-section analysis pertains to the com- mon assumption that "a rising tide lifts all boats," that is, that poverty rises and falls in concert with changes in national per capita income. Ivanic shows that this is not always the case in the near term because trade reform generates uneven gains in the economy. One sector gains and another loses, so it matters greatly where the poverty is concentrated. If most of the poor work in agriculture, and agriculture is hurt by trade reform, poverty may rise even if real national income rises. This is the case in Malawi, where 40 percent of the population is specialized in agricultural self-employment. Effects on Productivity and Economic Growth Sustained reductions in poverty require economic growth, which leads naturally to the question of how a prospective DDA might affect the growth rates of coun- tries currently experiencing the highest levels of poverty. This is a challenging area of research--worthy of an entire volume in its own right--but the final section of this book offers two country case studies and a global synthesis chapter oriented toward this theme. Chapter 15 on Bangladesh focuses on the growth question by emphasizing the impact of trade reform on capital accumulation. Nabil Annabi, Bazlul Khandker, Selim Raiham, John Cockburn, and Bernard Decaluwé begin with a short-run analysis in which they find that Bangladesh experiences an aggregate loss, as well as a small rise in poverty, under the Doha scenario. There are two reasons to expect such short run losses. First, Bangladesh is a net agricultural importer and, as such, will suffer from higher world prices of agricultural products. Second, as an LDC, Bangladesh currently enjoys tariff-free access to many of the rich coun- tries' markets. When tariffs in these markets fall, Bangladesh is expected to suffer from preference erosion--that is, the value of these tariff preferences diminishes. The analysis in this book suggests that the first explanation is the relevant one, with the main losses associated with imports of cotton, wheat, and oilseeds. There is no evidence of net preference erosion adversely affecting the terms of trade for Bangladesh. Because the apparel exports displaced by erosion from the EU are absorbed in the North American market, where, de facto, most apparel exports from Bangladesh do not enjoy preferential market access, Bangladesh benefits from the tariff cuts. The terms of trade losses facing Bangladesh under Doha are magnified under full liberalization. In addition to the above, to pay for additional imports, Bangladesh must expand the volume of its textile and apparel exports, 22 Poverty and the WTO: Impacts of the Doha Development Agenda which account for nearly 80 percent of export revenues. This tends to depress their prices. However, these short-run losses are transitory and the authors of chapter 15 estimate that after two to three years, the economy will be better off under full lib- eralization than under the business-as-usual scenario. The reason is that the cost of investment goods will fall, and increased investment will flow to the more com- petitive sectors, thereby stimulating additional growth. The authors estimate that in the long run (15 years), gross domestic product (GDP) will be 1.44 percent higher and poverty 6.1 percent lower under the full-liberalization scenario. A closer look at these results reveals that most of the stimulus for the increased investment and economic growth comes from the reduction in Bangladesh's own tariffs, which would be missing under the Doha scenario. These authors also explore an issue that has been the subject of much discus- sion recently in the context of the WTO: remittances from overseas workers. They formally explore the implications of a 50 percent increase in the flow of remit- tances to Bangladesh--and specifically to those households currently receiving these transfers. As a result, the domestic labor supply is reduced. This develop- ment has a favorable impact on poverty, reducing it by 0.8 percent in the short run and 4.0 percent in the long run. To the extent that rich countries are concerned about the impact on Bangladesh of higher food prices and preference erosion, a policy that permits increased temporary migration appears to be a good way to offset some of these negative effects, because the benefits of increased remittances dominate the short-run costs of trade liberalization. Chapter 16, by Thomas Rutherford, David Tarr, and Oleksandr Shepotylo, explores one of the key trade-growth linkages in the case of Russia. They focus particularly on the potential for international trade and foreign direct invest- ment (FDI) in the services sector to bring new varieties of goods and new tech- nologies to Russia, thereby enhancing productivity, generating economic growth, and lifting households out of poverty. The role of services sector reforms--an important aspect of future WTO agreements--is often neglected in analyses of trade and poverty. Yet, as Mattoo, Rathindran, and Subramamian (2001) demon- strate, such reforms, particularly in telecommunications and financial services, can boost long-run growth rates. The chapter on Russia begins by analyzing the Doha scenario explored by other authors. The impact of this scenario is mixed, but most of the households experience a small welfare loss. The full-liberaliza- tion scenario shifts the distribution of welfare impacts in the positive direction, so that most Russian households now gain and poverty falls, but again the changes are quite modest. The authors then turn to domestic reforms in the services sectors--a part of the economy that the DDA is not expected to affect to any great degree, but an Poverty Impacts of a WTO Agreement 23 area that is currently receiving a great deal of attention in the context of Russia's WTO accession negotiations. The authors show that the liberalization of barriers to FDI greatly enhances the potential welfare gains. The main vehicle for this enhancement is the provision of new varieties of services, which improve produc- tivity, not only in the services sector, but also in services-using sectors as well. Indeed, the added productivity boost from the elimination of services FDI barri- ers alone is sufficient to generate a per capita income increase of 5.3 percent, ensuring that virtually all Russian households benefit from the reform. There are two lessons to be drawn from this work. First, productivity growth is essential for generating widespread gains from trade reforms, and second, one way of obtain- ing such growth is through ambitious services sector reforms, such as those that have been a part of recent WTO accession negotiations, most notably in China, but also now in Russia.6 The final chapter in the book provides an integrated, global analysis of the potential for multilateral trade reforms to reduce poverty in the long run (by 2015). In this chapter, Kym Anderson, Will Martin, and Dominique van der Mensbrugghe use the latest version of the World Bank's LINKAGE model, along with the same GTAP dataset used in chapters 3 and 14, to project the growth path of the global economy from 2001 to 2015. They find that trade reforms have a modest impact on capital accumulation and thereby boost the projected global gains from multilateral trade reform by about one-quarter. However, they devote most of their attention to the potential impacts of increased trade on productivity growth. (It should be noted, however, that the authors focus entirely on produc- tivity growth associated with increased manufactures exports, not services trade or investment as in the Russia study). There is now a rapidly growing literature on the impacts of trade and trade policy reforms on productivity, and Anderson, Martin, and van der Mensbrugghe draw on this in their chapter. When they incorporate the additional impact of openness on labor productivity, they find a substantial boost to the global gains (40 percent larger gains in 2015) with a disproportionate share accruing to the South and East Asia developing economies. The poverty impacts of these alterna- tive scenarios are elicited by first estimating the income gains to the poorest households and then applying to this an estimated elasticity of poverty reduction with respect to income growth at the poverty line. Instead of using real per capita income for the region as a whole, the authors use the unskilled wage rate, deflated by an index of food and clothing prices, reflecting the dual facts that the main endowment of the poor is their own labor and they spend the bulk of their income on nondurable goods. Another critical assumption is that the poor do not pay taxes, so that any increase in tax rates required to offset forgone tariff revenues does not affect them. 24 Poverty and the WTO: Impacts of the Doha Development Agenda Applying these estimates of earnings at the poverty line to the poverty elastic- ity of income in each region, which varies depending on the regional distribution of income, the authors predict the extent of poverty reduction in developing countries. Of course, this depends on the poverty line. It also depends on the base- line poverty projections, which decline considerably between 2001 and 2015. For US$1 per day poverty, the estimated reduction in 2015, in the absence of addi- tional productivity gains, is 2.5 million people for the Doha scenario and 31.9 mil- lion people for full liberalization. When applied to current (2001) poverty levels, the authors' calculations result in poverty reductions of 9.7 million and 80.5 mil- lion people under the Doha and full-liberalization scenarios, respectively. The 2015 poverty reductions are increased to 4.3 million and 43.5 million, for Doha and full liberalization, respectively, when productivity gains are factored in. For US$2 per day poverty, the reduction in the number of the poor is larger, but the percentage reduction is smaller (see table 1.1).7 Based on the Doha­full liberalization comparison, it is clear that the (rather ambitious) Doha scenarios capture only a relatively small portion of the total poverty reduction possible under trade reforms. When the authors consider the Doha-all scenario, they find that implementing deeper cuts in the developing countries enhances the poverty outcome, nearly doubling the poverty reduction obtained under the central Doha scenario. This finding reinforces Ivanic's conclu- sions, in chapter 14, with respect to the beneficial poverty impacts of developing- country tariff cuts under the DDA. It is also hardly surprising in light of the increasing importance of south-south trade and the relatively high level of devel- oping country tariffs, as reported in chapter 3. Another important finding from the Anderson, Martin, and van der Mens- brugghe chapter relates to sensitive agricultural products, as well as special prod- ucts, in developing countries. Industrial countries have proposed that certain sen- sitive products be exempt from steep tariff reductions, instead being liberalized through a combination of quota expansion and tariff reduction. In chapter 2, Anderson and Martin suggest that a cut in bound tariffs might be most effective, and they consider the case in which these commodities, limited to 2 percent of industrial-country tariff lines in agriculture, face a modest 15 percent cut in bound tariffs. In the case of developing countries, an additional category of exemptions is provided for in the Framework Agreement. These special products, identified "based on criteria of food security, livelihood security and rural devel- opment needs," will be eligible for more flexible treatment as well (WTO 2004). Allowing for this additional category, the scenario outlined by Anderson and Mar- tin permits developing countries to exempt 4 percent of agricultural tariff lines from the tiered cuts, facing instead just a 15 percent cut in bound tariffs. Poverty Impacts of a WTO Agreement 25 Of course, both special and sensitive products invariably have the highest tar- iffs, so that exempting them can make a big difference in the results. The authors find that merely introducing these modest exemptions for a maximum of 2 per- cent of the industrial tariff lines in agriculture (4 percent for developing coun- tries) virtually eliminates the poverty impacts of a Doha agreement. Therefore, to have a significant poverty impact, the DDA must not only have ambitious numer- ical targets, it must also seek to limit--indeed, eliminate--the use of sensitive and special product exemptions.8 Summary and Conclusions Assessing the impact of multilateral trade liberalization on poverty is a challeng- ing assignment. As Winters (2000, p. 43) notes, "Tracing the links between trade and poverty is going to be a detailed and frustrating task, for much of what one wishes to know is just unknown. It will also become obvious that most of the links are very case specific." This book represents an attempt to make known a few more of these unknown linkages. As such, the approach has been heterogeneous and opportunistic, calling on experts in this field to undertake in-depth studies in countries for which appropriate data and analytical infrastructure are available. All of this research capacity has been directed toward the analysis of the trade pol- icy question that is central in many policy makers' minds today: What are the likely poverty impacts of a successful DDA? And what elements could be added to enhance this outcome? As noted previously, the approach taken in this book ensures consistency of methods in the global analysis of the multilateral trade reform scenarios, as well as in the methodology for incorporating these results into the national analyses. However, at the country level, different authors have had the liberty to take a vari- ety of approaches depending on the particular circumstances facing their coun- tries and their own analytical interests. This is why there are two studies of the Brazilian economy--one of which focuses on near-term impacts across heteroge- neous individuals, households, and regions in Brazil, and one of which focuses on longer-term impacts, particularly in light of the barriers to intersectoral labor mobility. In the case of China, one study focuses on market failure at the village level, and another focuses on labor mobility at the national level. Similarly, there are differences in methodology taken across country case studies, with a mix of partial and general equilibrium approaches, and static and dynamic frameworks. The base years differ across studies, and even the poverty lines chosen are not uni- form across all studies. Their findings, therefore, are not strictly comparable. Finally, because the choice of countries to include in this volume was made on the 26 Poverty and the WTO: Impacts of the Doha Development Agenda basis of preexisting work that laid a foundation for the current research project, this is not a random sample of developing countries. With these qualifications in mind, let us take an overview of the findings. Table 1.1 summarizes the poverty results from each of the national studies (subnational studies are not reported here) for both the Doha and full-liberaliza- tion scenarios, distinguished by length of run for the analysis. The long-term studies factor in the impact of trade policy on investment and capital accumula- tion--and in the case of the global analysis, productivity as well--whereas the short-term studies do not. The national poverty changes are reported in two dif- ferent ways: first, as the change in number of persons in poverty, and second, as the percentage change in the poverty headcount. Thus, a negative number in table 1.1 means that the number of poor has fallen as a result of multilateral trade reform, and a positive number indicates that the number of poor has risen. Table 1.1 suggests several tentative conclusions. First, the near-term analyses are mixed in terms of their outcomes, with poverty rising in some cases and falling in others. We view this diversity as correct and as a strength of the country-based approach. Even setting aside the methodological differences between studies, the case specificity alluded to above leads us to expect differences between countries' interests in the DDA, and the chapters explain exactly why this is so. The largest poverty reductions in table 1.1, in both absolute and relative terms, are in countries with agricultural export potential to the markets that liberalize most (that is, East Asia and Europe). The strong poverty reduction in Brazil is driven by increased agricultural production, which tends to be concentrated in regions with relatively higher poverty incidence. In China, the poverty reduction is fueled by increased agricultural exports to the highly protected agricultural markets of East Asia. However, the poverty increases tend to be in countries that are net importers of agricultural products (for example, Bangladesh) and that may eventually benefit from preferential market access (for example, Mozam- bique). Thus, the strongest difference between countries concerns their exposure to the shocks generated by the DDA. Even holding this constant, however, poverty impacts can vary with, for example, the degree of transmission of world prices to rural households, the barriers to the mobility of workers between sectors of the economy, and the incidence of national tax instruments used to replace lost tariff revenue. Taken as a whole, the number of countries where poverty declines under the Doha scenario is about the same as the number of countries where it falls, although looking at the absolute number of poor, it can be seen that poverty declines in several of the most populous countries (Brazil, China, and Indonesia) and therefore declines overall in this nonrandom sample of countries. As for the long-run results, all of the studies that consider the impact of trade on capital accumulation, productivity, or both predict a reduction in poverty Poverty Impacts of a WTO Agreement 27 Table 1.1. Poverty Impacts of a Prospective DDA Change in poverty headcount Change in poverty headcount Country (chapter number) Near term: fixed capitalLong term: investment Long term: Near term: fixed capital investment impacts Full Full Doha liberalization Doha liberalization Country (chapter 1,000 1,000 1,000 1,000 number) people % people % people % people % Bangladesh (15) 38 0.3 1,354 1.1 0 0 -5,758 -4.6 Brazil (7) -236 -0.4 -482 -0.8 Brazil (9) -380 -1.1 -1,030 -2.9 Cameroon (12) -22 -0.4 303 4.8 China (10) -4,590 -1.1 -8,271 -2.0 -5,378 -1.3 -11,170 -2.7 Indonesia (11) -48 -0.1 -1,384 -3.5 Mexico (4) 4 0.0 127 1.0 Mozambique (5) 27 0.3 60 0.6 Philippines (13) 12 0.0 -7 0.0 Russia (16) 209 0.9 -122 -0.5 All developing countries (17) US$1 per day: 2001a -7,000 -66,300 -9,700 -80,500 2015b -1,700 -0.3 -23,800 -3.8 -2,500 -0.4 -31,900 -5.1 US$2 per day: 2001 -8,700 -103,900 -12,600 -123,200 2015 -4,100 -0.2 -52,300 -2.7 -6,200 -0.3 -65,600 -3.3 Productivity effects addedc US$1 per day: 2001 -20,400 -126,500 2015 -4,300 -0.6 -43,500 -6.5 US$2 per day: 2001 -29,600 -193,200 2015 -12,100 -0.6 -94,700 -4.9 Source: Studies reported in this book. a. Based on percentage changes in 2015, but applied to 2001 poverty headcount. b. Computed for the year 2015 when the total number of poor is projected to be significantly lower. c. Productivity gains from increased openness to trade apply to both manufactures and agriculture. (Earlier versions assumed only productivity gains in manufactures. See chapter 17 for details.) 28 Poverty and the WTO: Impacts of the Doha Development Agenda (with the exception of Doha-Bangladesh, where there is no long-run measurable impact). Trade stimulates investment, investment stimulates growth, and growth reduces poverty. When productivity impacts are also considered (bottom group of rows in table 1.1), this effect is even stronger. This distinction between the short run and the long run is particularly striking in the case of the full-liberalization scenarios for Bangladesh, where the short-run impacts of trade reform translate into a rise in headcount poverty, and the long-run impacts of trade reform suggest a substantial decline. In addition to the quantitative summary reported in table 1.1, the research documented in this book has generated some additional insights. First, the liber- alization targets under the DDA have to be ambitious if the round is to have a measurable impact on world markets and hence poverty. Second, assuming an ambitious DDA, the near-term poverty impacts are likely to be mixed. The analysis suggests, however, that countries can enhance the impact on poverty by pursuing complementary domestic reforms to enable households to take advantage of market opportunities created by the DDA.9 These include improved infrastructure and the reform of domestic marketing institutions to improve price transmission to rural areas, rural education reform to enhance labor mobility between the farm and nonfarm sectors, and extension outreach to permit farmers to take advantage of new export opportunities opened up by the DDA. Of course, sustained poverty reduction depends on stimulating economic growth. Here, the impact of the DDA on productivity is critical. Empirical evi- dence suggests that increased merchandise trade will likely bring with it produc- tivity gains through disciplinary effects of import competition on domestic firms as well as, possibly, learning by doing on the export side. To fully realize potential productivity gains, however, trade reforms need to be far reaching and should include reducing barriers to services trade and investment in addition to mer- chandise tariffs, which lie mainly or wholly outside the DDA. Thus, even if the DDA is very successful, a major agenda of unilateral reform and further rounds of multilateral talks remains. Only through such comprehensive reforms can long- term growth and poverty reduction be ensured. Notes 1. The forthcoming book, Globalization and Poverty, edited by Ann Harrison, adopts the same strategy, combining a set of cross-country econometric studies with several individual country case studies. 2. For example, a tariff of, say, 100 percent is cut by 66­95 percent: [15%*0.45 + (90-15)%*0.70 + (100-90)%*0.75]. Applying the cuts at the margin avoids the discontinuities implied by the July Framework. Poverty Impacts of a WTO Agreement 29 3. Of course, some assumption about tax replacement is required in each of the studies in this volume. The standard assumption used is one of replacement of lost tariff revenue with an equiproportional (distribution-neutral) income tax. Although not a realistic assumption in most cases, it facilitates the comparability of results across regions. In those cases where country case study authors have emphasized the treatment of the domestic tax system, they have been encouraged to explore the impacts of replacing the lost tariff revenue with the most likely instrument (usually the value added tax). 4. Dimaranan, Hertel, and Keeney (2004) demonstrate that many developing countries have become much more dependent on imports of subsidized crops from OECD countries over the past 40 years. Removing these subsidies will obviously have an adverse effect in the near term. 5. In Ivanic's analysis, most of these gains come from improved market access to other developing countries. This is due to the relatively high optimal tariff in the underlying GTAP model, which makes unilateral reform relatively unattractive (see also chapter 3). This implies that to reap the benefits, developing countries must liberalize together--the multilateral aspect of reform is important. 6. Similarly dominant welfare effects from services reforms have been found in the case of China's WTO accession agreement (Walmsley, Hertel, and Ianchovichina Forthcoming). 7. These estimates of poverty reduction are considerably smaller than earlier predictions using the World Bank's LINKAGE model. The difference is due to the fact that these estimates are based on the most recent (Version 6) GTAP database, which is further updated to account for EU enlargement as well as the WTO accession of China and others. These recent trade reforms have reduced the overall level of protection worldwide, thereby lessening the gains from reform. In addition, the Version 6 data- base has a complete treatment of preferential tariffs, including the EU's 2001 Everything but Arms ini- tiative, which means that gains to the LDCs from trade reform are considerably reduced. 8. It is not argued that individual developing countries couldn't improve their poverty outcomes by exempting a few special products from liberalization. But given the multitude of products and coun- tries, such cases cannot be identified here. Besides, it is implausible that developing countries could leave the exemptions door open in any significant fashion without industrialized countries also squeezing their sensitive products through the same opening. 9. The volume edited by Harrison (Forthcoming) reaches a similar conclusion, based on a set of ex post analyses of trade reform and poverty. References Arndt, Channing, Henning Tarp Jensen, Sherman Robinson, and Finn Tarp. 2000."Marketing Margins and Agricultural Technology in Mozambique." Journal of Development Studies 37 (1): 121­37. Francois, J. F., and C. R. Shiells. 1994. "Applied General Equilibrium Models of North American Free Trade: An Introduction." In Modelling Trade Policy: Applied General Equilibrium Assessments of North American Free Trade, 3­46. Cambridge: Cambridge University Press. Harrison, G. W., T. F. Rutherford, D. G. Tarr, and A. Gurgel. 2003. "Regional, Multilateral, and Unilat- eral Trade Policies of Mercosur for Growth and Poverty Reduction in Brazil." Policy Research Working Paper 3051, World Bank, Washington, DC. Hertel, Thomas W., Maros Ivanic, Paul V. Preckel, and John Cranfield. 2004. "The Earnings Effects of Multilateral Trade Liberalization: Implications for Poverty." World Bank Economic Review 18 (2): 205­36. Hertel, Thomas W., and Will Martin. 2000. "Liberalizing Agriculture and Manufactures in a Millen- nium Round: Implications for Developing Countries." World Economy 23: 455­70. Jorgenson, D. W. 1984. "Economic Methods for Applied General Equilibrium Analysis." In Applied General Equilibrium Analysis, ed. H. E. Scharf and J. B. Shoven, 139­203. Cambridge: Cambridge University Press. Kehoe, Timothy J. 2005. "An Evaluation of the Performance of Applied General Equilibrium Models of the Impact of NAFTA." In Frontiers in Applied General Equilibrium Modeling: Essays in Honor of 30 Poverty and the WTO: Impacts of the Doha Development Agenda Herbert Scarf, ed. Timothy J. Kehoe, T. N. Srinivasan, and John Whalley, 341­77. Cambridge: Cam- bridge University Press. Mattoo, A., R. Rathindran, and A. Subramamian. 2001."Measuring Services Trade Liberalization and Its Impact on Economic Growth." Policy Research Working Paper 2655, World Bank, Washington, DC. McKitrick, R. R. 1998."The Econometric Critique of Computable General Equilibrium Modeling: The Role of Functional Forms." Economic Modeling 15 (4): 543­73. Valdes, A., and A. McCalla. 2004. "Where the Interests of Developing Countries Converge and Diverge." In Agriculture in the New Trade Agenda, ed. M. Ingco and L. A. Winters, 136­150. Cam- bridge: Cambridge University Press. Walmsley, T., T. W. Hertel, and E. I. Ianchovichina. Forthcoming. "Assessing the Impact of China's WTO Accession on Investment." Pacific Economic Review. Winters, L. A. 2000. "Trade and Poverty: Is There a Connection?" In Trade, Income Disparity and Poverty, ed. D. Ben David, H. Nordstrom and L. A. Winters, Special Study No. 5. Geneva: WTO. Winters, L. A., N. McCulloch, and A. McKay. 2004. "Trade Liberalisation and Poverty: The Evidence So Far." Journal of Economic Literature 42: 72­115. World Bank. 2004. Global Economic Prospects 2004--Realizing the Development Promise of the Doha Agenda. Washington, DC: World Bank. WTO. 2004. Doha Work Programme: Decision Adopted by the General Council on 1 August, 2004. Geneva: WTO. 2 Scenarios for Global Trade Reform Kym Anderson and Will Martin Introduction Since the failure of the Trade Ministerial meeting in Seattle in late 1999, the WTO membership has stressed continually that the organization's first multilateral trade negotiation round will have development at its heart (WTO 2001, 2004b). Simultaneously, the United Nations has emphasized that trade reform is crucial for achieving its first Millennium Development Goal of halving, between 1990 and 2015, the proportion of people earning less than US$1 a day (Zedillo, Messer- lin, and Nielson 2005). This chapter examines the decisions that have been made since the launch of the WTO's DDA in late 2001, and it draws out their implications for genuine trade reform. A set of numerical scenarios is developed to provide the basis for explor- ing the DDA's potential impacts on the economy and on poverty in the various developing countries considered in subsequent chapters. The chapter begins by examining what is at stake. It shows where the major potential gains from complete trade liberalization would come from in terms of sectors and, within agriculture, in terms of the three key classes of policy measures (import restrictions, export subsidies, and domestic support). How far the DDA will go toward reaching that potential is then explored. A sensible starting point is to begin with what the Doha Declaration and the Doha Work Program (DWP) Framework promise in terms of generalities and then examine the specifics of the annexes to that decision of August 1, 2004, beginning with agriculture and then turning to nonagricultural trade reforms and the special provisions for develop- ing countries. These are then brought together to provide a series of overall Doha scenarios, the world market effects of which can be estimated using global trade 31 32 Poverty and the WTO: Impacts of the Doha Development Agenda models. These world market effects, outlined in chapter 3, will then form the basis for the individual country case studies of potential poverty impacts of the DDA. What Is at Stake? Merchandise trade barriers have come down a long way in developed countries since the signing of the General Agreement on Tariffs and Trade (GATT) in 1947, with the notable exceptions of agriculture and textiles. Services trade barriers remain high, though, and many merchandise trade barriers in developing coun- tries have only recently begun to be lowered. What would happen if all those bar- riers and agricultural subsidies were to be removed? In particular, how important are the various types of barriers currently in place? Those questions are addressed by Hertel and Keeney (2006), using a medium- run closure of the GTAP­AGR model1 in comparative static, perfectly competitive mode. That is, they assume full employment and partially mobile factors (with some segmentation between agriculture and nonagriculture), no imperfect com- petition or economies of scale, and no dynamic gains from trade (drawing on Hertel 1997). The model uses Version 6.05 of the GTAP database (see www.gtap.org), which in turn owes much to the protection estimates in the MAcMap database assembled by the Centre d'Etudes Prospectives et d'Informa- tions Internationales (CEPII) in Paris (Bouët and others 2004). That database is far superior for present purposes to earlier versions in that it includes agricultural tariff rate quotas (TRQs), ad valorem equivalents of specific tariffs, all preferential tariffs (reciprocal, as in free-trade areas, as well as nonreciprocal, as in various developing country preference schemes), and is compatible with measures of bound tariffs which we obtain from Behir, Jean, and Laborde. Shifting to zero agricultural subsidies and complete free trade in goods and services from that post­Uruguay Round base is conservatively estimated by Hertel and Keeney to boost comparative static, global welfare by US$151 billion per year.2 Developing countries would enjoy a disproportionately large share of those gains (more than one-quarter, well above their one-sixth share of global GDP).3 The reason is twofold: they have relatively high tariffs themselves and, more important (as discussed see below), their exporters face much higher tariffs than do exporters from the high-income countries themselves. These full-liberalization numbers provide a benchmark against which to compare the gains likely from any partial reform to emerge from the Doha Round. What are the policy measures contributing most to those potential gains from full trade liberalization? Table 2.1 focuses just on merchandise trade policies. It decomposes them into measures affecting markets for agriculture and food, for textiles and clothing, and for other goods. Scenarios for Global Trade Reform 33 Table 2.1. Sectoral Contributions to Comparative Static Estimates of Economic Welfare Gains from Completely Removing Merchandise Trade Barriers Globally, Post-Uruguay Round (percent of total global gains) Benefiting Agriculture Textiles and Other All region and food clothing merchandise merchandise High-income countriesa 52 1 21 74 Low- and middle- income countries 14 10 2 26 All countries 66 11 23 100 Source: Drawn from several tables in Hertel and Keeney (2006). a. "High-income" here refers to developed countries, the four East Asian "tiger" economies, and all European transition economies. The results in table 2.1 are striking. Although agriculture contributes only 4 percent to global GDP, policies for that sector are responsible for an enormous two-thirds of the global cost of merchandise trade protection. More than three- quarters of that contribution is due to high-income countries' policies, so not sur- prisingly, high-income countries would gain most from the removal of farm pro- grams--but developing countries also gain a sizable portion. Indeed, agriculture would contribute more than half the total gains to developing countries from removing all merchandise trade restrictions globally (14 of its total 26 percentage points). Second in importance is textiles and clothing liberalization. Although it would contribute only one-fifth as much to global welfare as agricultural reform, its con- tribution to welfare in developing countries would be considerably greater, equal to more than two-thirds of that from farm trade reform and accounting for most of developing countries' gains from nonfarm merchandise reform (middle row of table 2.1). What would this freeing up of merchandise trade do to developing country exports? It turns out that every region in the Hertel-Keeney study would expand its exports of both farm and textile products by more than 10 percent, with the exception of Argentina's and Brazil's textile exports, which fall slightly in the wake of those countries' large increase in farm exports. Global annual exports of those two product groups each expand by about one-fifth in real terms. For developing countries, the increase is US$48 billion in agricultural goods and US$69 billion in textiles and clothing, valued at 2001 base prices. Other merchandise exports grow 34 Poverty and the WTO: Impacts of the Doha Development Agenda by only 5 percent globally, but that still represents an additional US$50 billion per year for developing countries, with only Brazil and Indonesia not enjoying an increase. Unsurprisingly, exports of textiles and clothing grow more for develop- ing than for high-income countries, but in the case of farm products, they each expand by a similar amount (almost US$50 billion per year). The changes for individual developing countries and regions are summarized in table 2.2. What happens when services trade reform also is included? Estimates are very much more difficult to obtain for this category, especially when it potentially involves FDI (commercial presence) and temporary labor migration (movement of natural persons). Hertel and Keeney (2006) do not attempt to include the latter services,4 but they do provide an estimate (based on distortion measures from Francois, van Meijl, and Tongeren [2005]) for direct trade in a few services, Table 2.2. Change in Exports Valued at 2001 Base Prices from Completely Removing Merchandise Trade Barriers Globally, Post-Uruguay Round (2001 US$ billions) Agriculture Textiles and Other Country/region and food clothing merchandise China 7.7 36.7 7.8 Indonesia 1.1 3.2 -0.3 Other Southeast Asia 6.7 8.8 6.5 India 6.0 4.8 14.6 Other South Asia 1.3 4.2 1.3 Argentina 1.5 -0.2 1.4 Brazil 8.3 -0.6 -0.7 Other Latin America 5.6 3.9 4.9 North Africa and Middle East 2.2 7.4 10.4 South African Customs Union 1.8 0.1 0.2 Other Southeastern Africaa 1.3 -0.0 0.4 Other Sub-Saharan Africa 1.8 0.5 3.3 All low- and middle-income countries 48.3 68.7 49.9 All high-income countriesb 47.7 18.3 178.7 All countries 96.0 87.0 228.6 Source: Summarized from Hertel and Keeney (2006, table 6). a. Botswana, Madagascar, Malawi, Mozambique, Tanzania, Uganda, Zambia, and Zimbabwe. b. "High-income" here refers to developed countries, the four East Asian "tiger" economies, and all European transition economies. Table 2.3. Sectoral and Regional Contributions to Comparative Static Estimates of Economic Welfare Gains from Completely Removing Merchandise and Services Trade Barriers Globally, Post- Uruguay Round (percent of total global gains) Agriculture and food Textiles and other services Liberalizing Market Domestic Export Total region access support subsidies agriculture Clothing Merchandise Services sectors High-income countriesa 30 2 1 33 3 6 Low- and middle-income countries 4 0 0 4 4 6 All countries 34 2 1 37 7 12 44 100 Scenarios Source: Drawn from several tables in Hertel and Keeney (2005). a. "High-income" here refers to developed countries, the four East Asian "tiger" economies, and all European transition economies. for Global Trade Reform 35 36 Poverty and the WTO: Impacts of the Doha Development Agenda including transportation, trade, and business services.5 Table 2.3 shows the huge potential importance of liberalizing services trade. Even with just this small subset of services included, they enhance considerably the potential gains from trade reform, accounting for 44 percent of those total gains from goods and services reforms. That exceeds agriculture's share of that total, namely 37 percent (with other merchandise accounting for just 19 percent). Table 2.3 also exposes the relative importance of the three separate pillars of agricultural support programs: import market access inhibited by tariffs and TRQs, domestic support measures, and export subsidies. According to these results, import market access measures deliver by far the greatest prospects for gains from agricultural reform--12 times the combined contribution of domestic support and export subsidies. Farm export subsidies are now of relatively minor importance globally, thanks to their cuts following the Uruguay Round. But devel- oping countries would lose slightly from their removal because as a group they are net importers of the subsidized program commodities.6 The loss is equal to -0.7 percent of the global gains included in table 2.3 (or -0.8 percent if implicit export subsidies in the form of food aid and export credits are also included). All develop- ing countries would gain from the removal of developed-country domestic subsi- dies. Four West African countries made headlines by stressing the plight of their cotton producers--that cotton has been explicitly targeted for domestic subsidy cuts in the Doha Round, even though the potential contribution to global welfare is only a small fraction of the gains from liberalizing all farm subsidies and tariffs. This again underlies the importance of making each of these reforms part of a comprehensive liberalization package so that there is scope for all countries to gain. Because of the differing signs of the welfare effects of the various policy meas- ures, it is necessary to look more closely to see if there are some countries that could lose from a move to free trade. Hertel and Keeney (2006, table 2.11) show that all the big developing countries would gain from complete global farm trade reform. But in North Africa and the Middle East, as well as in Sub-Saharan Africa (other than the southeastern African countries separately identified in the tables), there are more losses than gains. For some of those countries, this is because the prices of their imports would rise (for example, importers of temperate foods no longer to be dumped on international markets), but more commonly it is because the prices of their agricultural exports fall (for example, because their tariff pref- erence margin disappears as developed countries move to zero tariffs; another rea- son for the losses under free trade is that they and their neighbors expand exports so much as to depress their price in international markets enough to offset effi- ciency gains from their own reform). Likewise, Hertel and Keeney show some farm-exporting developing countries (Argentina, Brazil and Sub-Saharan Africa other than South Africa) losing from textiles and clothing reform for terms of Scenarios for Global Trade Reform 37 trade reasons. The terms of trade also explain the loss to smaller Latin American countries from other merchandise trade reform. But for services, each of the regions shown would gain. When all sectors are considered together, the only region not shown to benefit from a move to free trade is Sub-Saharan Africa other than the identified countries of southeastern Africa. But even that latter group loses less than US$1 billion per year, which could be easily offset by extra foreign aid. And the Hertel-Keeney study shows that their loss becomes a gain as soon as some trade facilitation is provided. Whether poor people within those or other developing countries would gain or lose is not answered directly in that study, but is instead the subject of chapters 4­17 of this volume. Given these results, a key question is: how close can the Doha Round come to realizing this potential for the global economy and especially for developing coun- tries? In addressing that question, it needs to be kept in mind from the outset that WTO trade negotiators are focusing on reductions not to the applied tariffs used in the Hertel-Keeney analysis but rather to members' legally bound tariffs, export subsidies, and domestic support commitments. These are higher than applied rates in nearly all countries, but especially so in most developing countries. Hence, if cuts to bound rates are sufficiently small, or the gap between bound and applied rates sufficiently large, no actual reform need take place from an agreed set of bound rate reductions. Before turning to empirical modeling results that bear on this issue, it is necessary to first consider what the agreed framework and modali- ties for the negotiations are likely to deliver, particularly in the key policy areas identified in table 2.3. What the Doha Declaration and July 2004 Framework Decision Promise The DDA, as outlined in the Doha Ministerial Declaration of November 2001, is unequivocal in its aim to ensure that developing countries benefit from this first multilateral trade negotiation since the formation of the WTO. On its first page, the declaration states: "International trade can play a major role in the promotion of economic development and the alleviation of poverty. We recognize the need for all our peoples to benefit from the increased opportunities and welfare gains that the multilateral trading system gen- erates. The majority of WTO Members are developing countries. We seek to place their needs and interests at the heart of the Work Programme adopted in this Declaration. Recalling the Preamble to the Marrakesh Agreement, we shall continue to make positive efforts designed to ensure that developing countries, and especially the least-developed among them, secure a share in the growth of world trade commensurate with the needs of their economic development. In this context, enhanced market access, balanced 38 Poverty and the WTO: Impacts of the Doha Development Agenda rules, and well targeted, sustainably financed technical assistance and capacity-building programmes have important roles to play. We recognize the particular vulnerability of the least-developed countries and the spe- cial structural difficulties they face in the global economy. We are committed to addressing the marginalization of least-developed countries in international trade and to improving their effective participation in the multilateral trading system. We recall the commitments made by Ministers at our meetings in Marrakesh, Singapore and Geneva, and by the international community at the Third UN Conference on Least- Developed Countries in Brussels, to help least-developed countries secure beneficial and meaningful integration into the multilateral trading system and the global econ- omy. We are determined that the WTO will play its part in building effectively on these commitments under the Work Programme we are establishing." (WTO 2001) Since that meeting in late 2001, WTO members have sought to implement those commitments. However, the following Trade Ministerial meeting, in Can- cún in September 2003, ended with acrimony and without an agreement on how to proceed. At Cancún, developing countries made it abundantly clear that further progress would not be possible without a commitment by developed countries to significantly lower their import barriers and agricultural subsidies (including importantly for cotton, despite its relatively minor role in developed country agri- culture, see Sumner [2006]). An intense period of consultations in July 2004 ended with a decision on how the DWP should proceed (WTO 2004b). The deci- sion again stresses the importance of keeping development at the heart of the DDA and particularly stresses agricultural reform as key to that. In its annexes, the decision provides guidance as to how a Doha agreement might be structured, with frameworks for establishing modalities for agriculture and nonagricultural mar- ket access, and for negotiations on trade facilitation, as well as providing recom- mendations for trade in services. The analysis in this chapter begins with the three agricultural pillars. Agricultural Market Access The impacts of agricultural trade reforms cannot be understood without a detailed analysis of the structure of the protection to which they are being applied. For this analysis, the MAcMap database developed by the International Trade Centre (ITC), Geneva, and Paris-based CEPII. The latest version of the database covers tariffs for 2001 and takes into account ad valorem tariffs, specific tariffs, and tariff preferences (Bouët and others 2004). Bound tariffs have also been pre- pared on a comparable basis and documented in Bchir, Jean, and Laborde (forth- coming). Some important presimulation changes were required before the Doha analysis, including modification of China's tariffs to take account of its WTO Scenarios for Global Trade Reform 39 accession commitments, the phase-in of remaining commitments from the Uruguay Round (especially the elimination of quotas on textiles and clothing trade), and changes due to the accession in April 2004 of 10 new members to the EU. Fortunately, the MAcMap database permitted isolation of these changes so that the model-based analysis could implement these "pre-experiment" shocks before proceeding with the simulation of the DDA. Overview of Agricultural Tariffs A critical issue in analyzing tariff protection in agriculture has to do with the treatment of TRQs. Here, the effective tariff assumed to be applied on any given TRQ commodity depends on whether the quota is filled or not. If the quota is less than 90 percent filled, the in-quota tariff is assumed to apply on these commodi- ties. If the quota is between 90 and 99 percent filled, the effective tariff is assumed to be the average of the in- and the out-of-quota tariff. If the quota is more than 99 percent filled, then the out-of-quota tariff is applied.7 Another critical issue has to do with the role of specific tariffs (for example, US $10 per ton), which cannot be aggregated and used in global modeling exer- cises without first being converted to ad valorem equivalent form. To do so requires an assumption about the price of the product being traded. In the MAcMap database, the associated prices for a given commodity are allowed to vary between developed, developing, and LDCs. And because poor countries tend to export low-value products, the specific tariffs affect a larger share of their export value. Using the MAcMap database, table 2.4 shows that specific tariffs are indeed quite important. The global average tariff of 17 percent is made up of 11 percent- age points from ad valorem tariffs and an additional 6 percentage points from the ad valorem equivalents of non­ad valorem measures. However, there are large variations between countries and country groups around these levels. For devel- oped countries as a group, the average tariff of 14 percent is made up of only 4 percent contributed by ad valorem tariffs and 10 percent from the ad valorem equivalents of specific, mixed, or compound duties. As noted previously, this is a particular concern to developing countries, because specific tariffs tend to impose greater burdens on low-income country exports. Within the developed country group, there is considerable variation in average tariffs, with Japan having an aver- age agricultural tariff of 36 percent, mostly derived from non­ad valorem tariffs, and the European Free Trade Association (EFTA) having a tariff of 29 percent. The average agricultural tariff in the EU is considerably lower, at 12 percent; in the United States and Australia, they are lower again at 3 percent. 40 Poverty and the WTO: Impacts of the Doha Development Agenda Developing countries, at 20 percent, have a higher average tariff than devel- oped countries, but only 2 percentage points of this protection is provided by specific tariffs. Average tariffs are extremely high in the Republic of Korea,8 at 94 percent, and also high in China, India, Pakistan, and Sub-Saharan Africa. The net agricultural exporting Mercado Común del Sur (Mercosur) region has quite low tariffs, at an average of 5 percent. Interestingly, both LDCs as a group, and Sub- Saharan African LDCs in particular, have relatively low tariffs, consistent with the tendency noted in the political economy literature for poor countries to have low agricultural protection (see, for example, Anderson and Hayami [1986]). Another feature of agricultural protection evident in table 2.4 is the height of the barriers by countries protecting with TRQs. The analysis by de Gorter and Kli- auga (2006) indicates that these products cover 20 percent of agricultural tariff lines in the countries using TRQs and 52 percent of the value of production.9 The fact that average applied tariffs on these commodities are so high, although some imports are permitted at lower in-quota tariffs, is striking testimony to the impor- tance of protection on these commodities. Given their large import shares in countries such as the Republic of Korea, and their extremely high tariff levels, it is clear that protection on these commodities constitutes a very important part of total agricultural protection in developed countries and in those developing countries using these measures. Had these commodities been automatically treated as sensitive products, as was proposed in WTO (2004a), it is clear that a very large share of total protection would have been shielded from liberalization. Another key element of the geography of market access is the relationship between applied and bound tariffs. The higher are bindings relative to applied rates, the larger the reductions in bound rates that must be made before applied rates change to improve market access. The gap between applied and bound duties has two origins: the binding overhang, which is the gap between bound and most-favored-nation (MFN) tariffs, and preferential arrangements, which cause a difference between the MFN and applied rate.10 It is widely known that there was substantial binding overhang in many devel- oping countries after the Uruguay Round. Developing countries had the right to set their tariff bindings without reference to previous levels of protection, under the so-called ceiling binding option. Many developing countries used this right to set their bindings at high--and frequently uniform--levels, such as 150 or 250 percent. The effects are evident in table 2.5, where the bound tariff in developing countries is 2.4 times the average applied rate. While developed countries did not have the right to use ceiling bindings, nego- tiators used a highly protected base period (1986­88) and many members used so- called "dirty tariffication" to set their tariff rates well above the previously prevailing Scenarios for Global Trade Reform 41 Table 2.4. Key Features of Applied Agricultural Tariffs, by Country and Region, 2001 (trade weighted averages, percent) Ad Country/ Overall valorem Specific Tariff for TRQ region average tariffs tariffs TRQs share Australia 3 2 1 1 6 Bangladesh 14 14 0 0 0 Cameroon 17 17 0 0 0 Canada 10 8 1 31 21 China 39 39 0 6 22 Indonesia 32 5 1 14 12 Japan 36 10 26 103 9 Korea, Rep. of 94 94 0 226 39 Mexico 11 11 0 34 24 Pakistan 30 10 21 0 0 India 55 54 1 0 0 Turkey 14 14 0 0 0 United States 3 1 2 11 17 Mercosur 13 13 0 7 3 EFTA 29 2 27 58 34 Association of Southeast Asian Nations (ASEAN) 11 8 4 32 8 Philippines 47 10 0 30 7 Sub-Saharan LDCs 13 13 0 0 0 Other Sub- Saharan Africa 26 26 0 0 0 Maghreb 18 16 2 39 14 South African Customs Union 13 4 9 16 56 EU15 12 3 9 36 22 Russia 13 12 1 0 0 Developed countries 14 4 10 37 17 Low- and middle-income countries 21 19 2 64 12 LDCs 13 13 0 0 0 All countries 17 11 6 47 14 Source: Jean, Laborde, and Martin (2006). 42 Poverty and the WTO: Impacts of the Doha Development Agenda Table 2.5. Bound and Applied Agricultural Tariff Rates, by Country and Region, 2001 (trade weighted averages, percent) Country/region Bound MFN Applied CVa CV MFN tariff tariff tariff bound applieda Australia 6 4 3 2 2 Bangladesh 157 14 14 9 3 Cameroon 80 17 17 0 0 Canada 20 19 10 24 24 Chinab 16 51 39 11 19 Indonesia 58 7 32 14 0 Japan 62 52 36 81 90 Korea, Rep. of 104 120 94 43 58 Mexico 49 32 11 18 25 Pakistan 108 30 30 3 5 India 153 55 55 23 13 Turkey 50 16 14 13 7 United States 6 6 3 14 14 Mercosur 34 13 13 2 1 EFTA 71 48 29 22 24 ASEAN 60 12 11 25 10 Philippines 35 10 47 6 0 Sub-Saharan LDCs 63 15 13 2 1 Other Sub- Saharan Africa 104 27 26 1 7 Maghreb 38 19 18 11 5 South African Customs Union 52 14 13 12 5 EU15 21 17 12 41 36 Russia 16 13 13 6 0 Developed countries 27 22 14 37 38 All low- and middle-income countries 48 27 21 14 15 LDCs 78 14 13 4 2 All countries 37 24 17 26 27 Source: Jean, Laborde, and Martin (2006). a. CV is the weighted coefficient of variation for the power of this tariff. b. The bound average duty reported for China takes into account commitments not in effect in 2001, hence its lower level compared to the MFN rate. Scenarios for Global Trade Reform 43 average applied tariffs (Hathaway and Ingco 1996). Table 2.5 indicates that binding overhang is substantial in developing countries, and smaller, but by no means non- existent, in developed countries.11 These results are broadly consistent with the find- ings of Martin and Wang (2004), which are based on an entirely different dataset. For developed countries, the average bound rate was almost twice as high as the applied rate. But this difference mainly comes from the large gap between MFN and applied rates, reflecting the importance of preferential agreements and TRQs in reducing average applied rates below their MFN levels. The difference is large in relative terms for all developed countries. A key feature of table 2.5 is the sharp difference between countries in the extent of the binding overhang. Low- income countries, such as the LDC group, tend to have a very large degree of binding overhang, with bindings for the LDC group six times applied rates. For Bangladesh, the difference is more than 150 percentage points. In Japan, the United States, and the EU, average bound rates are more than 50 percent above the applied rates, suggesting that relatively large cuts in bound rates would be needed before there would be any reductions in applied rates. Tiered Formulas The DWP Framework proposes the use of a tiered formula with deeper cuts in higher tariffs. Attempts to apply higher rates of tariff reduction to higher tariffs confront a problem of discontinuities. This is evident in figure 2.1, which maps tariffs before application of the formula to postformula tariffs using the transition points of 15 and 90 percent, and cut rates of 40, 50, and 60 percent as suggested in the Harbinson proposal (WTO 2003). The discontinuity problem is most evident around the 90 percent transition point, where a tariff of 90 percent becomes a tar- iff of 45 percent, and a tariff of just over 90 percent becomes a tariff of 36 percent. This discontinuity would not only result in a change in the ordering of tariffs, but could potentially raise the costly variability of tariffs. Most important from a political economy perspective, such discontinuities would likely create major political resistance from firms just above each of the transition points. One way to deal with the discontinuity problem is to follow the approach of the progressive income tax, where the higher proportional rate is applied to the part of the tariff that lies above the limit of the lower band. Although it has the disadvan- tage of cutting high tariffs by less in absolute terms than a proportional cut (because the lower portion of the tariff is cut at a lower rate), it does impose the higher cut on higher tariffs required by the DWP Framework. Further, it provides a continuous mapping from the old tariffs to the new, as depicted in figure 2.2. With this approach, Jean, Laborde, and Martin (2006) examine a range of tariff- cutting scenarios. Before these scenarios were developed, a priori adjustments were 44 Poverty and the WTO: Impacts of the Doha Development Agenda performed to introduce the developments already agreed to before any tariff reductions arising from the DDA. These pre-experiment changes include the expansion to the EU25, the phase-in of remaining agricultural commitments by developing countries,12 the abolition of the quotas on exports of textiles and cloth- ing originally imposed under the Multifibre Arrangement, and the tariff reforms agreed to by countries that recently acceded to the WTO, particularly China. The Doha scenario of central interest in this book is performed by cutting the bound tariffs but reducing applied rates only when and to the extent that the new bound rate is below the initial applied rate. This Doha scenario uses a tiered for- mula with inflexion points at 15 and 90 percent and marginal tariff cuts13 of 45, 70, and 75 percent in developed countries. For developing countries, the inflexion points were placed at 20, 60, and 120 percent and the marginal cuts at 35, 40, 50, and 60 percent. Also consistent with the special and differential treatment provi- sions in the DWP Framework, LDCs are not required to undertake any reduction commitments. Jean, Laborde, and Martin (2006) perform several alternative experiments exploring the consequences of different specifications of possible variants of the tiered formula. In particular, they examine the consequences of including flexibil- ity in the form of so-called sensitive products and, for developing countries addi- tionally, so-called special products. They allow countries to classify 2 percent (or 5 percent) of tariff lines as sensitive products. They also allow just developing coun- tries to classify a similar proportion of tariff lines as special products. Jean, Laborde, and Martin found that allowing sensitive and special products in these ways greatly diminished the extent of trade liberalization. They also compared the tiered formula with a proportional cut approach and examined the effects of a tar- iff cap of 200 percent, finding that it was quite effective in restoring the discipline lost by allowing sensitive and special products. Another policy experiment com- pared the Swiss formula, which reduces higher tariffs more sharply, with the tiered formulas.14 Finally, they examined whether the loss of market access resulting from flexibility is due to the inclusion of alcoholic beverages and tobacco--where revenue rather than protection might be the reason for high tariffs--and found that this was not the case. The implications of this full range of scenarios for poverty are explored in chapter 17 of this volume. Agricultural Domestic Support Reductions in domestic support are, like reductions in tariffs, to be undertaken using tiered formulas with deeper cuts in higher rates of protection. The DWP Framework sets out in some detail a number of constraints to be satisfied, includ- ing restrictions on the AMS and on this trade-distorting support plus blue box Scenarios for Global Trade Reform 45 Figure 2.1. Converting the Harbinson Formula into a Tiered Formula 60 50 cent) 40 (per 30 tariff 20 New 10 0 10 20 30 40 50 60 70 80 90 100 110 120 Old tariff (percent) Source: Authors. Figure 2.2. Tiered Tariff-Cutting Formula without Discontinuities 80 70 60 50 cent) (per 40 tariff 30 New 20 10 0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 Original tariff (percent) Source: Authors. Note: Upper line shows the 40 percent reduction. Lower line shows the 40:50:60 percent progressive tiered formula. 46 Poverty and the WTO: Impacts of the Doha Development Agenda and de minimis support in developed countries. In addition, product-specific support is to be capped at historic levels. As in the case of market access, special and differential treatment provisions allow for smaller cuts and longer implemen- tation periods in developing countries. Analyses of this aspect of the agreement by Jensen and Zobbe (2006) and by Hart and Beghin (2006) lead to the conclusion that there are two key features determining the impact of these disciplines. One is the ability of WTO members to "abolish" the market price support (MPS) component of their domestic sup- port without making any substantive changes in policy; the other is the gap between actual support and WTO commitments. For members such as the EU, Japan, and the Republic of Korea, where a very large fraction of domestic support is provided through administered market prices, the current AMS can be sharply reduced by abolishing (as Japan already has in the case of rice) the "administered prices" that are central to calculating the MPS component of AMS (see WTO [1995], annex 3, paragraph 8). Under current rules, this removes MPS from the reported current AMS, while leaving commitment levels unchanged. Once this feature of the AMS and the gaps between the latest notified levels of support and the binding commitment levels are taken into account, it becomes clear that very large reductions in domestic support commitments would be needed to bring about reductions in distorting support. Jensen and Zobbe (2006) examined the effects of a rule requiring developed countries providing domestic support above 20 percent of production to cut by 75 percent, while others cut by 60 percent, and developing countries cut by 40 percent. As shown in table 2.6, they conclude that this would require reductions in domestic support in only six WTO members: Australia, the EU, Iceland, Norway, Thailand, and the United States. Thus, these reductions in domestic support, as reported in table 2.6, are the only ones used in the Doha simulations underlying the analysis in this book. (Of course, the above full-liberalization scenarios consider the impact of abolishing domestic support for agriculture.) Clearly, unless the loophole provided by the ability to abolish the current AMS without removing it from the binding is addressed, very large reductions in domestic support commitments will be required to bring about reductions in actual domestic support, and most of those reductions will be in lightly protected countries. Agricultural Export Subsidies Farm export subsidies are inconsistent with GATT rules and for that reason alone deserve to be eliminated (Hoekman and Messerlin 2006). As noted by Hertel and Keeney (2006) and reported in table 2.3, the overall impact of eliminating export Scenarios for Global Trade Reform 47 Table 2.6. Cuts in Domestic Support under a Tiered Formula with 75 Percent Cuts in High-Supporting Countries Country/region Required cut in domestic support (%) United States 28 EU 16 Iceland 1 Australia 10 Norway 18 Thailand 30 Source: Jensen and Zobbe (2006), tables 9.6 and A9.1. subsidies would be very small relative to the impact of abolishing barriers to mar- ket access. Additional analysis showed that to be true even when implicit export subsidies in the form of food aid and export credits are included. These subsidies are heavily concentrated in a few high-income countries, with the EU alone accounting for almost 90 percent of the total (although Switzerland has the high- est average export subsidy rate, at three times the EU average). EU export subsi- dies are heavily concentrated in just a few commodities, with beef, sugar, dairy products, wheat, coarse grains, and processed foods accounting for more than 80 percent of total export subsidies. However, the DWP Framework does envisage complete abolition of export subsidies, although the outcome for tariffs and domestic support is likely to be considerably less than full abolition. Furthermore, as will be seen in chapter 14, among others, export subsidies have a disproportionate impact on the poor. For these reasons, and because of their historical importance in the policy debate, export subsidies rightly represent a key part of any DDA in agriculture. Nonagricultural Market Access Given the importance of food and agriculture to the poor--both as a source of earnings, as well as a large expenditure item--this chapter, and indeed the entire book, focuses heavily on reforms to these sectors. This is also an area where the decision of August 1, 2004, is most explicit. However, developing countries as a group are far more reliant on nonagricultural exports for foreign exchange earn- ings, thus this area must also be considered in the scenarios. Lacking explicit direc- tion from the 2004 decision, the core Doha scenario simply assumes that developed countries cut bound tariffs on nonagricultural products by 50 percent and devel- oping countries by two-thirds of this amount (and again no cut by the LDCs). 48 Poverty and the WTO: Impacts of the Doha Development Agenda Implications of Doha for Market Access To obtain some insights into the implications of Doha reform for market access, table 2.7 presents estimates of the average tariffs recently and prospectively facing each country. The first column of this table reports the average tariffs in 2001 for agriculture and food alone (table 2.7a) and for total merchandise trade (table 2.7b). It then reports those averages after the baseline pre-experiment in which significant policy changes to the end of 2004 are implemented (EU expansion, Uruguay Round completion, WTO accession by China and other countries). The tariff averages are based on 2015 trade flows, because this is the end of the baseline period used in the closing chapter (17) of this book. The pre-experiment tariffs are then compared with what they would be following either of two Doha scenar- ios--the core scenario involving lower reform commitments for developing coun- tries and none for LDCs (Doha), and another scenario in which those countries forgo that element of special and differential treatment and liberalize to the same degree as developed countries (Doha-All). First, note in table 2.7b that the global average tariff declines about one-tenth (from 5.2 to 4.7 percent) in the pre-experiment, thanks to the above-mentioned actual reforms (especially those of China) during the 2001­04 period. That is also true of average import tariffs applying to food and agriculture alone. Second, the Doha scenario reduces average tariffs globally by a further one- quarter, from 4.7 to 3.5 percent (table 2.7b). Most of that is concentrated in agri- culture and clothing, whose average tariffs fall about one-third. (Average agricul- tural tariffs fall from 15.2 to 9.9 percent [table 2.7a].) A number of countries, particularly in East Asia and Western Europe show very sharp tariff reductions in agriculture, because the top-down nature of the tiered formula substantially reduces the tariffs on farm products that are subject to particularly high tariff rates. Third, as a result of those cuts to tariffs, the average rate of protection facing many exporters falls substantially. Table 2.8 reports this information, showing the average tariffs applied to exports from each region for food and agriculture alone (table 2.8a) and for all merchandise trade (table 2.8b). For example, food exporters, such as those that are members of the Cairns Group, experience reduc- tions in the average tariffs applied to their exports that amount to as much as five percentage points or more under the Doha scenario. But even exporters in some net-food-importing economies--such as China, the EU, and even the Republic of Korea--experience large declines in the agricultural tariffs they face. And the same is true even for some low-income food exporters, such as Cambodia, Cameroon, and Zimbabwe, although they have been previously eligible for prefer- ences into the EU. Scenarios for Global Trade Reform 49 Table 2.7a. Agricultural and Food Tariffs, Percent Baseline Doha Doha- Country/region 2001 2015 2015 All Australia and New Zealand 2.6 2.6 1.7 1.7 EU25 plus EFTA 13.9 13.9 7.0 7.0 United States 2.4 2.4 1.7 1.7 Canada 9.0 9.0 4.9 4.9 Japan 29.4 29.3 14.7 14.7 Korea, Rep. of, and Taiwan, China 55.0 53.0 27.9 18.7 Hong Kong, China, and Singapore 0.1 0.1 0.1 0.1 Argentina 7.1 7.1 6.9 6.1 Bangladesh 12.7 12.7 12.7 11.9 Brazil 5.0 5.0 4.9 4.4 China 37.6 10.3 7.9 6.9 India 50.3 49.9 45.5 37.4 Indonesia 5.0 5.0 4.9 4.5 Thailand 29.7 16.7 13.5 11.0 Vietnam 37.1 37.1 37.1 37.1 Russia 13.5 13.5 8.7 6.5 Mexico 11.6 10.3 8.6 6.5 South Africa 8.8 8.6 8.1 6.6 Turkey 16.7 16.6 13.8 10.6 Rest of South Asia 21.3 21.1 20.9 16.5 Rest of East Asia 13.7 13.4 12.7 11.2 Rest of Latin America and the Caribbean 11.0 10.8 9.8 8.9 Rest of Europe and Central Asia 16.0 15.7 14.3 12.9 Middle East and North Africa 14.1 13.1 11.5 10.4 Selected Sub-Saharan African countries 11.9 11.8 11.5 11.0 Rest of Sub-Saharan Africa 21.4 21.2 19.6 16.1 Rest of the world 12.1 11.8 11.5 9.4 High-income countries 16.0 15.9 8.2 7.5 Quad (Canada, EU, Japan, and U.S.) plus Australia and New Zealand 13.6 13.6 7.0 7.0 Other high-income countries 31.3 30.2 15.9 10.7 Low- and middle-income countries 17.7 14.2 12.4 10.6 Developing countries--WTO definition 20.0 16.9 13.0 10.7 Middle-income countries 16.5 12.1 10.3 8.9 Low-income countries 22.2 22.0 20.7 17.5 World total 16.7 15.2 9.9 8.8 50 Poverty and the WTO: Impacts of the Doha Development Agenda Table 2.7b. Aggregate Merchandise Trade Tariffs, percent Baseline Doha Doha- Country/region 2001 2015 2015 All Australia and New Zealand 4.8 4.7 4.0 4.0 EU25 plus EFTA 3.2 3.1 1.7 1.7 United States 1.8 1.8 0.9 0.9 Canada 1.4 1.4 0.8 0.8 Japan 5.2 5.1 2.7 2.7 Korea, Rep. of, and Taiwan, China 7.6 7.3 5.0 3.9 Hong Kong, China, and Singapore 0.0 0.0 0.0 0.0 Argentina 10.0 10.0 9.9 9.2 Bangladesh 18.4 18.4 18.4 18.3 Brazil 9.5 9.5 9.2 8.5 China 13.6 6.1 4.3 3.3 India 28.1 26.9 23.1 19.6 Indonesia 4.8 4.7 4.6 4.5 Thailand 10.2 8.6 8.0 7.3 Vietnam 16.7 16.7 16.7 16.7 Russia 9.7 9.7 6.5 4.8 Mexico 5.1 5.0 4.7 4.3 South Africa 6.6 6.6 6.0 4.9 Turkey 2.5 2.4 2.2 2.0 Rest of South Asia 14.6 14.4 14.3 13.3 Rest of East Asia 4.6 4.5 4.1 3.7 Rest of Latin America and the Caribbean 9.1 9.1 8.3 7.7 Rest of Europe and Central Asia 5.0 4.9 4.7 4.4 Middle East and North Africa 9.8 9.4 8.8 8.3 Selected Sub-Saharan African countries 8.7 8.5 8.3 8.2 Rest of Sub-Saharan Africa 16.2 16.3 15.9 15.0 Rest of the world 9.1 9.0 8.8 8.3 High-income countries 2.9 2.9 1.6 1.6 Quad (Canada, EU, Japan, and U.S.) plus Australia and New Zealand 2.7 2.7 1.5 1.5 Other high-income countries 4.1 3.9 2.6 2.1 Low- and middle-income countries 9.9 8.4 7.5 6.8 Developing countries--WTO definition 8.5 7.3 6.3 5.6 Middle-income countries 8.9 7.2 6.3 5.6 Low-income countries 15.9 15.6 14.6 13.4 World total 5.2 4.7 3.5 3.2 Source: GTAP 6.05 and CEPII scenario file; authors' aggregation. Note: EU tariffs are net of intra-EU trade. Scenarios for Global Trade Reform 51 Table 2.8a. Agricultural and Food Tariffs Faced by Exporters, percent Baseline Doha Doha- Country/region 2001 2015 2015 All Australia and New Zealand 17.4 16.8 9.5 8.7 EU25 plus EFTA 15.5 14.7 11.2 9.7 United States 23.4 20.5 12.2 10.7 Canada 9.5 9.2 6.0 5.6 Japan 9.6 8.1 6.1 4.9 Korea, Rep. of, and Taiwan, China 11.5 10.2 6.9 6.4 Hong Kong, China, and Singapore 21.4 20.0 18.2 17.0 Argentina 23.0 13.7 9.6 8.1 Bangladesh 3.9 2.9 2.8 2.7 Brazil 24.2 19.5 11.1 9.6 China 26.8 26.2 14.0 11.2 India 10.4 9.5 6.9 6.2 Indonesia 14.1 13.0 11.6 9.9 Thailand 19.6 19.2 11.5 10.4 Vietnam 9.9 9.9 9.9 9.9 Russia 7.0 6.2 4.6 4.3 Mexico 4.9 4.8 3.4 3.3 South Africa 15.0 14.8 10.4 9.5 Turkey 11.1 10.9 8.0 7.3 Rest of South Asia 15.0 14.5 10.9 9.1 Rest of East Asia 17.2 16.6 15.0 12.4 Rest of Latin America and the Caribbean 11.4 11.2 6.3 5.8 Rest of Europe and Central Asia 8.8 8.8 7.5 7.3 Middle East and North Africa 8.8 8.8 7.1 6.9 Selected Sub-Saharan African countries 10.1 9.3 6.0 5.6 Rest of Sub-Saharan Africa 11.1 10.9 6.8 6.5 Rest of the world 23.0 21.7 14.3 13.7 High-income countries 17.8 16.3 10.6 9.4 Quad (Canada, EU, Japan, and U.S.) plus Australia and New Zealand 17.9 16.4 10.5 9.3 Other high-income countries 15.5 14.1 11.4 10.7 Low- and middle-income countries 15.6 14.1 9.2 8.1 Developing countries--WTO definition 15.6 14.1 9.3 8.3 Middle-income countries 16.5 14.7 9.4 8.2 Low-income countries 12.4 11.7 8.7 8.0 World total 16.7 15.2 9.9 8.8 52 Poverty and the WTO: Impacts of the Doha Development Agenda Table 2.8b. Aggregate Merchandise Trade Tariffs Faced by Exporters, percent Baseline Doha Doha- Country/region 2001 2015 2015 All Australia and New Zealand 8.0 7.6 4.7 4.3 EU25 plus EFTA 6.7 6.1 5.1 4.6 United States 4.7 4.2 2.9 2.7 Canada 1.3 1.2 0.9 0.8 Japan 5.5 4.5 3.5 3.1 Korea, Rep. of, and Taiwan, China 6.1 4.9 3.8 3.4 Hong Kong, China, and Singapore 5.7 4.6 3.5 3.2 Argentina 14.1 9.4 7.2 6.2 Bangladesh 5.1 4.9 2.9 2.8 Brazil 9.8 8.4 5.6 5.0 China 6.1 5.9 4.1 3.7 India 6.6 6.3 4.9 4.6 Indonesia 6.5 5.9 4.8 4.4 Thailand 7.6 7.0 4.8 4.4 Vietnam 7.6 7.6 7.6 7.6 Russia 2.1 1.8 1.5 1.4 Mexico 1.0 0.9 0.8 0.7 South Africa 5.5 5.4 4.3 3.9 Turkey 5.5 5.0 4.1 3.8 Rest of South Asia 9.6 9.3 6.7 6.1 Rest of East Asia 3.7 3.1 2.6 2.3 Rest of Latin America and the Caribbean 5.2 5.1 3.4 3.2 Rest of Europe and Central Asia 3.2 3.0 2.8 2.7 Middle East and North Africa 2.8 2.8 2.6 2.5 Selected Sub-Saharan African countries 4.8 4.4 3.1 2.9 Rest of Sub-Saharan Africa 3.4 3.5 2.4 2.4 Rest of the world 10.4 9.9 6.4 6.2 High-income countries 5.4 4.7 3.6 3.3 Quad (Canada, EU, Japan, and U.S.) plus Australia and New Zealand 5.3 4.7 3.6 3.3 Other high-income countries 6.0 4.8 3.7 3.4 Low- and middle-income countries 4.9 4.5 3.4 3.1 Developing countries--WTO definition 5.1 4.6 3.4 3.2 Middle-income countries 4.7 4.3 3.2 2.9 Low-income countries 6.3 6.0 4.6 4.3 World total 5.2 4.7 3.5 3.2 Source: GTAP 6.05 and CEPII scenario file; authors' aggregation. Note: EU tariffs are net of intra-EU trade. Scenarios for Global Trade Reform 53 The Doha-All scenario reported in the final columns of tables 2.7 and 2.8 shows what would happen to average tariffs if developing countries and LDCs were to fully reciprocate the tariff cuts made by developed countries. This would lower their average applied tariffs by one-tenth (slightly more so for farm prod- ucts and slightly less so for other merchandise), or by about as much again as under the central Doha scenario. Although that may seem a nontrivial additional reduction, it is still proportionately smaller than for developed countries because of the higher degree of tariff binding overhang in those poorer countries. Yet it would boost substantially the likely economic benefits of Doha to developing countries not only because of more own-country reform but also because it would mean more market access abroad for developing country exporters, in two respects: (a) because a significant share of developing countries' trade is with other developing countries and (b) because such an increased willingness to par- ticipate in the Doha Round would provide an opportunity for developing coun- tries to demand greater access to developed-country markets for farm and textile products (and possibly more foreign aid for trade facilitation investments). Conclusions The two objectives of this chapter were to examine the potential impacts of removing trade barriers and agricultural subsidies (what's at stake) and specify a core Doha scenario that can be used in subsequent chapters as representative of the July 2004 DWP Framework agreement. In terms of potential reform, the analysis by Hertel and Keeney (2006) shows that increased agricultural market access is the key to successful liberalization of merchandise trade, accounting for well over half the potential economic welfare gains to developing countries and the world as a whole from removing all merchandise trade distortions and farm subsidies. Within agriculture, the potential gains from market access are shown to be far more important than those from abolition of domestic support and export subsidies, accounting for 93 percent of the gains from total agricultural liberalization. The chapter also lays out the approach used to assess the potential impact of partial liberalization of the type envisaged in the DWP Framework agreement. In the case of market access, it highlights a potentially serious problem with tiered formulas: the problem of discontinuities at the transition point from one depth of cut to another. In the simulations used in this book, that problem is resolved by applying higher cuts at the margin--that is, applying only to the part of each tar- iff above the transition point from one tariff to another. Meaningful evaluation of tariff reforms using a tiered formula can only be undertaken using detailed data on WTO tariff bindings, applied tariff data that 54 Poverty and the WTO: Impacts of the Doha Development Agenda reflect the importance of specific tariffs, and tariff preferences granted either uni- laterally or in the context of bilateral or regional arrangements. In this chapter, the data on applied tariffs are drawn from the CEPII-ITC MAcMap database, and the tariff bindings are based on comparable data prepared by CEPII. Before the policy experiments are performed, the base data from 2001 are modified to take into account commitments reached before 2005, including China's WTO accession commitments, remaining Uruguay Round tariff cuts, the abolition of the export quotas introduced under the Multifibre Arrangement, and the incorporation of 10 additional countries into the EU in May 2004. Finally, the tariff cuts are per- formed by applying a relatively deep-cutting tiered formula (with maximum tariff cuts of 75 percent) to the bound rates. Impacts of these simulations on prices and other economic variables are presented in chapter 3 of this book. Subsequent chapters of the book will explore the poverty impacts of the core Doha scenario and compare them with the potential impacts of complete global trade reform. Both chapters 14 and 17 will also examine the market and welfare implications of developing countries and LDCs becoming full participants in this round's tariff cuts as in the Doha-All scenario. Notes 1. The GTAP-AGR is a new variant on the standard GTAP model of the global economy that fine- tunes the specification of the food and agricultural sectors in line with recent econometric work. See Keeney and Hertel (2006) for details. 2. This is considerably below the estimate reported in Anderson and others (2001), based on the GTAP Version 5.4 database for 1997, despite the inclusion of liberalization of commercial services in the results presented here from Version 6.05 for 2001. The reasons for the differences include the reductions in global protection between 1997 and 2001, the inclusion of nonreciprocal tariff prefer- ences for low-income countries in the latest dataset, and structural changes in the global economy. This estimate is also below that reported in Anderson, Martin, and van der Mensbrugghe (2005) for merchandise trade. Those authors reconcile the difference as due (in decreasing order of importance) to their reporting for the year 2015 when the world economy (and especially the economies of more highly protected developing countries) is larger, their inclusion of some of the dynamic gains from trade, and their use of larger (longer-run) trade elasticities. 3. In 2001, based on World Bank (2003), where the term "developing countries" here excludes East Asia's four newly industrialized "tigers" and the transition economies of Eastern Europe and the for- mer Soviet Union. 4. The gain to developing countries from the temporary movement of labor could be enormous, though, swamping the welfare gains from product trade liberalization. See Winters and others (2003) for an early estimate using the GTAP model. 5. The distortion estimates are highly speculative. They have been obtained from an econometric model that aims to predict the volume of services imports into each region in the absence of trade bar- riers. The tariff equivalent of this barrier is obtained by increasing import prices until import volumes are scaled back to their observed level. 6. See, for example, table 3 in Dimaranan, Hertel, and Keeney (2004). 7. In those cases where there are quota administration reasons for believing this 90 percent rule underestimates TRQ protection, the following discussion will underestimate the gains from Doha. Scenarios for Global Trade Reform 55 8. Korea is a self-declared developing country in the WTO, although a member of the OECD and a high-income country by World Bank standards. 9. This percentage corresponds to products that are at least partly protected by a TRQ (see de Gorter and Kliauga [2006] for details). It should therefore be considered as an upper bound. 10. Note, however, that given the methodology used here, TRQs are also a source of difference between MFN and applied rates, because the MFN duty is the out-of-quota tariff rate in this database, but this is not the case for the applied duty as soon as the quota is less than 99 percent filled. 11. Computing perfectly comparable information on MFN and bound ad valorem equivalent tar- iffs is a complex task. Because treating the information concerning MFN tariffs sometimes involves specific difficulties, such as incomplete raw information, it is possible that the extent of the binding overhang found here for developed countries, although already small, is still overstated (because the level of MFN duties might have been slightly understated in some cases). This is likely to be the case in particular for the EU and for Japan. 12. Developing countries had 10 years from 1994 to implement their Uruguay Round commit- ments on agriculture. 13. An initial simulation was undertaken with cuts of 35, 60, and 65 percent and is reported in Jean, Laborde, and Martin's agricultural scenario 7 (2006). It was not chosen as the base for further simula- tions because it created insufficient liberalization to allow evaluation of the effects of liberalization erosion through the addition of sensitive and special products. 14. For details of the results, see Jean, Laborde, and Martin (2006). References Anderson, K., B. Dimaranan, J. Francois, T. Hertel, B. Hoekman, and W. Martin. 2001. "The Burden of Rich (and Poor) Country Protectionism on Developing Countries." Journal of African Economies 10 (3): 227­57. Anderson, K., and Y. Hayami. 1986. The Political Economy of Agricultural Protection: East Asia in Inter- national Perspective. Boston, London, and Sydney: Allen and Unwin. Anderson, K., and W. Martin, eds. 2006. Agricultural Trade Reform and the Doha Development Agenda. Basingstoke, U.K.: Palgrave Macmillan; Washington, DC: World Bank. 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. Basingstoke, U.K.: Palgrave Macmillan; Washington, DC: World Bank. Bchir, M., S. Jean, and D. Laborde. Forthcoming. "Binding Overhang and Tariff-Cutting Formulas: A Systematic, Worldwide Quantitative Assessment." Working Paper, CEPII, Paris. http://www.cepii.fr/anglaisgraph/news/accueilengl.htm Bouët, A., Y. Decreux, L. Fontagné, S. Jean, and D. Laborde. 2004. "A Consistent, ad valorem Equivalent Measure of Applied Protection across the World: The MAcMap-HS6 Database." Working Paper, CEPII, Paris: http://www.cepii.fr/anglaisgraph/news/accueilengl.htm de Gorter, H., and E. Kliauga. 2006. "Reducing Tariffs Versus Expanding Tariff Rate Quotas." In Agri- cultural Trade Reform and the Doha Development Agenda, ed. K. Anderson and W. Martin. Bas- ingstoke, U.K.: Palgrave Macmillan; Washington, DC: World Bank. Dimaranan, B., T. Hertel, and R. Keeney. 2003. "OECD Domestic Support and the Developing Coun- tries." In The WTO, Developing Countries and the Doha Development Agenda, ed. B. Kuha-Gasno- bis. New York: Palgrave Macmillan. Francois, J. F., H. van Meijl, and F. van Tongeren. 2005. "Trade Liberalization in the Doha Round." Eco- nomic Policy 20(42): 349­91. Hart, C. E., and J. C. Beghin. 2006. "Rethinking Agricultural Domestic Support under the World Trade Organization." In Agricultural Trade Reform and the Doha Development Agenda, ed. K. Anderson and W. Martin. Basingstoke, U.K.: Palgrave Macmillan; Washington, DC: World Bank. 56 Poverty and the WTO: Impacts of the Doha Development Agenda Hathaway, D., and M. Ingco. 1996. "Agricultural Liberalization and the Uruguay Round." In The Uruguay Round and the Developing Countries, ed. W. Martin and L. A. Winters, pp. 30­58. Cam- bridge and New York: Cambridge University Press. Hertel, T. W., ed. 1997. Global Trade Analysis: Modeling and Applications. Cambridge and New York: Cambridge University Press. Hertel, T. W., and R. Keeney. 2006. "What Is at Stake: The Relative Importance of Import Barriers, Export Subsidies, and Domestic Support." In Agricultural Trade Reform and the Doha Development Agenda, ed. K. Anderson and W. Martin. Basingstoke, U.K.: Palgrave Macmillan; Washington, DC: World Bank. Hoekman, B., and P. Messerlin. 2006. "Removing the Exception of Agricultural Export Subsidies." In Agricultural Trade Reform and the Doha Development Agenda, ed. K. Anderson and W. Martin. Bas- ingstoke, U.K.: Palgrave Macmillan; Washington, DC: World Bank. Jean, S., D. Laborde, and W. Martin. 2006."Consequences of Alternative Formulas for Agricultural Tar- iff Cuts." In Agricultural Trade Reform and the Doha Development Agenda, ed. K. Anderson and W. Martin. Basingstoke, U.K.: Palgrave Macmillan;Washington, DC: World Bank. Jensen, H., and H. Zobbe. 2006. "Consequences of Reducing Limits on Aggregate Measures of Sup- port." In Agricultural Trade Reform and the Doha Development Agenda, ed. K. Anderson and W. Martin. Basingstoke, U.K.: Palgrave Macmillan; Washington, DC: World Bank. Keeney, R., and T. Hertel. 2005. "GTAP-AGR: A Framework for Assessing the Implications of Multilat- eral Changes in Agricultural Policies." GTAP Technical Paper 24, Purdue University, West Lafayette, IN. https://www.gtap.agecon.purdue.edu/resources/tech_papers.asp Martin, W., and Z. Wang. 2004. "Improving Market Access in Agriculture." Unpublished paper,World Bank, Washington, DC. Sumner, D. A. 2006. "Reducing Cotton Subsidies: The DDA Cotton Initiative." In Agricultural Trade Reform and the Doha Development Agenda, ed. K. Anderson and W. Martin. Basingstoke, U.K.: Pal- grave Macmillan; Washington, DC: World Bank. Winters, L. A., T. Walmsley, Z. K. Wang, and R. Grynberg. 2003. "Liberalizing Temporary Movement of Natural Persons: An Agenda for the Development Round." The World Economy 26 (8): 1137­61. World Bank. 2003. Global Economic Prospects and the Developing Countries. Washington, DC: The World Bank. WTO. 1995. "Agreement on Agriculture." In The Results of the Uruguay Round of Multilateral Trade Negotiations, 39­68. Geneva: WTO. ------. 2001. "Doha Ministerial Declaration." WT/MIN(01)/DEC/1, WTO Secretariat, Geneva. ------. 2003. "Negotiations on Agriculture: First Draft of Modalities for the Further Commitments." (The Harbinson draft.) TN/AG/W/1/Rev.1, WTO, Geneva. ------. 2004a. "Doha Development Agenda: Draft General Council Decision of July 2004." (The Groser draft.) JOB(04)/96, WTO, Geneva. ------. 2004b."Decision Adopted by the General Council on 1 August 2004." WT/L/579, WTO Secre- tariat, Geneva. Zedillo, E., P. Messerlin, and J. Nielson. 2005. Trade for Development. Report of the Task Force on Trade for the UN Millennium Development Goals Project, London: Earthscan. 3 Assessing the World Market Impacts of Multilateral Trade Reforms Thomas W. Hertel and Maros Ivanic Introduction To deduce the national poverty impacts of the Doha scenarios outlined in chap- ter 2, a modeling framework is needed. One approach would be to use a global model with sufficiently disaggregated households at the national level to say something about potential poverty impacts. This is the approach pursued by Ivanic in chapter 14, where he explores the poverty impacts of global trade reform in 15 focus countries. It is a useful way to offer an integrated analysis of the poverty impacts of multilateral trade reforms, but it is also a very demanding exercise that ultimately must abstract from many of the country-specific features that may play an important role in determining the poverty impacts of trade lib- eralization. Therefore, for most of the studies reported in this volume, a two-step approach is used. The first step uses a global model to estimate, for each of the tar- get countries in turn, the price changes and estimated changes in export volumes, as well as import price changes arising from the liberalization of trade in other countries. In the second stage, the authors who deal with specific countries use these as inputs into their national models, which are tailored to address particu- larly important features of their focus economy. This chapter describes the first step in the process, the approach to estimating country-specific price and export volume changes and the methodology for passing these on to the country level. 57 58 Poverty and the WTO: Impacts of the Doha Development Agenda Modeling World Price and Volume Changes Choice of Model There is always a tradeoff between the use of detailed, partial equilibrium models of trade reform and less detailed, but more comprehensive, general equilibrium approaches. A good example of the partial equilibrium approach is offered by papers collected in Aksoy and Beghin (2004). They provide a detailed exploration of the impacts of trade reform in nine different agricultural commodity markets. However, the goal of this book is to assess the poverty impacts of comprehensive trade reforms that affect not just a few agricultural commodity markets, but also food processing, textiles and apparel, and other manufacturing trade. Also, because this book assesses the impacts of such reforms on poverty, it must con- sider not only the impact on traded commodities, but also the resultant changes in the prices of services and primary factors of production. Finally, to the extent that trade reform has adverse impacts on fiscal revenue, the impacts of replacing lost tariff revenue with other tax instruments must be considered. For all of these reasons, a general equilibrium approach is required. Nevertheless, this chapter draws on the commodity-specific studies for insights into the market impacts of farm-specific support policies and also seeks to validate the general equilibrium approach used here against observed behavior in specific commodity markets. Nearly all the global general equilibrium models today draw on the global pro- duction, use, and trade database collected and maintained by Global Trade Analy- sis Project (GTAP).1 This database provides a snapshot of the world economy for a single year (Version 6 represents 2001), with complete bilateral trade flows and interindustry sales for a 57-sector disaggregation of each region (of which there are 87 in Version 6). The models built on the GTAP database often differ in their choice of parameters and model closure. Some are comparative static and some are dynamic. The simplest of these models, and therefore one of the most widely used, is the standard GTAP model (Hertel and Tsigas 1997). It has the virtue of being well documented, transparent, and highly flexible in the aggregation of sec- tors and countries. However, this very flexibility can be a limitation in the analysis of specific commodity markets. As shown in the chapter 2, a disproportionate share of the trade distortions arise in agriculture. Furthermore, as documented in chapter 15, a disproportionate share of the poor is employed in farming, particu- larly in the lowest-income economies, and the poor spend a disproportionate share of their income on food. Therefore, it is important to "get the story right" in discussing farm and food impacts of trade reform. For this reason, this chapter uses a special-purpose version of the GTAP model, GTAP-AGR, which has been designed to draw on the latest econometric research on the underlying parameters World Market Impacts of Multilateral Trade Reforms 59 governing supply and demand in the food and agricultural markets (Keeney and Hertel 2005). Model Description As documented in Hertel (1997) and on the GTAP Web site,2 the standard GTAP model includes demand for goods for final consumption, intermediate use, and government consumption; demands for factor inputs; supplies of factors and goods; and international trade in goods and services. The model uses the simplis- tic but robust assumptions of perfect competition and constant returns to scale in production activities. Bilateral international trade flows are handled using the Armington assumption, by which products are exogenously differentiated by ori- gin. This means that there is not a single "world price" for a given product--for example, rice. Rather, the price varies with the origin of the product. For example, rice exported from Thailand is differentiated from rice produced in and exported from the United States. This is a useful approach for several reasons. First, even at the level of rice (or processed sugar), commodities are heterogeneous. For exam- ple, processed sugar trade includes high-fructose corn syrup as well as processed cane sugar and processed beet sugar. The category of rice includes many different varieties, some of which substitute only imperfectly for one another. A second, closely related reason for adopting this product differentiation assumption is that it permits the explanation, within the model, of why an individual country is both an exporter and an importer of a given product. For example, the United States is an exporter of high-quality beef, but it also imports low-quality beef for use in fast food. Of course, an important and well-known consequence of the Armington assumption is the presence of relatively high optimal tariffs, even for small coun- tries. This is because every country faces a downward-sloping export demand schedule, with the small country elasticity being approximated by the elasticity of substitution among imports in the destination regions. Consequently, the optimal import tariff is roughly equal to the inverse of this Armington elasticity. In the model discussed below, the simple average of these elasticities is about seven, so that the optimal tariff for the countries analyzed in this book is in the neighbor- hood of 16 percent. Because many developing countries already have average tar- iffs in this range, further tariff cuts on a unilateral basis will likely generate aggre- gate welfare losses. This does not mean that poverty will necessarily increase, but it certainly biases the model against showing favorable developments in the wake of own-liberalization. This will be an important point to bear in mind in consid- ering the impacts of developing-country tariff cuts on poverty. Using this standard modeling framework as a starting point, GTAP-AGR incorporates some alternative representations to bring focus to the intricacies of 60 Poverty and the WTO: Impacts of the Doha Development Agenda agricultural production and markets. Several structural features have been high- lighted in the agricultural economics literature for their importance in the analy- sis of agricultural policy changes: intersectoral factor mobility and factor substitu- tion in production, crop-livestock sector interactions, consumer food demand, and trade elasticities. The manner in which these features are introduced into the model is detailed in Keeney and Hertel (2005) and is discussed briefly below. Work by the OECD (2001) on the cost and world market impacts of agricul- tural support highlights the role of factor market issues in an empirical partial equilibrium model. This work focuses on the segmentation that occurs in land, labor, and capital markets between the agricultural and nonagricultural economies, and it provides the region-specific factor supply elasticities used in the GTAP-AGR model. Keeney and Hertel (2005) also follow the OECD's factor sub- stitution regime for primary agriculture, focusing on substitution possibilities among farm-owned and purchased inputs, as well as between the two. They cali- brate the constant elasticity of substitution cost functions for farm-level sectors to the region-specific Allen elasticities of substitution provided by the OECD. The interaction between livestock and crop sectors received considerable atten- tion in the literature after European Common Agricultural Policy reform in 1992 and has continued to be an area of concern (Peeters and Surry 1997). The primary concern has to do with the ability of the livestock industry to change the mix of feedstuffs demanded in response to changing relative prices induced by farm sup- port policies. The GTAP-AGR model follows the approach of Rae and Hertel (2000) in modeling the substitution possibilities for feedstuffs in livestock pro- duction as an additional part of the firm's cost minimization problem, governed by a constant elasticity of substitution among ingredients. They calibrate this region-generic parameter to an average substitution elasticity calculated from Surry's (1990) three-stage model describing the behavior of European livestock producers, composite feed mixers, and grain producers. The importance of consumer demand for foodstuffs is prominent in the agri- cultural economics literature. The unique role of food in the consumer budget has been emphasized in much of this work, especially as it relates to the distribution of incomes (Cranfield and others 2003; Seale, Regmi, and Bernstein 2003). Inelas- tic demand in many markets, coupled with volatile supplies, translates into volatile prices for staple products, which can have a significant impact on house- holds near or below the poverty line. The GTAP-AGR model uses a recent set of estimates from a cross-country study of demand, keying on own-price and income elasticities of demand for food (Seale, Regmi, and Bernstein 2003). In the GTAP-AGR model, Keeney and Hertel (2005) calibrate the parameters of the GTAP demand system to the elasticities for the eight food aggregates and an addi- tional nonfood aggregate. World Market Impacts of Multilateral Trade Reforms 61 International trade elasticities that describe the substitution possibilities between goods differentiated by origin have received considerable attention for the important role they play in simulation models determining the terms of trade (TOT) impacts of liberalization. Hertel and others (2004) provide recent esti- mates of this substitution relationship at the same level of disaggregation as the sectors in the GTAP model. Those authors also show how the estimated gains from trade liberalization hinge critically on the value of these parameters. The research reported in this chapter makes use of their region-generic estimates of the elasticity of substitution among imported goods from different sources as specified in the model's Armington demand structure.3 Model Validation Although each of these individual model modifications is supported by the cur- rent literature, when they are used together in the context of a general equilibrium model, there remains the unanswered question of how well the model as a whole performs relative to the historical record. Valenzuela and others (2005) address this question in a validation exercise aimed at investigating how well the model performs in reproducing historical price volatility in world markets for agricul- tural products. Their approach makes use of the fact that most of the annual volatility in crop commodity markets is induced by shocks to supply--the major- ity of which are induced by exogenous natural phenomena (for example, droughts, heat waves, floods, and so on). The authors' validation experiment focuses specifically on wheat and begins by estimating a time series forecasting model to predict output trends, attributable to either technical advancements or year-to-year market signals, in wheat-producing regions. The remaining year-to-year changes are attributable to nature-induced supply-side shocks. These random shocks are used to build a distribution of annual supply shocks with which to perturb the model. The ensuing distribution of price predictions may then be compared to that observed historically in order to validate the model. Their specific validation criterion is the predicted versus observed standard deviation in wheat prices for each country, where the model prediction is obtained by solving the model repeatedly, each time sampling from the historical distribution of supply-side shocks for each region of the world. The most recent results from the work of Valenzuela and others (2005) are summarized in figure 3.1. Here, individual countries represented in the model are given by single points. Regions including multiple countries are represented by brackets, because the model predicts only one price per region, but the data from the UN Food and Agriculture Organization (FAO) data show a range of prices--differing for each country in the region. This figure shows that the 62 Poverty and the WTO: Impacts of the Doha Development Agenda Figure 3.1. Validating the GTAP-AGR Model: Predicted versus Observed Standard Deviations for Wheat 45 Other Europe 40 Rest of Latin America 35 Argentina price, 30 25 deviation Mexico Australia calculated 20 Sub-Saharan Africa Canada data United States standard15 Brazil AOF Middle East and China North Africa 10 South Asia EU15 5 Japan 0 0 5 10 15 20 25 30 35 40 GTAP-AGR Model price, standard deviation Source: Valenzuela and others (2005). model-predicted standard deviation of prices is virtually the same as that observed historically for wheat in the cases of Australia, China, the EU, the Mid- dle East and North Africa, and South Asia. For Brazil and Japan, the model over- predicts price volatility. This is hardly surprising, because both of these countries have had domestic policies in place to stabilize prices, and these policies have not been captured in the model simulations. In the remaining individual countries shown in this figure--Argentina, Canada, Mexico, and the United States--the model predicts too little volatility in wheat prices. With the exception of Mexico, which is heavily influenced by the U.S. market, these countries are major wheat exporters. As such, their domestic prices are influenced by other countries' trade policies, which often serve to destabilize markets by dumping excess production on world markets (export subsidies) in times of surplus and restricting sales in times of shortage. Because the model abstracts changes in trade policies from year to year over this period, it appears to understate the variability of prices in the major export markets. Finally, two regions stand out as having a very large range of price variation. These are the rest of Latin America and Central and Eastern Europe. Hertel, World Market Impacts of Multilateral Trade Reforms 63 Keeney, and Valenzuela (2004) attribute some of this price variation to policy volatility (particularly in Eastern Europe), as well as macroeconomic instability and the difficulty of properly deflating year-to-year price changes to obtain esti- mates of real price variation over this period. Overall, it appears that the GTAP- AGR model shows reasonably valid behavior for many regions. Improving its per- formance will likely require more attention to how individual commodity policies operate at the margin. This goes well beyond the scope of this book, which focuses instead on the impacts of eliminating or sharply reducing farm and food support policies. Although the authors have thus far conducted this validation analysis only for a single commodity, this evidence is a positive indication that GTAP-AGR is indeed a valid framework for analyzing impacts of global agricultural liberaliza- tion on world markets. Therefore, the chapter turns next to an analysis of the price impacts of multilateral agricultural trade reform. Results This section considers the global and national impacts of the full-liberalization scenario, as well as the central Doha scenario outlined in chapter 2. The first of these scenarios offers a simple, free-trade benchmark by assessing the impact of removing all merchandise trade barriers as well as domestic support in the OECD countries. The Doha scenario represents an aggressive interpretation of the July 2004 Framework Agreement for the Doha Development Agenda (DDA), includ- ing tiered formulae for reductions in tariffs and domestic support and full elimi- nation of agricultural export subsidies. This scenario also embodies special and differential treatment for developing countries whereby cuts in bound commit- ments are only two-thirds of those in developed countries, and the least developed countries (LDCs) make no cuts whatsoever. Tariff Landscapes in the Focus Countries The implications of these two scenarios for average tariffs in the focus countries are reported in table 3.1. Here, average import tariffs on products coming into the country in question4 are reported in the first column of each block under the heading "original." Thus, in Bangladesh, the average tariff on primary agriculture imports is 6 percent, whereas it is 20 percent on other primary products and processed foodstuffs. Table 3.1 also reports information on the average rate of protection facing a given country's exports. In the case of Bangladesh, this is 8 percent for agriculture, 2 percent for other primary products, and 1 percent for foodstuffs. Comparing tariff entries within a column shows which sectors are more heavily protected and therefore likely to contract under unilateral abolition 64 Poverty and Table 3.1. Tariff Rates on Imports to and from the Focus Regions the WTO: Bangladesh Brazil China To country From country To country From country To country From country Impacts of the Sector Original Doha Doha-All Original Doha Doha-All Original Doha Doha-All Original Doha Doha-All Original Doha Doha-All Original Doha Doha-All Doha Primary agriculture 6 6 6 8 8 8 2 2 2 15 10 10 8 6 6 39 22 22 Other primary 20 20 20 2 2 2 0 0 0 1 1 1 0 0 0 2 1 1 Development Food 20 20 20 1 1 1 9 9 8 21 11 11 11 9 8 19 12 11 Textile 30 30 30 6 3 3 16 16 14 7 5 5 10 6 5 10 6 6 Other manfactures 16 16 16 3 2 2 10 10 9 4 4 3 6 4 3 3 3 2 Agenda Table 3.1. (Continued) W Indonesia Mexico Mozambique orld To country From country To country From country To country From country Market Impacts Sector Original Doha Doha-All Original Doha Doha-All Original Doha Doha-All Original Doha Doha-All Original Doha Doha-All Original Doha Doha-All Primary agriculture 2 2 2 6 5 5 11 9 9 2 2 2 7 7 7 9 8 8 of Other primary 0 0 0 2 1 1 5 5 5 1 1 1 5 5 5 0 0 0 Multilateral Food 9 9 9 17 15 15 11 9 9 8 6 6 16 16 16 3 2 2 Textile 9 9 9 11 8 8 7 5 4 1 1 1 22 22 22 15 14 10 Other manfactures 5 4 4 4 3 3 4 4 4 1 1 1 8 8 8 0 0 0 Trade Reforms 65 66 Poverty and the WTO: Table 3.1. (Continued) Impacts Philippines Russia Vietnam To country From country To country From country To country From country of the Doha Sector Original Doha Doha-All Original Doha Doha-All Original Doha Doha-All Original Doha Doha-All Original Doha Doha-All Original Doha Doha-All Development Primary agriculture 6 6 6 11 10 10 5 3 3 4 4 4 11 11 11 7 7 7 Other primary 3 3 3 1 1 1 1 1 0 0 0 0 4 4 4 2 2 2 Food 11 10 10 9 7 6 17 11 11 7 5 5 44 44 44 12 12 12 Textile 7 7 7 11 6 6 15 10 7 8 5 5 31 31 31 10 10 10 Other manufactures 2 2 2 1 1 1 8 6 4 3 2 2 12 12 12 7 7 7 Agenda Source: GTAP database, Version 6.0. World Market Impacts of Multilateral Trade Reforms 67 of trade barriers. Comparing the two "original" columns in a given row shows whether imports are relatively protected when compared to exports and hence whether a given industry is likely to expand or contract global free trade in a given product. Of course, the actual full-liberalization scenario will be a combination of these two forces. The next column in table 3.1 (labeled "Doha") reports the average tariffs after the Doha reforms. Note that in the case of most of the focus countries, the post- Doha tariffs are the same as the original tariffs. In the cases of Bangladesh and Mozambique, both of which are LDCs, this is due to the fact that the DDA does not require them to make tariff cuts. In the case of many of the other developing countries in this table, there is also no difference. The Doha-negotiated cut in the tariff bindings has no impact because the bound tariffs are far in excess of their applied rates (recall the discussion of binding overhang in chapter 2), thus cutting the bindings still leaves them above the applied rates in many cases. Indeed, the Doha scenario has a measurable impact only in the cases of Brazil, China, Mexico, and Russia,5 where the binding overhang (the amount by which the bound rate exceeds the applied tariff rate) is less pronounced. The change in tariffs levied on a given country's exports can also be observed in table 3.1. This is smallest for the LDCs, which already receive preferential access into many markets, as well as for Mexico, which already has nearly free trade with its largest trading partners. For China, the post-Doha tariffs are quite a bit lower--nearly 50 percent in the case of primary agriculture exports. Full Liberalization Tables 3.2­3.6 report the results from the full-liberalization experiment. These tables summarize changes in import and export prices, import and export vol- umes, and commodity trade balances, respectively.6 Although the model itself contains 29 regions, because of space constraints, and in keeping with the poverty emphasis of this book, these changes are reported only for the world as a whole and for the 10 focus countries for which general equilibrium, national case studies are provided in this book. Begin with the impact of global trade liberalization on import prices as reported in table 3.2. Price changes in this table are grouped into six broad cate- gories: primary agriculture, other primary products, manufactured food products, textiles and apparel, other manufactures, and services. The first thing to note is the relatively larger price changes for primary agriculture. As shown in chapter 2, this is where the bulk of the commodity market interventions occur, when OECD countries remove domestic support for farm commodities, supply is reduced and world prices tend to rise. When coupled with a reduction in protection for 68 Poverty and the WTO: Impacts of the Doha Development Agenda Table 3.2. Full Liberalization: Import Prices for All Regions a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Import price index 0.5 1.1 0.3 0.7 0.9 1.0 0.6 0.9 0.8 0.4 0.7 Primary agriculturea 6.1 9.5 5.8 5.3 10.8 13.4 17.0 5.1 9.0 1.2 9.2 Paddy rice 22.2 3.0 6.2 4.0 3.0 9.4 52.6 14.4 8.3 3.8 9.2 Wheat 9.0 2.7 8.2 9.1 8.8 7.8 11.4 8.1 8.7 1.3 3.7 Cereal grains 12.2 5.3 7.4 8.0 14.5 8.2 12.6 4.8 8.4 12.7 5.0 Fruits and vegetables 1.6 ­0.7 4.1 ­1.3 2.3 2.8 1.1 4.0 2.2 0.8 3.0 Oilseeds 15.3 14.8 5.8 7.6 15.4 14.8 19.2 7.8 12.2 5.2 8.5 Raw sugar 3.8 2.2 1.7 ­2.1 12.3 2.4 2.7 2.3 2.3 2.4 2.6 Plant fibers 23.1 10.4 4.0 1.2 18.6 15.2 57.4 1.9 16.1 1.9 24.2 Other crops 0.3 ­1.3 ­1.1 ­1.6 1.2 0.1 0.7 2.3 0.1 ­0.8 0.4 Cattle 2.5 ­1.3 2.3 ­2.6 5.0 6.5 4.1 4.7 6.6 ­0.1 3.3 Animal products 2.2 0.9 1.9 0.1 2.2 ­2.3 2.7 3.5 1.1 0.8 ­3.6 Raw milk 0.8 0.1 ­0.3 ­0.7 ­0.5 ­0.6 ­0.3 ­0.6 ­0.6 ­0.8 ­0.8 Wool 6.3 3.1 ­2.7 ­1.0 6.6 ­1.9 4.0 ­3.6 1.0 ­0.5 3.9 Other primarya 0.5 0.5 0.6 0.1 0.5 0.1 0.4 1.7 0.7 0.6 1.2 Forestry 0.3 1.9 ­0.2 ­0.5 0.1 0.5 ­0.1 0.9 1.8 ­0.1 1.6 Fishing 1.4 1.1 ­0.2 ­0.2 1.3 1.3 0.8 1.8 2.1 0.8 1.1 Coal 1.0 ­2.4 0.5 0.2 1.3 0.3 0.7 0.6 1.6 2.2 0.6 Crude oil 0.5 0.4 0.6 0.1 0.5 ­0.2 0.2 ­0.4 0.5 0.1 0.6 Natural gas 0.5 ­1.3 ­2.7 0.5 0.8 0.4 0.3 1.6 0.4 0.4 0.4 Other minerals 0.7 ­0.1 0.1 0.1 0.6 1.0 0.4 2.3 1.1 0.2 1.0 Fooda 2.8 4.8 2.4 6.0 2.2 4.2 2.5 2.5 3.8 7.4 3.3 Bovine meat products 8.4 1.1 3.5 9.0 4.7 4.8 2.4 4.7 1.5 16.3 6.3 Other meat 3.4 3.8 1.5 3.4 1.3 ­2.0 0.6 2.7 ­2.2 3.2 ­0.6 Vegetable oils and fats 3.4 4.7 3.1 2.1 1.2 ­0.1 3.8 3.4 ­6.0 0.7 1.6 Milk 11.8 5.7 7.9 17.0 8.4 8.6 5.4 1.9 8.1 7.6 7.6 Processed rice 7.7 5.2 1.7 9.8 15.6 10.6 3.8 7.2 6.8 4.9 5.4 Sugar 4.6 3.3 6.0 4.7 3.5 5.1 0.4 2.6 6.7 4.0 4.4 Food products 0.4 ­0.9 0.3 0.6 ­1.4 ­0.2 0.4 1.0 0.6 ­0.1 ­0.2 Beverages and tobacco 0.1 0.9 0.2 ­0.4 0.4 ­0.1 0.1 0.6 2.0 ­0.6 2.5 Textilea 0.0 0.7 0.5 ­0.1 0.8 0.7 ­0.1 ­0.2 0.9 ­0.4 0.8 Textiles 0.2 0.8 0.4 0.0 0.7 0.7 0.3 ­0.1 0.8 ­0.6 0.8 Wearing apparel ­0.2 0.1 0.1 ­0.3 1.2 ­0.1 ­0.8 ­0.8 0.6 ­0.2 0.0 World Market Impacts of Multilateral Trade Reforms 69 Table 3.2. (Continued) a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Other manufacturesa 0.1 0.1 0.0 ­0.1 0.5 0.5 ­0.1 0.5 0.5 ­0.1 0.6 Leather products 0.4 ­0.7 0.2 0.4 ­1.2 ­0.8 0.0 0.4 ­0.3 0.1 ­1.2 Wood products 0.5 1.1 0.2 0.0 1.5 0.6 0.1 0.1 1.2 0.0 1.3 Paper products 0.3 0.9 0.0 0.4 1.0 0.8 0.0 3.0 1.0 0.0 1.6 Petroleum, coal 0.1 0.4 ­0.3 0.2 ­0.6 0.2 0.0 0.3 ­0.2 0.2 0.4 Chemical, plastic 0.0 ­0.4 ­0.2 ­0.6 0.3 0.3 ­0.1 0.8 0.4 ­0.2 0.3 Mineral products 0.4 1.2 0.3 ­0.2 1.1 1.0 0.2 4.0 1.1 0.0 1.8 Iron and steel 0.2 0.1 0.2 ­0.5 0.6 0.6 0.2 0.5 0.5 ­0.4 0.7 Metals 0.0 ­1.4 ­0.3 ­1.3 0.4 0.7 ­0.2 0.5 0.7 ­1.1 0.9 Metal products 0.2 ­0.3 0.1 ­0.6 0.7 0.8 0.0 0.2 0.6 0.0 1.1 Motor vehicles ­0.3 0.1 ­0.2 ­0.3 0.2 0.1 ­0.1 ­1.7 0.1 ­0.2 ­0.2 Transport equipment 0.0 ­0.1 ­0.2 0.3 ­0.5 0.1 ­1.2 ­0.6 0.0 ­0.3 0.6 Electronic equipment 0.4 0.1 0.1 0.1 0.6 0.6 0.2 0.3 0.4 0.1 0.8 Machinery 0.1 ­0.4 0.1 ­0.1 0.5 0.6 0.0 0.2 0.6 ­0.1 0.8 Manufactures 0.0 1.0 0.2 ­0.1 0.1 0.6 0.4 0.0 0.5 ­0.1 0.7 Electricity 0.1 ­1.5 ­0.1 ­1.2 1.5 0.0 0.0 0.2 0.2 0.0 0.3 Gas 0.1 0.3 0.1 0.1 0.1 0.1 0.0 0.0 0.2 0.0 0.1 Water 0.6 0.5 0.5 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 Construction 0.3 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 Servicesa 0.7 0.7 0.5 0.5 1.6 0.6 0.5 0.5 0.8 0.4 0.4 Trade 1.5 2.3 0.9 0.9 2.3 1.0 1.0 0.9 1.9 0.5 0.9 Transport 0.5 0.5 0.4 0.4 0.3 0.3 0.5 0.1 0.4 0.4 0.4 Water transport 0.4 0.2 0.3 0.3 0.3 0.4 0.2 0.2 0.4 0.3 0.4 Air transport 0.4 0.3 0.3 0.3 0.3 0.4 0.3 0.3 0.3 0.3 0.3 Communication 0.6 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.4 0.5 0.4 Financial services 0.6 0.5 0.5 0.5 0.6 0.6 0.5 0.5 0.5 0.5 0.4 Insurance 0.5 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 Business services 0.6 0.6 0.5 0.5 0.5 0.5 0.7 0.5 0.5 0.4 0.4 Recreational services 0.4 0.4 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.2 Government services 0.3 0.2 0.1 0.1 0.2 0.2 0.2 0.2 0.2 0.2 0.1 Dwelling 1.5 0.2 6.0 ­2.2 2.6 4.3 ­0.2 ­0.6 2.8 ­0.7 22.8 Source: Authors' simulations. a. Calculated as a weighted average of respective price changes, excluding intra-EU trade 70 Poverty and the WTO: Impacts of the Doha Development Agenda imported products that simultaneously boosts demand, prices rise significantly. The world average price for primary agricultural products rises by 6.1 percent, rel- ative to the numeraire price, which is an index of world primary factor prices. However, there is considerable variation within this broad category of goods, with world average prices rising more than 20 percent for plant-based fibers (primarily cotton) and paddy rice. These are followed by oilseeds and cereals. Wheat and wool also show above-average price rises. Processed food products show the next highest world average price increase, and this category is led by dairy and beef products, as well as processed rice. Most other world average price changes are quite modest. This variation in commodity-specific price changes also gives rise to regional variation in import price indexes--particularly in food and agriculture. For example, Bangladesh is a heavy importer of cotton and oilseeds, the prices of which rise substantially following global trade reform. Accordingly, Bangladesh's national import price index for primary agriculture rises by 9.5 percent. The largest country-specific average agriculture import price rise is 17 percent in the case of Mexico. As a result of geographic proximity, as well as the success of the North American Free Trade Agreement, the Mexican economy is heavily inte- grated with that of the United States. And the United States has very large subsi- dies for cotton, rice, oilseeds, and grains. Mexico sources the vast majority of her imports of these products from the United States and thus feels the full force of this subsidy removal. Indeed, Mexican import prices for paddy rice and plant- based fibers rise by about two-thirds. These import price rises for Mexico are far larger than those for Bangladesh, which sources only a modest share of imports of cotton and oilseeds from the United States--or even the EU, for that matter. In summary, the commodity-specific import price indexes depend impor- tantly on the source of a given country's imports. This in turn may be traced back to the Armington assumption of product differentiation. Of course, the degree to which prices for a particular commodity are permitted to diverge will depend on the elasticity of substitution among imports within that particular commodity group. As noted above, these elasticities are taken from the recent econometric study by Hertel and others (2004). Their estimates of this key parameter range from 2.6 for other animal products to 34.4 for natural gas. Table 3.3 reports the export price changes for the same full-liberalization experiment. Although the world price changes are nearly the same as for imports--the difference being due to the presence of international trade and transport margins--the regional price changes are quite different. The difference between the changes in the national export and import price indexes is a measure of the change in each country's TOT, which is also reported at the top of table 3.3. World Market Impacts of Multilateral Trade Reforms 71 Bangladesh, Cameroon, Mexico, Mozambique, the Philippines, and Russia all experience negative TOT shocks from full trade liberalization, with Bangladesh's loss standing out in particular. The strong deterioration in the Bangladeshi TOT is driven by three factors. The first is the fact that Bangladesh is a net importer of food products, and the world price of these products has risen (recall table 3.2). The second stems from the fact that the country already faces very low tariffs in its export markets--particularly the industrial countries, where Bangladesh has tariff-free access as an LDC (recall table 3.1). When the industrial countries liberalize, this preference is eroded and export prices fall. The third factor derives from Bangladesh's own-liberalization in the full-liberalization scenario, which involves elimination of the relatively high tariffs imposed on imports into Bangladesh (table 3.1). When these tariffs are eliminated, imports increase. Therefore, exports must also increase to restore external trade balance. The bulk of Bangladeshi exports is in textiles and apparel, and because a substantial increase in these exports is required, the boost in demand following liberalization elsewhere in the world is insufficient to accom- plish this, so a large price reduction ensues (-6.4 percent on average for textiles and apparel). Brazil, China, Indonesia, and Vietnam all experience TOT gains, with the largest gains by far going to Brazil. Global liberalization generates a substantial increase in the demand for Brazilian agricultural exports (recall the large tariffs on Brazilian exports reported in table 3.1). As a result, Brazilian agricultural export prices rise by an average of 13.7 percent--twice the rate of increase worldwide. To restore external trade balance, the prices of all primary factors in Brazil must rise relative to the world average factor price index. This in turn boosts prices for nonagricultural products and services, thus the associated export prices are seen rising across the board in table 3.3. A similar situation, although much more muted, occurs in China, which takes advantage of falling import tariffs in East Asia to increase its farm and food exports (recall the high protection against China's agriculture exports in table 3.1). In Indonesia, the improved TOT are driven by natural resource­based products and light manufactures. Recall also from table 3.1 that Indonesian tariffs are low relative to those imposed against its exports, so full liberalization tends to boost demand for its goods and hence raise the prices for its products. Tables 3.4 and 3.5 report trade volume changes for the full-liberalization experiment. Worldwide, trade volume (including services trade) rises by 7.2 per- cent in this scenario. The largest increases come in food products, followed by tex- tiles and apparel, then primary agriculture and other manufactures. Other pri- mary product markets have few trade policy distortions and so experience only a 72 Poverty and the WTO: Impacts of the Doha Development Agenda Table 3.3. Full Liberalization: Export Prices for All Regions a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Terms of trade ­6.1 5.1 ­2.8 0.2 1.1 ­2.1 ­2.2 ­0.5 ­0.8 0.8 Export price index 0.5 ­5.1 5.4 ­2.2 1.1 2.2 ­1.4 ­1.3 0.3 ­0.4 1.5 Primary agriculturea 6.4 ­1.0 13.7 ­2.6 4.5 1.1 ­1.9 ­0.7 ­1.0 0.2 2.7 Paddy rice 23.8 ­3.6 15.7 ­4.2 10.0 2.3 42.5 ­1.2 ­4.5 1.3 7.0 Wheat 9.4 ­3.2 11.7 ­3.4 3.2 0.9 18.7 0.0 0.5 0.4 ­12.4 Cereal grains 12.8 ­3.2 16.2 ­4.1 5.1 2.4 12.0 ­2.0 ­1.6 ­0.8 3.9 Fruits and vegetables 1.8 ­3.5 13.1 ­2.7 4.4 2.9 ­2.8 ­2.1 ­1.9 ­2.2 5.3 Oilseeds 16.1 ­4.6 13.5 3.0 6.6 10.2 52.2 1.1 4.6 1.6 22.5 Raw sugar 3.6 ­4.0 14.1 ­4.3 3.8 0.7 ­2.2 2.2 ­1.4 ­2.0 8.7 Plant fibers 24.8 1.8 14.3 0.1 3.5 12.5 2.3 0.3 7.7 11.9 19.7 Other crops 0.3 ­4.3 13.2 ­3.4 2.8 0.6 ­4.2 ­1.0 1.4 ­3.1 0.5 Cattle 2.9 ­3.3 25.1 ­3.8 6.3 2.7 ­1.0 ­2.0 ­0.7 ­0.3 7.0 Animal products 2.3 ­3.1 17.6 ­3.9 5.0 4.4 ­2.0 ­2.7 ­1.5 ­2.0 6.1 Raw milk 0.8 ­3.3 16.4 ­3.1 5.3 2.7 ­2.5 ­1.5 ­1.8 ­0.4 ­1.4 Wool 6.6 ­0.2 11.9 ­3.7 5.9 2.5 ­3.2 ­3.0 ­3.6 0.1 7.1 Other primarya 0.6 ­2.6 2.2 ­0.2 0.5 1.7 0.2 ­0.7 1.1 0.5 3.0 Forestry 0.2 ­0.6 6.5 ­2.0 1.9 3.6 ­0.6 ­0.9 2.0 ­0.6 12.3 Fishing 1.7 ­3.0 3.4 ­5.7 3.4 2.5 ­1.2 ­4.3 1.4 ­1.8 4.9 Coal 1.1 2.3 ­0.2 ­8.0 0.6 2.1 ­0.5 ­0.9 ­2.5 0.7 8.7 Crude oil 0.5 ­15.0 1.1 0.1 ­0.2 1.0 0.2 ­0.4 ­1.5 0.7 2.2 Natural gas 0.5 ­5.4 ­6.6 ­3.5 0.7 1.1 ­0.8 44.2 2.8 0.4 35.1 Other minerals 0.8 ­3.2 2.2 ­1.9 0.2 3.2 ­0.4 0.3 1.0 0.4 3.4 Fooda 2.9 ­2.4 9.0 ­3.7 3.6 3.9 ­0.8 ­1.3 1.0 ­2.0 7.5 Bovine meat products 8.9 ­4.8 15.2 ­4.0 3.1 3.4 ­0.6 ­1.8 ­0.3 ­0.4 2.4 Other meat 3.6 ­6.9 14.6 ­3.6 4.3 4.0 ­1.1 ­2.4 ­0.4 ­2.0 8.1 Vegetable oils and fats 3.6 ­3.4 8.6 ­2.8 3.6 4.9 11.2 ­1.7 2.6 ­0.9 6.2 Milk 12.5 ­3.9 8.4 ­3.9 3.1 3.6 ­1.1 ­1.5 2.6 ­0.8 ­0.6 Processed rice 9.1 ­2.7 7.9 ­4.2 5.4 2.6 0.5 ­1.1 ­2.1 ­0.8 7.0 Sugar 5.1 ­4.0 7.1 ­3.8 3.0 2.7 ­1.0 ­1.5 0.3 ­13.1 7.2 Food products 0.4 ­2.4 5.9 ­3.8 3.4 3.5 ­0.4 ­1.3 0.5 ­2.1 7.7 Beverages and tobacco 0.1 ­2.7 5.1 ­3.6 2.8 2.9 ­1.4 ­2.4 0.5 ­1.7 5.5 Textilea 0.0 ­6.4 1.9 ­7.6 0.9 2.1 ­1.0 ­2.1 ­0.8 ­3.0 ­7.7 Textiles 0.2 ­3.5 1.7 ­7.2 1.0 1.9 ­1.0 ­1.6 ­0.3 ­2.8 ­4.8 Wearing apparel ­0.2 ­8.1 2.6 ­8.3 0.9 2.3 ­1.0 ­2.6 ­1.0 ­3.2 ­8.7 World Market Impacts of Multilateral Trade Reforms 73 Table 3.3. (Continued) a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Other manufacturesa 0.1 ­3.5 2.5 ­4.7 0.9 2.1 ­1.7 ­1.4 0.3 ­0.7 2.0 Leather products 0.4 ­3.8 3.7 ­4.7 1.6 3.1 ­2.0 ­7.4 ­0.3 ­2.6 ­0.2 Wood products 0.5 ­3.0 4.1 ­3.6 1.4 3.5 ­0.6 ­1.9 0.9 ­0.8 5.7 Paper products 0.3 ­4.3 3.1 ­4.3 1.6 2.7 ­1.1 ­2.2 ­0.9 ­1.3 1.9 Petroleum, coal 0.1 ­12.5 1.6 ­2.0 0.4 0.9 0.0 ­0.4 ­2.0 0.5 11.6 Chemical, plastic 0.0 ­4.8 2.4 ­5.0 0.4 1.2 ­1.5 ­1.6 ­0.6 ­1.1 3.7 Mineral products 0.4 ­1.7 3.4 ­4.7 1.4 2.7 ­0.7 ­1.8 0.0 ­1.0 6.9 Iron and steel 0.2 ­5.8 2.9 ­7.1 1.1 1.8 ­1.6 ­1.4 ­0.5 ­0.5 3.2 Metals 0.0 ­4.8 1.5 ­4.9 0.8 2.5 ­1.3 ­1.2 1.3 ­0.7 8.2 Metal products 0.2 ­4.1 3.0 ­6.2 1.1 0.7 ­1.8 ­3.0 0.6 ­0.9 2.9 Motor vehicles ­0.3 ­2.4 2.1 ­6.9 ­0.6 ­0.5 ­2.3 ­5.2 ­3.9 ­1.8 1.6 Transport equipment 0.0 ­2.5 2.8 ­5.9 0.7 0.3 ­1.9 ­3.9 ­0.3 ­1.5 ­9.7 Electronic equipment 0.4 ­2.1 ­0.3 ­4.5 0.3 2.1 ­1.8 ­4.0 0.5 ­1.8 2.6 Machinery 0.1 ­2.3 2.5 ­5.1 0.7 0.4 ­1.5 ­3.3 0.1 ­1.3 4.2 Manufactures 0.0 ­2.5 3.7 ­4.8 1.5 2.9 ­1.3 ­4.1 0.9 ­1.7 1.8 Electricity 0.1 ­3.7 3.7 ­4.0 1.2 2.5 ­0.4 ­3.3 1.0 0.3 6.4 Gas 0.1 ­1.7 2.1 ­3.0 1.5 3.1 ­0.4 ­1.9 0.3 ­0.3 20.8 Water 0.6 ­0.6 3.7 ­3.6 2.0 3.3 ­0.5 ­2.1 1.6 ­0.8 7.9 Construction 0.3 ­2.2 4.3 ­5.8 1.5 2.5 ­0.9 ­1.4 1.2 ­1.4 8.0 Servicesa 0.7 ­0.8 4.5 ­3.1 2.0 3.4 ­0.4 ­1.3 1.4 ­0.7 10.8 Trade 1.5 ­0.6 4.3 ­2.7 2.2 4.2 ­0.4 ­0.6 2.2 ­1.0 12.5 Transport 0.5 ­0.8 3.5 ­3.6 2.0 3.4 ­0.4 ­1.5 0.9 ­0.6 1.8 Water transport 0.4 ­0.8 3.5 ­3.7 1.4 2.2 ­0.4 ­1.3 0.9 ­0.8 ­3.7 Air transport 0.4 ­0.8 3.5 ­3.4 1.0 2.6 ­0.5 ­1.4 0.6 ­0.8 1.7 Communication 0.6 0.1 4.7 ­3.3 1.8 4.1 ­0.3 ­2.3 2.3 ­0.6 8.8 Financial services 0.6 ­0.1 4.6 ­2.4 2.2 4.0 ­0.3 ­0.3 2.5 ­0.6 13.6 Insurance 0.5 ­0.1 4.5 ­2.5 2.1 4.0 ­0.3 ­0.3 2.2 ­0.6 10.8 Business services 0.6 ­0.5 5.0 ­2.6 2.0 3.5 ­0.4 ­1.3 1.7 ­0.6 16.0 Recreational services 0.4 ­0.7 4.5 ­2.7 1.4 3.4 ­0.4 ­1.1 1.0 ­0.9 13.8 Government services 0.3 ­0.9 4.2 ­3.3 1.9 3.8 ­0.4 ­1.5 1.8 ­1.2 12.1 Dwelling 1.5 0.2 6.0 ­2.2 2.6 4.3 ­0.2 ­0.6 2.8 ­0.7 22.8 Source: Authors' simulations. a. Calculated as a weighted average of respective price changes, excluding intra-EU trade 74 Poverty and the WTO: Impacts of the Doha Development Agenda Table 3.4. Full Liberalization: Import Quantities for All Regions a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Import quantity index 7.2 31.7 22.3 9.3 16.6 8.5 7.7 3.7 4.0 9.0 23.2 Primary agriculturea 11.2 1.9 25.2 7.6 6.1 6.5 5.2 2.6 6.6 3.5 21.2 Paddy rice 222.6 ­18.6 49.7 8.0 43.8 43.5 8.7 ­42.9 53.2 34.3 51.0 Wheat 9.4 ­4.0 4.8 ­3.9 ­12.6 ­3.0 17.5 ­7.0 2.6 ­2.9 ­6.0 Cereal grains 5.5 ­4.8 19.4 1.8 38.2 ­6.1 17.6 ­8.7 17.4 ­9.7 5.0 Fruits and vegetables 12.4 22.6 19.4 41.3 30.7 9.4 7.9 26.5 7.7 6.8 58.0 Oilseeds 6.3 ­30.9 15.6 9.7 ­2.7 ­1.1 ­1.9 ­7.5 3.1 ­1.1 26.7 Raw sugar 50.4 ­18.6 44.7 11.3 2.7 ­6.4 ­4.0 40.4 ­7.2 ­4.4 23.0 Plant fibers 5.3 ­6.2 17.3 ­9.9 ­12.6 12.7 ­19.7 ­30.0 10.1 0.1 8.6 Other crops 9.3 83.5 84.8 17.9 11.8 15.3 19.6 8.1 21.1 4.1 17.5 Cattle 1.6 0.9 194.9 30.0 13.4 ­0.9 ­4.6 ­3.8 ­6.1 4.7 ­0.8 Animal products 3.7 18.0 80.9 11.2 14.2 17.7 ­6.0 ­9.8 10.3 ­1.7 21.7 Raw milk ­2.0 ­9.9 90.5 ­9.5 22.9 12.6 ­10.5 ­4.0 ­2.5 3.4 ­4.3 Wool 1.9 6.2 131.6 15.5 8.7 45.3 ­9.7 ­2.0 3.4 2.3 ­6.2 Other primarya 0.9 ­10.3 ­1.3 2.4 ­3.1 3.2 3.1 2.8 0.7 3.9 12.5 Forestry 2.6 0.5 25.6 14.1 3.2 6.4 5.4 ­0.2 ­1.5 11.0 20.1 Fishing 3.6 22.8 12.5 4.7 14.8 5.9 16.7 ­3.1 5.1 6.8 14.7 Coal 0.9 25.4 ­7.4 ­2.3 9.3 22.5 ­0.3 ­4.6 0.6 ­3.1 44.4 Crude oil 1.0 ­19.2 0.4 0.6 ­5.8 3.6 2.2 1.9 1.2 8.3 29.1 Natural gas 0.2 ­52.1 ­54.9 128.2 77.4 11.7 ­8.4 34499.7 57.0 1.2 17424.7 Other minerals 0.2 ­1.2 ­6.7 13.0 ­3.6 0.0 4.8 6.5 ­1.7 2.5 8.0 Fooda 35.9 27.4 37.2 25.3 38.9 21.2 28.3 15.3 24.3 18.1 60.6 Bovine meat products 87.8 31.3 76.9 14.1 18.1 20.8 ­0.1 32.5 6.0 ­5.2 18.7 Other meat 68.9 94.0 166.0 59.0 86.6 56.2 103.9 46.5 179.5 51.8 139.0 Vegetable oils and fats 61.8 21.0 49.6 39.5 50.2 30.3 39.8 12.4 85.7 23.8 44.7 Milk 41.7 53.4 26.6 ­4.6 17.3 1.0 44.1 7.5 1.0 ­2.1 14.0 Processed rice 123.0 ­6.7 17.0 27.1 ­17.6 42.4 5.3 ­0.4 115.1 6.8 51.1 Sugar 86.5 44.5 57.4 23.1 23.2 45.7 33.9 1.0 99.0 9.1 61.8 Food products 13.7 31.2 32.2 21.5 34.0 18.5 8.1 24.8 10.3 14.9 62.1 Beverages and tobacco 16.4 35.4 26.6 37.8 20.3 22.8 19.5 13.5 4.1 12.2 87.4 Textilea 23.0 113.7 64.2 26.5 40.9 34.7 21.6 19.8 34.5 6.7 99.7 Textiles 23.1 114.5 53.7 21.2 38.6 33.9 9.3 14.4 35.7 7.5 101.5 Wearing apparel 22.9 94.9 114.1 49.8 60.1 47.5 51.8 34.5 23.0 6.1 78.4 World Market Impacts of Multilateral Trade Reforms 75 Table 3.4. (Continued) a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Other manufacturesa 6.3 17.0 27.6 10.0 16.7 7.8 7.2 2.7 1.4 12.6 19.0 Leather products 17.2 63.0 54.9 66.5 56.6 31.5 57.4 1.0 7.3 32.0 50.6 Wood products 4.2 78.5 61.7 71.6 8.2 20.5 15.4 27.0 7.3 27.4 53.4 Paper products 4.4 29.1 31.9 10.0 9.9 5.9 1.9 0.4 3.4 10.8 27.2 Petroleum, coal 4.9 8.6 1.1 20.7 13.5 4.3 3.7 3.6 2.0 10.0 4.1 Chemical, plastic 8.9 17.4 18.9 12.7 23.2 10.6 4.1 5.6 7.4 11.4 24.9 Mineral products 11.4 35.0 44.5 18.3 38.2 10.1 18.3 3.1 5.2 22.0 74.6 Iron and steel 5.9 ­2.5 32.6 ­1.0 10.9 6.1 15.5 2.5 0.8 6.4 ­2.1 Metals 5.9 13.5 18.1 24.4 12.8 ­2.6 10.3 18.8 7.6 18.0 27.2 Metal products 10.9 69.7 80.4 31.4 35.4 19.5 10.1 ­0.3 13.1 25.0 6.1 Motor vehicles 9.1 3.5 50.3 0.7 28.1 22.6 6.4 ­0.3 10.4 29.0 8.6 Transport equipment 5.5 30.1 11.6 2.3 16.6 4.3 15.7 ­0.7 6.3 29.5 20.8 Electronic equipment 1.2 3.2 21.0 2.3 3.5 3.8 6.1 ­0.5 ­2.2 2.8 18.0 Machinery 6.0 3.2 34.0 1.2 20.8 3.1 5.0 ­0.4 2.5 3.8 10.8 Manufactures 10.1 105.1 93.9 62.9 60.9 26.7 29.4 16.4 14.9 40.7 74.2 Electricity 1.2 ­4.9 3.4 0.4 ­2.1 6.3 ­0.8 ­0.3 3.3 8.6 26.2 Gas 0.8 ­4.4 1.7 ­11.1 2.6 8.2 ­2.0 9.3 0.6 ­0.2 76.0 Water 0.0 8.6 5.6 ­11.9 3.7 1.9 ­1.9 ­7.2 3.3 ­2.4 14.6 Construction 0.7 ­4.2 17.6 ­11.6 2.5 7.9 ­0.6 ­4.1 3.8 ­3.6 22.2 Servicesa 0.1 ­0.1 5.7 ­5.6 0.6 3.6 ­0.9 ­2.5 1.7 ­1.0 12.6 Trade ­1.2 ­4.8 6.9 ­5.8 ­0.2 5.5 ­1.5 ­2.2 2.1 ­2.5 23.1 Transport 0.7 ­2.2 5.5 ­7.8 2.7 4.6 ­0.6 ­1.0 2.4 ­0.9 6.4 Water transport 2.5 ­2.3 2.0 ­5.6 2.1 1.6 ­1.1 ­3.0 0.7 ­0.5 1.0 Air transport 0.2 0.4 3.1 ­4.9 0.7 3.4 ­0.5 ­2.9 1.3 ­0.9 12.1 Communication 0.3 5.2 0.7 ­7.3 1.8 5.0 ­1.3 ­5.5 2.1 ­1.5 7.0 Financial services ­0.1 15.8 6.2 ­6.2 2.5 6.5 ­1.0 ­1.9 1.5 ­0.8 8.1 Insurance 0.2 6.6 6.8 ­5.4 2.1 6.2 ­0.4 ­1.5 2.5 0.4 3.2 Business services ­0.1 0.1 8.1 ­4.1 1.8 2.1 ­0.6 ­3.2 1.3 ­0.3 12.6 Recreational services 0.8 ­2.4 4.7 ­5.3 1.6 6.1 ­1.4 ­2.0 1.1 ­1.3 16.5 Government services 0.3 ­3.9 3.3 ­8.5 3.0 6.3 ­1.7 ­2.7 2.4 ­1.7 20.2 Dwelling 0.2 ­0.5 1.5 ­0.8 ­0.2 0.6 ­0.2 ­1.2 ­0.4 0.6 0.7 Source: Authors' simulations. a. Calculated as a weighted average of respective price changes, excluding intra-EU trade 76 Poverty and the WTO: Impacts of the Doha Development Agenda Table 3.5. Full Liberalization: Export Quantities for All Regions a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Export quantity index 7.2 49.5 6.5 17.2 13.4 4.1 8.1 10.4 4.4 6.8 31.2 Primary agriculturea 11.2 53.0 ­13.3 30.0 76.4 ­2.3 9.4 47.4 7.0 28.0 ­2.2 Paddy rice 222.6 580.7 428.6 66.4 7,053.0 246.2 129.1 99.5 1,657.8 30.1 113.4 Wheat 9.4 4,409.3 ­59.6 106.7 61.6 ­3.9 ­87.3 45.0 457.1 64.0 372.3 Cereal grains 5.5 151.4 18.2 9.2 81.4 17.9 ­4.7 5.2 36.1 28.9 ­31.8 Fruits and vegetables 12.4 5.7 ­30.2 28.0 52.5 18.0 9.0 33.9 ­6.9 24.7 11.4 Oilseeds 6.3 84.3 6.5 202.3 88.6 7.7 ­74.7 78.2 48.3 81.0 51.7 Raw sugar 50.4 ­4.3 ­17.3 ­2.0 370.4 755.6 ­24.6 ­40.8 ­33.9 ­26.6 16.9 Plant fibers 5.3 83.5 17.5 76.6 81.9 8.0 87.0 46.5 53.7 ­7.9 17.0 Other crops 9.3 24.0 ­43.0 8.8 ­3.8 ­4.9 14.1 49.1 48.8 ­27.7 ­10.5 Cattle 1.6 2.3 ­52.1 25.3 ­6.7 7.1 22.1 ­2.6 40.1 13.7 ­31.5 Animal products 3.7 8.2 ­28.2 9.9 ­4.9 9.3 18.2 5.7 9.8 4.4 ­0.5 Raw milk ­2.0 14.6 ­69.7 19.3 ­35.9 ­22.4 13.2 5.1 7.8 ­3.3 4.9 Wool 1.9 143.2 ­38.0 149.8 13.7 20.5 98.6 37.4 116.2 11.3 20.3 Other primarya 0.9 1.9 ­2.2 3.8 5.4 ­1.9 1.9 5.6 2.1 ­0.1 ­8.1 Forestry 2.6 8.4 ­17.9 12.0 0.8 3.9 9.3 6.3 2.1 3.7 ­25.1 Fishing 3.6 0.6 ­1.2 10.1 8.8 4.0 6.3 17.4 2.5 11.1 ­4.2 Coal 0.9 ­14.3 3.1 29.4 6.0 ­1.6 3.4 7.7 17.4 0.0 ­27.0 Crude oil 1.0 447.5 7.3 3.0 6.0 ­1.3 1.8 ­0.1 22.7 ­0.9 ­6.2 Natural gas 0.2 704.5 1,066.3 226.7 5.5 ­2.7 63.2 ­100.0 ­48.2 0.3 ­100.0 Other minerals 0.2 7.0 ­2.1 4.1 1.8 ­3.8 4.2 2.4 1.9 1.5 ­4.4 Fooda 35.9 16.2 107.3 13.1 36.3 23.2 21.8 0.6 21.9 15.8 ­2.5 Bovine meat products 87.8 240.3 828.8 64.8 7.9 118.3 89.0 671.9 186.6 587.1 60.9 Other meat 68.9 1,082.8 ­2.4 227.0 ­0.3 97.2 129.5 ­9.5 169.5 70.6 ­29.9 Vegetable oils and fats 61.8 55.8 ­39.9 ­14.1 ­12.7 63.8 ­34.3 18.0 ­5.2 17.1 104.4 Milk 41.7 232.8 49.0 273.4 114.5 60.7 214.6 135.8 82.3 174.4 278.5 Processed rice 123.0 255.3 19.9 63.8 501.6 126.6 73.9 15.9 ­25.1 41.9 64.1 Sugar 86.5 ­30.3 36.5 2.6 95.1 17.4 ­7.9 78.8 177.9 1,15.3 211.9 Food products 13.7 0.1 16.4 7.0 0.2 ­11.2 8.5 ­7.4 16.7 5.3 ­24.8 Beverages and tobacco 16.4 60.3 ­6.8 7.5 18.4 26.2 11.3 ­0.8 63.5 5.1 21.0 Textilea 23.0 67.7 ­9.0 77.5 24.5 22.2 ­26.6 31.4 49.5 44.3 148.7 Textiles 23.1 32.8 ­6.4 73.7 25.0 18.8 ­19.4 16.4 48.3 29.6 108.7 Wearing apparel 22.9 89.0 ­18.9 85.0 24.2 27.6 ­33.0 47.1 50.0 61.7 161.3 World Market Impacts of Multilateral Trade Reforms 77 Table 3.5. (Continued) a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Other manufacturesa 6.3 25.7 ­7.9 38.7 7.9 1.3 10.7 9.3 0.0 10.2 39.6 Leather products 17.2 16.6 ­25.3 21.4 21.5 21.0 ­30.1 66.8 21.3 3.5 51.9 Wood products 4.2 53.8 ­14.8 25.6 0.2 ­1.6 4.5 7.0 0.2 10.8 ­22.4 Paper products 4.4 28.8 ­14.2 23.6 ­0.8 ­5.9 10.3 47.5 17.8 9.1 15.1 Petroleum, coal 4.9 69.6 4.7 13.2 6.1 2.1 ­0.1 17.2 13.5 7.7 ­36.0 Chemical, plastic 8.9 36.1 ­12.2 49.9 11.8 12.9 13.7 136.4 28.8 8.2 183.9 Mineral products 11.4 26.6 ­4.0 ­0.7 12.9 9.5 1.5 4.4 6.7 15.4 ­19.7 Iron and steel 5.9 64.3 ­11.1 18.7 3.0 ­2.6 21.0 0.3 8.8 4.9 ­4.5 Metals 5.9 77.9 1.6 46.7 6.3 ­16.5 8.7 6.8 ­6.0 12.4 ­50.6 Metal products 10.9 47.5 ­8.1 40.2 12.0 5.6 8.3 ­1.3 0.0 11.5 ­4.8 Motor vehicles 9.1 19.9 16.3 40.7 ­6.4 0.2 9.2 35.6 30.7 26.0 ­5.2 Transport equipment 5.5 14.5 ­21.4 58.3 27.9 30.5 14.8 32.0 20.2 23.2 301.4 Electronic equipment 1.2 23.8 9.8 59.3 5.9 ­5.6 16.7 46.6 ­3.4 37.9 ­11.8 Machinery 6.0 20.9 ­13.9 44.0 6.2 6.6 7.9 11.3 6.3 13.0 11.4 Manufactures 10.1 40.4 ­22.4 41.1 1.1 ­14.0 4.9 40.5 ­6.4 0.0 ­1.1 Electricity 1.2 23.6 ­18.2 18.0 4.0 ­11.8 2.5 22.9 ­4.6 ­4.3 ­28.3 Gas 0.8 11.8 ­9.7 20.1 ­6.7 ­14.6 2.2 12.8 1.0 2.8 ­64.5 Water 0.0 7.4 ­15.7 26.4 ­8.2 ­14.4 5.8 16.0 ­6.1 7.5 ­32.8 Construction 0.7 11.6 ­13.0 27.0 ­4.3 ­7.9 5.0 7.0 ­3.1 7.0 ­24.6 Servicesa 0.1 4.6 ­13.5 15.1 ­4.8 ­10.1 3.8 7.7 ­3.0 5.1 ­29.2 Trade ­1.2 6.3 ­11.5 15.3 ­4.7 ­11.4 5.0 5.5 ­4.8 7.6 ­33.8 Transport 0.7 5.9 ­10.1 17.5 ­5.3 ­10.0 4.0 7.6 ­1.0 4.6 ­4.6 Water transport 2.5 7.1 ­8.4 19.8 ­1.9 ­4.4 5.2 9.3 0.4 7.4 20.1 Air transport 0.2 4.8 ­10.8 15.6 ­2.4 ­8.0 3.6 7.0 ­0.6 4.5 ­4.9 Communication 0.3 2.0 ­14.1 15.9 ­4.7 ­12.5 3.3 11.6 ­6.3 4.5 ­26.0 Financial services ­0.1 2.8 ­14.0 12.1 ­6.0 ­12.4 2.9 3.5 ­7.1 4.3 ­37.4 Insurance 0.2 3.0 ­13.8 12.1 ­5.8 ­12.1 3.2 3.6 ­6.3 4.3 ­31.1 Business services ­0.1 4.1 ­15.2 12.9 ­5.1 ­10.2 4.1 7.2 ­4.2 4.5 ­41.9 Recreational services 0.8 5.0 ­13.7 13.1 ­3.3 ­10.2 3.6 6.2 ­1.8 5.3 ­37.7 Government services 0.3 4.3 ­13.9 14.6 ­5.7 ­12.2 2.8 7.1 ­5.3 5.8 ­34.6 Dwelling 0.2 ­0.5 1.5 ­0.8 ­0.2 0.6 ­0.2 ­1.2 ­0.4 0.6 0.7 Source: Authors' simulations. a. Calculated as a weighted average of respective price changes, excluding intra-EU trade 78 Poverty and the WTO: Impacts of the Doha Development Agenda small increase. In the case of services trade, as noted in chapter 2, there are no solid measures of trade barriers, and the DDA does not appear to promise much liberalization, so any liberalization analysis in these markets is ignored. The country-specific import volume increases range from 3.7 percent, in the case of Mozambique, to 31.7 percent, in the case of Bangladesh. In the case of pri- mary agriculture, Brazil and Vietnam experience import volume changes above the world average, and Vietnam also shows a very strong increase in imports of nonagricultural primary products. In the case of processed food imports, all countries show a strong increase in import volume, with the largest increases aris- ing in nonruminant meat products, beverages, and tobacco. The rise in textiles and apparel imports is even higher for a number of the focus countries. Here, Bangladesh stands out, with more than a doubling of imports as import tariffs fall and apparel exports expand. Other manufactures' import volume changes are quite heterogeneous. In the case of Mozambique, there are declines in import vol- ume for many of these sectors. The change in volume of services imports is small, in keeping with the absence of any liberalization in these markets. Table 3.5 reports the change in export volumes by country and commodity. Whereas national imports, which reflect a composite of exports from many differ- ent sources, showed relatively uniform changes, national export volumes are much more heterogeneous--often showing a mix of positive and negative signs. For example, in the case of Brazil, where processed food exports increase very strongly, primary agriculture exports decline, as do exports of most nonfood manufactures. This is due to the finite primary factor endowments in the country. It is not possible to increase production (and hence exports) of all products simul- taneously. As manufactured food production increases, more agricultural prod- ucts are required as inputs, thereby reducing the amount available for export. Textiles and apparel exports increase sharply for Bangladesh, Cameroon, and Vietnam. Of course, as shown in table 3.4, imports also increase in these coun- tries. To see the impact on the countries' net trade position, turn to table 3.6, which reports the change in the value of exports less imports (net trade), by com- modity, in hundreds of millions of U.S. dollars. Here, the trade balance in textiles and apparel improves markedly for Bangladesh, China, Indonesia, the Philip- pines, and Vietnam. Because the macroeconomic closure in the analysis fixes the ratio of the aggregate trade balance to national income, a strong increase in net trade in one sector forces some other sectors to experience a deteriorating trade balance. In the economies with a strongly expanding net trade position in textiles and apparel, declines in the net trade balance for nonmanufactures are seen. In the case of Brazil, it is the US$11 billion increase in net exports of food products that drives the net trade story, with compensating reductions in primary agriculture, other manufactures, and services. World Market Impacts of Multilateral Trade Reforms 79 Table 3.6. Full Liberalization: Trade Balance (US$ hundreds of millions) a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Trade balance 0.6 ­81.3 10.5 48.1 1.9 ­15.0 0.2 ­2.5 ­0.8 ­13.1 Primary agriculturea 0.0 0.4 ­7.8 15.0 43.5 ­3.5 ­6.8 0.3 ­0.8 1.2 ­0.7 Paddy rice ­2.2 0.0 ­0.4 0.0 33.2 0.0 ­0.3 0.0 0.1 0.0 0.1 Wheat ­0.6 0.1 ­1.4 ­0.2 0.4 ­0.2 ­1.9 0.0 ­0.5 1.1 0.0 Cereal grains ­0.3 0.0 2.3 0.0 4.2 0.0 ­4.1 0.0 ­0.1 0.3 0.0 Fruits and vegetables ­2.7 ­0.3 ­1.9 1.6 10.1 0.1 1.2 0.0 ­0.6 ­0.6 0.0 Oilseeds ­0.8 0.2 6.5 5.1 0.0 ­0.3 ­2.1 0.0 ­0.1 0.8 0.2 Raw sugar 0.0 0.0 0.0 0.0 0.2 0.0 0.0 0.0 0.0 0.0 0.0 Plant fibers 0.2 0.7 0.4 7.2 0.8 ­1.8 ­0.5 0.1 0.0 0.0 ­0.1 Other crops ­2.5 ­0.3 ­11.9 1.1 ­1.7 ­1.3 ­0.4 0.1 0.5 ­0.4 ­0.9 Cattle 0.2 0.0 ­0.2 0.0 0.0 0.0 0.9 0.0 0.0 0.0 0.0 Animal products ­0.3 0.0 ­1.0 0.0 ­2.6 0.1 0.2 0.0 0.0 0.1 0.0 Raw milk 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Wool 0.0 0.0 ­0.1 0.2 ­1.2 0.0 0.0 0.0 0.0 0.0 0.0 Other primarya 0.0 0.2 0.4 6.6 6.1 ­0.9 2.0 0.0 ­0.3 0.6 ­0.8 Forestry ­0.3 0.0 ­0.1 1.2 ­0.6 0.2 0.0 0.0 0.0 0.5 0.0 Fishing ­0.6 0.0 0.0 0.0 0.4 0.2 0.0 0.0 0.0 0.0 0.0 Coal ­0.5 0.0 0.5 0.1 0.8 0.0 0.0 0.0 0.0 0.1 ­0.2 Crude oil ­2.0 0.2 ­0.4 5.3 3.8 ­0.5 2.1 0.0 ­0.4 ­1.1 ­0.5 Natural gas ­0.1 0.0 0.0 0.0 0.1 ­0.6 0.1 0.0 0.0 1.2 0.0 Other minerals ­0.5 0.0 0.4 0.0 1.6 ­0.2 ­0.1 0.0 0.1 ­0.1 0.0 Fooda 0.0 ­1.9 110.8 ­14.0 12.7 8.0 ­8.6 ­0.3 ­2.8 ­11.3 ­6.6 Bovine meat products ­8.8 0.0 100.5 0.0 ­0.8 0.2 0.3 0.0 0.0 1.2 0.0 Other meat ­7.0 0.3 1.3 ­1.0 ­7.8 3.0 ­4.1 ­0.1 ­0.7 ­8.6 ­0.3 Vegetable oils and fats ­6.2 ­0.9 ­2.8 ­1.7 ­2.6 9.4 ­1.0 0.0 ­0.2 ­0.6 ­0.2 Milk ­3.1 ­0.5 ­0.5 0.3 ­0.7 0.4 ­4.1 0.0 ­0.2 1.5 ­0.5 Processed rice ­9.3 0.1 ­0.1 ­4.8 31.0 ­0.7 0.0 0.0 ­2.4 ­0.1 2.7 Sugar ­5.9 ­0.3 6.5 ­0.6 ­0.6 ­0.8 ­0.3 0.0 ­0.1 ­1.6 0.1 Food products ­10.6 ­0.5 6.4 ­3.6 ­6.7 ­3.7 ­0.5 ­0.2 0.7 ­2.3 ­4.8 Beverages and tobacco ­2.8 0.0 ­0.5 ­2.7 0.9 0.2 1.1 0.0 0.1 ­0.8 ­3.5 Textilea 0.0 7.4 ­9.1 ­4.0 145.8 15.0 ­35.9 ­0.1 7.6 0.7 10.3 Textiles ­23.2 ­13.4 ­6.2 ­2.7 ­9.9 4.5 ­12.0 0.0 ­1.3 ­0.2 ­8.9 Wearing apparel ­17.0 20.8 ­2.9 ­1.2 155.7 10.5 ­23.9 0.0 8.9 0.9 19.2 80 Poverty and the WTO: Impacts of the Doha Development Agenda Table 3.6. (Continued) a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Other manufacturesa 0.0 ­5.8 ­159.6 ­6.0 ­149.6 ­9.8 29.3 0.1 ­5.6 0.8 0.6 Leather products ­8.7 0.4 ­7.7 ­2.3 57.2 6.5 ­7.3 0.0 0.9 ­4.0 12.6 Wood products ­5.2 ­0.1 ­3.9 ­0.3 ­1.4 0.8 ­1.4 ­0.1 ­0.1 ­2.0 ­1.6 Paper products ­4.2 ­0.7 ­6.2 ­0.9 ­8.0 ­1.8 0.2 0.0 0.0 ­0.3 ­1.0 Petroleum, coal ­3.0 ­0.2 0.1 ­2.7 ­5.4 ­0.7 ­0.8 0.0 0.2 6.2 ­0.6 Chemical, plastic ­24.4 ­0.8 ­25.9 ­2.5 ­70.1 0.6 1.1 0.0 ­0.8 ­2.7 2.4 Mineral products ­12.2 ­0.8 ­5.4 ­2.9 ­3.5 0.7 ­4.0 0.0 ­0.2 ­2.2 ­3.0 Iron and steel ­5.9 0.1 ­5.2 0.4 ­11.9 ­1.1 ­1.6 0.0 ­0.1 2.5 0.1 Metals ­2.4 ­0.1 ­1.6 5.7 ­8.5 ­3.1 ­1.4 0.2 ­0.7 11.1 ­0.9 Metal products ­7.3 ­1.0 ­7.4 ­3.6 2.6 ­1.1 ­2.6 0.0 ­0.9 ­1.6 ­0.2 Motor vehicles ­9.8 ­0.1 ­11.6 0.4 ­26.2 ­3.8 6.6 0.0 0.0 ­6.6 ­0.4 Transport equipment ­3.4 ­1.1 ­10.9 3.9 4.1 0.2 ­1.2 0.0 0.1 4.2 0.6 Electronic equipment ­3.3 ­0.1 ­12.6 0.6 21.4 ­4.3 36.1 0.0 ­3.7 0.1 ­2.3 Machinery ­16.3 ­0.2 ­53.2 0.9 ­89.8 0.3 10.7 0.0 0.6 0.6 ­2.0 Manufactures ­4.7 ­1.1 ­7.5 ­3.3 ­9.4 ­2.7 ­5.2 0.0 ­0.8 ­4.1 ­1.6 Electricity 0.0 0.0 ­0.7 0.1 0.1 0.0 0.0 0.1 0.0 ­0.6 0.0 Gas 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Water 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Construction 0.0 0.0 ­0.1 0.4 ­0.6 ­0.1 0.2 0.0 ­0.1 0.2 ­1.3 Servicesa 0.0 0.3 ­16.0 12.9 ­10.4 ­6.9 5.1 0.2 ­0.7 7.2 ­15.9 Trade 0.0 0.0 ­1.5 1.3 ­7.0 ­1.9 0.5 0.0 ­0.4 0.5 ­3.0 Transport 45.2 0.0 ­0.9 2.1 ­2.3 ­0.7 1.3 0.0 0.0 1.5 ­0.5 Water transport 142.7 0.1 ­0.1 1.0 3.2 0.2 0.5 0.0 0.3 1.9 0.3 Air transport 29.7 0.0 ­0.6 1.8 0.1 ­0.4 0.7 0.0 0.0 1.4 ­1.1 Communication 0.0 0.0 ­0.3 0.5 ­0.3 ­0.1 0.3 0.0 ­0.2 0.2 ­0.6 Financial services 0.0 ­0.1 ­0.8 0.3 ­0.4 ­0.3 0.2 0.0 0.0 0.1 ­1.1 Insurance 0.0 0.0 ­0.7 0.3 ­0.5 ­0.4 0.5 0.0 ­0.1 0.0 ­0.5 Business services 0.0 0.0 ­9.2 2.9 ­1.3 ­2.1 0.4 0.1 ­0.2 0.6 ­5.1 Recreational services 0.0 0.0 ­0.6 0.5 ­0.5 ­0.7 0.4 0.0 ­0.1 0.4 ­1.7 Government services 0.0 0.3 ­1.3 2.2 ­1.5 ­0.6 0.3 0.0 ­0.1 0.7 ­2.7 Dwelling 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Source: Authors' simulations. a. Calculated as a weighted average of respective price changes, excluding intra-EU trade World Market Impacts of Multilateral Trade Reforms 81 A very useful feature of table 3.6 is that it highlights the cases where the extraordinarily large percentage changes in trade volumes are irrelevant as a result of the extremely small size of the initial flow. Brazilian natural gas imports pro- vide a case in point. There is more than a 1,000 percent increase in exports, but the change in net trade is negligible, according to table 3.6. Doha Scenario Tables 3.7­3.10 report the price and trade volume results from the core Doha sce- nario developed in chapter 2 of this book. It represents an aggressive implementa- tion of the framework agreed upon by WTO members on August 1, 2004. The first thing to note is that the impacts on prices and trade volumes are much smaller. Compared with the full trade liberalization, world average agricultural prices rise by only one-third as much under Doha, while world trade volumes for primary agriculture are virtually unchanged. In the latter case, the trade-diminishing effect of export subsidy elimination offsets the trade-enhancing impact of tariff reduc- tions. In the case of food products, the rise in world average prices is more than two thirds of the full-liberalization case, but the rise in import volume is much less--again as a result of the elimination of export subsidies in the United States and the EU. Because of the nonlinearity in the tariff reduction formulae, as well as the great differences in binding overhang across commodities and regions, the composition of the price differences is quite different between full liberalization and Doha. Under the Doha scenario, Vietnam has a TOT deterioration instead of an improvement: as a nonmember of the WTO, Vietnam does not enjoy the benefits of tariff cuts in other countries. Instead of a dramatic decline in its TOT, Bangladesh now shows a much smaller change. Recall that much of the sharp decline under full liberalization was a result of Bangladesh's own-liberalization and the subsequent increase in export volumes in the face of modest increases or even decreases in export demand. As an LDC, Bangladesh does not reduce its tar- iffs at all under the Doha scenario (recall table 3.1), so this source of TOT deterio- ration is not present. In contrast, Brazil still experiences a TOT improvement amounting to about half of the full-liberalization case. Communicating Global Results to the National Models As noted at the outset, the goal of these model simulations is to provide country case study authors with a picture of how their external environment is likely to change as a result of multilateral trade reform under the DDA. However, the indi- 82 Poverty and the WTO: Impacts of the Doha Development Agenda Table 3.7. Doha: Import Prices for All Regions a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Import price index 0.1 0.5 0.1 0.4 0.3 0.4 0.2 0.5 0.3 0.2 0.2 Primary agriculturea 1.1 2.4 1.7 1.6 2.9 3.4 4.3 1.3 1.8 0.4 2.2 Paddy rice 7.4 0.1 2.8 0.1 1.3 0.8 17.7 3 3.6 1.4 1.4 Wheat 1.6 1.3 2 2.8 1.3 1.9 1.6 1.7 1.4 0.2 1.4 Cereal grains 3.4 1.3 2.4 4.1 5.8 1.6 2.3 1.8 1.9 8.9 1.2 Fruits and vegetables ­0.1 0.4 1.6 ­0.2 0.8 1 1.1 1.1 0.9 0.4 1 Oilseeds 4.1 2.7 2.5 1.7 4.8 4.7 5.2 2.3 4 1.5 2.8 Raw sugar 0.7 1.5 1 0.2 3.8 1.5 1.6 1.5 1.5 1.5 1.5 Plant fibers 5.7 4.4 2.4 1.1 5 6.7 18.2 1.3 6.9 0.1 10 Other crops ­0.2 0 ­0.3 ­0.4 0.9 0.1 0.6 0.9 0.4 ­0.3 ­0.1 Cattle ­0.1 0 1 ­0.2 1.8 2.3 1.2 1.9 2.3 ­1.1 1.1 Animal products 0 0.3 0.4 ­0.1 0.7 ­0.8 1 1.4 0 0 ­1.6 Raw milk 0.3 0.9 0.6 0.2 0.4 0.2 0.4 0.2 0.2 0.1 0 Wool 1.9 1.4 0.4 0.8 2.3 0.5 1.8 0.4 1 0 1.5 Other primarya 0.1 0.1 0.1 0.1 0.2 0.1 0.2 1 0.1 0.3 0.3 Forestry 0 0.3 0 ­0.2 0 0.1 0 0.1 0.6 ­0.2 0.3 Fishing 0.1 0.1 0.2 0.1 0.2 0.1 0.1 0.4 0.5 ­0.1 0.1 Coal 0.3 ­0.6 0.2 0.2 0.3 0 0.2 0 0.2 1.8 0.2 Crude oil 0.1 0 0.1 0.1 0.1 0 0 0.1 0.1 0 0.2 Natural gas 0 0 ­0.2 0 0 0 0 0.3 ­0.1 0 ­0.1 Other minerals 0.4 0 0.1 0.2 0.4 0.3 0.2 1.7 0.4 ­0.1 0.3 Fooda 0.7 1.7 1.3 3.2 0.8 1.7 1 1.1 1.5 3 1.3 Bovine meat products 2.2 0.9 2.3 7.3 2.3 2.7 0.8 2.2 1.3 14.5 5.5 Other meat 0.8 1 1.8 3 1.2 0.2 0.5 1.1 0 2.1 1.1 Vegetable oils and fats 0.6 2 0.9 0.6 0.5 0.3 1.5 1.3 ­0.6 0.3 0.5 Milk 3.9 3.9 4.5 17.5 5.5 5.7 3.9 1.4 4.8 6.5 4.4 Processed rice 1.6 1.1 1.3 1.7 2.6 1.5 1.1 3.2 0.1 0.7 1.6 Sugar 1.9 1.7 5.2 4.3 1.5 2.2 0.7 1.4 4.8 2.2 2 Food products 0 0.3 0.7 1.2 ­0.3 0.2 0.3 0.6 0.5 0.5 0.2 Beverages and tobacco ­0.1 0.4 0.3 0.1 0.2 0.3 0.2 0.3 0.5 ­0.1 0.6 Textilea 0.1 0.4 0.7 0.2 0.3 0.4 0 0.7 0.4 ­0.1 0.4 Textiles 0.1 0.4 0.6 0.2 0.3 0.4 0.1 1 0.4 ­0.2 0.4 Wearing apparel 0.1 0.1 0.8 0.1 0.4 0.2 ­0.2 0 0.8 ­0.1 0.2 World Market Impacts of Multilateral Trade Reforms 83 Table 3.7. (Continued) a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Other manufacturesa 0 0.1 0 0 0.1 0.2 0 0.4 0.1 ­0.2 0.2 Leather products 0 ­0.1 0.3 0.1 ­0.2 ­0.2 ­0.2 0.2 ­0.1 ­0.1 ­0.2 Wood products 0 0.3 0.1 ­0.1 0.2 0.1 0 0 0.3 ­0.2 0.2 Paper products 0 0.3 0 0.3 0.3 0.4 0 2.6 0.3 ­0.2 0.4 Petroleum, coal 0 0.1 0.1 0 0.1 0.1 0 0 0.1 0 0.1 Chemical, plastic ­0.1 0.1 0 0 0.1 0.1 ­0.1 0.9 0.1 ­0.2 0.2 Mineral products 0 0.2 0.1 0.1 0.3 0.3 0 3.6 0.3 ­0.1 0.4 Iron and steel 0 0.2 0 ­0.1 0.2 0.2 0.1 0.1 0.2 ­0.2 0.2 Metals 0 0.1 0 ­0.1 0.2 0.4 0 0.1 0.3 ­0.2 0.4 Metal products 0 0.1 0 ­0.1 0.2 0.3 0 0 0.2 ­0.2 0.3 Motor vehicles ­0.1 0.1 0.3 ­0.1 0 0.2 ­0.1 0 0.2 ­0.2 0.1 Transport equipment ­0.1 0.1 ­0.2 0 ­0.2 0.2 ­0.2 ­0.2 0.1 ­0.2 0.2 Electronic equipment 0.1 0.1 0 ­0.1 0.2 0.2 0 ­0.1 0.1 ­0.1 0.2 Machinery 0 0 0 ­0.1 0.1 0.2 ­0.1 0 0.2 ­0.2 0.2 Manufactures 0 0.3 0 ­0.1 0 0.2 0 0 0.1 ­0.1 0.2 Electricity ­0.1 ­0.1 0 0 0.3 ­0.1 ­0.1 0 0 ­0.2 ­0.1 Gas ­0.2 ­0.1 ­0.2 ­0.3 ­0.2 ­0.2 ­0.2 ­0.2 ­0.2 ­0.3 ­0.3 Water 0 0 0 0 0 0 0 0 0 0 0 Construction ­0.1 0 0.1 ­0.1 ­0.1 0 0 ­0.1 ­0.1 ­0.1 ­0.1 Servicesa 0 0.1 0 0 0.3 0 0 0 0.1 0 0 Trade 0.2 0.4 0.1 0.1 0.4 0.1 0.1 0.1 0.3 0 0.1 Transport 0 0 0 0 0 0 0 ­0.1 0 0 0 Water transport 0 0 0 0 0 0 0 0 0 0 0 Air transport 0 0 0 0 0 0 0 0 0 0 0 Communication 0 0 0 0 0 0 0 0 0 0 0 Financial services 0 0 0 0 0 0.1 0 0 0 0.1 0 Insurance 0 ­0.1 ­0.1 ­0.1 ­0.1 ­0.1 ­0.1 ­0.1 ­0.1 0 ­0.1 Business services 0 0 0 0 0 0 0.1 0 0 0 0 Recreational services ­0.1 0 ­0.1 ­0.1 0 ­0.1 ­0.1 0 ­0.1 ­0.1 ­0.1 Government services 0 0 0 0 0 0 0 0 0 0 0 Dwelling 0.1 ­0.2 2.4 ­0.2 0.8 0.7 ­0.6 ­0.2 0.4 ­0.2 ­3.3 Source: Authors' simulations. a. Calculated as a weighted average of respective price changes, excluding intra-EU trade 84 Poverty and the WTO: Impacts of the Doha Development Agenda Table 3.8. Doha: Export Prices for All Regions a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Terms of trade ­6.1 5.1 ­2.8 0.2 1.1 ­2.1 ­2.2 ­0.5 ­0.8 0.8 Export price index 0.1 0.2 2.5 ­0.1 0.3 0.6 ­0.4 ­0.1 0.2 ­0.2 ­1.1 Primary agriculturea 1.1 0.6 5.1 ­0.1 1.2 0.2 0.1 0.4 ­0.1 ­0.2 ­1.1 Paddy rice 7.9 0.2 5.9 0 1.5 0.9 4.5 0.3 0 0.2 ­1 Wheat 1.6 0.4 4.2 0.4 0.9 0.7 ­0.2 0.5 0.5 ­0.2 0.6 Cereal grains 3.5 0.3 6 ­0.1 1.3 0.6 ­0.2 0.1 0.2 ­0.1 ­2 Fruits and vegetables ­0.1 0.1 5 ­0.2 1.2 0.8 0.1 0.1 ­0.2 ­0.4 ­1.1 Oilseeds 4.3 0.6 5 0.1 1.9 1.4 1.6 0.8 1.5 0.3 1.4 Raw sugar 0.8 0.3 5.4 ­0.2 1.1 0.9 0 1.2 0.7 ­0.3 ­1.5 Plant fibers 6.1 1 5.6 1 1.2 4.8 1.5 1.1 3.5 0.5 2.8 Other crops ­0.2 0.1 4.8 ­0.3 0.6 0.1 0 0.1 0 ­0.7 ­1.2 Cattle ­0.1 0.2 9.8 ­0.2 1.8 1.1 0.2 0.2 0.5 ­0.3 ­1.6 Animal products 0 0.3 6.7 ­0.1 1.3 1.1 0.1 0.1 0.1 ­0.5 ­1.5 Raw milk 0.3 0.2 6.3 ­0.1 1.5 2 ­0.1 0.5 0.3 ­0.4 ­0.6 Wool 2 0.1 3.1 ­0.1 1.8 0.9 0.5 0.4 0.3 0 0.4 Other primarya 0.1 ­0.2 1.2 0 0.1 0.4 0 ­0.2 0.6 0.1 ­0.6 Forestry 0 ­0.2 2.6 ­0.3 0.5 0.5 ­0.4 ­0.3 0.2 ­0.3 ­1.9 Fishing 0.1 ­0.2 1.8 ­0.4 0.8 0.6 ­0.5 ­0.7 0.4 ­0.3 ­2.5 Coal 0.3 0 0 0.8 0.1 0.4 0.1 0.1 0.4 0.4 ­0.2 Crude oil 0.1 0.1 0.4 0.1 ­0.1 0.4 0 0 0.2 0.1 ­0.5 Natural gas 0 ­0.1 ­1.9 ­0.1 0 0.2 ­0.2 1.2 0.5 0 ­4.5 Other minerals 0.4 ­0.2 1.2 ­0.1 0 1 ­0.4 0 0.7 0.1 ­1.2 Fooda 0.8 0.2 3.8 ­0.1 1 0.8 ­0.3 0 0.4 ­0.6 ­1.4 Bovine meat products 2.3 0.2 6.1 ­0.2 1.1 0.9 ­0.4 0.3 0.5 0.1 0.6 Other meat 0.8 0.4 5.8 ­0.2 1.1 0.9 ­0.3 0.1 0.1 ­0.5 ­1.5 Vegetable oils and fats 0.7 0.9 3.5 ­0.1 1.2 0.8 2.6 0.2 0.4 ­0.3 ­1 Milk 4.1 0.2 3.5 1.9 0.9 0.8 ­0.4 0 2.7 ­0.3 1.3 Processed rice 1.8 0.1 3.3 0 1.1 0.8 ­0.3 0.3 0.1 ­0.3 ­1 Sugar 2.1 0.1 3.2 ­0.2 1 0.8 ­0.3 0.1 0.5 ­4.4 ­1.4 Food products 0 0.2 2.6 ­0.1 1 0.7 ­0.3 0 0.4 ­0.6 ­1.5 Beverages and tobacco ­0.1 ­0.1 2.4 ­0.2 0.8 0.5 ­0.4 0 0.3 ­0.5 ­1.3 Textilea 0.1 0.3 1.9 0 0.3 0.9 ­0.5 0 0.5 ­0.9 ­0.5 Textiles 0.1 0.4 1.9 0 0.3 0.9 ­0.5 0.1 0.7 ­0.9 ­0.7 Wearing apparel 0.1 0.3 1.9 ­0.1 0.3 0.8 ­0.5 ­0.1 0.4 ­1 ­0.4 World Market Impacts of Multilateral Trade Reforms 85 Table 3.8. (Continued) a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Other manufactuersa Leather products 0 0 2.2 ­0.2 0.5 0.7 ­0.6 0 0.3 ­0.7 ­1 Wood products 0 ­0.1 2.1 ­0.2 0.4 0.6 ­0.5 ­0.2 0.3 ­0.3 ­1.2 Paper products 0 ­0.1 1.8 ­0.2 0.5 0.6 ­0.4 0.5 0.3 ­0.4 ­1 Petroleum, coal 0.1 0 0.6 0 0.1 0.3 ­0.1 ­0.3 0.1 0.1 ­1.7 Chemical, plastic ­0.1 ­0.1 1.7 ­0.1 0.2 0.4 ­0.4 ­0.1 0.2 ­0.4 ­0.9 Mineral products 0 ­0.1 1.8 ­0.1 0.4 0.5 ­0.5 0.1 0.3 ­0.3 ­1.4 Iron and steel 0 ­0.1 1.7 ­0.1 0.3 0.4 ­0.4 ­0.1 0.2 ­0.1 ­0.3 Metals 0 0 1.4 ­0.1 0.3 0.5 ­0.4 ­0.2 0.3 ­0.2 ­1.4 Metal products 0 ­0.1 1.8 ­0.2 0.4 0.5 ­0.4 0 0.3 ­0.3 ­0.7 Motor vehicles ­0.1 ­0.1 1.6 ­0.2 ­0.2 0.5 ­0.3 0 0.2 ­0.5 ­0.9 Transport equipment ­0.1 ­0.1 1.8 ­0.2 0.2 0.5 ­0.5 ­0.1 0.3 ­0.5 ­0.5 Electronic equipment 0.1 ­0.1 1.4 ­0.2 0.1 0.5 ­0.3 ­0.1 0.2 ­0.6 ­0.8 Machinery ­0.1 ­0.1 1.8 ­0.2 0.2 0.4 ­0.4 ­0.1 0.2 ­0.4 ­1 Manufactures 0 ­0.1 1.9 0.1 0.5 0.6 ­0.4 ­0.1 0.4 ­0.5 ­0.9 Electricity ­0.1 ­0.1 1.7 ­0.1 0.3 0.4 ­0.2 0 0.3 0 ­1.3 Gas ­0.2 ­0.1 1.2 ­0.2 0.5 0.5 ­0.5 ­0.1 0.2 ­0.1 ­2.8 Water 0 0 1.7 ­0.2 0.6 0.6 ­0.5 ­0.1 0.3 ­0.2 ­1.5 Construction ­0.1 ­0.1 2.1 ­0.2 0.5 0.6 ­0.5 0 0.3 ­0.4 ­1.6 Servicesa 0 ­0.1 2.1 ­0.2 0.6 0.6 ­0.5 ­0.1 0.3 ­0.2 ­1.9 Trade 0.2 ­0.1 2 ­0.2 0.7 0.7 ­0.6 ­0.2 0.3 ­0.3 ­2 Transport 0 ­0.1 1.7 ­0.2 0.6 0.6 ­0.5 ­0.1 0.3 ­0.2 ­1.2 Water transport 0 ­0.1 1.7 ­0.2 0.4 0.4 ­0.5 ­0.1 0.3 ­0.2 ­1.1 Air transport 0 ­0.1 1.7 ­0.2 0.3 0.5 ­0.5 ­0.1 0.3 ­0.2 ­1 Communication 0 ­0.2 2.1 ­0.2 0.6 0.7 ­0.6 ­0.1 0.3 ­0.2 ­1.4 Financial services 0 ­0.1 2.1 ­0.2 0.7 0.7 ­0.6 ­0.3 0.3 ­0.2 ­2 Insurance 0 ­0.1 2.1 ­0.2 0.6 0.7 ­0.6 ­0.3 0.3 ­0.2 ­1.7 Business services 0 ­0.1 2.2 ­0.2 0.6 0.7 ­0.6 ­0.1 0.3 ­0.2 ­2.5 Recreational services ­0.1 ­0.1 2.1 ­0.2 0.4 0.7 ­0.6 ­0.1 0.3 ­0.3 ­2 Government services 0 ­0.1 2 ­0.2 0.6 0.7 ­0.6 ­0.1 0.3 ­0.4 ­1.8 Dwelling 0.1 ­0.2 2.4 ­0.2 0.8 0.7 ­0.6 ­0.2 0.4 ­0.2 ­3.3 Source: Authors' simulations. a. Calculated as a weighted average of respective price changes, excluding intra-EU trade 86 Poverty and the WTO: Impacts of the Doha Development Agenda Table 3.9. Doha: Import Quantities for All Regions a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Import quantity index 1.3 ­0.7 3.4 ­0.4 4.6 0.5 ­0.3 ­0.6 0.0 2.4 ­1.9 Primary agriculturea 0.0 ­3.0 6.3 ­1.0 0.5 ­0.1 ­1.7 ­2.1 0.2 1.5 ­2.3 Paddy rice 35.7 0.3 14.0 1.1 1.5 0.9 ­32.2 ­10.7 ­11.3 4.8 ­11.9 Wheat 0.3 ­2.6 1.6 ­2.6 ­1.1 ­0.7 ­4.3 ­1.4 0.5 ­0.5 ­1.1 Cereal grains ­0.7 ­0.9 6.8 ­4.6 6.5 ­1.1 1.8 ­2.7 1.0 ­6.9 ­3.2 Fruits and vegetables 1.5 ­0.6 4.7 4.4 4.8 ­0.4 1.5 ­1.9 0.2 2.4 ­3.0 Oilseeds ­1.5 ­2.5 6.5 ­0.1 ­1.5 ­4.9 ­1.3 ­0.6 ­0.9 ­0.6 ­0.4 Raw sugar ­1.7 ­2.6 9.7 ­0.1 ­1.7 ­1.4 ­1.8 9.1 ­0.4 ­0.8 ­5.6 Plant fibers ­1.5 ­5.3 2.7 0.5 ­3.5 2.3 ­11.4 0.9 1.1 1.1 ­4.8 Other crops ­2.2 0.5 15.0 0.2 2.0 0.4 2.6 ­2.2 ­0.4 1.8 ­0.3 Cattle ­0.5 ­0.2 69.6 1.5 3.5 ­0.5 ­2.0 ­1.8 ­2.7 8.6 0.2 Animal products 0.1 ­0.1 28.4 1.9 3.1 4.1 ­0.6 ­1.4 2.7 0.6 ­0.3 Raw milk ­6.9 ­2.7 27.4 ­0.3 4.9 5.1 ­1.5 1.9 0.1 0.4 0.5 Wool ­3.8 ­0.9 32.5 ­4.2 0.4 4.2 ­4.2 0.1 ­0.3 1.2 2.7 Other primarya 0.0 ­0.2 ­0.2 0.0 ­1.1 0.5 ­0.2 ­0.7 ­0.1 0.5 ­3.0 Forestry ­0.1 ­1.3 6.6 2.8 0.8 0.9 ­0.4 ­1.3 ­1.8 2.3 ­3.8 Fishing 0.8 ­0.5 1.6 ­0.2 4.0 0.3 2.1 ­1.9 0.3 3.1 ­3.6 Coal ­0.1 1.7 ­2.7 ­0.4 2.6 0.9 ­0.3 1.5 0.0 ­3.6 ­1.2 Crude oil 0.1 0.1 0.7 0.0 ­1.7 0.8 ­0.2 0.3 0.1 2.2 ­2.2 Natural gas 0.0 ­0.6 ­27.6 ­1.2 14.2 3.1 ­2.8 16.3 10.2 0.4 ­55.1 Other minerals ­0.3 0.0 ­2.9 ­0.2 ­1.6 ­0.4 0.1 ­0.8 ­0.9 0.8 ­2.5 Fooda 5.9 ­2.6 5.1 ­2.3 6.6 ­0.1 0.3 ­1.2 1.2 2.4 ­3.5 Bovine meat products 24.9 ­0.3 22.5 ­16.5 4.5 4.0 ­2.4 ­5.1 0.5 ­12.4 0.5 Other meat 10.8 0.3 15.9 ­8.7 11.8 4.1 ­2.3 ­2.9 30.2 11.7 ­9.1 Vegetable oils and fats 7.4 ­2.3 8.2 0.1 8.2 1.9 2.7 ­2.0 4.0 5.1 ­2.6 Milk ­4.4 ­7.4 ­0.7 ­15.1 ­1.3 ­4.1 5.8 ­0.6 ­0.5 ­9.3 ­3.9 Processed rice 22.6 ­2.6 5.0 ­2.7 ­3.6 ­1.5 ­2.1 ­2.0 0.0 2.1 ­6.8 Sugar 29.3 ­3.6 ­1.6 ­4.0 0.1 ­2.3 ­1.6 ­0.6 4.9 2.7 ­6.6 Food products 1.7 ­0.1 4.6 ­0.6 7.8 1.3 ­0.6 ­1.1 0.2 3.7 ­3.3 Beverages and tobacco 2.6 ­0.5 2.3 8.3 4.0 3.0 1.2 ­0.3 ­0.2 3.5 ­3.1 Textilea 6.2 0.1 3.7 ­0.2 12.6 3.3 0.7 ­0.8 6.8 2.0 ­6.0 Textiles 5.5 0.1 3.6 ­0.2 12.2 3.2 ­1.4 ­0.9 7.5 2.2 ­6.1 Wearing apparel 7.1 0.4 4.6 ­0.1 16.1 3.9 6.0 ­0.5 ­0.7 1.9 ­4.1 World Market Impacts of Multilateral Trade Reforms 87 Table 3.9. (Continued) a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Other manufacturesa 1.1 ­0.3 3.7 ­0.1 4.6 0.2 ­0.2 ­0.4 ­0.5 3.8 ­1.4 Leather products 4.2 ­0.6 2.8 ­0.2 16.0 9.8 3.9 ­0.5 0.1 9.3 ­2.8 Wood products 0.1 ­1.2 8.1 ­0.2 2.9 0.1 ­0.6 ­0.5 ­0.9 8.0 ­3.8 Paper products 0.4 ­0.5 3.6 ­0.9 3.1 ­0.4 ­0.7 ­1.0 ­0.2 3.6 ­1.8 Petroleum, coal 0.5 ­0.1 0.1 0.0 3.1 0.2 ­0.2 0.2 0.0 2.9 0.1 Chemical, plastic 1.9 ­0.2 2.8 0.1 5.8 1.3 ­0.5 ­0.5 0.8 3.6 ­1.1 Mineral products 2.1 ­0.8 5.1 ­0.5 10.2 ­0.3 ­0.6 ­4.6 ­0.2 6.3 ­5.8 Iron and steel 0.5 ­0.1 1.6 ­0.3 2.9 ­0.3 ­0.5 ­0.2 ­0.5 2.0 0.6 Metals 0.8 ­0.6 0.9 0.2 3.1 ­2.1 ­0.3 ­0.7 ­0.8 4.6 ­4.6 Metal products 1.9 ­0.5 5.3 ­0.2 10.3 0.4 ­0.5 0.2 0.0 7.3 0.4 Motor vehicles 2.2 ­0.1 7.7 ­0.2 8.4 0.6 ­0.8 ­0.2 0.0 8.5 ­1.4 Transport equipment 1.1 ­0.7 3.2 ­0.2 5.1 0.2 ­0.6 ­0.2 0.2 7.6 ­0.4 Electronic equipment ­0.1 ­0.2 2.2 ­0.2 0.5 ­0.2 0.4 ­0.2 ­1.0 0.8 ­2.2 Machinery 1.4 ­0.1 4.7 ­0.2 6.5 ­0.7 ­0.2 ­0.1 ­0.2 1.2 ­1.6 Manufactures 0.5 ­1.1 7.4 0.8 14.4 0.9 ­0.1 ­0.2 0.7 11.8 ­3.4 Electricity 0.0 ­0.3 1.9 ­0.4 ­0.3 1.2 ­0.1 0.5 1.0 2.9 ­4.1 Gas 0.5 0.2 2.6 0.3 1.5 1.9 ­0.7 1.2 1.2 0.6 ­7.9 Water ­0.2 0.0 3.5 ­0.6 1.5 0.3 ­1.3 ­0.6 0.8 ­0.5 ­3.0 Construction ­0.1 ­0.4 6.2 ­0.5 1.2 1.4 ­1.5 ­0.1 0.8 ­0.7 ­4.8 Servicesa ­0.1 ­0.3 2.9 ­0.4 0.5 0.7 ­0.7 ­0.3 0.4 ­0.2 ­1.8 Trade ­0.3 ­1.0 3.9 ­0.5 0.3 0.7 ­1.1 ­0.5 0.1 ­0.5 ­2.3 Transport 0.1 ­0.3 2.8 ­0.5 1.0 0.9 ­0.7 0.0 0.8 ­0.1 ­0.9 Water transport 0.2 ­0.4 0.9 ­0.4 0.7 0.3 ­0.9 ­0.2 0.3 ­0.1 ­0.4 Air transport 0.0 ­0.1 1.5 ­0.3 0.3 0.8 ­0.6 ­0.2 0.6 ­0.2 ­0.8 Communication 0.0 ­0.2 0.6 ­0.5 0.8 0.9 ­1.1 ­0.2 0.4 ­0.2 ­0.1 Financial services ­0.2 0.3 3.5 ­0.5 0.9 1.0 ­0.7 ­0.4 0.2 ­0.2 ­0.8 Insurance 0.0 0.0 3.6 ­0.3 0.9 1.2 ­0.5 ­0.2 0.5 0.1 ­0.5 Business services ­0.1 ­0.2 4.1 ­0.3 0.8 0.5 ­0.9 ­0.3 0.3 0.0 ­2.2 Recreational services 0.0 ­0.2 2.3 ­0.5 0.8 1.4 ­0.9 ­0.2 0.7 ­0.1 ­3.2 Government services 0.0 ­0.2 1.6 ­0.5 1.1 1.3 ­1.1 ­0.4 0.6 ­0.4 ­3.9 Dwelling 0.0 ­0.1 0.6 ­0.2 ­0.1 0.1 ­0.1 ­0.4 0.0 0.2 ­1.0 Source: Authors' simulations. a. Calculated as a weighted average of respective price changes, excluding intra-EU trade 88 Poverty and the WTO: Impacts of the Doha Development Agenda Table 3.10. Doha: Export Quantities for All Regions a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Export quantity index 1.3 ­0.2 ­1.8 0.1 3.7 0.1 0.5 0.0 ­0.1 1.8 0.8 Primary agriculturea 0.0 6.2 ­11.8 ­0.7 14.0 ­1.6 0.9 1.4 ­4.8 6.2 0.6 Paddy rice 35.7 72.4 16.4 1.4 462.5 60.4 36.5 ­4.8 203.6 13.4 17.6 Wheat 0.3 328.0 ­24.0 7.9 10.7 ­0.3 ­29.1 ­7.4 32.3 8.2 ­8.4 Cereal grains ­0.7 31.1 0.5 0.5 20.1 3.7 8.3 1.5 3.7 11.2 ­19.6 Fruits and vegetables 1.5 ­5.4 ­16.2 ­3.9 22.7 2.5 1.4 ­4.0 ­6.4 3.5 0.9 Oilseeds ­1.5 20.5 ­4.6 8.2 27.0 9.2 9.1 11.6 7.0 16.2 6.2 Raw sugar ­1.7 ­10.5 10.7 ­13.0 74.5 95.0 ­14.8 ­20.7 ­22.9 ­14.4 7.0 Plant fibers ­1.5 15.0 ­1.0 18.1 30.2 2.0 28.7 8.7 12.9 17.0 14.3 Other crops ­2.2 ­2.1 ­23.1 ­5.1 ­2.4 ­2.1 ­3.6 ­2.3 ­2.0 ­6.6 ­0.1 Cattle ­0.5 ­6.9 ­27.9 ­0.3 ­3.5 ­2.8 5.1 ­8.5 ­7.8 ­5.8 ­1.7 Animal products 0.1 ­1.7 ­13.4 ­1.2 ­1.8 ­1.5 2.5 ­2.0 0.0 ­1.0 3.1 Raw milk ­6.9 ­5.8 ­38.8 ­3.0 ­14.5 ­17.0 ­3.8 ­7.4 ­5.8 ­1.5 0.1 Wool ­3.8 16.6 ­26.3 17.5 ­4.7 5.4 8.6 4.6 11.1 6.6 13.8 Other primarya 0.0 0.9 ­1.6 0.2 1.6 ­0.3 0.5 1.0 0.1 ­0.1 1.3 Forestry ­0.1 0.2 ­12.0 0.6 ­1.6 ­1.0 2.3 1.3 0.4 0.9 8.7 Fishing 0.8 1.0 ­1.8 ­1.1 2.7 1.7 2.3 0.6 0.3 2.2 5.1 Coal ­0.1 0.3 0.7 ­3.6 1.6 ­0.2 0.7 0.3 ­2.1 ­1.0 2.1 Crude oil 0.1 ­0.1 ­3.2 0.1 2.0 0.0 0.5 0.2 0.8 ­0.3 0.8 Natural gas 0.0 3.0 91.9 1.4 1.4 ­0.5 8.5 ­33.5 ­13.7 ­0.1 400.2 Other minerals ­0.3 1.1 ­1.5 0.5 0.4 ­1.0 0.8 0.7 0.0 0.3 1.9 Fooda 5.9 ­3.5 43.4 ­1.4 0.3 ­0.5 6.1 ­2.7 4.1 5.2 0.5 Bovine meat products 24.9 66.9 373.9 15.3 5.6 44.6 3.2 ­8.7 86.3 184.4 12.9 Other meat 10.8 11.0 ­4.9 21.5 ­21.4 16.6 53.1 ­13.9 36.0 19.4 11.7 Vegetable oils and fats 7.4 ­0.8 0.4 ­9.2 ­8.0 ­5.4 ­10.0 4.2 ­2.7 1.4 3.0 Milk ­4.4 61.7 30.9 115.0 48.6 51.4 51.0 49.3 22.3 47.8 13.0 Processed rice 22.6 13.0 ­0.9 ­9.4 49.5 46.5 0.8 ­0.9 ­35.8 16.7 6.1 Sugar 29.3 ­32.0 6.8 ­12.7 27.5 ­7.2 ­5.5 16.7 52.6 30.6 ­16.0 Food products 1.7 ­4.0 ­1.4 ­2.9 ­0.5 ­3.3 2.1 ­4.6 3.7 2.6 ­1.3 Beverages and tobacco 2.6 3.8 ­2.0 2.8 2.1 5.6 1.9 ­2.2 3.5 0.3 0.1 Textilea 6.2 0.1 ­4.4 2.9 10.0 4.1 ­13.5 ­7.7 13.3 15.8 ­9.6 Textiles 5.5 ­1.9 ­4.5 3.6 8.4 2.4 ­10.0 ­6.9 12.0 11.4 ­2.0 Wearing apparel 7.1 1.3 ­3.8 1.6 10.9 6.8 ­16.6 ­8.6 13.9 21.0 ­12.0 World Market Impacts of Multilateral Trade Reforms 89 Table 3.10. (Continued) a orld Sector W Bangladesh Brazil Cameroon China Indonesia Mexico Mozambique Philippines Russia ietnamV Other manufacturesa 1.1 ­1.9 ­9.3 0.0 1.5 ­0.3 1.0 0.2 ­1.3 2.7 1.8 Leather products 4.2 ­5.6 ­14.9 ­6.2 8.6 14.7 ­20.6 ­7.3 ­2.4 0.3 ­2.7 Wood products 0.1 9.6 ­10.0 0.1 ­1.5 ­0.3 2.6 ­0.5 ­2.2 1.3 6.8 Paper products 0.4 0.2 ­9.0 1.0 ­1.7 ­1.5 3.3 ­3.2 ­0.9 2.5 5.5 Petroleum, coal 0.5 0.4 ­0.9 0.2 0.0 ­0.7 0.4 1.9 0.9 2.0 6.2 Chemical, plastic 1.9 ­0.3 ­8.4 2.3 2.0 0.5 1.7 23.6 3.4 2.0 2.2 Mineral products 2.1 ­0.3 ­5.4 1.1 3.4 ­0.4 ­0.2 ­1.1 ­1.7 0.4 5.8 Iron and steel 0.5 3.7 ­8.2 ­0.9 0.0 ­1.0 5.2 0.0 ­0.8 1.3 2.6 Metals 0.8 ­9.9 ­5.2 0.7 0.8 ­2.9 2.0 ­0.1 ­2.2 4.2 10.6 Metal products 1.9 1.8 ­9.4 0.4 3.0 ­2.6 ­0.1 ­1.4 ­2.2 4.3 4.0 Motor vehicles 2.2 ­1.6 ­5.1 ­1.3 ­1.7 ­1.9 ­1.3 1.0 ­0.2 2.2 2.5 Transport equipment 1.1 ­2.3 ­14.3 ­1.0 2.1 ­2.2 2.3 ­1.9 ­1.6 7.5 1.8 Electronic equipment ­0.1 1.3 ­9.5 1.6 0.3 ­2.7 2.7 0.0 ­1.3 7.7 7.2 Machinery 1.4 0.9 ­11.6 0.1 0.7 ­2.4 0.8 ­0.1 ­1.3 2.6 7.5 Manufactures 0.5 6.7 ­12.7 ­3.1 ­0.5 ­5.2 0.6 ­0.6 ­5.1 1.6 3.3 Electricity 0.0 0.7 ­9.4 0.7 0.7 ­2.2 0.9 1.4 ­1.3 ­1.1 7.7 Gas 0.5 ­0.1 ­7.2 0.2 ­3.2 ­3.5 1.9 ­0.4 ­1.5 ­0.3 16.5 Water ­0.2 0.5 ­8.8 1.1 ­3.4 ­3.2 3.2 0.7 ­1.7 1.5 8.9 Construction ­0.1 1.2 ­7.4 0.5 ­2.1 ­2.4 1.7 ­0.1 ­1.4 1.5 6.0 Servicesa ­0.1 0.5 ­7.5 0.8 ­2.1 ­2.3 2.2 0.6 ­1.0 1.0 7.6 Trade ­0.3 0.8 ­7.0 1.1 ­2.3 ­2.3 2.4 1.2 ­1.0 1.5 8.3 Transport 0.1 0.8 ­6.0 0.8 ­2.2 ­2.2 2.3 0.5 ­0.9 0.9 5.0 Water transport 0.2 1.0 ­5.7 1.1 ­1.2 ­1.1 2.4 0.9 ­0.6 1.4 4.7 Air transport 0.0 0.7 ­6.0 0.8 ­1.0 ­1.9 2.2 0.5 ­0.9 1.0 4.2 Communication 0.0 0.7 ­7.6 0.8 ­2.0 ­2.4 2.3 0.3 ­1.1 0.7 5.7 Financial services ­0.2 0.6 ­7.6 0.9 ­2.4 ­2.6 2.1 1.2 ­1.2 0.7 7.7 Insurance 0.0 0.7 ­7.7 0.8 ­2.4 ­2.6 2.2 1.2 ­1.3 0.7 6.6 Business services ­0.1 0.4 ­8.1 0.8 ­2.3 ­2.5 2.2 0.5 ­1.2 0.8 10.2 Recreational services 0.0 0.4 ­7.6 0.8 ­1.7 ­2.8 2.0 0.4 ­1.3 0.9 7.8 Government services 0.0 0.4 ­7.4 0.6 ­2.2 ­2.6 2.2 0.5 ­1.2 1.4 7.3 Dwelling 0.0 ­0.1 0.6 ­0.2 ­0.1 0.1 ­0.1 ­0.4 0.0 0.2 ­1.0 Source: Authors' simulations. a. Calculated as a weighted average of respective price changes, excluding intra-EU trade 90 Poverty and the WTO: Impacts of the Doha Development Agenda vidual country case study authors are the experts in assessing the likely responses of their respective national economies to these external developments and to their own trade policy reforms. Accordingly, there is a need to blend the global and national analysis into a single, coherent story. This section summarizes the approach used for the studies reported in this volume.7 The idea of first solving a global model and passing the world price or volume results (or both) on to a national model is rather intuitive, but when faced with implementation, several problems arise. The first problem is that national policy reforms should not want be implemented twice--once in the global model and once in the national model. The global analysis should instead capture only the impact of policy reforms in the rest of the world. Consider, for example, the case of Brazil. It is desirable for the national model to receive results from the two sce- narios discussed above, omitting Brazilian reforms in the process. The world price changes passed on to the national model would then reflect the impact on world markets of policy reform in all countries except for Brazil. The national model then takes these world price changes as exogenous and implements the Brazilian portion of the reform package. In the special case where the national model is identical to the Brazilian portion of the global model, this approach should give the same results for Brazil as were obtained under the comprehensive reform sim- ulations reported in the preceding tables.8 Thus, the first problem is resolved by solving each scenario with the focus economy omitted, thereupon passing the resulting world market effects onto the national studies, to be implemented as exogenous shocks in the country case studies. A second problem that arises in linking the two models is how specifically to pass the information on world markets to the national model. Because all of the national models take import price as given, and import supply facing the focus economies in the global model is very elastic, it is easy to handle this side of the story. The import price changes generated in the global model are appropriately aggregated and then passed on to the national model, where they are applied as exogenous shocks. Exports, however, are more challenging. The fundamental problem with exports is that the global model treats prod- ucts as being differentiated by origin--the Armington assumption discussed ear- lier in this chapter. Therefore, Brazil's export prices are not exogenous, even for products where their world market share is relatively small. Accordingly, the impact of world market changes on both price and quantity and the models' dif- ferences in the connections between these two variables must be considered. Specifically, there are different export supply schedules for commodities in the global and national models due to the differences in the two models' representa- tions of the Brazilian economy. It is often found that the national models of devel- oping countries have less elastic export supply schedules, reflecting domestic World Market Impacts of Multilateral Trade Reforms 91 constraints on export capacity not captured in the global model. However, the export demand schedule is treated as having the same slope in the two models. Nearly all of the country case study authors have built downward-sloping export demand schedules into their models, and they have taken the elasticity of export demand from the global model.9 Figure 3.2 illustrates this point for a specific product--for example, processed sugar. The initial equilibrium for exports of Brazilian sugar is at point A, where the supply schedule (SG in the global model and SN in the national model) inter- sects the global demand schedule, D. When a Doha scenario is implemented in the Figure 3.2. Transmission of Global Results to a National Model SN' SN SG SG' D' D PH C PG B A D' SG' SG SN SN' D QN QG Source: Fan Zhai. 92 Poverty and the WTO: Impacts of the Doha Development Agenda rest of the world, the first thing that happens to Brazilian exports of sugar is that the reduction in protection in the industrialized countries results in an outward shift in export demand to D'D'. But the export supply curve may also be affected--owing, for example, to the limited endowments of land, labor, and cap- ital in Brazil and the simultaneous shifts to import prices and other export demands. Here, it is assumed that there is a reduction in demand in manufactur- ing, thereby releasing additional resources for use in agriculture and causing export supply to shift to the right. The global model finds a new equilibrium at point B, with price PG and quantity QG. Referring back to the tables of results for Doha shows that point B embodies a small increase in Brazil's sugar price (3.3 percent) and a somewhat larger increase in quantity (6.8 percent). To communicate these changes in global markets to the national model, it is assumed that the national model will adequately take care of the supply shift due to domestic, general equilibrium changes in response to liberalization in the rest of the world. This makes sense, because both models typically draw on the same underlying social accounting matrix and both embody the same general equilib- rium restrictions. Thus, there is only a need to identify the extent to which the demand curve shifts outward in figure 3.2. As shown in the appendix to this chap- ter, this demand shift can be readily established with three pieces of information: the change in price, the change in quantity, and the slope of the export demand schedule. One can then solve the demand function for the shift necessary to ensure that the new equilibrium still lies on the demand curve. Of course, this approach is limited by the fact that the two models have differ- ent characterizations of supply. For example, in figure 3.2, even assuming the same shifts in demand and supply, the national model generates a different equi- librium at point C, with higher price, PN, and lower quantity, QN, than in the global model as a result of the less elastic nature of supply. In summary, the price and quantity changes generated by the national model in response to these exoge- nous shocks will not be exactly the same as the global model. A detailed compari- son of the price and quantity outcomes in the global and national models has been conducted for Brazil and for China, and they yield correlations of nearly 0.9 for quantities and somewhat less for prices. So, although the two sets of results are different, they are indeed highly correlated. In summary, there is an inbuilt inconsistency: the global and national models treat exports differently. This cannot be resolved perfectly, but the approach used here to communicating between the global and national models permits use of the strengths of each of the two models while ensuring broad consistency in results. It could be further perfected by modifying the global model to better reflect the national models, but this is well beyond the scope of the present project--and maybe not even desirable, because, for most purposes, a global model needs a World Market Impacts of Multilateral Trade Reforms 93 degree of consistency in its treatment of individual countries and so cannot reflect all the country details precisely.10 In closing, one word of caution to the reader is advisable. In the subsequent chapters, authors report price changes for full liberalization and for the core Doha scenario, obtained from this global exercise. It will seem natural to compare these price changes to those reported in this chapter. However, recall that the price changes provided to the individual country case study authors omit the impacts of that country's own actions. Also, in some cases, the country case study authors have rescaled the price changes by normalizing them on the domestic consumer price index, for example. Although this rescaling doesn't affect the results in their models, which depend only on relative price changes, it does make direct compar- isons of price changes between chapters more difficult. 94 Poverty and the WTO: Impacts of the Doha Development Agenda Chapter 3 Annex A: Shocking a Single-Country CGE Model with Export Prices and Quantities from a Global Model Mark Horridge and Fan Zhai This annex explores the following problem: suppose a GTAP simulation has pro- duced percent changes in import and export quantities and border prices for a particular country, say Brazil. How are the GTAP results applied to a single-coun- try CGE model of Brazil (assuming it has the same commodity aggregation as the GTAP simulation)? Discussion of this issue distinguishes between the two most common types of single-country CGE models: models where exports and domes- tically produced goods are perfect substitutes (type A) and those where they sub- stitute only imperfectly (type B). Type A Single-Country CGE Model The type A single-country CGE model has capital and labor mobile among sec- tors, and export goods are identical to those domestically used. In type A models, individual export supply functions tend to be very flat, especially for nonprimary goods. The (small) slope derives from economy-wide factor constraints and, per- haps, sector-specific fixed factors such as land. Figure 3A.1. Demand and Supply for Single Export in Type A Model P Supply Flat demand Q Source: Authors. Some slope is needed for export demand functions in the type A model. If export prices were fixed (small country assumption), quite small shifts in supply func- tions could cause export quantities to fluctuate wildly (the overspecialization or flip-flop problem). Indeed, at first order, each commodity price will be a share- weighted average of the prices of factors or imports. Hence, with more goods than World Market Impacts of Multilateral Trade Reforms 95 factors (and import prices fixed), not all export prices can vary independently. Thus, in a type A model, attempts to exogenously fix all export prices will fail or will simply produce ridiculous results. To prevent this problem, Type A models usually postulate a downward-sloping constant-elasticity demand curve for each export good, as shown in figure 3A.1. This means that export expansion will be accompanied by falling export prices and a TOT loss. Indeed, at modest tariff lev- els, this TOT loss will dominate the efficiency gains obtained from unilateral tariff reduction and aggregate welfare will fall. This is simply evidence of a nonzero optimal tariff. Type B Single-Country CGE Model In a type B single-country CGE model, export prices are not identical to prices of domestically used goods. The two are related via a constant elasticity of transfor- mation (CET) transformation frontier. This gives individual export supply func- tions a marked upward slope. Type B models are therefore compatible with fixed export prices (the small country assumption) and therefore zero optimal tariffs. Figure 3A.2. Demand and Supply for Single Export in Type B Model P Flat demand Supply Q Source: Authors. For each good, the export price is related to the export and domestic quantity ratio for that good, thus export prices can be shocked independently and export quantities will adjust to suit. Both types A and B models normally assume that cost, insurance, and freight (CIF) inclusive import prices are fixed, and that users substitute between imports and domestic goods via a constant elasticity of substitution (CES) nest, with the ease of substitution governed by an Armington elasticity. Therefore, there is no difficulty about shocking import prices. Here, the concentration is on the prob- lem of how to shock exports. 96 Poverty and the WTO: Impacts of the Doha Development Agenda Single Country within GTAP The individual countries (or regions) embedded within the GTAP model are akin to type A models (there is no export and domestic CET). The downward slope on export demand schedules derives from the Armington assumption applied in other regions. Indeed, the export demand elasticity for good i facing a country with small world market share will be approximately equal to the (interimport) Arm- ington elasticity of substitution. Thus, in the global model, the import and export demand elasticities are inextricably intertwined. Figure 3A.3. Demand and Supply for a Single Export in a GTAP Simulation P F final Demand Supply p fp E initial q Q Source: Authors. Figure 3A.3 shows how shifts in export supply and demand schedules lead to observed changes in price (p) and quantity (q). Here, the focus is on the vertical shift in the demand schedule, fp, because that will prove crucial in the subsequent methodology. Note that fp is not equal to the price change, p. Depending on the supply shift, p and fp may even be of opposite sign. Also, note that even if the GTAP simulation shows only the effect of other countries' actions, the supply curves would still be expected to shift, because all sectors use the same mobile fac- tors, and an expansion of economic activity in another sector will raise produc- tion costs in the focus sector. What Should the GTAP Model Communicate to the Single-Country Model? In using the GTAP model to drive a single-country model, should an attempt be made to match the GTAP export prices or quantities or both of these? The aim here is to let the single-country model determine export supply behavior and take world demand changes from the GTAP model. Figure 3A.3 shows that the GTAP World Market Impacts of Multilateral Trade Reforms 97 export prices and quantities are simultaneously determined by the slopes and shifts of the GTAP export demand and supply curves. The same results would not be expected if the GTAP supply behavior were replaced by a supply curve from another single-country model. Rather, the numbers to take from the GTAP model are the slope and shift (fp) of the world demand schedule. Of course, there are alternative methods of communicating the global model results to the national models. The authors experimented with many of these and found them deficient in one way or another.11 Calculating the Vertical Shift in the GTAP Model's Export Demand Curve If export prices and quantities from a GTAP simulation and the slope of the export demand curve are known, fp, the vertical shift in the demand curve, can be calculated as follows: The GTAP export demand curve can be written: (3A.1) Q = [FP/P] ESUBM and ESUBM is the (positive) slope of the demand curve, approximately equal to the GTAP elasticity of substitution among imports. In proportional (log-change, percent) form this becomes: (3A.2) q = - ESUBM*(p - fp) or (3A.3) p = fp - q/ESUBM where lowercase variables denote percentage changes in their uppercase counter- parts. Hence (3A.4) fp = p + q/ESUBM For example, suppose that the country model for Brazil was type A, based on the same input-output table as used for the GTAP database, used the same factor mobility assumptions as GTAP, and used the same trade elasticities. Further sup- pose that the export demand elasticities were equal to the GTAP intercountry elas- ticity of import substitution. In short, suppose that the Brazil model was essen- tially the same as the Brazil part of the GTAP model. Then it would be expected 98 Poverty and the WTO: Impacts of the Doha Development Agenda that appropriate shocks to FP would produce very similar price and quantity changes to the GTAP model. In practice, the similarity criteria just listed will not all be satisfied. Thus, tak- ing the slope and shift (fp) of the world demand schedule from the GTAP model will yield export prices and quantities different from the GTAP simulation. That could be desirable if the Brazil single-country model represented Brazil better than the Brazil part of the GTAP model. This is the operating assumption used in this book. Summary of Recommended Approach For Type B Models, Add an Export Demand Curve for Each Good Mimic the GTAP export demand curve by adding equation (3A.1) to the model for each exported good. Type A models already have such an equation: the elastic- ity, ESUBM, should be taken from the GTAP parameter file. The Shock from the GTAP Model is a Change in FP (Export Demand Curve Shift) Given ESUBM, and percent changes q and p from the GTAP simulation, the percent change, fp, can be computed at first order as using equation (3A.4), or exactly as: (3A.5) fp = 100*[a - 1] where a = [1+0.01*p]*([1+0.01*q]1/ESUBM) Tailor the Single-Country Model to Resemble the GTAP Simulation and Model This includes choosing the trade elasticities, closure, and method of tax redistrib- ution that matches the GTAP treatment. For type B models, the CET should be set to a high value, or eliminated altogether, because the role it has played (to prevent flip-flopping of results) is no longer necessary. What about the Import Side? A similar argument could be made about import prices and quantities: The GTAP model presents upward-sloping import supply curves to the single-coun- try model, GTAP changes in import prices and quantities are again simultane- ously determined by world import supply curves (which are borrowed from the World Market Impacts of Multilateral Trade Reforms 99 GTAP model), and GTAP import demand curves will be replaced with those from the single-country model. Should the GTAP import supply curve indeed mimic and shift? No. It seems that merely shocking import prices is likely to be adequate, because, in the GTAP model, the import supply curves to a small coun- try are really very flat, and in all the models, the import demand curves (which use the import-domestic Armington elasticities) are comparatively steep. Hence, vertical shifts in import supply are well proxied by exogenous price changes (see figure 3A.4). Figure 3A.4. Demand and Supply for a Single Import P Supply Demand Q Source: Authors. Numerical Examples Tables 3A.1 and 3A.2 illustrate some of the above points using results from a GTAP Doha-all simulation driving single-country models for Brazil and China. The rows, corresponding to commodities, are ordered by the initial value of exports in the GTAP model. The tables show percent price and quantity changes from the GTAP simulation and the implied shift in the GTAP export demand curve, assuming that it has slope dictated by the GTAP model's Armington elastic- ity of substitution. These tables also report resulting percent price and quantity changes from the single-country models. Import prices were also shocked but are not shown in the table. (The focus country's tariffs were not changed for this sim- ulation.) Several points are worth making about the tables. First, the GTAP price change is a poor proxy for the GTAP demand shift. Second, the prices and quantities from the single-country model, although highly correlated with those from the GTAP model, are rather different in magnitude (and sometimes sign) because supply behavior is different in the two models. The correlation is higher for the changes in quantities (0.87 for both countries) than for the changes in prices (0.7 for Brazil 100 Poverty and the WTO: Impacts of the Doha Development Agenda Table 3A.1. Interaction of the GTAP Model with the Brazil Model Price Export Demand Price Export Export Elasticity of change change shift change change values substitution Sector (GTAP) (GTAP) (GTAP) (model) (model) (GTAP) (GTAP) Automobiles 1.5 ­3.6 0.9 0.7 1.0 5,316.9 5.6 Electric Materials 1.8 ­11.9 0.3 0.8 ­4.1 5,250.6 8.8 Enterprise services 2.2 ­8.3 ­2.2 0.9 ­6.1 4,475.9 2.0 Chemicals 1.6 ­8.9 0.2 0.9 ­4.2 4,289.6 6.6 Other food 2.4 ­0.6 2.2 0.8 5.9 3,969.1 4.0 Mineral extraction 1.1 ­1.4 0.4 0.9 ­1.0 3,833.8 1.8 Soybeans 3.8 ­9.6 1.7 2.1 ­1.7 3,747.7 4.9 Machinery 1.8 ­15.2 ­0.7 0.8 ­10.0 3,656.9 6.7 Iron products 1.7 ­8.4 0.2 0.8 ­3.3 3,352.2 5.9 Other agriculture 3.8 ­19.0 ­0.4 ­0.1 ­1.7 3,250.7 5.0 Slaughtering 5.0 169.7 17.6 7.8 102.5 2,714.7 8.8 Footwear 2.1 ­15.6 0.0 1.5 ­11.8 2,537.0 8.1 Electronic equipment 1.4 ­9.9 0.2 0.7 ­4.0 2,513.6 8.8 Paper products 1.8 ­9.8 0.0 0.8 ­4.6 2,502.6 5.9 Wood furniture 2.0 ­10.3 0.4 0.8 ­2.7 2,248.6 6.8 Nonferrous metals 1.4 ­6.3 0.6 0.8 ­1.4 2,185.0 8.4 Transportation 1.6 ­6.0 ­1.5 0.8 ­4.5 1,635.0 2.0 Sugar refining 2.7 4.6 3.6 1.2 13.0 1,435.0 5.4 Nonmetallic minerals 1.8 ­5.3 0.8 0.9 ­0.7 1,320.3 5.8 Textiles 1.8 ­7.2 0.8 0.7 0.3 906.3 7.5 Other metal products 1.8 ­10.3 0.2 0.8 ­4.3 762.0 7.0 Wholesale-retail trade 2.0 ­7.2 ­1.8 1.0 ­5.5 713.8 2.0 Public administration 2.0 ­7.6 ­2.0 0.9 ­5.8 700.7 2.0 Corn 5.1 10.5 9.2 1.8 19.8 690.9 2.6 Miscellaneous industries 1.9 ­14.6 ­0.4 1.0 ­9.1 671.7 7.0 Financial institutions 2.0 ­7.9 ­2.1 0.8 ­5.7 636.2 2.0 Vegetable oils 2.8 ­2.3 2.5 1.5 6.5 521.3 6.6 Petroleum refining 0.5 ­1.0 0.3 0.8 ­2.0 407.9 4.2 Communications 2.1 ­8.0 ­2.1 0.9 ­5.9 284.5 2.0 Apparel 1.8 ­8.1 0.7 0.9 ­1.5 243.0 7.4 Farm services 2.0 ­7.8 ­2.1 1.3 ­6.5 173.6 2.0 Cotton 4.0 ­13.4 1.1 ­1.4 12.9 173.3 5.0 Poultry 5.5 ­12.3 0.3 8.6 ­19.6 170.0 2.6 Coffee processing 2.2 ­2.3 1.2 0.7 1.1 76.0 2.3 Dairy 3.0 35.9 7.4 0.7 55.8 30.6 7.3 Civilian construction 2.1 ­7.8 ­2.0 1.0 ­5.8 30.2 2.0 Petroleum and gas extraction 0.3 ­2.8 ­0.1 0.9 ­7.8 16.7 8.0 Livestock 8.3 ­22.8 1.5 13.8 ­39.8 5.3 4.0 Paddy rice 4.6 17.6 6.3 1.0 61.2 2.1 10.1 Milk 4.9 ­31.9 ­13.4 0.4 ­26.2 1.1 2.0 Wheat 3.5 ­17.1 1.3 1.6 ­2.3 0.8 8.9 Sugar cane 4.2 18.0 7.7 1.7 32.2 0.0 5.0 Building rental 2.4 0.3 2.6 1.0 3.1 0.0 2.0 Source: Authors' calculations. World Market Impacts of Multilateral Trade Reforms 101 Table 3A.2. Interaction of the GTAP Model with the China Model Price Export Demand Price Export Export Elasticity of change change shift change change values substitution Sector (GTAP) (GTAP) (GTAP) (model) (model) (GTAP) (GTAP) Electronics 0.6 ­2.8 0.2 0.2 0.3 74,798.7 8.8 Apparel 0.9 8.9 2.1 0.9 9.2 69,587.1 7.4 Electrical machinery 0.7 ­2.1 0.5 0.3 1.5 53,484.2 8.1 Instruments 0.7 ­2.1 0.5 0.3 1.3 53,484.2 8.1 Textiles 1.0 5.5 1.7 0.9 6.4 39,431.7 7.5 Social articles 0.9 ­2.2 0.6 0.4 1.6 39,173.8 7.5 Other manufactures 0.9 ­2.2 0.6 0.5 0.7 39,173.8 7.5 Leather 1.1 7.2 2.0 0.7 10.5 30,420.6 8.1 Chemical 0.8 ­0.1 0.7 0.6 0.8 22,775.9 6.6 Medicine 0.8 ­0.1 0.7 0.6 1.0 22,775.9 6.6 Synthetic fibers 0.8 ­0.1 0.7 0.3 3.1 22,775.9 6.6 Rubber and plastic 0.8 ­0.1 0.7 0.4 2.5 22,775.9 6.6 Transport 0.9 ­2.3 0.3 0.3 ­0.3 13,476.4 3.8 Furniture 0.8 ­2.8 0.4 0.4 0.1 11,505.9 6.8 Metal products 0.8 1.2 0.9 0.6 2.3 10,991.9 7.5 Building materials 0.8 2.6 1.3 0.9 2.3 7,983.8 5.8 Commerce 0.9 ­2.5 0.3 0.4 ­0.5 7,792.5 3.8 Food processing 1.5 ­0.5 1.4 1.0 2.3 7,766.2 5.6 Other transport equipment 0.7 ­0.9 0.6 0.3 2.8 5,351.4 8.6 Social services 0.9 ­2.4 0.2 0.3 ­0.3 5,156.0 3.8 Other crops 1.6 14.7 4.5 2.9 7.5 4,033.6 4.9 Paper and printing 0.9 ­2.9 0.4 0.4 ­0.1 3,112.6 5.9 Nonferrous mining 0.7 ­2.2 0.4 0.6 ­1.0 3,041.3 8.4 Nonferrous metals 0.7 ­2.2 0.4 0.4 ­0.2 3,041.3 8.4 Ferrous mining 0.8 ­1.3 0.5 0.5 0.0 2,942.2 5.9 Iron and steel 0.8 ­1.3 0.5 0.3 1.0 2,942.2 5.9 Crude oil 0.1 ­0.5 0.1 0.1 ­0.4 2,739.7 14.9 Machinery 0.7 ­6.8 ­0.6 0.1 ­3.3 2,671.2 5.6 Special equipment 0.7 ­6.8 ­0.6 0.0 ­3.3 2,671.2 5.6 Automobiles 0.7 ­6.8 ­0.6 0.0 ­2.9 2,671.2 5.6 Education, science and health services 0.9 ­2.4 0.2 0.4 ­0.6 2,249.0 3.8 Public administration 0.9 ­2.4 0.2 0.4 ­0.6 2,249.0 3.8 Other livestock 1.8 ­1.3 1.3 1.1 0.7 1,528.2 3.1 Refined petroleum 0.4 0.2 0.4 0.3 0.4 1,264.3 4.2 Coal mining 0.6 0.8 0.7 0.7 0.3 1,250.8 6.1 Quarrying 0.5 0.1 0.6 0.5 0.3 1,033.9 1.8 Beverages 1.2 2.1 2.1 1.0 2.6 961.9 2.3 Tobacco 1.2 2.1 2.1 1.0 2.6 961.9 2.3 Financial services 0.8 ­2.3 0.2 0.3 ­0.4 886.0 3.8 Corn 2.0 21.1 9.7 4.5 13.7 721.9 2.6 Construction 0.8 ­2.6 0.1 0.3 ­0.5 690.5 3.8 Grain milling and forage 1.4 52.9 10.1 4.6 30.1 568.5 5.2 Fishing 1.2 2.9 2.4 1.3 2.6 504.6 2.5 102 Poverty and the WTO: Impacts of the Doha Development Agenda Table 3A.2. (Continued) Price Export Demand Price Export Export Elasticity of change change shift change change values substitution Sector (GTAP) (GTAP) (GTAP) (model) (model) (GTAP) (GTAP) Telecommunications 0.8 ­2.2 0.2 0.2 0.0 482.2 3.8 Utilities 0.7 ­1.4 0.5 0.4 0.2 290.7 5.6 Wool 2.2 ­7.1 1.6 1.1 6.7 210.5 12.9 Vegetable oils 1.7 ­9.2 0.2 0.8 ­3.7 190.5 6.6 Forestry 0.7 ­1.4 0.4 0.7 ­1.3 102.6 5.0 Cotton 1.8 28.6 7.0 4.4 13.2 88.5 5.0 Other agriculture 1.8 28.6 7.0 4.2 14.0 88.5 5.0 Wheat 1.4 6.8 2.1 0.0 0.0 48.3 8.9 Rice 2.1 452.5 20.9 15.3 61.5 42.5 10.1 Sugar 1.4 24.8 5.6 1.5 24.2 27.6 5.4 Source: Authors' calculations. and 0.61 for China).12 Nevertheless, if the single-country model best describes the particular country, and world demand changes are well summarized by the demand shift in the global model, then the single-country model price and quan- tity changes are the best estimates available. Notes 1. For the details on the database, see Dimaranan and McDougall (2006). 2. http://www.gtap.agecon.purdue.edu/products/models/. 3. Unfortunately, because of a lack of data on domestic purchases and prices, those authors are unable to estimate the elasticity of substitution between domestic goods and imports. As with the stan- dard GTAP model, these parameters are still obtained using the "rule of two" (that is, the import- import elasticities are assumed to be twice as large as the import-domestic elasticities). Liu, Arndt, and Hertel (2004) formally test this hypothesis in a model-based analysis of changing trade shares in East Asia over the 1980s and early 1990s. They fail to reject this hypothesis. 4. Note that two of the focus countries, Cameroon and Zambia, are omitted from table 3.1. Cameroon is omitted because this country is not broken out in the GTAP database. Therefore, price impacts of liberalization in the rest of the world are inferred from the impacts on the "rest of SSA" region. In the case of Zambia, the country study focuses solely on cotton and therefore draws on par- tial equilibrium studies of trade reform. 5. Because Russia is still in the process of joining the WTO, some assumptions about accession were made to obtain this binding. 6. Price changes for disaggregated products are Divisia indexes. The composite price and quantity indexes reported in these tables represent aggregates of individual bilateral prices or trade flows that have been aggregated using base-period trade weights--free on board in the case of exports and cost, insurance, and freight (CIF) in the case of imports. Intra-EU trade is excluded from the world price and volume changes. 7. This section relies on research conducted by Mark Horridge and Fan Zhai. The authors of this chapter thank those country case study authors for their valuable insights and guidance on this topic. The appendix to this chapter offers a detailed description of the methodology they developed. World Market Impacts of Multilateral Trade Reforms 103 8. Of course, if this were the case, then there would be no reason to separate the two analyses. The reason for the two-step approach is that the national models are more complete and accurate in their representation of the focus economies. 9. In the global model, the price elasticity of export demand facing a small country is well approx- imated by the intercountry elasticity of substitution among imports. These have been econometrically estimated and are reported in table 1 of Hertel and others (2004). 10. A model is, by definition, a simplification of reality--that is the point of models. 11. For instance, the most obvious thing would be to simply perturb the export demand schedule (type A model) or the export price (type B model) by the amount of the GTAP price change. However, as will be seen, this produces perverse results in the type A model case, because fp and p frequently move in opposite directions. In the type B model, this produces reasonable price changes, but it can produce quantity changes that are far too small. Other strategies involve imposing some sort of tech- nical change or export tax or subsidy in the national model, but these have undesirable welfare conse- quences. 12. The relatively smaller correlation for China's prices is likely due to the fact that this model retained the CET assumption on exports (type B model), whereas the Brazil model dispensed with this assumption as per the preferred approach outlined in this appendix. References Aksoy, M. A., and J. C. Beghin, eds. 2004. Global Agricultural Trade and Developing Countries. Washing- ton, DC: World Bank. Cranfield, J. A. L., P. V. Preckel, J. S. Eales, and T. W. Hertel. 2003. "Estimating Consumer Demand Across the Development Spectrum: Maximum Likelihood Estimates of an Implicit Direct Additiv- ity Model," Journal of Development Economics (68): 289­307. Dimaranan, B. V., and R. A. McDougall. 2005. Global Trade, Assistance, and Production: The GTAP 6 Data Base. West Lafayette, IN: Center for Global Trade Analysis, Purdue University. Hertel, T. W., ed. 1997. Global Trade Analysis: Modeling and Applications. Cambridge, MA: Cambridge University Press. Hertel, T. W., D. Hummels, M. Ivanic, and R. Keeney. 2004. "How Confident Can We Be in CGE-Based Assessments of Free Trade Agreements?" GTAP Working Paper 26, Center for Global Trade Analy- sis, Purdue University, West Lafayette, IN. http://www.gtap.agecon.purdue.edu/resources/ working_paper.asp. Hertel, T. W., and M. E. Tsigas. 1997. "Structure of the GTAP Model." In Global Trade Analysis: Model- ing and Applications, ed. T. W. Hertel. Cambridge, MA: Cambridge University Press. Keeney, R., and T. W. Hertel. 2005. "GTAP-AGR: A Framework for Assessing the Impacts of Multilat- eral Changes in Agricultural Policies." GTAP Technical Paper 24, Center for Global Trade Analysis, Purdue University, West Lafayette, IN. https://www.gtap.agecon.purdue.edu/resources/ tech_papers.asp. Liu, J., T. C. Arndt, and T. W. Hertel. 2004. "Parameter Estimation and Measures of Goodness of Fit in a Global General Equilibrium Model." Journal of Economic Integration 19 (3): 626­49. OECD. 2001. Market Effects of Crop Support Measures. Paris: OECD. Peeters, L., and Y. Surry. 1997. "A Review of the Arts of Estimating Price-Responsiveness of Feed Demand in the European Union." Journal of Agricultural Economics 48: 379­92. Rae, A. R., and T. W. Hertel. 2000. "Future Developments in Global Livestock and Grains Markets: The Impacts of Livestock Productivity Convergence in Asia-Pacific." Australian Journal of Agricultural and Resource Economics 44: 393­422. Seale, James, Anita Regmi, and Jason Bernstein. 2003. International Evidence on Food Consumption Pat- terns. Technical Bulletin 1904, Economic Research Service, U.S. Department of Agriculture, Wash- ington, DC. 104 Poverty and the WTO: Impacts of the Doha Development Agenda Surry, Y. 1990. "Econometric Modeling of the European Compound Feed Sector: An Application to France." Journal of Agricultural Economics 41: 404­21. Valenzuela, E., T. W. Hertel, R. Keeney, and J. J. Reimer. 2005. "Assessing Global CGE Model Validity Using Agricultural Price Volatility." GTAP Working Paper No. 33, Center for Global Trade Analysis, Purdue University, West Lafayette, IN. https://www.gtap.agecon.purdue.edu/resources/ working_papers.asp. Part II price linkages 4 Multilateral Trade Liberalization and Mexican Households: The Effect of the Doha Development Agenda Alessandro Nicita* Summary Empirical evidence suggests that global trade reforms are unlikely to produce analogous results across countries, especially when analyzing their effect on poverty. This implies that the analysis of trade reform on social welfare cannot be generalized and needs to be conducted on a country-by-country basis. Moreover, even within the same country, geographic areas, households, and individuals are likely to be differentially affected, some of them benefiting more than others, while others might lose. With this in mind, this chapter provides a quantitative estimate of the effect on Mexican households from the implementation of the Doha Development Agenda (DDA). The analysis uses a two-step approach for which changes in prices and factors are estimated through a computable general equilibrium (CGE) model (Global Trade Analysis Project--GTAP) and then *The author wishes to thank Thomas Hertel, Marcelo Olarreaga, and George Rapsomanikis and par- ticipants of the Conference on the Poverty Impacts of the Doha Development Agenda in The Hague (December 2­4, 2004) for helpful comments and discussions. Also, the author is indebted to Maros Ivanic for estimating and providing the GTAP results. 107 108 Poverty and the WTO: Impacts of the Doha Development Agenda mapped into the welfare function of the household using household survey data. The empirical approach used in this study measures the impact of Doha imple- mentation by tracing changes in the household prices of goods and factors and their impact on household welfare, taking particular account of the role of domestic price transmission. The findings suggest that multilateral trade liberalization alone would have a negative effect on Mexican households, even though very small. However, when the implementation of the DDA is complemented by domestic policies intended to increase productivity and improve domestic price transmission, the overall effect becomes positive. The results point to the importance of domestic price transmission in determining the variance of the effects across households. Given the existing structure of markets in Mexico, most of the effects of multilateral trade liberalization would be felt in the northern states, which are more connected to international markets. Conversely, households living in the southern states are isolated from most effects--not because of the composition of their consumption or income bundle, but because of the marginal effect of trade reforms on prices in those areas. An alternative scenario explores the impact of complementary poli- cies (for example, improved extension services) that might enable farm house- holds to respond to increased market opportunities without having to incur addi- tional costs. This enhances the welfare outcome, especially for the poorest rural households in the south of the country, when accompanied by reforms designed to increase price transmission within the Mexican economy. Introduction Trade negotiations have recently occupied center stage in multilateral policy dis- cussions. The belief is that international trade, and the reduction of protectionist barriers as a means of increasing it, is a powerful tool to spur economic growth and reduce poverty in developing countries. However, the evidence of the positive effects of international trade reform on poverty in developing countries is frag- mentary.1 In practice, the consensus is that trade policies are only one ingredient in the development recipe, and other policies are generally needed to ensure that trade will enhance welfare for the majority of the poor. Therefore, it is important to investigate the factors that influence the relationship between trade reform and poverty alleviation. This chapter focuses specifically on the role of the marketing system in transmitting price changes from the border to rural and urban house- holds throughout Mexico. Empirical evidence suggests that similar trade reforms are unlikely to produce analogous results across countries, especially when analyzing their effect on poverty. This implies that the analysis of trade reform on social welfare cannot be Multilateral Trade Liberalization and Mexican Households 109 generalized and needs to be conducted on a country-by-country basis. Moreover, even within the same country, geographic areas, households, and individuals are likely to be differentially affected, some of them benefiting more than others, while others might lose. Therefore, it is necessary to analyze the impact of trade policies on poverty using a microeconomic framework so as to identify likely win- ners and possible losers. In particular, the analysis of the distribution of benefits and costs across regions, communities, and individuals is important when think- ing about complementary and compensatory policies. This chapter provides a quantitative estimate of the effect on Mexican house- holds from the implementation of the DDA. The analysis uses a variant of the two-step approach outlined in chapter 3. However, in this case, a national CGE model is not used. Instead, the changes in prices and factors estimated through the GTAP global CGE model (Hertel 1997) are transformed based on the econo- metrically estimated price transmission relations and then mapped directly into the welfare function of the household.2 The contribution of this chapter rests in the translation of the national price changes to the local level. In analyzing the poverty effect of multilateral trade liberalization, this chapter takes into account the changes in factor returns (labor and land) and the cost of the consumption basket and value of income sources of poor households to meas- ure changes in real income and poverty. This study is enriched by the analysis of domestic price transmission to investigate the magnitude of the effect of trade policies at the local level. Simply put, it measures the effect of trade policies on poverty not only on the basis of what the poor produce and consume, but also taking into account the geographic area where this production and consumption take place. To summarize the main results, the findings suggest that multilateral trade lib- eralization alone would have a negative effect on Mexican households, albeit a very small one. However, when the implementation of the DDA is complemented by domestic policies intended to increase productivity and improve domestic price transmission, the overall effect become positive. The results point to the importance of domestic price transmission in determining the variation in impacts across households. Given the existing structure of markets in Mexico, the results indicate that the effects of multilateral trade liberalization would concentrate in the northern states, which are more closely connected to international markets. Conversely, households living in the southern states are largely insulated from these effects-- not because of the composition of their consumption or income bundle, but rather because of the very limited effect of trade reforms on prices in those areas. The remainder of this chapter is organized as follows. Section 1 describes the extent and distribution of poverty in Mexico. First is provided a description of the 110 Poverty and the WTO: Impacts of the Doha Development Agenda extent and distribution of poverty in Mexico, followed by the implication of the DDA for Mexican households. Next, the empirical framework is presented. Finally, the last two sections discuss the results and draw some conclusions. More detail on the household database as well as its reconciliation with, and mapping to, the macroeconomic data (GTAP) is available in the appendix to Nicita (forthcoming). Poverty in Mexico Despite Mexico's status as a middle-income country and a member of the OECD, poverty in Mexico is widespread. Poverty levels moved substantially during the 1990s, decreasing in periods of economic growth and increasing in economic downturns. Extreme poverty3 was estimated to be about 24 percent in the early 1990s. Economic reforms and growth produced a reduction of about 3 percentage points by 1994. The economic crisis of 1995 and the sharp devaluation of the peso then led to a sharp increase in poverty (to 37 percent in 1996 and 34 percent in 1998). Finally, economic recovery in the late 1990s produced the largest decline in poverty, with extreme poverty falling to precrisis levels in 2000 and declining thereafter to about 20 percent in 2002. The incidence of poverty in Mexico varies widely by region. Table 4.1 illustrates the incidence of poverty in five Mexican regions for the year 2000.4 Poverty in Mexico is fundamentally a rural phenomenon. More than half of the households living in rural areas are extremely poor. With the exception of the Federal District of Mexico City, northern states register the lowest incidence of poverty. The states in the central regions and especially the southernmost states register the highest percentage of poor. Although extreme poverty rates are rela- tively low, especially in urban areas, moderate poverty is more widespread. At the national level, more than 50 percent of the population is moderately poor, with Table 4.1. Poverty in Mexico (Headcount) Extreme poverty Moderate poverty Region Total Urban Rural Total Urban Rural Federal district 11.2 11.2 43.3 43.3 U.S. border 15.9 7.5 32.9 51.2 35.4 67.6 North 23.1 16.1 43.2 52.8 41.4 73.3 Center 27.6 16.5 55.7 59.2 49.4 90.0 South 45.4 25.3 78.6 71.3 56.8 93.8 Total 24.2 13.7 58.5 53.7 43.7 83.8 Source: Author's calculations. Multilateral Trade Liberalization and Mexican Households 111 peaks of about 90 percent in rural areas in the central and southern states. Given these premises, it appears that in order to have the greatest effect on poverty, trade policies need to reach the rural poor in the central and southern regions. Exposure of the Poor to International Price Shocks The extent to which international trade policies will result in a decrease in poverty in Mexico depends in particular upon the exposure of poor Mexican households to trade shocks. The easiest way to think about how poor rural households are affected by trade policies is in terms of the farm household, which produces goods and services, sells its labor, and consumes goods and services. In this setup, an increase in the price of something of which the household is a net seller increases its real income, and a decrease reduces it. Therefore, this section examines, in turn, the net sales positions of poor Mexican households, as well as the anticipated price shocks in the wake of Doha. Net Sales Position of the Poor Figure 4.1 summarizes the income sources of Mexican households. Households are categorized as very poor (those below the extreme poverty line), poor (those below the poverty line but above the extreme poverty line), and nonpoor. This fig- ure points to the importance of labor earnings for Mexican households. Labor earnings represent about 50 percent of income for poor households and slightly less than 40 percent for very poor households. Moreover, the very poor are tied to the performance of the agricultural sector because more than half of their income is related to agriculture (own-consumption plus agricultural sales and agricul- tural wages).5 Figures 4.2a and 4.2b present the composition of the expenditure basket of Mexican households. The consumption basket of very poor households is roughly equally divided among own-consumption, food purchases, and purchases of non- food goods and services. A similar consumption basket is found in the case of poor households, which exchange a lower share in autoconsumption with a higher share of other expenses (especially services). Among food purchases (Fig- ure 4.2b), cereals (mainly maize) take about one-fourth of expenditures. Other vegetables take about 20 percent of purchases, and animal-based products account for about 30 percent. Poor households tend to purchase more animal- based products and fewer cereals and vegetables relative to very poor households. In summary, the analysis of income sources and expenditure baskets of poor households reveals that (a) Mexican households rely greatly on labor earnings; (b) the income of very poor households is strongly related to the agricultural sectors; 112 Poverty and the WTO: Impacts of the Doha Development Agenda Figure 4.1. Household Income 40 35 cent) 30 (per 25 share 20 Income 15 10 5 0 Nonpoor Poor Very poor Agricultural wages Agricultural sales Other Income Skilled wages Unskilled wages Autoconsumption Source: Author's calculations. (c) there is a net distinction in the labor earnings of different household groups, with the nonpoor relying mostly on skilled labor income and the poor relying mostly on unskilled labor earnings; and (d) on the consumption side, poor house- holds spend most of their income on food purchases, and among those, most is spent on cereal (maize) and animal-based products (meat, dairy). Given these premises, the effect on poverty of the DDA will depend mostly on its effect on the prices of some key products (namely cereals and meats) and on labor earnings. The next section analyzes the impact of a successful Doha implementation on prices and factors important for poor households in Mexico Doha Implications for Mexico In this chapter, the implications of the DDA for Mexican households and poverty reduction are estimated analyzing four factors:6 1. Impact on prices of goods produced and consumed by Mexican households 2. Impact on the demand for Mexican exports 3. Impact on labor and land earnings in Mexico 4. Extent to which those effects are transmitted to each household Multilateral Trade Liberalization and Mexican Households 113 Figure 4.2a and 4.2b. Household Consumption 40 35 cent) 30 (per 25 share 20 15 10 Consumption 5 0 Nonpoor Poor Very poor Autoconsumption Food Goods Services/other 4040 3535 cent) (per 3030 2525 share 2020 1515 Expenditure 1010 5 0 Nonpoor Poor Very poor Cereal Vegetables Animal-based Other food Source: Author's calculations. 114 Poverty and the WTO: Impacts of the Doha Development Agenda The change in average prices, the return to labor, and export supply for the average Mexican household (items 1­3) are estimated through the GTAP model and are discussed below. Price transmission (item 4) is discussed in section 4.4, "Simulation Results." The change in prices and quantities and returns to labor and land consequent to trade reforms are obtained from the GTAP model, as described in chapter 3. Unlike most of the country studies in this volume, the GTAP results used here are generated by trade reform simulations that include Mexican cuts in tariffs and domestic support. This is because a national CGE model is not introduced in this chapter. Two scenarios are considered, as discussed in chapter 2: the full-liberal- ization scenario, which assumes full tariff removal, removal of all export subsi- dies, and domestic support, and the core Doha scenario. Table 4.2 reports the change in prices and factor returns as estimated for the Mexican economy by the GTAP model for both of these scenarios.7 From these results, it is clear at the DDA is expected to produce only small changes in the prices of goods and factor returns in Mexico.8 The largest effect for the Doha sce- nario is estimated in the return to natural resources, which is expected to increase about by 1.6 percent in real terms. The return to land is expected to increase by 1 percent, and wages (both skilled and unskilled) are expected to decrease mini- mally. Prices, with the exception of oils and fats, are expected to rise by between 0 and 1 percent. More generally, prices are expected to rise only for agricultural products and not for manufacturing. Larger effects are estimated for the full-liberalization scenario. In this scenario, return to land is expected to decrease substantially (by about 16 percent) as domestic support for Mexican agriculture is fully removed. Labor earnings are expected to decline by about 0.1 percent (unskilled) and increase by 0.1 percent (skilled). The effects on prices are more interesting. The price of cereals is expected to rise by almost 15 percent, oils and fats by another 15 per- cent, and the price of dairy products is expected to decline by about 0.6 percent, but little or no effect is found in the price of meat products and sugar. Finally, prices for vegetables and other agricultural products are expected to decline by nearly 2 percent. Smaller changes are estimated for the prices of manufactures, which change between -0.6 percent (household items) and +0.3 percent (food products). In addition to the change in prices, the trade reforms are estimated to result in a change in production. In the case of the Doha scenario, Mexico's aggregate produc- tion is estimated to increase by about US$850 million. Those increases are mostly concentrated in manufacturing and services. In the case of full liberalization, production (and especially exports) is expected to decrease substantially. This is driven by the erosion of Mexico's preferential access to the U.S. market. Multilateral Trade Liberalization and Mexican Households 115 Table 4.2. Scenarios--Doha Implementations and Full Trade Liberalization Change in factor returns Sector Doha (%) Full liberalization (%) Return to land 1.0 ­16.4 Unskilled labor ­0.1 ­0.1 Skilled labor ­0.2 0.1 Capital ­0.2 0.0 Natural resources 1.6 1.1 Doha (%) Full liberalization (%) Change in Change in production production Price (US$ Price (US$ Product group change (%) million) change (%) million) Cereals 0.4 18.9 14.6 ­351.6 Dairy 0.2 ­35.6 ­0.6 ­418.0 Meat products 0.2 135.6 0.1 ­495.8 Oils and fats 3.0 12.3 15.2 ­57.9 Sugar 0.2 ­2.2 0.0 ­27.2 Vegetables 0.6 11.9 ­1.8 80.7 Other agricultural 0.5 61.4 ­1.9 222.3 Food products 0.0 44.4 0.3 ­62.6 Household items 0.0 746.9 ­0.6 147.4 Textiles and apparel 0.0 ­565.1 0.1 ­2,506.0 Other manufacturing 0.0 122.6 ­0.3 ­1,760.9 Other products 0.3 31.0 0.8 81.4 Services ­0.1 241.4 0.6 361.3 Source: Results based on chapter 3 of this book. The next section illustrates the empirical strategy used to measure how the changes in the prices and demand for Mexican products, as well as the return to factors for Mexico, translate into household welfare and ultimately affect poverty. Empirical Framework The approach used here to estimate the effect of trade liberalization on household welfare can be summarized in three steps. First, the effects of the Doha implementa- tion estimated by the GTAP model are translated into local prices (and quantities) using a pass-through model that allows the transmission from border prices to 116 Poverty and the WTO: Impacts of the Doha Development Agenda domestic prices to vary by local markets. Second, the changes in the prices of goods at the local level are used to investigate the movement in earnings and quantities supplied. Last, those changes are mapped to the household survey and fed into the household welfare function using a farm household model to measure the changes in real income. International Prices and Domestic Prices As seen in section 2, a successful Doha implementation would have an effect, albeit small, on the prices of various products important in both the consumption baskets and the income sources of Mexican households. However, it is widely rec- ognized that the international prices of products and their retail prices are only loosely linked, because internal factors such as transportation costs and local sup- ply of substitute products act as filters between the two (Frankel, Parsley, and Wei 2005; Winters, McCulloch, and McKay 2004). The isolation of local markets is particularly evident in rural areas, where marketing infrastructure is poorly devel- oped or altogether missing. Given that domestic price transmission is imperfect, to measure poverty effects of trade reforms, it is necessary first to estimate the magnitude of changes in local retail prices consequent to changes in world prices. In other words, movement in average prices consequential to trade policies (those estimated by the CGE model) need to be translated into changes in retail prices (those faced by the households). The model used here to measure the extent to which local prices vary relative to the international prices follows the approach of Nicita (2004) and is based on the tariff and exchange rate pass-through literature (Goldberg and Knetter 1997; Campa and Goldberg 2002). In the pass-through estimation, all product groups are aggregated into two main categories: agriculture, and manufacturing. Within these two broad cate- gories, all products are assumed to have the same domestic price pass-through coefficient. This model allows changes in prices to be different across the 32 Mex- ican states, which are further differentiated by urban and rural areas. To ensure compatibility with the CGE estimates, the changes in regional prices consequent to movement in the international prices are rescaled while keeping the change faced by the average household equal to the one estimated by the GTAP model.9 The model used to estimate domestic price pass-through is based on the effect of tariff liberalization on domestic prices as they vary with distance from the U.S. bor- der. In this model, the effect of a change in tariff is perceived in local markets in the same way as a movement in the world price, therefore the extent to which domestic prices move in function of movement of the tariff can be interpreted as the degree of correlation between border prices and retail prices. In summary, the model tracks Multilateral Trade Liberalization and Mexican Households 117 the effect of a change in price at the U.S. border (produced by the change in tariff) to changes in the price at the regional level so as to capture how much of the move- ment in the border prices is reflected in each of the retail prices in different geo- graphic areas.10 To capture differences in pass-through across states, the pass- through coefficient is interacted with the distance variable.11 This interaction term is further interacted with a rural and an urban dummy to investigate possible differ- ences in pass-through between urban and rural areas.12 Referring to Nicita (2004) for a detailed explanation of the model, the estimating equation is given by: (4.1) where Xgtis the primary control variable (the international price of good g expressed in domestic currency), and Zgtr is a vector of control variables that includes local supply and regional income, R and U denote rural and urban dummies, and gtris an error term. The coefficients of interest are , which represents the tariff pass-through elasticity, and 1 and 2, which are its adjustment for distance from the U.S. border. The pass-through is "full" or "complete" if = 1 and the pass-through is "incom- plete" if = <1. Similarly, the effect of the pass-through will be identical in all urban areas if 1 = 0. However, if local prices vary as a consequence of movement in the tariff, then 0. Similar reasoning is applied in the case of rural areas, where the coefficient of interest is 2. The econometric estimation of equation (1) combines a time series of cross- sectional data set into a pseudo panel.13 The data consist of domestic prices for 63 regions and six time periods. Average prices for each region are arranged into a panel dataset, and the estimation is performed separately for agriculture and for manufacturing. Table 4.3 reports the results of the pass-through model, which indicate a pass-through between the international price and the border price of about 26 percent for agriculture and 67 percent for manufacturing. The negative sign on the interacted terms indicates that, as the distance from the U.S. border increases, price pass-through coefficients decline, suggesting the possibility of missing markets. Moreover, changes in prices may be internalized by intermediaries or absorbed by trade costs. Therefore, retail prices in the states closer to the U.S. market tend to better "feel" the effect of movement in the tariff. Conversely, southern states seem to be the least connected to the international markets. Another result is the difference between urban and rural areas. Urban areas in all regions "feel" the movement in the tariff to a larger extent, especially in the case of agricultural products. Finally, movement in the tariffs of agricultural products tends to be reflected to a lesser extent in domestic prices relative to man- ufacturing products (especially in rural areas). This is not surprising and is likely 118 Poverty and the WTO: Impacts of the Doha Development Agenda Table 4.3. Pass-Through Variable Agriculture Manufacturing International price 1.449 *** (0.165) 0.004 (0.007) Regional consumer price index 0.284 * (0.149) 1.174 *** (0.247) Local supply ­0.036 *** (0.011) ­0.016 (0.017) Urban or rural 0.131 *** (0.043) 0.510 *** (0.064) Distance 0.002 (0.012) ­0.030 * (0.016) Tariff pass­through 0.260 * (0.155) 0.671 *** (0.101) Urban transmission 0.003 (0.033) ­0.091 *** (0.034) Rural transmission ­0.054 ** (0.027) ­0.108 *** (0.027) Constant 9.498 *** (1.021) 5.201 *** (0.623) Observations 378 378 R2 0.58 0.64 Source: Author's calculations. Note: All variables, except distance, are in log. White corrected standard errors are shown in parentheses. Significance level of 1 percent, 5 percent, and 10 percent are indicated by ***, **, and *, respectively. driven by a greater presence of domestic substitutes and stronger consumer pref- erence for domestically produced varieties. Production and Export Supply A successful Doha implementation is estimated to produce an increase in overall production of about US$850 million per year, mostly driven by increases in inter- national demand for Mexican products. It is important to note that an increase in demand for Mexico's exports will not necessarily have a substantial effect on poverty. The reason is twofold. First, poor households may not be directly employed in producing (and marketing) products for which there are increases in export demand. Second, there is a cost associated with the increase in supply, with net gains likely to be much smaller than the change in production volume. The increase in sales can be decomposed into the quantity effect (the actual value of the increase in production) and the price effect (the increase in value of this quantity due to the higher price). The base case simulation of the Doha Multilateral Trade Liberalization and Mexican Households 119 scenario assumes that there are real costs associated with the increase in produc- tion required to meet increased agricultural demands in the wake of policy reform. Therefore, the net gains to households originate only from the increase in the prices, now applied to the increased production. A second assumption is required to allocate the increase in production to individual households. This is assumed to be proportional to the marketed production of households, and it is also assumed to follow the price pass-through mechanism, with weaker effects in the more remote rural areas. This implies that households producing only for auto-consumption will not be allowed to increase production and households that will not observe any price signals will not adjust their production to fill the increase in demand. Labor Earnings The link between trade reforms and labor earnings goes through the price mecha- nism. International trade reforms operate through changes in prices, and changes in prices will consequentially affect labor earnings. In estimating the impact on wages of Doha implementation, this chapter makes the assumption that move- ments in wages are directly affected by movements in prices. A more sophisticated approach would require the estimation of price-wage elasticities for different prod- ucts and different types of labor. However, this would require additional data and would make the analysis more cumbersome while adding little to the overall analy- sis. Moreover, labor markets in developing countries are seldom integrated, and empirical evidence suggests that returns to labor vary greatly across different geo- graphic areas (Hanson 1997, 2003), calling for a model that allows wage response to vary across geographic areas. The GTAP model estimates an average change in wages (skilled and unskilled) across scenarios that falls between -0.2 and 0.1 per- cent. Given these small changes, and for the sake of simplicity, wages are assumed to follow the price pass-through mechanism on a regional level. Arguably, this is a reasonable assumption that implies that wages are assumed to move more in regions where price pass-through is greater relative to regions where price pass- through is smaller. As in the case of the prices of goods, the movement of the aver- age wage is kept at the level estimated by the CGE model. Changes in Household Welfare Having illustrated the channels used to investigate the effect on households result- ing from the implementation of Doha, it is now possible to calculate changes in household welfare.14 In developing countries, most households are simultane- ously consumers and producers of goods and services. Therefore, in analyzing the 120 Poverty and the WTO: Impacts of the Doha Development Agenda effect on household welfare from any policy, it is important to recognize this dual role of the household.15 The farm household model fits this purpose (Singh, Squire, and Strauss 1986). The approach used here to measure the change in real income (dyh) can be expressed as follows: (4.2) where dPh are the changes in prices of good g faced by households h, h is the g g share of income obtained from the sale of good g by household h, h is the share of income obtained in the labor market, h is the share of the consumption basket g devoted to good g; yh is the income of the household16, and (4.3) where Prodg is the total change in the production of good g, and yh is the income g originating from the sale of good g by household h.17 In this setup, equation (2) suggests that a change in the price of good g favors or harms the household based on the net exposure of its budget to that particular good. Moreover, an increase in the international demand for a particular good favors households in proportion to their marketed production of the good, and movement in wages affects households relative to their share of wage income. Finally, the change in welfare is distributed across household members, expendi- tures are determined by the new level of income, and new welfare indicators are calculated at the new level of consumption. Simulation Results The first scenario examined in this section looks at the effects of Doha and revolves around the status quo in which price transmission is kept at the estimated level and increases in farm output are costly. The second scenario builds on the first but mimics an improvement in the Mexican economy, assuming that any increase in agricultural production and exports is achieved at no additional cost to produc- ers.18 This could be due to an increase in productivity, or it could be a consequence of the household using surplus labor to achieve the increased production. The third scenario builds on the second and adds the assumption that domestic price transmission is improved by half.19 Finally, a fourth scenario measures the results of full international trade liberalization on Mexican households, while still assum- ing the status quo in the domestic economy (no complementary reforms). Multilateral Trade Liberalization and Mexican Households 121 The change in real income is used as the welfare indicator for each scenario and household group. Results are differentiated by region and presented for three household groupings: all households, all poor (those living below the asset poverty line), and very poor (those living below the food poverty line). Doha Scenario The results of the Doha implementation based on the absence of domestic reforms suggest that these trade reforms would have a small negative impact on overall real income in Mexico. Table 4.4 reports the change in real income for the total population, the poor, and the very poor, further differentiated by region and urban and rural areas. The only exceptions to the negative impacts are the positive effects for the very poor in the northern and U.S. border regions. However, the effects, both positive and negative, are in all cases within 0.3 percent of change in real income. Complementary Reform Scenario 1 ("Doha Plus") The results of the Doha implementation in the presence of facilitating increases in productivity (or the use of surplus labor) are reported in table 4.5. The results from this scenario, although small, show a positive effect from Doha implementa- tion. On average, Doha is expected to raise real income in Mexico by about 0.4 Table 4.4. Change in Real Income (Doha) (percentage) Urban Rural All Urban Rural Very very very Region Total Urban Rural poor poor poor poor poor poor Federal District ­0.1 ­0.1 ­0.1 ­0.1 0.0 0.0 Border ­0.3 ­0.3 0.1 ­0.1 ­0.2 0.1 0.2 0.1 0.3 North ­0.1 ­0.1 0.1 0.0 ­0.1 0.1 0.2 0.3 0.2 Center ­0.1 ­0.2 0.0 ­0.1 ­0.2 0.0 0.0 ­0.1 0.0 South ­0.1 ­0.2 0.0 ­0.1 ­0.2 0.0 0.0 ­0.2 0.0 National ­0.1 ­0.2 0.0 ­0.1 ­0.2 0.0 0.0 ­0.1 0.0 Source: Author's simulations. 122 Poverty and the WTO: Impacts of the Doha Development Agenda Table 4.5. Change in Real Income ("Doha Plus") (percentage) Urban Rural All Urban Rural Very very very Region Total Urban Rural poor poor poor poor poor poor Federal District 0.1 0.1 0.3 0.3 1.2 1.4 Border 0.6 0.4 1.5 1.0 0.8 1.6 2.0 1.8 2.4 North 0.9 0.7 1.2 1.1 0.9 1.2 1.8 2.5 1.5 Center 0.3 0.2 0.5 0.4 0.3 0.5 0.6 0.7 0.6 South 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.2 0.0 National 0.4 0.2 0.6 0.5 0.4 0.5 0.7 0.9 0.6 Source: Author's simulations. percentage points. However, the poor, and in particular the very poor, gain sub- stantially more, especially in the northern and U.S. border regions. Average gains are about 0.7 percent for the very poor and 0.5 percent for the poor as a whole. Urban areas are expected to gain less relative to rural areas because of the smaller share of agricultural production in total income in the broadly defined urban areas. However, the poor in the remote southern states still gain little from this trade reform. Complementary Reform Scenario 2 ("Doha Plus-Plus") The results of the Doha implementation in the presence of both increases in pro- ductivity and improved domestic price transmission are reported in table 4.6. The results show the role of domestic price transmission in distributing the effects of Doha implementation across income groups and regions. Improved domestic price transmission is expected to redistribute gains from the northern regions of the country to the south and at the same time from the non-poor to the poor. This scenario estimates a change in the real income of the poor and very poor of 0.6 and 1.1 percent, which when compared to the Doha-plus scenario, translate into an additional increase of 0.1 and 0.4 percent, respectively. This reflects the fact that poor households are generally more remotely located and therefore experi- ence fewer of the gains from increased trade opportunities because of incomplete markets. Overall, the gains from this scenario are more uniformly distributed across regions. Multilateral Trade Liberalization and Mexican Households 123 Table 4.6. Change in Real Income ("Doha Plus-Plus") (percentage) Urban Rural All Urban Rural Very very very Region Total Urban Rural poor poor poor poor poor poor Federal District 0.2 0.2 0.4 0.4 1.5 1.4 Border 0.3 0.2 1.1 0.7 0.5 1.1 1.4 1.1 1.9 North 0.7 0.5 1.1 0.9 0.7 1.1 1.4 1.9 1.3 Center 0.5 0.3 0.9 0.7 0.4 1.0 1.1 0.9 1.2 South 0.4 0.2 0.7 0.5 0.3 0.7 0.8 0.6 0.9 National 0.4 0.2 0.9 0.6 0.4 0.9 1.1 1.0 1.1 Source: Author's simulations. Based on the results from this Doha-plus-plus scenario, the change in real income at the household level is regressed on household characteristics so as to better investigate the variance of the gains. Clearly, these variables are not purely exogenous to the welfare gains; however, this regression is of descriptive interest in helping to isolate covariates of relevance. This information may prove useful when thinking about compensatory policies. Results of the regression and a sum- mary of descriptive statistics are presented in table 4.7. The share of variance in gains that is accountable to those covariates is about 21 percent. The regression results give a rough indication of how the gains are distributed.20 As seen in table 4.5, the results suggest that urban areas gain relatively less from the Doha imple- mentation. Moreover, other things being equal, the geographic distribution of gains suggests that northern regions and regions closer to the U.S. border gain substantially more, and southern regions gain the least. The coefficients on house- hold characteristics suggest that larger families gain less than smaller ones. Simi- larly, lower gains are estimated for female-headed households as well as house- holds where the household head is relatively less educated. Finally, the coefficients on the decile variables suggest that low-income households obtain the largest gains in percentage terms. Full Liberalization The results from a hypothetical, multilateral full-liberalization scenario are reported in table 4.8. Full multilateral trade liberalization is expected to produce a 124 Table 4.7. Improved Scenario: Variance of the Gains Regression results (Dependent variable: percentage gains) Descriptive statistics Poverty Household Standard Household Standard characteristic Coefficient error characteristic Mean deviation and Gender (1 = female) 0.0001 (0.0004) Gender (1 = female) 0.18 0.39 Age 0.0001 *** (0.0000) Age 46.31 15.38 the Share child 0.0006 *** (0.0001) Share child 1.49 1.50 Share elderly 0.0006 * (0.0003) Share elderly 0.30 0.60 WTO: Household size ­0.0049 *** (0.0004) Household size 1.30 0.54 Urban dummy ­0.0050 *** (0.0004) Urban dummy 0.77 Binary Impacts Region Mexico City (dropped) 0.23 Binary U.S. border 0.0015 *** (0.0004) U.S. border 0.16 Binary North 0.0031 *** (0.0005) North 0.10 Binary Center 0.0010 *** (0.0004) Center 0.39 Binary of South ­0.0006 (0.0005) South 0.13 Binary the Education No education (dropped) 0.15 Binary Primary ­0.0009 ** (0.0004) Primary 0.43 Binary Doha Middle ­0.0013 *** (0.0005) Middle 0.20 Binary Secondary ­0.0024 *** (0.0006) Secondary 0.21 Binary College ­0.0033 ** (0.0013) College 0.01 Binary Development Income deciles Income decile 1 (dropped) 0.10 Binary Income decile 2 ­0.0022 *** (0.0006) Income decile 2 0.10 Binary Income decile 3 ­0.0037 *** (0.0006) Income decile 3 0.10 Binary Income decile 4 ­0.0032 *** (0.0006) Income decile 4 0.10 Binary Income decile 5 ­0.0037 *** (0.0007) Income decile 5 0.10 Binary Income decile 6 ­0.0037 *** (0.0007) Income decile 6 0.10 Binary Agenda Income decile 7 ­0.0034 *** (0.0007) Income decile 7 0.10 Binary Income decile 8 ­0.0026 *** (0.0007) Income decile 8 0.10 Binary Income decile 9 ­0.0023 *** (0.0007) Income decile 9 0.10 Binary Income decile 10 ­0.0042 *** (0.0008) Income decile 10 0.10 Binary Constant 0.0125 *** (0.0010) Observations 10108 R2 0.21 Source: Author's estimates. Note: All variables, except distance, are in log. White corrected standard errors are shown in brackets. Significance levels of 1 percent, 5 percent, and 10 percent are indicated by ***, **, and *, respectively. Multilateral Trade Liberalization and Mexican Households 125 Table 4.8. Change in Real Income ("Doha Plus-Plus") (percentage) Urban Rural All Urban Rural Very very very Region Total Urban Rural poor poor poor poor poor poor Federal District ­0.8 ­0.8 n.a. ­1.1 ­1.1 n.a. ­1.0 ­2.1 n.a. Border ­1.4 ­1.3 ­2.2 ­2.1 ­2.0 ­2.5 ­3.3 ­3.7 ­2.8 North ­1.7 ­1.8 ­1.5 ­2.1 ­2.6 ­1.5 ­2.5 ­4.2 ­1.7 Center ­0.9 ­1.2 ­0.2 ­1.1 ­1.9 ­0.2 ­0.9 ­2.5 ­0.2 South ­0.7 ­1.2 0.0 ­0.8 ­1.9 0.0 ­0.5 ­2.5 0.0 National ­1.0 ­1.2 ­0.5 ­1.3 ­1.9 ­0.4 ­1.0 ­2.7 ­0.3 Source: Author's simulations. Note: n.a. = not applicable. negative impact on Mexican households. Losses are on the order of about 1 per- cent, with peaks of about 4 percent for the very poor living in urban areas in the northern regions. The negative outcome of this scenario is the result of the increase in the prices of consumption goods for Mexican households and the reduction in exports driven by the sharp erosion of Mexican preferences in the U.S. market. Summary The results that emerge from the four scenarios can be summarized as follows: (a) Doha alone, without any complementary reform, is likely to have a negative, albeit very small, impact on Mexican households; (b) Doha implementation with com- plementary reforms aimed at increasing productivity is expected to produce posi- tive small gains for Mexican households; (c) without improvement of domestic price transmission, the effects of the Doha implementation are expected to be concentrated in the northern regions; and (d) improvement in the domestic price transmission results in a more uniform distribution of the effects and larger ben- efits for the poorest households. Most important, the findings suggest that the variance of the gains largely depends on domestic price transmission. When price transmission is assumed to stay at the estimated level, households living in the southern regions (especially rural regions) are largely unaffected by Doha, either positively or negatively. This happens because price signals originating from the Doha-induced movement in 126 Poverty and the WTO: Impacts of the Doha Development Agenda world prices subsequent to trade reforms are perceived only marginally in those areas. When domestic price transmission is assumed to improve, the distribution of the gains is more uniform across geographic areas and households. What emerges from the analysis of the simulation exercise is that without com- plementary reforms, Mexico is not expected to gain from Doha and is expected to lose in the case of full multilateral trade liberalization. This outcome is not sur- prising, considering that Mexico has already liberalized trade with its most impor- tant trading partner through its membership in NAFTA. Thus, multilateral tariff reductions in the United States result in an erosion of those preferences currently enjoyed by Mexican exports. Conclusions This chapter provides a quantitative estimate of the effect on Mexican households from the implementation of the DDA, taking into account the role of domestic price transmission. The findings suggest that multilateral trade liberalization is likely not beneficial for most Mexican households unless it is complemented by domestic reforms aimed at facilitating the response of households to these new market opportunities. The findings suggest that the poor would likely share in the benefits (and costs) of such trade reforms. The differences in impacts across households are more closely associated with geographic areas of residence than income level. The results point to the importance of domestic price transmission in determining the variance of the effects across households. Given the existing structure of markets in Mexico, most of the effects of multilateral trade policies would be felt in the northern states, which are more connected to international markets. Conversely, households living in the southern states are isolated from most effects, not because of the composition of their consumption or income bundle, but because of the marginal effect of trade reforms on prices in those areas. Measures aimed at enhancing domestic price transmission could ensure that any gains from trade reform (when accompanied by productivity-enhancing policies) would be more evenly spread throughout the country. Notes 1. A review of the findings of the literature is given by Winters, McCulloch, and McKay (2004); Hertel and Reimer (2004); and Berg and Kruger (2003). 2. The economics involved in this approach is generally well known and has found numerous applications. See Hertel and Reimer (2004) for a review. 3. This corresponds to the food poverty line. The food poverty line is defined as the minimum expenditure necessary to guarantee a 2,200-calorie intake per day. Multilateral Trade Liberalization and Mexican Households 127 4. The extreme poverty line corresponds to the food poverty line. The moderate poverty line is the asset poverty line. The difference corresponds to nonfood components in the consumption bundle. Both poverty lines are calculated by CEPAL (2001). 5. Other sources of income include government transfers (about 4 percent of total income) and remittance (about 3 percent of total income). Given their low weight on the overall income source of poor households, income from these categories is assumed to be fixed and therefore not affected by trade policies. 6. The GTAP model was estimated keeping the impact of trade reforms on government revenues neutral (for example, compensated by internal taxation). It is also assumed that transfer payments are indexed such that they do not play a role in the welfare calculation. 7. Within the model, the impact of tariff changes on government revenues and redistribution is kept neutral, adjusting prices by the effect of compensating changes in direct income taxation. 8. The small changes are found because trade is already largely free within the North American Free Trade Agreement (NAFTA) countries. 9. The GTAP Armington specification produces average market prices already adjusted for imper- fect price pass-through. 10. Prices are corrected for quality issues following the standard methodology of Prais and Houthakker (1955). 11. The variable used in the model to capture differences in the movement in the retail prices is the driving distance from the U.S. border. Distance is measured in 1,000 kilometers. 12. Higher transport costs and local supply suggest a lower price pass-through in rural areas, espe- cially on agricultural products. 13. From a time series of six cross-section surveys (from 1989 to 2000), synthetic cohorts are defined as groups with fixed membership, whose individuals (or households) can be identified as they show up in the surveys. For this reason, groups are defined according to some time-invariant variables. Means within each cohort are calculated and followed for each temporal unit under examination: this cohort aggregation is defined as a pseudo panel. 14. The change in household welfare is calculated by taking into account only first-order effects. 15. For details and applications, see Deaton (1997). 16. Income is equated to expenditures. 17. Prices are different by region. The subscript r for region is omitted. 18. This implies that the value of the increase in exports is directly transferred to the income of the household through agricultural production. 19. That is, the coefficients in the interaction terms are divided by two. 20. The constant represents low-education households in the Federal District of Mexico City in the first income decile. References Berg, A., and A. Kruger. 2003. "Trade, Growth and Poverty: A Selective Survey." IMF Working Paper WP/03/30, International Monetary Fund, Washington, DC. Campa, J. M., and L. S. Goldberg. 2002 "Exchange Rate Pass-Through into Import Prices: A Macro or Micro Phenomenon?" NBER Working Paper 8934, National Bureau of Economic Research, Cam- bridge, MA. CEPAL (Comisión Económica par América Latina y el Caribe). 2001. Panorama social de América Latina 2000­2001. Santiago de Chile: United Nations. Deaton, A. 1997. The Analysis of Household Surveys. A Microeconometric Approach to Development Pol- icy. Baltimore and Washington DC: Johns Hopkins University Press for the World Bank. Frankel, J., D. Parsley, and S. J. Wei. 2005. "Slow Passthrough around the World: A New Import for Developing Countries?" NBER Working Paper 11199, National Bureau of Economic Research, Cambridge, MA. 128 Poverty and the WTO: Impacts of the Doha Development Agenda Goldberg, P., and M. Knetter. 1997. "Good Prices and Exchange Rates: What Have We Learned?" Jour- nal of Economic Literature 35: 1243­72. Hanson, G. 1997."Increasing Returns, Trade and the Regional Structure of Wages." The Economic Jour- nal 440: 113­33. ------. 2003. "What Happened to Wages in Mexico since NAFTA?" NBER Working Paper 9563, National Bureau of Economic Research, Cambridge, MA. Hertel, T. W. 1997. Global Trade Analysis: Modeling and Applications. New York: Cambridge University Press. Hertel, T. W., and J. J. Reimer. 2004."Predicting the Poverty Impacts of Trade Reform." Policy Research Working Paper 3444, World Bank, Washington, DC. Nicita, A. 2004."Who Benefited from Trade Liberalization in Mexico? Measuring the Effects on House- hold Welfare." Policy Research Working Paper 3265, World Bank, Washington, DC. Nicita, A. Forthcoming. "Multilateral Trade Liberalization and Mexican Households: The Effect of the DDA." Policy Research Working Paper, World Bank, Washington, DC. Singh, I., L. Squire, and J. Strauss. 1986. Agricultural Household Models, Extensions, Applications and Policy. Baltimore: World Bank and the Johns Hopkins University Press. Prais, S. J., and H. S. Houthakker. 1955. The Analysis of Family Budgets. Cambridge: Cambridge Uni- versity Press. Winters, L. A., N. McCulloch, and A. McKay. 2004. "Trade Liberalization and Poverty: The Evidence So Far." Journal of Economic Literature. 42: 72­115. 5 The Doha Trade Round and Mozambique Channing Arndt Summary This chapter considers the potential implications of the Doha Development Agenda (DDA), as well as other trade liberalization scenarios, for Mozambique. An applied general equilibrium (AGE) model, which accounts for high marketing margins and home consumption in the Mozambique economy, is linked to results from the Global Trade Analysis Project (GTAP) model of global trade. In addition, a microsimulation module is used to consider the subsequent implications of trade liberalization for poverty. The implications of trade liberalization, particularly the Doha scenarios, are found to be relatively small. Presuming that a more liberal trading regime will positively influence growth in Mozambique, an opportunity exists to put in place such a regime without imposing significant adjustment costs. Introduction The Doha Round of trade negotiations seeks explicitly to involve developing countries. In terms of process, developing countries are expected, as a group, to be much more engaged in the actual negotiations. Wealthier nations, on their side, are expected to place greater emphasis on the implications of any agreement for the developing countries, particularly for poverty. The hope is to reach an agreement that enhances opportunities for developing countries to achieve poverty-reducing economic growth through stronger trade linkages with the world economy. 129 130 Poverty and the WTO: Impacts of the Doha Development Agenda As the region with the highest rate of poverty and relatively weak linkages to the global economy, it seems logical to carefully consider the role of Africa within the DDA. The African continent is both enormous and highly diverse. As a result, implications of any given global trade agreement will differ across economies on the continent. This chapter considers the potential implications of trade liberalization scenarios for the case of Mozambique. Like all African economies, Mozambique has distinguishing features that render it unique. However, as will be discussed, it also shares many structural features with other African countries. The logic of some of the ideas developed here can therefore be applied to a number of other countries across the continent. Section 1 presents a brief description of Mozambique. Section 2 considers implications of various trade liberalization scenarios derived from an analysis that marries outputs from the GTAP model of global trade with a more detailed country computable general equilibrium (CGE) model of Mozambique. Poverty analysis proceeds using a separate household microsimulation module. Section 3 provides a critique of the main results that come out of the models. Section 4 concludes. The transmission of prices to low-income households is a theme that is developed in particular detail throughout the chapter. Mozambique Mozambique is located along the southeastern coast of Africa. In terms of total area, coastline, and shape, it is roughly similar to (a mirror image of) the combined areas of the states of California, Oregon, and Washington that make up the western coast of the United States. Exploitation of natural resources underpins a substantial share of economic activity. Fisheries are a major export industry. The stock of arable land is large, and much arable land remains unexploited. Important agricultural exports include cotton and tobacco. Forestry is also important. With its long coastline and abundance of natural harbors, Mozambique provides port and transport services to neighboring countries. Exploitation of natural gas, uranium, titanium, and other mineral resources has also begun. Finally, Mozambique's natural beauty, particularly its beaches and coral reefs, attracts tourists. These favorable attributes are spread out over a relatively small population of not quite 19 million people. Nevertheless, more than half of the population is categorized as absolutely poor. This implies that slightly more than one person in two has difficulty in meeting very basic survival needs in terms of calorie consumption and basic nonfood necessities such as housing and clothing. This pervasive poverty is the result of a complex historical legacy that included weak human capital development over the colonial period even by African standards, failed socialist policies initiated shortly after independence in 1975, and The Doha Trade Round and Mozambique 131 finally, a brutal civil war that endured for more than a decade. The cessation of hostilities in 1992 coincided with one of the worst droughts on record. The cumulative effect of these disasters earned Mozambique the unwanted label of "poorest country in the world" in the early 1990s. Since then, the economic record has been considerably more positive. From a low base, economic growth has averaged in the range of 7-8 percent per year for more than a decade. This growth coincided with the implementation of a fairly standard structural adjustment program. Very considerable flows of external assistance clearly helped to fuel growth and provided major funding for social initiatives, with particularly large investments in basic health and education.1 By most objective indicators, living conditions for the Mozambican population have improved considerably. In 1996­97, using real consumption as a metric, about 69 percent of the population was characterized as absolutely poor. By 2002­03, this number had fallen by 15 percentage points to about 54 percent, using the same metric. Indicators such as crop production, asset ownership, income of rural households, school enrollments, infant mortality, and child vaccination coverage rates also showed improvements (Mozambique Ministry of Planning and Finance, International Food Policy Research Institute, and Purdue University 2004). Because initial development levels were so low, a decade-plus of rapid growth and rapid improvement in many social indicators have placed Mozambique near Sub-Saharan African averages for a range of indicators. In short, the trends are positive, but the absolute levels of such indicators remain dismal. The clear challenge is to maintain the positive momentum developed over the past decade. Over the coming decade, international trade will likely play a prominent role if growth is to continue. Growth in the past decade has been driven in large measure by internal reconstruction needs (usually donor-funded) and production of basic goods and services that often have been consumed at very local levels, frequently within the household where they are produced.2 These sources of demand are likely to continue to be important, at least through the medium term, but there is also a clear need to strengthen links to international markets, particularly with respect to exports of labor-intensive products. This thumbnail sketch illustrates many aspects of Mozambique that are unique on the African continent. However, Mozambique also shares many essential structural features that are quite common. A nonexhaustive list includes: · A predominantly rural population with economic and social indicators typi- cally at less favorable levels in rural areas--hence, the large majority of the poor reside in rural areas, making improvements in the well-being of current rural dwellers practically a condition sine qua non of any significant reduction in overall poverty levels. 132 Poverty and the WTO: Impacts of the Doha Development Agenda · An overwhelming dependence on agriculture in rural areas. · Large distances and poor transport infrastructure result in substantial transport costs, particularly between distant regions. These weaken or even sever entirely market linkages across disparate regions of the country. For example, the cost of transporting maize by truck from growing regions in the north to the capital city, located in the far south, is so high as to be effectively prohibitive. The chapter now turns to an assessment of the implications of the DDA derived from a formal applied general equilibrium (AGE) model of Mozambique that is linked to outputs from the GTAP model of global trade. Modeling the Implications of Doha The goal of trade liberalization is to redirect productive resources to areas of comparative advantage. At the global level, this implies that production patterns will shift across countries. Within countries, some industries are likely to contract, thereby freeing productive resources that, at least in principle, might allow other industries to expand. Typically, after trade liberalization, one expects productive patterns within individual countries to concentrate in particular industries that have a comparative advantage. Surplus production is sold on global markets, and the resulting income permits countries to import products that were formerly produced at home. Because the goal of trade liberalization frequently involves the reallocation of resources across productive sectors, CGE models have come to be the workhorses for analyses of trade agreements. The global CGE model (the GTAP model) used to analyze the implications of various Doha scenarios at the global level has been well-described in other chapters of this book. This chapter focuses on the Mozambique model, including the microsimulation module for poverty analysis. The first subsection below provides a description of the basic features of the Mozambique model. The second subsection discusses structural features of the economy that can be expected to drive model results. The third subsection presents salient model results. The Mozambique CGE Model Analysis begins from a standard, trade-focused CGE model, which contains three basic elements: (a) specification of economic behavior for firms and households, (b) operation of markets, and (c) macroeconomic closure.3 Novel features particularly relevant for this analysis are then discussed. The Doha Trade Round and Mozambique 133 Behavioral Specification. The model assumes profit maximization by pro- ducers under a sectoral constant elasticity of substitution (CES) technology. Consumers are assumed to demand commodities according to a linear expendi- ture system (LES) utility function formulation. Investment and government expenditures are allocated in a Leontief fashion, with fixed real coefficients rather than fixed expenditure shares. Foreign trade is specified using the Armington assumption. There are CES functions for sectoral imports. Armington import elasticities are taken from Hertel and others (2004). A constant elasticity of transformation (CET) function is used on the export side. However, to remain consistent with the GTAP model, the sectoral export transformation elasticities were set to a high value (5). Also, a downward-sloping demand function for Mozambican exports was developed, again using elasticities from Hertel and others (2004). The presence of these downward-sloping demand functions permits the Mozambique country model to simulate both the world price changes and the shifts in demand generated by the GTAP model under various global trade liberalization scenarios.4 Operation of Markets. A CGE model simulates the operation of product and factor markets, solving for market-clearing prices and wages. It is a closed general equilibrium system, incorporating all elements of the circular flow of income and expenditure and the corresponding real flows. Characteristic features of this type of model include (a) households must respect their budget constraint; (b) the domestic price of imports equals the cost, insurance, and freight (CIF) price mul- tiplied by the exchange rate and the prevailing tariff rate plus any marketing mar- gins or additional domestic sales taxes; (c) the value of imports cannot exceed the availability of foreign exchange; (d) supply of commodities must equal demand for commodities (with inventory accumulation counted as demand); (e) firms collectively cannot use more of any factor than the total availability in the econ- omy; (f) investment must be financed via foreign or domestic savings; and (g) government consumption must be financed through tax revenue, foreign grants (aid), or borrowing on domestic or foreign markets. Also, in this model, aggregate employment of all factors of production is exogenous, and factor returns adjust to clear factor markets. Finally, the model numeraire is the consumer price index (CPI), so all price changes reported in this chapter are relative to the CPI. Macro Closure. All CGE models incorporate macro balances. How equilib- rium is achieved between savings and investment, the government deficit, and the trade deficit constitutes the macro closure of the model. In the Mozambican model, aggregate investment is determined by savings (private plus government plus foreign), so the model is savings driven. Private savings are endogenous, 134 Poverty and the WTO: Impacts of the Doha Development Agenda depending on fixed savings rates by households and enterprises. Government expenditure is set as a fixed share of aggregate absorption in the economy, and the government deficit is exogenous. Direct tax rates across institutions (households and enterprises) vary in order to maintain a constant deficit. Foreign savings and aid are fixed exogenously, and the real exchange rate adjusts to achieve external balance through changes in aggregate exports and imports. More Novel Features. Importantly for this analysis, the CGE model specifi- cally accounts for the substantial costs required for products to reach commercial markets. This is particularly important in the case of agricultural products. These marketing margins reflect storage and transportation costs, as well as risk associ- ated with trading activities and limited opportunities for diversification. Market- ing margins are introduced into the static CGE model by assuming that each unit of a given production good requires a fixed amount of marketing services to reach the market. Because the current model framework treats imported and exported goods as inherently different from domestically consumed production, marketing margins related to exports, imports, and domestic goods are accounted for sepa- rately. A single production activity provides the commercial services associated with the marketing of commodities. Transaction costs vary across sectors. They are zero in the case of service sectors, by definition, and they are nonzero (and sometimes quite large) in other goods sectors--particularly agricultural sectors, where products are bulky and distances between points of production and consumption can be large. Almost all Mozambican households have some money income, either from goods sales or from factor remunerations. This income is used for purchases of essential goods that cannot be produced by the households themselves. Nevertheless, the possibility of home consumption enables households to bypass the market in so far as they can produce consumption goods themselves. The presence of high marketing margins implies the existence of significant differences between farm gate (and factory gate) sales prices, on the one hand, and prices in the commercial markets, on the other hand. Rather than sell at a low price and purchase at a high price, households--particularly rural agricultural households--often opt to consume at least some of what they produce. In some cases, these marketing margins are so large as to isolate the household from commercial markets altogether. Therefore, explicit modeling of the interaction between marketing costs and home consumption becomes essential for assessing important aspects of the economy. All home-consumed commodities and market consumption of all commodities are captured in the LES formulation mentioned above. Appropriate modeling of home consumption and marketing margins has been shown to be important (Arndt and others 2000). The Doha Trade Round and Mozambique 135 The Mozambique Microsimulation Model A microsimulation model in the spirit of Chen and Ravallion (2004) is developed to examine the poverty implications of the trade liberalization scenarios analyzed. The model relies upon data from the 2002­3 Mozambican National Household Budget Survey (INE 2004). The survey provides detailed information on consumption patterns for a nationally representative sample of 8,700 households. The survey also provides detail on household members, including sector of economic activity and education level. The analysis examines the first-order implications of the changes in commodity prices and factor prices generated by the Mozambican CGE model for each of the 8,700 households in the sample. Specifically, changes in commodity prices are multiplied by individual household consumption shares, and changes in factor prices are multiplied by the corresponding share of earnings from each factor in total household income. The factor price effect less the commodity price effect yields a money metric indicator of the first-order change in utility due to the trade reforms for each household. Importantly, in first order analysis, the net effect of price changes for commodities that are home produced and consumed is zero because commodity price changes are exactly offset by gains or losses in factor income. This tends to blunt the impact of trade policy reform on rural households. As mentioned above and detailed in the next section, home consumption is very important in the Mozambican context. In addition, the overwhelming predominance of informal activities implies that wage information is scarce. As a result, earnings by labor category are inferred from educational attainment data combined with econometric estimates of returns to education (Maximiano 2005). Similarly, for the large majority of households, it is practically impossible to separate overall household earnings into labor and capital components. This is less of an issue for poor households because the large majority of earnings can reasonably be assumed to be derived from labor income. In the microsimulation model, 5 percent of total income is assumed to come from capital earnings for households living at less than twice the absolute poverty line. Structure of the Mozambican Economy Tables 5.1, 5.2, and 5.3 provide an overview of the structure of the Mozambican economy. Table 5.1 reports the macroeconomic aggregates. For a very poor country, Mozambique allocates fairly substantial resources to government consumption and government investment. The relatively high level of government expenditure is enabled by substantial inflows of external assistance, which are typically used to support government spending and public investment. These same 136 Poverty and the WTO: Impacts of the Doha Development Agenda Table 5.1. Components of GDP GDP Component Share (%) Private consumption 72.4 Private investment 11.2 Government 28.9 Exports 20.6 Imports -33.0 Total 100.0 Source: Author's calculations. foreign inflows permit Mozambique to run a trade deficit, with the value of imports substantially exceeding the value of exports. Table 5.2 indicates the sectoral structure of production and trade. Agriculture, forestry, and fisheries amount to about 25 percent of GDP at factor cost. Trade and transport amount to another 25 percent and construction to nearly 10 percent. More than half of total exports come from two primarily foreign-owned island sectors. Aluminum smelting alone accounted for 48 percent of the value of total exports in 2001. Exports of electricity from the Cahora Bassa dam in northern Mozambique accounted for another nearly 10 percent of total exports. Unfortunately, the large majority of these export revenues are used to pay for imported intermediates, salaries for expatriate personnel, and repatriation of profits. Hence, the links to the Mozambican economy are relatively small.5 Fisheries provide the next most important source of export revenue. Imports tend to be concentrated in processed food, fuel, and manufactures, particularly transport equipment and other capital goods. Average tariff rates by commodity are also included in table 5.2. The rates implied by the social accounting matrix (SAM) originally developed for this analysis are presented under the heading "average tariffs," and the rates used in the GTAP model of global trade are presented under the heading "GTAP tariffs." Generally, the tariffs implied by the SAM correlate well with those employed in the GTAP model (the correlation is about 0.58), even though the methodologies for developing these tariff aggregates have been rather different. Table 5.3 provides a better sense of the degree of competition between imports and domestic production. The results in the table are derived from an analysis of local production and imports comprising all economic activity divided into 144 sectors. Each of the 144 sectors was put into one of three groups. The first group contains sectors where production accounts for at least 90 percent of total availability (production plus imports). The second group contains sectors where imports account for at least 90 percent of total availability. The third group The Doha Trade Round and Mozambique 137 Table 5.2a. Sectoral Shares in Value Added, Exports, and Imports Value Export Import Average GTAP Sector added share share tariff tariff share (%) (%) (%) rate rate Paddy rice 1.0 0.0 0.0 0.0 2.3 Wheat 0.0 0.0 1.9 2.4 2.1 Cereal grains nec 2.1 0.2 0.3 2.0 2.3 Vegetables, fruit, nuts 3.8 1.9 0.1 23.0 23.0 Oilseeds 0.8 0.0 0.1 7.8 9.9 Sugar cane, sugar beet 0.2 0.0 0.0 0.0 0.0 Plant-based fibers 1.1 0.1 0.0 23.2 0.0 Crops nec 9.7 2.6 0.4 3.2 5.2 Bovine cattle, sheep, goats, horses 0.6 0.0 0.1 1.9 6.1 Animal products nec 1.1 0.0 0.5 10.4 4.7 Forestry 2.7 1.5 0.0 2.5 2.7 Fishing 2.5 12.6 0.0 22.4 6.8 Minerals nec 0.3 0.3 0.2 5.3 7.1 Bovine meat products 0.4 0.0 0.0 23.2 15.7 Meat products nec 1.2 0.2 1.0 8.9 19.4 Vegetable oils and fats 0.3 1.1 1.1 16.0 13.6 Processed rice 0.1 0.0 4.5 5.8 7.1 Sugar 0.1 0.5 0.6 5.3 7.5 Food products nec 2.5 0.6 3.4 9.2 18.3 Beverages and tobacco products 0.8 0.1 1.6 9.4 24.2 Textiles 0.4 2.6 3.8 11.5 20.7 Wearing apparel 0.6 0.6 0.5 21.7 24.0 Leather products 0.1 0.1 0.3 29.9 22.6 Wood products 0.7 0.4 1.1 14.6 18.0 Paper products, publishing 0.0 0.0 0.8 9.5 6.5 Petroleum, coal products 0.2 2.5 4.4 12.0 4.8 Chemical, rubber, plastic products 0.4 0.3 19.0 6.7 9.4 Mineral products nec 0.5 0.1 2.4 6.4 8.8 Ferrous metals 4.5 49.0 0.2 9.6 6.3 Metal products 0.2 0.4 6.3 5.1 9.9 Motor vehicles and parts 0.0 0.0 6.1 7.9 8.6 Transport equipment nec 0.0 0.2 9.5 7.8 11.5 Electronic equipment 0.0 0.0 6.0 2.4 6.9 continued 138 Poverty and the WTO: Impacts of the Doha Development Agenda Table 5.2b. Sectoral Shares in Value Added, Exports, and Imports (Continued) Value Export Import Average GTAP Sector added share share tariff tariff share (%) (%) (%) rate rate Manufactures nec 0.0 0.2 1.6 21.6 21.9 Electricity 1.9 7.8 4.2 0.0 0.0 Water 0.3 0.0 0.0 0.0 0.0 Construction 9.4 0.0 0.0 0.0 0.0 Trade 17.3 0.0 0.0 0.0 0.0 Transport nec 7.2 6.5 0.0 0.0 0.0 Water transport 0.2 0.0 0.0 0.0 0.0 Air transport 0.4 0.0 0.0 0.0 0.0 Communication 1.8 0.0 0.0 0.0 0.0 Financial services nec 2.0 0.5 0.2 0.0 0.0 Insurance 0.1 0.0 0.3 0.0 0.0 Business services nec 3.7 4.7 16.3 0.3 0.0 Public administration, defense, education, and health 16.3 2.5 1.1 0.0 0.0 Dwellings 0.4 0.0 0.0 0.0 0.0 Total 100.0 100.0 100.0 n.c. n.c. Source: Author's calculations. Note: nec = not elsewhere classified; n.c. = not calculated. contains all remaining products. This third group contains sectors where neither domestic supply nor imports dominate the total supply of the commodity. The first two groups are considered to be specialized, and the third group is considered nonspecialized. Table 5.3 indicates that, in general, sectors tend rather strongly to be either dominated by imports or by domestic production. Overall, about 89 percent of the value of domestic production is specialized, with the large majority of these facing minor to no import competition in their particular product category.6 The sectors that compete most directly with imports are in primary product processing, which includes processed foods. According to the table, 53 percent of sales in this category come from sectors that are specialized (dominated by either imports or by domestic production). This implies that slightly less than half of sales in these sectors are in sectors where both imports and domestic production account for a significant volume of total domestic supply. These sectors also benefit from fairly substantial tariff protection (see table 5.2). However, these The Doha Trade Round and Mozambique 139 Table 5.3. Indications of Import Competition Specializeda Overall Share of Share of production value total supply production Sector share (%) (%) (%) Total economy 100.0 82.1 88.8 Agriculture, forestry and fisheries 15.1 98.2 98.5 Primary product 12.9 46.1 53.4 processing Other goods 8.1 74.6 74.5 Services 63.9 89.1 95.5 Source: Author's calculations. a. The figures in this table are drawn from production and import information for 144 sectors represent- ing all commodities. The intent is to discover which productive sectors compete intensively with imports and which are specialized, meaning that either commodity supply comes 90 percent from domestic production or 90 percent from imports. sectors make up only about 13 percent of the value of total sales and a smaller percentage of value added. Generally, the volume of resources located in sectors where import competition could be expected to be keen is relatively small. There is little to no possibility for substitution between domestic production and imports in sectors where imports are dominant, such as oil, vehicles, and capital goods. Mozambique quite simply has very little to no productive capacity in these areas. Consequently, imports are expected to dominate under any scenario. Similarly, where production values for tradeables are large, such as in primary agriculture and fisheries, import volumes tend to be minor. Import volumes are also minor in most service sectors. With respect to households, home consumption of basic food items represents a very important element of total expenditure. The importance of home consumption, from various perspectives, is presented in table 5.4. According to the macroeconomic accounts, home consumption amounts to 22 percent of total consumer expenditure on commodities. Home consumption is much more prevalent in rural than in urban areas. It amounts to about 36 percent of total rural consumer spending and only about 8 percent of total urban consumer expenditure. Wealthy households whose population weight is small but whose economic weight is large tend to dampen significantly the importance of home consumption in the macroeconomic accounts. Wealthy individuals tend to engage 140 Poverty and the WTO: Impacts of the Doha Development Agenda Table 5.4. Share of Value of Home Consumption in Total Consumption Share-weighting scheme Urban Rural Total Macroeconomic share 7.8 35.7 22.0 Population weight share 15.7 58.2 44.6 Poor population weight share 19.5 59.2 47.1 Source: Author's calculations. in very little home consumption as a share of total consumption and have large economic weight, thus their presence drives down the share of home consumption in the macroeconomic data. When home consumption shares are derived using population weights (for example, the share of home consumption for the average household), the share of home consumption grows considerably. At the national level, the average household obtains 45 percent of the value of total consumption from home consumption. The average rural household share remains considerably higher than the urban household share, at 58 percent and 16 percent, respectively. The population categorized as poor tends to home consume proportionately somewhat more than the national average. Nevertheless, in terms of share of goods that are home consumed, households characterized as poor are not all that different from the population average. This is not surprising when one considers that the poor represent more than half the population. In addition, a further large fraction of the population consumes at levels above, but still near, the poverty line. For example, 90 percent of the population consumes at levels less than twice the poverty line. The tendency to home consume apparently remains relatively constant across these basic levels of income. Inequality James, Arndt, and Simler (2005) conduct a detailed analysis of inequality based on the 2002­3 National Household Budget Survey (INE 2004) for Mozambique. They estimate a national Gini coefficient of 0.42, which represents a fairly high degree of inequality, though not out of line with other Sub-Saharan African countries.7 Table 5.5 shows an index of real consumption by quintile. Families in the highest quintile consume about eight times the value for the poorest quintile. Inequality varies by region, with consumption tending to be more evenly distributed in rural than in urban zones (a standard result). Regional differences also exist with the south, especially the capital city, Maputo, exhibiting much greater degrees of inequality. The Doha Trade Round and Mozambique 141 Table 5.5. Consumption by Quintiles Real As ratio of highest Population quintile consumption quintile's index consumption 0-20% 0.39 7.97 21-40% 0.66 4.63 41-60% 0.94 3.29 61-80% 1.32 2.34 81-100% 3.08 1.00 Mean 1.28 2.41 Source: Author's calculations. Simulations and Results Table 5.6 describes the shocks applied in the simulations analyzed, and table 5.7 describes the simulations. Results from the GTAP model of global trade are transmitted to the Mozambique model via changes in import prices and export prices and quantities faced by Mozambique. Import price changes are simply applied to the exogenous import prices in the Mozambique model. Export price and quantity changes derived from the GTAP model are applied in the manner developed by Horridge and Zhai in the appendix to chapter 3 of this volume. Specifically, an export demand function of the form: (5.1) Q = [FP/P]^ESUBM (where Q is the quantity exported, P is the export price, ESUBM is the elasticity of demand for exports, and FP is a shift parameter) has been added to the Mozambique model to mimic the global GTAP model. In the appendix to Chapter 3, Horridge and Zhai show that export price and quantity changes generated by the GTAP can be mimicked in a country through shocks to the shift parameter FP. Using lowercase to indicate percentage change, the percentage change in FP applied to the Mozambique model can be derived as follows: (5.2) fp = p + q/ESUBM. The four simulations presented are detailed in table 5.7. These are unilateral complete trade liberalization (UniLib), global trade liberalization with Mozambique not participating (Global), complete global trade liberalization 142 Poverty and the WTO: Impacts of the Doha Development Agenda Table 5.6a. Export and Import Price Changes and Tariff Cuts for Simulations Global liberalization Doha Sector Export Import Export Export Import Export prices prices quantity prices prices quantity Paddy rice n.a. 12.8 n.a. n.a. 2.9 n.a. Wheat n.a. 6.7 n.a. n.a. 1.5 n.a. Cereal grains nec 1.6 3.4 -5.2 0.0 1.6 1.8 Vegetables, fruit, nuts 1.4 2.7 14.6 0.0 0.9 -4.3 Oilseeds 3.3 6.4 56.5 0.7 2.2 11.7 Sugar cane, sugar beet n.a. n.a. n.a. n.a. n.a. n.a. Plant-based fibers 3.5 1.1 26.9 1.0 1.2 9.0 Crops nec 2.0 0.7 20.9 0.0 0.7 -2.4 Bovine cattle, sheep, goats, horses n.a. 3.3 n.a. n.a. 1.7 n.a. Animal products nec 1.6 2.1 -6.3 0.1 1.2 -1.7 Forestry -0.9 -0.2 3.0 -0.3 0.1 1.8 Fishing -2.4 0.4 9.5 -0.7 0.4 0.5 Minerals nec -0.8 1.1 2.2 0.0 1.7 0.7 Bovine meat products n.a. 3.4 n.a. n.a. 2.0 n.a. Meat products nec 1.2 1.4 -37.7 0.1 1.0 -12.6 Vegetable oils and fats 0.5 2.6 -16.2 0.2 1.2 4.0 Processed rice 2.2 5.6 -6.8 0.2 3.0 -2.1 Sugar 0.0 1.3 54.9 0.0 1.3 17.0 Food products nec 0.1 -0.1 -16.1 -0.1 0.6 -4.4 Beverages and tobacco products -0.7 -0.7 -6.5 -0.1 0.2 -2.1 Textiles -0.1 -1.3 -2.4 0.1 0.7 -3.8 Wearing apparel -1.0 -2.0 22.7 -0.2 -0.4 1.7 Leather products -0.8 -0.9 -8.6 0.0 0.2 -8.8 Wood products -1.0 -1.1 -5.3 -0.2 -0.2 -1.1 Paper products, publishing -0.4 1.6 25.2 0.4 2.5 -3.9 Petroleum, coal products -1.0 -0.8 16.0 -0.3 0.0 1.7 Chemical, rubber, plastic products -1.0 -0.4 112.5 -0.2 0.8 39.9 Mineral products nec -0.8 2.8 -8.3 0.0 3.5 -2.9 Ferrous metals -1.0 -0.7 -7.6 -0.2 0.0 -0.8 Metal products -0.9 -1.0 -21.9 0.0 -0.1 -3.6 continued The Doha Trade Round and Mozambique 143 Table 5.6b. Export and Import Price Changes and Tariff Cuts for Simulations (cont.) Global liberalization Doha Sector Export Import Export Export Import Export prices prices quantity prices prices quantity Motor vehicles and parts n.a. -2.9 n.a. n.a. -0.4 n.a. Transport equipment nec -1.0 -0.9 -1.0 -0.1 -0.1 -0.1 Electronic equipment n.a. -1.0 n.a. n.a. -0.1 n.a. Manufactures nec -1.0 -1.1 1.5 -0.1 -0.1 -0.3 Electricity -0.9 -1.0 2.1 -0.1 -0.1 1.6 Water n.a. n.a. n.a. n.a. n.a. n.a. Construction n.a. n.a. n.a. n.a. n.a. n.a. Trade n.a. n.a. n.a. n.a. n.a. n.a. Transport nec -1.0 n.a. 1.3 -0.2 n.a. 0.4 Water transport n.a. n.a. n.a. n.a. n.a. n.a. Air transport n.a. n.a. n.a. n.a. n.a. n.a. Communication n.a. n.a. n.a. n.a. n.a. n.a. Financial services nec -1.2 -0.7 2.4 -0.3 -0.1 1.0 Insurance n.a. -0.8 n.a. n.a. -0.2 n.a. Business services nec -1.0 -0.7 1.2 -0.2 -0.1 0.2 Public administration, defense, education, health -0.8 -0.8 0.2 -0.2 -0.1 0.4 Dwellings n.a. n.a. n.a. n.a. n.a. n.a. Source: Based on results from chapter 3 of this book. Note: nec = not elsewhere classified; n.a. = not applicable (commodity has import or export volume of zero). including Mozambique (FL), and the Doha scenario (Doha). These scenarios are described in detail in chapter 2. Because of its status as an LDC, Mozambique does not have to reduce its tariffs under the Doha scenario. Results are presented in tables 5.8, 5.9, and 5.10. Focusing first on the macroeconomic results in table 5.8, one notes that unilateral trade liberalization generates a substantial real exchange rate depreciation. With tariffs removed, imports become more attractively priced and import volumes increase. To obtain the foreign currency to purchase these additional imports, exports must increase more than proportionately because of the large initial trade deficit. As mentioned above, to remain consistent with the GTAP, downward-sloping export demand 144 Poverty and the WTO: Impacts of the Doha Development Agenda Table 5.7. Simulations Simulation Description UniLib Unilateral complete trade liberalization by Mozambique uniquely Global Complete global trade liberation excluding Mozambique FL Complete global trade liberalization including Mozambique Doha Doha Source: Scenarios based on chapter 2 of this book. functions are specified. Therefore, the growth in export volume results in somewhat lower prices for export commodities, leading to a deterioration in the terms of trade (TOT). Devaluation helps to attenuate the import surge and provides additional incentives to exporting sectors. Global trade liberalization with Mozambique not participating operates through shifts in world demand curves for Mozambican export commodities as described above. It turns out that global trade liberalization tends to improve the terms of trade for Mozambique, permitting increased imports even though exports remain flat. The results for the third scenario, FL, are essentially an additive combination of the first two simulations. In the Doha scenario, the TOT effect is negative for Mozambique as a consequence of the elimination of export subsidies and the erosion of Mozambican tariff preferences in industrial countries. The negative TOT shock is accommodated primarily through compression of imports (recall that initial import values are much larger than export values). A relatively large decline in the export price for the fisheries sector, an important exporter, helps to explain both the direction of the TOT shock and the compression of import values. Overall household welfare as calculated from the CGE model (table 5.9) is driven largely by the TOT. The presence of downward-sloping export demand functions is a particularly important element in the TOT changes when domestic trade liberalization is considered. By contrast, with the small country assumption (constant world prices) and operatively small export transformation elasticities, unilateral trade liberalization tends to improve household welfare (scenario not shown). In all scenarios, the impacts on welfare are not particularly large. Microsimulation analysis generally points to similarly small results. Table 5.10 summarizes the implications of trade liberalization on household welfare for the lower four income quintiles. It shows the mean, minimum, and maximum household level welfare impact (in percentage change from the base) for each simulation. The mean effect in the microsimulation model tends to be closer to zero than the equivalent welfare calculation provided in table 5.9. This is due The Doha Trade Round and Mozambique 145 Table 5.8. Macroeconomic Indicators (percentage change, relative to base) Variable UniLib Global FL Doha Total absorption -0.7 0.6 0.0 -0.2 Real exports 4.4 0.0 4.4 0.2 Real imports 0.5 1.9 2.4 -0.4 Real exchange rate 4.3 -3.4 0.8 0.4 TOT -1.4 0.8 -0.6 -0.7 Source: Author's simulations. Table 5.9. Equivalent Variation for Households (percentage change, relative to base) Variable Base UniLib Global FL Doha Urban 2,538.74 -0.552 0.489 -0.088 -0.219 Rural 2,631.26 -0.75 0.527 -0.192 -0.173 Total 5,170.00 -0.653 0.508 -0.141 -0.195 Source: Author's simulations. Note: UniLib, Global, FL, and Doha simulations are given as a percentage change from the figures in the "Base" column. primarily to the insulating effects of the high value of home consumption in the lower 80 percent of the consumption distribution (see table 5.4). Nevertheless, concentration of earnings sources in certain factors and consumption on certain commodities exposes some households to stronger than average effects of trade liberalization. The range of the distribution is captured by the maximum and minimum values. The worst affected household would be one specialized in the factor with least favorable change in factor prices and specialized in consumption of commodities whose prices have tended to rise. The range of outcomes for the Doha scenario is presented in figure 5.1. Outcomes for both urban and rural households tend to concentrate near the mean. Nevertheless, impacts tend to be much more heterogeneous in urban than in rural areas. This result also holds in all of the other scenarios (histograms not shown). This occurs as a result of more heterogeneous factor endowments across households in urban areas (rural households tend to depend very heavily on unskilled labor), as well as substantially greater reliance on the market for the purchase of commodities (that is, less own-consumption). For rural households, homogeneity in income sources tends to concentrate welfare outcomes near the mean, and the prevalence of home consumption implies that this mean effect is typically quite small. 146 Poverty and the WTO: Impacts of the Doha Development Agenda Table 5.10. Microsimulation, Percentage Changes in Welfare by Quintile Quintile Statistic UniLib Global FL Doha Rural 0­20% Mean -0.65 0.14 -0.49 -0.10 21­40% Mean -0.62 0.11 -0.48 -0.09 41­60% Mean -0.55 0.14 -0.38 -0.09 61­80% Mean -0.43 0.15 -0.24 -0.09 0­20% Maximum 1.99 2.16 2.64 0.17 21­40% Maximum 2.61 2.56 3.29 0.14 41­60% Maximum 1.71 2.05 2.87 0.17 61­80% Maximum 3.19 1.31 4.21 0.16 0­20% Minimum -1.37 -0.69 -1.70 -1.06 21­40% Minimum -1.90 -0.66 -1.89 -0.96 41­60% Minimum -1.43 -0.85 -2.16 -0.90 61­80% Minimum -1.72 -0.90 -2.62 -0.93 Urban Quintile Statistic UniLib Global FL Doha 0­20% Mean -0.29 0.08 -0.23 -0.18 21­40% Mean -0.27 0.13 -0.16 -0.19 41­60% Mean -0.10 0.17 0.05 -0.18 61­80% Mean -0.02 0.31 0.27 -0.20 0­20% Maximum 2.39 1.53 3.38 0.25 21­40% Maximum 3.05 1.65 4.02 0.29 41­60% Maximum 2.61 2.27 3.17 0.37 61­80% Maximum 2.64 2.15 3.48 0.20 0­20% Minimum -1.78 -0.89 -1.95 -0.96 21­40% Minimum -2.21 -1.09 -2.36 -1.25 41­60% Minimum -2.03 -0.99 -2.47 -1.17 61­80% Minimum -1.91 -0.91 -1.89 -1.07 Source: Author's simulations. Note: The top-earning quintile is not presented because of difficulties in separating labor and capital income for this group of households. Because nearly three of four poor Mozambicans live in rural areas, the overall implications for poverty rates in all of the scenarios tend to be small. In the scenario with the largest effect, unilateral trade liberalization (UniLib), the poverty rate edges up from 54.1 percent nationwide to 54.4 percent. Impacts in the remaining scenarios are much smaller. The Doha Trade Round and Mozambique 147 Figure 5.1. Distribution of Changes in Household Welfare Urban Rural 30 20 t rcen Pe 10 0 ­0.5 ­1.0 ­0.5 0 0.5 ­1.5 ­1 ­0.5 0 0.5 Welfare change Source: Author's simulations. Limitations of the Analysis Price Transmission As reviewed in Winters, McCulloch, and McKay (2004), marketing costs between the frontier of a country (the port, for example) and the point of production cause the price of an export good at the point of production to be considerably more variable in proportional terms than the free on board (FOB) price. For example, consider a good with an export price at the border of 100 and a marketing wedge between the border and the farm or factory gate of 50. If the FOB price increases by 10 percent to 110 and the marketing wedge remains constant, then the farm or factory gate price also increases by 10, from 50 to 60 for a proportionately double price increment of 20 percent. The inverse happens with respect to importation. Consider an imported good that is available at the border for a price of 50. Marketing costs of 50 are incurred to get the product to the point of final consumption. If the border price increases by 10 percent and marketing costs remain constant, then the price of the imported good at the point of consumption increases by only 5 percent. Therefore, in terms of 148 Poverty and the WTO: Impacts of the Doha Development Agenda proportional price changes, marketing wedges tend to expand the impact of changes in export prices (free on board [FOB] minus export taxes) and dampen the impact of changes in import prices (CIF plus import tariffs). If border price changes are transmitted in the manner described above, it seems likely that past assessments of the implications of past global trade negotiation rounds may have given undue weight to the implications of import price changes and insufficient weight to the implications of export price changes when considering the implications of trade agreements for poverty and well-being for many parts of Africa. The current model, with its explicit addition of margins for exports, imports, and domestics, partially captures these effects. This represents an important step forward; however, there remains much to do. The impact of trade liberalization on poverty depends crucially upon where the poor are living and the strength of the ensuing links to regional, national, and global markets. Distance and poor transport infrastructure alone may sever links to both import and export markets. Imperfect competition within the marketing system may also sever market linkages (Moser and Minten 2004). Thus, particularly in large countries such as Mozambique, the analysis of trade and poverty forces one to consider building models with finer levels of spatial detail. This is true for both commodity and factor markets. Unfortunately, attaining enhanced spatial detail is easier said than done. Attempts have been made (see, for example, chapter 7 by Ferreira Filho and Horridge in this volume); however, these attempts tend to be partial and tend not to generate a spatial price map that reflects the appropriate distribution of prices over space.8 This is crucial because more distant regions often exhibit higher rates of poverty and very high marketing wedges. Although a partial approach to regionalization (for example, regional detail in the production of some agricultural commodities) within an AGE model seems attractive initially, the incompleteness might actually hamper the goal of more faithfully modeling the role of geography in shaping the impact of policy change. Therefore, despite formidable information lacunae on the spatial distribution of economic activity and the complete absence of information on inter-regional trade, it may be better to develop regional SAMs that account for what is known about the regional distribution of economic activity, estimating the remainder under plausible assumptions.9 Revenue Replacement In the case of Mozambique, the GTAP model uses average tariffs obtained by multiplying applied tariff rates by import weights. To remain consistent, the country CGE model also uses these average tariff values. However, as discussed in Arndt and Tarp (2004), published tariff rates are generally larger than the tariff rate implied by The Doha Trade Round and Mozambique 149 the average tariff rate because of official exemptions, smuggling, or both. If the marginal import pays published tariff rates, then the published tariff rate and not the average rate is the operative one for trade policy analysis. In addition, the rents associated with smuggling and official tariff exemptions may be large. Elimination or reduction of these rents through trade liberalization can have substantial distributional effects, often with positive welfare implications for the poor (because the poor typically do not profit from these rents in the initial situation). Gaps between average and published tariff rates also have implications for revenue. Pritchett and Sethi (1994) find that the gap between these rates tends to fall as published tariff rates decline. Hence, higher collection ratios may substantially attenuate declines in revenues as a result of lower tariff rates. The heavy dependence of Mozambique and many other African countries on VAT applied at the border implies that even complete trade liberalization (tariff rates at zero) may have offsetting revenue implications if a higher share of import volumes pass through official channels and hence pay VAT. Examination of these revenue issues in the Mozambican context goes beyond the scope of this chapter (though it is an important topic for future research). The use of a neutral income tax for revenue replacement is a poor substitute for realistic modeling of revenue replacement options; however, the complexities of the revenue replacement issue (see Arndt and Tarp [2004]) precluded modeling of options that are effectively more realistic within the time frame available for this analysis. Downward-Sloping Export Demand Functions In the analysis undertaken in this chapter, trade liberalization by Mozambique results in increased export volumes. Because the country is presumed to face downward-sloping export demand functions, increases in exports result in lower prices and a deterioration in the nation's TOT. This formulation permits consis- tency with the GTAP model. Unfortunately, the formulation is the major driver of welfare results in the scenarios where Mozambique undertakes own-liberalization. Although this is perhaps a reasonable specification for some sectors, exports from many sectors are likely to be constrained by supply factors. In this view, more could be exported at a constant price if more could be produced. In fact, for many sec- tors, low export volumes are often pointed to as a cause of low prices, particularly at the farm or factory gate. Low volumes are viewed as a cause of high marketing costs and diminished confidence of potential importers in the quality and reliabil- ity of supply of Mozambican products. As indicated before, changing the modeling assumption to that of supply-constrained exports and constant world prices switches the sign on the welfare result for unilateral trade liberalization, although the implications remain relatively small for the same reasons discussed above. 150 Poverty and the WTO: Impacts of the Doha Development Agenda Despite these limitations of the analytical framework used in this chapter, a few robust conclusions may effectively be drawn. These are discussed in the final section. Conclusions To rise out of poverty, Mozambique must achieve rapid growth over a long period. Even with rapid growth, it will take some time, perhaps decades, to lift the bulk of the Mozambican population out of poverty. Seen from this perspective, the static results presented in this chapter are disappointing because they do not contribute to the growth required for such sustained poverty reduction. Nevertheless, as pointed out by Winters, McCulloch, and McKay (2004), most economists believe that more liberal trading regimes tend to be associated with higher rates of economic growth. Difficulties, in their view, come about in making the transition from more restrictive to more open trade regimes. In this respect, the results of this chapter may be viewed in a more positive light. For Mozambique, the short-term poverty impacts of moving to a liberal trade regime appear to be relatively small. Hence, Mozambique has the opportunity to set in place the liberal trade element of a growth strategy at relatively low short-term adjustment cost. It is well recognized that, especially in the Mozambican context, low or zero barriers to imports are not a sufficient condition for ensuring poverty-reducing economic growth. A key element to sustaining growth over the coming decades very likely involves substantially expanding the volume of exports in sectors where volumes are currently very small--or breaking into new export markets entirely. A liberal import regime helps set the stage for export expansion; however, such expansion will not occur without appropriate complementary policies aimed at improving price transmission to rural areas, as well as facilitating producer supply response. Only after such reforms will the vast majority of the poor in Mozambique be able to take advantage of the improved world market opportunities that are expected to follow from global trade reforms. Notes 1. For a more complete historical review, see Arndt, Jensen, and Tarp (2000). 2. "Big projects," such as the Mozal aluminum smelter, have contributed considerably to GDP but very little to gross national product. 3. Löfgren, Harris, and Robinson (2001) and Tarp and others (2002) provide detailed explanations of the basic CGE model that was revised for the purposes of this analysis. 4. Downward-sloping export demand functions offer the considerable advantage of consistency with the global modeling framework. Disadvantages are discussed in detail in the penultimate section of this chapter, which presents a critique of the current model. The Doha Trade Round and Mozambique 151 5. Aluminum smelting is modeled as an island sector. Nearly 100 percent of production is exported. Returns to capital from aluminum smelting are assumed to be repatriated abroad. 6. Substitution across commodities would amplify competition. Thus, for example, maize production faces little direct import competition in the form of imported maize. However, significant volumes of wheat and rice are imported. Because maize meal and bread are substitutes, domestic maize competes indirectly with imports through the potential for consumers to alter dietary choices. 7. For example, the Gini coefficient is 0.43 in Uganda (Uganda Bureau of Statistics 2003). 8. The distribution of prices over time is another important element. 9. Another option is to link the results of a CGE model to a partial equilibrium model(s) to flesh out in more detail implications for important sectors. References Arndt, C., H. T. Jensen, S. Robinson, and F. Tarp. 2000. "Agricultural Technology and Marketing Margins in Mozambique." Journal of Development Studies 37 (1): 121­37. Arndt, C., H. T. Jensen, and F. Tarp. 2000. "Stabilization and Structural Adjustment in Mozambique." Journal of International Development 12: 299­323. Arndt, C., and F. Tarp. 2004. "Trade Policy Reform and the Missing Revenue: An Application to Mozambique." National Directorate of Planning and Budget, Mozambique Ministry of Finance, Maputo. Chen, S., and M. Ravallion. 2004. "Welfare Impacts of China's Accession to the World Trade Organization." World Bank Economic Review 18 (1): 29­57. Ferreira Filho, J. B. S., and M. Horridge. 2005. "The Doha Round, Poverty, and Regional Inequality in Brazil." In Poverty and the WTO: Impacts of the Doha Development Agenda, ed. T. W. Hertel and L. A. Winters. Basingstoke, U.K.: Palgrave Macmillan; Washington, DC: World Bank. Hertel, T., D. Hummels, M. Ivanic, and R. Keeney. 2004. "How Confident Can We Be in CGE-Based Assessments of Free Trade Agreements?" GTAP Technical Paper 26, Purdue University, West Lafayette, IN. http://www.gtap.agecon.purdue.edu/resources/working_paper.asp INE (Instituto Nacional de Estatística). 2004. Inquérito Nacional aos Agregados Familiares sobre Orçamento Familiar 2002/3 [National Household Budget Survey]. Maputo. http://www.ine.gov.mz James, R. C., C. Arndt, and K. R. Simler. 2005. "Has Economic Growth in Mozambique Been Pro- Poor?" National Directorate of Planning and Budget, Ministry of Planning and Finance, Maputo, Mozambique. Löfgren, H., R. L. Harris, and S. Robinson. 2001."A Standard Computable General Equilibrium (CGE) Model in GAMS." Trade and Macroeconomics Discussion Paper 75, International Food Policy Research Institute, Washington, DC. Maximiano, N. 2005. "A Dinamica dos Determinantes de Pobreza" ["The Dynamics of Poverty Determinants"]. National Directorate of Planning and Budget, Ministry of Planning and Finance, Maputo, Mozambique. Moser, C., and B. Minten 2004. "Missed Opportunities and Missing Markets: Spatio-Temporal Arbitrage of Rice in Madagascar." Paper presented at the Trade and Industrial Policy Strategies Forum 2004: "African Development and Poverty Reduction: The Macro-Micro Linkage," October 13, Somerset West, South Africa. Mozambique Ministry of Planning and Finance, International Food Policy Research Institute, and Purdue University. 2004. Poverty and Well-Being in Mozambique: The Second National Assessment (2002­2003). Maputo: National Directorate of Planning and Budget, Ministry of Planning and Finance, Mozambique. Pritchett, L., and G. Sethi. 1994. "Tariff Rates, Tariff Revenue, and Tariff Reform: Some New Facts." World Bank Economic Review 8 (1): 1­16. 152 Poverty and the WTO: Impacts of the Doha Development Agenda Tarp, F., C. Arndt, H. T. Jensen, S. Robinson, and R. Heltberg. 2002. "Facing the Development Challenge in Mozambique: An Economy-Wide Perspective." IFPRI Research Report 126, International Food Policy Research Institute, Washington, DC. Winters, L. A., N. McCulloch, and A. McKay. 2004. "Trade Liberalization and Poverty: The Evidence So Far." Journal of Economic Literature 42: 72­115. Part III household impacts oF price changes 6 The WTO Doha Round, Cotton Sector Dynamics, and Poverty Trends in Zambia Jorge F. Balat and Guido G. Porto Summary The Zambian cotton sector went through significant reforms during the 1990s. After a long period of parastatal control, a process of liberalization in cotton production and marketing began in 1994. These reforms were expected to benefit farmers. In Zambia, these are rural, often vulnerable, smallholders. This chapter investigates the connection between the dynamics of the cotton sector and the dynamics of poverty and evaluates to what extent cotton can work as a vehicle for poverty alleviation. The findings suggest that cotton can indeed act as an effective mechanism for increased household welfare, and income gains associated with cotton production are likely to have positive impacts on the long-run nutritional status of Zambian children. These impacts, however, are relatively small. Introduction The Zambian cotton sector has been profoundly reformed during the last 10 years. Traditionally, cotton production was controlled by the government through public firms and parastatal organizations. Until 1994, Lintco (Lint Company of Zambia) sold inputs on loan and purchased cotton seeds from farmers. The 155 156 Poverty and the WTO: Impacts of the Doha Development Agenda reforms comprised a broad liberalization of the sector that included the privatiza- tion of Lintco and the encouragement of market entry. The dynamics of the sector include an initial phase of regional private monopsonies and a later phase of more active competition. At present, the market is relatively unregulated, and several firms seem to freely compete (but may collude) for locally produced cotton seeds. Poverty in Zambia is a deep phenomenon, particularly in rural areas. In 1998, for example, the head count was 82.1 percent. Moreover, poverty in some regions exceeded 90 percent. Cotton is one of the key agricultural activities in rural Zam- bia. In cotton-growing provinces, a large share of the cash income of rural farmers comes from the sale of cotton seeds. This chapter studies the dynamics of cotton production and marketing and the links with rural poverty. It has long been claimed that cash crops can work as an effective vehicle for poverty alleviation in rural areas. Hence, this chapter explores whether cotton can actually achieve significant poverty reduction in Zambia. The chapter begins with a review of the reforms in the cotton sector. Because cotton can be produced only in selected regions in Zambia, the trends in poverty and cotton income in different provinces are examined. Two different empirical exercises are performed to investigate the role of cotton as a source of household income. First, the evolution of the share of income generated by cotton is exam- ined, and the poverty impacts brought about by an increase in cotton prices and an expansion of the sector are simulated. Second, the potential impact of subsis- tence households switching to cotton production is evaluated. Income differen- tials, anthropometric measures, and educational outcomes are also examined. The findings are mixed. On the one hand, cotton production is associated with higher household income, lower poverty, and higher welfare. This result has two components. First, higher cotton prices would benefit rural farmers directly. Second, cotton farmers enjoy income gains and long-run nutritional gains compared to subsistence farmers. On the other hand, the estimated mag- nitudes are not as large as expected. Very large price changes of cotton seeds or very large supply responses would be needed to estimate empirically meaningful reductions in poverty rates. Thus, international trade in cotton is indeed a promising activity for farmers, but there is a long way to go to achieve the full benefits from increased market access and higher prices. Reforms in the provi- sion of infrastructure, access to credit, extension services, and social services are essential complementary policies to market access and liberalization of agricul- tural markets in the developed world. Section 1 reviews the main reforms in cotton markets, discusses the charac- teristics of world cotton markets, and provides a poverty profile of Zambian households. Section 2 looks at cotton as a source of cash income and provides a simulation of the impacts of market access and higher international prices. Sec- tion 3 provides an estimation of income differential gains in cotton over subsis- The WTO Doha Round, Cotton ... Poverty Trends in Zambia 157 tence agriculture. Nonmonetary outcomes, such as educational and nutritional status, are studied as well. Section 4 reviews the main results and summarizes the conclusions. Cotton Reforms and Poverty Trends in Zambia Zambia is a landlocked country located in southern central Africa. Clockwise, its neighbors are Congo, Tanzania, Malawi, Mozambique, Zimbabwe, Botswana, Namibia, and Angola.1 In 2000, the population was 10.7 million inhabitants. With a per capita GDP of only US$302, Zambia is one of the poorest countries in the world and is considered a least developed country (LDC). Zambia achieved independence in 1964. A key characteristic of the country is its abundance of natural resources, particularly mineral deposits (for example, copper) and land. As a result of high copper prices, the new republic did quite well in the initial stages of development. Poverty and inequality, however, were wide- spread, and this raised concerns among the people and the policymakers. Soon, the government began to adopt interventionist policies, with a much larger par- ticipation of the state in national development. Interventions included import substitution, price controls on all major agricultural products (such as maize), and nationalization of manufacturing, agricultural marketing, and mining. In the 1970s and 1980s, the decline in copper prices and negative external con- ditions led to stagnation and high levels of external debt. A crisis emerged, and a structural adjustment program was implemented between 1983 and 1985. Riots in 1986 forced the government to abandon the reforms in 1987. A second Interna- tional Monetary Fund program failed in 1989, when the removal of controls in maize led to significant price increases. In 1991, the Movement for Multiparty Democracy was elected. Faced with a sus- tained, severe recession and a meager future, the new government began economy- wide reforms including macroeconomic stabilization, exchange rate liberalization, fiscal restructuring, removal of maize subsidies, decontrol of agricultural prices, pri- vatization of agricultural marketing, and new trade and industrial policy. For a more detailed description of the reforms, see World Bank (1994), McCulloch, Baulch, and Cherel-Robson (2001), and Litchfield and McCulloch (2003). Cotton Reforms The cotton sector was significantly affected by the agricultural reforms adopted by Zambia during the 1990s.2 Before 1994, intervention in cotton markets was wide- spread and involved setting prices for sales of certified cotton seeds, pesticides, and sprayers; providing subsidized inputs to producers, facilitating access to credit, and so forth. From 1977 to 1994, Lintco acted as a nexus between local 158 Poverty and the WTO: Impacts of the Doha Development Agenda Zambian producers and international markets. Lintco had a monopsony in seed cotton markets and a monopoly in inputs sales and credit loans to farmers. The reforms of the mid-1990s eliminated most of these interventions, and markets were liberalized. After Lintco was sold to Lonrho Cotton in 1994, a domestic monopsony developed soon after liberalization. As market opportuni- ties arose, several firms (private ginners such as Swarp Textiles and Clark Cotton) entered the Zambian cotton market. This initial phase of liberalization, however, did not succeed in introducing much competition in the sector because the three major firms segmented the market geographically. In consequence, liberalization gave rise to geographic monopsonies rather than national oligopsonies. At that time, Lonrho and Clark Cotton developed an out-grower scheme with the Zambian farmers. This scheme allowed ginners to expand production and take advantage of economies of scale and idle capacity. In these out-grower pro- grams, firms provided seeds and inputs on loans together with extension services to improve productivity. The value of the loan was deducted from the sales of cot- ton seeds to the ginners at picking time. Prices paid for the harvest supposedly depended on international prices. Initially, repayment rates were high (roughly 86 percent), and cotton production significantly increased. By 1997, the expansion of the cotton production base attracted new entrants, such as Amaka Holdings and Continental Textiles. Entrants and incumbents started competing in many districts, doing away with localized monopsonies. As a result, the capacity for ginning increased beyond production levels. This caused an excess demand for cotton seeds and tightened the competition among ginners for Zambian cotton. In addition, some entrants that were not using out-grower schemes started offering higher prices for cotton seeds to farmers who had already signed contracts with other firms. This caused repayment problems and increased the rate of loan defaults. The relationship between ginners and farmers started to deteriorate. On top of all this, world prices began to decline, and farm gate prices declined as a result. After many years of high farm gate prices, and with limited information on world market conditions, farmers started to mistrust the ginners, and suspi- cions of exploitation arose. In consequence, farmers felt that out-growers' con- tracts were being breached, and default rates increased. This led firms to increase the price of the loans charged to farmers, who, in the end, received a lower net price for their crops. Partly as a result of this failure of the out-grower scheme, Lonrho announced its sale in 1999, and Dunavant Zambia Limited entered the market. Today, the major players in Zambian cotton markets are Dunavant Zambia Limited, Clark Cotton Limited, Amaka Holdings Limited, Continental Ginneries Limited, Zam- bia-China Mulungushi Textiles, and Mukuba Textiles. The WTO Doha Round, Cotton ... Poverty Trends in Zambia 159 At present, most cotton production in Zambia is carried out under the out- grower scheme. Farmers and firms understood the importance of honoring con- tracts and the benefits of maintaining a good reputation. The out-grower pro- grams were perfected, and there are now two systems used by different firms: the farmer group system and the farmer distributor system. In the latter, firms desig- nate an individual farmer as the distributor and provide inputs. The distributor prepares individual contracts with the farmers. The distributor is also in charge of assessing reasons for loan defaults--being able, in principle, to condone default in special cases--and responsible for renegotiating contracts in incoming seasons. In the farmer group system, small-scale producers deal with the ginneries directly, purchasing inputs on loan and repaying at the time of harvest. Both systems seem to work well. International Markets and World Prices The production of cotton and the international trade of cotton products are, and have traditionally been, subject to significant interventions. The distortions include taxes (either directly or indirectly through state marketing monopsonies [parastatals]), border intervention (tariffs, quotas), production and export sup- port, and input subsidies. Aksoy and Beghin (2004) summarize the markets for different commodities, including cotton. They suggest that the combined support in cotton production in the major world producers (Brazil, China, Egypt, Greece, Spain, Turkey, and the United States) between 1997­98 and 2001­2 ranged from US$3.8 billion to US$5.3 billion. The EU and the United States have historically intervened extensively in cotton markets (Baffes 2004). In the United States, intervention in cotton production is regulated by farm bills, such as those passed in 1996 and 2002. They establish price and income support (usually decoupling payments based on historical areas planted), tariffs, quotas, public agricultural research, provision of infrastructure (irrigation), export subsidies, export credit, subsidized loans and insurance, and so forth. The 2002 farm bill is expected to be in place for the next six years. The EU intervenes in cotton production to provide support to Spanish and Greek producers. Under the Common Agricultural Policy of the European Union, support is given to cotton growers based on the difference between the market price and a guide (support) price. Advance payments are calculated on the basis of estimated cotton production. Ginners receive these payments and pass them through to producers in the form of higher prices (Baffes 2004). The effects of the removal of cotton distortions in world markets have been widely researched. Some of the literature is reviewed by Baffes (2004). He reports results from the Food and Agricultural Policy Research Institute (2002), which 160 Poverty and the WTO: Impacts of the Doha Development Agenda showed that under global agricultural liberalization, world cotton prices would increase on average by 12.7 percent. Using similar methods, Sumner (2004) reports price increases according to different scenarios of cotton reforms. On average, his findings indicate an expected increase of 11.58 percent in world cot- ton prices. These numbers are in line with those reported by Baffes (2004). Notice, however, that the latter work focuses mostly on the impact on world prices of the elimination of U.S. domestic support. Comparison of the Baffes and Sumner results indicates that most of the price changes that can be expected from the lib- eralization of world markets would be generated by U.S. policies. This is con- firmed by Hoekman, Nicita, and Olarreaga (2004), who report much lower cotton price changes from a Doha scenario that considers the elimination of trade barri- ers without changes in domestic support. It is unclear whether the reforms needed to achieve these increases in prices are reasonable or even feasible. Nevertheless, in the subsequent sections of this chapter, these estimates are used to simulate the poverty effects on Zambian farmers. But before doing that, it is useful to review the poverty trends observed in Zambia during the 1990s. Poverty Trends In spite of the significant reforms adopted by the Zambian government and of the significant intervention in international agricultural markets, Zambia is one of the poorest countries in the world. Furthermore, poverty rates tended to increase during the 1990s. The description of the poverty trends in this chapter is based on three house- hold surveys, the 1991 Priority Surveys and the 1996 and 1998 Living Conditions Monitoring Surveys (LCMSs).3 The Priority Survey of 1991 is a Social Dimension of Adjustment Survey. It was conducted during October and November and cov- ered a total of 9,886 households. Sample sizes were increased to 11,750 and 16,800 households in the 1996 and 1998 LCMSs. Table 6.1 reports the poverty dynamics. In 1991, the poverty rate at the national level was 69.6 percent. Poverty increased in 1996, when the headcount reached 80 percent, and then declined toward 1998, with a headcount of 71.5 per- cent. In rural areas, poverty is widespread; the headcount was 88.3 percent in 1991, 90.5 percent in 1996, and 82.1 percent in 1998. Urban areas fared better, with a poverty rate of 47.2 percent in 1991, 62.1 percent in 1996, and 53.4 percent in 1998. In what follows, the focus is on rural areas. Poverty trends by provinces are reported in table 6.2. Zambia is a large country, and provinces differ substantially in such basic characteristics as land quality, dis- tance to the capital, roads, and so forth. In particular, cotton, the commodity under investigation here, can be produced (because of soil characteristics) in only The WTO Doha Round, Cotton ... Poverty Trends in Zambia 161 the Southern, Central, and Eastern provinces. At the national level, poverty increased from 1991 to 1996 and then declined in 1998. Poverty trends in Lusaka, the Copperbelt, and the Northwestern province are similar to those at the national level. In the Central province, poverty first declined in 1996 and then increased in 1998. In the remaining provinces, particularly in the Eastern and Southern provinces, poverty has declined throughout the period. Table 6.1. Poverty in Zambia (headcount) Region 1991 1996 1998 National 69.6 80.0 71.5 Rural 88.3 90.5 82.1 Urban 47.2 62.1 53.4 Source: Authors' calculations based on the 1991 Priority Survey and 1996 and 1998 LCMSs. Note: The headcount is the percentage of the population below the poverty line. Table 6.2. Rural Poverty Trends: 1991, 1996, and 1998 (headcount) Region 1991 1996 1998 National 88.3 90.5 82.1 Central 83.9 80.3 82.3 Copperbelt 66.2 78.8 82.1 Eastern 92.0 85.5 80.6 Luapula 90.8 86.1 84.6 Lusaka 70.9 78.1 75.7 Northern 94.2 90.6 83.3 Northwestern 86.3 87.3 77.4 Southern 85.4 82.5 73.0 Western 94.1 92.9 90.3 Source: Authors' calculations based on the 1991 Priority Survey and 1996 and 1998 LCMSs. Note: Rural poverty only. The headcount is the percentage of the population below the poverty line. 162 Poverty and the WTO: Impacts of the Doha Development Agenda Cotton Income and Higher Export Prices This section investigates the potential effects of higher cotton prices on household income in rural Zambia. As argued by Deaton (1989, 1997), the short-run effects of price changes can be assessed by looking at income shares. Table 6.3 reports the aver- age income shares for different sources of income. At the national level, the main sources of income are income from home consumption (42.5 percent), income from nonfarm businesses (16.8 percent), sales of food crops (9.1 percent), livestock and poultry (8.1 percent), and wages (6.9 percent). Note that the differences in income sources between poor and nonpoor households are not very significant. Because of regional variation in soil, climate, and infrastructure, the relevant sources of income may be different for households residing in different provinces. To clarify this, table 6.4 reports the main sources of agricultural household income in the rural areas of the nine Zambian provinces. The table shows the average share of total income accounted for by a given activity. In the Central, Eastern, and Southern provinces, the most relevant cash-crop activity is cotton. In the remaining provinces, cotton is not a feasible option, and income shares are negligible or zero. The remainder of this chapter investigates the impacts of cotton reforms only in the relevant provinces. Table 6.3. Sources of Income in Rural Areas, 1998 (percentage) Income Source Total Poor Nonpoor Own production 42.5 42.9 42.0 Sales of food crops 9.1 9.5 7.6 Sales of nonfood crops 3.8 4.0 2.9 Livestock and poultry 8.1 8.7 5.9 Wages 6.9 5.9 10.3 Income, nonfarm 16.8 16.3 18.3 Remittances 5.3 5.0 6.1 Other sources 7.5 7.7 6.9 100.0 100.0 100.0 Source: Authors' calculations based on the 1998 LCMS. The WTO Doha Round, Cotton ... Poverty Trends in Zambia 163 Table 6.4. Income Shares from Agricultural Activities, Rural Zambia, 1998 (percentage) 1 2 3 4 5 6 7 8 9 Total Cotton 8.4 0 9.5 0 0.8 0 0.1 2.8 0.2 3.1 Vegetables 1.1 2.8 0.3 0.2 1.2 0.7 0.5 1.7 0.3 0.8 Tobacco 0.2 0.1 2.3 0 0 0 0.1 0.2 0.1 0.5 Groundnuts 0.9 0.7 2.4 2 0.2 1.4 1.1 0.4 0.2 1.2 Paprika 0 0 0.1 0 0 0 0 0 0 0 Industrial maize 6.1 2 0.7 0.3 1.7 0.6 0.3 1.4 0.5 1.3 Cassava 0.3 0.2 0.1 4.1 0 2.4 2.2 0.1 1.3 1.2 Maize 4.4 3.1 3.2 0.5 1.1 0.9 3.8 0.9 2.6 2.2 Rice 0 0 0.3 0.1 0 0.1 0 0 1.2 0.2 Millet 0.9 0.1 0.2 0.3 0 1.3 0 0 0.2 0.4 Sorghum 0.1 0.2 0 0 0.1 0.2 0.5 0 0.2 0.1 Beans 0.2 0.1 0 0.5 0 2 0.8 0 0 0.5 Soya beans 0.4 0 0.4 0.1 0 0 0 0 0 0.1 Sweet potatoes 0.9 2.8 0.1 1 0 0.3 1.6 0.1 0.5 0.7 Irish potatoes 0 0.1 0 0.1 0 0 0.3 0.1 0 0.1 Sunflowers 0.1 0 0.5 0 0 0.1 0.1 0.2 0 0.1 Livestock 2.9 1.3 4.3 0.6 3.8 2 2.3 8 6.9 3.8 Poultry 6.4 2.2 4.5 2.7 5.9 3.4 2.8 4.6 6.7 4.3 Source: Authors' calculations based on the 1998 LCMS. Note: Provinces are indexed as follows: 1, Central; 2, Copperbelt; 3, Eastern; 4, Luapala; 5, Lusaka; 6, Northern; 7, Northwestern; 8, Southern; 9, Western. Figure 6.1 displays the dynamics of cotton shares. The average cotton share is estimated conditionally on household per capita income. These averages are esti- mated with nonparametric, locally weighted regressions, which are local linear regressions that weigh the data using kernel methods following those of Fan (1992) and Pagan and Ullah (1999). A different nonparametric regression is esti- mated using data for 1996 and 1998.4 Figure 6.1 provides details of the evolution of cotton as a cash crop in rural Zambia. In both 1996 and 1998, the cotton income shares increase at the very bottom of the income distribution and then decline with income. The maximum share is roughly more than 10 percent. At the upper tail, the average share is quite small, about 2 percent. The average share of cotton dynamics is higher in 1998 at the bottom of the distribution (mostly poor households) and at its upper tail. In the middle of the income distribution, the cotton income share is higher in 1996 than in 1998. 164 Poverty and the WTO: Impacts of the Doha Development Agenda Figure 6.1. Dynamics of Cotton Income Shares 12.5 10.0 7.5 cent Per 5.0 2.5 0.0 6 8 10 12 Log per capita income Source: Authors' calculations based on LCMS (1996, 1998). Note: The graph shows the average cotton income shares in total income. These averages are estimated with nonparametric regressions (Fan 1992; Pagan and Ullah 1999). The solid curve corresponds to cotton shares in 1998, and the broken line to the shares in 1996. It is instructive to look at the evolution of cotton shares across the different cotton-producing provinces. Figures 6.2­6.4 plot the nonparametric averages for the Central, Eastern, and Southern provinces, respectively. In the Central province, for instance, cotton income shares in 1998 track the shares in 1996 at the bottom of the distribution but become smaller at the middle. The Central province resembles the pattern at the national level, with higher average shares in 1998 at to bottom and at the top of the distribution, and lower shares at the mid- dle. In contrast, cotton shares in the Southern province are higher across the entire income distribution in 1998 than in 1996. The results in figures 6.2­6.4 indicate a regionally differentiated effect of the reforms on the pattern of income sources in rural Zambia. In particular, the increase in cotton shares in the Southern province is remarkable. This is the region that is perhaps closest to the capital, thus the result indicates that access to markets and infrastructure are key variables in adoption and in the deepening of cotton production. The link between these dynamics of cotton shares and the timing of the cotton reforms is straightforward. It would seem that this indicates that the increase in cotton shares, particularly among the poorest farmers and in the Southern province, can be attributed to those reforms. This, however, does not necessarily The WTO Doha Round, Cotton ... Poverty Trends in Zambia 165 Figure 6.2. Dynamics of Cotton Income Shares, Central Province 15.0 12.5 10.0 cent Per 7.5 5.0 2.5 0.0 6 8 10 12 Log per capita income Source: Authors' calculations based on LCMS (1996, 1998). Note: The graph shows the average cotton income shares in total income. These averages are estimated with nonparametric regressions (Fan 1992; Pagan and Ullah 1999). The solid curve corresponds to cotton shares in 1998, and the broken line to the shares in 1996. Figure 6.3. Dynamics of Cotton Income Shares, Eastern Province 15.0 12.5 10.0 cent Per 7.5 5.0 2.5 0.0 6 8 10 12 Log per capita income Source: Authors' calculations based on LCMS (1996, 1998). Note: The graph shows the average cotton income shares in total income. These averages are estimated with nonparametric regressions (Fan 1992; Pagan and Ullah 1999). The solid curve corresponds to cotton shares in 1998, and the broken line to the shares in 1996. 166 Poverty and the WTO: Impacts of the Doha Development Agenda Figure 6.4. Dynamics of Cotton Income Shares, Southern Province 10.0 7.5 cent Per 5.0 2.5 0.0 6 8 10 12 Log per capita income Source: Authors' calculations based on LCMS (1996, 1998). Note: The graph shows the average cotton income shares in total income. These averages are estimated with nonparametric regressions (Fan 1992; Pagan and Ullah 1999). The solid curve corresponds to cotton shares in 1998, and the broken line to the shares in 1996. follow. There are other factors simultaneously affecting cotton shares. Key factors include the collapse of the copper sector, the adoption of macroeconomic reforms, and the exogenous changes in international cotton prices. Given the available data, it is impossible to disentangle the contribution of these different factors to the observed trends in cotton shares. However, it is very likely that the marketing reforms are the major determinants. The collapse of the copper sector (mainly as a result of a declining trend in international copper prices) is perhaps an urban, rather than rural, phenomenon. In addition, copper is mainly associated with the Copperbelt province (where the copper mines and the industrial belt are located) and with Lusaka (the country's administrative center). It seems, therefore, that cotton-producing provinces would be relatively unaffected by changes in the copper sector. Note, however, that the general equilibrium effects of the dramatic changes in a traditionally impor- tant sector of the Zambian economy cannot be ignored. Similar remarks apply to the adoption of other economic policies. There is one important reform that has to be carefully considered, though. Zambia used to have a maize marketing board that set producer prices for maize grain and con- sumer prices for maize meals. As in cotton, the marketing board was eliminated. The WTO Doha Round, Cotton ... Poverty Trends in Zambia 167 This is a major rural reform, and it should be expected to have significant effects on the allocation of agricultural resources and cotton shares. It is important to note that the maize reforms took place in 1993, well before the 1996­98 period investigated in this chapter. Accordingly, for purposes of the present analysis, it is assumed that the effects of the maize reforms have already taken place in 1996, the baseline period for this chapter.5 To the extent, however, that these reforms have long-lasting effects on farmers, the observed dynamics in figures 6.1­6.4 will cap- ture them. There is another element that favors the role of the reform as a major determi- nant of cotton dynamics. An important observation is that the increase in cotton income shares is larger at the bottom of the distribution of per capita expendi- ture--that is, among the poorest households. If macroeconomic and aggregate shocks, with magnitudes that affect all households simultaneously, are size neu- tral, then differences in the impacts at different points along the income distribu- tion shouldn't be expected. This indicates, at least, that the relative changes in cot- ton dynamics are mostly generated by the marketing reforms. For instance, the larger increase in shares at the bottom of the distribution can be due to expanded access to seeds and fertilizers among the poor. Finally, it is important to have in mind that cotton prices have continuously declined during this period. From 1996 to 1998, in particular, the real price of cot- ton in international markets declined by as much as 20 percent.6 There are two implications for the analysis in this chapter. First, the dynamics in figures 6.1­6.4 show that cotton shares could have been even higher, in the face of the reform, had cotton prices been higher or remained stable. In other words, it would be reason- able to expect larger increases in cotton shares due to the cotton reforms if the change in cotton prices could be controlled for. Second, the decline in the real prices of agricultural products has been similar for other commodities, such as maize, the major alternative crop in rural Zambia. This suggests that the changes in the price of cotton relative to maize have been mild.7 In summary, the available information prevents the econometric identification of the effects of the cotton reforms on the dynamics of cotton shares among Zam- bian farmers. But the observed trends can be mostly linked to those reforms, and the induced increase in cotton shares could have been larger. This is an interesting instance of a domestic policy that had crucial impacts on the farmers and their ability to reap most of the benefits from further trade liberalization. In what follows, this chapter investigates the effects of the complementarities between domestic reforms and international agricultural liberalization on house- hold welfare. As shown in section 1, international cotton markets are subject to strong intervention, particularly by developed countries. Therefore, the analysis begins by looking at the welfare effects of the increase in world prices that would 168 Poverty and the WTO: Impacts of the Doha Development Agenda take place if agricultural markets were liberalized. This involves merging the analysis of cotton shares with the projected increases in world cotton prices. Define the household income as (6.1) yh= hc + hj j where hc is the income (profit) from cotton sales and hj is income from activity j (wages, vegetables, maize, groundnuts, and so on). The change in income caused by an increase in the price of cotton pc is (6.2) dlnyh= hc dlnpc. The proportional change in household income is given by the share of cotton income, hc , multiplied by the proportional change in prices. Section 1 reviewed some evidence indicating that a full liberalization of agri- cultural markets along the lines of the Doha Round negotiations would bring about an increase in world cotton prices of around 12.7 percent. Similar results are obtained if the price increase predicted by Sumner (2004) of 11.6 percent is used. Using the data for 1998, it is possible to estimate the average welfare effect given by equation 6.2. These averages, estimated again with a Fan regression, are shown in figure 6.5. It can be observed that the increase in prices would benefit farmers across the income distribution. The effects would range from more than 0.75 percent to nearly 1.5 percent, at the bottom of the distribution, to roughly 0.5 percent at the top. The unconditional average gain would be about 1 percent of the initial household per capita income. Figure 6.6 explores the regional differences in cotton gains. The solid line cor- responds to the nonparametric average in the Central province; the broken line, to the nonparametric average in the Eastern province; and the dotted line, to the nonparametric averages in the Southern province. The figure shows that larger gains would take place in the Eastern province, particularly at the bottom of the distribution. In the Central province, the gains track the average gains in the Southern province at the bottom of the distribution; from the middle to the top, however, the gains in the Central province remain high, whereas the gains in the Southern province sharply decline with income. These findings show that a 12.7 percent increase in the price of cotton would cause household income to increase by, at most, 1.5 percent (among the poorest households in the Eastern province, for instance). Although these are positive effects associated with welfare gains, it is clear that the magnitudes are quite small. There are several reasons that help explain this fact. The WTO Doha Round, Cotton ... Poverty Trends in Zambia 169 Figure 6.5. Cotton Prices and Household Income 1.5 1.0 Percent 0.5 6 8 10 12 Log per capita income Source: Authors' calculations based on LCMS (1998). Note: The graph shows the average welfare effects (defined as the cotton shares multiplied by the change in world cotton prices) at different levels of household per capita income. The curves are estimated with nonparametric locally weighted regressions (Fan 1992; Pagan and Ullah 1999). Figure 6.6. Cotton Prices and Household Income, Regional Analysis 2.0 1.5 1.0 Percent 0.5 0.0 6 8 10 12 Log per capita income Source: Authors' calculations based on LCMS (1998). Note: The graph shows the average welfare effects (defined as the cotton shares multiplied by the change in world cotton prices) at different levels of household per capita income. The curves are estimated with nonparametric locally weighted regressions (Fan 1992; Pagan and Ullah 1999). The solid, broken, and dotted lines correspond to the Central, Eastern, and Southern provinces, respectively. 170 Poverty and the WTO: Impacts of the Doha Development Agenda As discussed in section 1, the reforms in cotton markets have been relatively successful but have not been smooth. Initially, the public monopoly was trans- formed into a private monopoly. Entry was useful in early stages, but the failure of the out-grower scheme limited the expansion of the sector. In fact, the out-grower scheme was still failing in 1998, when the household survey data were collected. This means that the evolution in income shares that can be captured with the available data does not reveal the whole benefits of the reforms. Using additional farm data, Brambilla and Porto (2005) report that the cotton reforms had two distinctive effects on cotton farming. First, there is a decline of the land area devoted to cotton in 1998­99 (when the out-grower scheme was fail- ing), followed by a significant increase in area planted in 2000­1, when the out- grower scheme was perfected with the entrance of Dunavant. Second, Brambilla and Porto find that farm productivity in cotton showed a similar pattern, declin- ing in 1998­99 and increasing in 2000­1. In addition, their findings indicate that income shares increased, on average, by roughly 10­20 percent. These additional factors could help increase the average gains from a 12 percent price increase to roughly 2 percent of household income, on average. Although these figures are higher, the effects seem still fairly low. The study by Brambilla and Porto (2005) reveals another reason why the increase in cotton prices may not have large impacts. Productivity of smallholder cotton production traditionally has been very low in Zambia. Although there is some evidence that the reforms have caused productivity to increase in recent years, there is still room for improvements in this area. One additional reason for the small impacts is that cotton activities are not really widespread in Zambia. More important, this chapter has considered only a first-order approximation to the welfare gains. The next section captures some supply responses and second-round effects. This chapter explores only the poverty alleviation effect of cotton production in rural areas by smallholders. There are vertical linkages in cotton production that suggest that significant additional poverty effects could be secured through the domestic production of textiles and garment products for exports. These important issues, which deserve further consideration and analysis, are not the focus of this investigation. Cotton Production and Household Outcomes This section explores the impacts of cotton production on household outcomes. If free trade and cotton liberalization bring about renewed incentives for cotton pro- duction in Zambia, farmers could be expected to switch from subsistence to cot- ton production (and, more generally, to market-oriented agriculture). This sec- The WTO Doha Round, Cotton ... Poverty Trends in Zambia 171 tion investigates these supply responses with the help of matching methods: by matching households in subsistence agriculture with households in cotton, the average effects of participating in cotton markets on several household outcomes are estimated. The focus here is on income differentials, child anthropometry, and education outcomes.8 The Method The aim here is to estimate the differences in outcomes linked to the production of a cash crop, such as cotton, and explore the poverty alleviation effects of allow- ing for an expansion of cotton activities among Zambian farmers. Matching methods based on the propensity score are used.9 This approach begins by esti- mating a probit model of participation in cotton, which defines the propensity score p(x), for a given vector of observables x. Subsistence farmers are matched with cotton farmers based on this propensity score, and the outcome differential is estimated using kernel methods. Start with the income gains. Let y m be the income per hectare in cotton of household h. Let-ysh be the home-produced own-consumption per hectare. The h average income differential (per hectare) for those involved in cotton production is defined as (6.3) = E [ y -ysh|C = 1] m h The task is to estimate the counterfactual quantity E [ y -ysh|C = 1], the average m h return in subsistence agriculture among cotton farmers. This is done by using matching methods. The main assumption of matching methods is that the participation in market agriculture can be based on observables. This is the ignorability of treatment assignment. Define an indicator variable C, where C = 1 if the households derive most of their income from cotton. In practice, most Zambian households in rural areas produce something for own-consumption. As a consequence, C = 1 is assigned to households that derive more than 50 percent of their income from cotton. Households that derive most of their income from home production are assigned C = 0. The propensity score p(x) is defined as the conditional probability of participating in cotton, p(x) = P(C = 1|x). The ignorability of treatment assignment requires that y -ysh|C |x. When the m h propensity score is balanced, it can be asserted that, conditional on p(x), the par- ticipation in cotton C and the observables x are independent. In other words, observations with a given propensity score have the same distribution of observ- ables x for households involved in cotton as in subsistence. The importance of the 172 Poverty and the WTO: Impacts of the Doha Development Agenda balancing property, which can be tested, is that it implies that, conditionally on p(x), the returns in cotton and in subsistence are independent of market partici- pation, which implies that households in subsistence and cotton are comparable. The decision to participate in market agriculture depends on three main vari- ables: access to markets, food security and risk, and tradition in subsistence agri- culture. These effects are captured by including in the propensity function several key control variables, such as regional (district) dummies, the size of the house- hold, the demographic structure of the family, the age and the education of the household head, and the availability of agricultural tools. These variables x com- prise a comprehensive set of observables to explain the selection mechanism. Once the propensity score is estimated, the balancing condition is tested. This requires partitioning the estimated p(x) and testing that, within each stratum, the mean and variances of the covariates are not statistically different.10 In the current case of cotton, the balancing property was always satisfied.11 Monetary Outcomes This chapter investigates a constrained model of household agricultural produc- tion. This means that households are assumed to face significant constraints in terms of land, family labor supply, or inputs, and expanding cash crop activities would mean forgone income. In this model, if a family were to plant an additional acre of cotton, then an acre of land devoted to own-consumption (and all other relevant resources) should be released. Table 6.5 reports the results. The first column shows the gains per hectare. In the second column, the constrained household is assumed to expand cotton pro- duction by the average size of the plots devoted to cotton in Zambia. The results indicate that there are gains from cotton production: farmers growing cotton are expected to gain 18,232 kwachas (K) on average, more than similar farmers engaged in subsistence agriculture. The gain is equivalent to 19.9 percent of the average expenditure of a representative poor farmer. To get a better sense of what these numbers mean, notice that the food poverty line in 1998 was estimated at K 32,233 per month and the poverty line at K 46,287 per month (per equivalent adult). Further, because the exchange rate in December 1998 was about K 2,200, the gains are equivalent to just over US$8 (in 1998 prices). The actual gains will depend on the land area allocated to cotton. If farmers are allowed to plant the average size of a typical cotton plot, which is estimated at 1.2 hectares, the estimated gains increase to K 21,878. This is equivalent to 23.9 per- cent of the income of the poor. Notice that because the average size of the land plots allocated to home production ranges from 1.5 to 5 hectares, with an uncon- ditional average of around 2 hectares, it would be feasible for an average house- hold to switch from own-consumption to cotton-growing activities. The WTO Doha Round, Cotton ... Poverty Trends in Zambia 173 Table 6.5. Income Gains from Cotton Production Constrained model Constrained model (per hectare) Monthly % of Monthly % of kwachas expenditure kwachas expenditure Cotton 18,232 19.9 21,878 23.9 (7,456) (8,947) Source: Authors' simulations. Note: Results are from propensity score matching of cotton farmers and subsistence farmers using kernel methods. Standard errors in parentheses are estimated with bootstrap methods. The constrained model (per hectare) assumes that the household has to give up 1 hectare of land to produce an additional hectare of cotton. The constrained model assumes that the farmer moves from subsistence to cotton and allocates the average plot size of cotton farmers (1.2 hectares). The matching results suggest that there might be additional gains from switch- ing to cotton. A natural question is why these opportunities are not exploited by the farmers. Although there are many reasons that can explain this fact, the key role of complementary policies should be emphasized here. Access to interna- tional markets is a basic prerequisite. This requires openness and export-oriented incentives on behalf of Zambia and a liberalization of agricultural markets in developed countries. Price and income support, and export and input subsidies, should be eliminated. Other domestic complementary policies should be imple- mented as well. There are several key policies. Extension services to farmers-- including transmission of information and know-how about cropping, crop diversification, fertilizer and pesticide use, and so on--are critical. The provision of infrastructure to reduce transport and transaction costs is also essential. Irriga- tion may also help. The development of a stronger financial and credit market can also help farmers reduce the costs of the out-grower programs. Finally, education (both formal education and labor discipline) and the provision of better health services will surely help increase farm productivity in cotton. It is generally difficult to assess the role of complementary policies empirically, but some sense of their importance can be gained by looking at evidence reported in the related literature. For example, Brambilla and Porto (2005) find that farm- ers that received extension services are 8.4 percent more productive in cotton than farmers that did not receive any technical assistance. Other things being equal, this would imply an 8.4 percent increase in household well-being. This clearly shows that complementary policies can indeed be useful in improving the living condi- tions of poor farmers in rural Zambia. Of course, a comprehensive assessment of such policies would require an evaluation of the costs of providing these improved services. 174 Poverty and the WTO: Impacts of the Doha Development Agenda In the authors' view, the role of Doha is not only to provide a higher price for cotton, but also to facilitate market access. Complementary policies can help farmers to fully exploit these opportunities. So far, this chapter has explored the effects of increasing prices on household income and provided a quantification of the potential gains of switching from subsistence to cotton. Doha and the comple- mentary policies are considered as vehicles to make these gains feasible. To look at these links more closely, the following experiment is performed. The increase in prices caused by the Doha Round would induce an expansion of quantities pro- duced. If these quantities could be produced and sold by Zambian farmers, then the realization of the gains becomes feasible. To quantify these effects, the analysis proceeds as follows. First, some quantity changes are induced by Doha. This could be estimated, for instance, by multiply- ing the price changes reported in above by an export supply elasticity. As an exam- ple, if this elasticity were 1, an increase in price of 12.7 percent would cause quan- tities to react by 12.7 percent, too.12 Given these quantity changes, the issue is how to allocate them to the different households. To do this, notice that the estimated propensity score indicates the probability of being a cotton producer. One reason- able scenario is thus to allocate the quantity changes on the basis of the relative propensity score. It is important to notice that proceeding in this way allows households in subsistence to switch to some production of cotton. However, this switch can be minor if the relative probability is small for particular farmers. In addition, farmers who are already producing cotton are more likely to have higher estimated probabilities, which can make them better candidates to absorb larger fractions of the export opportunities. Figure 6.7 plots the average relative probability of being a cotton producer across the income distribution. The curve is estimated with nonparametric meth- ods, as before. The relative probability, and therefore the gains from any expansion in quantities, slightly increases with income at the bottom of the distribution and then remains relatively constant. This finding can be interpreted as indicating that everyone across the entire income distribution would benefit about the same from the Doha market opportunities. Figure 6.8 plots the relative probabilities for the three provinces. As before, the solid line corresponds to the Central province; the broken line, to the Eastern province; and the dotted line, to Southern province. Although there are differ- ences in the level of the relative probability across provinces (which should resem- ble the regional differences in the likelihood of cotton production), the distribu- tional effects are the same in all three provinces. No discernible differences across the income distribution can be expected. It is tempting to use this framework to allocate the potential export opportuni- ties brought about by Doha across Zambian farmers and identify households that would actually switch from subsistence to cotton. In addition, it would be possible The WTO Doha Round, Cotton ... Poverty Trends in Zambia 175 Figure 6.7. Relative Probability of Cotton Production 0.036 0.034 Percent 0.032 0.03 6 8 10 12 Log per capita income Source: Authors' calculations based on LCMS (1998). Note: The graph shows the average relative probability of being a cotton producer at different levels of household per capita income. The curves are estimated with nonparametric locally weighted regressions (Fan 1992; Pagan and Ullah 1999). Figure 6.8. Relative Probability of Cotton Production, Regional Analysis 0.08 0.06 Percent 0.04 0.02 0 6 8 10 12 Log per capita income Source: Authors' calculations based on LCMS (1998). Note: The graph shows the average relative probability of being a cotton producer at different levels of household per capita income. The curves are estimated with nonparametric locally weighted regressions (Fan 1992; Pagan and Ullah 1999). The solid, broken, and dotted lines correspond to the Central, Eastern, and Southern provinces, respectively. 176 Poverty and the WTO: Impacts of the Doha Development Agenda to guess how much substitution would take place for different households. To do this, however, would require recourse to ad hoc rules that would dictate the pat- tern of agricultural switching. Although some interesting attempts to do this (in the context of labor markets) are being developed, this chapter reports only the estimated relative probabilities of switching. This approach provides estimates of expected gains, which is as far as one can go with the available data and methods. It has the virtue of being based on estimates derived from econometric models rather than from ad hoc rules. Nevertheless, the estimates can be used to shed additional light on the impacts of Doha and complementary reforms in Zambia. As reported in table 6.5, the esti- mated gain from switching from subsistence agriculture to cotton would be 19.9 percent. This is a measure of the gains from switching, even in the absence of trade reforms. The price effect (roughly a 12.7 percent increase in cotton prices due to trade reforms) would cause the average income of a cotton producer to increase by approximately 1 percent (figure 6.6). If trade reforms induce a switch from subsistence to cotton, then the gains of a switcher would be of about 20.9 percent (19.9 percent due to higher cotton returns plus 1 percent due to higher prices). An expansion of quantities produced and exported (supply responses) would gener- ate additional gains. An example would be the productivity gains of around 8.4 percent (Brambilla and Porto 2005) induced by successful extension services and made feasible by world markets (through Doha). The average income of a cotton producer would increase by 9.4 percent and that of a subsistence farmer by 29.3 percent. It can thus be concluded that, although the price effect of trade reforms (tariffs plus subsidies in cotton) would be generally small, the combination of new market opportunities and domestic reforms (so that switching and productivity gains become viable) can work as very effective vehicles for poverty alleviation. Nonmonetary Outcomes The analysis now turns to the nonmonetary effects of cotton production. The effects on two household outcomes are examined--the nutritional status of infants and young children (from 0 to 60 months old) and education performance of children in primary and secondary school. Malnutrition remains a widespread problem in developing countries, as it does in Zambia. Nutritional status is assessed on the basis of anthropometric indicators (such as height or weight). The three most commonly used anthropometric indi- cators for infants and children are used: weight-for-height, height-for-age, and weight-for-age. Weight-for-height (whz) measures body weight relative to height. It is nor- mally used as an indicator of current nutritional status, and it can be useful for The WTO Doha Round, Cotton ... Poverty Trends in Zambia 177 measuring short-term changes in nutritional status. Extreme cases of low whz rel- ative to a child of the same sex and age in a reference population are commonly referred to as "wasting." Wasting may be the consequence of starvation or severe disease (in particular, diarrhea), but it can also be due to chronic conditions. Height-for-age (haz) reflects cumulative linear growth. Deficits in haz indicate past or chronic inadequacies nutrition, chronic or frequent illness, or both, but cannot measure short-term changes in malnutrition. Extreme cases of low haz are referred to as "stunting." Weight-for-age (waz) reflects body mass relative to age. This is, in effect, a composite measure of haz and whz, making interpretation dif- ficult. The term "underweight" is commonly used to refer to severe or pathologi- cal deficits in waz. A problem arises because weight and height depend on both age and gender (and other factors such as genetic variation), but it is possible to use physical measurements by comparing indicators with the distribution of the same indica- tor for a "healthy" reference group. This chapter uses z-scores (standard deviation scores), the most common way of expressing anthropometric indexes.13 Table 6.6 presents some summary statistics. The value of the mean of the haz z-score is -2.21, reflecting long-term cumula- tive inadequacies of health, nutrition, or both.14 There seems to be no wasting problem: the mean of the whz is 0.23. Using the summary measure of nutritional status (waz), there is mild underweight, probably caused by long-term nutritional problems. For the education outcome, an index of school performance for children between ages 7 and 18 was generated--that is, children in primary and secondary Table 6.6. Child Nutrition in Rural Areas (0­60 months old) Constrained model Constrained model (per hectare) Nutrition Mean Standard Moderate Severe indicator deviation Stunting (haz) -2.21 1.77 23 33 Wasting (whz) 0.23 1.40 5 1 Underweight (waz) -1.21 1.24 20 6 Source: Authors' simulations. Note: Accumulated undernutrition is measured by haz. Levels of current undernutrition are measured by whz. The summary measure of nutritional status is waz. In medicine, the prevalence rate is the proportion of individuals suffering a disease. "Moderate" refers to those individuals with a z-score between -3 and -2, and "severe" refers to a z-score below-3. 178 Poverty and the WTO: Impacts of the Doha Development Agenda school. The index is the ratio of years of education completed by an individual and the years of education this individual should have for his or her age.15 The mean of this index for rural areas is 0.49, including children not attending school (approximately 45 percent of the sample). The analysis now turns to assessing the effects of participation in cotton pro- duction in dimensions other than monetary income. This entails using the same matching methods as before. The effects on child nutrition and education of switching from subsistence to cotton are estimated. Table 6.7 reports the results. Differences in outcomes for the sample of all infants and young children (0 to 60 months old) are estimated for the subsample of males, and for the subsample of females. It is interesting that statistically significant effects are found only in terms of stunting, or long-run nutritional gains. On average, a cotton family would enjoy a higher z-score of 0.64 points. This is equivalent to 30 percent of the average haz z-score for households in subsistence. There are no significant differ- ences in wasting and underweight between cotton and subsistence households. Also, there is no differential effect between females and males, although the mag- nitudes for males are much larger and marginally significant. Similar results have been found in, for example, von Braun and Kennedy (1994). These are very interesting results. They indicate that there are no differences in current nutrition among children living in cotton-producing or subsistence households. However, there are statistically significant benefits in terms of long- term nutrition among those children living on cotton farms. One interpretation is that whereas doing cotton or subsistence allows children to be currently well fed, through the consumption of maize or sweet potatoes for instance, cash income derived from cotton allows farmers to purchase milk, fish, or dairy products that have longer-term benefits. Another hypothesis argues that the movement from subsistence to agricultural commercialization implies a change in the use of fertil- izers and pesticides that helps prevent health hazards and improve the long-term nutritional status of the children.16 In the case of education, our findings indicate that educational outcomes are similar in households involved in cotton and in households involved in subsis- tence. This result holds for the whole population (all children between 7 and 18 years old), children in primary school, and children in secondary school. Conclusions This chapter examines the relationship between cotton reforms and poverty in Zambia. Cotton is one of the main cash crops of smallholders in suitable provinces in rural Zambia. Further, rural poverty is pronounced and widespread. The sector has experienced significant reforms that involved the movement from a publicly controlled parastatal firm to privatization and competition. In this The Table 6.7. Effects on Child Nutrition and Education from Market Agriculture, Cotton versus Subsistence WTO Total Males Females Doha Stunting Wasting Underweight Stunting Wasting Underweight Stunting Wasting Underweight (haz) (whz) (waz) (haz) (whz) (waz) (haz) (whz) (waz) Round, 0.64 -0.004 0.34 1.07 -0.25 0.45 0.14 -0.004 0.07 (0.34) (0.33) (0.24) (0.63) (0.52) (0.45) (0.85) (0.66) (0.37) Cotton Total Males Females ... All Primary Secondary All Primary Secondary All Primary Secondary Poverty -0.02 -0.01 0.01 -0.03 -0.07 -0.02 0.01 -0.01 0.13 (0.04) (0.05) (0.05) (0.04) (0.06) (0.07) (0.05) (0.07) (0.10) Trends Source: Authors' simulations. in Zambia Note: Results are from propensity score matching of cotton farmers and subsistence farmers using kernel methods. Standard errors in parentheses are estimated with bootstrap methods. 179 180 Poverty and the WTO: Impacts of the Doha Development Agenda context, cotton is claimed to be a major market agricultural activity for vulnerable families in rural areas. Two angles of the cotton-poverty connection have been explored. On the one hand, the chapter has presented analysis of the welfare effects that would take place if agricultural cotton markets were liberalized and the world price would thereby increase. On the other hand, the differences in several outcomes between households involved in cotton and households involved in subsistence agriculture have been estimated. The first finding shows that the domestic reforms have caused cotton shares to increase at the bottom of the income distribution. These are poor farmers. Regarding international market access, it is estimated that the increase in world price would benefit cotton producers across the entire income distribution. An estimated 12.7 percent increase in prices would bring about welfare gains reaching roughly 1 percent of household income. In addition, it is found that households involved in cotton enjoy income gains over households involved in subsistence. This implies that a movement from subsistence to market agriculture would ben- efit rural farmers and would lead to a further decline in poverty rates. After world trade reforms, for instance, the welfare gain of a switcher was estimated at approx- imately 21 percent. Further, productivity gains induced by extension services (improved during the marketing reforms) and made feasible by expanded inter- national markets (due to Doha) would lead to welfare gains of 9 percent among cotton producers and 30 percent among switchers. In terms of nonmonetary out- comes, higher long-run nutritional status among children residing in cotton pro- ducing farms is found, but no significant differences in educational attainments. These results highlight promising avenues for poverty alleviation through cash agricultural activities such as cotton. It is important to notice that the estimated magnitudes are relatively small. This shows that to take full advantage of the access to international markets (with a liberalization of world agricultural mar- kets), complementary policies are essential. These policies include extension serv- ices (information), infrastructure (transport), irrigation, access to credit and finance, education, and health services. Notes 1. This section relies heavily on Balat and Porto (forthcoming). 2. For more details on cotton reforms in Zambia, see Food Security Research Project (2000) and Cotton News (2002). 3. The Zambia Central Statistical Office is in charge of conducting the surveys. The data from the 2001 LCMS were under preparation when this chapter was being written. 4. Unfortunately, the 1991 Priority Survey does not include separate data on cotton income. The WTO Doha Round, Cotton ... Poverty Trends in Zambia 181 5. Similarly, the collapse of Zimbabwe has had significant impacts on the Zambian rural economy. It is observed, for instance, that tobacco has been increasingly adopted in Zambia, mainly as a result of the migration of neighboring peasants from Zimbabwe. However, the Zimbabwean crisis took place in recent years, after 1998, and should not affect this analysis. 6. The decline around the long-term trend seems to be, however, lower. See Food Security Research Project (2000). 7. See Food Security Research Project (2000). 8. The estimation of supply responses has proved very difficult. The survey in Winters, McCulloch, and McKay (2004) highlights these issues and reports some of the available methods and results. For the case of income gains, see Lopez, Nash, and Stanton (1995) and Heltberg and Tarp (2002). For non- monetary outcomes, see Edmonds and Pavcnik (Forthcoming). 9. Seminal papers on matching methods include Rubin (1977); Rosenbaum and Rubin (1983); Heckman, Ichimura, and Todd (1997, 1998); Heckman and others (1996); and Dehejia and Wahba (2002). 10. In general, this involves setting up a series of F-tests for the equality of means, for instance. See Dehejia and Wahba (2002) for more details. 11. The balancing property is a minor requirement that is imposed in this procedure. In many applications, the property is not necessarily satisfied. Balat and Porto (forthcoming), for example, found that the balancing did not hold in cases including cassava or sunflowers. Notice that the ignora- bility requirement cannot be tested, which is an assumption of the matching method. 12. See Hoekman, Nicita, and Olarreaga (2004) for a nice attempt along these lines. 13. A z-score is defined as the difference between the value for an individual and the median value of the reference population for the same age or height, divided by the standard deviation of the refer- ence population. 14. The World Health Organization uses a z-score cutoff point of -2 to classify low waz, low haz, and low whz as moderate and severe undernutrition, and -3 to define severe undernutrition. 15. Then, for an individual with no education, the index takes a 0, and if she is in the grade that corresponds to her age, the index takes 1. 16. It is also possible that the sample year is one of relatively good subsistence but a relatively bad cotton season. There is actually no evidence that this was so in 1998, however. References Aksoy, A., and J. Beghin, eds. 2004. Global Agricultural Trade and Developing Countries. Washington, DC: World Bank. Baffes. J. 2004. "Cotton: Market Setting, Trade Policies, and Issues." In Global Agricultural Trade and Developing Countries, ed. A. Aksoy and J. Beghin, 259­274. Washington, DC: World Bank. Balat, J., and G. Porto. Forthcoming. "Globalization and Complementary Policies. Poverty Impacts in Rural Zambia." In Globalization and Poverty, ed. A. Harrison. Chicago: University of Chicago Press. Brambilla, I., and G. Porto. 2005. "Farm Productivity and Market Structure. Evidence from Cotton Reforms in Zambia." Yale University, New Haven, CT. Cotton News. 2002. Cotton Development Trust. Lusaka. Deaton, A. 1989. "Rice Prices and Income Distribution in Thailand: A Non-Parametric Analysis." Eco- nomic Journal 99: 1­37. ------. 1997. The Analysis of Household Surveys. A Microeconometric Approach to Development Policy. Baltimore and Washington DC: Johns Hopkins University Press for the World Bank. Dehejia, R., and S. Wahba. 2002. "Propensity Score Matching Methods for Non-Experimental Causal Studies." Review of Economic Studies 84 (1): 151­61. Edmonds, E., and N. Pavcnik. Forthcoming."The Effects of Trade Liberalization on Child Labor." Jour- nal of International Economics. 182 Poverty and the WTO: Impacts of the Doha Development Agenda Fan, J. 1992. "Design-Adaptive Nonparametric Regression." Journal of the American Statistical Associa- tion 87 (420): 998­1004. Food and Agricultural Policy Research Institute. 2002. "The Doha Round of the World Trade Organi- zation: Liberalization of Agricultural Markets and Its Impacts on Developing Economics," Univer- sity of Missouri, Columbia. Food Security Research Project. 2000. "Improving Smallholder and Agribusiness Opportunities in Zambia's Cotton Sector: Key Challenges and Options." Working Paper 1. Lusaka. Heckman, J., H. Ichimura, J. Smith, and P. Todd. 1996. "Sources of Selection Bias in Evaluating Social Programs: An Interpretation of Conventional Measures and Evidence on the Effectiveness of Matching as a Program Evaluation Method." Proceedings of the National Academy of Sciences 93 (23): 13416­20. Heckman, J., H. Ichimura, and P. Todd. 1997. "Matching as an Econometric Evaluation Estimator: Evi- dence from Evaluating a Job Training Programme." Review of Economic Studies 64 (4): 605­654. ------. 1998."Matching as an Econometric Evaluation Estimator." Review of Economic Studies 65 (2): 261­94. Heltberg, R., and F. Tarp. 2002. "Agricultural Supply Response and Poverty in Mozambique." Food Pol- icy 27 (2): 103­24. Hoekman, B., A. Nicita, and M. Olarreaga. 2004. "Can Doha Deliver the Development Round?" World Bank, Washington, DC. Litchfield, J., and N. McCulloch. 2003. "Poverty in Zambia: Assessing the Impacts of Trade Liberaliza- tion in the 1990s." Poverty Research Unit, Institute for Development Studies, Sussex University, Sussex, U.K. Lopez, R., J. Nash, and J. Stanton. 1995. "Adjustment and Poverty in Mexican Agriculture. How Farm- ers' Wealth Affects Supply Response." Policy Research Working Paper 1494, World Bank, Washing- ton, DC. McCulloch, N., B. Baulch, and M. Cherel-Robson. 2001. "Poverty, Inequality and Growth in Zambia during the 1990s." Presented at the UN World Institute for Development Economics Research Development Conference on Growth and Poverty, Helsinki, May 25­26. Pagan, A., and A. Ullah. 1999. Nonparametric Econometrics. New York: Cambridge University Press. Rosenbaum, P., and D. Rubin. 1983. "The Central Role of the Propensity Score in Observational Stud- ies of Causal Effects." Biometrika 70 (1): 41­55. Rubin, D. 1977. "Assignment to a Treatment Group on the Basis of a Covariate." Journal of Educational Statistics 2 (1): 1­26. Sumner, D. 2004. "Reducing Cotton Subsidies: The DDA Initiative." In Agricultural Trade Reform and the Doha Development Agenda, ed. K. Anderson and W. Martin. Basingstoke, U.K.: Palgrave Macmillan; Washington, DC: World Bank. von Braun, J., and E. Kennedy. eds. 1994. Agricultural Commercialization, Economic Development, and Nutrition. Baltimore and London: John Hopkins University Press for the International Food Policy Research Institute. Winters, A., N. McCulloch, and A. McKay. 2004. "Trade Liberalization and Poverty: The Evidence So Far." Journal of Economic Literature 42: 72­115 World Bank. 1994. Zambia Poverty Assessment. Washington, DC: World Bank. 7 The Doha Round, Poverty, and Regional Inequality in Brazil Joaquim Bento de Souza Ferreira Filho and Mark Horridge Summary This chapter addresses the potential effects of the Doha Round of trade negotia- tions on poverty and income distribution in Brazil, using an applied general equi- librium (AGE) and microsimulation (MS) model of Brazil tailored for income distribution and poverty analysis. It is particularly important that the representa- tive household hypothesis is replaced by a very detailed representation of house- holds. The model distinguishes 10 different labor types and has 270 different household expenditure patterns. Income can originate from 41 different produc- tion activities (which produce 52 commodities) in 27 different regions inside the country. The AGE model communicates to a MS model comprising 112,055 Brazilian households and 263,938 adults. Economic activity in Brazil, a large country, is spread unevenly across the terri- tory. Manufacturing industries are concentrated in the southeastern region, and agriculture, although more evenly distributed geographically, is the main source of income of the center-west states. Poverty, however, is a pervasive phenomenon in the country, which has one of the worse income distributions in the world. The poorest states in Brazil (defined based on the share of population below the poverty line) are concentrated in the northeastern states. 183 184 Poverty and the WTO: Impacts of the Doha Development Agenda Poverty and income distribution indexes are computed over the entire sample of households and persons, before and after the policy shocks. Model results show that even important trade policy shocks, such as those applied in this study, do not generate dramatic changes in the structure of poverty and income distribution in the Brazilian economy. The simulated effects on poverty and income distribution are positive but rather small. The benefits are, however, concentrated in the poor- est households. The study also suggests that the poverty reductions would arise from the income-earning changes and not from the fall in the consumption bundle prices. This outcome is highly correlated to the agriculture and related industries, which have their activity levels increased in all simulations. Finally, the bulk of the poverty impacts can be attributed to the liberalization by other countries, rather than to changes in Brazil's tariff structure. From a methodological point of view, the study emphasizes the need to approach poverty analysis by the household (rather than the personal) dimension, by tracking changes in the labor market from individual workers to households. In the Pesquisa Nacional por Amostragem de Domicílios (National Household Survey [PNAD]; Instituto Brasileiro de Geografia e Estatística [IBGE] 2001) data used here, the head-of-household income accounts for only 65 percent of aggregated household income in Brazil. As a consequence, using head-of- household income as a proxy for household income may poorly predict the effect of policy changes. Introduction One of the most striking aspects of the Brazilian economy is its high degree of income concentration. Despite the changes the economy has faced in the last 20 years--ranging from the country's redemocratization, trade liberalization, hyperinflation, many currency changes, and finally to the macroeconomic stabilization in the mid-1990s--the country still shows one of the worst patterns of income distribution in the world. The resilience of this income distribution problem has attracted the attention of researchers worldwide, and it is the central point of a lively debate in Brazil. The problem is, of course, extremely complex and related to a great number of socioeconomic variables, which makes it a particularly difficult analytical issue, because the effect of many variables upon poverty is uncertain. At the same time, new changes in the external environment face the Brazilian economy. The Doha Round of international trade negotiations may be one of the most important. A complex phenomenon in itself, the economic integration poses new questions relating to the prospects for the poor. This chapter addresses The Doha Round, Poverty, and Regional Inequality in Brazil 185 these questions with a systematic and quantitative approach. For this purpose, an AGE model of Brazil tailored for income distribution and poverty analysis will be used. The model also uses a regional breakdown to examine the associated issue of regional inequality. The next section shows some figures about the problem of poverty and income distribution in Brazil, with a brief review of the recent literature on the topic. Then, the methodological approach is presented with a discussion of the relevant literature on the many different approaches. Next, the model itself is presented with a discussion of its main aspects and of the database. Finally, results and con- clusions are presented. Poverty and Income Distribution Evolution in Brazil: An Overview Although Brazil is a country with a large number of poor people, its population is not among the poorest in the world. Drawing on the 1999 Report on Human Development, Barros, Henriques, and Mendonça (2001) show that about 64 percent of the countries in the world have per capita income less than that in Brazil, a figure that mounts to 77 percent if the number of persons in the same condition is considered. The same authors show that although 30 percent of the total population in Brazil is poor, only 10 percent on average are poor in other countries with similar per capita income. Indeed, based on the same report, the authors define an international norm that, based on per capita income, would impute only 8 percent of poor for Brazil. That is, if the inequality of income in Brazil were to correspond to the world average inequality for countries in the same per capita income range, just 8 percent (rather than 30 percent) of the Brazilian population would be expected to be poor. Taking the concept of poverty in its particular dimension of income insufficiency, Barros Henriques, and Mendonça (2001) show that in 1999, about 14 percent of the Brazilian population lived in households with income below the line of extreme poverty (indigence line, about 22 million people), and 34 percent of the population lived in households with income below the poverty line (about 53 million people). Even though the percentage of poor in the population declined from 40 percent in 1977 to 34 percent in 1999, this level is still very high and seemingly stable. The size of poverty in Brazil, measured either as a percentage of the population or in terms of a poverty gap, stabilized in the second half of the 1980s, although at a lower level than was observed in the previous period. Barros and Mendonça (1997) have analyzed the relations between economic growth and reductions in the level of inequality upon poverty in Brazil. Among their main conclusions, these authors point out that an improvement in the 186 Poverty and the WTO: Impacts of the Doha Development Agenda distribution of income would be more effective for poverty reduction than economic growth alone if growth maintained the current pattern of inequality. According to these authors, as a result of the very high level of income inequality in Brazil, it is possible to dramatically reduce poverty in the country, even without economic growth, by turning the level of inequality in Brazil close to what can be observed in a typical Latin American country. Brazilian poverty also has an important regional dimension. According to calculations by Rocha (1998) in a study of the 1981­95 period, the richer southeastern region of the country, although accounting for 44 percent of total population in 1995, had only 33 percent of the poor. These figures were 15.4 percent for the southern region (8.2 percent of the poor) and 6.8 percent for the center-west region (5.2 percent of the poor). For the poorer regions, on the contrary, the share of population in each region is lower than the share of the poor: 4.6 percent (9.3 percent of the poor) for the northern region and 29.4 percent (44.3 percent of the poor) for the northeastern region, the poorest region in the country. In terms of evolution of regional inequality, Rocha (1998) concludes that no regular trend could be observed in the period. The author also concludes that the yearly observed variations in concentration are mainly related to what happens in the state of São Paulo (southeastern region) and in the northeastern region. This reinforces the position of these two regions in the extremes of the regional income distribution in Brazil. Rocha also points out that once the effects of income increase that followed the end of the hyperinflationary period in 1995 run out, reduction of national and regional poverty will depend mainly on the macroeconomic determinants related to investment. Also, the author concludes that even keeping unchanged the actual level of poverty, reductions in regional inequality will require reallocation of industrial activity to peripheral regions. Finally, Rocha (1998) concludes that opening of the economy to the external market (mainly in relation to the formation of Mercosur) would help reduce regional inequality in Brazil. This would happen through reduced consumer prices in the poorest regions, which are lacking in the industries most threatened by new trade flows. The behavior of wages and the allocation of labor throughout the 1980­99 trade liberalization period in Brazil were analyzed by Green, Dickerson, and Arbache (2001). Among their main findings, the authors point out that wage inequality remained fairly constant for the 1980s and 1990s, with a small peak in the mid-1980s. The main conclusion of the study is that the egalitarian consequences of trade liberalization were not important in Brazil for the period under analysis. As caveats, Green, Dickerson, and Arbache note the low trade exposure of the Brazilian economy (about 13 percent in 1997), as well as the low The Doha Round, Poverty, and Regional Inequality in Brazil 187 share of workers that have completed college studies in total (1 in 12 workers at that time). Gurgel and others (2003) present a computable general equilibrium (CGE) analysis of the effects on Brazil of trade liberalization. They used a Global Trade Analysis Project (GTAP)-derived multicountry model with additional Brazilian detail. For Brazil, 10 urban and 10 rural household income types are recognized. The paper compares the effects of the Free Trade Area of the Americas (FTAA), the EU-Mercosur, and multilateral trade agreements on Brazil. Among the scenarios examined are the effects of the FTAA or the EU-Mercosur if the United States or the EU did not offer free access to Brazilian farm products and the interaction between trade deals--for example, whether the FTAA makes the EU- Mercosur less attractive. Gurgel and others (2003) conclude that the trade deals are in varying degrees good for Brazil--and especially good for Brazil's poor. The poor benefit more because they tend to work in agriculture, which is export oriented and currently suffers from both foreign trade barriers and indirect taxation through the protection of Brazilian manufacturing. Methodology CGE models have long been used for poverty analysis. Many CGE models use a single representative household (RH) to denote consumer behavior in the model. This formulation, although adequate for many purposes, limits the investigation of poverty and income distribution analysis. More recent approaches were developed to deal with these constraints in this study. Savard (2003) provides a thorough discussion of the topic. He groups CGE models dealing with poverty and income distribution analysis into three main categories: models with a single RH, models with multiple households (MH), and the microsimulation (MS) approach that links a CGE model to an econometric household MS model. The RH model is the traditional method, and it has been widely used in the literature. The main drawback of this model for income distribution and poverty analysis is that there are no intragroup income distribution changes because the households are all aggregated into a single representative. The second approach, the MH model, consists of multiplying the number of households. For example, the study by Gurgel and others (2003) distinguished 20 household types. Because they have varying expenditure and income source shares, the households are affected differently by economic changes. However, differences within a particular household group are ignored. Increasing computation capacity allows a large number of households in an MH model. To take an extreme case, the total number of households in a 188 Poverty and the WTO: Impacts of the Doha Development Agenda household survey could be used. This approach then allows the model to take into account the full detail in household data and avoids prejudgment about aggregating households into categories. The main disadvantages of this type of approach are that data reconciliation can be difficult and the size of the model can become a constraint. The third approach, the MS model, draws on microsimulation techniques. Here, a CGE model generates aggregate changes that are later communicated to an MS model based on a large unit-record database. Savard (2003) points out that the drawbacks to the approach are coherence between models, because the causality usually runs from the CGE model to the MS model, with no feedback between them. Overview of the Modeling Approach The approach pursued in this chapter takes advantage of the same general idea raised by Savard (2003) to overcome the difficulties posed by the three first options mentioned above: the use of a CGE model linked to an MS model, but with a bidirectional linkage between them that would guarantee a convergence of solution for both models. Savard (2003) links the models by running them in a repeated sequence of CGE-MS model runs, first computing the CGE simulation, then the MS model simulation, in a looping way, until convergence occurs. The main advantages of this approach are avoidance of scaling the microeconomic data to match the aggregated macroeconomic data, accommodation of more households in the MS model, and the ability of the MS model to incorporate discrete choice or integer behavior that might be difficult to incorporate in the CGE model. The CGE model used here is a static inter-regional model of Brazil based on the well-known ORANI-G model of Australia (Horridge 2000). The model's structure is quite standard: consumption is modeled through the linear expenditure system over composite commodities (domestic and imported), exporters of each commodity face constant-elasticity1 foreign demand schedules, production for exports or domestic markets is regulated by constant elasticity of transformation2 (CET) functions for each firm, production is a nested Leontief­constant elasticity of substitution (CES) structure for primary factors and composite inputs, and labor is a CES function of 10 different types of labor. This nonlinear model is solved with GEMPACK software, and it distinguishes among 42 sectors, 52 commodities,3 and 10 labor occupational categories. All quantity variables in the model are disaggregated according to 27 regions within Brazil, using an elaboration of the top-down regional modeling method described in chapter 6 of Dixon and others (1982). This methodology recognizes The Doha Round, Poverty, and Regional Inequality in Brazil 189 local multiplier effects: many service goods are little traded between regions, so that local service output must follow local demand for services. The CGE model is calibrated with data from the Brazilian economy for 1996, which was obtained from two main sources: the 1996 Brazilian Input-Output Matrix (IBGE http://ibge.gov.br) and the Brazilian Agricultural Census (IBGE 1996a). On the income generation side of the model, workers are divided into 10 different categories (occupations) according to their wages. These wage classes are then assigned to each regional industry in the model. Together with the revenues from other endowments (capital and land rents), these wages are used to generate household incomes. Each activity uses a particular mix of the 10 different labor occupations (skills). Changes in activity level change employment by sector and region. This drives changes in poverty and income distribution. Using the Pesquisa de Orçamentos Familiares (Household Expenditure Survey [POF]; IBGE 1996b) data mentioned below, the CGE model is extended to cover 270 different expenditure patterns, composed of 10 different income classes in 27 regions. In this way, all the expenditure-side detail of the MS dataset is incorporated within the main CGE model. There are two main sources of information for the household MS model: the PNAD (IBGE 2001) and the POF (IBGE 1996b). The PNAD contains information about households and persons, with a total of 331,263 records. The main information extracted from the PNAD was wage by industry and region, as well as other personal characteristics such as years of schooling, sex, age, position in the family, and other socioeconomic characteristics. The POF is an expenditure survey that covers 11 metropolitan regions in Brazil. It was undertaken during 1996 and covered 16,014 households with the purpose of updating the consumption bundle structure. The main information drawn from this survey was the expenditure patterns of 10 different income classes for the 11 regions. One such pattern was assigned to each individual PNAD household, according to each income class. As for the regional dimension, the 11 POF regions were mapped to the larger set of 27 CGE regions. Here, it must be stressed that the POF contains just information about urban areas (the metropolitan areas of the main state capitals). Model Running Procedures and Highlights As noted above, the model consists of two parts, a CGE model and a household model, the MS. The models are run sequentially. There are two strategies to ensure consistency between the two models. First, the CGE model is sufficiently detailed, and its categories and data are close enough to those of the MS model, that the 190 Poverty and the WTO: Impacts of the Doha Development Agenda CGE model very closely predicts MS behavior (which is also included in the CGE model, such as household demands or labor supplies). The role of the MS model is to provide extra information--for example, about the variance of income within income groups or about the incidence of price and wage changes in groups not identified by the CGE model, such as groups identified by ethnic type, educational level, or family status. A second consistency strategy is that, if the MS model predicts household demands or labor supplies at variance with the CGE model, there is the option of feeding back corrections into the CGE model and running the two models iteratively until they agree. That option was not exercised in the simulations reported here.4 The analysis starts with a set of trade shocks generated by a GTAP model simulation that excludes the effect of Brazil's own tariff reductions. These shocks consist of changes in import prices and in export demands. The Brazilian tariff shocks (the trade liberalization in Brazil) are added to these shocks. Import prices and tariffs are naturally exogenous to the Brazil model. The export demand changes are applied via vertical shifts in the export demand curves facing Brazil (table 7.6). The trade shocks are applied and the results calculated for 52 commodities, 42 industries, 10 households, and 10 labor occupations--all of which vary by 27 regions. Next, the results from the CGE model are used to update the MS model. At first, this update consists basically in updating wages and hours worked for the 263,938 workers in the sample. These changes have a regional (27 regions) as well as sectoral (42 industries) dimension. The model then relocates jobs according to changes in labor demand.5 This is done by changing the PNAD weight of each worker to mimic the change in employment--the "quantum weights method."6 In this approach, then, a true job relocation process occurs. Although the job relocation has very little effect on the distribution of wages among the 270 household groups identified by the CGE model, it may have considerable impact on the variance of income within a group because although the jobs move, the workers do not. Thus, regional adjustment is achieved by workers moving into or out of employment. One final point about the procedure used in this chapter should be stressed. Although the changes in the labor market are simulated for each adult in the labor force, the changes in expenditures and in poverty are tracked back to the household dimension. This is possible because the PNAD has a key that links persons to households. Each household contains one or more adults, either working in a particular sector and occupation or unemployed, as well as dependents. In the model used here, then, it is possible to recompose changes in the household income from the changes in individual wages. This is a very The Doha Round, Poverty, and Regional Inequality in Brazil 191 important aspect of the model, because it is likely that family income variations are cushioned, in general, by this procedure. If, for example, one person in some household loses his job but another in the same household gets a new job, household income may change little. Because households are the expenditure units in the model, household spending variations would be expected to be smoothed by this income-pooling effect. However, the loss of a job will increase poverty more if the displaced worker is the sole earner in a household. The Base Year Picture This section extends the above description of poverty and income inequality in Brazil. The reference year for the analysis is 2001. Some general aggregated information about poverty and income inequality in Brazil can be seen in table 7.1. The rows of table 7.1 correspond to household income classes, grouped according to POF definitions7 such that POF[1] is the lowest income class and POF[10] the highest. A fair picture of income inequality in Brazil emerges from the table. It can be seen that the first five income classes, although accounting for 52.6 percent of total population in Brazil, get only 17 percent of total income. The highest income class accounts for 11 percent of the population and about 45 Table 7.1. Poverty and Income Inequality in Brazil, 2001 Income Ave Unemp Pr Ave Pr group PrPop PrInc HouInc Rate White Wage Child POF[1] 10.7 0.9 0.1 32.6 35.2 0.2 46.2 POF[2] 8.0 1.8 0.4 17.3 38.3 0.3 37.2 POF[3] 16.0 5.2 0.6 10.4 42.0 0.4 35.1 POF[4] 7.3 3.1 0.8 8.8 45.1 0.4 32.5 POF[5] 11.0 5.8 1.0 7.5 49.2 0.5 28.7 POF[6] 7.9 5.1 1.2 7.4 53.4 0.6 26.4 POF[7] 12.9 11.1 1.7 6.8 60.3 0.8 24.5 POF[8] 7.5 8.7 2.3 6.1 66.3 0.9 21.5 POF[9] 7.7 12.7 3.1 5.9 71.2 1.4 20.5 POF[10] 10.9 45.7 7.9 4.2 81.6 3.2 17.7 Total 100.0 100.0 n.c. n.c. n.c. n.c. n.c. Source: IBGE (2001). Note: n.c. = not computed. PrPop = % in total population; PrInc = % in country total income; AveHouInc = average household income; UnempRate = unemployment rate; PrWhite = % of white population in total; AveWage = average normalized wage; PrChild = share of population under 15 by income class. 192 Poverty and the WTO: Impacts of the Doha Development Agenda percent of total income. The Gini index associated with the income distribution in Brazil in 2001, calculated using an equivalent household8 basis, is 0.58, placing Brazil's income distribution among the world's worst. The unemployment rate is also relatively higher among the poorer classes. This is a very important point because of its relevance for modeling. The opportunity to get a new job is probably the most important element lifting people out of poverty: hence the importance for poverty modeling of allowing the model to capture the existence of a switching regime (from unemployment to employment) and not just changes in wages. As can be seen in table 7.1, the unemployment rate reaches 36.5 percent among the lowest-income group (persons older than 15 years) and just 7.7 percent among the richest. The percentage of white people also increases considerably with household income, and the percentage of children decreases markedly. Although the analysis does not specifically focus on gender and ethnic aspects, these are important indicators to take into account when analyzing results. To further describe the state of income insufficiency in Brazil, a poverty line defined as one-third of the average household income was set.9 According to that criterion, 30.8 percent of the Brazilian households in 2001 would be poor.10 This would make up 96.2 percent, 76.6 percent, and 53.5 percent, respectively, of households in the first three income groups,11 or 34.5 million of 112 million households in 2001. The first columns in table 7.11 report two overall measures of poverty following Foster, Greer, and Thorbecke ([FGT] 1984)--FGT0, the proportion of poor households (that is, living below the poverty line), and FGT1, the average poverty gap ratio (proportion by which household income falls below the poverty line), and so forth--for each POF group.12 These figures reveal a large average poverty gap for the two lowest income classes. Together, these two income classes contribute to about half of the general average poverty gap index of the economy. The first income class, for example, falls below the poverty line by about 70 percent. Thus, large income increases for the poor are needed to significantly change the number in poverty. As stated before, this general poverty and inequality picture also has an important regional dimension in Brazil because economic activity is located mainly in the southeastern region. This is particularly true of manufacturing; agriculture is more dispersed among regions. Figure 7.1 summarizes the regional variation of poverty and income inequality, shading them according to proportions of households in poverty. The states in the northeastern region plus the states of Tocantins and Pará in the north show the highest poverty rates. If, however, regional population is taken into account, the populous regions of The Doha Round, Poverty, and Regional Inequality in Brazil 193 Figure 7.1. Brazilian States, Shaded According to Proportion in Poverty Roraima Amapa Amazonas Para Maranhao Ceara RGNorte Paraiba Piaui Pernambuco Acre Alagoas Tocantins Sergipe Rondonia Bahia MtGrosso DF Goias MinasG EspSanto MtGrSul SaoPaulo RioJaneiro Parana 0.14 (minimum) StaCatari 0.24 0.35 (median) 0.51 RGSul 0.58 (maximum) Proportion below poverty line Source: Authors' calculations from PNAD 2001 data (IBGE, 2001). 194 Poverty and the WTO: Impacts of the Doha Development Agenda Ceará, Pernambuco, Bahia, Minas Gerais, and São Paulo play a larger role in the overall poverty picture.13 Tables 7.2 and 7.3 report important information about the labor structure of the Brazilian economy. In these tables, sectoral wage bills are split into the model's 10 occupational groups. The occupational groups are defined in terms of a unit wage ranking. More skilled workers, then, would be those in the highest income classes and vice versa. As can be seen in Table 7.2, agriculture is the activity that uses more unskilled labor (40.5 percent of that sector's labor bill), and petroleum and gas extraction and petroleum refinery are the most intensive users of skilled labor (10th labor class) activities, with financial institutions coming next. If labor inputs were measured in hours (rather than in values), the concentration of low- skill labor in agriculture would be even more pronounced. Agriculture is also the sector that hires the most unskilled labor in Brazil, about 41 percent of total workers in income class 1 (table 7.3). The trade sector is the second largest employer of this type of labor. As for the higher income classes, the Table 7.2. Share of Occupations in Each Activity's Labor Bill (percentage) Occupations (wage class) Sector 1 2 3 4 5 6 7 8 9 10 Total Agriculture 40.5 30.2 5.8 6.0 5.2 3.3 3.7 1.8 1.9 1.6 100 Mineral extraction 12.0 19.4 6.8 6.9 8.4 6.1 12.8 9.9 10.8 6.9 100 Petroleum and gas extraction 0.0 0.0 0.0 0.9 0.9 6.1 16.1 12.1 22.8 41.1 100 Non-metallic minerals 7.1 18.8 7.4 8.9 11.5 11.8 14.1 7.6 7.4 5.3 100 Iron production 1.9 6.8 4.0 6.3 10.2 9.7 22.7 14.0 15.4 9.1 100 Non-ferrous metals 1.9 6.8 4.0 6.3 10.2 9.7 22.7 14.0 15.4 9.1 100 Other metals 1.9 6.8 4.0 6.3 10.2 9.7 22.7 14.0 15.4 9.1 100 Machinery and tractors 0.5 4.6 1.9 4.8 6.8 9.0 19.6 17.2 16.8 18.8 100 Electric materials 0.4 3.8 2.6 3.3 10.3 11.6 20.4 15.5 17.0 15.1 100 The Doha Round, Poverty, and Regional Inequality in Brazil 195 Table 7.2. (Continued) Occupations (wage class) Sector 1 2 3 4 5 6 7 8 9 10 Total Electronic equipment 0.4 3.8 2.6 3.3 10.3 11.6 20.4 15.5 17.0 15.1 100 Automobiles 0.3 2.5 1.0 2.4 7.7 8.6 19.6 15.7 22.4 19.8 100 Other vehicles and spare parts 0.3 2.5 1.0 2.4 7.7 8.6 19.6 15.7 22.4 19.8 100 Wood and furniture 8.2 11.7 6.6 8.8 12.4 11.9 16.6 9.3 9.6 5.0 100 Paper 2.3 7.8 3.7 6.2 8.4 8.1 18.7 13.0 16.7 15.1 100 Rubber manufactures 0.8 4.7 3.2 4.6 14.4 5.5 24.0 13.6 16.6 12.5 100 Chemicals 2.1 7.8 3.0 4.2 9.1 11.8 14.2 15.6 16.4 15.8 100 Petrol refining 0.5 1.5 2.7 0.3 9.0 5.7 13.1 7.2 10.5 49.5 100 Miscellaneous chemicals 0.0 6.8 9.6 13.4 25.3 0.0 14.5 2.8 7.9 19.7 100 Pharmaceutical products 1.7 5.7 3.1 6.8 4.1 7.5 13.5 11.3 18.7 27.4 100 Plastics 1.6 6.3 2.3 8.5 12.8 12.1 24.6 10.3 9.0 12.6 100 Textiles 14.7 9.0 4.9 7.2 12.5 11.0 17.6 11.3 6.2 5.5 100 Apparel 3.2 17.3 7.5 15.1 16.1 9.7 15.7 5.4 4.5 5.5 100 Footwear 4.1 16.2 6.5 13.5 18.2 13.0 14.4 5.7 4.8 3.6 100 Coffee processing 8.6 14.3 6.1 9.6 13.2 11.3 15.1 8.3 7.4 6.0 100 Vegetable processing 8.6 14.3 6.1 9.6 13.2 11.3 15.1 8.3 7.4 6.0 100 Slaughtering 8.6 14.3 6.1 9.6 13.2 11.3 15.1 8.3 7.4 6.0 100 Dairy 8.6 14.3 6.1 9.6 13.2 11.3 15.1 8.3 7.4 6.0 100 Sugar refining 8.6 14.3 6.1 9.6 13.2 11.3 15.1 8.3 7.4 6.0 100 Vegetable oils 8.6 14.3 6.1 9.6 13.2 11.3 15.1 8.3 7.4 6.0 100 Other food 8.6 14.3 6.1 9.6 13.2 11.3 15.1 8.3 7.4 6.0 100 196 Poverty and the WTO: Impacts of the Doha Development Agenda Table 7.2. Share of Occupations in Each Activity's Labor Bill (percentage) (Continued) Occupations (wage class) Sector 1 2 3 4 5 6 7 8 9 10 Total Miscellaneous industries 16.8 13.4 6.6 6.2 11.4 7.4 13.1 7.8 10.7 6.5 100 Public utilities 1.7 17.5 5.3 8.6 7.1 6.0 12.9 12.2 14.2 14.5 100 Construction 6.3 13.4 8.6 10.1 12.5 9.0 20.2 9.6 6.9 3.4 100 Trade 10.0 14.2 6.6 8.2 10.7 8.2 15.1 8.3 10.0 8.7 100 Transport 4.6 7.0 4.4 4.7 7.5 7.1 19.0 16.1 18.1 11.6 100 Communi- cations 1.4 4.6 2.4 5.1 7.9 9.4 18.6 13.9 17.2 19.4 100 Financial institutions 0.9 3.5 1.3 3.5 6.6 4.2 10.0 11.8 23.3 34.9 100 Farm services 16.4 20.3 7.4 8.4 9.6 6.8 12.1 6.5 7.2 5.4 100 Enterprise services 2.9 8.1 4.3 5.7 8.1 6.4 13.0 8.6 15.7 27.2 100 Building rentals 2.0 4.3 2.7 4.8 9.9 6.3 17.1 8.8 18.4 25.7 100 Public administration 1.7 13.1 3.6 7.2 7.6 6.8 13.0 12.1 19.3 15.6 100 Other private Services 7.6 16.6 6.0 9.2 9.3 10.9 13.7 8.2 11.6 6.9 100 Source: Authors' computations. financial institutions and public administration sectors hire the largest numbers of well-paid workers. Table 7.4 shows the distribution of occupational character class (OCC) wages among the household income classes (POF classes). In this table, the rows show household income classes, and the columns show the wage earnings by occupation. It is evident from this table that the wage earnings of the higher wage occupations (OCC10, for example) are concentrated in the higher income households and vice versa. Most of the wages earned by workers in the first wage class (OCC1) accrue to the three poorest households, POF[1]­[3]. All the workers in the highest wage class are in households from the eighth income class and above. It can be seen, then, that the household income classes are highly positively correlated with the occupational wage-earning classes. The Doha Round, Poverty, and Regional Inequality in Brazil 197 Table 7.3. Share of Each Activity in Total Labor Bill, by Occupation (percentage) Occupations (wage class) Sector 1 2 3 4 5 6 7 8 9 10 Agriculture 41.0 17.8 9.8 6.9 4.8 3.8 2.2 1.4 1.1 0.9 Mineral extraction 0.5 0.4 0.4 0.3 0.3 0.3 0.3 0.3 0.2 0.1 Petroleum and gas extraction 0.0 0.0 0.0 0.0 0.0 0.1 0.2 0.2 0.3 0.5 Non-metallic minerals 0.5 0.8 0.9 0.8 0.8 1.0 0.6 0.5 0.3 0.2 Iron production 0.1 0.1 0.2 0.2 0.3 0.3 0.4 0.3 0.3 0.2 Non-ferrous metals 0.0 0.1 0.1 0.1 0.2 0.2 0.2 0.2 0.1 0.1 Other metals 0.3 0.7 1.2 1.3 1.7 1.9 2.4 2.0 1.5 0.9 Machinery and tractors 0.1 0.5 0.5 0.9 1.1 1.7 2.0 2.3 1.6 1.8 Electric materials 0.0 0.1 0.2 0.2 0.5 0.7 0.7 0.7 0.5 0.5 Electronic equipment 0.0 0.1 0.2 0.2 0.4 0.6 0.5 0.5 0.4 0.4 Automobiles 0.0 0.1 0.1 0.1 0.3 0.4 0.5 0.5 0.5 0.5 Other vehicles and spare parts 0.0 0.2 0.2 0.3 0.8 1.1 1.3 1.3 1.4 1.2 Wood and furniture 0.9 0.7 1.1 1.0 1.2 1.4 1.0 0.8 0.6 0.3 Paper 0.3 0.6 0.8 0.9 1.0 1.2 1.4 1.3 1.2 1.1 Rubber manufactures 0.0 0.1 0.1 0.1 0.3 0.1 0.3 0.2 0.2 0.1 Chemicals 0.1 0.1 0.2 0.1 0.3 0.4 0.3 0.4 0.3 0.3 Petrol refining 0.0 0.1 0.3 0.0 0.5 0.4 0.5 0.3 0.4 1.7 Miscellaneous chemicals 0.0 0.3 1.1 1.0 1.6 0.0 0.6 0.2 0.3 0.8 198 Poverty and the WTO: Impacts of the Doha Development Agenda Table 7.3. Share of Each Activity in Total Labor Bill, by Occupation (percentage) (Continued) Occupations (wage class) Sector 1 2 3 4 5 6 7 8 9 10 Pharmaceutical products 0.1 0.2 0.3 0.4 0.2 0.5 0.5 0.5 0.6 0.9 Plastics 0.1 0.2 0.2 0.5 0.6 0.7 0.8 0.4 0.3 0.4 Textiles 0.7 0.2 0.4 0.4 0.5 0.6 0.5 0.4 0.2 0.1 Apparel 0.3 0.9 1.1 1.5 1.3 1.0 0.8 0.4 0.2 0.3 Footwear 0.2 0.4 0.4 0.6 0.7 0.6 0.3 0.2 0.1 0.1 Coffee processing 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.0 0.0 Vegetable processing 0.5 0.4 0.5 0.6 0.6 0.7 0.5 0.3 0.2 0.2 Slaughtering 0.4 0.3 0.4 0.5 0.5 0.5 0.4 0.3 0.2 0.1 Dairy 0.1 0.1 0.1 0.2 0.2 0.2 0.1 0.1 0.1 0.0 Sugar refining 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.1 0.1 0.1 Vegetable oils 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.0 0.0 Other food 1.0 1.0 1.2 1.2 1.4 1.5 1.0 0.7 0.5 0.4 Miscellaneous industries 0.7 0.3 0.5 0.3 0.5 0.4 0.3 0.3 0.3 0.2 Public utilities 0.5 3.2 2.8 3.0 2.0 2.1 2.4 3.0 2.5 2.6 Construction 2.7 3.3 6.1 4.8 4.9 4.3 5.0 3.2 1.6 0.8 Trade 13.5 11.2 14.8 12.6 13.3 12.5 12.0 8.7 7.5 6.6 Transport 2.6 2.3 4.1 3.0 3.8 4.4 6.2 7.0 5.6 3.6 Communi- cations 0.2 0.4 0.6 0.8 1.0 1.5 1.6 1.6 1.4 1.6 Financial institutions 1.0 2.3 2.4 4.4 6.9 5.3 6.7 10.5 14.6 22.3 Farm services 21.0 15.1 15.8 12.1 11.2 9.8 9.0 6.5 5.1 3.9 Enterprise services 1.6 2.6 4.0 3.6 4.1 4.0 4.2 3.8 4.8 8.5 The Doha Round, Poverty, and Regional Inequality in Brazil 199 Table 7.3. (Continued) Occupations (wage class) Sector 1 2 3 4 5 6 7 8 9 10 Building rentals 0.1 0.2 0.3 0.3 0.6 0.4 0.6 0.4 0.6 0.9 Public administration 6.4 29.4 23.3 31.2 26.7 29.3 29.2 36.3 40.8 33.7 Other private Services 2.2 2.8 2.9 3.0 2.4 3.5 2.3 1.8 1.8 1.1 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Authors' computations. Model Closure In choosing a model closure, the analysis in this chapter aims to mimic the GTAP model that generated the foreign price scenario. On the supply side, total national employment was fixed by occupation, with jobs moving freely among sectors and regions.14 The model allows substitution between occupations, driven by relative wages. Similarly, capital is fixed nationally but is mobile between sectors and regions. The land stock (used just in the agriculture activity) is fixed.15 Because agriculture is an activity that produces 11 products, land is allocated to these com- peting products through relative prices, allowing the crop mix to change. On the demand side, government and investment spending are fixed in real terms, and a fixed trade balance enforces the national budget balance, which is accommodated by changes in real consumption. The trade balance, then, drives the level of absorption. As noted previously, the numeraire is the Brazilian consumer price index, so all prices reported here are relative to the CPI. Finally, tax revenue losses because tariff cuts are replaced: real aggregate revenue from all indirect taxes is kept fixed via a uniform endogenous change in the power of indirect taxes on sales to households. This mechanism is equivalent to a lump-sum tax of value proportional to each household's spending.16 It also mimics the traditional method of raising tax revenues in Brazil, through indirect tax collection. 200 Table 7.4. Wage Bill Distribution According to Occupational Wages and Household Income Classes (1996 Brazilian reais, millions) Household Occupational wages classes (personal) Poverty income classes OCC1 OCC2 OCC3 OCC4 OCC5 OCC6 OCC7 OCC8 OCC9 OCC10 Total and POF[1] 1,531 1,637 0 0 0 0 0 0 0 0 3,168 the POF[2] 538 2,409 1,632 783 0 0 0 0 0 0 5,362 WTO: POF[3] 1,804 3,996 1,201 2,460 4,327 3,728 342 0 0 0 17,859 POF[4] 766 1,513 861 1,380 1,077 616 5,020 0 0 0 11,233 Impacts POF[5] 932 2,787 1,147 1,649 2,746 2,254 5,945 3,526 0 0 2,0985 POF[6] 537 1,811 795 1,410 2,133 2,127 4,305 5,517 405 0 19,039 POF[7] 576 2,315 1,178 2,012 3,038 3,102 8,717 7,654 12,773 0 41,365 of POF[8] 201 1,137 524 1,045 1,819 1,969 4,896 5,585 13,211 1,427 31,814 the POF[9] 123 695 401 762 1,312 1,449 4,571 5,218 15,864 16,994 47,388 Doha POF[10] 83 527 301 576 1,135 1,185 3,939 5,086 18,480 134,499 165,811 Total 7,091 18,827 8,040 12,077 17,586 16,430 37,734 32,586 60,732 152,920 364,024 Development Source: Authors' computations. Agenda The Doha Round, Poverty, and Regional Inequality in Brazil 201 Results The CGE Model Results The Brazilian economy has a limited exposure to external trade. The shares of exports and imports in total GDP were 7.0 percent and 8.9 percent, respectively, in the 1996 base year. These shares have increased recently, but not by enough to sig- nificantly change this picture.17 Table 7.5 shows more information about the struc- ture of Brazilian external trade as well as of related parameters and production structure; table 7.6 shows the nature and size of the shocks applied to the model. As stated before, the shocks applied to the model were generated by a previous run of the GTAP model in which the Doha scenarios were implemented. The GTAP effects on the Brazilian economy were then transmitted to the Brazil CGE model through changes in tariffs and import prices changes and shifts in the demand schedules for the Brazilian exports.18 An inspection of tables 7.5 and 7.6 can give an idea of the importance of these shocks combined with the importance of each commodity in Brazilian external trade. Brazilian exports are spread among many different commodities, with no specialized trend. Raw agricultural products have a very small share in total exports, composed almost entirely of soybeans. Processed food and agriculturally based exports (including wood and furniture, rubber, paper, textiles, and apparel), however, account for a significant 0.369 share of total exports in the base year, highlighting the importance of agriculture in the Brazilian economy. Imports as a share of each domestic production are concentrated in wheat, oil, machinery, electric materials, electronic equipment, and chemical products. In terms of total import shares, however, oil products (raw and refined), machinery, electric materials, electronic equipment, and chemical products are the most important products. Table 7.5 also shows some relevant parameters and other production charac- teristics of the model. The Armington elasticities are borrowed from the GTAP database. The same is true for the export demand elasticities (not shown in the table), which are made equal to the GTAP region-generic elasticity of substitution among imports in the Armington structure. The agriculture sector is modeled as a multiproduction sector, producing 11 commodities. Thus, the capital to labor ratio (a ratio of values) in table 7.5 is the same for every agricultural product. The value of land is not included in the value of capital here. If land were included, the value of the capital to labor ratio in agri- culture would rise to 0.99. The value added and value of production column, however, includes the returns to land for agriculture. The presentation of Brazilian impacts due to multilateral trade reform begins with some macro results to establish a benchmark for the regional and 202 Table 7.5. Brazilian External Trade Structure External trade Production Poverty Value Share in Exported Import added and Commodity Armington total Brazilian share of share in Share in Capital to value of and elasticities exports total output local markets total imports labor ratio production the Coffee 2.38 0 0 0 0 0.64 0.61 WTO: Sugar cane 2.2 0 0 0 0 0.64 0.61 Paddy rice 2.2 0 0 0.02 0.001 0.64 0.61 Impacts Wheat 2.2 0 0 0.68 0.020 0.64 0.61 Soybeans 2.2 0.019 0.170 0.06 0.004 0.64 0.61 Cotton 2.2 0 0 0.02 0 0.64 0.61 of Corn 2.2 0.001 0.015 0.01 0.001 0.64 0.61 the Livestock 2.8 0 0 0.01 0.001 0.64 0.61 Doha Raw milk 2.2 0 0 0 0 0.64 0.61 Poultry 2.8 0 0.002 0.01 0 0.64 0.61 Development Other agriculture 2.38 0.022 0.019 0.02 0.015 0.64 0.61 Mineral extraction 2.8 0.059 0.398 0.09 0.006 0.44 0.28 Petroleum and gas extraction 2.8 0 0.002 0.41 0.063 4.19 0.51 Non-metallic minerals 2.8 0.014 0.033 0.04 0.009 1.58 0.38 Agenda Iron production 2.8 0.073 0.154 0.03 0.009 2.99 0.18 Non-ferrous metals 2.8 0.041 0.196 0.1 0.014 2.99 0.23 Other metals 2.8 0.018 0.037 0.06 0.018 0.32 0.36 Machinery and tractors 5.2 0.038 0.077 0.22 0.088 1.53 0.56 Electric materials 2.8 0.027 0.086 0.19 0.040 0.86 0.27 Table 7.5. (Continued) External trade Production Value Share in Exported Import added and Commodity Armington total Brazilian share of share in Share in Capital to value of The elasticities exports total output local markets total imports labor ratio production Doha Electronic equipment 2.8 0.018 0.047 0.36 0.123 3.04 0.38 Automobiles 5.2 0.029 0.057 0.1 0.034 2.60 0.25 Round, Other vehicles and spare parts 5.2 0.068 0.144 0.2 0.057 0.69 0.30 Wood and furniture 2.8 0.026 0.078 0.02 0.004 0.66 0.40 Poverty Paper 1.8 0.032 0.067 0.06 0.018 0.45 0.28 Rubber industry 1.9 0.012 0.071 0.1 0.010 2.41 0.32 ,and Chemicals 1.9 0.016 0.066 0.15 0.032 3.61 0.35 Petrol refining 1.9 0.031 0.034 0.11 0.083 6.08 0.31 Regional Miscellaneious chemicals 1.9 0.015 0.039 0.1 0.028 1.11 0.28 Pharmaceuticals 1.9 0.007 0.021 0.15 0.028 1.84 0.46 Inequality Plastics 1.9 0.004 0.021 0.07 0.010 1.46 0.43 Textiles 2.2 0.020 0.052 0.11 0.031 1.98 0.26 Apparel 4.4 0.003 0.011 0.03 0.005 0.37 0.38 in Footwear 4.4 0.043 0.294 0.10 0.006 0.71 0.35 Brazil Coffee processing 3.1 0.033 0.237 0 0 2.64 0.21 Vegetable processing 2.2 0.058 0.105 0.04 0.012 1.69 0.22 Slaughtering 2.2 0.025 0.055 0.02 0.004 1.45 0.19 203 Dairy 2.2 0.001 0.003 0.05 0.007 2.99 0.22 204 Table 7.5. Brazilian External Trade Structure (Continued) External trade Production Poverty Value Share in Exported Import added and Commodity Armington total Brazilian share of share in Share in Capital to value of and elasticities exports total output local markets total imports labor ratio production the Sugar refining 2.2 0.029 0.217 0 0 0.32 0.16 WTO: Vegetable oils 2.2 0.065 0.229 0.04 0.006 2.72 0.11 Other food 2.2 0.022 0.029 0.05 0.020 1.03 0.27 Impacts Miscellaneous industries 2.8 0.010 0.049 0.22 0.028 1.22 0.43 Public utilities 1.9 0 0 0.03 0.014 0.91 0.59 of Construction 1.9 0 0 0 0 4.06 0.66 the Trade 1.9 0.009 0.016 0.01 0.011 0.18 0.53 Doha Transport 1.9 0.053 0.084 0.04 0.022 0.19 0.49 Communications 1.9 0.005 0.014 0.01 0.003 1.97 0.78 Development Financial institutions 1.9 0.007 0.006 0.01 0.006 0.23 0.64 Farm services 1.9 0.016 0.010 0.05 0.067 0.36 0.67 Enterprise services 2.1 0.019 0.027 0.05 0.029 0.52 0.72 Building rentals 1.9 0 0 0 0 51.56 0.95 Agenda Public administration 1.9 0.010 0.003 0.01 0.012 0.00 0.73 Other services 2.1 0 0 0 0 0.01 0.93 Source: Authors' computations The Doha Round, Poverty, and Regional Inequality in Brazil 205 Table 7.6. Shocks to the CGE Model Implied Import tariffs Import CIF prices export price shifta Full Full Full liberali- liberali- liberali- Commodity Doha zation Doha zation Doha zation Coffee -0.04 -6.43 0.74 1.92 -0.74 -0.73 Sugar cane 0 -4.99 1.02 1.80 7.73 9.65 Paddy rice 0 -0.17 2.8 6.47 7.58 38.41 Wheat 0 -0.12 1.95 8.49 0.94 -1.80 Soybeans 0 -0.09 2.54 5.92 3.90 15.49 Cotton 0 -5.55 2.45 4.26 5.37 18.13 Corn 0 -0.55 2.41 7.56 6.32 25.24 Livestock 0 -0.37 1.05 2.40 0.24 -4.50 Raw milk 0 0 0.73 -0.26 -1.11 -9.08 Poultry 0 -4.53 0.45 1.9 0.47 0.39 Other agriculture -0.04 -6.43 0.74 1.92 -0.74 -0.73 Mineral extraction 0 -2.95 0.16 0.12 0.48 1.40 Petroleum and gas extraction 0 0 0.14 0.6 0.20 1.70 Non-metallic minerals -0.01 -9.82 0.13 0.26 0.78 2.76 Iron production -0.07 -10.72 0.04 0.19 0.25 0.88 Non-ferrous metals -0.23 -7.57 0.03 -0.27 0.80 1.70 Other metals -0.04 -14.25 -0.01 0.13 0.45 1.76 Machinery and tractors -0.02 -2.59 -0.17 -0.27 -0.09 -0.45 Electric materials -0.1 -10.92 -0.02 0.05 0.19 0.36 Electronic equipment -0.01 -10.84 0 0.05 0.28 0.67 Automobiles -2.14 -16.91 0.24 -0.16 0.53 5.13 Other vehicles and spare parts -0.02 -2.59 -0.17 -0.27 -0.09 -0.45 Wood and furniture -0.84 -11.81 0.06 0.24 0.49 1.54 Paper 0 -8.54 0 -0.04 0.21 0.28 Rubber industry -0.28 -7.98 0 -0.25 0.35 0.30 206 Poverty and the WTO: Impacts of the Doha Development Agenda Table 7.6. Shocks to the CGE Model (Continued) Implied Import tariffs Import CIF prices export price shifta Full Full Full liberali- liberali- liberali- Commodity Doha zation Doha zation Doha zation Chemicals -0.28 -7.98 0 -0.25 0.35 0.30 Petrol refining 0 -0.41 0.14 -0.31 0.45 2.65 Miscellaneous chemicals -0.28 -7.98 0 -0.25 0.35 0.30 Pharmaceu- ticals -0.28 -7.98 0 -0.25 0.35 0.30 Plastics -0.28 -7.98 0 -0.25 0.35 0.30 Textiles 0 -13.6 0.65 0.33 1.34 0.79 Apparel 0 -17.18 1.00 0.25 1.46 -0.67 Footwear -0.14 -11.64 0.43 0.21 0.26 -0.32 Coffee processing -0.05 -16.54 0.25 0.2 1.50 1.66 Vegetable processing -0.21 -7.66 0.74 0.46 2.20 10.32 Slaughtering 0 -4.02 2.17 2.91 18.02 38.79 Dairy -0.02 -6.39 4.43 6.74 7.56 15.41 Sugar refining 0 -13.18 5.22 5.93 4.30 14.73 Vegetable oils 0 -7.18 0.88 3.39 3.50 -0.70 Other food -0.21 -7.66 0.74 0.46 2.20 10.32 Miscellaneous industries -0.05 -15.39 0.05 0.1 0.11 -0.15 Public utilities 0 0 -0.05 0.07 -0.08 -0.32 Construction 0 0 0.03 0.15 -0.03 0.02 Trade 0 0 0.05 0.89 0.01 0.52 Transport 0 0 -0.01 0.3 0.03 0.51 Communica- tions 0 0 -0.03 0.4 -0.06 0.03 Financial institutions 0 0 -0.07 0.38 -0.10 -0.01 Farm services 0 0 -0.10 0.21 -0.11 -0.01 Enterprise services 0 0 -0.06 0.29 -0.04 0.16 Building rentals 0 0 2.53 7.76 2.69 8.29 The Doha Round, Poverty, and Regional Inequality in Brazil 207 Table 7.6. (Continued) Implied Import tariffs Import CIF prices export price shifta Full Full Full liberali- liberali- liberali- Commodity Doha zation Doha zation Doha zation Public admin- istration 0 0 -0.05 0.07 -0.08 -0.32 Other services 0 0 -0.10 0.21 -0.11 -0.01 Source: Based on results from chapter 3. a. Vertical shift in export demand schedule calculated from GTAP results. poverty analysis. When interpreting these results, one should bear in mind that the model has a top-down inter-regional specification, meaning that regional results depend on national results, but not vice versa. National macro results can be seen in table 7.7. Because the closure fixes total supply of all primary factors (land, the 10 categories of labor, and capital), the GDP shows only a slight increase in the simulations. The real exchange rate rises (revaluation) as a result of the shocks, with corresponding gains in the external terms of trade. For factor market results, recall that land is used only by agriculture, and capital and the 10 types of labor are fixed nationally but mobile between sectors. The average (aggregated) capital rental increases in all scenarios. With capital stocks and labor fixed in total, the expanding industries would attract capital and labor from the contracting ones. In these industries, those with falling capital to labor ratios increase the marginal productivity of capital, and hence capital returns, determining an increase in aggregated results. The price of land also shows a strong increase, reflecting the increase in production of activities using this factor (agriculture). National changes in industry output are shown in table 7.8. As can be seen in table 7.8, agriculture and agriculture-related industries (the food industry in general) are expanding industries in all scenarios. The only exception is the vegetable oils industry, which contracts under full liberalization. Model results show a general fall in activity in the Brazilian manufacturing sectors after trade liberalization. This suggests that regions specializing in manufacturing would fare worse. The Doha results are similar, just differing in size (but not sign) when compared with the full-liberalization scenario, with a few exceptions. 208 Poverty and the WTO: Impacts of the Doha Development Agenda Table 7.7. Selected Macroeconomic Results Scenarios (percentage changes) Macros Doha Full liberalization Real household consumption 0.22 0.61 Real investment 0.00 0.00 Real government expenditure 0.00 0.00 Exports volume 0.91 13.21 Imports volume 1.98 12.39 Real GDP 0.04 0.26 Aggregated employment 0.00 0.00 Real wage 0.02 -0.22 Aggregated capital stock 0.00 0.00 Average rate of return 0.24 1.36 CPI (numeraire) 0 0 GDP price index 0.05 -0.33 Export price index 0.11 -0.38 Imports (CIF) price index -1.10 -1.65 Imports (domestic prices) price index -1.23 -7.63 Real devaluation -1.15 -1.32 TOT 1.22 1.28 Nominal exchange rate -1.26 -1.99 Balance of trade as a GDP share 0.00 0.00 Price of agricultural land 7.7 20.97 Source: Authors' simulation results. Table 7.9 shows regional results. In this table, states are grouped according to their macroregions inside Brazil. For each of the 10 labor types, total employment is fixed, thus labor demand (and unemployment) will be redistributed among regions according to changes in regional industry output. Employment falls in São Paulo and Rio de Janeiro in the southeast (the most populous and industrialized states) and also in Amazonas, Distrito Federal (Brasília), and Amapá. The states of São Paulo and Rio de Janeiro are industrial states, hosting the bulk of Brazil's manufacturing. As seen before, manufacturing is contracting in general in all three scenarios. The same effect drives the result for Amazonas, where there is a free exporting zone, and reduced mining activity drives the results for Amapá. The Doha Round, Poverty, and Regional Inequality in Brazil 209 Table 7.8. Activity Level Variation by Industry: Percentage Change Activity level Doha Full liberalization Agriculture 1.35 3.60 Mineral extraction -1.00 -1.21 Petroleum and gas extraction -1.45 -0.99 Non-metallic minerals -0.36 -1.13 Iron production -2.13 -3.75 Non-ferrous metals -1.55 -0.50 Other metals -1.19 -4.11 Machinery and tractors -2.25 -4.95 Electric materials -1.27 -5.22 Electronic equipment -0.60 -3.36 Automobiles -1.06 -6.35 Other vehicles and spare parts -3.32 -6.55 Wood and furniture -0.33 0.01 Paper -0.58 -1.14 Rubber manufactures -1.60 -4.76 Chemicals -0.86 -3.81 Petrol refining -0.39 -0.48 Miscellaneous chemicals -0.23 -1.23 Pharmaceutical products -0.05 -0.01 Plastics -0.49 -2.16 Textiles 0.27 -3.06 Apparel 0.20 -1.52 Footwear -4.94 -10.86 Coffee processing 0.39 0.72 Vegetable processing 0.79 4.52 Slaughtering 7.78 18.81 Dairy 0.71 0.86 Sugar refining 4.52 19.08 Vegetable oils 1.95 -5.61 Other food 0.34 1.36 Miscellaneous industries -1.08 -7.75 Public utilities -0.07 -0.15 Construction 0.00 0.00 Trade 0.09 0.41 Transport -0.10 0.52 Communications -0.01 0.01 Financial institutions -0.07 -0.28 Farm services -0.05 -0.14 210 Poverty and the WTO: Impacts of the Doha Development Agenda Table 7.8. Activity Level Variation by Industry: Percentage Change (Continued) Activity level Doha Full liberalization Enterprise services -0.30 -0.35 Building rentals 0.14 0.23 Public administration -0.03 -0.03 Other private Services 0.14 -0.01 Source: Authors' simulation results. The trade liberalization scenarios seem to redistribute economic activity toward poorer regions. This occurs because higher-value-added sectors (manufacturing) shrink, and relatively lower-value-added sectors (agriculture) grow. Poverty and Income Distribution Results In the previous section, it was seen that model results are differentiated among regions and industries. The outcome of these changes on income and income inequality measures, as well as over income group­specific consumer price indexes (CPIs) are presented in table 7.10. In this table, the POF groups are groups of household income, with POF[1] the lowest and POF[10] the highest. The Gini index fell by 0.21 percent in the Doha scenario and 0.52 percent in the full liberalization. These results confirm the general understanding that the Gini index usually changes very little with policy measures in the short run and accord with observed facts in Brazil in the last 15 years. Even though the country faced a strong trade liberalization process in the 1990s, it was observed that the Gini index changed very little in the period. The CPI column in each scenario is the particular CPI change for each household income class, because the consumption bundle of each class is different. It is interesting to notice that the bulk of the real income effect comes from the income generation side and not from the fall in prices. Actually, there is a strong increase in some food products, such as meats, in all scenarios, driven mainly by the liberalization in the rest of the world. This is in contrast to what was expected by Rocha (1998), who predicted that opening the Brazilian economy to the external market would help reduce inequality in Brazil through reductions in prices in the poorest regions. The results in this chapter suggest that the CPI The Doha Round, Poverty, and Regional Inequality in Brazil 211 Table 7.9. Regional Results in 27 Regions (percentage change) Aggregate employment Gross regional product State Doha Full liberalization Doha Full liberalization Rondônia 0.75 1.92 0.99 2.53 Acre 0.40 1.02 0.54 1.42 Amazonas -0.14 -0.35 -0.19 -0.58 Roraima 0.48 1.33 0.74 2.09 Pará 0.32 0.89 0.44 1.25 Amapá -0.03 0.00 -0.02 0.06 Tocantins 2.04 5.34 2.33 6.14 Maranhão 0.86 2.35 1.14 3.12 Piauí 0.86 2.15 1.24 3.16 Ceará 0.33 0.77 0.50 1.18 Rio Grande do Norte 0.17 0.45 0.24 0.64 Paraíba 0.23 0.56 0.35 0.85 Pernambuco 0.14 0.34 0.19 0.47 Alagoas 0.43 1.56 0.51 1.80 Sergipe 0.15 0.34 0.21 0.50 Bahia 0.24 0.62 0.22 0.64 Minas Gerais 0.07 0.24 0.07 0.24 Espírito Santo 0.07 0.25 0.10 0.37 Rio de Janeiro -0.15 -0.26 -0.11 -0.13 São Paulo -0.21 -0.60 -0.25 -0.75 Paraná 0.27 0.70 0.34 0.86 Santa Catarina 0.21 0.50 0.21 0.54 Rio Grande do Sul 0.01 0.09 0.02 0.21 Mato Grosso do Sul 1.49 3.82 1.74 4.41 Mato Grosso 1.06 2.76 1.24 3.11 Goiás 0.71 1.80 0.85 2.14 Distrito Federal -0.04 -0.09 0.01 0.05 Source: Authors' simulation results. would actually go up more in the lowest income classes but these increases are more than compensated by the income elevation. It is important to notice that the highest positive changes in household income are concentrated on the lowest-income households, decreasing monotonically as household income increases. As can be seen in table 7.11, the reduction in the number of poor households is concentrated in the poorest groups. High positive 212 Poverty and the WTO: Impacts of the Doha Development Agenda Table 7.10. Average Household Income, CPI by Household Income Class and Gini Index Percentage Change Doha Full liberalization Income CPI Income CPI POF[1] 6.45 0.16 16.54 0.44 POF[2] 1.23 0.14 2.82 0.41 POF[3] 0.69 0.11 1.39 0.33 POF[4] 0.29 0.08 0.46 0.22 POF[5] 0.19 0.08 0.01 0.23 POF[6] -0.02 0.06 -0.35 0.18 POF[7] -0.10 0.04 -0.60 0.12 POF[8] -0.26 0.00 -0.93 0.04 POF[9] -0.27 -0.01 -0.91 -0.03 POF[10] -0.33 -0.08 -1.09 -0.27 Gini -0.21 -0.52 Source: Authors' simulation results. figures in POF groups 7 and 8 are percentage changes over very low numbers, because there are very few poor households in these income classes.19 The headcount ratio index (FGT0 in table 7.11) captures the extension of poverty but is insensitive to its intensity (Hoffmann 1998). The change in the intensity of poverty can be seen through the FGT1 index, poverty gap, or insufficiency of income ratio. A reduction in FTG1 means a reduction in the severity of poverty in each household income class. As seen in table 7.11, the FGT1 index decreases more than the headcount ratio in all scenarios. This means that there was actually an income distribution improvement, but not enough to drive a large number of persons (or households) out of poverty. This is due to the high value of those indexes in the base year. In addition, the effects on Brazil of its own liberalization (assuming other countries did not liberalize) were computed (but are not tabulated here due to space constraints). The Brazilian own-tariff reduction contributes very little to the Doha scenario and is dominated by the other countries' actions even in the full- liberalization scenario. Finally, table 7.12 shows model results relating to the regional breakdown inside Brazil. These results summarize at regional level the outcome of the simulated scenarios, as a net effect of the regional industries. They reflect, then, the pattern of regional specialization in production. The Doha Round, Poverty, and Regional Inequality in Brazil 213 Table 7.11. Percentage Changes in the Proportion of Poor Households (FGT0) and in the Poverty Gap Ratio (FGT1) by Household Income Groups Original value Doha Full liberalization Household income class FGT0 FGT1 FGT0 FGT1 FGT0 FGT1 POF[1] 0.9617 0.7334 -0.52 -1.45 -1.55 -3.74 POF[2] 0.7657 0.3047 -0.48 -1.31 -1.26 -2.91 POF[3] 0.5355 0.1496 0.00 -0.89 -0.91 -1.39 POF[4] 0.2837 0.0539 -1.72 0.39 -2.28 2.67 POF[5] 0.1143 0.0189 -1.03 2.71 1.55 9.85 POF[6] 0.0390 0.0054 1.78 11.22 9.44 33.32 POF[7] 0.0082 0.0009 10.96 57.55 32.24 156.86 POF[8] 0.0008 0.0001 92.35 417.68 247.52 1107.21 POF[9] 0.0000 0.0000 0.00 0.00 0.00 0.00 POF[10] 0.0000 0.0000 0.00 0.00 0.00 0.00 Brazil 0.308 0.145 -0.37 -1.08 -0.78 -2.43 Source: Authors' simulation results. Note: FGT0: Foster-Greer-Torbecke proportion of poor households index (headcount ratio). FGT1: poverty gap ratio. Table 7.12 shows that the states of Amazonas, Amapá, São Paulo, and Rio de Janeiro would be the only ones where the number of households below the poverty line would increase in both simulations, although only slightly. Amazonas and Amapá have small populations, but São Paulo and Rio de Janeiro are the most densely populated and industrialized states in Brazil. As noted before, the result is related to the high concentration of contracting (high-value-added) industries in the regions of São Paulo, Rio de Janeiro, and Amazonas, mainly automobiles, machinery and tractors, electric materials, electronic equipment, and other vehicles and spare parts; the result for Amapá is driven by the mining industries. Concluding Remarks One of the main objections raised by some opponents to multilateral trade reform in agriculture is that the bulk of the benefits will go to rich landowners in farm export­oriented economies such as Brazil, thereby worsening an already skewed income distribution. The findings in this chapter refute this hypothesis. In fact, multilateral trade reform under the Doha Development Agenda is found to 214 Poverty and the WTO: Impacts of the Doha Development Agenda Table 7.12. Percentage Changes in Number of Poor Households by Region and Total Number Change Scenario State Doha Full liberalization Rondônia -0.73 -1.58 Acre -0.36 -0.47 Amazonas 0.42 0.80 Roraima -0.60 -1.74 Pará -0.24 -0.82 Amapá 2.41 2.06 Tocantins -1.34 -3.94 Maranhão -0.87 -2.04 Piauí -0.34 -1.32 Ceará -0.32 -0.88 Rio Grande do Norte -0.53 -0.83 Paraíba -0.82 -1.56 Pernambuco -0.35 -1.09 Alagoas -0.38 -1.35 Sergipe -0.41 -0.61 Bahia -0.45 -1.04 Minas Gerais -0.48 -1.08 Espírito Santo -0.71 -1.29 Rio de Janeiro 0.77 0.99 São Paulo 0.72 1.97 Paraná -1.19 -2.47 Santa Catarina -1.79 -2.08 Rio Grande do Sul -0.54 -2.12 Mato Grosso do Sul -2.77 -6.44 Mato Grosso -2.32 -6.06 Goiás -1.06 -2.80 Distrito Federal 0.11 0.18 Change in total number of households -55,908 -139,874 Change in total number of persons -235,886 -481,989 Source: Authors' simulation results. The Doha Round, Poverty, and Regional Inequality in Brazil 215 reduce inequality as well as poverty in Brazil. Although wealthy farmers may indeed gain, largely through higher returns to their land, the poor gain proportionately more. Their gains are derived through the labor market. Because 40 percent of the lowest-skill group work in agriculture, an expansion of that sector benefits the poorest households, which rely heavily on earnings from low- skill work. More generally, the model results show that even important trade policy shocks do not generate dramatic changes in the structure of Brazilian poverty and income distribution. The simulated effects on poverty and income distribution are positive, but very small. This is partly due to the fact that the Brazilian economy is not very oriented toward external trade. The domestic market is far bigger and more important for the general economy than the external market, as researchers have long understood. This makes Brazil naturally less sensitive to tariff structure changes, as well as to changes in export demands. There are modest impacts of trade reform on poverty because poverty is approached here through the household dimension, tracking the changes in the labor market from individual workers to households. This tends to blunt the impacts that trade may have on the employment of any individual and on poverty at the household level. In the PNAD 2001 data used in this chapter, the head of the family's income accounts for about 65 percent of aggregated household income in Brazil. Therefore, using head-of-household income as a proxy for household income may poorly predict the effect of policy changes, as convincingly argued by Bourguignon, Robilliard, and Robinson (2003). If spending (and welfare) is in any sense a household phenomenon, this is the appropriate method. Even though there may be a somewhat higher computational cost associated with this procedure, it seems worthwhile. The role played by the agriculture sector in the analysis should also be stressed. As seen before, agriculture still accounts for a large share of employment for the poorest in Brazil. Despite the steady decline over time of agricultural employment as a share of total employment, the importance of agricultural policies for poverty alleviation in Brazil should not be overlooked. Finally, it should be noted that this study assesses only the static impact of trade liberalization scenarios. The research methodology used here fails to capture many other effects generally associated with external trade liberalization, such as endogenous technology improvements and other dynamic effects. Indeed, the results of this study suggest that if any strong impact on poverty is supposed to arise from trade liberalization, it must be expected to arise from these other aspects. 216 Poverty and the WTO: Impacts of the Doha Development Agenda Notes 1. For the simulations reported here, the export demand elasticities were set to values derived from the GTAP model to increase consistency between results for the world and Brazil models. 2. To fit with the assumptions of the GTAP model, the constant elasticity of transformation between domestic and exportable goods was set to infinity for the simulations reported in this chapter. 3. One of the activities (agriculture) produces 11 commodities; a CET function determines output mix. 4. As in the GTAP, labor supplies were fixed. Further, each household in the micro dataset used here had 1 of the 270 expenditure patterns identified in the main CGE model. There is very little scope for the MS to disagree with the CGE model. 5. The methodology is described in more detail in the appendix to the World Bank Policy Research Working Paper version of this chapter (Ferreira-Filho and Horridge, 2005). Only the main ideas are presented here. 6. Mark Horridge developed this method for this project. 7. POF[1] ranges from 0 to 2 minimum wages, POF[2] from 2+ to 3, POF[3] from 3+ to 5, POF[4] from 5+ to 6, POF[5] from 6+ to 8, POF[6] from 8+ to 10, POF[7] from 10+ to 15, POF[8] from 15+ to 20, POF[9] from 20+ to 30, and POF[10] is above 30 minimum wages. The minimum wage in Brazil in 2001 was around US$76 per month. 8. The equivalent household concept measures the subsistence needs of a household by attributing weights to its members: 1 to the head, 0.75 to the other adults, and 0.5 to the children (for example, to feed 2 persons does not cost double). Because poverty is defined here on an equivalent basis, a few (very large) families in middle-income groups fall below the poverty line. 9. This poverty line was equivalent to US$48 per month in 2001. 10. Barros, Henriques and Mendonça (2001), working with a poverty line that takes into account nutritional needs, find that 34 percent of Brazilian households were poor in 1999. 11. The proportion of households below the poverty line in the other income groups are 0.284 percent for the fourth, 0.14 percent for the fifth, 0.04 percent for the sixth, 0.008 percent for the seventh, and 0.001 percent for the eighth. There are no households below the poverty line for the two highest income classes. 12. The poverty gap and poverty line values are constructed with "adult equivalent" per capita household income. 13. See the World Bank Policy Research Working Paper version of this chapter for more details on the breakdown of poverty by region (Ferreira-Filho and Horridge, 2005). 14. There is a tension between the GTAP-like closure in this chapter and Brazilian reality. The microdata show substantial unemployment of less skilled groups in all regions. For the MS, it was assumed that jobs created (or lost) in a region were allotted to (or taken from) households in that region. An alternate scenario, in which fixed real wages replaced national labor constraints, yielded results similar to those reported here. 15. The factor market closure causes the model to generate changes in prices for 10 labor types, capital, and land--that is, price changes are uniform across regions. Changes in demand for each of the 12 factors also vary by sector and region. Each adult in the PNAD microdata is identified by region and labor type; those employed are also identified by sector. Changes in microdata poverty levels are driven by wage changes and by the redistribution of jobs between sectors and regions (and hence between households). 16. That is, neither the distribution of spending nor relative prices facing households is altered. With fixed labor supplies, distortion of any labor-leisure choice does not arise. 17. The share of imports plus exports in Brazilian GDP in 2001 and 2002 were, respectively, 22.3 percent and 23.4 percent. The Doha Round, Poverty, and Regional Inequality in Brazil 217 18. The shifts in the demand schedules for Brazilian exports were calculated using export price and quantity results from the GTAP with export demand elasticities drawn from GTAP data. 19. Some middle-income households have many family members. With low per capita income, they fall below the poverty line. References Barros, R. P., and R. Mendonça. 1997. "O Impacto do crescimento econômico e de reduções no grau de desigualdade sobre a pobreza." Texto para discussão 528, Instituto de Pesquisa Econômica Aplicada (IPEA), Rio de Janeiro. Barros, R. P., C. H. Corseuil, and S. Cury. 2001. "Salário mínimo e pobreza no Brasil: Estimativas que consideram efeitos de equilíbrio geral." Texto para Discussão 779, IPEA, Rio de Janeiro. Barros, R. P., R. Henriques, and R. Mendonça. 2001."A Estabilidade inaceitável: desigualdade e pobreza no Brasil." Texto para Discussão 800, IPEA, Rio de Janeiro. Bourguignon, F., A. S. Robilliard, and S. Robinson. 2003."Representative versus Real Households in the Macro-Economic Modeling of Inequality." Working Paper 2003-05. Département et Laboratoire d'Économie Théorique et Appliquée, Centre National de la Recherche Scientifique, École des Hautes Études en Sciences Sociales, Paris. Dixon, P., B. Parmenter, J. Sutton, D. Vincent. 1982. ORANI: A Multisectoral Model of the Australian Economy. Amsterdam: North-Holland. Ferreira-Filho, J.B. and M. Horridge. 2005. "The Doha Round, Poverty and Regional Inequality in Brazil." Policy Research Working Paper 3701, World Bank, Washington, DC. Foster, James, Joel Greer, and Erik Thorbecke. 1984. "A Class of Decomposable Poverty Measures." Econometrica 52: 761­65. Green, F., A. Dickerson, and J. S. Arbache. 2001. "A Picture of Wage Inequality and the Allocation of Labor through a Period of Trade Liberalization: The Case of Brazil." World Development 29 (11): 1923­39. Gurgel, A., G. W. Harrison, T. F. Rutherford, and D. G. Tarr. 2003. "Regional, Multilateral, and Unilat- eral Trade Policies of MERCOSUR for Growth and Poverty Reduction in Brazil." Research Working Paper 3051, World Bank, Washington, DC. Hoffmann, R. 1998. Distribuição de renda: Medidas de desigualdade e pobreza. São Paulo: Editora da Universidade de São Paulo. Horridge, J. M. 2000. "ORANI-G: A General Equilibrium Model of the Australian Economy." Working Paper OP-93, Centre of Policy Studies, Monash University, Melbourne. IBGE (Instituto Brasileiro de Geografia e Estatística). 1996a. Censo agropecuário do Brasil. Rio de Janeiro. ------.1996b. Pesquisa de orçamentos familiares. Rio de Janeiro.. ------. 2001. Pesquisa Nacional por Amostra de Domicílios. Rio de Janeiro. Rocha, S. 1998. "Desigualdade regional e pobreza no Brasil: a evolução--1985/95." Texto para Dis- cussão 567, IPEA, Rio de Janeiro. Savard, L. 2003. "Poverty and Income Distribution in a CGE-Household Sequential Model." Interna- tional Development Research Centre, University of Laval, Quebec. 8 Growing Together or Growing Apart? A Village-Level Study of the Impact of the Doha Round on Rural China Marijke Kuiper and Frank van Tongeren Summary Most studies of the opening of the Chinese economy focus on the national level. The few existing disaggregated analyses are limited to analyzing changes in agri- cultural production. This chapter uses an innovative village equilibrium model that accounts for nonseparability of household production and consumption decisions. This allows analysis of the impact of trade liberalization on household production, consumption, and off-farm employment; the interactions among these three aspects of household decisions; and the interactions among house- holds in a village economy. The village model is used to analyze the impact of price changes and labor demand, the two major pathways through which interna- tional trade affects households. Analysis of the impact of trade liberalization for one village in the Jiangxi province of China reveals that changes in relative prices and outside-village employment have opposite impacts on household decisions. At the household level, the impact of price changes dominates the employment impacts. When full trade liberalization and the more limited Doha scenario are compared, reactions are more modest in the latter case for most households, but the response is nonlinear to increasing depth of trade reforms. This is explained by household-specific transaction (shadow) prices in combination with endoge- nous choices to participate in the output markets. 219 220 Poverty and the WTO: Impacts of the Doha Development Agenda Rising income inequalities are a growing concern in China. Whether trade lib- eralization allows incomes to grow together or to grow apart depends on whether one accounts for the reduction in aggregate consumption demand when house- hold members migrate. Assessment of the net effect on the within-village income distribution shows that even poorer households are able to catch up. Poorer households that own draft power gain most from trade liberalization. The house- holds that have to rely on the use of own-labor for farm activities, and are endowed with neither traction power nor a link to employment opportunities in the prospering coastal regions, have fewer opportunities for adjustment. Introduction A gradual integration into the global economy, combined with far-reaching domestic reforms, has made China a showcase of attaining rapid economic growth through market-based reforms. The rapid economic growth during the past decades, however, has been accompanied by an increasing disparity between coastal and interior regions and between rural and urban areas. The coastal cities benefited most from the increasing export opportunities because of a combina- tion of geographic factors and deliberate policies (Démurger and others 2002). The extent of reforms in combination with the sheer size of the Chinese econ- omy has resulted in a body of literature with a growth rate rivaling China's GDP growth. Most studies of the opening of the Chinese economy to the rest of the world focus on the national-level impacts. One exception is a study by Diao, Fan, and Zhang (2003) of the regional impact of China's recent WTO accession. As in other national-level studies, they find a positive aggregate effect for China as a whole. This aggregate effect obscures differences across regions. Reflecting past trends of diverging growth between coastal and inland provinces (see, for exam- ple, Démurger and others [2002] and Jones, Li, and Owen [2003]), income gaps among provinces are found to widen following WTO accession. They also find that rural-urban migration provides an important mechanism for transmitting urban growth to the rural areas. Central provinces bordering the booming coastal provinces (Anhui, Jiangxi, Hubei, and Hunan) especially benefit from an increase in rural-urban migration. A second exception to the common use of a national-level analysis is a study by Huang, Li, and Rozelle (2003) analyzing the impact of WTO accession on farm households. They find that despite a small positive aggregate impact of the WTO, the distribution of benefits gives cause for concern. Households in richer coastal areas benefit most, having higher-yielding lands and cultivating internationally competitive crops. Another interesting finding of their household-level analysis is an increase in aggregate agricultural production, despite a decrease in the agricul- A Village-Level Study of the Impact of the Doha Round on Rural China 221 tural price index. Farmers respond to the changes in relative prices induced by the WTO by shifting to more competitive activities (livestock, fish, vegetables, and rice), resulting in a net increase in agricultural production. These two studies show that a disaggregated analysis of the impact of trade reform yields insights that diverge from insights gained at the national level. The objective of this chap- ter is to contribute a village-level perspective on trade reform in China. Adjust- ment responses of different household types and within-village interactions are central to this analysis. This study complements earlier studies by concentrating on the differential impact of trade liberalization on households within a village. A new methodology that accounts for family-farm production specifics in a village- economy setting provides a unique perspective on the impact of trade reform. Going beyond the household-level study of Huang, Li, and Rozelle (2003), the analysis in this chapter also takes account of the impact of farm income on rural household consumption, as well as the impact of rural-urban migration, thereby extending the household analysis beyond developments in the agricultural sector. This study combines a macro-level analysis of trade reform with a village-level general equilibrium model of a rice-producing village in Jiangxi province. To be able to combine the macro- and village-level analyses of trade reform, it is neces- sary to ascertain how trade reform affects households. Prices and labor demand form the key transmission mechanisms through which macro-level trade reform affects rural households (Winters 2002). Thus, the focus is on changes in prices for consumed goods, agricultural inputs and outputs, and the increased demand for labor by labor-intensive sectors in which China has a comparative advantage. This analysis of further trade reform builds on a baseline that encompasses China's recent WTO accession, phasing out of the export quota for textiles under the Agreement on Textiles and Clothing. This chapter draws on the global results from chapter 3 for an assessment of the macro-level impacts of further reform under the Doha Round and feed these into the village-level model. Specifically, the impacts of the standard Doha scenario described in chapter 2, as well as of the full-liberalization scenario, are simulated. In both cases, the effects of changing prices for inputs and outputs from the effects of increases in off-farm employ- ment and wages are disentangled. The full-liberalization scenario provides a useful benchmark against which the less ambitious Doha scenario can be compared. Analysis of consumption reveals that the poorer households face a stronger rise in expenditures. This is due to a larger share of agricultural goods in their consumption and a shift from being net sellers to self-sufficiency in the case of some households. Analysis of the impact of changes in agricultural input and output prices (with unchanging employment and wages) shows an increased village supply of rice and other livestock, which corresponds to the findings of national-level studies and 222 Poverty and the WTO: Impacts of the Doha Development Agenda the results of the farm household analysis in Huang, Li, and Rozelle (2003). This increased supply of rice, however, is the net result of three household groups increasing rice production and one group reducing rice production. The net impact of more off-farm employment (at constant prices) after global trade reform is a decrease in rice supply caused by an increasing scarcity of labor. Again, there are divergent household responses, with some households increasing rice production because of lower costs of animal traction rented within the vil- lage. Because this is opposite in sign to the impact of higher prices, it is interesting to ask which dominates. The analysis presented here shows that the impact of price changes thus dominates the impact of increased employment. A more modest liberalization within the context of the Doha Round has a sim- ilar, but less pronounced, impact to that observed under full trade liberalization. The notable exception is the village rice economy. Under the full-liberalization scenario used in this chapter, rice production becomes more intensive in land and labor, and village-marketed surplus increases. A Household Perspective on General Equilibrium Modeling A disaggregated perspective can lead to new insights in the impact of trade reform. Separating farm production decisions from household consumption decisions is standard practice in general equilibrium models, both macro as well as existing village general equilibrium models. Ignoring the interdependency of household production and consumption decisions, however, can be misleading when market imperfections render these two aspects of household decisions non- separable. Nonseparability of household production and consumption decisions occurs when the effective price of a commodity used in both production and consump- tion is not exogenous to the household but determined endogenously by house- hold demand and supply. In this case, production decisions will affect supply of the commodity, which affects its shadow price and hence consumption decisions, and vice versa. Such nonseparability occurs if households are not price takers in a market, if markets are missing, or if there is a gap between buying and selling prices (Löfgren and Robinson 1999). The seminal work of de Janvry, Fafchamps, and Sadoulet (1991) shows how rational behavior of farmers in combination with market failures may give rise to sluggish or counterintuitive household responses. A number of points are essential to farm household modeling. First, standard economic rules for production and consumption remain valid. Differences with a separate analysis of production and consumption decisions occur because of A Village-Level Study of the Impact of the Doha Round on Rural China 223 endogenous prices, not because of different behavior by the household. Conse- quently, standard approaches to modeling production and consumption deci- sions can be followed. But endogenous household prices complicate empirical work. The endogenous household shadow prices are an analytical construct and thus cannot be directly observed. This complicates the estimation of demand and supply functions for nonseparable household models. Second, household models tend to generate ambiguous results and quickly become analytically intractable. Ambiguous results may already occur with per- fect markets. Assume that prices for food increase. This will increase food produc- tion and thus household income. The higher income prompts an increase in con- sumption, which may outweigh the increase in food production, depending on the preferences of the household. The food price increase then does not lead to an unequivocal increase in marketed surplus. In fact, if the income effect is strong enough, sales by the household will actually fall. Thus, in those cases where an analytical solution of the household model can be obtained, it will generally be difficult to sign the effects because of counteracting effects on the production and consumption sides of the household. Models with multiple missing markets com- plicate things even more. A third point about household modeling is the importance of accounting for different levels of market integration of households from a policy perspective, as can be illustrated with a price band model (figure 8.1). Starting from an exoge- nous market price, transaction costs increase the effective purchase price and decrease the effective sales price faced by the household. Household demand and supply then determine the household-specific shadow price of the commodity, with effective purchase and sales prices forming upper and lower boundaries. Fig- ure 8.1 shows the supply curve for three different types of households. Depending on the intersection of the demand and supply curve, a household is (a) a net buyer, (b) self-sufficient, or (c) a net seller of the commodity. If the household is a net buyer or seller, the household shadow price equals the effective purchase or sales price. If the household is self-sufficient (case b), the household shadow price is endogenously determined within the price band and decisions become nonsep- arable. A missing market can be conceptualized in this model as a wide price band (in the most extreme case, a sales price of zero and an infinite purchase price) such that all households always operate within it. Household response then consists of two decisions a discrete decision on mar- ket position, determining their position as net buyer, net seller, or not participat- ing, and a continuous decision on production and consumption levels, determin- ing supply response. The position of the household in the market determines the effective decision-making prices for the second decision. Net buyers will respond 224 Poverty and the WTO: Impacts of the Doha Development Agenda Figure 8.1. Household Supply Response with Price Bands Supplya Purchase price Purchased Supplyb Transaction costs of purchasing Exogenous market price Transaction costs of selling Sold Sale price Supplyc Demand Quantity Source: Adapted from Sadoulet and de Janvry (1995). a. Supply of a net buyer. b. Supply of a self-sufficient household. c. Supply of a net seller. differently to a price increase from net sellers, and households operating within their price band will not show any response to the price change. The position of the household in the market thus has an important impact on the household response to price incentives. Nonseparability has become an important feature of household models, but it is absent from AGE models. A study by Löfgren and Robinson (1999) provides a stylized application of including farm-household models in a general equilibrium model, but this approach has not previously been implemented in an empirical analysis. Given the recent transformations in China, markets are still developing and imperfections can be expected to abound. Studies of factors influencing migra- tion decisions (Hare 1999; Murphy 2000; Rozelle and others 1999; Rozelle, Taylor, and de Brauw 1999) and of patterns in inequality (Benjamin and Brandt 1999) refer to imperfect land, labor, and credit markets as being relevant in the Chinese context. Such a partial integration in markets may give rise to nonseparability of household decisions or may create (thin) local markets through which household decisions affect each other. A Village-Level Study of the Impact of the Doha Round on Rural China 225 If households interact with each other in local village markets, and these mar- kets are not integrated with markets outside the village, local general equilibrium effects occur. Studies of market integration in China find villages to be integrated in markets for major outputs (Huang, Li, and Rozelle 2003) and for fertilizer (Qiao and others 2003). Although villages may thus be assumed to be integrated into agricultural input and output markets, integration of factor markets is still limited. Labor markets are highly segmented (Gilbert and Wahl 2003), resulting in a rural labor surplus (Cook 1999), which is only partly absorbed in township-vil- lage enterprises. The local village labor markets are limited in rural areas, includ- ing the labor market in this chapter's case study village in Jiangxi province. A prime reason for limited development of a rural labor market is the collective ownership of land, which grants all households access to land. Consequently, there are no landless households that would specialize in wage-earning activities and hence little scope for local labor markets. In spite of land tenure reforms that have granted household user rights for 30 years (and which were recently changed to permit inheritance), land rights remain ambiguous (Huang and Rozelle 2004). Land is allocated on the basis of demo- graphic criteria, and readjustments occur to adjust for changes in household size, despite formal household user rights. The result is an ambiguous land tenure situ- ation (Ho 2001) in which households have an incentive to keep their land culti- vated to avoid losing it during the next readjustment. Households that migrate to urban areas rent their land to other local households, seeking to maintain their claim to the land in case they are unable to secure a living in the urban areas. Given the ambiguity of land tenure, land rental markets are inherently local in nature. Village interactions may also arise through informal credit markets. Govern- ment intervention in the formal banking sector remains strong. Regulated interest rates are well below market clearing levels, and soft loans to state enterprises seize a large share of available funds. Rural households are thus rationed out of credit markets. In the late 1980s, rural cooperative funds developed, targeted at rural households. These funds proved to be too successful competitors with existing rural credit cooperatives, and they were dissolved in 1999 (Park, Brandt, and Giles 2003). As a result of the lack of formal credit options, households have to rely on local, informal credit markets. This study therefore uses a village-level general equilibrium model to account for the interactions among households in the local markets for land and capital, while paying due attention to nonseparability of household decisions flowing from the presence of significant transaction costs. Taylor and Adelman (1996; 2003) pioneered the use of general equilibrium models at the village level. Their model closely follows the structure of macro- level models--for example, by modeling production at the sector level, which misses the impact of nonseparability of household decisions. This chapter takes a 226 Poverty and the WTO: Impacts of the Doha Development Agenda different approach to village modeling, placing differences in household response due to nonseparability at the center of the model. This household perspective on general equilibrium modeling results in a model structure different from that used in macro-level general equilibrium models and existing village-level models. In this chapter, production activities are modeled as being household specific. This allows for idiosyncratic household responses consistent with nonseparability of production and consumption decisions. A second major difference is the nest- ing structure used for modeling production decisions. For each activity, the pro- duction structure is calibrated based on the household survey data.1 As a result, there are household-specific production functions capturing differences in house- hold access to inputs.2 By placing households at the center of the model, the village model used in this study is able to capture differences in production decisions reflecting differences in access to inputs, interactions between household production and consumption decisions, and interactions among different households within a village economy. Models and Data Linking Macro Results to the Village Model This chapter analyzes the village-level impact of the two liberalization scenarios discussed in part 1 of this book: full trade liberalization and a Doha Round sce- nario. Because the Chinese economy is not modeled at the national level, the Global Trade Analysis Project (GTAP) results used in this chapter are those that include China's tariff cuts as well as liberalization in the rest of the world (recall chapter 3). To link the macro shocks for China to the village model, this chapter follows the conceptual framework of Winters (2002) and focus on prices and labor demand. Studies of market integration of three main staple crops and fertilizer (Huang, Li, and Rozelle, 2003; Qiao and others 2003) show integrated regional and national markets, with village-level prices responding to changes at the national level. These integrated commodity markets allow direct translation of relative price changes derived from the macro-level analysis to the village level. In addition to the transmission of price changes, this chapter analyzes the impact of changes in employment opportunities. Studies of trade liberalization find an expansion of labor-intensive sectors in which China has a comparative advantage. Thus there is a need to link aggregate expansion of employment to village-level changes in temporary migration to urban areas. Lacking data to quantify the link between national-level changes in employment and household decisions, percent- A Village-Level Study of the Impact of the Doha Round on Rural China 227 age changes in aggregate labor demand are assumed to be completely transmitted to the village. Percentage changes in outside village employment are thus set equal to changes in aggregate labor demand. Such a one-on-one relation between national- level employment and off-farm activities of the households in the case study village seems justified, given the findings in Diao, Fan, and Zhang (2003). Comparing the change in rural-urban migration across regions after WTO accession, they find the fastest increase in rural-urban migration in the central provinces. The case study vil- lage is located in Jiangxi, one of these central provinces, and migration outside the province plays an important role in the village economy, justifying the assumption of a complete transmission of the demand for labor to the village level. (The case study village is more thoroughly described below.) Similarly, changes in outside-village wages and changes in nonagricultural wages are taken from the GTAP simulations. As will be seen below, the rural- urban wage differential is endogenously determined through modeling of house- hold-specific shadow wages. Mapping macro-level changes computed with the GTAP model to the village level results in the shocks summarized in table 8.1. The Case Study Village The case study village has been selected to be representative of rice-producing vil- lages in the plains area of Jiangxi province, one of the poorer provinces in China. Data on production and consumption of 168 households were collected for 2000, using standard household questionnaires with questions on source and destina- tion of commodities added to allow construction of a village social accounting matrix (SAM). These surveyed households account for about a quarter of the village population, totaling 729 households. Differences among households are at the center of the village equilibrium model. Four groups of households are distinguished, using ownership of draft power (cattle or tractor) and access to extraprovince employment as grouping cri- teria. The resulting groups represent households with differential capacity for earning a living from agriculture and from (transitory) migration to coastal cities. The upper part of table 8.2 presents the activities of each household type in terms of contribution to value added. In the first column for the unlinked house- holds with no draft power, it can be seen that crops are the dominant source of farm value added. One-season rice contributes 9.4 percent to value added, and the more intensive two-season rice contributes as much as 28.5 percent. Other crops, such as vegetables, contribute another 21.2 percent to household value added. The share of livestock is rather limited, with pigs and other livestock each contributing 0.1 percent. For this household group, the total contribution of agricultural activities to value added is 59.2 percent, and the remainder is coming 228 Poverty and the WTO: Impacts of the Doha Development Agenda Table 8.1. Shocks Administered to Village Model (% change with respect to base) Full liberalization Doha Scenario Prices Employment Prices Employment (scenario A) (scenario B) (scenario A) (scenario B) Agricultural outputs One-season rice 7.4 n.a. 1.3 n.a. Two-season rice 7.4 n.a. 1.3 n.a. Other crops 4.1 n.a. 1.1 n.a. Pigs 5.0 n.a. 1.3 n.a. Other livestock 5.0 n.a. 1.3 n.a. Agricultural inputs Fertilizer 0.6 n.a. 0.2 n.a. Herbicides 0.6 n.a. 0.2 n.a. Pesticides 0.6 n.a. 0.2 n.a. Seed 5.1 n.a. 1.3 n.a. Purchased feed 5.1 n.a. 1.3 n.a. Other inputs 0.9 n.a. 0.3 n.a. Consumption goods Food 3.3 n.a. 0.9 n.a. Processed food 3.3 n.a. 0.9 n.a. Nonfood 1.8 n.a. 0.6 n.a. Durables 0.0 n.a. 0.0 n.a. Other expenditures 1.8 n.a. 0.6 n.a. Wages Nonagricultural employment n.a. 2.2 n.a. 0.7 Migration, inside province n.a. 2.2 n.a. 0.7 Migration, outside province n.a. 0.6 n.a. 0.3 Outside-village employment Nonagricultural employment n.a. 1.8 n.a. 1.5 Migration, inside province n.a. 1.8 n.a. 1.5 Migration, outside province n.a. 2.2 n.a. 1.9 Source: Authors' simulations Note: n.a. = not applicable under this scenario. Each of the complete liberalization experiments applies a combined shock of prices and employment (that is, scenario A + scenario B = complete liberalization ­ either Full Liberalization or Doha). A Village-Level Study of the Impact of the Doha Round on Rural China 229 from engaging in off-farm activities. Hiring out its labor to other villagers con- tributes 1.3 percent to value added, and working in local businesses earns 19.4 percent of value added. Outside-village employment is a significant income source for this household group, with a share of 18.1 percent. For this unlinked household, migration opportunities are restricted to moving inside the province, which contributes 1.8 percent of value added. For all household types, off-farm employment contributes a significant share of income, but there are important differences in the nature of nonagricultural income sources. Comparison of the two household groups that have no link out- side the province shows that households lacking draft power are oriented more toward local off-farm employment. The households with draft power obtain 71.4 percent of value added from agriculture, which is similar to the importance of agriculture for the other household group owning draft power. The household group with access to migration but lacking draft power derives about 40 percent of value added from outside-province migration and derives only about 45 per- cent of value added from agricultural activities. Differential access to agricultural and migration opportunities is thus reflected in the composition of household value added. Differences in activities result in differences in income patterns across the four household groups. The bottom part of table 8.2 presents income per adult con- sumer equivalent3 to allow a direct comparison across households. As a crude poverty assessment for each household group, available income is computed in terms of U.S. dollars per day. Three of the household groups fall in between the one dollar and two dollars per day poverty lines. The notable exception is the household group with an outside link and lacking draft power, with just over two dollars per day. This is in line with the rural-urban income differences because this household group specializes in outside-province migration. To get a clear view of differences in household endowments that affect house- hold response, the middle part of table 8.2 details income sources. Specific fea- tures of the village social accounting matrix (SAM) and equilibrium model arise in this table. For example, income from labor and irrigated land is split between shadow income4 and above shadow income. The household survey data reveal imperfect labor and land markets. Households are involved in a variety of off- farm activities with different wages. These wages are well above the estimated shadow wage, indicating restricted access to off-farm employment and suggesting a situation of labor surplus at the local level. This is not unexpected, given the high population density in rural China and similar findings in a study by Bowlus and Sicular (2003). 230 Poverty and the WTO: Impacts of the Doha Development Agenda Table 8.2. Activities and Income by Household Group Link outside province: No link Link Owning draft power: No Yes No Yes Village N = 78 100 256 295 729 Composition of activities (% value added) Agriculture One-season rice 9.4 10.8 8.2 10.5 9.5 Two-season rice 28.5 28.2 18.6 27.1 23.9 Other crops 21.2 24.7 18.4 22.4 20.9 Cattle n.a. 7.5 n.a. 8.1 4.1 Pigs 0.1 0.1 0.2 0.4 0.3 Other livestock 0.1 0.1 0.0 0.0 0.0 Village employment Agricultural labor 1.3 2.2 n.a. 0.1 0.4 Local business 19.4 13.5 2.4 2.7 5.2 Outside village Outside employment 18.0 12.9 9.1 5.4 8.9 Migration Inside province 2.1 n.a. 4.2 0.9 2.3 Outside province n.a. n.a. 38.8 22.3 24.6 100 100 100 100 100 Sources of household income (% total income) Labor Shadow wage income 57.4 52.8 52.6 62.1 57.4 Above shadow wage income 17.5 12.3 23.3 6.0 13.5 Land Irrigated land shadow income 12.4 13.3 12.8 11.4 12.1 Irrigated land, above shadow income 4.8 6.5 1.4 4.5 3.7 Nonirrigated land 7.3 7.9 7.7 6.6 7.2 Capital Cattle n.a. 3.3 n.a. 3.1 1.8 Tractor n.a. 1.6 n.a. 0.8 0.6 Transfer Within-village transfers 0.6 0.3 0.01 0.5 0.3 Receipts from outside the village 0.0 1.9 2.2 5.0 3.3 100 100 100 100 100 Income per adult consumer equivalent Annual income in 1,000 yuan (Y) 2.2 2.3 3.0 2.7 2.7 Income in U.S. dollars per day 1.5 1.6 2.1 1.9 1.8 Source: Authors' simulations. Note: n.a. = not applicable. The household is not involved in this activity. A Village-Level Study of the Impact of the Doha Round on Rural China 231 In the SAM and village equilibrium model, the demand-constrained labor market is accounted for by valuing labor against household-specific shadow wages, estimated using the household survey data. In the case of off-farm activi- ties, labor then earns revenue above the shadow wage, which is tracked in a sepa- rate account of the SAM. For example, for the unlinked households with no draft power, labor is the most important endowment, contributing 74.9 percent to its income, broken down into 57.4 percent coming from shadow wages and 17.5 per- cent from above shadow wages. Although there is a rental market of sorts for irrigated land (paddy fields), the village model does not include a land market. Analysis of the village trade in land showed that all four household groups are net renters of land. This is due to a bias in the surveyed sample of households, which excludes households that have migrated from the village. These households are renting out land for a price below its productive value. This difference can be interpreted as an insurance premium the households are willing to pay to maintain access to their land, which is collec- tively owned, in case they need to return to the village. The households remaining in the village thus get an indirect transfer of money from the migration of entire households, through having to pay less than the productive value for land rented. Analysis of the migration of entire households is beyond the scope of this chap- ter's model. Therefore the supply of land is fixed at the level observed in the SAM, effectively removing the land market from the model. Taking again the example of the unlinked households with no draft power, the return earned on irrigated land endowment contributes 17.2 percent to its income, broken down into a shadow rent component of 12.4 percent and the above shadow rent component of 4.8 per- cent. This above shadow rental income results from renting in land at a price below its marginal production value from migrant households. Nonirrigated land contributes another 7.3 percent to the household income. A last remark on the village SAM pertains to the lack of data for modeling capi- tal flows in the village. The SAM shows that the household group most involved in migration is a net supplier of capital to the other three groups of households. Although the survey contains some data on the conditions in which such funds are loaned, insufficient information is available to model a village-level capital market. It is therefore assumed that the household group lacking draft power but having an outside link spends a fixed share of its income on within-village transfers. These transfers are allocated to the three household groups based on their share of trans- fers in the SAM. The model thus includes a rather simple mechanism through which the income from migration is transmitted through village linkages. To summarize: this chapter analyzes the response of four different types of households, distinguished on the basis of their access to agricultural income and income from outside-province migration. Analyzing sources of income pointed to 232 Poverty and the WTO: Impacts of the Doha Development Agenda imperfect labor and land markets. These are accommodated by estimating house- hold-specific shadow prices, introducing profits earned from off-farm employ- ment and renting of land, and modeling household production and consumption decisions as nonseparable. The Village Equilibrium Model Despite introducing nonseparability of household production and consumption decisions, the mathematical structure of the model closely resembles macro-level general equilibrium models. Consumption decisions are modeled through a lin- ear expenditure system, and production is modeled by nested constant elasticity of substitution (CES) functions. Table 8.3 summarizes the key substitution elas- ticities for each activity. The estimation procedure for obtaining these substitution elasticities exploits the interhousehold variation in the survey data. The nesting structure differs across activities and is determined by statistical testing based on pairwise comparisons. Kuiper (2005) provides full details of this method. The village model does not attempt to treat two-way flows of commodity trade with the outside world. Households consume farm output but do not purchase these goods from outside the village or other households in the village. Household sales to outside-village markets are thus equal to total production minus house- hold consumption. Village markets exist for traction by draft animals or tractors and locally pro- duced consumption goods. Of these village markets, only animal traction has an endogenous village price in the model. The SAM indicates that only limited use is made of the tractors. This underuse of available tractors is therefore modeled through fixed prices for tractor services, the volume of which adjusts endoge- nously to demand. Off-farm employment options were found to be restricted, resulting in wages exceeding the shadow price of labor. This is handled in the village equilibrium model by fixing the levels of outside-village employment and having households earn a profit above labor costs from off-farm activities. Levels of village employ- ment (agricultural and nonagricultural) cannot be fixed, although for these activ- ities, wages also exceed shadow wages. Agricultural employment is therefore assumed to be demand driven, with prices being exogenously fixed.5 Demand for nonagricultural labor in the village economy is linked to local business activities, to which the chapter now turns. Because of lack of data on other inputs, local business activities use only labor (village nonagricultural labor), yielding a return that exceeds the shadow wage. All households are involved in local business activities, and all of them purchase locally produced goods. This reflects a heterogeneity in goods not captured by the A Village-Level Study of the Impact of the Doha Round on Rural China 233 Table 8.3. Substitution Elasticities for Cropping Activities (Village Average) Animal Other Land Labor traction Tractor inputs One-season rice Labor 0.39 Animal traction 0.39 1.87 Tractor 0.39 1.87 79.21 Other inputs 1.72 2.09 2.09 2.09 1.84 Two-season rice Labor 0.66 Animal traction 0.66 0.66 Tractor 0.66 0.66 53.65 Other inputs 0.98 0.98 0.98 0.98 1.35 Other crops Labor 0.33 Animal traction 0.33 2.88 Tractor n.a. n.a. n.a. Other inputs 0.57 1.41 1.41 n.a. 1.06 Labor Crop residues Purchased feed Pigs Crop residues 1.53 Purchased feed 1.53 1.53 Other inputs 1.53 1.53 1.49 Other livestock Crop residues 0.87 Purchased feed 0.87 0.87 Other inputs 0.87 0.87 0.87 Source: Kuiper (2005). Note: n.a. = not applicable. Elasticities as well as the structure of the production functions are calibrated with the survey data. For details, see Kuiper (2005). Because of differences in cost shares, substitution elasticities vary slightly by household group. aggregates used in the SAM and village model. Because of a lack of data, village prices of local goods are fixed to deal with the gap between product prices and costs of labor. Assuming fixed prices seems justified, because prices of village-pro- duced goods are common knowledge, and shadow wages cannot be observed. Given the unobservable character of shadow wages, it seems unlikely that a 234 Poverty and the WTO: Impacts of the Doha Development Agenda change in labor costs will be reflected by a change in the village price. A second reason for fixing prices of local business activities is the absence of a peak season. Production can therefore be shifted to times when little labor is needed in agricul- ture, limiting the need to increase the price when shadow wages increase. For all demand-driven activities (local consumption goods, hired agricultural labor, tractor services), market equilibrium is established by allocating demand to suppliers based on the initial market shares recorded in the SAM. Finally, all surveyed households are net sellers of agricultural production--that is, they begin in regime 3 of figure 8.1. The simulations may result in a regime change for households, possibly turning some household groups into net buyers. Lacking observations from the survey, this study uses an estimate of transaction costs for rice from Park and others (2002) to set the width of the price band at 25 percent of the selling price. Thus, households will become net buyers if their shadow price rises 25 percent above the initial selling price. Thus, to summarize, the village equilibrium model resembles macro-level gen- eral equilibrium models (of the sort used throughout this book) in the way in which consumption and production are modeled. A major difference with macro- and existing village-level models is household-specific production, which is affected by household consumption decisions through endogenous household shadow prices. Lack of data resulted in most village markets being modeled as fixed-price, demand-driven equilibria. The only exception is the village market for animal traction, which is balanced through an endogenous village price. House- hold production and consumption decisions are calibrated on the household sur- vey data, resulting in household-specific demand and supply functions. Full Liberalization Impacts There are two major pathways through which trade liberalization affects house- holds: changes in prices of consumed goods, agricultural inputs and outputs, and changes in off-farm employment and wages. This section first analyzes each of these two pathways separately before looking at the combined impact of the full- liberalization scenario. The Impact of Price Changes with Full Liberalization The discussion of price changes focuses on the changes in production. Prices of consumption goods increase as the overseas demand for China's products increases strongly and the country experiences a real appreciation (recall chapter 3). The price increases in agricultural output, however, outstrip the increased cost of consumption. More important, whereas households have limited opportunities A Village-Level Study of the Impact of the Doha Round on Rural China 235 to change their consumption patterns, they have much more flexibility in chang- ing their mix of production. With full liberalization, all household groups increase other livestock produc- tion, and three of four intensify rice production and shift toward two-season rice (table 8.4). Two forces account for these shifts in production. First, livestock pro- duction is cash constrained because of the absence of a credit market. The rise in output prices increases the availability of cash for all households, resulting in an expansion of previously constrained livestock production. The switch to other livestock instead of pigs is due to differences in input use. Pig production uses purchased feed, which experiences a strong price increase of 4.1 percent; external inputs used in other livestock production increase by only 0.4 percent, which is well below the rise in output prices. The second driving force behind the shift in production patterns is the increase in rice prices, making more intensive rice production attractive. Rice production can be intensified by switching from one-season to two-season rice, thus doubling the use of the available irrigated land. Having two cropping Table 8.4. Household Production and Marketed Surplus with Full-Liberalization Scenario A (exogenous increase in prices for outputs, inputs and consumption) (percentage change) Link outside province: No link Link Owning draft power: No Yes No Yes Village Production Crops One-season rice -18.8 90.3 -46.6 -67.4 -31.8 Two-season rice 8.5 -32.5 23.6 32.7 17.3 Other crops -1.1 -1.9 0.0 -1.7 -1.1 Livestock Pig production -31.1 -83.8 -5.2 -24.6 -22.9 Other livestock 432.1 3093.4 56.5 367.6 650.0 Marketed surplus Crops One-season rice -39.5 125.6 -90.6 -100.0 -52.4 Two-season rice 19.4 -100.0 68.0 91.7 45.4 Other crops -72.7 -36.4 0.4 -23.0 -16.1 Livestock Pig production -32.9 -100.0 -5.8 -38.6 -27.9 Other livestock 946.4 6539.5 207.1 1187.0 1923.5 Source: Authors' simulations. 236 Poverty and the WTO: Impacts of the Doha Development Agenda seasons strongly increases the demand for labor, which explains the opposite production response of the household group with draft power but lacking an outside link. This household starts to rent out draft power and invests the pro- ceeds in intensive (other) livestock production, and it reallocates labor from two- season rice production toward intensive livestock production. It ceases to be a seller of two-season rice and pigs, moving into regime 2 in figure 8.1. It almost becomes a buyer of pigs, with its household-specific price for pigs rising by 21 percent, but this price rise falls just within the 25 percent price band, and hence it does not yet become a buyer. Similarly, the fourth household in table 8.4 ceases to sell one-season rice. To summarize, changes in agricultural input and output prices increase the availability of cash, allowing an expansion of previously constrained livestock production. It further leads to an intensification of rice production for those household groups that have sufficient labor resources. The Household as a Supplier of Labor: The Impact of Increasing Off-Farm Employment Off-farm employment is an important source of income. For the village as a whole, 41 percent of income is generated from off-farm sources, both inside the village in local business activities and outside the village and even from employ- ment outside the province. For the household lacking draft power but having an outside link, for example, migration accounts for close to 40 percent of value added. An increase in off-farm employment opportunities is simulated in full-lib- eralization scenario B through rising wages and increased employment demand. To clarify the impact of this second pathway through which trade liberalization affects households, the analysis next abstracts from the price changes of inputs and outputs associated with full liberalization. Of course, per capita consumption is increased by the additional income. This holds especially for the households involved in migration, because the number of household members present in the village decreases. This also leaves more income for the remaining household members. Increased off-farm employment decreases the available agricultural labor force, which leads to less labor-intensive agricultural production for three of the household groups, resulting in a slight decrease of two-season rice and a marked increase of other livestock production (table 8.5). The driving force behind this response is the village market for animal draft services. With "linked" households moving to one-season rice, demand for animal traction is reduced and its price falls. This means that the renting in of draft power becomes cheaper for the household group with no animal traction and no outside employment. A Village-Level Study of the Impact of the Doha Round on Rural China 237 Table 8.5. Household Production and Marketed Surplus with Full-Liberalization Scenario B (exogenous increase in off-farm employment) (percentage change) Link outside province: No link Link Owning draft power: No Yes No Yes Village Production Crops One-season rice -1.4 -0.3 1.2 37.2 17.3 Two-season rice 0.3 0.2 -0.4 -16.2 -7.7 Other crops -0.1 -0.1 -0.3 -0.1 -0.1 Livestock Pig production 1.0 1.9 1.3 -1.6 0.5 Other livestock -15.4 -47.0 -9.4 52.4 4.0 Marketed surplus Crops One-season rice -3.7 -0.6 2.4 54.6 27.9 Two-season rice -0.3 -0.1 -1.2 -47.9 -22.1 Other crops -17.6 -1.8 -1.1 -0.8 -1.3 Livestock Pig production 1.0 2.2 1.4 -2.5 0.6 Other livestock -34.7 -100.0 -34.4 169.9 11.6 Source: Authors' simulations. The third household receives about 40 percent of its income from migrant labor. It responds to the employment opportunities generated by full liberaliza- tion by shifting resources out of agriculture and concentrating more on off-farm employment. It does, however, keep some rice production and pigs. Pig produc- tion uses less labor than other livestock and is thus a more attractive option with increasing shadow wages resulting from a rise in off-farm employment. The driving forces behind the diverging response of the fourth household group, those owning draft power and having a link outside the province, are endowments of labor and access to migration. The access to migration outside the province provides an important source of cash, just as it does for the other house- hold group with an outside link. The fourth household, however, has the largest labor endowment of all households, and this tempers the rise in its shadow wages. As a result, the labor and cash-intensive other livestock production is more attrac- tive than pig production for this one household group. It is interesting that the second household, with no outside link but with own- ership of draft power, chooses to stop selling other livestock. Its shadow price of 238 Poverty and the WTO: Impacts of the Doha Development Agenda other livestock rises just above the market price, and it becomes more attractive to use the output for own-consumption. To summarize the results in this section, an increase in off-farm employment reduces the agricultural labor force. Although the wage hike encourages a switch toward less labor-intensive pig production on the part of some households, the village as a whole shows little change in pig production, but a large increase in other livestock production. Combining Price and Employment Effects The above discussion shows that price effects of liberalization may move opposite to the effects of improved off-farm employment opportunities, depending on the initial household endowments and their links with the economy outside the vil- lage. The combined effect is summarized in table 8.6. At the village level, price and employment changes have an opposite impact on rice and pigs exported from the village. Where rising output prices promote rice Table 8.6. Household Production and Marketed Surplus with the Full-Liberalization Scenarioa Link outside province: No link Link Village Owning draft power: No Yes No Yes average Production Crops One-season rice -22.8 90.7 -42.0 -45.0 -20.4 Two-season rice 9.7 -32.1 21.5 23.0 12.3 Other crops -1.2 -2.1 -0.2 -1.5 -1.1 Livestock Pig production -30.5 -83.7 -4.2 -22.8 -21.8 Other livestock 419.8 3088.5 48.0 360.3 642.0 Marketed surplus Crops One-season rice -48.5 125.9 -81.6 -67.5 -34.1 Two-season rice 21.2 -100.0 62.0 61.9 30.5 Other crops -91.3 -40.1 -0.6 -23.8 -17.7 Livestock Pig production -32.3 -100.0 -4.7 -36.2 -26.7 Other livestock 918.6 6528.3 175.9 1162.1 1898.9 Source: Authors' simulations. a. Full-liberalization scenario = exogenous changes in prices (A) and exogenous increase in off-farm employment (B). A Village-Level Study of the Impact of the Doha Round on Rural China 239 and other livestock production, increasing off-farm employment opportunities tend to reduce the marketed surplus of labor-intensive agricultural output. When these two elements of the full-liberalization scenario are combined, the price effects dominate and the result is a more labor-intensive package of outputs. The net effect is a marked 37 percent rise in village-marketed surplus, mainly driven by expansion of other livestock production. To put this huge increase in perspective, it is important to be aware that the contribution of this commodity to the village's marketed surplus is just 2 percent in the base, whereas it becomes 29 percent in the wake of full liberalization. The other important surplus commodity is two- season rice, which contributes 62 percent of the village-marketed surplus in the base and still accounts for 50 percent in the full-liberalization scenario. Doha Impacts Having established the maximum potential impacts of trade reform on this village economy, the analysis now turns to the Doha scenario as outlined in chapters 2 and 3. The second set of bars in figure 8.2 summarizes production responses at the village level. For pig production and other livestock, the Doha scenario produces a less pronounced response than the full-liberalization scenario. This is to be Figure 8.2. Village Production Response under Alternative Liberalization Scenarios 70 60 Full liberalization 50 Doha 40 30 cent 20 Per 10 0 ­10 ­20 ­30 One-season Two-season Other crops Pig Other livestock rice rice production (per 1,000) Source: Authors' simulations. 240 Poverty and the WTO: Impacts of the Doha Development Agenda expected, because the exogenous price shocks are smaller under Doha. But in both cases, pig production decreases and other livestock production increases remark- ably (note that changes in other livestock production are expressed as per 1,000 instead of per 100 for expositional clarity). For rice production, a more interesting aggregate response emerges. Under the full-liberalization scenario, rice production becomes more labor- and land-inten- sive, with an increase in two-season rice and a decrease in one-season rice. This is due to the combined effect of households three and four moving into more inten- sive rice production in response to rising prices (table 8.6) while household two specializes in renting out traction power and consequently has to switch to less intensive own­rice production. Under the Doha scenario, however, there is a de-intensification of rice produc- tion in the village and a drop in aggregate rice output. Rice prices do not increase enough and shadow wages do not rise enough to induce the "linked" households to specialize in more labor-intensive forms of farming. For these linked house- holds, economic developments outside the village are paramount. Those linked households without draft power are the most engaged in outside-province migra- tion and are already able to realize substantial gains from the more modest Doha scenario by seizing the improved employment opportunities. Impacts on Inequality Growing income inequalities are now at the top of the policy agenda in China. The rural-urban income inequalities are transmitted to the village by asymmetric access to migration. The household group with the strongest involvement in migration also has the highest income per adult equivalent (see table 8.2). The increase in employment after trade liberalization may therefore be expected to increase within-village income inequality. The impact of an increase in agricul- tural output prices, however, may be expected to benefit the households owning draft power but lacking an outside link, because the activities of this household group are concentrated in agriculture. Table 8.7 summarizes income effects in terms of equivalent variation per adult equivalent. The simulated income gains are substantial, with an average increase of income over base levels as high as 21 percent under the full-liberalization experiment. Under the Doha scenario, the income gains are reduced to about 5 percent as a result of the smaller price changes facing the village economy. In both cases, above-average gains from price effects are observed for the household without an outside link but owning draft power. Under full liberaliza- tion, this household group gains Y 725 per adult equivalent--33 percent of its A Village-Level Study of the Impact of the Doha Round on Rural China 241 Table 8.7. Equivalent Variation per Adult Equivalent by Household Group (in yuan) and as Percentage of Base Adult Equivalent Income Link outside province: No link Link Village Owning draft power: No Yes No Yes average Full trade liberalization Prices 378 752 199 601 465 17% 33% 7% 22% 17% Employment 59 29 123 47 71 3% 1% 4% 2% 3% Prices and employment 441 806 321 702 563 20% 35% 11% 26% 21% Doha Prices 80 166 42 105 90 4% 7% 1% 4% 3% Employment 33 16 90 34 50 1% 1% 3% 1% 2% Prices and employment 121 168 131 133 136 5% 7% 4% 5% 5% Source: Authors' simulations. Note: Y 1 US$0.25; adult equivalents are corrected for the absence of migrants. base-level income per adult equivalent. Ownership of capital in the form of draft power is decisive for the relative size of the gains. Most gains from increased employment opportunities fall on the households engaged in outside employment and not owning draft power. Under full liberal- ization, this amounts to Y 123, 4 percent of its base household income. Employ- ment contributes 38 percent of the total gains under this scenario, and price changes contribute 62 percent (table 8.8). Under the more modest Doha scenar- ios, with limited price changes, the employment component contributes as much as 70 percent of the gains for this household group. In general terms, outside-vil- lage employment effects after trade liberalization indeed increase income inequal- ity within the village, but the welfare gains from employment are substantially smaller than the gains from prices changes. Combining both effects, it can be noted that the rising income inequality may be compensated by gains from spe- cialization for those who stay behind. The net effect on the within-village income distribution is determined by the interplay of initial endowments, village markets for inputs and outputs, and market imperfections. As a result, it appears that even 242 Poverty and the WTO: Impacts of the Doha Development Agenda poorer households begin to catch up. The households that have to rely on the use of own-labor on the household farm and are not endowed with traction power or a link to employment opportunities in the prospering coastal regions have fewer opportunities for adjustment. In fact, the only option for them is to farm rice more intensively and shift into labor-intensive other livestock production. Conclusions This study used an innovative village equilibrium model, which fully accounts for nonseparability of household production and consumption response. This allowed analysis of the impact of trade liberalization on agricultural supply response and off-farm employment, simultaneously accounting for household consumption decisions. The village model is used to analyze the impact of trade liberalization, which was quantified through macro-level shocks to the Chinese economy obtained from GTAP model simulations. The impact of price changes and labor demand, the two major pathways through which international trade affects households, were analyzed. The full-liberalization benchmark shows results that are well in line with the findings of national-level studies and the results of the household analysis in Huang, Li, and Rozelle (2003). Analysis of the impact of changes in agricultural input and output prices shows an increased village supply of rice and livestock Table 8.8. Contribution of Price Changes and Employment to Income Gains Link outside province: No link Link Village Owning draft power: No (%) Yes (%) No (%) Yes (%) average (%) Full trade liberalization Prices 86 93 62 86 83 Employment 13 4 38 7 13 Interaction effects 1 3 0 8 5 Total 100 100 100 100 100 Doha Prices 66 99 32 79 66 Employment 27 10 69 26 37 Interaction effects 7 -8 -1 -5 -3 Total 100 100 100 100 100 Source: Authors' simulations. A Village-Level Study of the Impact of the Doha Round on Rural China 243 other than pigs. As the cash constraint is lifted in the wake of rising incomes after liberalization, the households invest the proceeds in the capital-intensive activity of livestock production. The increased supply of rice is the result of more complex interactions, however, because some household groups increase rice production and others reduce rice production. Apart from influencing agricultural input and output prices, trade liberaliza- tion increases off-farm employment opportunities. The net impact of more off- farm employment is a decrease in rice supply, caused by an increasing scarcity of labor. Again, there is diverging household response, with some households increasing rice production as a result of lower costs of animal traction rented within the village. Employment and migration leads to less intensive rice production and a drop in village-marketed surplus. Combined with the price effects from full liberaliza- tion, increases in rice surplus can be observed. This is interesting because one household specializes in renting out traction services to the households engaged in migrant employment and decreases its own rice production. In terms of the vil- lage supply response, the impact of the change in prices thus dominates the impact of increased employment. The two pathways through which trade affects households thus have an oppo- site impact on household production response. Assessing the combined effect at the household level shows that the dominant aspect of trade liberalization depends on household endowments and production activities. Changes in intra- village specialization were observed, depending on the households' endowments and the strength of their linkages with the outside economy. It was also found that a strong involvement in off-farm employment does not necessarily imply that the employment aspect of trade liberalization dominates household response, thereby hampering ex ante judgments about the most relevant aspect of trade liberaliza- tion for a specific household type. A more modest liberalization within the context of the Doha Round has a dif- ferent impact on the village rice economy from that of full trade liberalization. The reason for this nonlinear response to increasing depth of trade reforms lies in the household-specific transaction (shadow) prices in combination with endogenous choices to participate in the output markets. Taking transaction costs into account, some households choose to withdraw from the market if their own shadow price is greater than the market price. Clearly, a partial reform scenario, such as Doha, leads to less pronounced output price changes than full liberalization. Under the Doha scenario, average income gains amount to about 5 percent, but under full liberalization, the gains are four times as high. However, the impacts vary by household type, and the question arises whether such changes will reduce or exacerbate existing inequalities in China. Whether trade liberaliza- 244 Poverty and the WTO: Impacts of the Doha Development Agenda tion allows incomes to grow together or grow apart depends on whether one accounts for the reduction in consumption demand when household members migrate. Assessing the net effect on the within-village income distribution shows that even poorer households are able to catch up. The households that have to rely on the use of own-labor on the household farm and are not endowed with trac- tion power or a link to employment opportunities in the prospering coastal regions have fewer opportunities for gains. Thus, although rural-urban migration can transfer benefits from economic growth in the coastal provinces to inland provinces, asymmetric access to migration implies that the rising rural-urban income differences are transferred as well. Notes 1. The estimation in this chapter also rejected the commonly assumed separability of factors and intermediate inputs, and this assumption was therefore dropped in the village equilibrium model. 2. Detailed description of the village equilibrium model and calibration procedures can be obtained from the authors: Marijke.Kuiper@wur.nl. 3. Adult equivalent instead of per capita consumption is used to account for differences in con- sumption between males and females and between age groups. Lacking survey data, conversion factors were taken from detailed consumption data of a study in Bangladesh (Zeller and others 2001). In addi- tion to differences in age and gender, consumer equivalents were corrected for the length of absence of household members due to temporary migration. 4. Nonseparability results in household-specific shadow prices that balance households' unobserv- able demand and supply. Therefore, an agricultural production function was estimated, explaining the total value of household output in terms of labor, land, manure, feed, and external inputs. The shadow prices are derived from this estimated production function as the marginal value product of each input. Specifically, for each household in the sample, the household-specific shadow prices for house- hold nontradables are derived as the marginal value product of each input. Averaging over the house- holds within a household group yields a shadow price for each household nontradable and each household group. These shadow prices are used in constructing the SAM and in calibrating the village equilibrium model. 5. Agricultural wages are in excess of shadow wages. Households are therefore always willing to supply additional agricultural labor when it is demanded. References Benjamin, D., and L. Brandt. 1999. "Markets and Inequality in Rural China: Parallels with the Past." The American Economic Review 89 (2): 292­95. Bowlus, A. J., and T. Sicular. 2003. "Moving towards Markets? Labor Allocation in Rural China." Jour- nal of Development Economics 71: 561­83. Cook, S. 1999. "Surplus Labour and Productivity in Chinese Agriculture: Evidence from Household Survey Data." The Journal of Development Studies 35 (3): 16­44. de Janvry, A., M. Fafchamps, and E. Sadoulet. 1991. "Peasant Household Behavior with Missing Mar- kets: Some Paradoxes Explained." Economic Journal 101: 1400­17. A Village-Level Study of the Impact of the Doha Round on Rural China 245 Démurger, S., J. D. Sachs, W. T. Woo, S. Bao, and G. Chang. 2002. "The Relative Contributions of Loca- tion and Preferential Policies in China's Regional Development: Being in the Right Place and Hav- ing the Right Incentives." China Economic Review 13: 444­65. Diao, X., S. Fan, and X. Zhang. 2003. "China's WTO Accession: Impacts on Regional Agricultural Income--a Multi-Region, General Equilibrium Analysis." Journal of Comparative Economics 31: 332­51. Gilbert, J., and T. Wahl. 2003. "Labor Market Distortions and China's WTO Accession Package: An Applied General Equilibrium Assessment." Journal of Comparative Economics 31: 774­94. Hare, D. 1999. "`Push' versus `Pull' Factors in Migration Outflows and Returns: Determinants of Migration Status and Spell Duration among China's Rural Population." The Journal of Develop- ment Studies 35 (3): 45­72. Ho, P. 2001. "Who Owns China's Land? Policies, Property Rights, and Deliberate Institutional Ambi- guity." The China Quarterly 166: 394­420. Huang, J., N. Li, and S. Rozelle. 2003. "Trade Reform, Household Effects, and Poverty in Rural China." American Journal of Agricultural Economics 85 (5): 1292­98. Huang, J., and S. Rozelle. 2004. "Agricultural Development and Policies in China." In China's Food Economy in the Early 21st Century; Development of China's Food Economy and Its Impact on Global Trade and on the EU, ed. F. W. van Tongeren and J. Huang, pp. 26­78. The Hague: Agricultural Eco- nomics Research Institute (LEI). Jones, D. C., C. Li, and A. L. Owen. 2003."Growth and Regional Inequality in China during the Reform Era." China Economic Review 14: 186­200. Jones et al., 2003 Kuiper, M. 2005. "Village Modeling--a Chinese Recipe for Blending General Equilibrium and House- hold Modeling." Ph.D. diss., Wageningen University. Löfgren, H., and S. Robinson. 1999. "Nonseparable Farm Household Decisions in a Computable Gen- eral Equilibrium Model." American Journal of Agricultural Economics 81 (3): 663­70. Murphy, R. 2000. "Migration and Inter-Household Inequality: Observations from Wanzai County, Jiangxi." China Quarterly 164: 965­82. Park, A., L. Brandt, and J. Giles. 2003. "Competition under Credit Rationing: Theory and Evidence from Rural China." Journal of Development Economics 71 (2): 463­95. Park, A., H. Jin, S. Rozelle, and J. Huang. 2002. "Market Emergence and Transition: Arbitrage, Transac- tion Costs, and Autarky in China's Grain Markets." American Journal of Agricultural Economics 84 (1): 67­82. Qiao, F., B. Lohmar, J. Huang, S. Rozelle, and L. Zhang. 2003. "Producer Benefits from Input Market and Trade Liberalization: The Case of Fertilizer in China." American Journal of Agricultural Eco- nomics 85 (5): 1223­27. Rozelle, S., L. Guo, M. Shen, A. Hughart, and J. Giles. 1999. "Leaving China's Farms: Survey Results of New Paths and Remaining Hurdles to Rural Migration." China Quarterly 158: 367­93. Rozelle, S., J. E. Taylor, and A. de Brauw. 1999. "Migration, Remittances, and Agricultural Productivity in China." The American Economic Review 89 (2): 287­91. Sadoulet, E., and A. de Janvry. 1995. Quantitative Development Policy Analysis. Baltimore and London: Johns Hopkins University Press. Taylor, J. E., and I. Adelman. 1996. Village Economies: The Design Estimation and Use of Villagewide Eco- nomic Models. Cambridge: Cambridge University Press. ------. 2003. "Agricultural Household Models: Genesis, Evolution, and Extensions." Review of Eco- nomics of the Household 1: 33­58. Winters, L. A. 2002. "Trade, Trade Policy and Poverty: What Are the Links?" The World Economy 25: 1339­67. Zeller, M., M. Sharma, A. U. Ahmed, and S. Rashid. 2001. "Group Based Financial Institutions for the Rural Poor in Bangladesh. An Institutional and Household Level Analysis." Research Report 120, International Food Policy Research Institute, Washington, DC. Part IV a focus on labor markets 9 Structural Change and Poverty Reduction in Brazil: The Impact of the Doha Round Maurizio Bussolo, Jann Lay, and Dominique van der Mensbrugghe Summary Over the medium time horizon, skill upgrading, intersectoral technological progress differentials, and migration of labor out of farming are some of the major structural adjustment factors shaping the evolution of an economy and its connected poverty trends. Our main focus is understanding, for the case of Brazil, how a trade shock interacts with these structural forces, and ascertaining whether it enhances or hinders medium term poverty reduction. A recursive-dynamic computable general equilibrium model simulates Doha scenarios and compares them against a Business as Usual scenario. The poverty effects are estimated using a microsimulation model that primarily takes into account individuals' labor sup- ply decisions. Our analysis shows that trade liberalization does indeed contribute to structural poverty reduction. However, unless increased productivity and stronger growth rates are attributed to trade reform, its contribution to medium term poverty reduction is rather small. Introduction In their review of the relationship between trade liberalization and poverty, Win- ters, McCulloch, and McKay (2004) conclude that trade liberalization "may be 249 250 Poverty and the WTO: Impacts of the Doha Development Agenda one of the most cost-effective anti-poverty policies available to governments" although they go on to note that it may not be the most powerful policy and its effectiveness is likely to vary substantially from case to case. In the medium- to long-run time horizon, economies adjust not only to trade policy reforms but also to many other changes, including technological progress, changes in the skill composition of the population, and varying consumption patterns. This chap- ter's main objective is to assess the role of trade liberalization in poverty reduc- tion over a time horizon during which these other structural trends are operat- ing. In particular, this chapter assesses the poverty impact of a Doha Round (and a full-liberalization) scenario on Brazil against a baseline scenario that incorpo- rates some of the main features of medium-run structural change but no changes in trade policies. Recent research has demonstrated that growth can differ tremendously in its potential to reduce poverty both across countries and over time.1 In high-inequal- ity countries such as Brazil, even a slight worsening of the income distribution can imply that growth has very little impact on poverty. Ascertaining how trade liber- alization affects the pattern of income growth is therefore a core part of the analy- sis of the nexus of trade and poverty in the longer run. The labor market is a key factor determining such impacts. Both changes in relative factor prices and changes in endowments play an important role in the medium to long run. Changes in sectoral employment can also contribute significantly to poverty reduction, because they may enable people to escape low-wage poverty traps. There is considerable evidence of the existence of such poverty traps that can arise in the presence of discrete occupational and technology choices and fixed costs (Barrett 2004). Moving out of agriculture--where poverty rates are often much higher than in other sectors--is one example of this type of occupational choice, and one of particular interest in the Brazilian context, where there has been a mas- sive reduction in agricultural employment in recent years. This reduction in agri- cultural employment may have contributed to poverty reduction, because poverty rates among agricultural households are considerably higher than among nona- gricultural households. Trade liberalization is expected to favor agriculture in Brazil. By retaining workers in agriculture, it may thus work against the "natural" forces of structural change with an adverse impact on poverty reduction. However, trade liberaliza- tion may also relieve some of the pressure on nonagricultural incomes resulting from out-migration from agriculture as incomes in that sector rise. This ambigu- ity in the poverty impacts of trade reform illustrates the necessity of quantifying each of these transmission channels to evaluate the overall poverty and distribu- tional impact of trade reform. The methodology used here combines a dynamic computable general equilibrium (CGE) model with a microsimulation (MS) Structural Change and Poverty Reduction in Brazil 251 model for Brazil. Using a time horizon of 15 years, a business-as-usual (BaU) scenario and two counterfactual trade reform scenarios are developed in the CGE model, and aggregate results on relative factor prices and resource movements from agricultural to nonagricultural sectors are linked to an MS. This macro- micro modeling framework enables analysis of the medium- to long-term poverty and distributional impact of different growth patterns. The chapter is structured as follows. It begins by providing some background information on the Brazilian case and motivation for the chapter's approach. Then, the macro and micro modules of the model are described. The results of the simulations are reported and commented on in the following section. The last sec- tion summarizes and concludes. Background and Motivation The main objective of this chapter is to assess whether trade reform favors the Brazilian poor. It is therefore important to know who the poor are, where they live, and especially how they earn their living. In addition, it should prove helpful to identify economic trends that have been particularly important for the poor. Brazil's per capita income has remained stagnant for much of the past 25 years, and the very unequal distribution of income has remained more or less unchanged. Accordingly, poverty in Brazil has remained fairly constant over the past 25 years (Bourguignon, Ferreira, and Lustig 2005; Verner 2004). In light of the substantial structural changes that have occurred over this period, especially increasing urbanization, a massive decline in agricultural employment, increasing unemployment, educational expansion, and demographic changes, this outcome appears "paradoxical" in the words of Bourguignon, Ferreira, and Lustig (2005). Ferreira and Paes de Barros (2005) explore this apparent paradox using an MS approach and show that these various features of structural change have tended to offset one another when it comes to poverty and inequality impacts between the years 1976 and 1996. Poverty in Brazil varies considerably among regions, rural and urban areas, and city sizes, with poverty rates being particularly high in rural areas, small and medium-size towns and the metropolitan peripheries of the north and northeast (Ferreira, Lanjouw, and Neri 2001). In 1996, the north and northeast accounted for 55 percent of the poor and 34 percent of the Brazilian population. At the national level, about 20 percent of the population lived in rural areas, contributing 35 percent of total poverty.2 The high poverty rates in rural areas, particularly in the north and northeast, are related to the predominance of agricultural employ- ment in these regions. The northeast had the highest share of agriculture in aggre- gate employment in 2001, with 34 percent compared to only 11.5 percent in the 252 Poverty and the WTO: Impacts of the Doha Development Agenda southeast.3 According to Ferreira, Lanjouw, and Neri (2001), 20 percent of all households had a household head employed in agriculture, and these households contributed 34 percent to overall poverty in 1996. Changes in poverty also differ widely across regions and activities. Verner's (2004) figures based on the Pesquisa Nacional por Amostragem de Domicílios (National Household Survey; PNAD)4 suggest that the poverty headcount in the northeast declined from almost 60 percent in 1990 to 42.3 percent in 2001, whereas poverty in Brazil's most populous state, São Paulo, rose slightly, from 8.6 to 9.4 percent during the same period. For urban areas, Ferreira and Paes de Barros (2005) show that extreme poverty increased between 1976 and 1996. In contrast, Paes de Barros (2004) reports that the poverty incidence among both rural households and those households engaged in agricultural activities declined from levels of about 60 percent to about 50 percent between 1992 and 2001. One important factor for understanding these developments is the structural change in Brazilian agriculture in the 1980s and 1990s. This has had a profound impact on both rural livelihoods and poverty in Brazil and living conditions in the urban areas through the migration of rural labor to the cities. With the exception of Paes de Barros (2004), research efforts in this direction, however, have focused on agricultural performance rather than on how this performance affects people's livelihoods. In their assessment of the impact of sector-specific as well as economywide reforms on Brazilian agriculture, Helfand and Rezende (2004) conclude that agri- culture became one of the most dynamic sectors in the Brazilian economy. Between 1980 and 1998, real GDP grew by about 40 percent and real agricultural output by about 70 percent. In many subsectors, agricultural yields increased sig- nificantly and the area devoted to export crops, in particular soybeans and sugar- cane, was expanded. Agriculture benefited from a favorable macroeconomic envi- ronment and trade reforms that led to less industrial protection coupled with elimination of taxes and quantitative restrictions on agricultural exports. In addi- tion, specific agricultural reforms--the reform of agricultural credit and price support policies, an agrarian reform program that included land reform, and the deregulation of domestic markets for agricultural goods--were important drivers of the observed agricultural performance.5 However, the increase in agricultural productivity was accompanied by a mas- sive layoff of hired labor and important changes in the size distribution of farms. According to the agricultural census from 1996, the number of small farms declined dramatically, and agricultural employment shrank by 23 percent between 1986 and 1996, although these figures should be taken with some caution (Helfand and Rezende 2004). Structural Change and Poverty Reduction in Brazil 253 Nonagricultural activities appear to have compensated for the loss in agricul- tural employment in rural areas, but unemployment rates in urban areas have risen in that period (Dias and Amaral 2002). The analysis in this chapter based on the 1997 and 2001 PNADs suggests that this decline in agricultural employment has continued after 1996. In 2001, agriculture accounted for 20.6 percent of employment in Brazil, down from 24.2 percent in 1997. Unemployment in rural areas has stayed constant at about 2.5 percent during this period, whereas urban unemployment has risen from 9.44 to 10.6 percent, an increase that may be related to the decline in agricultural employment.6 Fewer agricultural employment opportunities may also be one of the reasons for further urbanization in Brazil, although it is difficult to establish this link empirically, as explained in more detail elsewhere in this chapter. The rural popu- lation declined sharply in the past decade, falling from 24.41 percent in 1991 to 21.64 percent in 1996 (IBGE 1997) and 16 percent in 2001 (PNAD 2001). The trends in rural poverty mentioned above suggest that the described developments have improved rural livelihoods. Nevertheless, poverty rates in rural areas remain well above urban poverty rates. Future developments in agriculture are a subject of some debate, but it is likely that many of the recent trends, in particular the decline in agricultural employ- ment and the modest increase in incomes from agriculture, will continue. They are therefore incorporated in the BaU scenario in this chapter, against which the trade reform scenarios are to be judged. The analysis here addresses the poverty and distributional impact of some of the structural changes that are particularly relevant for Brazil. The focus is partic- ularly on structural change in agriculture, and how this interacts with trade poli- cies. Of course, the reader should bear in mind that more than two-thirds of the Brazilian poor either live in urban areas or derive their income from nonagricul- tural activities, and this model devotes relatively less attention to how structural change might affect them. The Modeling Framework The analytical framework consists of a sequentially dynamic CGE model linked to an MS. The MS takes the changes in factor and goods prices as given; hence, there is no feedback between these two parts of the model. This framework is particu- larly well suited for the questions at hand, because the CGE model captures some of the main features of structural change and the relative price changes accompa- nying them. The MS, in turn, then allows for a detailed empirical assessment of the household responses to these changes. 254 Poverty and the WTO: Impacts of the Doha Development Agenda The Macro Model A 1997 social accounting matrix (SAM) has been used as the initial benchmark equilibrium for the CGE model. This SAM has been assembled from various sources, including the 1997 input-output table, the earlier SAM assembled by Harrison and others (2003), and the 2001 PNAD. For purposes of this model, the full SAM--which includes 41 sectors, 41 commodities, 12 factors (skilled and unskilled labor by gender and by farm and nonfarm occupation, agricultural and nonagricultural capital, and land and natural resources), an aggregate household account, and other accounts (government, savings and investment, and rest of the world)--has been aggregated to a smaller size of 17 sectors and commodities and 7 factors (skilled and unskilled labor by farm and nonfarm occupation, capital, and land and natural resources). The CGE model is a standard neoclassical, recursive-dynamic general equilib- rium model, and the next subsections describe its main features. Given this chap- ter's focus on labor markets and dynamic structural trends, the following discus- sion emphasizes the modeling of factor markets and growth.7 Production Output is produced using nested constant elasticity of substitution (CES) func- tions that, at the top level, combine intermediate and value added aggregates. At the second level, intermediate inputs are obtained by combining all products in fixed proportions (Leontief structure), and value added is produced by aggregat- ing the primary factors. At this level, primary factors are a capital-labor bundle and an aggregate land input. Lower levels of the production function disaggregate capital and labor, and then labor, into different categories. Income Distribution and Absorption Labor income and capital earnings are allocated to households according to a fixed coefficient distribution matrix derived from the original SAM. As will be shown below, one of the main advantages of using the micromodule is the enrich- ment of this rather crude macro distribution mechanism. Private consumption demand is obtained through maximization of household-specific utility functions following the linear expenditure system (LES). Private savings are a fixed propor- tion of income. Once the total value of private consumption is determined, gov- ernment and investment demands8 are disaggregated into sector demands accord- ing to fixed coefficient functions. Structural Change and Poverty Reduction in Brazil 255 International Trade The model assumes imperfect substitution among goods originating in different geographic areas.9 Import demand results from a CES aggregation function of domestic and imported goods. Export supply is symmetrically modeled as a con- stant elasticity of transformation (CET) function. Producers decide to allocate their output to domestic or foreign markets responding to relative prices. The assumptions of imperfect substitution and imperfect transformability grant a cer- tain degree of autonomy of domestic prices with respect to foreign prices and pre- vent the model from generating corner solutions. To facilitate the incorporation of shocks from the global CGE model (recall chapter 3), export demand functions were added so that the increased market access accompanying multilateral trade liberalization scenarios can be simulated more precisely.10 No international import supply functions have been added; Brazil is treated as a price taker for its imports. The balance of payments equilib- rium is determined by the equality of foreign savings (which are exogenous) to the value of the current account. Factor Markets Two types of labor are distinguished, skilled and unskilled. These categories are considered imperfectly substitutable inputs in the production process. Moreover, some degree of factor market segmentation is assumed: capital and land are per- fectly mobile across sectors, natural resources are sector specific, and labor mar- kets for the unskilled are segmented between agriculture and nonagriculture, whereas skilled workers are fully mobile. The labor market specification is a key element of the model and an important driver of poverty and distributional results. Therefore, its specification calls for some clarification and justification. The segmentation of the labor market by skill has become a standard assumption in CGE modeling, and it is easily justifiable for the case of Brazil. The inequalities of Brazilian society in terms of educational endowments and, more important, access to education and on-the-job training certainly support this assumption, even over a longer time horizon. The assumption that the market for unskilled labor is further segmented into agricultural and nonagricultural activities is more controversial, particularly in light of its importance for the poverty and distributional results. To test the valid- ity of this assumption, the authors check whether incomes in agriculture are still below incomes in other sectors once the following wage determinants are con- trolled for: education, experience, gender, racial dummies, and employment status variables such as self-employment, seasonal employment, and employment in the informal sector. Additionally, to take into account price differentials across space, 256 Poverty and the WTO: Impacts of the Doha Development Agenda geographic variables capturing differences among Brazilian regions as well as a rural-urban dummy variable are included in the wage estimation. The largest nonagricultural sector (in terms of employment), "other services," is taken as the reference group. Regression analysis shows that, relative to this ref- erence group, agricultural labor incomes are significantly lower for individuals in similar circumstances.11 Underreporting of income, externalities linked to work- ing in agriculture, and other factors may partially explain this negative bias in agricultural incomes; however, the authors believe this earnings gap is also due to barriers to mobility between agricultural and nonagricultural employment that prevent individuals from moving out of the agricultural sector. The econometric analysis (see also section "Who Moves Out of Agriculture") identifies two such barriers that are relevant over a medium-run time horizon: land ownership and the specificity of human capital acquired in agricultural occupations. With this empirical support for the hypothesis that the Brazilian labor market for unskilled labor is segmented into agricultural and nonagricultural employ- ment, the dual labor market for unskilled workers is modeled following the stan- dard Harris-Todaro specification, whereby the decision to migrate is a function of the expected income in the nonagricultural (urban) segment relative to the expected income in the agricultural (rural) segment. Model Closure The equilibrium condition on the balance of payments is combined with the other closure conditions so that the model can be solved for each period. The gov- ernment budget surplus is fixed, and the household income tax schedule shifts, to achieve this predetermined net fiscal position for the government. Second, invest- ment must equal savings, which originate from households, corporations, govern- ment, and the rest of the world. Aggregate investment is set equal to aggregate sav- ings, and aggregate government expenditures are exogenously fixed. Growth Equations Sectoral shifts among agriculture and nonagriculture and human capital upgrad- ing are two of the main features that have characterized recent growth processes in Brazil and in most developing nations. To capture these features in a transparent and simple dynamic framework, productivity growth rates are calibrated sepa- rately for the agriculture and nonagriculture sectors. Brazilian agriculture has his- torically recorded high productivity growth, and this exogenous historical growth rate for productivity in agriculture is imposed uniformly across all factors in that sector. In contrast, the growth rate of productivity for the nonagriculture sector is calibrated by imposing an exogenous growth path for real GDP. This dynamic cal- ibration results in the observed labor savings in agriculture production trends of Structural Change and Poverty Reduction in Brazil 257 the past decade continuing in the forecasting period.12 Other elements of simple dynamics include exogenous growth of labor supply, with skilled labor growing faster than unskilled labor, and investment-driven capital accumulation. The MS Model The micro model is linked to the macro model through changes in the following set of endogenous variables: (a) changes in agricultural and nonagricultural labor income of unskilled labor (two variables), (b) changes in labor income of skilled labor (one variable), and (c) changes in the sectoral (agriculture versus nonagri- culture) composition of the unskilled workforce (one variable). In addition, the fact that unskilled and skilled labor supplies grow at different rates is taken into account. The MS does not produce a series of cross-sections through time, but only simulates one cross-section that reflects the cumulative changes in the afore- mentioned exogenous and endogenous variables over the entire period from 2001 to 2015. In accordance with the structure of the CGE model, the micro model simulates the decision to move from agriculture into nonagriculture sectors only for unskilled workers. The MS module consists of a set of equations that describe the income genera- tion process of the household. It includes logit equations for moving out of agri- culture, estimated separately for household heads and nonheads. The wage-profit equations are estimated separately for unskilled agricultural, unskilled nonagri- cultural, and skilled labor using ordinary least squares. Together, the mover-stayer model and the wage-profit equations provide the basis for the MS of household- level outcomes. The left-hand-side variable of the mover-stayer model is a dichotomous variable that assumes a value of 1 if an individual has moved out of agriculture during the past 12 months and 0 otherwise. The model is estimated on a sample that includes stayers in agriculture along with last year's movers. An overview of key estimation results is provided below. The wage-profit equations explain between 30 and 50 percent of variability of log wages-profits using a rela- tively short list of explanatory variables, including education, work experience, gender, racial, and regional dummies. The estimation of agricultural wages and profits also controls for the number of nonremunerated household members. The 2001­15 MS involves three steps. First, households are reweighted to reflect the change in the skilled to unskilled labor ratio, as predicted by the CGE model over this period. In the second step, unskilled labor moves out of agriculture until the new share of unskilled labor in agriculture given by the CGE is repro- duced. Third, wages and profits are adjusted according to the CGE results, taking into account the changes in the skill composition of the workforce as well as the sectoral movements of unskilled labor from agriculture into nonagricultural 258 Poverty and the WTO: Impacts of the Doha Development Agenda sectors. In sum, by using the estimated equations, the MS is "forced" to reproduce the aggregate results for employment and wage changes generated by the CGE model. Technically, this requires changing the constants in each of the equations.13 To account for total household income in addition to labor income, transfer and capital income as reported in the PNAD are considered. Transfer income is scaled up or down according to the GDP per capita growth rate, and capital income is adjusted according to the change in the rental rate on capital as reported in the CGE model. The sum of all household members' individual incomes is divided by the number of household members to give the household income per capita. Regional poverty lines are developed by taking the R$80 per capita poverty line (in current 2001 prices) for urban Rio de Janeiro as a basis and adjusting it for regional price differences following Paes de Barros (2004). Who Moves Out of Agriculture? The "employment history" section of the PNAD is key to the analysis of the deci- sion to move out of agriculture. This data, which is nonexistent in most other countries' household surveys, offers the information needed for estimating the intersectoral migration choice model. In this section, the PNAD provides addi- tional data that allows identification of the movers out of agriculture and, very important for this undertaking, the characteristics of these individuals at the time of moving. For example, the PNAD reports which type of land right they had and whether they were self-employed before they moved out of agriculture. This information has not been previously exploited by researchers, and it is key to the findings here. Estimation of the mover-stayer model using these data allows highlighting of the main factors affecting the propensity to move out of agriculture. For both heads and nonheads of households, a higher level of educational attainment pos- itively influences this propensity, whereas age is one of the most significant factors that negatively affects the choice of moving. As one would expect, older individu- als are less likely to move out of agriculture. Owning land or other agricultural production factors, such as livestock, also appears to act as an important barrier to intersectoral movements. Finally, household heads from the north are more likely to move out of agriculture than those elsewhere in Brazil. Household heads appear to respond to intersectoral wage differentials to a lesser degree than other family members, thus showing a tendency to be "trapped" in agricultural activi- ties, possibly as a result of factor market imperfections. However, their decision to stay or move is of great importance for the choice of other household members. For these individuals, the strongest determinant of moving out of agriculture is a dummy indicating whether the household head is employed in a nonagricultural Structural Change and Poverty Reduction in Brazil 259 sector. Furthermore, the decision on the part of the household head to leave agri- culture also strongly influences the choice of the nonheads. Nonremunerated nonheads of households are less likely to move out of agriculture, a finding that points toward the importance of positive externalities associated with this type of agricultural employment. Brazil in the Next Decade: A Baseline Scenario A central question of this chapter involves assessing the poverty effects of trade policy reforms over the longer run when the forces of structural adjustment shape the income generation process. The starting point is the CGE model used to build a BaU scenario depicting the evolution of the Brazilian economy over the next decade. This baseline scenario should not be considered as a statistical forecast, but rather as a consistent "projection" of the economy into a future where inter- sectoral productivity growth differentials, skill upgrading, and migration of labor out of farming activities play major roles. This BaU scenario sets the backdrop against which the alternative scenarios involving trade policy reforms can be eval- uated. The next subsections describe in detail the macro and micro results for the BaU and trade scenarios. Macroeconomic Characteristics of the Baseline In the BaU scenario, real GDP for Brazil is projected to grow (from 2005 onward) at the annual rate of 3.3 percent; this is optimistic when compared to the 1980­2000 rate of 2 percent. The projected GDP growth performance is sup- ported by strong factor productivity growth rates. As explained above, productiv- ity in the agriculture sector is assumed to be factor neutral, and its growth rate is exogenously set at 2.9 percent per year; in the nonfarm sectors, growth of labor productivity is calibrated at 1.02 percent per year and growth of capital produc- tivity at 0.82 percent per year. The changes in the structure of labor markets, shown in table 9.1, are of partic- ular relevance for poverty and income distribution trends. As can be seen, the dif- ferences in productivity growth rates across sectors, combined with faster growth in the supply of skilled versus unskilled labor (education increases the supply of skilled workers, which is growing at a 2.0 percent annual rate versus a yearly 1.6 percent growth rate for the unskilled labor supply), generate structural adjustments in line with those observed for the last decade. This includes contin- ued out-migration of unskilled workers from agriculture. The declining labor demand in agriculture is driven by three factors: the relatively higher rate of labor 260 Poverty and the WTO: Impacts of the Doha Development Agenda productivity growth in agriculture relative to the rest of the economy, an income elasticity of private consumption for agricultural commodities that is less than one, and international prices for traded agricultural products decreased through time in the BaU scenario. These trends in the supply and demand for labor are equilibrated by move- ments in relative wages. Over the next decade, real wages of skilled labor are pro- jected to increase at 1.3 percent annually. In nonagricultural sectors, wages for unskilled workers increase at the annual rate of 0.9 percent; however, their upward trend is dampened by migration of unskilled workers from agriculture. The latter contributes to a five-percentage-point reduction in agricultural labor supply, leading to higher agricultural wages, which are growing at an annual rate of 1.7 percent over the baseline period, thereby narrowing the agriculture-nonagricul- ture wage gap. The BaU macroeconomic market trends are linked to developments at the sec- toral level (shown in table 9.2). Output growth rates are slightly lower for the agri- cultural sectors than for the nonagricultural ones. Agriculture exports grow at a slightly slower pace than nonagriculture exports because of falling primary com- modity international prices in the BaU scenario. In addition, productivity gains dictate that fewer workers are needed to achieve the same output. Meanwhile, ris- ing wages, in particular for unskilled workers, induce producers to substitute skilled workers for unskilled ones. The rightmost panel of the table shows the rel- ative skill intensities and employment sizes of each sector. Services are the largest employers of both skilled and unskilled workers, but, on average, they use skilled labor more intensively. Agriculture employs almost one-third of unskilled work- ers and uses this factor quite intensively, whereas manufacturing labor intensities fall in between agriculture and services. Distributional and Poverty Results for the BaU Scenario MS of these structural trends using the linking variables described above and Brazilian household data results in a moderate decrease in poverty between 2001 and 2015. Considering the full sample of households, the headcount poverty ratio (P0) declines by about 6 percentage points under the BaU scenario (see table 9.3). The reductions in the average normalized poverty gap (P1) and the poverty sever- ity index (P2) indicate that those who remain poor also become better off, thereby reducing the gap to the poverty line.14 Inequality changes very little, as indicated by the 0.1 decrease in the Gini coefficient. These indexes all indicate that some progress in reducing aggregate poverty and inequality would be achieved in a BaU scenario, but these aggregate measures may conceal relevant distributional changes at a more disaggregated level. Table 9.1. Medium-Term Labor Market Structural Adjustments Sector Productivity of Income Employment Wages Unskilled labor Cumulative labor elasticity of migration as % of: migration demand Receiving Skilled Unskilled Skilled Sending Unskilled 2001­15 population population Yearly growth Constant Yearly growth rate Yearly % rate Millions Structural Agriculture 2.9 .54 n.a. 0.0 n.a. 1.7 1.7 n.a. -4.0 Nonagriculture 1.0 1.05 n.a. 2.2 n.a. 0.9 n.a. 0.5 4.0 Economywide -- -- 2.0 1.7 1.3 n.a. n.a. n.a n.a. Change Source: Author's calculations. and Note: -- = not available; n.a. = not applicable. Poverty Reduction in Brazil 261 262 Table 9.2. The BaU Scenario's Output and Trade Sectoral Growth Rates and Employment Intensities Annual average growth rates Employment percentages Labor demand By sector By skill Poverty Sector Output Imports Exports Skilled Unskilled Skilled Unskilled Skilled Unskilled Cereal and grains 3.2 2.5 2.3 0.3 0.1 0 5 2 98 and Oilseeds 3.1 2.2 2.4 0.1 -0.1 0 1 6 94 the Raw sugar 3.2 n.a. n.a. 0.2 0.1 0 1 4 96 Other crops 2.9 1.3 2.5 0.0 -0.1 1 12 3 97 WTO: Livestock 3.2 1.5 n.a. 0.3 0.1 2 4 10 90 Raw animal products 3.3 2.5 1.6 0.4 0.3 0 3 1 99 Impacts Oil and minerals 3.3 3.0 2.9 1.5 1.7 0 0 15 85 Light manufacturing 3.3 0.8 3.7 1.0 1.2 1 2 16 84 Food industries 3.2 0.5 3.4 1.0 1.2 2 3 16 84 of Wood products the and paper 3.3 0.9 3.5 1.0 1.2 2 2 15 85 Doha Chemicals and petroleum products 3.3 1.8 2.9 1.1 1.3 2 1 30 70 Development Metals and mineral products 3.5 1.8 3.3 1.2 1.4 2 2 17 83 Machinery and equipment 3.6 1.9 3.5 1.4 1.6 3 2 28 72 Other services 3.0 2.6 1.7 2.1 2.3 58 30 33 67 Agenda Construction 3.2 n.a. n.a. 2.3 2.5 2 8 6 94 Trade and communications 3.1 2.4 1.8 2.2 2.4 15 18 17 83 Public services 3.1 2.7 1.7 2.2 2.4 9 4 41 59 Agriculture 3.0 1.9 2.4 n.a. 0.0 4 27 6 94 Nonagriculture 3.2 2.0 3.1 n.a. 2.2 96 73 26 74 Economywide 3.2 2.0 3.1 2.0 n.a. 100 100 24 76 Source: Authors' calculations. Note: n.a. = not applicable. Structural Change and Poverty Reduction in Brazil 263 Table 9.3. Poverty and Inequality in the BaU Scenario, by Sectors Variable All households Nonagricultural Agricultural households households 2001 2001­15 2001 2001­15 2001 2001­15 level change level change level change per capita income 314.9 1.5 351.9 1.2 148.3 2.3 Gini 58.6 -0.1 57.1 0.6 56.6 -0.7 P0 23.6 -5.6 18.6 -3.1 46.2 -13.8 P1 9.6 -3.0 7.1 -1.6 21.0 -8.0 P2 5.3 -1.8 3.7 -0.9 12.3 -5.2 Population (%) 100 n.a. 81.8 3.3 18.2 -3.3 Contribution to P0 n.a. n.a. 64.4 8.8 35.6 -8.8 Source: Authors' calculations. Note: n.a. = not applicable. Per capita income is 2001 reals, and the change is given as annual growth rate. All levels are in percent and changes in percentage points. Perhaps the most obvious way to gather more detailed information is to ana- lyze the poverty and inequality impacts separately for the agricultural and nona- gricultural households. A household is classified as "agricultural" when its head or at least two of its members are employed in agriculture. According to this classifi- cation, in 2001, agricultural households accounted for 18.2 percent of the Brazil- ian population, poverty incidence among them reached nearly 50 percent, and their contribution to total poverty was about 36 percent (see table 9.3). Between 2001 and 2015, the share of agricultural households in the population is projected to shrink by 3.3 percentage points after the decline in agricultural employment of more than 5 percentage points. Poverty among agricultural households falls by more than 13 percentage points (of agricultural population), whereas poverty among nonagricultural households decreases by only 3.1 percent. Accordingly, the contribution of agricultural households to the headcount falls by almost 9 per- centage points. A more detailed analysis also shows that the lack of progress in aggregate inequality is due to the fact that the agricultural and nonagricultural groups' individual inequality indicators move in opposite directions. Among nonagricul- tural households, inequality rises because skilled labor earnings, a major source of income for these households, grow faster than earnings from unskilled labor. Conversely, inequality among agricultural households falls, mainly because richer agricultural households earn a higher share of their income from nonagri- cultural labor. 264 Poverty and the WTO: Impacts of the Doha Development Agenda Another way of analyzing detailed distributional effects is to consider growth incidence curves. These curves plot per capita income growth at income per- centiles (Ravallion and Chen 2003) and are shown in figure 9.1 for all households as well as for the agricultural and nonagricultural subgroups.15 Per capita income growth is much higher for agricultural households, reflecting the increase in unskilled agricultural wages from the CGE model's results. In addition, the agri- cultural growth incidence curve illustrates a strong pro-poor distributional shift. The agricultural households' distributional shifts also explain the pro-poor changes in the national income distribution, because only minor distributional changes are registered in the nonagricultural distribution. However, richer nona- gricultural households experience somewhat higher gains than poorer house- holds. Incomes for the poor nonagricultural households increase by a meager 1­1.5 percent annually. These more detailed analyses of the long-term evolution of the Brazilian income distribution highlight the different roles played by changes in inequality and shifts in the growth rates of the average incomes. The following two questions then arise: if the current (2001) distribution of income were to remain unchanged, to what extent would the additional growth under the BaU scenario Figure 9.1. Growth Incidence Curves, BaU Scenario: All, Agricultural, and Nonagricultural Households 3 % All Agricultural in Nonagricultural 5. growth income capita per 522. Annual 11 0 20 40 60 80 100 Cumulative population ranked by per capita income Source: Authors' calculations. Structural Change and Poverty Reduction in Brazil 265 contribute to reducing poverty? And what is the role of the BaU sectoral differen- tial in growth rates for agriculture and nonagriculture in reducing poverty? Answering these questions requires performing two additional MSs. The first simulation generates a counterfactual distribution under the assumption that all incomes of all sources grow by 1.5 percent annually. This implies shifting the entire income distribution "to the right," leaving its shape unchanged. Individuals do not change employment sectors, and hence households retain their initial nonagricultural or agricultural classification. Results from this simulation are pre- sented in table 9.4 and changes are given as a percentage share of the BaU change (BAU change I). In addition, a second set of counterfactual distributions was sim- ulated for agricultural and nonagricultural households separately with per capita incomes of the respective household types growing with the BaU rates, that is, by 1.3 percent annually for nonagricultural and 2.4 percent annually for the agricul- tural households (BAU change II). Comparison of the counterfactual simulations of the "completely" distribu- tionally neutral (BAU change I) and the "separately" neutral (BAU change II) sce- narios shows that the growth bias in favor of agricultural households is poverty reducing. Yet, the difference between the BaU and the completely neutral scenario does not seem pronounced. This is due to the fact that in the latter, poverty among nonagricultural households is reduced much more than in the BaU, where the income distribution among these households worsens. This "slight" worsening of the income distribution significantly hampers the potential of growth to reduce poverty among nonagricultural households. In addition, the differences between the two neutral scenarios for nonagricultural households illustrate that a 0.2 per- centage point difference in annual growth rates for 14 years can make a substantial difference in terms of poverty reduction. The last two columns of table 9.4 illustrate the importance of growth for reducing poverty among agricultural households as well. A 0.9 percentage point difference in annual income growth rates for 14 years implies a reduction of about 5 percentage points in the headcount over this period. In contrast to what is seen for nonagricultural households, the impact of the pro-poor distributional shift for agricultural households observed in the BaU is relatively small. In other words, had the income distribution among agricultural households not improved, growth would have reduced poverty by only a little less. The poverty reductions recorded in the BaU scenario are due to a combina- tion of factors, including: the change in skill endowments, the increase in real factor prices, and the intersectoral movement of workers. A main advantage of MS is the ability to decompose the total effect in different partial effects that can be attributed to single causes. A slight complication arises because of the interac- tion effect among these three factors because incomes increase at different rates 266 Table 9.4. Poverty and Inequality in a Distributionally Neutral Scenario All households Nonagricultural Agricultural households households Poverty 2001 BaU BaU 2001 BaU BaU 2001 BaU BaU level change I change II level change I change II level change I change II (%) (%) (%) (%) (%) (%) and per capita income 314.9 100.0 100.0 351.9 117.7 98.6 148.3 65.7 102.9 the P0 23.6 91.7 102.4 18.6 139.8 133.3 45.9 56.5 90.5 WTO: P1 9.6 90.9 97.7 7.1 132.5 119.7 20.8 61.9 93.2 P2 5.3 86.8 97.9 3.7 125.6 114.3 12.1 62.6 93.4 Impacts Source: Authors' calculations. Note: This table shows results for two MSs. The first simulation generates a counterfactual distribution under the assumption that all incomes of all sources grow by of 1.5 percent annually. This implies shifting the entire income distribution "to the right" leaving its shape unchanged. Individuals do not change employment sectors, the and hence households retain their initial nonagricultural or agricultural classification. Results are presented in the columns with heading "change I" as percentage share of the BaU change (where, in fact, households change occupations and experience different gains according to the structure of their income sources). In the Doha second counterfactual, distribution for agricultural and nonagricultural households is shifted separately, using per capita income growth rates of the respective household types (1.3 percent annually for nonagricultural and 2.4 percent annually for the agricultural households), and results are denoted by the heading Development "change II." Agenda Structural Change and Poverty Reduction in Brazil 267 in agricultural and nonagricultural sectors. By simulating counterfactual distri- butions with only one or two of these changes included, it is possible to decom- pose the total effect into individual or joint (interactive) contributions. This is the subject of the next investigation. Figure 9.2 displays the results of the poverty decomposition for the BaU sce- nario. Factor price changes account for the largest share of total poverty reduc- tion. The change in the composition of the workforce (skill upgrading) does not contribute much to poverty reduction, whereas the sectoral shifts in the workforce are quite important, in particular for the poorest of the poor, as the higher contri- bution of the sectoral change component with regard to P2 indicates. This is because households with members moving out of the agricultural sector tend to escape poverty. The interaction component hampers poverty reduction (negative contribution in figure 9.2) because people moving out of agriculture experience a lesser rate of increase in their incomes over the BaU time frame. Figure 9.2. Decomposition of Poverty Changes, BaU Scenario, All Households 100 80 60 40 20 0 P0 P2 ­20 ­40 Skill upgrading Factor price change Sector change Interaction Source: Authors' calculations. Note: The figure displays the contribution of the component to the total change in P0 and P2, respec- tively, in percent. The contributions add to 100. Contributions refer to reductions in the respective poverty indexes. 268 Poverty and the WTO: Impacts of the Doha Development Agenda In sum, the distributional and poverty analysis suggests that the BaU scenario leads to modest poverty reduction. Agricultural households fare relatively well, and the poverty incidence and intensity among them are substantially reduced. Decomposition analyses show that sectoral change contributes significantly to poverty reduction, although factor income growth is the most important source of poverty reduction. Microaccounting exercises underline the importance of growth for poverty reduction, but they also illustrate that small increases in inequality can considerably reduce the poverty reduction potential of growth in the context of a high-inequality country, such as Brazil. With this background, the chapter now turns to the central question of this book: Can this rate of poverty reduction be enhanced by global trade reforms? Macroeconomic Impacts of Trade Reforms The trade shocks simulated in the dynamic CGE model consist of changes in Brazilian tariff protection against imports from the rest of the world and of exoge- nous changes of international prices of traded goods and export quantities demanded by foreigners.16 The shocks are assumed to take place progressively through a gradual phasing in starting in 2005 and lasting 6 years. Table 9.5 dis- plays these shocks as percentage changes of the final year (2015) between the BaU and the trade reform scenarios. In keeping with the other chapters in this volume, the government fiscal balance remains unchanged, thus tariff revenue losses are compensated by an equiproportional direct tax paid by households. This tax is the least distortionary instrument that can be readily used in this model; however, in practice, the Brazilian government may choose other forms of compensatory taxes, which may alter relative prices and have significant income distribution effects, as explored in other chapters in this volume. The full liberalization scenario has the largest impacts: tariffs are completely eliminated, and Brazil enjoys strong terms of trade gains. The Doha shocks gener- ate almost no tariff cuts in Brazil because of the extensive binding overhang (recall chapter 2), and they are accompanied by fairly muted global price effects. To fully anticipate their final effects, these shocks need to be mapped to the economic structure of Brazil. Table 9.6 presents this structure. For instance, in the full-liberalization sce- nario, export-oriented sectors (those displaying high shares of export to domestic output), such as oilseeds, other crops, and the food industries, record considerable increases of their export prices. Conversely, import-competing sectors, such as chemicals and oil-derived products and capital goods, do not face high increases in their international prices. These combined export and import price movements result in strongly favorable terms of trade (TOT) gains, inducing significant Table 9.5. Trade Shock: Tariff Reductions and International Price Changes Own tariff reductions Change in import prices Change in export prices Commodity Full Full Full Doha Doha liberalization Doha liberalization liberalization Cereals and grains -100 n.a. 8 2.1 16 6.0 Oilseeds -100 n.a. 6 2.5 14 4.9 Raw sugar n.a. n.a. 2 1.0 14 5.4 Other crops -100 0 2 0.9 13 4.8 Livestock -100 n.a. 2 1.1 25 9.8 Structural Raw animal products -100 n.a. 2 0.4 18 6.7 Oil and minerals -100 n.a. 0 0.1 2 1.3 Light manufacturing -100 0 1 1.2 9 4.0 Change Food industries -100 -1 0 0.6 7 3.2 Wood products and paper -100 -2 0 0.0 4 2.0 Chemicals and petroleum products -100 -3 -1 0.0 3 1.7 and Metals and mineral products -100 -1 0 0.0 3 1.7 Poverty Machinery and equipment -100 -2 0 0.0 2 1.7 Other services n.a. n.a. 0 0.0 5 2.2 Reduction Construction n.a. n.a. 0 0.0 4 1.9 Trade and communications n.a. n.a. 0 -0.1 5 2.1 Public services n.a. n.a. 0 -0.1 5 2.3 in Agriculture -100 0 5 1.5 14 4.9 Brazil Nonagriculture -100 -2 0 0.1 4 2.1 Economywide -100 -2 0 0.1 5 2.4 269 Source: Authors' calculations. Note: n.a. = not applicable. 270 Poverty and the WTO: Impacts of the Doha Development Agenda reallocation of resources toward export-oriented sectors. Additional push for this reallocation comes from Brazil's own liberalization, which entails a reduction of the antiexport bias implicit in the higher protection rates for manufacturing of the initial tariff structure. The sectoral effects projected in the wake of trade reforms detailed in the complete elimination of tariffs in the full-liberalization case explain the large increase of imports (measured in volume), which, in the final year of this scenario, is 21 percent above the value in the same year of the BaU. Increases in imports of agricultural goods are much weaker: an aggregate 6 percent increase versus the 21 percent surge of the nonagriculture bundle. The combination of lower initial tariffs and stronger international price increases for agriculture, relative to nonagriculture, explains the difference in import response of these two broad sectors of the Brazilian economy. Given the very limited scope of tariff reductions under the Doha scenario, import changes are much smaller. With a relatively high elasticity of substitution in demand (set uniformly at 4), cheaper imports have the potential to displace domestic production, especially for those goods whose demand is fulfilled by a large share of foreign supply. For Brazil, this is the case for the chemicals and capital goods sectors. In the full-liber- alization scenario, domestic production experiences significant output reductions in these sectors; however, this does not happen in the Doha scenario, where Brazil- ian tariffs are hardly reduced. The competition from cheaper imports is also reflected--again, only for the full-liberalization case--in the decline of prices of domestic output. These import-demand side effects are linked to the supply response, to which the analysis now turns. For producers of exportable goods, the reduction of prices in local markets combined with unchanged or rising export prices creates incen- tives to increase the share of sales to foreign markets. This export response (shown in the columns "Export volumes" in table 9.7) varies across sectors and is linked to the pattern of Brazil's comparative advantage and the increase in international prices. Brazil's comparative advantage can be ascertained by considering the export orientation ("Exports to domestic output") column in table 9.6, which highlights three sectors in particular: oilseeds, other crops, and the agricultural transformation industry. These sectors, which also enjoy large jumps in their international price, experience export surges. As a result of the generally positive export price shocks, other sectors join in an overall expansion of supply to foreign markets. Rising export sales more than offset, or at least compensate, reductions of domestic sales and lead to changes observed in the columns labeled "Domestic output" in table 9.7. Given the foreign closure rule for the Brazilian model, econ- omywide increases of import volumes are balanced by a comparable increase in exports.17 Table 9.6. Initial (Year 2001) Structure of the Brazilian Economy Imports to domestic Exports to Commodity Tariff rates Sectoral demand of Sectoral Sectoral domestic imports composite ouput exports output Cereals and grains 7 1 15 1 0 1 Oilseeds 6 0 8 0 4 29 Raw sugar 0 0 0 0 0 0 Other crops 9 2 3 4 8 7 Structural Livestock 3 0 1 1 0 0 Raw animal products 8 0 1 1 0 1 Oil and minerals 4 7 33 1 7 25 Change Light manufacturing 17 4 5 5 3 2 Food industries 18 3 3 7 19 11 Wood products and paper 9 2 5 3 7 10 and Chemicals and petroleum products 9 15 10 9 8 3 Poverty Metals and mineral products 12 5 6 5 13 11 Machinery and equipment 19 37 27 8 20 11 Other services 0 11 3 23 5 1 Reduction Construction 0 0 0 8 0 0 Trade and communications 0 10 5 13 5 2 Public services 0 2 1 11 1 0 in Brazil Agriculture 8 4 4 7 12 6 Nonagriculture 11 96 6 93 88 4 Economywide 11 100 6 100 100 4 271 Source: Authors' calculations. 272 Table 9.7. Brazil's Structural Adjustment, Percent Changes in the Final Year between BaU Scenario and Trade Shocks Domestic demand Price of Price Poverty of domestic domestic output in of domestic Import volumes products domestic markets Export volumes Domestic output output Full Full Full Full Full Full and liberal- liberal- liberal- liberal- liberal- liberal- Commodity ization Doha ization Doha ization Doha ization Doha ization Doha ization Doha the WTO: Cereal and grains -6 -3 4 1 -2 1 68 13 5 1 -2 1 Oilseeds -18 -7 5 1 -6 0 60 8 20 3 -3 1 Raw sugar 0 0 -2 1 n.a. n.a. 0 0 -2 1 Impacts Other crops 23 2 1 0 -1 1 6 -3 1 0 -1 1 Livestock -4 1 3 1 -2 1 n.a. n.a. 3 1 -2 1 Raw animal products 22 5 2 1 -2 1 5 -1 2 1 -2 1 of Oil and minerals -6 1 1 -1 -5 1 26 1 7 0 -4 1 the Light manufacturing 48 -3 0 1 -5 0 159 61 5 3 -4 1 Doha Food industries 59 1 0 0 -4 1 30 4 3 1 -4 1 Wood products and paper 23 4 -1 0 -4 1 11 -1 0 0 -4 1 Development Chemicals and petroleum products 18 3 -2 0 -4 1 9 -1 -2 0 -4 1 Metals and mineral products 24 2 -4 -1 -5 1 15 -1 -2 -1 -4 1 Machinery and equipment 42 3 -12 -1 -6 1 11 -2 -10 -1 -5 1 Agenda Other services n.a. n.a. 1 0 -4 1 n.a. n.a. 1 0 -4 1 Construction -14 3 0 0 -3 1 8 -1 0 0 -3 1 Trade and communications -12 3 0 0 -3 1 6 -2 0 0 -3 1 Public services -13 3 0 0 -3 1 7 -2 0 0 -3 1 Agriculture 6 -1 2 1 -2 1 22 0 3 1 -2 1 Nonagriculture 21 3 -1 0 -4 1 21 2 0 0 -4 1 Economywide 21 3 -1 0 -4 1 21 2 0 0 -4 1 Source: Authors' calculations. Note: n.a.= not applicable. Structural Change and Poverty Reduction in Brazil 273 In summary, trade reforms promote a production structure specialized toward exportables, which in Brazil translates into a specialization toward primary or agricultural transformation sectors. This agriculture export-led boom is fully achieved only in the full-liberalization scenario, where domestic tariffs are fully eliminated and there are strong international price changes.18 From the point of view of poverty and income distribution, changes in factor markets are the most important aspect of the structural adjustment caused by trade reform. Changes in wages and sectoral employment are linked to changes of goods prices through the production technology and the functioning of the factor markets. A key aspect of the different production technologies is the difference in factor intensity across sectors shown in table 9.2. Recall that this chapter seeks to mimic realistic adjust- ment possibilities in the labor market by assuming that skilled workers can freely move across all sectors, whereas unskilled workers face two segmented markets and can just imperfectly migrate from the agriculture to the nonagriculture seg- ment. As a result of the boom in agriculture, which is very intensive in unskilled labor, the full trade liberalization induces a significant increase in the wage rate for unskilled workers, as reported in table 9.8. When compared with the BaU sce- nario, the yearly rate of growth of wages of unskilled workers in agriculture is 0.4 percentage point higher, and this results in a cumulative 14 year growth of 34 per- cent--much higher than the cumulative growth of 26 percent under the BaU sce- nario. Migration decreases with higher agricultural wages. About 340,000 workers who moved out of agriculture in the BaU scenario no longer do so in the full-lib- eralization case. This has some effect on the aggregate distribution of unskilled workers between agriculture and nonagriculture, as shown in the last column of table 9.8. The Doha effects are much weaker. Distributional and Poverty Impacts of Trade Reform Two fundamental results emerge from analyzing the micro impacts of the trade scenarios. First, the initial hypothesis that trade liberalization, by working against the "natural" forces of structural change, might weaken long-term poverty reduc- tion, has been soundly rejected. Although fewer people migrate toward more highly paid nonagricultural jobs, poverty is further reduced in the trade liberaliza- tion scenarios, largely through increased agricultural incomes. However (and this is the second fundamental result), trade reform as envisaged in the core Doha sce- nario for this book--and even in the hypothetical full-liberalization scenario-- pales in importance in the fight against poverty in the face of the overall assump- tions about productivity and economic growth that govern the BaU scenario. The full-liberalization scenario leads to a further reduction in the headcount poverty index of 0.5 percentage point, whereas for the Doha scenario the effects are almost 274 Table 9.8. Factor Market Effects Scenario Employment Wages Unskilled labor migration Cumulative Unskilled as % of: migration employment Poverty Sector Sending Receiving Skilled Unskilled Skilled Unskilled 2001­15 2015 population population and Yearly growth rates Yearly % Millions % the BaU WTO: Agriculture -- 0.02 -- 1.68 1.66 n.a. -4.04 21.51 Nonagriculture -- 2.20 -- 0.91 n.a. 0.53 4.04 78.49 Impacts Economywide 2.0 1.7 1.26 -- n.a. n.a. 0 100.0 Full liberalization Agriculture -- 0.18 -- 2.10 1.51 n.a. -3.71 21.99 of Nonagriculture -- 2.15 -- 1.07 n.a. 0.49 3.71 78.01 the Economywide 2.0 1.7 1.32 -- n.a. n.a. 0 100.0 Doha Doha Agriculture -- 0.06 -- 1.78 1.62 n.a. -3.96 21.64 Development Nonagriculture -- 2.19 -- 0.93 n.a. 0.52 3.96 78.36 Economywide 2.0 1.7 1.27 -- n.a. n.a. 0 100.0 Source: Authors' calculations. Agenda Note: -- = not available; n.a. = not applicable. Structural Change and Poverty Reduction in Brazil 275 negligible. Of course, such trade reforms may well affect the rate of productivity growth, and hence the fundamental determinants of the BaU outcome, but this linkage is not explored here. As for the BaU scenario, a thorough assessment of the trade scenarios needs to go beyond these aggregate indicators and should rely on more disaggregate poverty and distributional analyses. In search of trade-induced poverty effects, the remaining part of this section considers an array of indicators, from growth inci- dence curves to poverty statistics estimated on specific subsamples of the survey data. In particular, poverty and distributional impacts are separately measured for the agricultural and nonagricultural groups and the movers and stayers.19 Figure 9.3 shows the growth incidence curves for the poorest 30 percent of all households under the three scenarios. The curve for the Doha scenario lies slightly above the BaU curve. The full-liberalization reform also shifts the whole curve upward, but this shift is larger than that of the Doha case, and it seems to favor the poorest among the poor; in other words, full liberalization appears to induce an Figure 9.3. Growth Incidence Curves for the BaU and Trade Scenarios, Poorest 30 Percent of All Households Growth incidence curve (%) 2.2 growth 2 income 1.8 capita 6 per 1. Annual 1.4 0 10 20 30 Cumulative population ranked by per capita income BaU Doha Full Source: Authors' calculations. 276 Poverty and the WTO: Impacts of the Doha Development Agenda additional pro-poor distributional shift, resulting from Brazil's own-liberalization in the full-liberalization package of reforms. Table 9.9a-b shows the results for agriculture and nonagriculture groups of households. Compared to the BaU scenario, inequality for all households falls as a result of decreased inequality among agricultural households and lower inequal- ity increase among nonagricultural households, although inequality between these two groups may have risen somewhat. Despite declining inequality and slightly higher per capita income growth, the rate of poverty reduction for agri- cultural households barely changes. This is due to the lower migration levels induced by the trade shocks (see table 9.10c below). Indeed, in the Doha scenario, the reduction in the population share of agricultural households is only very slightly below that achieved in the BaU scenario. More remarkable is the addi- tional poverty reduction for nonagricultural households, which can largely be explained by a decrease in inequality because per capita income growth is only marginally higher under trade reform. Given its larger price and quantities shocks, the full-liberalization scenario yields more significant poverty changes, as shown in table 9.9b. In contrast to the Doha scenario, agricultural households gain considerably from full liberalization, and their headcount index is reduced by almost 1.5 percentage points. These sec- tor-specific income gains more than compensate the further (albeit small) reduc- tion of agricultural out-migration. For nonagricultural households, the full-liberalization scenario improves the income distribution, and the Gini increases by only 72 percent of the increase recorded in the BaU scenario. Growth is only slightly higher for this group of households but, as shown above, minor distributional shifts accompanied by slightly higher growth can result in significant poverty reduction. Trade shocks simultaneously increase agricultural incomes and reduce inter- sectoral migration; how these two contrasting forces affect poverty outcome depends on the income levels (and therefore on the socioeconomic characteris- tics) of those who decide to stay instead of moving. Table 9.10 sheds some light on this issue. It shows the poverty levels and changes under the BaU and trade sce- narios for agricultural households according to their migration decision. Table 9.10a shows those who remained in agriculture, the "stayers." First, consider the BaU case. With those households that will not move identified, it is possible to cal- culate the headcount for this group in the initial year (2001): their poverty head- count is equal to 44.1 percent, more than 2 percentage points below the 46.2 per- cent level20 calculated for all 2001 agricultural households (that is, the combination of stayers and potential movers). This lower level of poverty implies that moving households are on average poorer than those who remain in agricul- ture. Accordingly, the changes in P0 are 12.1 instead of 13.7 percentage points. In Table 9.9a. Poverty and Inequality in the Doha Scenario, by Sector All households Nonagricultural households Agricultural households BaU BaU BaU Variable 2001 2001­15 change 2001 2001­15 change 2001 2001-15 change levels changes (%) levels changes (%) levels changes (%) Per capita income 314.9 1.5 101.5 351.9 1.3 102.1 148.3 2.4 101.3 Gini 58.6 -0.2 194.4 57.1 0.5 81.8 56.6 -0.8 111.5 Structural P0 23.6 -5.8 103.4 18.6 -3.3 106.5 46.2 -14.0 101.5 P1 9.6 -3.1 102.7 7.1 -1.6 104.6 21.0 -8.2 102.1 P2 5.3 -1.9 102.5 3.7 -0.9 104.3 12.3 -5.3 102.0 Change Population (%) 100.0 n.a. n.a. 81.8 3.2 98.3 18.2 -3.2 98.3 Contribution to P0 n.a. n.a. n.a. 64.4 8.6 96.0 35.6 -8.6 96.0 and Source: Authors' calculations. Poverty Note: n.a. = not applicable. Reduction in Brazil 277 278 Table 9.9b. Poverty and Inequality in the Full Liberalization, by Sector All households Nonagricultural households Agricultural households Poverty 2001 2001­15 BaU 2001 2001­15 BaU 2001 2001­15 BaU Variable levels changes change levels changes change levels changes change (%) (%) (%) and Per capita the income 314.9 1.6 106.4 351.9 1.3 106.8 148.3 2.6 109.8 Gini 58.6 -0.3 312.2 57.1 0.5 72.0 56.6 -0.9 117.0 WTO: P0 23.6 -6.1 109.2 18.6 -3.6 116.3 46.2 -14.9 108.0 P1 9.6 -3.2 108.2 7.1 -1.8 113.7 21.0 -8.6 107.4 Impacts P2 5.3 -1.9 107.8 3.7 -1.0 113.0 12.3 -5.6 107.2 Population (%) 100.0 n.a. n.a. 81.8 3.1 93.0 18.2 -3.1 98.0 of Contribution the to P0 n.a. n.a. n.a. 64.4 8.4 96.0 35.6 -7.6 96.0 Doha Source: Authors' calculations. Note: n.a. = not applicable. Development Agenda Structural Change and Poverty Reduction in Brazil 279 Table 9.10a. Poverty Impact of Trade, by Migration Choices Households remaining in agriculture Variable 2001 BaU 2001­15 Full % of level of baseline Doha % BaU variable change in of BaU change change variable P0 44.1 -11.7 101.7 109.5 P1 20.0 -7.0 102.4 108.5 P2 11.7 -4.6 102.3 108.2 Population (%) n.a. 14.9 100.4 101.5 Source: Authors' calculations. Note: n.a. = not applicable. Table 9.10b. Poverty Impact of Trade, Nonagricultural Stayers Nonagricultural households, before and after Variable 2001 BaU 2001­15 Doha % Full % of level of baseline of BaU BaU variable change in change change variable P0 18.6 -3.8 104.0 110.7 P1 7.1 -1.8 103.3 109.8 P2 3.7 -1.0 103.2 109.5 Population (%) 82.4 n.a. n.a. n.a. Source: Authors' calculations. Note: n.a. = not applicable. Table 9.10c. Poverty Impact of Trade, Sectoral Movers Agricultural households that have become nonagricultural Variable 2001 BaU 2001­15 Doha % Full % of level of baseline of BaU BaU variable change in change change variable P0 56.6 -22.4 105.1 108.2 P1 26.0 -14.0 102.0 105.4 P2 15.2 -9.4 101.7 105.1 Population (%) n.a. 3.1 98.0 92.5 Source: Authors' calculations. Note: n.a. = not applicable. 280 Poverty and the WTO: Impacts of the Doha Development Agenda 2015, about 15 percent of the population still resides in agricultural households under the BaU scenario.21 The agricultural expansion after trade liberalization has only a minor effect on agricultural employment and not nearly enough to offset the reduction in agricultural employment under the BaU scenario. Accordingly, the change in the share of agricultural households due to trade liberalization is only minor, particularly for the Doha scenario. Yet, when translated into actual migrating individuals, this small share change means that almost 400,000 individ- uals (those who would have become members of nonagricultural households in the BaU scenario) remain in agricultural households under the full-liberalization scenario. Although these "potential mover households" are on average poorer than the typical "stayer household," as illustrated below, poverty among agricul- tural households decreases compared to the BaU scenario. Hence, it can be inferred that the relatively poor stayers gain under both trade scenarios, although this gain is very small for the Doha scenario. As indirectly inferred by the analysis of the stayers, the group of movers is expected to experience the largest welfare gains. As shown in table 9.10c, under the BaU scenario, agricultural households who become nonagricultural house- holds record a 22.4 percentage points reduction in their headcount index, down from a considerably higher initial level of 53.4 percent. This is a critical insight uniquely available through the use of the MS approach. The predicted additional poverty reduction for this group of mover households under the trade scenarios is modest and attributable to the income increases trade reforms induce in the nonagriculture sectors as well as due to the fact that the households that still move out of agriculture under the trade scenarios are actually poorer, on average. One final category needs to be examined: the nonagricultural stayers. This is a large group, representing 80 percent of the population; however, given the negligi- ble migration out of the nonagricultural sector observed in the data, this group is explicitly excluded from the migration choice. For these households, full liberal- ization brings about an additional reduction in the poverty headcount of 0.4 per- centage point,22 and the Doha scenario, through its favorable impact on nonagri- cultural unskilled wages, also makes a small but positive contribution. Conclusions The analysis in this chapter suggests that the economic effects of the Doha Round are rather limited for Brazil, in part as a result of the lack of tariff cuts in Brazil itself. Yet, through a slight improvement in the urban income distribution, the Doha scenario has some positive effect on poverty. In contrast, by adding domes- tic trade reforms and deepening reforms elsewhere, the full-liberalization scenario Structural Change and Poverty Reduction in Brazil 281 implies substantial welfare gains that are concentrated among some of the poorest groups in the country, particularly those in agriculture. Consequently, the rural poor in Brazil benefit more than the average. This result is driven by the export boom in agriculture and agricultural processing industries, growing labor demand, and associated higher wages. After full liberalization, a smaller number of workers remain in agriculture compared to the BaU scenario. Given that inter- sectoral migration substantially improves the income situation of many house- holds under the baseline, one might conjecture that full liberalization would weaken poverty reduction. However, this is not the case, because the gain in agri- cultural incomes more than compensates for the reduced benefits from lower migration flows. The positive impact of full liberalization is not limited to rural areas and nona- gricultural activities. The urban poor gain from higher unskilled wages, even in nonagricultural sectors. This is reflected in the pro-poor shift in the urban income distribution. In addition, the urban poor benefit indirectly from the gains in agri- culture because the pressure on nonagricultural unskilled workers is relieved somewhat. Trade reform, and particularly domestic trade reforms, may particu- larly help the poor Brazilian farmers, but only broad-based high growth will erad- icate urban poverty. An important limitation of the analysis in this chapter is that the potential interactions between trade liberalization and the rate of productivity growth in Brazil are not considered. The latter is assumed to be exogenous and fixed at its BaU level for all scenarios. This growth rate fuels the strong poverty reduction in the baseline scenario. Given the growing evidence of a beneficial impact of trade liberalization on productivity (see Winters, McCulloch, and McKay [2004]; see also chapter 17 in this book), it must be noted that this chapter's assessment of the potential for additional poverty reduction in the wake of a Doha Round is likely to be on the conservative side. Nonetheless, significant reductions in poverty beyond that achieved in the BaU scenario will likely require additional, complementary reforms. Based on the mover-stayer analysis in this chapter, policies aimed at facil- itating the movement of the poorest rural households out of agriculture could be particularly beneficial. Notes 1. See Bourguignon (2003), Ravallion (2001), Ravallion and Datt (1999), and Kappel, Lay, and Steiner (2005). 2. Poverty is measured by the headcount ratio. The poverty figures in this paragraph are taken from Ferreira, Lanjouw, and Neri (2001). 3. The figures on agricultural employment are own calculations based on the PNAD 1997 and the PNAD 2001. 282 Poverty and the WTO: Impacts of the Doha Development Agenda 4. The PNAD is a regularly conducted representative household survey. The sample had a size of about 380,000 individuals in 2001. 5. See Helfand and Rezende (2004) and Dias and Amaral (2002) for details. 6. Data from employment histories in the PNAD reveal that in both 1997 and 2001, about 6 per- cent of those who became unemployed in the last year were employed in agricultural sectors before. Taking into account the fact that approximately 20 percent of the workforce are employed in agricul- ture, this figure is rather low and may be taken as a sign that the rise in urban unemployment is not causally linked to the decline in agricultural employment. 7. An even more detailed documentation for the macro model is found in Bussolo, Lay and van der Mensbrugghe (2005). 8. Aggregate investment is set equal to aggregate savings, and aggregate government expenditures are exogenously fixed. 9. See Armington (1969) for details. 10. This chapter follows the Horridge and Zhai approach to shifting export demand. For more details, see the appendix to chapter 3. 11. Regression results are reported in Bussolo, Lay and van der Mensbrugghe (2005). 12. Additional support for this treatment of productivity comes from a recent panel study of sec- toral productivity growth in OECD and developing countries (Martin and Mitra 1999). In this study, depending on the estimation method, the average growth rate for total factor productivity in agricul- ture in middle-income developing countries ranges from 1.78 percent to 2.91 percent per year. 13. A complete description of the MS model, including the estimation of the wage-profit equations and the migration choice equations, as well as the reweighing procedure and the other steps, is found in Bussolo, Lay and van der Mensbrugghe (2005). 14. The income gap ratio (average income shortfall of the poor divided by the poverty line) can be calculated as P1/P0. This ratio is 0.4 for all households in this case--that is, the perfectly targeted cash transfer needed to lift every poor person out of poverty is 40 percent of the poverty line. Thus, 0.4 times the percentage point change in P0 (here 2.4) provides a percentage point change benchmark for evaluating the change in P1 as an indicator of the depth of poverty, because this would be the change in P1 that would be observed had the average income of the poor stayed constant while the headcount declined. 15. The household category (agricultural or nonagricultural household) is the category the house- hold belonged to in the base year 2001. 16. To mimic the global model results for increased demand for Brazilian exports and changes in international prices, a downward-sloping export demand function is introduced, as discussed in the appendix to chapter 3. During a shock, for obvious reasons, both prices and quantities cannot be tar- geted, and the shock is implemented by modifying both the international price index (the price shock) and the intercept (the quantity shock). The Brazil (single-country) model will then endogenously determine the quantity supplied. 17. Because of the closure rule of the external account (the fixing of foreign savings) and the full employment assumption, the slightly lower expansion of the volumes of exports with respect to import volumes is compensated with a real exchange rate appreciation that originates from rising domestic resource costs. 18. When they simulate analogous trade reforms, Harrison and others (2003) generate comparable sectoral reallocation results, as well as factor market outcomes similar to those shown in table 9.8. This consistency should not be surprising, given that the model in this chapter does not significantly differ from theirs and the initial sectoral bias in the Brazilian tariff structure as well as intersectoral factor intensities are very close in the two approaches. It should be stressed that in the model in this chapter, trade opening produces only allocative effi- ciency gains and not other, potentially stronger dynamic productivity gains, which are explored in chapter 17 in this volume. Structural Change and Poverty Reduction in Brazil 283 19. In Bussolo, Lay and van der Mensbrugghe (2005), detailed poverty impacts of trade reform are analyzed for a number of additional groupings, for example by educational attainments, occupational status, or region of residence. 20. Shown in Table 9.3. 21. The initial poverty levels among those who stay in agriculture under the trade scenarios are almost identical to the initial levels among the BaU stayers, so they are not reported here. The same holds for the movers, for whom results are reported later. 22. The 0.4 percentage point is calculated using table 9.9 figures: 0.4 = -3.8 - (-3.8/100 x 110.7/100) x 100. References Armington, P. S. 1969. "A Theory of Demand for Products Distinguished by Place of Production." IMF Staff Papers 16: 159­78. Barrett, C. B. 2004. Rural Poverty Dynamics: Development Policy Implications. Paper prepared for invited presentation at the 25th International Conference of Agricultural Economists, Durban, South Africa, August 17­23, 2003. Bourguignon, F. 2003. "The Growth Elasticity of Poverty Reduction: Explaining Heterogeneity across Countries and Time Periods," in Inequality and Growth, ed. T. S. Eicher and S. J. Turnovsky. Cambridge, MA: MIT Press. Bourguignon, F., F. H. G. Ferreira, and N. Lustig. 2005. "A Synthesis of the Results." In The Microeconomics of Income Distribution Dynamics in East Asia and Latin America, ed. F. Bourguignon, F. H. G. Ferreira, and N. Lustig, 357­406. New York: Oxford University Press; Washington, DC: World Bank. Bussolo, M., J. Lay, and D. van der Mensbrugghe. 2005. "Structural Change and Poverty Reduction in Brazil: The Impact of the Doha Round." Policy Research Working Paper, World Bank, Washington, DC. Dias, G. L. S., and C. M. Amaral. 2002. "Structural Change in Brazilian Agriculture, 1980­98." In Brazil in the 1990s--an Economy in Transition, ed. R. Baumann, 204­232. New York: Palgrave. Ferreira, H. G., P. Lanjouw, and M. Neri. 2001."A Robust Poverty Profile for Brazil Using Multiple Data Sources." Paper presented at the Latin American and Caribbean Economic Association meeting, Montevideo, October 18­20 2001. Ferreira, F. H. G., and R. Paes de Barros. 2005. "The Slippery Slope: Explaining the Increase in Extreme Poverty in Urban Brazil, 1976­96." In The Microeconomics of Income Distribution Dynamics in East Asia and Latin America, ed. F. Bourguignon, F. H. G. Ferreira, and N. Lustig, 83­124. New York: Oxford University Press; Washington, DC: World Bank. Harrison, G. W., T. F. Rutherford, D. Tarr, and A. Gurgel. 2003. "Regional, Multilateral, and Unilateral Trade Policies of MERCOSUR for Growth and Poverty Reduction in Brazil." Policy Research Working Paper 3051, World Bank, Washington, DC. Helfand, S. M., and G. C. de Rezende. 2004. "The Impact of Sector-Specific and Economy-Wide Reforms on the Agricultural Sector in Brazil: 1980­98." Contemporary Economic Policy 22 (2): 19­212. IBGE. 1997. Anuário estatístico do Brasil 57. São Paulo. Kappel, R., J. Lay, and S. Steiner. 2005. "Uganda: No More Pro-Poor Growth?" Development Policy Review, 23 (1): 27­53. Martin, W., and D. Mitra. 1999. "Productivity Growth and Convergence in Agriculture and Manufacturing." Policy Research Working Paper 2171, World Bank, Washington, DC. Paes de Barros, R. 2004. Pobreza rural e trabalho agrícola no Brasil ao longo da década de noventa. Unpublished paper. 284 Poverty and the WTO: Impacts of the Doha Development Agenda Ravallion, M., and S. Chen. 2003. "Measuring Pro-Poor Growth." Economics Letters 78 (1): 93­99. Ravallion, M. 2001. "Growth, Inequality, and Poverty: Looking beyond Averages." World Development 29 (11): 1803­15. Ravallion, M., and G. Datt. 1999. "When is Growth Pro-Poor? Evidence from the Diverse Experiences of India's States." Policy Research Working Paper 2263, World Bank, Washington, DC. Verner, D. 2004. "Making the Poor Count Takes More than Counting the Poor--a Quick Poverty Assessment of the State of Bahia, Brazil." Policy Research Working Paper 3216, World Bank, Washington, DC. Winters, L. A., N. McCulloch, and A. McKay. 2004. "Trade Liberalization and Poverty: The Evidence So Far." Journal of Economic Literature 152: 72­115. 10 Impacts of the DDA on China: The Role of Labor Markets and Complementary Education Reforms Fan Zhai and Thomas W. Hertel Summary This chapter offers an assessment of the implications of multilateral trade reforms for poverty in China by combining global results from chapter 3 with a national computable general equilibrium (CGE) model that features disaggregated house- holds in both the rural and urban sectors. Using the World Bank's $2/day poverty line, the findings indicate that multilateral trade reforms do in fact reduce poverty in China. The biggest reductions occur in the rural areas--largely as a result of higher prices for farm products. Urban poverty falls in two of the three household groups considered in this analysis, since the increased demand for China's products in world markets boosts factor earnings sufficiently to offset the impact of higher food prices. For the remaining group--which is heavily dependent on transfer payments--it is assumed that indexation of these payments will largely offset the adverse conse- quences of higher prices. However, a decline in other income sources is sufficient to cause a small increase in poverty, and this increase is large enough to boost the overall urban poverty headcount very modestly. But since the urban poor only represent 5 percent of the total poor in China and the national poverty head- count falls. 285 286 Poverty and the WTO: Impacts of the Doha Development Agenda This chapter also explores the implications of complementary reforms in China--in particular the impact of increased investments in rural education. These are aimed at increasing labor productivity, as well as enhancing the mobil- ity of the rural labor force, thereby putting these workers in a better position to benefit from trade reforms. The specific scenario considered is one in which rural enrollment rises by 16 percent. The analysis takes account of the cost of funding these additional students, as well as the reduction in the workforce that results from having more pupils in school. Nevertheless, these reforms generate very sub- stantial gains for China's economy. They also serve to boost rural incomes and reduce the incidence of rural poverty. Indeed, when combined with global trade liberalization, poverty in China is estimated to drop by about 55 million people. Introduction With its rapid economic growth and integration into the world economy over the last two decades, China has emerged as a global economic force. Now it is the fourth largest trader and the largest FDI recipient in the world. China's foreign trade and investment are expected to be further boosted by WTO accession, including the recent elimination of textile and apparel quotas, as well as prospec- tive multilateral trade liberalization in the context of the Doha Development Agenda (DDA). However, against the background of rapid economic growth and openness, the income distribution has deteriorated sharply in China. The ratio of urban to rural incomes increased from 2.2 in 1990 to 3.1 in 2002, which is extremely high by international standards. In the meantime, income inequality within rural areas, as measured by the Gini coefficient, rose from 0.31 in 1990 to 0.36 in 2001, and it increased from 0.23 to 0.32 in urban areas over the same period (Li and Yue 2004). This widening income disparity is the result of profound structural changes in the Chinese economy. The experience of the last decade suggests that trade liberal- ization might contribute to the increased inequality (Kanbur and Zhang 2001). China's WTO accession has further heightened the concern about the increasing rural-urban disparity, because most analyses suggest that accession will exacerbate inequality by lowering barriers to grain imports and increasing opportunities for manufacturing exports as well as foreign investment in urban-based services (Bhat- tasali, Li, and Martin, 2004). Hertel, Zhai, and Wang (2004) find that the poorest rural, agriculture-specialized households that have limited labor mobility out of farming might lose from WTO accession. How will these outcomes be affected by a potential Doha reform package? Do complementary reforms exist that might lessen adverse impacts on the poor? This chapter focuses on the potential for rural educa- tion reforms to enhance the poverty outcomes under a potential DDA. Impacts of the DDA on China 287 It has been widely recognized that education plays a critical role in creating human capital and subsequently prompting economic development and reducing poverty. However, investment in education is often inadequate relative to other investments, because of the presence of associated externalities, labor market dis- tortions that depress private returns to education, and the generally low level of public support for education (Heckman 2002). Moreover, disparities in funding for education have resulted in nonuniform access to education across regions and between urban and rural areas. In China, education spending has been dispropor- tionately directed toward urban areas at the cost of rural areas. In 2001, per capita spending on compulsory education in urban areas was a 16 percent higher than in rural areas (Wang 2003). This chapter uses a household-disaggregated applied general equilibrium (AGE) model to assess the differential household effects of multilateral trade liber- alization under the DDA of the WTO, as well as the additional impacts of increas- ing spending on rural education. The framework used here explicitly models the linkages between education and labor productivity improvement as well as off- farm labor mobility, which is a critical vehicle for poverty reduction in rural China. This chapter is organized as follows: section 1 describes the specification of the CGE model used in this study. Section 2 elaborates on how the educational expen- ditures and output are modeled in terms of the supply of labor force by skill and associated efficiency. Section 3 assesses the impact of Doha Round trade liberal- ization, as well as increasing rural educational expenditures, on rural-urban inequality. The final section offers conclusions. The CGE Model The CGE model of China used in this study is the latest in a long line of model developments based at the Development Research Center of the State Council in Beijing. The model has its intellectual roots in the group of single-country, AGE models used over the past two decades to analyze the impact of trade policy reform (Dervis, de Melo, and Robinson 1982; Shoven and Whalley 1992). Here, the focus is on the main features of the model. Household Behavior To come to grips with the poverty question, it is critical to disaggregate house- holds to the maximum extent possible, subject to the limitations posed by survey sampling, computational constraints, and human capacity for analysis. Following previous work (Hertel, Zhai, and Wang 2004; Hertel and Zhai 2004), rural and urban households are disaggregated into 40 rural and 60 urban representative 288 Poverty and the WTO: Impacts of the Doha Development Agenda households according to their primary source of income and relative income level. Recent analysis of trade and poverty by Hertel and others (2004) suggests the merit of distinguishing those households that are specialized (95 percent or more of their income from one source) in transfer payments, labor wages and salaries, or self-employment income. According to the available survey data, the rural households are stratified by agriculture-specialized and diversified (all other) and the urban households by three strata: transfer-specialized, labor-spe- cialized, and diversified. Within each stratum, households are ordered from poor- est to richest, based on per capita income, and then grouped into 20 vingtiles, each containing 5 percent of the stratum population. Household income derives from labor income, profits from family-owned agriculture and nonagriculture enterprises, property income, and transfers. Households consume goods and services according to a preference structure determined by the extended linear expenditure system. Through specification of a subsistence quantity of each good or service, this expenditure function generates nonhomothetic demands whereby the larger the relative importance of subsis- tence consumption (for example, it would be high for rice and low for automo- biles), the more income-inelastic the household's demand for that good. The other important dimension of household behavior is the supply of labor to off-farm activities. In China, the off-farm labor supply decision is complicated by institutional factors that have been built into the system keep the agricultural popu- lation in place (Zhao 1999b). During earlier years, the Chinese government sought to make it costly for individuals to leave the rural areas by tying incomes to daily participation in collective work. More recently, the absence of well-defined land tenure has raised the opportunity cost of leaving the farm (Yang 1997). Households that cease to farm the land may lose the rights to it, so they have a strong incentive to continue some level of agricultural activity, even when profitability is quite low (Zhao 1999a). With only modest growth in rural, nonfarm activities, this seriously limits the ability of households to obtain off-farm work (Zhao 1999b).1 A constant elasticity of transformation (CET) function is used to model the off- farm labor supply of rural households. The labor allocation between farm and off- farm jobs is determined by the ratio of the shadow value of labor in agriculture, rel- ative to the off-farm wage rate, and the elasticity of transformation,2 which reflects imperfect labor mobility. There are many reasons for this imperfect mobility of labor, including education, experience, and simple geography that can serve to iso- late farm households from the nonfarm labor market. Owing to the absence of an effectively functioning land market, the shadow value of labor in agriculture in this function takes into account the potential impact that reducing agricultural employment will have on the household's claim to farmland. This incremental fac- Impacts of the DDA on China 289 tor is calculated as the marginal value product of land, multiplied by the rate at which decreased on-farm labor reduces the household's land endowment.3 Rural-Urban Migration Despite the large income and poverty differential between rural and urban house- holds, permanent migration in China has been limited. This is due to a combina- tion of both direct and indirect measures. First, households must have an appropri- ate registration (hukou) to legally reside in an urban area. Without this registration, access to many of the urban amenities, including housing and education, is limited and quite expensive. In light of these barriers to moving the entire household to an urban area, rural-urban migration is largely a transitory phenomenon. For the modeling exercise, it is important to obtain an estimate of the wage gap motivating the temporary migration of workers from the rural to the urban sector in China. Zhao (1999a) documents an average annual wage gap between rural and urban work of 2,387.6 yuan for unskilled rural workers of comparable background and ability in Sichuan province in 1995. The majority of the wage gap is due to social costs associated with migration, including the disutility of being away from family, poor quality of housing, limited social services for migrants, and the general uncertainty associated with being an unregistered worker in an urban area (Zhao 1999a, 1999b). Although these transactions costs are unobservable, they clearly represent a very significant burden on the migrants and their families. If there were no barriers to the movement of labor between rural and urban areas, real wages would be expected to be equalized for an individual worker with given characteristics. Shi, Sicular, and Zhao (2002) explore the question of rural- urban inequality in greater detail for nine provinces using the China Health and Nutrition Survey (CHNS). The authors conclude that the apparent labor market distortion is about 42 percent of the rural-urban labor income differential and 48 percent of the hourly earnings differential.4 When applied to the average wage dif- ferential, this amounts to an ad valorem rate of apparent transaction "tax" on rural wages of 81 percent.5 These transaction costs are modeled as real costs assumed by the temporary migrants. Of course, these migrants are heterogeneous, and the extent of the bur- den varies widely. Those individuals who are single and live close to the urban area in which they are working are likely to experience minor inconvenience as a result of this temporary migration. They can be expected to be the first to migrate (all things being equal) in response to higher urban wages. However, some migrants have large families and come from a great distance. Their urban living conditions are often very poor, and it is not uncommon for them to be robbed on the train when they are returning home after their work. For such individuals, the decision 290 Poverty and the WTO: Impacts of the Doha Development Agenda to migrate temporarily is likely to be a marginal one--and one that they may not choose to repeat. With this heterogeneous population in mind, a transactions cost function is postulated. This function is increasing in the proportion of the rural population engaged in temporary work. It has a simple, constant-elasticity func- tional form, which begins at the origin and reaches the observed wage gap (adjusted for transport and living costs) at the current level of temporary migra- tion (about 70 million workers). Further increases in temporary migration are assumed to have only a modest impact on these transaction costs.6 Production, Exports, and Imports Since the 1990s, processing exports have grown rapidly as a result of their prefer- ential treatment, which includes duty-free imports. This sector now accounts for more than half of China's total exports. This is explicitly captured in the CGE model by incorporating two separate foreign-trading regimes. One is the export processing regime, which receives duty-free imports and is therefore extremely open, with considerable foreign investment. Under this regime, firms process and assemble the imported goods, turning them into finished goods for export; these imported intermediate goods are exempted from tariffs and VATs. Therefore, export-processing firms are more intensive in their use of imported intermediate inputs, and all of their output is exported. The other sector is the ordinary trade regime, which is carried out under traditional taxes and regulations. The firms under the ordinary trade regime sell products on the domestic market or export to the rest of the world, according to a CET function. Therefore, the ordinary exports are treated as differentiated products from those sold on the domestic market. It is also assumed that the buyers of the rest of the world choose a mix of the ordinary exports and the processing exports to minimize their costs. There are two types of imports in the model. The imports of duty-free process- ing goods are used by export-processing firms as their intermediate inputs. Ordi- nary imports are modeled using the Armington assumption, that is, they are assumed to be differentiated from Chinese products produced by ordinary firms. The small country assumption is assumed for imports, thus world import prices are exogenous in terms of foreign currency. Exports are demanded according to constant-elasticity demand curves. Therefore, the terms of trade (TOT) for China are endogenous in the simulation in this chapter. The values of export demand and Armington elasticities, which are reported in table 10.1, are based on the esti- mation by Hertel and others (2003). Table 10.1 also lists a variety of useful infor- mation on production and trade patterns of the Chinese economy, based on the 1997 Chinese social accounting matrix (SAM). Table 10.1. Elasticity Parameters for Trade and Sectoral Structure of Production and Trade Trade elasticity Sectoral structure Output Value added Capital- Exports, Imports, Export share share of labor outputs domestic Sector Armington CET demand (%) output (%) ratio (%) use (%) Rice 5.1 3.6 10.1 1.2 63.0 0.1 0.9 0.9 Wheat 4.5 3.6 8.9 0.7 62.9 0.1 0.0 7.9 Corn 1.3 3.6 2.6 0.4 63.0 0.1 7.9 1.1 Cotton 2.5 3.6 5.0 0.3 62.7 0.1 0.0 10.7 Other nongrain crops 2.5 3.6 4.9 3.8 62.9 0.1 2.3 0.8 Forestry 2.5 3.6 5.0 0.4 70.6 0.1 3.6 6.7 Wool 6.5 3.6 12.9 0.0 47.3 0.1 8.9 55.5 Other livestock 1.5 3.6 3.1 3.8 48.2 0.1 0.9 0.1 Fishing 1.3 3.6 2.5 1.1 59.0 0.2 1.2 0.2 Impacts Other agriculture 2.5 3.6 5.0 0.6 54.9 0.2 1.2 0.0 Coal mining 3.1 4.6 6.1 1.1 49.3 0.4 3.1 0.4 of Crude oil and natural the gas 7.4 4.6 14.9 0.8 64.6 3.8 14.4 23.9 Ferrous ore mining 3.0 4.6 5.9 0.2 27.2 0.7 0.0 29.6 DDA Nonferrous ore mining 4.2 4.6 8.4 0.4 31.7 0.7 1.1 6.5 on Other mining 0.9 4.6 1.8 0.9 36.5 0.5 4.3 5.5 China Vegetable oil 3.3 4.5 6.6 0.6 15.4 2.4 4.3 12.1 Grain mill and forage 2.6 4.5 5.2 1.6 15.7 2.4 1.6 4.0 Sugar 2.7 4.7 5.4 0.2 10.1 0.3 3.3 6.0 Processed food 2.8 4.5 5.6 2.6 21.9 1.1 9.7 2.5 291 292 Table 10.1. (Continued) Trade elasticity Sectoral structure Poverty Output Value added Capital- Exports, Imports, Export share share of labor outputs domestic Sector Armington CET demand (%) output (%) ratio (%) use (%) and Beverage 1.2 4.7 2.3 1.3 19.7 1.1 3.1 0.5 the Tobacco 1.2 4.7 2.3 0.7 12.1 1.7 3.2 1.3 WTO: Textile 3.8 5.4 7.5 4.6 22.8 1.0 18.4 10.1 Apparel 3.7 5.8 7.4 1.9 31.8 0.6 37.0 3.6 Impacts Leather 4.1 4.6 8.1 1.1 18.9 0.5 32.6 13.2 Sawmills and furniture 3.4 4.6 6.8 1.1 23.2 0.7 13.1 5.6 Paper and printing 3.0 4.6 5.9 1.7 27.4 0.6 2.0 9.9 of Social articles 3.8 5.6 7.5 1.1 25.9 0.9 40.0 8.9 the Petroleum refining 2.1 3.8 4.2 1.6 12.5 1.7 5.7 11.6 Doha Chemicals 3.3 3.8 6.6 4.1 20.1 1.0 8.3 16.5 Medicine 3.3 3.8 6.6 0.9 28.1 1.8 5.9 1.7 Development Chemical fibers 3.3 3.8 6.6 0.6 17.1 1.4 6.6 18.2 Rubber and plastics 3.3 3.8 6.6 2.1 19.2 0.9 15.7 6.9 Building materials 2.9 3.8 5.8 4.4 26.2 0.7 3.4 1.2 Primary iron and steel 3.0 4.6 5.9 2.7 16.3 0.6 5.5 8.6 Agenda Nonferrous metals 4.2 4.6 8.4 1.2 13.4 0.7 8.0 13.0 Metal products 3.8 4.6 7.5 2.5 19.3 0.7 13.1 7.0 Machinery 2.8 4.6 5.6 2.5 31.0 1.1 5.9 13.2 Special equipment 2.8 4.6 5.6 1.8 24.3 0.8 5.6 23.3 Automobile 2.8 4.6 5.6 1.7 21.2 1.0 1.9 4.1 Table 10.1. (Continued) Trade elasticity Sectoral structure Output Value added Capital- Exports, Imports, Export share share of labor outputs domestic Sector Armington CET demand (%) output (%) ratio (%) use (%) Other transport equipment 4.3 4.6 8.6 1.3 24.2 0.8 9.7 12.1 Electric machinery 4.1 4.6 8.1 2.8 18.1 0.9 15.9 9.5 Electronics 4.4 4.6 8.8 2.5 21.2 1.1 36.3 34.1 Instruments 4.1 4.6 8.1 0.4 27.3 0.9 49.5 43.2 Other manufacturing 3.8 3.8 7.5 0.8 53.3 1.0 8.1 3.1 Utilities 2.8 3.8 5.6 2.2 36.1 2.0 0.9 0.0 Construction 1.9 3.8 3.8 8.7 26.4 0.3 0.1 0.3 Transportation 1.9 2.8 3.8 2.5 50.4 1.0 9.4 1.8 Impacts Post and communications 1.9 2.8 3.8 1.0 53.5 3.5 5.7 1.3 Commerce 1.9 2.8 3.8 6.5 37.2 0.5 0.9 1.3 of Finance 1.9 2.8 3.8 1.8 37.9 1.1 0.5 1.2 the Social services 1.9 2.8 3.8 3.7 45.0 1.2 10.1 5.3 DDA Education and health 1.9 2.8 3.8 3.3 46.4 0.3 0.7 0.5 on Public administration 1.9 2.8 3.8 2.2 45.0 0.2 0.1 0.6 China Source: Hertel and others (2003); 1997 Chinese SAM. 293 294 Poverty and the WTO: Impacts of the Doha Development Agenda Production in each of the sectors of the economy is modeled using nested con- stant elasticity of substitution (CES) functions, and constant returns to scale are assumed. In the top level of the nest, value added and a composite of intermediate inputs produce outputs. Then a further CES function disaggregates the value added into a capital-labor composite and agricultural land. The capital-labor com- posite is further split into the capital­skilled labor composite and the aggregated less skilled labor (which is composed of semiskilled labor and unskilled labor.) The values of substitution elasticities in production functions are listed in table 10.2. A low substitution elasticity of 0.3 between capital and skilled labor is assumed here to introduce capital-skill complementarity. The elasticity of substitution between semiskilled labor and unskilled labor is set to 1.5, based on estimates for the United States by Katz and Murphy (1992) and Heckman and Lochner (1998). Within each labor category (unskilled, semiskilled, and skilled), rural and urban workers are distinguished. These two categories of workers substitute imperfectly in production. This is an indirect means of building into the model a geographic flavor because some sectors will be located largely in urban areas, and others will be predominantly in rural areas. By limiting the substitutability of rural and urban labor in each sector, the model proxies the economic effect of geographically distributed activity. This is particularly important in China, where significant barriers to rural and urban mobility remain (see section 1.2 above). Ideally, the geographic distribution of industrial activity would be modeled, but unfortunately the data do not exist to support this split. All commodity and nonlabor factor markets are assumed to clear through prices. With the exception of the farm or nonfarm labor supply decision, labor is Table 10.2. Substitution Elasticities in Production Type of elasticity Agriculture Nonagriculture Value added-aggregated intermediate input 0.10 0.10 Capital-labor composite-land 0.30 n.a. Capital-skilled labor composite- aggregate less skilled labor Old vintage 0.45 0.65 New vintage 0.95 1.10 Capital-skilled labor 0.30 0.30 Unskilled labor-semiskilled labor 1.50 1.50 Urban labor-rural labor 2.20 2.20 Source: DRC CGE model. Impacts of the DDA on China 295 assumed to be perfectly mobile across sectors. Capital is assumed to be partially mobile, reflecting differences in the marketability of capital goods across sectors. Recursive-Dynamic and Steady-State, Comparative Static Closures The CGE model is benchmarked to China's 1997 SAM, and it incorporates dynamics in two alternative ways. The recursive-dynamic version of the model is used to update the SAM to 2005 and assess the impact of intervening events on China's patterns of trade, production, and consumption. In this version of the model, the classical savings-investment mechanism determines the capital stock in the medium and long term. Dynamics originate from the accumulation of pro- ductive factors and productivity changes. The steady-state, comparative static ver- sion is used to assess the impact of trade and educational reforms, starting from the 2005 base. Here, the longer-term accumulation effects are taken into account by introducing a different capital market closure following Harrison, Rutherford, and Tarr (1997) and Francois and McDonald (1996). The aggregate capital stock is allowed to adjust to its long-term equilibrium based on an exogenous capital rental rate (fixed at the benchmark level). The theoretical underpinnings of this closure are based on the concept of an invariant capital stock equilibrium as pro- posed by Hansen and Koopmans (1972). Modeling Education Education affects the economy in several important ways. It improves the skills of workers, thereby enhancing their productivity and garnering them a higher wage. In the context of rural China, education is also a key determinant of an individ- ual's potential suitability for off-farm work (Yang 2004; Zhang, Huang, and Rozelle 2002). Because the farm­off-farm labor market linkage has proven key to the transmission of trade reform benefits to the rural poor (Hertel, Zhai, and Wang 2004), this mechanism receives special attention here. This section describes the framework through which education expenditure affects the pro- duction of human capital and its distribution among different household groups, as well as the linkage between schooling attainment and off-farm mobility of labor in rural areas. The Role of Education in the Model In the CGE model, each household is endowed with 17 groups of workers, distin- guished by their total years of schooling, ranging from 0 to 16. Based on this level of educational attainment, the skill level of household members in the workforce 296 Poverty and the WTO: Impacts of the Doha Development Agenda can be inferred. Unskilled labor refers to workers with 0 to 6 years of educational attainment. Semiskilled workers have 7 to 12 years of educational attainment, and skilled workers have an educational attainment from 12 to 16 years. For each household, the labor endowment by skill is determined by its age-specific school participation rates and labor force participation rates. The age-specific school participation rates are determined by the education costs per pupil-year and the total educational expenditure. It is assumed that a change in total education expenditure induces a proportional change in school participation rates across all ages. The education expenditure comprises private expenditure and government expenditure, which are assumed to be fully fungible. Government and private education expenditures enter the budget constraints of government and households, respectively. However, private decisions are not the result of household investment choices. Rather, throughout the analysis, private and public expenditures are assumed to be made in equal proportions. In the model, education enhances labor productivity through two channels. First, more education improves the skill composition of the labor force, resulting in a greater supply of skilled labor and lesser supply of unskilled labor. Second, for each skill level, more education yields a higher level of average schooling attain- ment, thereby improving its average labor productivity. The second channel is captured by linear increasing functions between labor efficiency and the average years of schooling for the three types of labor. As noted above, the off-farm labor supply decision of rural households is mod- eled as a CET function of the ratio of the shadow value of labor in agriculture, rel- ative to the off-farm wage rate. It is assumed that the elasticity of transformation is a linear, increasing function of average years of schooling within the labor skill groups. This specification, which implies that rural households with higher edu- cational attainment respond more effectively to farm-nonfarm wage gaps, is based on recent empirical evidence (Zhang, Huang, and Rozelle 2002). Model Calibration and Choice of Parameters Detailed data on the population and labor force are necessary to calibrate the edu- cation block of the model. The age distribution of the rural and urban populations, respectively, is calculated according to their mortality tables from the 2000 national population census, under the assumption of a stationary population, that is, no change in the age structure. The age distribution of each of the 40 rural household groups is assumed to be the same, and similarly with the 60 urban household groups. Age-specific labor participation rates by urban and rural classifications are also obtained from the 2000 national population census. They were scaled up or Impacts of the DDA on China 297 down to the aggregated labor force participation rate of each representative house- hold to obtain the age-specific labor participation rates by household. National average school participation rates by age are calculated from official enrollment and dropout rates for primary school, middle school, high school, and university or college. Then the age-specific school participation rates of each household are estimated by solving a quadratic program, which minimizes the difference between the school participation rates of each household and the national average school participation rates, subject to the constraints implied by the base year skill composition of each household's labor endowment. The productivity increments associated with educational attainment enhance- ment are derived from the study by Shi, Sicular, and Zhao (2002). These authors estimate wage (or shadow wage) equations for agricultural and rural, nonfarm workers. These equations include educational attainment as an explanatory vari- able. Their estimates suggest that additional education has the greatest impact on rural nonfarm wages, with one additional year of schooling boosting hourly wages (and presumably productivity) by 15 percent. Additional schooling also has an impact on agricultural productivity, with an additional year of schooling boosting total factor productivity on the farm by 2 percent. When adjusted for the share of labor in agricultural output, this is equivalent to a 2.5 percent increase in labor productivity. Specification of the values of the off-farm labor supply elastic- ity draws on the econometric work of Sicular and Zhao (2004) and Zhang, Huang, and Rozelle (2002). Sicular and Zhao report results from a household labor supply model estimated using labor survey data from the 1997 CHNS dataset for nine central provinces. From their labor supply equations for self- employed agricultural labor and self-employed nonagricultural labor, it is possi- ble to calculate elasticities of labor transfer from farm to nonfarm activities. They report a variety of elasticities in their paper.7 This chapter adopts their estimate of 2.67 for use in this work as the overall farm­off-farm transformation elasticity for the total rural labor force. To obtain separate estimates of the farm­off-farm transformation elasticity for three skill levels of labor, the rate at which increased schooling attainment enhances the transformation elasticity is used, based on the study by Zhang, Huang, and Rozelle (2002). These authors explore the labor supply behavior of a panel of 310 individuals in 109 families observed in four villages of Jiangsu province in 1988, 1992, and 1996. They find that for every additional year of edu- cation, farmers had a 14 percent greater chance of finding an off-farm job in 1996, all things being equal. Using the base year ratio of the shadow value of labor in agriculture relative to the off-farm wage rate, as well as the total farm and off-farm labor supplies, this increasing opportunity for access to off-farm jobs associated 298 Poverty and the WTO: Impacts of the Doha Development Agenda with higher educational attainment translates into an increment of 0.58 in trans- formation elasticity for each additional year of schooling. This is used to calculate the farm­off-farm transformation elasticities by skill level according to the aver- age years of schooling for each skill level of the rural labor force. The resulting val- ues for this elasticity are 0.68 for unskilled labor, 4.01 for semiskilled labor, and 7.49 for skilled labor. Simulations Simulation Design The baseline scenario from 1998­2005 is constructed by using the recursive- dynamic version of the model. The baseline scenario establishes a plausible growth trajectory for the Chinese economy, which takes into account events such as China's WTO accession and its recent dramatic surge in exports, which have doubled in the last four years. The updated 2005 database is then used as the new benchmark equilibrium from which the comparative steady-state model is used to conduct the policy simulations. A series of trade reform scenarios is then consid- ered, followed by a final scenario that is designed to capture the added impact of rural education reform. The first trade reform scenario (ROW-Lib) considers the impacts of global trade liberalization excluding China. In particular, this entails elimination of all import tariffs in the rest of the world. In addition, agricultural export subsidies are elimi- nated, as are subsidies for domestic agricultural production in the OECD. The sec- ond scenario (Uni-Lib) focuses on China's unilateral liberalization. All import tar- iffs and nontariff barriers of China are eliminated in this scenario. The third scenario (Full-Lib) considers the impact of global free trade by combining the first and the second scenarios. The fourth scenario is the standard Doha scenario. To reflect the impacts of multilateral trade liberalization in the single-country model, external market impacts are incorporated into the Chinese CGE model through exogenous shifts in import prices and export demand schedules. The sizes of these exogenous trade shocks are obtained from global simulations. The tariff reduction in China is excluded in this Global Trade Analysis Project (GTAP) simulation but is included in the simulation of the single-country model. Table 10.3 reports the results provided by the global analysis, detailed in chap- ter 3, for the Full-Lib and Doha scenario impacts on China (with China's own reforms excluded). In the case of Full-Lib, there are some enormous percentage increases in China's export volumes generated by the elimination of very high rates of protection elsewhere in Asia. Rice, corn, grain milling, and other food Table 10.3. Inputs from the Global Modela (percent change) Export volume Export price Import price Tariff Sector Full-Lib Doha Full-Lib Doha Full-Lib Doha Full-Lib Doha Rice 7,574.5 454.4 12.2 2.1 12.2 1.3 -100 0 Wheat 43.2 7.6 5.4 1.4 5.4 1.3 -100 0 Corn 88.8 21.3 7.6 1.9 7.6 5.5 -100 -28 Cotton 71.9 28.5 5.7 1.7 5.7 5.0 -100 -20 Other crops 37.9 14.9 6.7 1.6 6.1 3.6 -100 -21 Forestry 0.9 -1.4 2.5 0.7 2.5 0.1 -100 -24 Wool 3.0 -6.3 7.7 2.2 7.7 2.4 -100 -2 Other livestock -3.1 -1.3 7.1 1.8 7.1 0.7 -100 -23 Fishing 9.8 2.8 5.0 1.2 5.0 0.2 -100 -29 Other agriculture 71.9 28.5 5.7 1.7 5.7 5.0 -100 -20 Impacts Coal mining 1.9 0.5 2.0 0.5 2.0 0.3 -100 -30 Crude oil and gas -1.9 -0.3 1.0 0.1 0.7 0.1 -100 n.a. Ferrous ore mining -2.4 -1.4 2.4 0.7 2.4 0.2 -100 -30 of Nonferrous ore mining -3.6 -1.8 2.3 0.7 2.3 0.2 -100 -31 the Other mining 0.8 0.2 1.9 0.5 1.9 0.5 -100 -29 DDA Vegetable oil -18.8 -9.2 5.7 1.7 5.7 0.4 -100 -22 Grain mill and forage 525.3 53.2 6.9 1.4 6.9 2.6 -100 0 on Sugar 83.4 25.0 5.0 1.4 5.0 1.5 -100 -1 China Processed food 36.6 -0.5 5.9 1.4 5.5 0.9 -100 -25 Beverage 19.9 2.2 4.5 1.2 4.5 0.2 -100 -22 Tobacco 19.9 2.2 4.5 1.2 4.5 0.2 -100 -22 299 Textile 13.7 5.5 3.3 1.0 3.3 0.2 -100 -33 300 Table 10.3. (Continued) Export volume Export price Import price Tariff Poverty Sector Full-Lib Doha Full-Lib Doha Full-Lib Doha Full-Lib Doha Apparel 17.0 9.0 2.9 0.9 2.9 0.4 -100 -32 and Leather 15.2 7.0 3.7 1.1 3.7 -0.3 -100 -32 Sawmills and furniture -4.7 -2.8 2.8 0.8 2.8 0.2 -100 -32 the Paper and printing -5.1 -3.1 3.0 0.9 3.0 0.3 -100 -33 WTO: Social articles -4.3 -2.0 2.9 0.9 2.9 0.0 -100 -28 Petroleum refining 3.8 -0.5 1.2 0.3 1.2 0.1 -100 -30 Impacts Chemicals 0.6 -0.7 2.5 0.7 2.5 0.1 -100 -28 Medicine 0.6 -0.7 2.5 0.7 2.5 0.1 -100 -28 of Chemical fibers 0.6 -0.7 2.5 0.7 2.5 0.1 -100 -28 the Rubber and plastics 0.6 -0.7 2.5 0.7 2.5 0.1 -100 -28 Doha Building materials 8.8 2.1 2.6 0.8 2.6 0.2 -100 -32 Primary iron and steel -2.4 -1.4 2.4 0.7 2.4 0.2 -100 -30 Development Nonferrous metals -3.6 -1.8 2.3 0.7 2.3 0.2 -100 -31 Metal products 4.4 0.9 2.5 0.7 2.5 0.1 -100 -33 Machinery -18.3 -5.6 2.3 0.7 2.3 0.0 -100 -33 Special equipment -18.3 -5.6 2.3 0.7 2.3 0.0 -100 -33 Automobile -18.3 -5.6 2.3 0.7 2.3 0.0 -100 -33 Agenda Other transport equipment 15.7 -1.3 2.4 0.7 2.4 0.0 -100 -32 Electric machinery -4.1 -2.3 2.4 0.7 2.4 0.0 -100 -33 Electronics -4.8 -2.8 1.8 0.5 1.8 0.1 -100 -33 Instruments -4.1 -2.3 2.4 0.7 2.4 0.0 -100 -33 Other manufacturing -4.3 -2.0 2.9 0.9 2.9 0.0 -100 -28 Table 10.3. (Continued) Export volume Export price Import price Tariff Sector Full-Lib Doha Full-Lib Doha Full-Lib Doha Full-Lib Doha Utilities -2.4 -1.3 2.4 0.7 2.4 0.0 0 0 Construction -6.3 -2.6 2.7 0.8 2.7 -0.1 0 0 Transportation -5.5 -2.3 3.0 0.9 2.9 0.2 0 0 Post and communications -5.4 -2.2 2.6 0.8 2.6 0.0 0 0 Commerce -5.6 -2.5 3.1 0.9 3.1 0.3 0 0 Finance -5.6 -2.3 2.7 0.8 2.7 -0.1 0 0 Social services -6.2 -2.4 2.8 0.8 2.8 -0.1 0 0 Education and health -6.5 -2.3 2.8 0.8 2.8 0.0 0 0 Public administration -6.5 -2.3 2.8 0.8 2.8 0.0 0 0 Impacts Source: GTAP simulations. Note: n.a. = not applicable. a. These results are obtained by solving the global model with China's shocks omitted. The export price and volume changes are used, along with the export demand of elasticity, to compute the shift in the export demand schedule. the DDA on China 301 302 Poverty and the WTO: Impacts of the Doha Development Agenda products all show very large proportionate increases. Of course, the associated volume changes are often quite modest, because China is not a large exporter of most of these products. Moreover, after the China CGE model is solved with implied shifts in export demands and import prices, the resulting equilibrium change in export volumes predicted by the national model is much smaller than that suggested by the GTAP simulations. (See the appendix to Chapter 3 for a detailed comparison.) World food and agricultural prices facing China rise relative to nonfood prices, based on the global modeling exercise, with the increases ranging from 4 percent to 12 percent in the case of full liberalization, but considerably smaller for the Doha scenario. These are reported in the second and third sets of columns in table 10.3. The final set of columns in table 10.3 reports the percentage cut in the tariff rates in China under each scenario. In the case of Doha, they are in the range of one-quarter to one-third of Full-Lib (-100 percent). The final scenario of this chapter explores the potential poverty impacts of investing in rural education. This scenario equalizes the urban-rural imbalance in per capita government spending on education by increasing government spend- ing on rural education by 16 percent to bring per capita rural spending in line with that in urban areas.8 Since the emphasis here is on the impact of this reform in the context of multilateral trade liberalization, it is treated as a combined sce- nario, with the rural educational reforms added to the global liberalization sce- nario (Full-Lib). The shorthand for this combined scenario is Edu-Lib. Economywide Impacts of Multilateral Trade Liberalization The macroeconomic results from the trade reform scenarios are reported in the first four columns of table 10.4. The reported values are deviations from the base- line in 2005. The reduction in global trade barriers gives a substantial boost to trade in China, with both exports and imports rising by about 2­3 percent in the Doha Round trade liberalization and 5­6 percent in the scenario of global free trade (Full-Lib). Aggregate welfare, which is measured by the summation of indi- vidual household equivalent variation (EV) and reported as a percentage of GDP, would increase by about 0.4 percent under the Doha trade reforms and 0.8 per- cent in the scenario of global free trade, because of improved TOT and reduced distortions between world prices and domestic prices. China's welfare gain from global trade liberalization comes entirely from the liberalization of other coun- tries. Actually, in the scenario of unilateral trade liberalization, China experiences a welfare loss of 0.3 percent of its GDP as a result of a deterioration in its TOT, which are endogenous in this model. This reflects China's relatively low-level import protection after its WTO accession and its growing influence in world export markets, which in turn tends to reduce its export demand elasticities. Impacts of the DDA on China 303 Table 10.4. Aggregated Results (percent change) Edu-Lib ROW- Uni- Full- Incre- Cumu- Lib Lib Lib Doha mental lative Macroeconomic variables Welfare (EV) 1.0 -0.3 0.8 0.4 1.2 2.0 GDP 0.1 -0.1 0.1 0.0 1.2 1.2 Exports 0.2 4.4 4.8 2.2 1.0 5.8 Imports 1.9 3.8 5.9 2.9 0.8 6.7 TOT 1.4 -0.8 0.5 0.5 -0.3 0.3 CPI 4.6 -2.1 2.4 0.7 0.9 3.3 Capital stock 1.7 -0.6 1.1 0.3 1.1 2.2 Factor prices Returns to agricultural land 23.2 -7.2 15.5 5.2 7.0 23.5 Unskilled wages Urban 5.2 -1.9 3.3 1.0 14.8 18.6 Rural, nonagricultural 5.3 -1.9 3.3 1.0 16.7 20.6 Agricultural 6.1 -2.3 3.9 1.3 28.3 33.3 Semiskilled wages Urban 5.4 -1.9 3.4 1.1 -4.0 -0.7 Rural, nonagricultural 5.8 -2.0 3.7 1.2 -4.9 -1.4 Agricultural 5.2 -1.9 3.3 1.0 -4.2 -1.0 Skilled wages Urban 5.6 -1.7 3.8 1.2 -0.5 3.3 Rural, nonagricultural 5.6 -1.7 3.8 1.2 -0.5 3.3 Agricultural 5.9 -1.8 4.0 1.2 -3.9 -0.1 Labor migration (million) Off-farm labor -3.3 1.2 -2.2 -0.8 4.9 2.8 Unskilled -0.5 0.2 -0.4 -0.1 -10.2 -10.6 Semiskilled -2.6 1.0 -1.7 -0.6 13.3 11.6 Skilled -0.2 0.1 -0.1 0.0 1.9 1.8 Rural-urban -2.4 0.9 -1.5 -0.6 2.0 0.4 Unskilled -0.3 0.1 -0.2 -0.1 -10.4 -10.6 Semiskilled -1.9 0.7 -1.2 -0.5 10.5 9.2 Skilled -0.2 0.1 -0.1 0.0 1.9 1.8 Labor migration (%) Off-farm labor -2.5 0.9 -1.6 -0.6 3.8 2.1 Unskilled -0.9 0.3 -0.6 -0.2 -18.4 -18.9 Semiskilled -3.8 1.4 -2.5 -0.9 19.7 16.7 Skilled -2.3 1.0 -1.4 -0.5 27.9 26.1 Rural-urban -3.7 1.4 -2.3 -0.9 3.1 0.7 Unskilled -1.0 0.4 -0.7 -0.2 -38.2 -38.6 Semiskilled -6.1 2.4 -3.9 -1.4 34.2 29.0 Skilled -3.0 1.2 -1.8 -0.7 35.9 33.5 Source: Simulation results. 304 Poverty and the WTO: Impacts of the Doha Development Agenda With fixed labor endowments, full employment, and no productivity changes, China's GDP is little changed under trade reform. The small increase under Full- Lib is driven by the effects on labor reallocation and capital accumulation. In the case of Doha Round trade liberalization as well as global free trade, stronger export demand in agricultural products and larger cuts in tariff rates for manu- factured goods divert the labor force from high productivity, manufacturing sec- tors to lower productivity, agricultural sectors, when compared to the baseline outcome.9 Although capital stock rises slightly, spurred by the trade liberalization, the increased capital stock is largely offset by the productivity loss associated with more labor employed in agriculture and the rural sector, resulting in minimal gains in real GDP. However, in the scenario of China's unilateral liberalization, the deterioration in terms of trade reduces the profits of exports, which consequently discourages the capital accumulation, resulting in a lower level of steady-state cap- ital stock. Although some of the agricultural labor force is diverted to nonagricul- tural activities, this does not offset the adverse effect of less capital stock. As a con- sequence, GDP slightly declines in the scenario of China's unilateral liberalization. Turning to the changes in factor prices, it can be seen that the effects of global free trade and Doha Round trade liberalization on wages are largely neutral across skill levels and between rural and urban sectors. The increase in agricultural prof- itability, which is reflected in the rise of returns to agricultural land, increases the on-farm demand for labor and therefore reduces off-farm labor supply by about 0.8 million in the Doha Round trade liberalization and 2.2 million in the scenario of global free trade, relative to baseline. Urban and rural nonfarm wages are linked through the temporary migration of individuals to urban areas. In the multilat- eral trade liberalization scenarios, temporary migration from the rural to the urban sector is slowed, with about 1.5 million fewer migrants under Full-Lib than would be the case in the baseline. Because poverty and income distribution are central to this chapter, it provides several such measures for China as a whole in table 10.5. The urban-rural income ratio declines in all three global trade liberalization scenarios, although the mag- nitude of this change is very small--0.01 point in the case of global free trade. This is also reflected in a small improvement in urban-rural inequality, as meas- ured by the Gini coefficient. However, there are no discernible changes in inequal- ity within the urban and rural areas. Using the US$2 per day poverty line, Chen and Ravallion (2004) estimate that 45.2 percent of the rural population in China and 4.1 percent of the urban popu- lation are in poverty. Applying these figures to the benchmark data for 2005 pro- vides the poverty line of Y4,730 (1997 prices) for urban households and Y3,580 (1997 prices) for rural households. By assuming a uniform distribution of the population within each of the vingtiles, it is possible to estimate how the poverty headcount changes in the wake of these reforms. This information is also reported Impacts of the DDA on China 305 Table 10.5. Effects on Inequality and Poverty Edu-Lib Base Full-lib Doha Incremental Cumulative Inequality Urban-rural income 3.213 0.012 0.005 -0.230 -0.242 Ratio Gini 0.438 0.001 0.001 -0.015 -0.016 Urban 0.291 0.000 0.000 0.003 0.003 Rural 0.298 0.000 0.000 0.001 0.001 Poverty headcount ratio (%) Changes (percentage point) Total 31.3 -0.8 -0.4 -3.3 -4.2 Urban 4.1 -0.1 0.0 0.3 0.2 Transfer specialized 24.7 0.0 0.0 0.1 0.1 Labor specialized 3.8 -0.1 -0.1 0.3 0.2 Diversified 2.5 -0.1 0.0 0.3 0.2 Rural 45.2 -1.2 -0.6 -5.2 -6.4 Agriculture-specialized 54.3 -1.2 -0.5 -4.7 -5.8 Diversified 44.1 -1.2 -0.6 -5.3 -6.5 (millions of (% change) persons) Total 413.7 -2.7 -1.3 -11.0 -13.4 Urban 18.2 -2.1 -1.2 7.3 5.0 Transfer-specialized 5.3 -0.1 -0.1 0.5 0.4 Labor-specialized 6.7 -2.5 -1.4 8.5 5.8 Diversified 6.1 -3.5 -2.0 12.0 8.1 Rural 395.5 -2.7 -1.3 -11.8 -14.2 Agriculture-specialized 49.8 -2.1 -1.0 -8.8 -10.7 Diversified 345.7 -2.8 -1.3 -12.3 -14.7 Source: Simulation results. in table 10.5. In the Full-Lib scenario, the monetary poverty line increases by 2.4 percent following the change in the CPI (table 10.4). Nevertheless, higher factor earnings mean that the poverty headcount ratio declines for all household groups. Because transfer incomes are assumed to be constant in real terms and are indexed by the CPI, the urban transfer-specialized household group experiences only a modest decline in its poverty headcount, reflecting a decline in nontransfer income, which makes up less than 5 percent of this group's total income. The aggregate urban poverty headcount decreases by about 2.1 percent. Rural house- holds enjoy a 2.7 percent reduction in poverty headcount, which amounts to a 1.2 percentage point reduction in rural poverty (that is, the proportion of the entire 306 Poverty and the WTO: Impacts of the Doha Development Agenda Figure 10.1. Change in Sector Output, Full-Lib 20 15 10 steel y mining mining and ore metals change) 5 y ore iron machiner y y equipment (% 0 Ferrous Chemicals Cotton Primar Electric Forestr Nonferrous Nonferous Electronics Special Instruments Machiner Automobile Wheat Output ool Rice W forage Sugar extileT food Corn crops Apparel Leather ­5 and Other mill Processed ­10 Grain ­15 Sector Source: Authors' simulations. rural population in poverty falls by 1.2 percent). Given the large population base in rural China, this translates into a rural poverty reduction of 10.6 million people. The Doha scenario shows a similar pattern of poverty reduction across households, but with lesser absolute reductions. Overall, the impoverished share of the national population falls from 31.3 percent of total population to 30.5 percent in the sce- nario of global free trade, and to 30.9 percent under the Doha Round scenarios. Sector Impacts Figure 10.1 and figure 10.2 report a subset of the changes in sector output, in descending order, omitting the changes that are less than 2 percent for global free trade and less than 0.7 percent in absolute value for the Doha Round trade liber- alization, respectively. In both scenarios, the largest increases in output are due to the expansion of textiles and apparel exports, with these products, as well as the production of synthetic fibers, increasing by substantial amounts in the wake of tariff cuts in overseas markets. Some agricultural sectors, such as wool, corn and grain milling, and feedstuffs, also enjoy a boost in output, particularly under the Impacts of the DDA on China 307 Figure 10.2. Change in Sector Output, Doha 6 5 4 3 change) y steel mining 2 gas & (% mining metals ore & Oil y iron ore 1 machiner oil y equipment y Electric Crude Chemicals Forestr Primar Ferrous egetableV Nonferrous Nonferrous Machiner Special Automobile Output 0 ool ­1 W Leather extileT Sugar Corn forage fibers Apparel Cotton & ­2 mill Chemical Grain ­3 Sector Source: Authors' simulations. Full-Lib scenario (figure 10.1). On the other end of the spectrum, the most heav- ily protected sectors, with sizable trade exposure, experience declining output-- automobiles, machinery, special equipment, nonferrous metal products, and veg- etable oil. In the case of global free trade, wheat production shows the largest reduction in output, because of the very large reduction in China's tariff under that scenario. Trade volume changes associated with each of the trade reform experiments are reported in table 10.6. With the exception of a few mining products and trans- port services for which there is no cut in protection, import volumes increase for all sectors in the economy in the scenarios of Doha Round trade liberalization. The largest increases are for automobiles, as well as textiles, apparel, and leather products, where the demand for intermediate inputs increases strongly. Export volumes for most products also increase--especially for rice, corn, grain milling and feedstuffs, textiles, and apparel--fueled by increased demand in the global market. Those sectors with slight or negative increments in exogenous export demand, such as vegetable oil, nonferrous metals, some mining products, machinery, special equipment, and automobiles, experience reductions in export volumes under the Doha scenario. In the case of global free trade, the changes in both imports and exports are much more significant. Despite relatively large increases in the world price of 308 Poverty and the WTO: Impacts of the Doha Development Agenda Table 10.6. Sector Volume Impacts of Trade Liberalization: Percentage Deviation from Baseline Full-Lib Doha Sector Import Export Import Export Rice -29.3 290.5 -0.6 60.7 Wheat 83.7 n.a. 0.6 n.a. Corn 55.2 54.4 6.6 13.4 Cotton 54.5 28.8 5.6 12.3 Other nongrain crops 33.6 22.1 2.9 7.0 Forestry 16.9 -2.5 5.2 -1.7 Wool 2.7 27.9 2.9 5.8 Other livestock 2.2 4.3 3.1 0.3 Fishing 5.8 9.7 1.5 2.2 Other agriculture 17.3 34.9 2.5 13.5 Coal mining 15.2 -0.9 4.9 -0.6 Crude oil and natural gas 1.0 -1.8 0.1 -0.8 Ferrous ore mining -2.5 -1.0 0.1 -1.0 Nonferrous ore mining -3.6 -3.1 0.0 -1.4 Other mining 1.6 0.1 0.6 0.0 Vegetable oil 29.0 20.1 5.2 -4.8 Grain mill and forage -2.2 247.4 -3.2 29.5 Sugar 38.1 162.7 0.9 22.7 Processed food 18.1 32.8 7.7 1.7 Beverage 17.1 18.4 3.4 2.3 Tobacco 7.3 17.3 2.8 2.3 Textile 12.9 15.6 7.4 5.6 Apparel 15.2 13.0 7.1 8.5 Leather 13.1 12.9 9.8 9.5 Sawmills and furniture 4.5 -1.6 3.0 -0.5 Paper and printing 4.0 -1.7 2.6 -0.7 Social articles 0.7 -1.6 2.7 1.1 Petroleum refining 7.9 1.8 2.5 -0.2 Chemicals 6.6 0.8 3.2 0.1 Medicine 12.3 1.4 5.1 0.2 Chemical fibers 10.5 1.9 5.6 2.0 Rubber and plastics 6.1 0.2 3.4 1.5 Building materials 11.7 5.0 4.9 1.6 Primary iron and steel 3.2 -1.7 2.0 0.6 Nonferrous metals 1.2 -3.4 1.6 -0.4 Metal products 8.2 2.6 4.2 1.6 Machinery 7.1 -12.1 3.4 -3.0 Impacts of the DDA on China 309 Table 10.6. (Continued) Full-Lib Doha Sector Import Export Import Export Machinery 7.1 -12.1 3.4 -3.0 Special equipment 4.5 -12.0 2.4 -3.0 Automobile 26.8 -11.1 9.4 -2.6 Other transport equipment 6.9 11.5 3.8 1.9 Electric machinery 9.7 -3.3 4.9 0.8 Electronics -1.6 -5.8 0.9 -0.2 Instruments 3.0 -3.7 1.9 0.6 Other manufacturing 5.9 -2.2 3.6 0.3 Utilities -1.4 -0.2 1.1 -0.3 Construction -0.8 -2.4 1.4 -1.0 Transportation -1.2 -1.1 0.7 -0.7 Post and communications -1.7 -0.7 0.8 -0.3 Commerce -0.5 -1.5 1.0 -0.8 Finance -0.5 -1.8 1.4 -0.7 Social services -0.5 -1.6 1.4 -0.7 Education and health 0.2 -2.7 1.7 -0.9 Public administration 0.0 -2.8 1.6 -1.0 Source: Simulation results. Note: n.a. = not applicable (no trade flow). imports into China, import volumes grow by 20­50 percent in most crops and food sectors, because of their large reduction in import protection. One exception is imports of rice, which would decline as a result of the low initial protection and large increment in the import price. Similar to the cases of Doha Round trade lib- eralization, the rise of imports in automobiles and textiles and apparel are also large. The agricultural and food sectors, textiles and apparel, and the other trans- portation equipment sector are the major gainers in terms of export volume. However, there are also manufacturing sectors that would experience reductions in export volume. The large expansion in China's agricultural exports under the global free trade scenario can be better understood against the backdrop of the significant cuts in agricultural protection in the Republic of Korea and Japan. Given its close geo- graphic proximity to (and strong trade linkage with) these countries, China would benefit from the strong agricultural import growth in these markets after global 310 Poverty and the WTO: Impacts of the Doha Development Agenda trade liberalization. However, because of its small export volumes in most agricul- tural products, China is still a small agricultural exporter in the world market. In the case of grains, this chapter's baseline scenario predicts that China's exports of rice and corn will be Y 2.0 billion and Y 4.8 billion (1997 prices), respectively, in 2005. Even under the scenario of global free trade, its exports of rice and corn are only Y 7.9 billion and Y 7.5 billion (1997 prices). These changes are no larger than recent annual fluctuations in exports of these commodities.10 Household Impacts The analysis now turns to figures 10.3a and 10.3b, which report the household impacts of trade liberalization, by stratum, across the income spectrum. The first point to note from figure 10.3a is that global trade liberalization benefits all households, except those reliant on transfers. Because the transfers are held con- stant in real terms, and transfers make up most of their income, the transfer group is little affected by the trade liberalization. Among the other urban households, the smallest welfare increases in figure 10.3a are associated with urban, diversified households. This contrasts with the relatively larger gains made by the urban, labor-specialized households. The dif- ference occurs because the urban diversified households have significant income from capital earnings, particularly the wealthiest households. The increases in rates of return to other factors are larger than the increases in capital stock, thus the highest-income, diversified households benefit proportionately less than the other, labor-specialized urban household groups. The largest increases in welfare after global trade liberalization accrue to the rural households, especially the wealthier, agriculture-specialized households. They benefit because returns to agricultural land increase relative to other factor prices. Real income rises less for rural, diversified households because of the dominance of nonfarm wage earn- ings. Similar patterns of household incidence emerge from the Doha scenario. Impact of Investing in Rural Education As noted previously, one of the keys to enhancing the welfare of the rural poor-- particularly those reliant on agriculture for their income--is to enhance their off- farm employment opportunities. Econometric evidence suggests that education has proven to be one of the key determinants of off-farm employment. Therefore, this chapter turns next to a comparison of the impact of improved access to educa- tion for the rural households with the global free trade experiment implemented previously. The incremental aggregate effects of rural education reform are reported in the sixth column of table 10.4. Real GDP and welfare rise by 1.3 percent Impacts of the DDA on China 311 Figure 10.3a. Impacts on Households, Full-Lib 1.8 1.6 1.4 1.2 1.0 income 0.8 of 0.6 % as 0.4 EV 0.2 0.0 ­0.2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 ­0.4 Vingtile Urban transfer specialized Urban labor specialized Urban diversified Rural agriculture Rural diversified specialized Figure 10.3b. Impacts on Households, Doha 0.9 0.8 0.7 0.6 0.5 income 0.4 of 0.3 % as 0.2 EV 0.1 0.0 -0.1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 -0.2 Vingtile Urban transfer specialized Urban labor specialized Urban diversified Rural agriculture Rural diversified specialized Source: Authors' simulations. 312 Poverty and the WTO: Impacts of the Doha Development Agenda and 1.2 percent, respectively, as a result of increasing the rural education spending by 16 percent. Clearly, from an economywide point of view, rural education is a favorable investment, given the assumptions about productivity differentials, edu- cation costs, and financing mechanisms (a mix of public and private funding). Three factors contribute to the observed GDP growth after investment in rural education: (1) As a result of improved access to education, average schooling attainment of unskilled rural labor increases by 1.7 years.11 It results in higher productivity, which largely offsets the decline in the amount of unskilled labor supply. (2) Improved rural education increases the supply of rural semiskilled and skilled labor 16 percent, and unskilled labor declines by 23 percent. This favorable change in skill composition induces an economywide productivity gain. However, the forgone working hours from higher school participation rates are quite mod- est: the supply of aggregate rural labor declines by only 0.29 percent.12 (3) Higher educational attainment also improves off-farm labor mobility. As a result of improved rural education, 4.9 million additional workers leave agriculture and an additional 2.0 million temporal migrants move to urban areas. This movement of labor from relatively low-productivity sectors (agriculture and rural nonfarm employment) into higher-productivity activities (rural nonfarm work and urban employment, respectively) also boosts overall productivity. With an increase in the pool of semiskilled rural workers of 42.4 million, migration among this group out of agriculture increases by 13.3 million workers. Temporary migration of semiskilled workers to urban areas also rises by 10.5 mil- lion workers, contributing to a decline in urban semiskilled wage rates. Because this work abstracts from any transaction costs associated with the temporary rural-urban migration of skilled labor, the bulk of the increased supply of rural skilled workers (about 1.9 million, or 82 percent of the supply increment) migrates to urban areas. However, its impact on the urban skilled wage is very lim- ited, given the small size of temporary skilled migration compared to the stock of urban skilled workers. Also, with the combination of a diminished supply and an enhanced schooling attainment of unskilled workers in the rural areas, wages for this group rise sharply. As a consequence, both off-farm employment and tempo- rary migration of unskilled labor to urban areas actually decline. The distributional impacts of improved rural education can be seen in figure 10.4a, which reports the incremental welfare change of disaggregated household groups in urban and rural China. Most urban households lose under this scenario because they face more intense competition from increasingly well-educated and mobile rural workers. Furthermore, given the closure rules used in this chapter's model, the additional government expenditure on rural education is financed via a direct tax on household income. Therefore, urban households pay part of the costs of increased rural education.13 Lower-income households in the urban areas Impacts of the DDA on China 313 experience bigger losses because they rely more heavily on semiskilled labor income. As a consequence, the urban Gini index rises by 0.003 (table 10.5). How- ever, household welfare rises for all rural households. The largest proportional increase in welfare is for the agriculture-specialized rural households, which ben- efit from the strong increase in rural unskilled wages. Overall, the benefits from rural educational reform are spread relatively evenly across income levels, and the rural Gini index is hardly changed. The educational reform induces a 0.23 point decline in the urban-rural income ratio and a 0.015 decline in national Gini coef- ficient, indicating an improvement in urban-rural income distribution in China. Returning to table 10.5, it can be seen that the rural poverty headcount falls sig- nificantly, by 11.8 percent, after the investment in rural education. The largest fall is ascribed to diversified rural households. The poverty headcounts of urban labor-specialized and diversified urban household groups increase by 8.5 percent and 12.0 percent, respectively. However, given the share of urban poverty in the overall population, the deterioration of urban poverty is more than offset by the alleviation of rural poverty, and national poverty headcount falls by 44.3 million. The combined aggregate impact of both global free trade and improvement in rural education is reported in the final columns of tables 10.4 and 10.5. The results show that these reforms are potentially significant for the Chinese econ- omy. As a major indicator of overall efficiency, GDP increases by 1.2 percent, and aggregate welfare rises by 2.0 percent. Figure 10.4b shows the cumulative effect of global free trade and educational reform on disaggregate urban and rural household welfare. Here, the potential urban-rural redistribution of welfare is striking. The equivalent variation for agri- culture-specialized rural households is about 7­9 percent of initial income. Other rural households also benefit from these reforms. In contrast, urban household welfare falls by as much as 2 percent of initial income for the poorest urban households. Clearly, the reforms aiming at global free trade and promoting rural education would boost rural household welfare, but this does come at the expense of urban households, particularly the lower-income groups. However, when viewed in a historical context, this redistribution is quite modest. It does little more than undo the worsening of the urban-rural income disparity that has arisen since 1998. The combined education and trade reforms also contribute significantly to rural poverty reduction. The rural poverty headcount ratio declines by 14.2 per- cent, or 6.4 percentage points--from 45.2 percent in the base case to 38.8 percent in the Edu-Lib scenario--and the urban headcount ratio rises slightly, from 4.07 percent to 4.27 percent. Overall, the number of people in poverty nationwide declines by 55 million when rural education reforms are combined with global trade liberalization. 314 Poverty and the WTO: Impacts of the Doha Development Agenda Figure 10.4a. Incremental Impacts on Households, Edu-Lib 10.0 8.0 6.0 income 4.0 of % 2.0 as 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 EV 0.0 ­2.0 ­4.0 Vingtile Urban transfer specialized Urban labor specialized Urban diversified Rural agriculture Rural diversified specialized Figure 10.4b. Cumulative Impacts on Households, Edu-Lib 10.0 8.0 6.0 income 4.0 of % 2.0 as 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 EV 0.0 ­2.0 ­4.0 Vingtile Urban transfer specialized Urban labor specialized Urban diversified Rural agriculture Rural diversified specialized Source: Authors' simulations. Impacts of the DDA on China 315 Of course, an important question remains: How much interaction is there between the rural education reforms and multilateral trade liberalization? To iso- late this interaction, the Full-Lib scenario was repeated using the database and parameters that result from implementation of the education reforms. The results of the Full-Lib experiment in the wake of the education reform were nearly iden- tical, suggesting that there is little interaction between the two policies. In other words, the cumulative impact of undertaking both sets of reforms is essentially the sum of the two individual impacts. Conclusions and Policy Implications The goal of this chapter has been to assess the implications of multilateral trade reforms for poverty in China, which is done by combining results from a global modeling exercise with a national CGE model that features disaggregated house- holds in both the rural and urban sectors. Three different scenarios are examined: one involving global trade liberalization and two involving possible DDA reforms. Using the World Bank's US$2 per day poverty line, the results show that multilat- eral trade reforms do in fact reduce poverty in China. The biggest reductions occur in the rural areas, largely as a result of higher prices for farm products. Because this is where the bulk of the poor in China reside, an overall reduction in poverty follows. Urban poverty falls in two of the three household groups considered in this analysis, because the increased demand for China's products in world markets boosts factor earnings sufficiently to offset the impact of higher food prices. For the remaining group, which is heavily dependent on transfer payments, it is assumed that indexation of these payments will largely offset the adverse conse- quences of higher prices. However, a decline in other income sources is sufficient to cause an increase in poverty, and this increase is large enough to boost the over- all urban poverty headcount. However, the urban poor represent only 5 percent of the total poor in China, and thus the national poverty headcount falls. The implications of complementary reforms in China are also explored--in particular, increased investments in rural education aimed at equalizing per capita spending between rural and urban areas. This boosts rural enrollments by 16 per- cent, which has the twin benefits of increasing labor productivity as well as enhancing the mobility of the rural labor force, thereby putting these workers in a better position to benefit from trade reforms. The analysis in this chapter takes account of the cost of funding these additional students, as well as the reduction in the workforce that results from having more pupils in school. Nevertheless, these reforms generate very substantial gains for China's economy. They also serve to boost rural incomes and reduce the incidence of rural poverty. Indeed, when 316 Poverty and the WTO: Impacts of the Doha Development Agenda combined with global trade liberalization, poverty in China is estimated to drop by about 55 million people. Notes 1. However, as noted by Parish, Zhe, and Li (1995), the rural labor market is looking more like a market all the time. 2. See Hertel and Zhai (2004) for the details of the off-farm labor supply behavior in the model. 3. In this model, it is assumed that the elasticity of land with respect to on-farm labor is unitary. 4. There are likely other, unobserved factors inducing this rural-urban wage differential, in which case estimation of the labor market distortion via subtraction of known factors is biased in the direc- tion of overstating the hukou-related distortion. Therefore, it is useful to also estimate the direct impact of household registration status on the observed wage difference among households. Shi (2002) takes this approach to the problem, using the same CHNS dataset. He finds that only 28 percent of the rural-urban wage difference can be explained directly via the coefficient on the hukou registra- tion variable. This is quite a bit less than the 48 percent left unexplained via the subtraction approach of Shi, Sicular, and Zhao (2002). 5. See Hertel and Zhai (2004) for a detailed description of how this ad valorem distortion is obtained. 6. It is assumed that a doubling of temporary migration would increase the marginal cost of migration by only 10 percent. 7. Because of the variety of labor supply elasticities in response to the three different wages in their model, the authors obtain a variety of labor transfer elasticities, depending on the "thought experi- ment" being conducted. These are asymmetric, with the response to a change in shadow wages differ- ing from the response of labor supply to a change in the market wage. However, this response is treated as symmetric in the model in this chapter. This makes it difficult to choose the correct parameter for the analysis. The focus is on the transfer of labor from agriculture to market wage employment in response to a change in returns to agriculture, because this transfer accounts for the bulk of the labor flow in this analysis. 8. A caveat should been mentioned here. The scenario of education reform assumes that the pri- vate education spending of rural households increases proportionally to the public spending. It implies that the education demand in rural areas is constrained by the supply-side factors. This seems reasonable, given the low level of rural education in China, the potential benefits from education investment, and the long-term nature of the simulation. 9. In this model, there are exogenous differences in labor productivity across sectors, inferred from observed wage rates. 10. In the case of rice exports, this increase is comparable in size to that observed between 1997 and 1998, when rice exports increased from about Y 2 billion to nearly Y 9 billion, after which it steadily declined, returning to about Y 2 billion by 2004. In the case of maize, the projected change is smaller than recent annual export fluctuations. 11. The model predicts that only the unskilled labor force will experience an increment in school- ing attainment because it is assumed that the same proportional increase in the school participation rates occurs across grades. Thus, the increase in semiskilled labor at the low end is offset by a reduction at the high end as more semiskilled workers become skilled. However, at the low end, the increase in unskilled labor is fueled by a decline in the share of illiterate people in the unskilled labor force. 12. The relatively small reduction in the rural labor force is perhaps somewhat surprising. How- ever, this is the consequence of several factors. First, the 1.7 increment to schooling years applies only to the unskilled rural labor force. Second, the ages of the labor force in this model are from 15 to 70. The ages of pupils are from 7 to 25, thus increasing enrollment rates of primary school (0­6 schooling Impacts of the DDA on China 317 years for pupils at 7­12 years old) and middle school (7­9 schooling years for pupils at 13­15 years old) has no direct impact on total labor supply. 13. Because it is assumed that the income tax is levied on the nontransfer incomes, urban transfer- specialized households do not bear the costs of additional rural education. References Bhattasali, Deepak, Shantong Li, and Will Martin, eds. 2004. China and the WTO: Accession, Policy Reform, and Poverty Strategies. New York: Oxford University Press; Washington, DC: World Bank. Chen, Shaohua, and Martin Ravallion. 2004. "Welfare Impacts of China's Accession to the WTO." In China and the WTO: Accession, Policy Reform, and Poverty Strategies, ed. Deepak Bhattasali, Shan- tong Li, and Will Martin, 283­304.New York: Oxford University Press; Washington, DC: World Bank. Dervis, K., J. de Melo, and Sherman Robinson. 1982. General Equilibrium Models for Development Pol- icy. Cambridge, U.K.: Cambridge University Press. Francois, Joseph and Bradley McDonald. 1996. "Liberalization and Capital Accumulation in the GTAP Model." GTAP Technical Paper 7, Purdue University, West Lafayette, IN. www.gtap.org. Hansen, Terje, and Tjalling Koopmans. 1972. "On the Definition and Computation of a Capital Stock Invariant under Optimization." Journal of Economic Theory 5: 487­523. Harrison, Glenn, Thomas Rutherford, and David Tarr. 1997. "Quantifying the Uruguay Round." Eco- nomic Journal 107 (444): 1405­30. Heckman, James J. 2002."China's Investment in Human Capital." NBER Working Paper 9296. National Bureau of Economic Research, Cambridge, MA. Heckman, James J., and Lance Lochner. 1998. "Explaining Rising Wage Inequality: Explorations with a Dynamic General Equilibrium Model of Labor Earnings with Heterogeneous Agents." Review of Economic Dynamics 1 (1): 1­58. Hertel, Thomas W., ed. 1997. Global Trade Analysis: Modeling and Applications. Cambridge: Cambridge University Press. Hertel, Thomas, David Hummels, Maros Ivanic, and Roman Keeney. 2003. "How Confident Can We Be in CGE-Based Assessments of Free Trade Agreements?" GTAP Working Paper 26, Purdue Uni- versity, West Lafayette, IN. Hertel, Thomas W., Paul V. Preckel, John Cranfield, and Maros Ivanic. 2004. "Poverty Impacts of Mul- tilateral Trade Liberalization." World Bank Economic Review 18 (2): 205­36. Hertel, Thomas, and Fan Zhai. 2004. "Labor Market Distortions, Rural-Urban Inequality, and the Opening of China's Economy." Policy Research Working Paper 3455. World Bank, Washington, DC. Hertel, Thomas, Fan Zhai, and Zhi Wang. 2004. "Implications of WTO Accession for Poverty in China." In China and the WTO: Accession, Policy Reform, and Poverty Strategies, ed. Deepak Bhat- tasali, Shantong Li, and Will Martin, 283­304.New York: Oxford University Press; Washington DC: World Bank. Kanbur, R., and Xiaobo Zhang. 2001. "Fifty Years of Regional Inequality in China: A Journey through Revolution, Reform and Openness." CEPR Discussion Paper 2887. Centre for Economic Policy Research, London. Katz, L., and K. Murphy. 1992. "Changes in Relative Wages, 1963­1987: Supply and Demand Factors." Quarterly Journal of Economics 107 (1): 35­78. Li, Shi, and Ximin Yue. 2004. "Latest Changes in Individual Income Inequality in China." Working Paper, Economic Research Institute, China Academy of Social Science, Beijing. Parish, William L., Xiaoye Zhe, and Fang Li. 1995. "Nonfarm Work and Marketization of the Chinese Countryside." China Quarterly 143: 697­730. Shi, Xinzheng, 2002. "Empirical Research on Urban-rural Income Differentials: The Case of China." Unpublished manuscript, CCER, Beijing University. 318 Poverty and the WTO: Impacts of the Doha Development Agenda Shi, Xinzheng, Terry Sicular, and Yaohui Zhao. 2002. "Analyzing Urban-Rural Income Inequality in China." Paper presented at the International Symposium on Equity and Social Justice in Transi- tional China, Beijing, July 11­12. Shoven, J. B., and J. Whalley. 1992. Applied General Equilibrium Analysis. Cambridge, U.K.: Cambridge University Press. Sicular, Terry, and Yaohui Zhao. 2004. "Earnings and Labor Mobility in Rural China: Implications for China's Accession to the WTO." In China and the WTO: Accession, Policy Reform, and Poverty Strategies, ed. Deepak Bhattasali, Shantong Li, and Will Martin, 239­260. New York: Oxford Uni- versity Press; Washington, DC: World Bank. Wang, Dewen. 2003. "China's Rural Compulsory Education: Current Situation, Problems and Policy Alternatives." Working Paper Series 36, Institute of Population and Labor Economics, Chinese Academy of Social Sciences, Beijing. Yang, Dennis T. 1997."China's Land Arrangements and Rural Labor Mobility." China Economic Review 8 (2): 101­16. ------. 2004."Education and Allocative Efficiency: Household Income Growth during Rural Reforms in China." Journal of Development Economics 74 (1): 137­62. Zhang, Linxiu, Jikun Huang, and Scott Rozelle. 2002. "Employment, Recessions, and the Role of Edu- cation in China." China Economic Review 13 (2, 3): 313­28. Zhao, Yaohui. 1999a. "Labor Migration and Earnings Differences: The Case of Rural China." Economic Development and Cultural Change 47 (4): 767­82. ------. 1999b. "Leaving the Countryside: Rural-to-Urban Migration Decisions in China." American Economic Review 89 (2): 281­86. 11 The Social Impact of a WTO Agreement in Indonesia Anne-Sophie Robilliard and Sherman Robinson Summary Indonesia has experienced rapid growth and the expansion of the formal financial sector during the last quarter of the 20th century. Although this tendency was reversed by the shock of the financial crisis that spread throughout Asia in 1997 and 1998, macroeconomic stability has since been restored, and poverty has been reduced to precrisis levels. Poverty reduction nevertheless remains a critical chal- lenge for Indonesia, with more than 110 million people (53 percent of the popula- tion) living on less than US$2 per day. The objective of this study is to help identify ways in which the Doha Develop- ment Agenda (DDA) might contribute to further poverty reduction in Indonesia. To provide a good technical basis for answering this question, this chapter uses an approach that combines a computable general equilibrium (CGE) with a microsimulation (MS) model. This framework is designed to capture important channels through which macroeconomic shocks affect household incomes. It will allow recommendations to be made on specific trade reform options as well as on complementary development policy reforms. The framework presented in this study generates detailed poverty outcomes of trade shocks. Given the magnitude of the shocks examined here and the structural features of the Indonesian economy, only the full-liberalization scenario generates significant poverty changes. Their impact is examined under alternative specifica- tions of the functioning of labor markets. These alternative assumptions generate 319 320 Poverty and the WTO: Impacts of the Doha Development Agenda different results, and all of them agree that the impact of full liberalization on poverty would be beneficial, with wage and employment gains dominating the adverse food price changes that could hurt the poorest households. Two alterna- tive tax replacement schemes are examined. Although the direct tax replacement appears to be more desirable in terms of efficiency gains and translates into higher poverty reduction, political and practical considerations could lead the govern- ment of Indonesia to choose a replacement scheme through the adjustment of VAT rates across nonexempt sectors. Although the poverty reduction in terms of the number of people lifted out of poverty might appear impressive, the overall impact of trade liberalization is quite modest. One should also keep in mind that these results do not capture two opposing channels through which trade liberalization could affect poverty and income distribution. On the one hand, the results in this chapter are likely to underestimate the cost of temporary unemployment endured by displaced work- ers in some sectors. The reallocation could be all the more harmful because unem- ployment benefits are not available for most Indonesian workers. On the other hand, because it is static in nature, this model ignores positive dynamic gains from liberalization. Introduction Over the last 30 years, decreasing tariffs in both developed and developing countries, as well as declining transport costs and economic growth, have led to a sustained increase in the exports and imports of goods and services by developing countries. This might have benefited many developing countries, contributing to enhanced productivity growth and the development of the financial sector, but it is not obvious whether the poorest households have gained from increased trade liberalization. Such doubts are reinforced because this liberalization has so far been relatively asymmetric, with rich country protection still high in labor- intensive sectors such as wearing apparel and agriculture. This asymmetry is further exacerbated by the subsidies provided to OECD farmers by their governments. Indonesia has experienced rapid growth and the expansion of the formal financial sector during the last quarter of the 20th century. After a period of economic slowdown following the financial crisis, macroeconomic stability has been restored and growth has picked up, reaching 5.1 percent in 2004. Poverty reduction nevertheless remains a critical challenge for Indonesia, with more than 110 million people (53 percent of the population) living on less than US$2 per day. The objective of this study is to help identify ways in which the DDA might contribute to further poverty reduction in Indonesia. To provide a good technical The Social Impact of a WTO Agreement in Indonesia 321 basis for answering this question, an approach is used that combines a CGE with an MS model. This framework is designed to capture important channels through which macroeconomic shocks affect household incomes. It will allow recommendations to be made on specific trade reform options as well as on complementary development policy reforms. This chapter is organized as follows. First, some features of the Indonesian economy are presented, in particular with regard to trade. Second, the analytical framework developed to analyze the impact of further trade liberalization on poverty and income distribution is presented. Third, simulation results are described and commented on. In the fourth section, some of the key assumptions made in the reference scenario are examined, and the impact of trade reform on poverty when these are relaxed is explored. The last section concludes. Economic and Policy Environment Table 11.1 shows that Indonesia's overall openness to foreign trade is relatively high, with exports (imports) amounting to more than 37 percent (28 percent) of GDP. Not surprisingly, table 11.1 also shows that trade exposure is heterogeneous across sectors. Indonesia's trade appears to be concentrated in petroleum and manufactured products. However, agricultural sectors appear to be weakly exposed to trade, contributing very little to total exports and imports, a result of Indonesia's export diversification toward manufacturing products. The textile and wearing apparel industry's contribution to exports is significant (12.7 percent), and that sector's exposure to trade is important, with an export ratio of 41.4 percent and an import ratio of 20.0 percent. Overall, the most exposed nonoil sectors appear to be the wood and wood products industry; the paper printing, transport equipment, and metal products industries, and the chemical fertilization, clay products, cement, and basic metal manufacturing industries. Official data from the WTO Secretariat indicate that average applied most- favored-nation (MFN) tariffs have declined between 1998 and 2002, reflecting mainly unilateral cuts over this period (WTO 2003). Although more than 90 percent of Indonesia's tariffs are bound, a large gap remains between bound and applied rates (35 percentage points, on average). Almost all applied rates are ad valorem, and there are no tariff quotas. Nevertheless, the scope of import restrictions and licensing appears to have increased through new special import licensing from 2002. These affect sensitive products, such as rice and sugar, and are granted based on domestic needs. It should be noted that the new government is looking into removing or reducing these licenses. Import duties across commodities in the base year appear to be relatively low, with values ranging from 1.5 percent for nonfood crops to 6.6 percent for textile and wearing apparel 322 Table 11.1. Trade Structure of Indonesia Exports Imports Parameters Poverty Imports as Exports as a percent of Shares of a percent of Shares of composite Armington Import tariff and total exports sector output total imports good supply elasticities rates the Farm food crops 0.1 0.2 2.4 7.7 4.7 2.2 WTO: Farm nonfood crops 0.6 6.2 0.0 0.3 5.9 1.5 Livestock products 0.0 0.1 0.5 2.7 7.1 3.1 Impacts Forestry and hunting 0.1 2.0 n.a. n.a. 5.0 0.0 Fishery, drying, and salting of fish 0.3 3.1 0.0 0.1 2.5 6.2 Total agriculture 1.1 2.9 of Coal, metal ore, petroleum, the and natural gas 17.9 63.0 2.2 16.2 10.1 1.8 Doha Other mining and quarrying 0.1 3.1 0.4 6.6 1.8 1.5 Food, beverages, and Development tobacco manufacturing 7.3 11.2 3.7 4.9 4.3 6.2 Spinning, textile, leather, and wearing apparel manufacturing products 12.7 41.4 5.2 20.0 7.6 6.6 Wood and wood products 8.4 58.9 0.2 2.8 6.8 4.9 Agenda Paper printing, transport equipment, metal products, and other manufacturing products 15.7 37.9 42.7 57.2 7.6 3.0 Chemical fertilization, clay products, cement, and basic metal manufacturing products 21.3 36.1 27.1 37.0 6.8 2.9 Table 11.1. (Continued) Exports Imports Parameters Imports as Exports as a percent of Shares of a percent of Shares of composite Armington Import tariff total exports sector output total imports good supply elasticities rates Total industry 83.5 81.4 The Wholesale and retail trade, Social transport, storage, and warehousing 1.0 2.2 0.8 1.6 3.8 0.0 Restaurants 1.4 7.5 0.6 3.1 3.8 0.0 Impact Hotels and lodging places 4.0 44.0 1.1 17.1 3.8 0.0 Road transport and railways 1.6 8.5 1.0 5.7 3.8 0.0 Air and water transport and of communications 5.3 28.8 4.7 25.6 3.8 0.0 a WTO Banking and insurance 0.6 5.5 1.9 14.5 3.8 0.0 Real estate and business services 0.1 0.4 2.5 10.6 3.8 0.0 Agreement Public administration, defense, and social, recreational, and cultural services 1.2 3.0 2.0 5.1 3.8 0.0 Personal household and other services 0.2 1.7 1.2 8.4 3.8 0.0 in Total services 15.4 15.7 Indonesia Source: Updated SAM 2002, GTAP Version 6, and authors' calculations. Note: n.a. = not applicable. 323 324 Poverty and the WTO: Impacts of the Doha Development Agenda products. These low values hide somewhat higher values for very specific products that cannot be captured, given the level of aggregation of commodities in the database used here. Despite these higher tariffs, Indonesia has overall low tariffs, including in agriculture. The contribution of tariffs to Indonesia's tax revenue is also an important question. Tariffs accounted for 6.0 percent of government revenue, and corresponded to 0.8 percent of GDP in 2002, the base year chosen as the starting point of the model in this chapter. Indonesia's budget does not appear to be highly dependent on imports duties, but the tax replacement scheme could potentially alter the impact of trade liberalization, depending on the choice of replacement tax used. Despite Indonesia's economic recovery since the 1998 crisis, more than 110 million people (53 percent of the population) are still living on less than US$2 per day. According to Indonesia's official poverty lines, poverty incidence was 18.2 percent in 2002 at the national level, with higher levels in the rural sector (21.2 percent) than in the urban sector (14.5 percent). As a consequence, two-thirds of the poor in Indonesia live in the rural sector. Full liberalization in the rest of the world, particularly the removal of domestic agricultural support in OECD countries, is expected to lead to an increase in the prices of agricultural products. The impact on agricultural households should be positive, but the impact on poverty depends on whether poor agricultural households are net buyers or net sellers of food products. Of course, the size of the domestic market and the resulting low exposure of agricultural sectors to foreign trade mentioned above are likely to dampen the impact of world price changes on the domestic market. In the urban areas, poor households could suffer from an increase in the price of food. The resulting overall poverty impact is thus difficult to predict and depends on the relative size of the efficiency and hence income gains generated by the removal of the import duties and of the changes in the relative price of food. Analytical Framework: A Macro-Micro Model for Indonesia This section describes the specification of the Indonesia macro-micro model used to analyze the social impact of various trade liberalization scenarios. A more detailed discussion of the specification and econometric estimates of the various equations of the household income generation model and simulation methodology may be found in Bourguignon, Robilliard, and Robinson (2003). This chapter's approach combines an MS model with a CGE model in a sequential fashion. The term "microsimulation" spans a number of different approaches used in social sciences. Their common denominator is to focus primarily on the economic behavior of agents and investigate the impact of public The Social Impact of a WTO Agreement in Indonesia 325 policy and shocks at the microeconomic level. These models typically take representative samples of micro agents (households or firms) and measure the effect of government policy on these samples. Various MS techniques are described in more detail in the annex to the associated working paper (Robilliard and Robinson 2005; see also Cogneau, Grimm, and Robilliard 2003). Specifications of the Indonesia CGE model The CGE model is based on a standard social accounting matrix (SAM) and is designed to capture structural features of the economy as well as the general equilibrium effects of the macro constraints arising from macroeconomic shocks. The model was developed from the neoclassical-structuralist modeling tradition originally presented in Dervis, de Melo, and Robinson (1982). It is formulated as a set of simultaneous linear and nonlinear equations, which define the behavior of economic agents, as well as the economic environment in which these agents operate. Indonesia's economy is dualistic, which the model captures by distinguishing between formal and informal activities in each sector. The two subsectors differ in the type of factors they use. This distinction allows treating formal and informal factor markets differently. For all activities, the production technology is represented by a set of nested CES (constant elasticity of substitution) value added functions and fixed coefficient (Leontief) intermediate input relationships. On the demand side, consumers purchase a composite good, and imperfect substitutability is assumed between formal and informal products of the same commodity. Domestic prices of commodities are flexible, allowing markets to clear in a competitive setting where individual suppliers and demanders are price takers. Following Armington (1969), the model assumes imperfect substitutability, for each good, between imports and the corresponding composite domestic commodity. For export commodities, the allocation of domestic output between exports and domestic sales is determined on the assumption that domestic producers maximize profits subject to imperfect transformability between these two alternatives. The composite production good is a constant elasticity of transformation (CET) aggregation of sectoral exports and domestically consumed products. The trade elasticities used to calibrate the functions used in the CGE model were taken from Version 6 of the Global Trade Analysis Project (GTAP) database. These have recently been econometrically estimated for at the disaggregated GTAP commodity level (Hertel and others 2004). There are eight labor categories in the Indonesia CGE model: urban male unskilled, urban male skilled, urban female unskilled, urban female skilled, rural male unskilled, rural male skilled, rural female unskilled, and rural female skilled. The distinction between rural and urban labor implies that workers are not 326 Poverty and the WTO: Impacts of the Doha Development Agenda allowed to shift between rural and urban production sectors. Male and female, as well as skilled and unskilled labor, are assumed to be imperfect substitutes in the production activity of urban or rural sectors. Alternative specifications of the functioning of the labor markets can be implemented in the model. In the reference simulations, wages are assumed to adjust so as to clear all labor markets. This is consistent with the full employment assumption used in the global model. Two alternative specifications are also examined. Land appears as a factor of production in the agricultural sectors. Only one type of land is considered in the model, but capital markets are segmented into six categories: owner-occupied housing, other unincorporated rural capital, other unincorporated urban capital, domestic private incorporated capital, public capital, and foreign capital. Given the medium-term perspective of the model, it is assumed that land is activity specific and other types of capital are competitively allocated among the different sectors so that its marginal value product is equalized across activities. Equilibrium in a CGE model is defined by a set of constraints that need to be satisfied by the economic system but are not considered directly in the decisions of micro agents. Aside from the supply-demand balances in product and factor markets, three macroeconomic balances are specified in the Indonesia CGE model: (1) the fiscal balance, with government savings equal to the difference between government revenue and spending; (2) the external trade balance (in goods and nonfactor services), which implicitly equates the supply and demand for foreign exchange; and (3) savings-investment balance. We assume that savings are investment driven and adjust through flexible saving rates for firms, that foreign savings are in fixed supply with the demand for foreign exchange equated through an endogenous exchange rate, and that government income is also fixed, with lost tariff revenue replaced through a variable tax rate on households. This closure is expected to be relatively neutral in terms of the distributive impact of the shocks implemented. An alternative tax replacement scheme is also examined. The original CGE model was constructed to reflect Indonesia's economic precrisis situation and is based on a 1995 SAM. To capture the postcrisis structural features of the Indonesian economy, the 1995 Indonesia SAM was updated using cross-entropy methods (see Robinson, Cattaneo, and El-Said 2001). The updating procedure imposes the following pieces of information from 2002: value added by activity; the structure of imports and exports by commodity; and the data contained in a macro SAM. Table 11.2 summarizes the aggregate values of the resulting 2002 SAM.1 Table 11.2. A Macroeconomic SAM for Indonesia (2002 rupiah, billions) Activity Commodity Household Government World Investment Total Activity 1,610,012 1,610,012 Commodity 1,042,148 132,219 569,962 325,334 2,069,663 Household 1,538,826 19,246 1,558,072 Government 71,186 12,005 110,845 194,036 The World 447,626 447,626 Social Investment 405,079 61,817 -141,562 325,334 Total 1,610,012 2,069,663 1,558,072 447,646 325,334 Impact Source: Live Database on Line (World Bank); Essama-Nssah (2004). of a WTO Agreement in Indonesia 327 328 Poverty and the WTO: Impacts of the Doha Development Agenda Specification of the Indonesia microsimulation model The microsimulation model is based on household and individual level data from the Special Survey on Saving and Household Investment (SKTIR) for the year 1996 and simulates income generation mechanisms for 9,800 households corresponding to 42,400 individuals.2 Four occupational choices are distinguished at the individual level: (1) inactivity, (2) wage work, (3) self- employment, and (4) multiactivity (wage work and self-employment). All individuals aged 15 years and older are assumed to participate in the occupational choice. Following McFadden's approach to discrete choice behavior (1974), it is assumed that an individual chooses the outcome that maximizes the utility gained from that choice. It is also assumed that the occupational choice behavior of household heads is different from that of spouses or other members of the household. More specifically, for purposes of this analysis, it is posited that the decision process is sequential and that occupational choices for spouses and other members depend on characteristics of the household head and his or her occupational choice. The microeconomic database of the original Indonesia model is the 1996 SUSENAS SKTIR. All individuals aged 10 years and older were interviewed on their sources of income. The sample is updated using a cross-entropy approach (Robilliard and Robinson 2003). This method generates new sets of household statistical weights that are consistent with projected population and structure of the labor force for 2002. Changes in the labor force structure are based on observed changes in the 1996 and 1999 SAKERNAS national labor force surveys for Indonesia. The two models are solved separately. The macro or CGE model communicates with the MS model by generating a vector of prices, wages, and aggregate employment variables corresponding to a given shock or policy. Then the MS model is used to generate changes in individual wages, self-employment incomes, and employment status in a way that is consistent with the set of macro variables fed by the macro model. When this is done, the full distribution of real household income corresponding to the simulated shock or policy may be evaluated. Consistency of the MS model with the equilibrium of aggregate markets in the CGE model requires that three conditions hold. First, changes in average earnings with respect to the benchmark in the MS module must be equal to changes in wage rates provided by the CGE model for each labor market category. Second, changes in agricultural and nonagricultural self-employment income in the MS module must be equal to changes in the corresponding income per worker provided by the CGE model. Finally, changes in the number of wage workers and self-employed workers by labor market category in the MS model must match those same changes generated by the CGE model. The Social Impact of a WTO Agreement in Indonesia 329 Implementation and Analysis of Trade Policy Shocks Various scenarios are examined to inform the DDA negotiations. These scenarios are built upon the work laid out in chapters 2 and 3 of this volume and entail shocks to Indonesia's import prices, export prices, and tariffs.3 In the case of Indonesia, the pre-experiment outlined in chapter 3 was also quite important. Recall that it was designed to take into account China's accession to the WTO, the completion of the Uruguay Round Agreement on Textiles and Clothing commitment to abolish textiles and apparel quotas, and the EU's enlargement to 25 countries. This presimulation scenario (PRESIM) is given explicit attention in this chapter, because it generates a new base from which all subsequent liberalization simulations will start. The employment and earnings impacts generated by the CGE model are fed into the MS model.4 As described above, the MS is used to generate a new distribution of income that can then be compared to the base distribution. Both poverty and inequality indicators are presented, and poverty indicators are computed using official poverty lines.5 The macroeconomic and social impacts of the first set of scenarios are presented in tables 11.3 and 11.4. First, the impact of the PRESIM is examined. The aggregate real impact is small and negative. Private consumption decreases by 0.2 percent. Imports and exports also decrease by 1.4 and 1.0 percent, respectively. These results stem from a slight deterioration of the terms of trade (TOT) facing Indonesia as the demand for its exports fall in the wake of quota elimination and are driven by the fact that the estimated quota rents in the initial database are much larger for exports from China than for other textile exporting countries like Indonesia (Francois and Spinanger 2004). At a more disaggregated level, the shock leads to some reallocation of factors across sectors. Not surprisingly, the textile and wearing apparel sectors face the worst TOT shock, with export prices falling by 3.8 percent and import prices increasing very slightly. As a result, value added from the textile and wearing apparel sectors decreases by more than 30 percent, and factors of production are reallocated toward other manufacturing sectors. At the household level, the poverty and inequality impacts are relatively small. However, these results likely underestimate the cost of reallocation suffered by workers in the textile and wearing apparel sectors. Although these costs are not captured by the model, some displaced workers could suffer, at least temporarily, from unemployment, particularly because they are likely to come from informal sectors that do not provide unemployment benefits. Starting from the updated base year generated with the PRESIM scenario, the full trade liberalization impact is examined through a set of three simulations. The first examines the impact of unilateral liberalization of Indonesia (ULIB); the second considers the impact of full liberalization excluding Indonesia (FLIBX); 330 Table 11.3. Macroeconomic Results of Trade Liberalization Reference Scenarios Base PRESIM New base ULIB FLIBX Full-lib Doha Poverty Private consumption 1,042.1 -0.2 1,040.3 0.0 0.7 0.7 0.1 Investment demand 325.3 324.7 and Government consumption 132.2 132.2 the Total absorption 1,499.7 -0.1 1,497.3 0.0 0.5 0.5 0.1 WTO: Exports value 569.9 -1.0 566.0 4.6 2.1 5.9 0.4 Imports value 447.6 -1.4 443.1 5.9 4.3 9.1 0.8 Impacts Real GDP at factor costs 1,538.8 -0.1 1,537.8 0.0 0.0 0.0 0.1 Nominal GDP at factor costs 1,538.8 -0.1 1,537.8 0.8 0.4 1.2 0.0 Urban agricultural income 112.3 0.7 113.1 0.0 2.5 3.3 0.1 of Rural agricultural income 208.7 0.6 210.0 -0.2 2.5 3.0 0.5 the Urban nonagricultural income 182.5 0.3 183.0 1.4 0.4 1.8 0.6 Doha Rural nonagricultural income 112.3 -0.5 111.7 0.9 0.7 1.5 0.0 Urban unskilled wage income 223.3 -0.3 222.5 1.0 0.1 0.9 0.1 Development Rural unskilled wage income 43.6 -0.1 43.6 0.8 0.1 1.0 0.0 Urban skilled wage income 77.2 -0.2 77.0 1.0 0.3 1.3 0.0 Rural skilled wage income 67.0 -0.1 67.0 0.6 0.6 1.2 0.0 Nonlabor income 603.1 -0.2 601.8 1.0 -0.4 0.4 0.1 Agenda Source: Author's simulations. Note: Base and new base values in trillions of rupiah; percent changes for the nonbase columns. PRESIM=presimulation scenarios entailing China's WTO accession, the completion of the UR-ATC commitment to abolish textiles and apparel quotas, and the EU's enlargement to 25 members; ULIB=unilateral liberalization; FLIBX=full liberalization excluding Indonesia; Full-Lib=full liberalization including Indonesia; Doha=Doha Scenario. Table 11.4. Social Impact of Trade Liberalization Reference Scenarios Base PRESIM New base ULIB FLIBX Full-Lib Doha National level Per capita income 7,188.2 0.0 7,188.2 0.6 0.5 1.2 0.1 General entropy index (0) 35.7 -0.1 35.7 0.3 -0.1 -0.3 0.0 General entropy index (1) 48.4 -0.1 48.4 0.2 -0.1 -0.6 0.0 Gini index 45.7 -0.1 45.7 0.1 -0.1 -0.2 0.0 Poverty headcount 18.3 -0.2 18.3 -1.0 -2.0 -3.5 -0.1 Poverty gap 4.8 -0.2 4.8 -0.8 -1.6 -3.4 -0.1 Squared poverty gap 1.9 -0.2 1.9 -0.8 -1.3 -3.3 -0.1 Number of poor 39,253,480 -78,507 39,174,973 -374,680 -798,764 -1,384,164 -48,152 The Urban areas Per capita income 9,775.9 0.0 9,775.9 0.8 0.3 1.1 0.0 Social General entropy index (0) 38.5 -0.1 38.5 0.1 0.3 -0.4 0.1 General entropy index (1) 52.9 -0.1 52.8 0.0 0.3 -0.6 0.1 Impact Gini index 47.3 0.0 47.3 0.0 0.1 -0.2 0.0 Poverty headcount 14.5 -0.2 14.5 -2.2 -0.8 -3.4 0.0 Poverty gap 4.1 -0.1 4.1 -1.5 -0.1 -2.8 0.1 of Squared poverty gap 1.8 -0.1 1.8 -1.4 0.1 -2.9 0.1 a WTO Number of poor 13.322,340 -26,645 13,295,695 -287,148 -109,962 -454,021 0 Rural areas Agreement Per capita income 5,235.6 0.1 5,240.8 0.2 0.9 1.4 0.1 General entropy index (0) 25.0 -0.1 25.0 0.0 0.0 0.0 0.0 General entropy index (1) 30.3 -0.2 30.2 -0.2 0.1 -0.2 0.0 Gini index 38.4 -0.1 38.4 0.0 0.0 0.0 0.0 Poverty headcount 21.2 -0.2 21.2 -0.3 -2.7 -3.6 -0.2 in Poverty gap 5.3 -0.3 5.3 -0.4 -2.4 -3.8 -0.2 Indonesia Squared poverty gap 2.1 -0.3 2.1 -0.4 -2.1 -3.6 -0.2 Number of poor 25,931,138 -51,862 25,879,276 -87,530 -688,798 -930,140 -48,150 Source: Authors' simulations. Note: Base values percent changes for columns. Poverty indicators are computed using national poverty lines (World Bank 2003). PRESIM = presimulation scenarios 331 entailing China's WTO accession, the completion of the UR-ATC commitment to abolish textiles and apparel quotas, and the EU's enlargement to 25 members; ULIB = unilateral liberalization; FLIBX = full liberalization excluding Indonesia; Full-Lib = full liberalization including Indonesia; Doha = Doha scenario. 332 Poverty and the WTO: Impacts of the Doha Development Agenda and the third analyzes the combined impact of full liberalization including Indonesia (Full-Lib). Unilateral liberalization, whereby Indonesia cuts all the duties facing imports from the rest of the world, has some impact on nominal GDP6 and generates an increase in both exports and imports. At the sector level, there is some reallocation of factors out of textiles and wearing apparel and toward the paper printing, transport equipment, and metal products industries. Although the removal of the imports tariff hurts the sectors that benefited from higher relative protection levels, the total impact on household income is positive: as can be seen in table 11.4, per capita income increases by 0.6 percent at the national level, a number that is consistent with the increase in nominal GDP. Results suggest that unilateral liberalization would generate an increase in the average per capita household income of 0.6 percent, and the impact on the distribution of income would be negative but small. As a result of the average per capita income increase, poverty decreases modestly. The poverty headcount decreases by 1.0 percent at the national level (falling from 18.3 percent to 18.1 percent of the entire population), and the impact appears higher in the urban areas. Higher-order poverty indicators vary by the same magnitude, a result that indicates that the poorest of the poor also benefit from the unilateral liberalization, despite the slight worsening in the distribution of income. Overall, these changes translate into a total of 375,000 people escaping poverty. As a result of full trade liberalization in the rest of the world, the Indonesian economy faces decreasing import and export prices but an improvement in overall TOT: the export price index decreases by 0.6 percent, and the import price index decreases by 2.3 percent. Despite this average decrease, import prices for agricultural goods increase. The impact on nominal GDP at factor cost is positive, and both imports and exports increase by 4.3 and 2.1 percent, respectively. In terms of absorption components, both government consumption and investment are assumed to be fixed, and private consumption increases by 0.7 percent. At the sector level, changes are driven by the differential exposure to foreign trade and the TOT shocks, and some reallocation of factors occurs, mainly between manufacturing activities. These shocks translate into a 0.5 percent increase in per capita income at the household level. The increase is much smaller in the urban sector and is accompanied by a worsening in the income distribution. As a result, the poverty impact is small in terms of incidence, and both the poverty gap and squared poverty gap increase, suggesting that some of the poorest households fare badly.7 This result stems from the adverse impact on urban households of rising food prices. Results in the rural sector are quite different, with a 0.9 percent increase in the average per capita income and almost no change in inequality The Social Impact of a WTO Agreement in Indonesia 333 indicators. As a result, nearly 690,000 people are lifted out of poverty in the rural areas when the rest of the world liberalizes trade. The full liberalization scenario (Full-Lib) combines reforms in the rest of the world and Indonesia and generates more favorable aggregate results, with exports and imports increasing by 5.9 percent and 9.1 percent, respectively. Total employment is assumed to fixed, but the full liberalization entails some reallocation of labor toward self-employment for each labor category, namely a reallocation of labor out of formal sectors and toward agricultural sectors. This stems from the fact that import prices for nonagricultural goods decrease more than for agricultural products as a result of the full liberalization. As a consequence, there is an improvement in the distribution of income overall, as well as within urban and rural areas: Poverty falls, with nearly 1.4 million people escaping poverty, a number resulting from the drop in the incidence of poverty from 18.3 percent to 17.7 percent. The final scenario explored in tables 11.3 and 11.4 is the core Doha scenario. In the case of Indonesia, the impacts are very small--just a 0.1 percent impact on per capita consumption--and less than a 1 percent rise in aggregate imports and exports. There is a negligible impact on inequality, but, according to the model predictions, rising incomes boost about 50,000 people out of poverty. Examining Alternative Scenarios A number of alternative scenarios are examined in this section. They are aimed a exploring the importance of some of the assumptions regarding labor markets that are made in the reference simulations, as well as the choice of tax instrument for the replacement of tariff revenue. Tax Replacement Scheme In the reference simulation, direct taxes on household income are adjusted in an equiproportionate manner to compensate for the revenue loss due to the cut in import duties. This was done to permit comparability with other studies in this volume. This type of tax replacement scheme would be very efficient, but it would entail a major fiscal reform, which is unlikely to occur in the current political and practical context in Indonesia.8 Therefore, the analysis now turns to the impact of trade reform under an alternative tax replacement scheme whereby value added rates adjust to make up for the revenue loss.9 As mentioned above, import tariffs accounted for 6.0 percent of government income and 0.8 percent of GDP in 2002. The contribution of the VAT government 334 Poverty and the WTO: Impacts of the Doha Development Agenda revenue was six times higher, representing 36.5 percent of government income and 4.6 percent of GDP. The compensation for lost revenue through the adjustment of VAT results in a 17 percent increase in rates across nonexempt sectors. Tables 11.5 and 11.6 show that the increase in trade volume is comparable under the VAT replacement tax. However, the outcome in terms of efficiency gains is much smaller. In the urban sector, the lower per capita income gain is accompanied by a slight worsening of the distribution of income. The resulting poverty reduction amounts to 2.3 percent, with lesser rates of poverty reduction for the higher-order poverty indicators. Overall, 900,000 people are lifted out of poverty instead of the 1,400,000 with the direct tax replacement scheme. Table 11.5. Macroeconomic Results of Alternative Scenarios Variable Base FLIBVAT FLIB_2 FLIB_3 Private consumption 1,040.3 0.7 1.7 0.7 Investment demand 324.7 Government consumption 132.2 Total absorption 1,497.3 0.5 1.2 0.5 Exports value 566.0 5.8 6.5 6.0 Imports value 443.1 9.0 9.9 9.2 Real GDP at factor costs 1,537.8 0.0 0.7 0.0 Nominal GDP at factor costs 1,537.8 0.4 1.9 1.0 Urban agricultural income 113.1 3.3 5.0 1.8 Rural agricultural income 210.0 3.0 3.9 1.9 Urban nonagricultural income 183.0 1.4 2.1 2.3 Rural nonagricultural income 111.7 1.2 1.8 1.7 Urban unskilled wage income 222.5 0.0 1.3 0.8 Rural unskilled wage income 43.6 0.6 1.5 0.6 Urban skilled wage income 77.0 0.5 1.3 0.9 Rural skilled wage income 67.0 0.5 1.0 0.4 Nonlabor income 601.8 -0.9 1.6 0.4 Urban unskilled employment 15.0 1.5 Urban skilled employment 17.5 1.0 Rural unskilled employment 34.2 1.7 Rural skilled employment 9.8 1.1 Total employment 76.5 1.4 Source: Author's simulations. Note: Base values in trillions of rupiah, except employment outcomes in millions of workers; percent changes for the nonbase columns. FLIB_2 = full liberalization including Indonesia, with flexible unem- ployment; FLIB_3 = full liberalization including Indonesia, with sector-specific labor; FLIBVAT = full liberalization including Indonesia, with VAT rates adjustment as the tax replacement scheme. The Social Impact of a WTO Agreement in Indonesia 335 Labor Markets The results of the simulations examined thus far in the paper have rested on the assumption of fixed employment in all labor markets. That assumption led to modest changes in growth and welfare at the household level. In this section, the impact of the full liberalization scenario is examined with two alternative Table 11.6. Social Impact of Alternative Scenarios Base FLIBVAT FLIB_2 FLIB_3 National level Per capita income 7,188.2 0.8 1.7 1.0 General entropy index (0) 35.7 -0.1 0.1 0.3 General entropy index (1) 48.4 -0.1 -0.3 0.4 Gini index 45.7 -0.1 0.1 0.2 Poverty headcount 18.3 -2.3 -3.2 -2.4 Poverty gap 4.8 -2.3 -3.4 -2.1 Squared poverty gap 1.9 -2.0 -3.7 -2.2 Number of poor 39,174,973 -902,032 -1,236,696 -961,328 Urban areas Per capita income 9,775.9 0.6 1.4 1.1 General entropy index (0) 38.5 0.3 0.4 0.3 General entropy index (1) 52.8 0.3 0.0 0.4 Gini index 47.3 0.1 0.2 0.2 Poverty headcount 14.5 -1.3 -1.2 -2.5 Poverty gap 4.1 -0.5 -2.0 -2.1 Squared poverty gap 1.8 -0.5 -2.5 -2.2 Number of poor 13,295,695 -169,782 -164,989 -328,554 Rural areas Per capita income 5,240.8 1.2 2.0 0.8 General entropy index (0) 25.0 0.0 0.5 0.2 General entropy index (1) 30.2 0.0 0.0 0.1 Gini index 38.4 -0.1 0.3 0.1 Poverty headcount 21.2 -2.8 -4.1 -2.4 Poverty gap 5.3 -3.3 -4.3 -2.1 Squared poverty gap 2.1 -3.0 -4.4 -2.2 Number of poor 25,879,276 -732,250 -1,071,706 -632,774 Source: Author's simulations. Note: Base values in the first column and percent changes in the following columns. Poverty indicators are computed using national poverty lines (World Bank 2003). FLIB_2 = full liberalization including Indonesia, with flexible unemployment; FLIB_3 = full liberalization including Indonesia, with sector- specific labor; FLIB_VAT = full liberalization including Indonesia, with VAT rates adjustment as the tax replacement scheme. 336 Poverty and the WTO: Impacts of the Doha Development Agenda specifications of labor market functioning. In the first alternative closure, hourly wages are assumed to be fixed, with labor markets clearing through the adjustment of total employment (FLIB_2). This specification is expected to generate higher aggregate welfare effects as previously idle resources are brought into play. In a second specification, employment is assumed to be fixed--not only in the aggregate, but also by sector (FLIB_3). With less flexibility, this sector- specific labor scenario is expected to generate lower aggregate welfare gains. In the case of flexible unemployment, the growth impact is higher than under the full employment assumption. It is driven by an increase in employment of approximately 1.4 percent, ranging from 1.0 percent to 1.8 percent across labor categories. As a result, the employment changes fed into the MS model are bigger. This generates higher per capita income changes, but although the overall impact on distribution remains positive, it deteriorates in the urban areas. This leads to smaller changes in poverty in the urban areas, where the poverty headcount decreases by only 1.2 percent despite the higher per capita increase. As a consequence, the aggregate poverty reduction is somewhat smaller under this unemployment closure (-1,260,000) than under the reference scenario. How can it be that a scenario in which unemployment falls generates a smaller poverty reduction than one in which unemployment is fixed? The answer is that it all depends on who gets the jobs. If the jobs go to second or third earners in nonpoor households, then the income distribution can worsen, because the pool of unemployed keeps wages from rising and therefore mitigates the benefits to households for which the number of wage earners is fixed. Of course, the issue of who gets the new jobs is subject to considerable uncertainty, and this is reflected in the random draws for the error term associated with the occupational choice model. Therefore, there is clearly a need for Monte Carlo analysis, which will be discussed in the next section. Under the sector-specific labor assumption, there is no reallocation of labor across sectors and efficiency gains are smaller. The resulting poverty outcomes are also smaller and overall income distribution worsens as a result of a smaller improvement in per capita income in rural areas relative to urban areas. Monte Carlo Analysis Given the stochastic nature of the occupational choice model in the MS model, it makes sense to perform Monte Carlo experiments to examine the sensitivity of poverty and income distribution outcomes.10 Is it possible that the poverty outcome discussed above in the case of the unemployment closure is not robust? Monte Carlo experiments are performed on the full-liberalization scenario under the three alternative labor market closure specifications: the fixed employment The Social Impact of a WTO Agreement in Indonesia 337 closure (Full-Lib), the flexible employment closure (FLIB_2), and the sector- specific labor closure (FLIB_3). The Monte Carlo results presented in table 11.7 provide a much better idea of the robustness of the findings presented in tables 11.4 and 11.6. First, note that the magnitude of the standard deviation on the inequality indicators suggests that the changes in income distribution are not significantly different from zero in any simulation. Moreover, when the full range of possible outcomes in the occupational choice model is considered, the sign of difference in the poverty outcome between the fixed employment specification and the flexible employment closure is reversed. Under the flexible employment closure, the trade liberalization scenario generates higher poverty reductions. However, the size of the standard deviation on the poverty outcomes suggests that the difference in poverty outcomes under the reference scenario and the flexible unemployment closure may not be significant. Also, it is clear that the third case--that of fixed labor--does generate significantly smaller poverty reduction. Of course, any model is an abstraction of reality, and labor markets in Indonesia probably don't function precisely in the manner described under any of the three alternative specifications; rather, they are likely to reflect of combination of these polar views of the world. Therefore, these results should be viewed as providing a range of plausible poverty outcomes subject to the other assumptions embedded in the model. Summary and Conclusions The framework presented in this study has permitted us to generate detailed poverty outcomes resulting from international trade shocks. Given the magnitude of the shocks examined here and the structural features of the Indonesian economy, only the full liberalization scenarios generate significant poverty changes. Their impacts are examined under alternative specifications of the functioning of labor markets. These assumptions generate quite different results, but all conclude that full liberalization's impact on poverty would be positive, with efficiency and income gains dominating the adverse food price changes that could hurt the poorest households. Results also suggest that poverty reduction would be higher in the rural than in the urban sector. Two alternative tax replacement schemes are examined. Although the direct tax replacement appears to be more desirable in terms of efficiency gains and translates into higher poverty reduction, political and practical considerations could lead the government of Indonesia to choose a replacement scheme through the adjustment of VAT rates across nonexempt sectors. Such a move would dampen down the poverty-reducing potential of trade reform. 338 Table 11.7. Monte Carlo Simulations on the Social Impact of Alternative Labor Market Closures Standard Standard Standard Base Full-Lib deviation FLIB_2 deviation FLIB_3 deviation Poverty National level Per capita income 7,188.2 1.2 0.0 1.7 0.0 1.0 0.0 and General entropy index (0) 35.7 -0.3 0.6 -0.3 0.7 0.2 0.3 the General entropy index (1) 48.4 -0.9 1.8 -1.4 2.0 0.1 1.0 Gini index 45.7 -0.1 0.2 -0.1 0.3 0.1 0.1 WTO: Poverty headcount 18.3 -3.2 0.3 -4.0 0.5 -2.5 0.1 Poverty gap 4.8 -3.4 0.4 -3.9 0.6 -2.2 0.1 Impacts Squared poverty gap 1.9 -3.3 0.5 -3.8 0.9 -2.2 0.2 Urban areas Per capita income 9,775.9 1.1 0.1 1.4 0.1 1.1 0.0 of General entropy index (0) 38.5 -0.4 1.2 -0.6 1.4 0.1 0.7 the General entropy index (1) 52.8 -1.2 2.8 -1.8 3.2 -0.1 1.6 Doha Gini index 47.3 -0.2 0.5 -0.3 0.6 0.1 0.3 Poverty headcount 14.5 -2.5 0.9 -2.9 1.1 -2.6 0.3 Development Poverty gap 4.1 -2.6 0.7 -3.1 1.2 -2.2 0.3 Squared poverty gap 1.8 -2.6 0.9 -3.3 1.7 -2.3 0.4 Rural areas Per capita income 5,240.8 1.4 0.0 2.0 0.1 0.8 0.0 General entropy index (0) 25.0 0.1 0.2 0.5 0.4 0.2 0.0 Agenda General entropy index (1) 30.2 0.1 0.2 0.2 0.7 0.3 0.0 Gini index 38.4 0.0 0.1 0.2 0.2 0.1 0.0 Poverty headcount 21.2 -3.5 0.2 -4.5 0.6 -2.4 0.1 Poverty gap 5.3 -3.9 0.3 -4.4 0.7 -2.1 0.1 Squared poverty gap 2.1 -3.8 0.5 -4.1 1.1 -2.1 0.1 Number of experiments 98 98 99 Source: Authors' simulations. Note: Full-Lib = full liberalization including Indonesia, with fixed unemployment; FLIB_2 = full liberalization including Indonesia, with flexible unemployment; FLIB_3 = full liberalization including Indonesia, with sector-specific labor. The Social Impact of a WTO Agreement in Indonesia 339 As with any such study, there are a number of important limitations to this work. First of all, one should keep in mind that the results are likely to underestimate the cost of temporary unemployment endured by displaced workers in some sectors, particularly because unemployment benefits are not available for most Indonesian workers. Of course, these costs could be mitigated by gradually phasing in the trade reforms. And they would be further diminished if these trade reforms raised the overall growth rate of the Indonesian economy. Such dynamic growth gains--fueled by increased productivity and investment-- have been ignored here but are addressed in the final three chapters of this volume. Notes 1. The fully disaggregated SAM used has 39 activity accounts and 22 commodity accounts. Full detail is presented in the Policy Research Working Paper accompanying this chapter (Robilliard and Robinson, 2005). 2. The Special Survey on Saving and Household Investment (SKTIR) was integrated as a part of a module (submodule) of the SUSENAS survey. It was administered to only a subsample of the SUSE- NAS sample. 3. More details of the shocks fed into the CGE model are given in annex B of the World Bank Pol- icy Research Working Paper version of this chapter (Robilliard and Robinson 2005). Because the Indonesia model assumes export prices are exogenous and fixed, the approach used here diverge from the approach outlined in chapter 3 and simply shock export prices. This specification gives this model a zero optimal tariff, unlike that implicit in the GTAP simulations. 4. A total of 32 variables generated by the CGE model are fed into the microsimulation module. 5. The use of official poverty lines gives a much lower incidence of poverty than the US$2 per day mark and less scope for change. 6. Nominal GDP is computed with respect to the consumer price index (CPI). Because the CPI is the appropriate numeraire for the analysis conducted at the household level, nominal GDP is the macro aggregate that is consistent with the changes in per capita income at the household level. Note that poverty and inequality indicators in the MS module are based on total per capita earned income, with no deduction of direct taxes. The issue of tax replacement will be debated in subsequent simula- tions. 7. The poverty gap measures the distance between the average poor household income per capita and the poverty line; the squared poverty gap gives a measure of the distribution of income among poor households. 8. Personal tax rates range from 5 percent for the lowest bracket to 35 percent for the highest. All rates except the highest were lowered in 2001. 9. VAT rates are relatively homogeneous across nonexempt sectors. It is assumed that informal sec- tors are not subjected to VAT because of the difficulty of collecting these taxes at the level of informal production units. 10. The term "Monte Carlo experiments" refers here to the replication of MS results using different draws of the residuals for the occupational choice model as well as for the wage equation model. One hundred draws were performed for each simulation. The draws that did not generate a feasible solu- tion where dropped, which explains why the number of observations is smaller than 100 in table 11.7. 340 Poverty and the WTO: Impacts of the Doha Development Agenda References Armington, P. S. 1969. "A Theory of Demand for Products Distinguished by Place of Production." IMF Staff Papers 16 (2): 179­201. Bourguignon, F., A. S. Robilliard, and S. Robinson. 2003."Representative versus Real Households in the Macro-Economic Modeling of Inequality" In Frontiers in Applied General Equilibrium Modeling: Essays in Honor of Herbert Scarf, ed. T. Kehoe, T. N. Srinivasan, and J. Whalley, 219­254. Cambridge, U.K.: Cambridge University Press. Cogneau, D., M. Grimm, and A. S. Robilliard. 2003. "Evaluating Poverty Reduction Policies: The Contribution of Microsimulation Techniques." In The New International Strategies for Poverty Reduction, ed. J.P. Cling, M. Razafindrakoto, and F. Roubaud, 340­393. London: Routledge. Dervis, K., J. de Melo, and S. Robinson. 1982. General Equilibrium Models for Development Policy. New York: Cambridge University Press. Essama-Nssah, B. 2004. "Building and Running General Equilibrium Models in EViews." Policy Research Working Paper 3197. World Bank, Washington, DC. Francois, J., and D. Spinanger. 2004. "ATC Export Tax Equivalents." GTAP Version 6. https://www.gtap.agecon.purdue.edu/databases/v6/v6_doco.asp. Hertel, T., D. Hummels, M. Ivanic, and R. Keeney. 2004. "How Confident Can We Be in CGE-Based Assessments of Free Trade Agreements?" GTAP Working Paper 26. Purdue University, W. Lafayette, IN. McFadden, D. 1974. "Conditional Logit Analysis of Qualitative Choice Behavior." In Frontiers in Econometrics, ed. P. Zarembka, 105­42. New York: Academic Press. Robilliard, A. S., and S. Robinson. 2003. "Reconciling Household Surveys and National Accounts Data Using a Cross Entropy Estimation Method." Review of Income and Wealth 49 (3): 395­406. Robilliard, A. S., and S. Robinson. 2005. "The Social Impact of a WTO Agreement on Indonesia." Forthcoming Policy Research Working Paper, World Bank, Washington, DC. Robinson, S., A. Cattaneo, and M. El-Said. 2001."Updating and Estimating a Social Accounting Matrix Using Cross Entropy Methods." Economic Systems Research 13 (1): 47­64. WTO (World Trade Organization). 2003. Trade Policy Review--Indonesia 2003. Geneva. World Bank. 2003. Indonesia Development Policy Report: Beyond Macroeconomic Stability. Report 27374-IND. Washington, DC. Part V fiscal replacement of lost tariff revenue 12 The Poverty Impacts of the Doha Round in Cameroon: The Role of Tax Policy Christian Arnault Emini, John Cockburn, and Bernard Decaluwé* Summary This chapter assesses the possible impacts of the Doha Round on poverty in Cameroon. During the recent period of economic recovery, Cameroon has enjoyed a sharp decline in poverty, with the headcount index falling from 53.3 percent of inhabitants in 1996 to 40.2 percent in 2001, mostly thanks to economic growth rather than redistribution. Will the current trade negotiations under the Doha Round reinforce or curb this trend? The chapter applies a computable gen- eral equilibrium (CGE) microsimulation (MS) model that involves 10,992 house- holds to address this question. * The authors are grateful to Véronique Robichaud for her technical assistance, and to Nabil Annabi, Thomas Hertel, Alan Winters, Patrick Osakwe, and the participants at the conference, "Putting Devel- opment Back into the Doha Agenda," held in The Hague, December 2­4, 2004, for their valuable com- ments. The authors remain, of course, responsible for all possible errors and omissions. This work was carried out with financial assistance from the World Bank­Netherlands Partnership Program, as well as the Poverty and Economic Policy Research Network, which is financed by the International Devel- opment Research Centre. 343 344 Poverty and the WTO: Impacts of the Doha Development Agenda The Doha Round is found to be poverty reducing for Cameroon. For the whole country, the estimate of net number of people who are lifted out of poverty is 22,000, according to this scenario. Further investigations indicate that more ambi- tious world trade liberalization leads to greater poverty alleviation at the national level, and Cameroon's domestic trade liberalization has adverse poverty and inequality impacts, despite giving rise to higher aggregate welfare. Under the Doha scenario, the cuts in Cameroon's tariffs are very small (the average tariff rate moves from 11.79 percent in the base run to