Global 26759 Economic Prospects Realizing the Development Promise of the Doha Agenda 2004 Global Economic Prospects Realizing the Development Promise of the Doha Agenda 2004 © 2003 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 04 03 This volume is a product of the staff of the World Bank. The findings, interpretations, and conclusions expressed herein do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of the World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. ISBN 0-8213-5582-1 ISSN 1014-8906 Cover photo credit: AFP/CORBIS. Workers at the Los Ausoles coffee plantation in Ahuachapan, El Salvador, clean coffee beans, August 14, 2002. Contents Foreword ix Acknowledgments xi Overview xiii Abbreviations and Data Notes xxxi Chapter 1 Global Outlook and the Developing Countries 1 The industrial countries: Deficits, confidence, capital spending, and the dollar 4 The external environment for developing countries: Gradual improvement, but a bumpy road ahead 19 The developing countries: Back on track toward growth? 28 Trade, growth, and poverty in developing countries 38 Looking ahead to the Doha Round 47 Annex 1 Historical trade dynamics for developing countries 55 Notes 59 References 60 Chapter 2 Trade Patterns and Policies: Doha Options to Promote Development 63 Changing patterns in developing-country exports 65 Behind the patterns: Economic and policy determinants 73 Market access for development: The agenda 78 From Doha to Cancún and beyond: How should protection be reduced? 88 Notes 98 References 98 Chapter 3 Agricultural Policies and Trade 103 Poverty, rural households, and trade in agriculture 105 Trade and export growth in agriculture 109 Global agricultural protection: The bias against development 114 Proposals for reforms in the Doha Round 131 Notes 138 References 139 iii G L O B A L E C O N O M I C P R O S P E C T S Chapter 4 Labor Mobility and the WTO: Liberalizing Temporary Movement 143 The bigger picture: Global migration and remittance trends 145 Temporary movement of workers 150 Bilateral and regional approaches to labor mobility 152 Understanding the impact of temporary foreign workers 155 Mode 4 and the WTO 166 Notes 172 References 174 Chapter 5 Reducing Trading Costs in a New Era of Security 179 Why transport, trade facilitation, and logistics matter 181 The new international security dimension in trade 182 The anticompetitive effects of international transport regulations 188 Trade facilitation 191 Trade facilitation and the WTO agenda 195 Lowering transport costs, increasing security, and facilitating trade 198 Notes 200 References 202 Chapter 6 Development and the Doha Agenda 205 Special and differential treatment and the WTO 207 Market access for development 208 Toward a new regime for WTO rules 220 Putting development into the Doha agenda 227 Notes 228 References 229 Appendix 1 Regional Economic Prospects 233 Appendix 2 Global Commodity Price Prospects 257 Appendix 3 Global Economic Indicators 279 Figures 1.1 Growth in the OECD countries falters 4 1.2 OECD manufacturing shows a distinct “double dip” 5 1.3 Consumer confidence recovers from pre-war lows 6 1.4 The drop in U.S. household net worth has been offset by real estate appreciation 8 1.5 Capital spending has been hesitant in all industrial countries 9 1.6 Corporate profits have risen moderately in the United States and Japan 9 1.7 Business confidence remains poor, but better in the United States than in Europe 10 1.8 The U.S. fiscal deficit is widening quickly 11 1.9a The U.S. current account deficit is at record levels 11 1.9b The U.S. current account deficit is at record levels 13 1.10 Market interest rates have dropped 14 1.11 Is deflation a danger for Europe and the United States? 15 iv C O N T E N T S 1.12 Output gaps are widening, bringing deflationary pressures to bear 16 1.13 The dollar has fallen sharply since early 2002 16 1.14 OECD recovery begins in the United States 18 1.15 OECD-area imports have declined sharply since April 2000 21 1.16 China’s share of East Asian exports keeps rising 21 1.17 The price of oil fell sharply before the war in Iraq 23 1.18 Agricultural prices have begun to decline as crop prospects improve 25 1.19 Emerging-market spreads rallied sharply after late 2002 26 1.20 Bond issuance dominates capital market flows in 2003 27 1.21a Regional trends in industrial production are mixed 29 121b Inflation is moderating in the developing world 29 1.21c Major currencies in Latin America and East Asia are firming up 30 1.22 Developing countries are on track toward long-term growth 31 1.23 Growth rates in developing countries will rise through 2005 31 1.24 Before the SARS outbreak, East Asian GDP was growing robustly 32 1.25 Argentina, Brazil, and Chile see strong upturn in production 33 1.26 Growth will cool in CIS while picking up in Central and Eastern Europe 34 1.27 Middle East oil production has increased to prevent shortages 35 1.28 Indian production of food and automobiles recovered sharply in early 2003 37 1.29 Growth in Africa is expected to improve modestly 38 1.30 Income elasticity has risen globally, but particularly in the developing world 40 1.31 Export-to-GDP ratios have risen sharply in developing countries 41 1.32 Productivity will contribute more to GDP growth through 2015 than will capital or labor 44 1.33 The pro-poor reform scenario promises substantial income gains 50 1.34 Exports should rise sharply 52 1.35 Millions of people would be moved out of poverty 52 1.36 Gains for most, but adjustment costs for some 53 1.37 Significant shifts in global output patterns 54 2.1 Developing countries have become important exporters of manufactured products 65 2.2 Manufactures account for a growing share of exports in all regions 67 2.3 Technology-laden manufactures have increased as a share of exports from each group of countries, while the share of resource-based exports has diminished 70 2.4 Global production sharing is increasingly important for China and India 71 2.5 Soaring exports from China and India had only a moderate effect on China’s and India’s terms of trade 72 2.6 Many developing countries face an adjustment when quotas are lifted 80 2.7 Antidumping barriers by sector and by country group 88 3.1 Countries that produce more cash crops also produce more food 109 3.2 Import growth rates of nontraditional export commodities decreased in industrial countries but increased in developing countries 112 3.3 Developing countries’ exports of nontraditional products have surged, but industrial countries’ exports have changed little 114 3.4 Developing countries lowered tariffs on manufactured products more than on agricultural products 119 v G L O B A L E C O N O M I C P R O S P E C T S 3.5 Rich countries use non–ad valorem tariffs more often than do developing countries 122 3.6 Throughout the world, tariff rates escalate with degree of processing 123 3.7 The proportion of tariff lines containing non–ad valorem duties increases with degree of processing 125 3.8 Tariff rate quotas protect a substantial portion of output in many industrial countries 126 3.9 High protection of sugar and wheat has increased domestic production and reduced net imports 128 4.1 Workers’ remittances are an important source of income for many developing countries 149 5.1 Customs clearance takes longer in the developing world than in the OECD, lowering the competitiveness of developing-country trade 185 5.2 Higher trade costs reduce global welfare 186 5.3 Facilitating trade in less-efficient countries would bring significant gains 194 5.4 The impact of individual trade-facilitation measures differs significantly from region to region 195 5.5 Domestic reforms alone would produce many of the same gains as global reform 196 6.1 The benefits of U.S. trade preferences are distributed unequally 211 6.2 Countries “graduating” from U.S. generalized system of preferences have better export performance than those still in program 212 6.3 Preferences have not increased the share of the least developed countries in imports into the European Union and the United States 215 6.4 Market shares of countries eligible for three U.S. “deep preference” programs have not increased 215 6.5 Preferred countries’ apparel exports to the United States have risen 216 6.6 Agricultural exports from Mexico and Spain rose dramatically after the two countries joined regional trade blocs 217 6.7 The trade policies of countries in the U.S. generalized system of preferences are more protectionist than those of countries not in the program 218 6.8 Countries enjoying preferences have increased their exports of apparel to the United States 219 Tables 1.1 Global growth should accelerate, but risks persist 3 1.2 Weak fundamentals underlie sluggish growth in the rich countries 5 1.3 The difficult environment for developing-country growth should improve 20 1.4 Developing countries’ exports will grow faster than those of the high-income countries 22 1.5 GDP per capita will grow faster in the developing world than in the OECD area 43 1.6 Global poverty will decrease significantly, but not uniformly across regions 46 1.7 Tariffs could be cut clearly and simply 48 1.8 The pro-poor tariff scenario would significantly lower protection 49 1.9 A large share of real income gains comes from lowering of barriers in agriculture and food 51 1.A1 Sectoral export decomposition for developing countries 55 vi C O N T E N T S 1.A2 Regional export decomposition for developing countries 57 2.1 Developing countries are becoming exporters of high-value products 68 2.2 Developing countries’ exports became more competitive in the 1990s 74 2.3 Investment in people and in capital grew rapidly 75 2.4 Tariffs hurt exports—but less so in the 1990s than in the 1980s 77 2.5 Quota abolition in China will move resources from other activities to textiles and clothing 81 2.6 Industrial countries levy higher tariffs on imports from developing countries than from other industrial countries—and some regions have high tariff walls 82 2.7 Developing countries pay large amounts in tariffs to their neighbors 83 2.8 Most antidumping actions are filed by developing countries against other developing countries 86 2.9 Antidumping rates are much higher than tariff rates 86 2.10 Antidumping duties are high 87 2.11 Competing formulas make a big difference for tariffs 95 2A.1 The various liberalization proposals have very different features 97 3.1 Most of the world’s poor live in rural areas outside the least developed countries 106 3.2 Rural poverty is higher in poorer countries 107 3.3 Even in subsistence economies, cash is important 107 3.4 U.S. farmers earn less from farming than from other sources 107 3.5 Manufacturing exports grew much faster than agricultural exports 110 3.6 South-South exports in agriculture are rising as South-North export shares fall 110 3.7 Developing countries have shared unequally in export market gains 117 3.8 Agricultural tariffs are higher than manufacturing tariffs in both rich and poor countries 118 3.9 Agricultural tariffs: High peaks and deep valleys 119 3.10 Most subsidies go to producers—and come from border protection 120 3.11 Subsidies account for a large share of farmers’ revenues 121 3.12 Specific tariffs are higher than ad valorem rates 123 3.13 Tariffs rise with level of processing 124 3.14 The Harbinson proposals could greatly reduce applied tariffs in the European Union and the United States 133 3.15 The Harbinson proposals would not significantly reduce protection in the developing world—if reductions were taken from bound rates 133 3.16 U.S. trade preferences—a plethora of programs 136 4.1 Migration is rising in many OECD countries 147 4.2 Workers’ remittances are the second-largest source of external funding for developing countries 148 4.3 Remittances are a significant source of income in all regions of the developing world 149 4.4 Temporary movement is rising in rich countries 150 4.5 Foreign-born workers meet skill shortages in rich countries 152 4.6 The distribution of costs and benefits associated with Mode 4 trade 157 4.7 TMNP is the smallest of the four modes of international service supply 168 4.8 Most Mode 4 commitments by WTO members are in management categories 169 vii G L O B A L E C O N O M I C P R O S P E C T S 5.1 Elimination of anticompetitive private practices can cut costs drastically 190 6.1 Developing countries rarely receive significant preferences in sectors in which they would have a comparative advantage 209 6.2 Utilization rates for preference-eligible products with high MFN tariffs are low 210 6.3 Actual use of preference programs is declining 211 Boxes 1.1 Consumer confidence and U.S. private consumption 7 1.2 Financing the U.S. current account deficit: From equity to debt 12 1.3 OPEC struggles to achieve higher prices amid growing supply competition 24 1.4 Economic effects of Severe Acute Respiratory Syndrome (SARS) 33 1.5 AIDS is taking a rising toll in Sub-Saharan Africa 39 2.1 Poor export performance in 43 countries 69 2.2 Swimming upstream: The case of Vietnamese catfish 85 2.3 The scourge of the specific 89 2.4 “Average cuts,” the cut you have when you’re not having a cut 92 2.5 The implications of five tariff-cutting proposals 93 3.1 The impact of national trade integration and reform on poverty 106 3.2 Did agricultural exports slow down solely because of falling prices? 111 3.3 Decomposing export growth in manufacturing 113 3.4 Food safety standards: From barriers to opportunities 115 3.5 Decoupling agricultural support from production decisions 127 3.6 Fewer subsidies, stronger agricultural sector 132 3.7 The potential impact of real preferences 134 3.8 Rules of origin in preferential schemes are complicated—and often contradictory 136 3.9 Food aid principles 137 4.1 Population aging and migration 146 4.2 Temporary labor movement and the East Asian crisis of 1997–98 151 4.3 Recent initiatives to facilitate temporary movement of highly skilled workers 153 4.4 A trade facilitation approach to labor mobility: NAFTA and APEC 154 4.5 Initiatives to encourage return migration 160 4.6 Wages and conditions 163 4.7 E-commerce and temporary movement 164 4.8 Boosting intra-EU labor mobility 165 4.9 Measuring Mode 4 is still imprecise 167 4.10 Key impediments to Mode 4 trade 169 4.11 Elements of a possible GATS visa/permit regime 171 5.1 The evolving definition of trade facilitation 181 5.2 The logistics needs of a German car part manufacturer in Tunisia 192 5.3 Tackling corruption in customs: Peru 197 5.4 Customs reform in Lebanon 198 6.1 EU and U.S. preference programs 213 6.2 Major WTO provisions allowing developing countries greater freedom to use restrictive trade policies 221 6.3 A “development box” for the Agreement on Agriculture? 223 viii Foreword T he international community finds itself at a crossroads as it goes into the last quar- ter of 2003. Will the Doha Agenda regenerate the multilateral consensus that has been the hallmark of successive rounds of trade liberalization since 1947 and in doing so provide new impetus for global integration? Or will the Doha Agenda collapse in stalemate and perhaps be viewed as the moment when the international community retreated from multilateralism and opened the floodgates for less desirable bilateral and regional arrangements? The answers to these questions matter a great deal to the world’s poor. The round of trade talks launched in November 2001 in Doha, Qatar, is the first negotiation focused primarily on is- sues of concern to developing countries, and the first trade round since the birth of the World Trade Organization (WTO). Moreover, the Doha round is the first trade round for many new WTO members, including the world largest developing economy, China. Consequently, the round has the opportunity to remove many of the inequities in the global trading system that put de- veloping countries—and poor people in particular—at a disadvantage in their trade. Three trade barriers are of particular concern. Poor people work in agriculture, and agricul- tural products are subject to the highest barriers to trade. In addition, poor people produce labor- intensive manufactures, which are subject to peak tariffs in a world that has already reduced average tariffs in manufactures to historic lows. Poor people could benefit from greater tempo- rary migration. Governments everywhere have worked hard to create the opportunity to reduce these and other barriers. And they will have to work hard to capitalize on that opportunity. To fulfill the development promise of the Doha Agenda, rich countries will have to reduce protection of their (relatively wealthy) farmers. Their tariff walls and huge subsidies depress global prices of the products that poor farmers produce throughout the developing world. These subsidies cost the average working family in the European Union, Japan, and the United States more than $1,000 a year. Middle-income countries, though their protection is generally lower and less distorting in agriculture, have high average tariffs in all sectors, and are more restrictive in services. As south- south trade increases in importance, protection of sectors in middle-income countries undermines their poorer trading partners and often undercuts the countries’ own productivity growth. Fi- nally, low-income countries should look to the international system to meet their very reasonable demands—not for special preferences to some markets and exemptions from rules, but for nondiscriminatory market access to every market in products in which they have a comparative advantage, for appropriately phased introduction of international regulations, and for develop- ment assistance in implementing administratively costly WTO rules. Like other countries, low- ix G L O B A L E C O N O M I C P R O S P E C T S income countries will find it in their interest to reduce their own external levels of protection as part of an integrated development strategy aimed at reducing poverty. Reducing barriers to trade is not enough to fulfill the development promise of Doha. Trade must be part of a larger development strategy for each country, a strategy that includes attention to macroeconomic policy, infrastructure, education, and health as well as to accountable and responsible governance. These elements of investment climate take time to develop but are es- sential for growth and poverty reduction and are crucial to make a sound trade strategy pay its growth and poverty reduction dividends. The World Bank, working in partnership with the other international institutions and bilateral donors, is committed to supporting a pro-poor Doha outcome. Our objectives in trade are two- fold: promoting a world trading system in which global, regional, and bilateral rules are con- ducive to development and poverty reduction, and helping individual developing countries lever- age trade to promote their own growth. The latter objective hinges on integrating appropriately sequenced trade reforms into national development and poverty reduction strategies. The Bank is increasing its investment in research, technical assistance, and lending for trade. A casual perusal of the bibliography in each chapter of this report will give the interested reader an idea of the scope of the Bank’s research program. Moreover, in the last two years, the Bank has undertaken at the request of governments more than 20 diagnostic studies of obstacles to trade integration. In conjunction with six partner institutions, the Bank has led the Integrated Framework program—studies of trade obstacles in a dozen least-developed countries to date. It has completed several regional studies of trade. In addition to studies and policy advice, the Bank has provided technical assistance in the form of lending to improve trade-related institutions and transport logistics. The Bank has programs that finance activities in 49 countries (approximately one-third of its active client countries). These projects span all regions and range from export competitiveness projects in Ghana and Bangladesh, to transport and trade facilitation projects in Eastern Europe, to support for im- proving customs–border control agencies and training the trading community in Pakistan. The Bank is also implementing projects to improve quality standards and is leading the “Standards and Trade Development Facility,” an interagency partnership with the WTO, the FAO, and the World Health Organization, to deliver technical assistance for food safety and related standards. Should trade ministers reach an agreement on the Doha Agenda, the Bank will expand its lend- ing and technical assistance to help countries take advantage of new market access, to use trade to promote their domestic competitiveness, and to manage any transitional costs—such as those arising from erosion of trade preferences, changes in prices of imports, or reallocation of domes- tic resources from inefficient sectors to more efficient ones. A pro-poor outcome in the Doha Agenda is only one step toward a world more supportive of development. But this step is an important one. And it can be achieved only if everyone under- stands what is at stake in this historical moment—and moves purposefully and together to seize the opportunity. Nicholas Stern Chief Economist and Senior Vice President World Bank x Acknowledgments T HIS REPORT WAS prepared by the World Bank Development Prospects Group, drawing on re- sources throughout the Development Economics Vice Presidency and the World Bank’s op- erational units. Richard Newfarmer was the lead author and manager of the report, under the direction of Uri Dadush. The principal chapter authors were Richard Newfarmer (Overview); Elliot Riordan and Dominique Van der Mensbrugghe (Chapter 1); William Martin and Vlad Manole (Chapter 2); Ataman Aksoy (Chapter 3); Pierre Sauvé, drawing on work by the OECD (Chapter 4); John Wilson, Shweta Bagai, and Carsten Fink (Chapter 5); and Bernard Hoekman and Caglar Ozden (Chapter 6). We are grateful for the ideas and insights of several peer reviewers who provided comments at various stages: Bijit Bora (World Trade Organization); J. Michael Fin- ger (American Enterprise Institute); Gary Hufbauer (Institute for International Economics); Mari Pangestu (Center for Strategic and International Studies), Gary Horlick (Wilmer, Cutler, and Pick- ering), and Julia Nielson (OECD); Julio Nogues (United Nations Development Programme); and Olivier Cattaneo (Agence Française de Développement). The report was prepared under the gen- eral guidance of World Bank Chief Economist and Senior Vice President Nicholas Stern. Many staff from inside and outside the World Bank contributed to the report. In the Overview, Aart Kraay contributed a note on trade and poverty, and Carsten Fink, Bernard Hoekman, William Martin, and Aaditya Mattoo provided helpful suggestions. In Chapter 1, Hans Timmer, Caroline Farah, Himmat Kalsi, Robert Keyfitz, Annette I. De Kleine, Robert Lynn, Fernando Martel Garcia, Mick Riordan, and Bert Wolfe contributed to the global trends analysis; Do- minique Van der Mensbrugghe provided the long-term analysis; Shaohua Chen and Martin Ravallion contributed to the poverty analysis; and Katherine Rollins was the staff assistant. Chapter 2 benefited from background papers and other inputs from J. Michael Finger and Andri Zlate. For Chapter 3, John Beghin, Donald Mitchell, John Baffes, Harry De Gorter, Ndiame Diop, Paul Brenton, Steve Jaffee, and Mirvat Sewadeh provided background papers, and Baris Sivri, Tarek Soueid, Konstantin Senyut, and Gaston Gohou undertook data collection and analy- sis. Chapter 4 drew on research papers prepared by the OECD Trade Directorate and on the an- nual OECD report Trends in International Migration; the chapter reflects insights from Jeffrey Lewis, Julia Nielson (OECD), and Olivier Cattaneo (AFD). Tsunehiro Otsuki and Katherine Mann (IIE) worked closely with the team on Chapter 5, and Ranga Rajan Krishnamani provided research assistance. Chapter 6 draws on research by Bernard Hoekman, Constantine Michalopoulos, and L. Alan Winters. The regional annexes benefited from the written input of regional chief economists around the Bank and their staff, particularly Milan Brahmbhatt. John Baffes, Betty Dow, Donald Mitchell, and Shane Streifel prepared the commodity annex. The staff assistant for the report was Awatif Abuzeid. Steven Kennedy provided editorial assistance. Denis xi G L O B A L E C O N O M I C P R O S P E C T S Medvedev provided research assistance. Dorota Nowak coordinated the report’s publication and dissemination activities, working closely with the World Bank’s Office of the Publisher. Several experts provided written comments that have immeasurably improved the quality of the report at various stages: Paul Brenton, Robin Carruthers, Jean-Jacques Dethier, Shahrokh Fardoust, Coralie Gevers, Ian Goldin, Gary Horlick, Elena Ianchovichina, Mark Juhel, Hans Peter Lankes, Jeffrey Lewis, Patrick Low (WTO), Kunio Mikuriya (World Customs Organiza- tion), John Nash, David Rosenblatt, John Panzer, Luiz Perreira da Silva, Byungdoo Sohn, Mark Sundberg, Helena Tang, Yvonne Tsikata, and L. Alan Winters. xii Overview O N THE EVE of the World Trade Organi- dean countries, continues to weigh down re- zation’s (WTO) Fifth Ministerial Meet- gional performance. Africa, suffering from low ing in Cancún in September 2003, the commodity prices, is growing slowly; although world’s trade ministers—and the governments faster than in the 1980s and 1990s, today’s they represent—face enormous challenges. The growth is far short of the pace necessary to global trade talks are stalled in several policy make significant dents in the poverty head- domains vital to developing countries—agricul- count or to achieve the Millennium Develop- ture, nonfarm trade, access to patented drugs ment Goals in health and education. War has for countries without domestic drug industries, adversely affected regional performance in the special and differential treatment, and dispute Middle East and North Africa; sluggish per- settlement. Nor is there much progress in other formance in Europe, especially Germany, has contentious areas, such as the “Singapore is- adversely affected many countries in Central sues” of investment, competition, trade facilita- and Eastern Europe. Even though progress on tion, and government procurement. trade would undoubtedly boost investor confi- At the same time, the global recovery con- dence, politicians coping with slow growth tinues to sputter. Although some signs of a and high unemployment at home have been turnaround have been evident in the United finding it more difficult to risk alienating in- States, Europe seems to be losing momentum, fluential constituencies by accepting bold pro- and Japan appears positioned for another dis- posals in the world trade talks. appointing year. The Chinese economy, rein- The outlook for the remainder of this year forced by a positive performance in East Asia and for 2004, though somewhat improved, is in 2002, continues to bustle along, but con- unlikely to produce growth strong enough to cerns over Severe Acute Respiratory Syndrome cut sharply into unemployment rates (figure 1). (SARS) and lost export momentum in the face Uncertainty in the global environment remains of the world slowdown haunted the regional unusually high. Structural problems persist— outlook. South Asia continues to grow more overcapacity in high-tech industries globally, rapidly than the world average. Latin America rising twin deficits in the U.S. fiscal and cur- is showing signs of an upturn, driven in part by rent accounts, and lingering bad loans in renewed confidence in Brazil, a tentative re- Japanese and (to a lesser extent) European bound in Argentina, and an increase in Mex- banks. Other problems may prove more tran- ico’s growth; however, the recession in the sitory. The cessation of conflict in Iraq has not República Bolivariana de Venezuela, when yet produced complete calm, and the inability coupled with political difficulties in the An- to reach consensus at the UN Security Council xiii G L O B A L E C O N O M I C P R O S P E C T S in 1994, that tentative steps toward freeing up Figure 1 The recovery is building . . . trade in products of particular interest to de- but slowly veloping countries—notably agriculture and GDP growth, percent per annum textiles—were included. Consequently, many Forecast 5 of the hardest issues for rich countries have Developing been left to this negotiation. economies 4 ᮡ Realizing the development 3 promise of the Doha agenda T he challenge is daunting. But so is the re- High-income economies 2 ᮡ ward to success. With room for addi- tional fiscal and monetary stimulus rapidly 1 vanishing, progress on structural reforms such as trade is important. In addition to bolstering investor confidence in the short term, a Doha 0 2000 2001 2002 2003 2004 2005 Round agreement that slashed trade barriers, particularly in agriculture, would stimulate Source: World Bank data and projections. trade and raise incomes around the world, leading to a substantial reduction in global poverty. has created a lingering distrust among multi- The open question is whether a new multi- lateral partners that clouds the global business lateral agreement will live up to the develop- environment. Nonetheless, policy responses ment promise of the Doha Agenda. Several are promising. Governments in the United issues under discussion are pivotal to develop- States and Europe reacted to weak economic ment outcomes. They are the focus of this conditions with fiscal and monetary policy to report: stimulate their economies. And at the global political level, the June meeting of the G-8, to- • Because most poor people live in rural areas, gether with several subsequent bilateral meet- trade barriers in agriculture are among the ings, began to mend frayed multilateral rela- most important to poverty reduction. tions. It remains to be seen whether this new • Labor-intensive manufactures have been the positive momentum will extend into multilat- most dynamic market segment for every eral collaboration in trade. major region, including Africa, yet many The precarious international environment is developing countries find that their exports only one reason why the global trade talks meet obstacles in foreign markets—high have progressed slowly. Deeper explanations tariffs, quotas, specific duties, and “anti- can also be found in the history of multilateral development” tariff structures that discour- trade talks themselves. With the incorporation age adding value in poor countries. of ever more countries—mainly from the de- • In services, the potential for development- veloping world—the sheer number of actors promoting reciprocal gains is especially high. has expanded, making coalitions more difficult Regulations in some developing countries to build and consensus more elusive. More- still protect some inefficient state monopo- over, previous multilateral rounds produced lies from competition—a drag on growth. agreements in areas of primary interest to the (To be sure, proper regulation in some sec- rich countries that dominated these discus- tors must precede liberalization to avoid po- sions, particularly in manufactured goods. It tential disruptions in socially important mar- was only with the Uruguay Round, concluded kets, such as finance or basic services.) Also, xiv O V E R V I E W access for developing countries’ services ex- veloping countries, and low-income countries ports to industrial countries has yet to be alike. Rich countries account for two-thirds of fully bound in the General Agreement on world trade and comprise nearly three-quarters Trade in Services (GATS) (World Bank of world GDP, so their domestic policies—most 2001). Finally, national laws prevent greater evident in agriculture—have the greatest effect labor mobility that would otherwise con- on the global marketplace. Despite the fact that tribute to higher standards of living in both agricultural protection, tariff peaks, and anti- receiving and sending countries. dumping measures shield powerful lobbies, • Reducing the costs of trading by improving rich-country leadership in reducing this protec- international transportation services, cus- tion is a prerequisite for a pro-poor develop- toms and ports, and logistics management— ment outcome. trade facilitation—requires substantial new Today’s middle-income developing coun- investment, additional technical assistance, tries have increased their global market share and coordinated multilateral efforts. Trade in the last two decades. Because they include facilitation is fundamental to realizing the many of the most dynamic global economies, expanded trade promise of Doha, but the their domestic policies no longer have only WTO agenda constitutes a small part of minor consequences for trade. With protection the challenge. rates in manufactures three times the level of • Finally, the issue of special treatment for de- those in rich countries and with ubiquitous re- veloping countries cuts across all of these strictions on services, the middle-income coun- policy domains and affects trade preferences tries have ample scope for undertaking reduc- and exemptions from WTO regulations. tions in protection that will accelerate their The pursuit of trade preferences and exemp- growth and provide access and a growth im- tions from multilateral rules have not al- pulse to neighboring countries. High protec- ways served developing countries partic- tion in these countries taxes their growth and ularly well, both because preferences their poor in much the same way as protection have not proven reliable and because selec- in the North. tive coverage has often left productivity- Low-income countries have a special inter- detracting trade barriers in place. The resid- est in greater market access, but they cannot ual barriers sap growth in the protected succumb to the siren calls of preferential mar- economies and in developing-country trad- ket access nor opt out of reducing border pro- ing partners that are denied access. Perhaps tections at home, which tax exports and cut most important, the majority of the world’s into productivity growth. Preferences for poor do not live in the least developed coun- LDCs can help, but would be more effective if tries (LDCs). Trade preferences targeted at they were made less restrictive and more reli- these countries do not benefit the three- able than at present—and if benefiting coun- quarters of the world’s poor that live on tries take the necessary policy steps, including US$1 per day in other countries. In imple- reductions in border protection, to promote a menting new WTO rules, new accords will supply response. Moreover, because other de- be most effective if they recognize differ- veloping countries are unlikely to be granted ences among individual countries’ capacity new trade preferences, global reciprocal re- to undertake new, resource-intensive rules. duction in trade barriers holds the most These differences require a new approach to promise for the world’s poor. special and differential treatment. Market access is not the whole develop- ment story. Even if developing countries suc- These areas pose difficult political chal- ceed in obtaining access to new markets, they lenges for all segments of the international will have to adopt complementary policies— community—rich countries, middle-income de- removing obstacles to private investment, im- xv G L O B A L E C O N O M I C P R O S P E C T S proving public investment in infrastructure, Uruguay Round trade agreements to reduce and providing education—to ensure that do- protection, agriculture is among the most dis- mestic firms respond to new opportunities as- torted sectors in international trade. Even sociated with greater integration, and that the though levels of average tariff protection are benefits of integration are transmitted to the comparable in rich and poor countries, the ex- poor. Put differently, trade policies must be tensive use of producer subsidies in the OECD embedded in a coherent national development countries and the fact that the OECD consti- strategy—they are not a substitute for it. For tutes two-thirds of world agricultural trade un- all of these reasons, realizing the development derscore the centrality of their policies to de- promise of the Doha Agenda requires the par- velopment outcomes. Reducing protection in ticipation of all groups of the international agriculture alone would produce roughly two- community. thirds of the gains from full global liberaliza- tion of all merchandise trade. This report: toward a pro-poor A few facts are enough to establish the con- Doha outcome text: protection facing developing country ex- This report analyzes central elements of the porters in agriculture is four to seven times Doha Agenda that are important to developing higher than in manufactures in the North and countries. Chapter 1 describes the prospects two to three times higher in developing coun- for the global economy that form the back- tries (IMF-World Bank 2002). Tariff peaks are drop to the Doha trade negotiations. Chapters particularly high in rich countries against 2–6 focus on agriculture, nonagricultural trade, products from poor countries. Tariff escalation services, transport and trade facilitation, and that discourages development of further pro- special development provisions. In each area, cessing is more pronounced in agriculture in we expand on themes that have received less both rich and poor countries (figure 2). Hefty analysis in previous World Bank reports— specific duties are particularly common in rich among them specific duties in agriculture, an- countries; they automatically increase protec- tidumping in manufactures trade, temporary movement of labor in services, security issues in trade facilitation, and trade preferences and exemptions from rules as part of special and Figure 2 Escalating tariff rates discourage differential treatment (SDT). The remainder of development Tariff rates this overview weaves these findings together with those of previous Bank studies1 to lay out 25 the principal elements of a pro-poor outcome for the Doha Agenda. 20 Final ᮡ 15 A Doha deal for development Intermediate ᮡ 10 Agriculture is at the heart Raw of a development round ᮡ Agriculture is central to the development 5 promise of this trade round for two reasons: most of the world’s poor work in agriculture 0 and most of the world’s protection is directed QUAD Canada Japan United European States Union at agriculture. Some 70 percent of the world’s poor live in rural areas and earn their income Sources: World Bank staff. from agriculture. Largely exempt from pre– xvi O V E R V I E W tion when commodity prices fall, throwing the tent, the United States. The net effect of subsi- burden of adjustment onto global prices and dizing the relatively rich in wealthy countries at poor countries. Subsidies in OECD countries the expense of adverse price penalties for the amount to US$330 billion—of which some products of the relatively poor in developing US$250 billion goes directly to producers. The countries is to aggravate global income inequal- effect is to stimulate overproduction in high- ities. Said differently, subsidies make the rela- cost rich countries and shut out potentially tively rich even richer and the poor even poorer. more competitive products from poor coun- Realizing the development potential of tries. It is no wonder that agricultural exports Doha requires phased reductions of border from developing countries to rich countries protection and subsidies. Of these, border pro- grew in the 1990s at just half the rate they did tection is the most important. These reductions to other developing countries. ought to be done in a way that cuts off anti- Consider how agricultural protection plays development tariff peaks, reduces tariff escala- through individual commodity markets. Sugar tion, and phases out specific duties. A pro- in the European Union (EU), Japan, and the poor reform also means reforming policies that United States is commonly protected through distort particular commodities of importance a combination of quotas, tariffs, and subsidies to developing countries—sugar, cotton, rice, allowing domestic sugar producers in those wheat, and dairy products. countries to receive more than double the Because global prices may rise in some com- world market price. OECD governments sup- modities, the international community may port sugar producers at the rate of US$6.4 bil- want to design—and help finance—a program lion annually—an amount nearly equal to all of adjustment in vulnerable countries that suf- developing country exports. Prices are so high fer deterioration in their terms of trade. These that it has become economic to grow sugar effects are likely to be confined to a few coun- beets in cold climates and to convert corn to tries for several reasons: many food importers high-fructose corn syrup. Sugar imports in the also export other agricultural products that OECD have shrunk to next to nothing. U.S. will experience positive terms-of-trade changes subsidies to cotton growers totaled US$3.7 bil- from liberalization; others now have tariffs on lion last year, three times U.S. foreign aid to those same food imports, tariffs that can be re- Africa. These subsidies depress world cotton duced to offset any increase in global prices; prices by an estimated 10–20 percent, reducing some food importers will gain access to new the income of thousands of poor farmers in markets in nonagricultural products and be West Africa, Central and South Asia, and poor able to export; and, because prices will change countries around the world. In West Africa relatively slowly, some food importers will in- alone, where cotton is a critical cash crop for crease domestic production in response to many small-scale and near-subsistence farmers, higher prices and become self-sufficient or even annual income losses for cotton growers are net exporters. Nonetheless, even though the about US$250 million a year. Rice support in changes are likely to be manageable at the Japan amounts to 700 percent of production at global level, the issue requires study and in world prices, stimulating inefficient domestic some countries may require action. production, reducing demand, and denying ex- Because rich and poor countries alike will port opportunities to India, Thailand, Vietnam, benefit from liberalization, all must make the and other countries. policy changes necessary to realize its develop- More than 70 percent of subsidies in rich ment promise. The rich countries, whose poli- countries are directed to large (often corporate) cies arguably distort international trade the farmers. These farmers have incomes that are most, cannot escape leadership on agriculture. higher—often substantially so—than average Moreover, leadership among donors to fi- incomes in Europe, Japan, and, to a lesser ex- nance a program to cushion adjustment is xvii G L O B A L E C O N O M I C P R O S P E C T S Box 1 Trade and poverty: what are the links? C ountries that trade more grow faster, according to evidence emerging from case studies of trade liberal- ization and from large cross-country and time-series Box Figure 2 Changes in trade have little relation to inequality econometric studies. Although the links from specific Average annual change in Gini coefficient trade policy instruments to trade outcomes and growth Trade and the Gini coefficient is less clear, the basic association between increased trade 4 and growth is clear (box figure 1).a Series 1 3 Even when trade raises average incomes, its ᮡ effects on poverty will depend on whether poverty 2 Linear in a given country is sensitive to growth in average in- 1 (Series 1) comes, and on how the increase in trade affects the dis- ᮡ 0 tribution of income in the country. The first of these is- –0.08 –0.06 –0.04 –0.02 0 0.02 0.04 0.06 0.08 sues is empirically well understood. The sensitivity of –1 poverty to growth in average incomes depends in an im- –2 y = 2.5227x + 0.0139 portant way on initial inequalities in a country (Raval- R2 = 0.0013 lion 1997). When incomes and opportunities are distrib- –3 uted relatively equally, the effect of growth on poverty –4 is larger than when initial inequality is high. Thus, Average annual change in trade/GDP growth associated with increased trade (or from any Note: This figure shows changes in trade as a fraction of GDP and other source) is likely to have larger proportional effects changes in the Gini measure of income inequality, for a large on poverty in countries where initial inequality is low. sample of growth episodes of at least five years in duration. More interesting and potentially more important Source: Dollar and Kraay (2001). are the effects of increased trade on the distribution of income. Almost by definition, if increased trade dispro- in many developing countries are likely to be relatively portionately benefits the poor, poverty will fall faster well off, but will benefit poorer consumers of their than if trade disproportionately benefits the nonpoor. products by lowering prices. Understanding the likely distributional consequences of At the same time, however, the distributional conse- trade liberalization is therefore crucial to understanding quences of trade liberalization can also work against the overall effects of trade on poverty. In many cases, poor people. For example, reductions in tariffs imply re- there are very direct channels through which trade liber- ductions in trade tax revenues that can be important in alization is likely to disproportionately benefit the poor. developing countries that rely disproportionately on this For example, agricultural trade liberalization that al- source of revenue. To the extent that public spending dis- lows previously suppressed prices of agricultural goods proportionately benefits poor people (and this is by no to rise to world levels will benefit farmers, who are net means universal), reductions in tax revenues that accom- producers, but will hurt consumers. If farmers are more pany trade liberalization can have adverse distributional likely to be poor, the liberalization will be, on average, consequences. pro-poor. Similarly, reductions in tariffs on manufactur- The likely distributional consequences of trade lib- ers will hurt previously protected urban workers, who eralization, therefore, are complex and country-specific. Determining whether a given action would be pro- or anti-poor requires careful analysis. Looking back across Box Figure 1 Integration with global countries, there is little evidence that increased trade is markets is associated with faster growth systematically associated with either increases or de- Average annual per capita growth, 1980–99 creases in inequality (box figure 2). 3.5 On average, trade can be a powerful force for 3 poverty reduction, especially over longer horizons where 2.5 the cumulative effects of growth on incomes of the poor 2 are large. But this will not be true for all countries at all 1.5 times—underscoring the importance of complementary 1 pro-poor policies at the country level to ensure maximum 0.5 positive effects in every situation. 0 Decreasing export Increasing export a share in GDP share in GDP For contrasting views on the state of the evidence on trade, trade policies, and growth, see Srinivasan and Bhagwati (2000), Source: World Bank (2001). Rodriguez and Rodrik (1997), Bernanke and Rogoff (2001). Source: World Bank staff. xviii O V E R V I E W essential; their technical assistance to help markets. Because import tariffs indirectly tax implement standards and facilitate trade is exports, reducing trade barriers in developing needed to help developing countries take ad- countries stimulated trade. The burden of im- vantage of new trade opportunities. Middle- port protection on all export activities in de- income countries, whose own policy reforms veloping countries declined, but more so for would produce a large share of the benefit to manufactures than for agriculture and natural developing countries from global liberaliza- resources. At the same time, the fact that suc- tion in agriculture, have to move more as- cessive multilateral trade rounds liberalized sertively than in the past. Their high tariffs global manufactures, while rich countries con- have an adverse impact on growing South- tinued to protect their agriculture (and devel- South trade, especially with neighboring oping countries eventually began to follow countries. In a pattern common to all regions, suit) meant that developing countries’ exports agricultural exporters in East Asia, for exam- of manufactures were free to grow more ple, paid one-third of all their tariff duties to rapidly than those in agriculture. other East Asian governments (second only Today, trade in manufactures is still im- to tariffs paid to get into rich countries). Agri- peded. Although tariffs on manufacturing in cultural exporters in the Middle East paid rich countries are on average lower than in de- 44 percent of their tariff duties to regional veloping countries, the tariffs rich countries neighbors. charge developing countries are substantially higher than those they charge other industrial Nonfarm trade is increasingly essential countries. For example, exporters of manufac- to growth in poor countries tures from industrial countries face, on aver- Over the past two decades, developing coun- age, a tariff of 1 percent on their sales to other tries have increased their share of global trade industrial countries; exporters in developing from just under one-quarter to about one- countries pay anywhere from 2 percent if they third. As a group, they have moved beyond are from Latin America (where NAFTA weighs their traditional specialization in agricultural heavily) to 8 percent if they are from South and resource exports into manufactures trade. Asia. Overall, rich countries collect from de- Exports of manufactures have grown at nearly veloping countries about twice the tariff rev- twice the rate of agriculture, and now consti- enues per dollar of imports that they collect tute nearly 80 percent of exports from all de- from other rich countries. However, the prob- veloping countries. Countries that were low in- lem is not solely a North-South issue. Latin come in 1980 managed to raise their exports of American exporters of manufactures, for ex- manufactures from roughly 20 percent of their ample, face tariffs in neighboring Latin Ameri- total exports to more than 80 percent (figure can markets that are seven times higher than in 3). As a result, many grew quickly and entered industrial countries. In Sub-Saharan Africa, the the ranks of today’s middle-income countries. same multiple is six; in South Asia, two. The middle-income group of 1980 also in- Protection takes forms other than tariffs— creased its manufactures share, but somewhat among them quotas, specific duties, and con- less rapidly, to reach nearly 70 percent. This tingent protection measures such as antidump- dramatic change in trade magnitudes and com- ing duties. As with tariffs, these measures tend position has given developing countries a new to be used more frequently against labor- interest—and a powerful voice—in the ongo- intensive products from developing countries. ing Doha Round. The quota arrangements in the WTO Agree- One reason for this change was the dra- ment on Textiles and Clothing (ATC) still matic reduction in border barriers in develop- shackle the exports of many poor countries. ing countries since the mid-1980s, in combi- Although these arrangements are scheduled to nation with increased access to rich-country be removed in only 15 months, rich countries xix G L O B A L E C O N O M I C P R O S P E C T S Figure 3 Developing countries have become important exporters of manufactures In middle-income countries, manufactures make up 70 percent of exports Middle-income countries’ share of world exports, 1981–2001 (percent) 80 70 Manufacturing exports (%) 60 ᮡ 50 40 Resources exports (%) 30 ᮡ 20 ᮡ 10 Agricultural exports (%) 0 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 In low-income countries, manufactures make up 80 percent of exports Low-income countries’ share of world exports, 1981–2001 (percent) 90 80 Manufacturing exports (%) ᮡ 70 Resources exports (%) 60 ᮡ 50 40 30 Agricultural exports (%) 20 ᮡ 10 0 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Source: UN COMTRADE. to date have freed up only 15 percent of the Realizing the development promise of quotas, obliging them to implement major Doha depends particularly on three efforts. changes at the end of the phase-in period. Av- erage antidumping duties are seven to ten • First, rich countries desirous of promoting times higher than tariffs in industrial coun- development can do so by ensuring that the tries, and about five times higher in developing now lagging phase-out of the ATC is com- countries. Today’s protection remains heavily pleted according to the agreement—and not concentrated in the most politically sensitive reversed through antidumping actions. The areas—textiles, clothing, and other labor-in- ATC phase-out will also require reforms by tensive manufactures, as well as agriculture— some exporters facing increased competi- in both rich and poor countries. tion, many of which are LDCs, to ensure a xx O V E R V I E W smooth adjustment; trade-related develop- No less important, developing countries ment assistance could play a role in easing have an interest in locking in market access for the transition. their services exports to rich-country markets— • Second, in both rich and poor countries, ef- exports that are growing more rapidly than forts to cut back on antidumping measures merchandise exports. Examples include China’s that create a patchwork of ad hoc protection incipient software industry as well as software are essential if market access granted by the and back-office services from India. right hand of quota elimination and tariff re- The Doha Round has the potential of lock- ductions is not to be withdrawn by the left ing in access to foreign markets for services hand of antidumping suits. Developing coun- exports. Just as many rich countries have not tries themselves have become accomplished yet bound access for developing countries’ ser- practitioners of contingent protection. vices exports, many developing countries have • Third, moving forward in nonfarm trade re- yet to schedule with the WTO liberalizing re- quires a Swiss-type formula approach that forms that have already been undertaken. Of- will require disproportionately greater re- fering to bind unilateral reforms can be used ductions in high tariffs so as to mitigate the to lock in existing access to overseas services antidevelopment bias embedded in most markets. Active participation in the services tariff structures around the world. The negotiations could help accelerate these twin choice of the formula, and of its coefficients processes (Mattoo 2003). of reduction, is important. Applying these The GATS process allows governments to cuts to bound rates will effectively credit de- liberalize services at their own pace. It does not veloping countries that have unilaterally re- require that a government forgo its regulatory duced their applied tariffs since the end of responsibilities. Nor does the GATS frame- the Uruguay Round. work require a cessation of subsidies or pre- empt pro-poor regulation on universal service Services liberalization could raise access. The main requirement is that, once a productivity sector is scheduled, governments are required Services are the fastest-growing component of to have transparent regulations, treat domestic the global economy. Even in developing coun- and foreign companies alike, and permit all tries, services exports grew more rapidly than foreign companies access to the domestic mar- manufactures in the 1990s (World Bank 2001, ket on the same terms as domestic companies. chapter 3). More efficient backbone services— In fact, many governments have chosen to in finance, telecommunications, domestic trans- liberalize—but not to make commitments portation, retail and wholesale distribution, with the GATS that would bind this opening. and professional business services—improve Some two-thirds of the WTO members have the performance of the whole economy because scheduled fewer than 60 sectors of the approx- they have broad linkage effects. Yet most devel- imately 160 sectors covered by the GATS. For oping regions trail the industrialized world in example, only 12 developing countries have exposing service sectors to competition. Figure made commitments in education. None have 4 shows that only Latin American countries made commitments in the provision of water. are beginning to approximate the high-income Why the reluctance? Liberalization in ser- countries in their degree of competition. Esti- vices is more complicated than in goods mar- mates suggest that, after controlling for other kets. Privatization without competition and determinants of growth, countries that fully lib- proper regulation may achieve nothing more eralized trade and investment in finance and than transforming a public monopoly into a telecommunications grew on average 1.5 per- private monopoly—with no improvement in centage points faster than other countries over services. And too many developing countries the past decade (Mattoo, et al. 2001). have been content to change ownership xxi G L O B A L E C O N O M I C P R O S P E C T S Figure 4 Developing countries lag behind rich countries in services liberalization Financial services South Asia ᮡ East Asia Middle East and North Africa Europe and Central Asia Latin America and the Caribbean High income Middle East and North Africa Telecommunications Europe and Central Asia ᮡ South Asia East Asia Latin America and the Caribbean High income 0 2 4 6 8 10 Greater competitiveness ᮡ Source: World Bank Global Economic Prospects 2002, based on data from Mattoo, et al. (2001). through privatization while retaining limits on 2001, chapter 3). Trade ministers wishing to entry that buttress monopolies. harness the reciprocal negotiating framework Effective regulation is critical to ensure that of the GATS to spur domestic reforms while the poor have access to basic services (World leveraging market access abroad must ensure Bank 2002a, 2002b). Some sectors, such as that sectoral ministries have properly se- retail and wholesale services, can be opened quenced regulations to support liberalization. expeditiously because competition can be re- lied on to discipline firms’ pricing and invest- Liberalized trade in labor services could ment decisions. Others, however, require well- contribute much more formulated regulations before liberalization to To date, virtually all GATS commitments have ensure proper market functioning and ade- focused on the first three “modes” of interna- quate access for low-income groups to ser- tional service delivery. Most trade in services vices. In China’s financial sector, for example, has occurred through those same modes. the World Bank recommended that financial Twenty-eight percent of the value of services markets be opened gradually to allow regula- trade, for example, has been in Mode 1, “cross- tions and institutional developments to pre- border supply of services.” Another 14 percent cede liberalization. The goal was to avoid has been in Mode 2, “consumption abroad,” destabilizing financial losses by state banks such as tourism. Fifty-six percent has been saddled with poor portfolios as efficient in Mode 3, “commercial presence,” such as banks, domestic and foreign, entered the mar- through foreign direct investment in services. ket (World Bank 1996). China’s WTO acces- Mode 4, which involves the temporary sion agreement generally reflected this phased movement of labor to provide services, ac- approach. In network sectors, such as counts for only 1.4 percent of services trade telecommunications and water, ensuring ade- (figure 5). Temporary movement has some ad- quate pricing and universal access are simi- vantages over permanent migration for both larly important if the poor are to benefit from developed and developing countries. Rich the expansion of the system (World Bank countries can obtain workers whose skills are xxii O V E R V I E W to higher-level personnel. More than 40 per- Figure 5 Temporary labor mobility is an cent of workers covered by existing Mode 4 underused mode of trade in service commitments are intracorporate transferees Value of world trade in services by mode, (percent) whose mobility is intimately related to foreign Mode 4 direct investment (often in services); another (movement of natural persons) 50 percent are executives, specialists, and sales 1% personnel who are business visitors. To date, Mode 1 therefore, Mode 4 has been of limited signifi- (cross-border cance for developing countries, whose compar- supply) 28% ative advantage lies in the export of medium and low-skilled, labor-intensive services. In addition to other concerns associated with broader migration issues, two fundamen- tal tensions hamper progress on Mode 4 tem- porary labor mobility. The first is that govern- Mode 3 Mode 2 ments are reluctant to undertake permanent (commercial presence) (consumption commitments when employment demand varies abroad) 57% 14% with cyclical conditions. Wanting to maintain policy flexibility, immigration and labor market Source: IMF, Balance of Payments Yearbook. officials have made GATS commitments far below the degree of TMNP access already af- forded under domestic laws and regulations. in short supply, with minimal disruption of TMNP liberalization has been greatest in sec- labor markets and without taxing social ser- tors (and for categories of workers) where labor vices. Temporary migration allows develop- demand routinely exceeds supply—tourism, in- ing countries to obtain access to new, higher- formation technology, health services. The sec- paying jobs without necessarily suffering the ond tension stems from the fact that regional “brain drain” that would occur with perma- patterns of migration create domestic political nent migration. Poor countries also gain from support for programs that favor neighboring remittances sent home by temporary migrants, countries, whereas Mode 4 programs necessar- and returning workers bring new skills back to ily are open to all countries on a most-favored- the sending country. In 2001, remittances from nation (MFN) basis. Preferential migration permanent as well as temporary migrants pro- schemes are commonly negotiated at the bilat- vided some US$71 billion to developing coun- eral and regional levels—and MFN-based liber- tries, nearly 40 percent more than all official alization would undermine these. Because development assistance and significantly more many bilateral labor agreements are usually not than net debt flows to developing countries. If tied to trade policy or other agreements, they temporary movement of labor up to 3 percent afford governments a greater degree of flexibil- of the total labor force in rich countries were ity to adjust programs to evolving migration permitted, developing countries would stand trends and labor-market needs. to gain as much as US$160 billion in addi- Tensions notwithstanding, present levels of tional income (Walmsley and Winters 2003). Mode 4 use fall far short even of Mode 4’s rel- To date, however, even after the significant atively modest potential. To rectify this, devel- liberalization of trade in services during the oping countries should expand their requests Uruguay Round, little has been done to loosen and offers in the Doha Round. Only six re- conditions governing the temporary movement quests had been tabled by June 2003, and only of natural persons (TMNP) supplying services. two from developing countries (India and Co- Present commitments refer almost exclusively lómbia). Also, WTO members should adopt xxiii G L O B A L E C O N O M I C P R O S P E C T S rules that would provide greater clarity and To counter any trade-reducing effects of predictability. To help regularize entry and security measures, every effort to cut trade- exit while improving security, countries could related costs in other areas is imperative. Reg- adopt a GATS visa system that would facili- ulatory restrictions on international air and tate national visas for up to one year, subject maritime transport services inflate transport to appropriate security checks and oversight costs—on some routes by amounts that dwarf (see Hatcher 2003 and Self and Zutshi 2003). the value of tariffs. International air transport, which carries about 30 percent of developing Reducing transport costs and facilitating countries’ exports by value, is heavily pro- trade can have a powerful effect tected from international competition. Bilat- The cost of moving goods across international eral air service agreements commonly bar entry borders is often as important as formal trade to efficient outside carriers, thereby raising ex- barriers in determining the cost of landed port costs for developing countries. City-pair goods—and ultimately of market share. One routes on which more than two passenger air- study estimated that every day spent in cus- lines or dedicated freight airlines operate can toms adds nearly 1 percent to the cost of goods cut costs by an average of 10.7 percent. Mar- (Hummels 2001). In developing countries, itime transport, too, is often subject to prac- transit costs are routinely two to four times tices, such as cargo-reservation schemes and higher than in rich countries. Transparent cus- limitations on port services, which protect in- toms regimes, modern port facilities, dense efficient service providers. Such competition- transportation networks, and access to infor- restricting practices among shipping lines mation and telecommunications systems—all and port-terminal operators can increase can help lower transit costs. freight rates up to 25 percent on some routes. Since September 11, 2001, security has be- Rising concentration in the market for port- come a dominant issue in international trade. terminal services has increased the risk that Border inspections, cargo review, and other private firms may capture the benefits of gov- measures have increased transport times and ernment reforms. Abusive practices by private driven up costs. Each 1 percent increase in operators are of special concern in develop- costs to trade from programs to tighten border ing countries, where traffic volumes are lower security reduces world income by US$75 bil- and competitive forces inherently more limited. lion per year. Developing countries, too, are Regulations governing such practices are now vulnerable to security threats and terrorism, outside the WTO mandate, but logically they but limited budgets, dependence on foreign should be reviewed for reformulation. trade and investment, and outdated infrastruc- Facilitating trade by eliminating delays in ture and technology present serious challenges developing countries would lower trading for these countries. New security protocols costs significantly, particularly if accompanied being deployed at ports, customs offices, and by liberalization of transport and telecommu- border posts around the world have the poten- nications, and streamlined regulations to pro- tial to add costs and diminish market access mote domestic competition. Trade facilitation for developing countries—at least in the short requires modernizing customs, improving port term. But managed correctly, the same mea- facilities, and making investments in trade- sures can streamline trade transactions while related information technology—a huge insti- promoting safety and security. To achieve this tutional and infrastructural agenda. Countries trade-expanding result, a global framework display wide variation in customs efficiency must be established to ensure that the needs and clearance times, for example (figure 6). If of developing countries are addressed as en- those whose trade-facilitation capacity was hanced security regimes take shape. below average could be brought halfway up to xxiv O V E R V I E W Aspects of trade facilitation are part of the Figure 6 Clearing customs takes longer WTO’s trade-related disciplines, particularly in developing countries the provisions that encourage uniform treat- Average number of days to clear customs for sea cargo ment of transit trade and transparency of fees. Strengthening provisions related to transit, Developed fees, and transparency, issues originally in the General Agreement on Tariffs and Trade East Asia and (GATT), would be helpful. However, best Pacific practice cannot be established in a vacuum; it Latin America has to be gradually created in sound domestic and the Caribbean laws, regulations, and practices. A sustained program of institutional reform must be tai- Africa lored to each country, and it often requires technical assistance. The bilateral donors and multilateral development banks and agen- South Asia cies are best positioned to provide the thor- ough diagnostics and technical assistance 0 2 4 6 8 10 12 required to promote needed institutional Note: This is based on a sample of countries in each area; see figure 5.1 in chapter 5. change. Source: International Exhibition Logistics Associates. If the dynamics of the Doha negotiations Available at http://www.iela.org. propel the WTO into a role in the broader trade-facilitation agenda, any agreement, if it is to be effective, should recognize limitations in the global average, international trade would domestic capacity for implementation. An increase by US$380 billion annually. agreement would be most effective if it in- Multilateral efforts are under way outside cluded a serious commitment by developed na- the WTO to promote—and in some cases fi- tions to finance new trade-facilitation systems. nance—institutional changes in trade facilita- Development assistance delivered under the tion. Key players include the World Customs commitment could be provided by the World Organization, the regional development banks, Customs Organization, the multilateral devel- and the World Bank. Their efforts focus on opment banks, and bilateral donors. The obli- policy reform, technical assistance, and infra- gations of developing countries should be tai- structure modernization. lored to their implementation capacity. And because the WTO’s dispute settlement provi- Should trade facilitation, investment, and sions are largely inappropriate to promoting competition be the subject of new institutional changes, conventional enforce- multilateral disciplines in the WTO? ment of dispute settlement through trade sanc- As one of the four Singapore issues, trade tions ought to be set aside. facilitation is under discussion in Geneva for Other Singapore issues would stretch the possible inclusion in the Doha Agenda. Al- WTO mandate into yet new areas, probably ready the WTO, through the GATS, has a po- with only marginal development benefits if tentially important role to play in interna- taken up in isolation. As discussed in Global tional transport and trade logistics—many of Economic Prospects 2003, there is no evidence the transport service sectors could be immedi- that an investment agreement would, by itself, ately scheduled with the GATS if countries promote new foreign investment. Similarly, saw fit to do so. However, few countries have adopting an agreement in competition policy— taken advantage of its provisions. as currently framed in the negotiations—would xxv G L O B A L E C O N O M I C P R O S P E C T S have minimal effects on the terms of trade of developing countries, unless the agreement Figure 7 Fewer exporters to the Quad were to establish new disciplines on national countries are taking advantage of export cartels and illegal international cartels preferences (World Bank 2002a, chapter 4). Finally, a new Share of potential imports under GSP that entered with preferential access, 1994–2001 (percent) agreement on government procurement that 60.0 focuses on transparency is unlikely to improve market access substantially (Evenett 2002). 55.0 Virtually all of the disciplines proposed in these arrangements would require new policy actions 50.0 only in developing countries. Although some of these may promote development, the main ben- 45.0 efits of WTO agreements in these areas would be in the market access that new agreements 40.0 leverage (Newfarmer 2003). 35.0 Securing the benefits of trade for the poorest countries 30.0 More favorable and differential treatment of 1994 1995 1996 1997 1998 1999 2000 2001 developing countries is a prominent feature Source: Inama (2003). of multilateral trade rules. Selected subsets of countries have been granted trade preferences. Some countries were granted exemptions or al- “deep preference” programs, such as the EU’s lowed to defer implementing some multilateral Everything But Arms program and the U.S. agreements; many have benefited from techni- African Growth and Opportunity Act, but each cal assistance to help implement mandates. has different rules and exceptions. For these The present patchwork system has not reasons, preferences cover only a portion of ex- worked especially well. Countries benefiting ports from even poor developing countries— from trade preferences have generally under- and among eligible countries and products, performed in exports. One reason is that rich only a fraction of preferences are actually used. countries grant preferences voluntarily rather Even when effective, preferences tend to divert than as part of a binding multilateral negotia- trade away from other poor countries, effec- tion. Those preferences often come laden with tively “robbing Peru to pay Panama.” restrictions, product exclusions, and adminis- Existing preferences do relatively little for trative rules that prevent beneficiaries from tak- most of the world’s poor people (those living ing full advantage of them. For example, only on less than US$1 per day), most of whom live 39 percent of potentially preferred imports in China, India, Nigeria, Pakistan, Northeast under the Generalized System of Preferences Brazil, and the ASEAN countries, which may (GSP) into the Quad countries—Canada, the enjoy only partial preferences at best. Al- EU, Japan, and the United States—actually though some of these countries enjoy limited took advantage of preferential access—and preferential access to some markets, all would usage rates are declining (figure 7). At times, be better off with across-the-board, non- protectionist lobbies have weighed in to pres- discriminatory binding access. sure for reductions in the preference, either be- Finally, the extensive use of voluntary pref- fore a country was deemed eligible or even erence schemes has created perverse incentives later, when the first signs of export success for in both rich and poor countries to avoid liber- developing countries become evident. Beyond alization that would otherwise benefit the GSP, the Quad countries sponsor their own poor. Too often, rich countries have offered xxvi O V E R V I E W differential treatment to a subset of poor • Finally, the WTO membership must learn countries instead of arriving at MFN reduc- which of its policies promote, and which de- tions in trade barriers that would benefit all feat, the interests of developing countries. developing countries. And, too often, develop- Getting the rules right is arguably the major ing countries have sought preferential access challenge confronting WTO members from and exemptions from agreed MFN reductions a development perspective. Among other in trade barriers that would benefit themselves things, getting the rules right means limiting and other developing countries. In other new rule-making to cases in which the payoff words, the present system of preferences re- for developing countries is clearly positive. duces the incentives to negotiate effectively for reductions in trade barriers abroad and with TRIPS and Public Health domestic protectionist constituencies at home. Negotiations at the WTO on patents and pub- Making trade regime more supportive of lic health have stalemated over the question of development therefore involves four important improving access to generic drugs for poor policy directions. countries. The WTO’s Agreement on Trade Re- • Central to any new regime is improvement lated Aspects of Intellectual Property Rights in market access for all developing countries (TRIPS), which took effect in 1995, obliges on an MFN basis, especially in products that countries to extend patent protection to phar- have hitherto escaped WTO disciplines, maceutical products and processes after a such as agriculture and labor-intensive prod- phase-in period linked to level of development ucts. Broad market access would allow trade (World Bank 2001, chapter 5). Under these reform to reach the 70 percent of the world’s rules, countries that are able to manufacture the poor not living in the 49 least developed drugs themselves would continue to have legal countries. access to generics if they chose to issue compul- • Trade preferences would be more effective sory licenses. These tend to be the larger and if they were consistent and uniform, shorn better-off developing countries such as Brazil, of restrictions that raise the cost of tak- China, India, and Thailand. Countries that lack ing advantage of preferences. WTO rules sufficient manufacturing capability—typically that require institutional improvements— the world’s poorest and often most disease- especially “behind the border” policies, as ravaged states—may be barred from importing distinct from trade policy changes that can generic versions of patent-protected drugs, once be implemented at the “stroke of the pen”— rules take effect. Hence, the Doha mandate on would be more effective if they were cali- TRIPS and Public Health included finding a brated to developing countries’ capacity mechanism by which such countries can import to implement them. As countries move up generic drugs protected by patents abroad. the ladder of development, they should be These rules are important for poor people. expected to assume the full obligations of For example, one day’s supply of patented an- WTO members. tiretrovirals to treat a single HIV/AIDS patient • Integrating technical assistance into the na- can cost as much as US$30 in rich countries. tional priorities for development while in- Such prices are prohibitive for the nearly 3 bil- creasing “aid for trade”—a part of the lion people who live on less than US$2 per Monterrey consensus—could help poor de- day. Generics are not always cheaper, but the veloping countries identify and address threat of competition has helped to reduce trade-capacity priorities. Increased develop- prices of patented antiretrovirals supplied to ment assistance—for ports, customs, and developing-country governments (Fink 2003). logistics management—would augment the Patents create incentives for research by of- capacity of developing-country firms to fering temporary monopolies on new drugs, benefit from market-access opportunities. and developing countries need that research as xxvii G L O B A L E C O N O M I C P R O S P E C T S much as the rest of the world. Indeed, in- countries. Of equal importance to the health of creased R&D for medicines to treat diseases the poor is undertaking the large investments that are more prevalent in developing coun- in complementary health infrastructure, in- tries is desperately needed. Yet poor countries cluding hospitals, roads, warehouses, and doc- that lack pharmaceutical manufacturing capa- tors and nurses. For example, even in some bility form only a tiny portion—perhaps less countries that manufacture anti-AIDS generics than 1 to 2 percent—of the global pharma- or that get AIDS drugs free, governments have ceuticals market. In the 12 months to October not succeeded in providing medicines to signif- 2002, developed countries accounted for more icant shares of the needy population. Second, than 95 percent of the US$270 billion of sales funding for fighting the developing world’s in the world’s leading 20 country markets health crisis needs to be scaled up—and mas- worldwide. The group of developing countries sively. For example, the latest projections by that may benefit from a WTO agreement on UNAIDS put the cost of the global struggle importing generic drugs under compulsory against AIDS at US$10.5 billion a year by licensing probably accounts for less than 1 or 2005 and US$15 billion a year by 2007; even 2 percent of global pharmaceutical sales. Per- if governments in affected countries cover part mitting the export to these markets of generic of this amount, estimated aid flows of about versions of patented medicines developed for US$3 billion in 2002 are still insufficient. The rich-country markets is unlikely to erode in- Global Fund to Fight Aids, Tuberculosis, and centives for research and development (Fink Malaria remains cash-strapped. The recent 2003). Despite this unlikelihood, the negotia- U.S. commitment of US$15 billion to fight tions going into the summer of 2003 were HIV/AIDS will, when disbursed, partially re- stalemated on possible restrictions of the list lieve resource constraints, but a substantial of diseases that would be covered by any new funding gap remains. The TRIPS issue is small agreement. when compared with the real obstacles pre- Governments everywhere have potentially venting access to better health in developing competing interests. They have an interest in countries, and it concerns a small corner of the maintaining R&D and in preventing illegal global pharmaceutical market—two reasons generics entering rich-country markets from why the international community should move undercutting patent rights that finance it. swiftly to resolve it. Strengthening mechanisms that prevent such illegal trade is important, such as by prohib- iting generic manufacturers from mimicking Delivering the Doha deal the packaging of patented drugs. At the same for development time, governments everywhere have an inter- est in ensuring that limited budgets for drugs to improve health in poor countries go as far T he potential for reciprocal reductions in trade protection holds the promise of bet- ter lives for everyone. To illustrate, we consider as possible, and this means that all developing the effects of a pro-poor agreement in which countries have access to drugs at the cheapest rich countries cut tariff peaks to 10 percent in most competitive prices. In balancing these agriculture and to 5 percent in manufacturing, objectives, any eventual agreement should put and in which these reductions are reciprocated the developing countries with insufficient with cuts to 15 and 10 percent in developing manufacturing capacity on the same footing countries (table 1). This program, combined as those countries that have manufacturing with reductions in prevailing tariff averages, a capacity. decoupling of agricultural subsidies, and an end Resolving the Doha issue is only one small to agricultural subsidies could realize nearly piece of the larger problem of delivering drugs three-quarters of the gains that might be antic- and health care to sick people in developing ipated from full merchandise liberalization. xxviii O V E R V I E W Table 1 A pro-poor tariff reduction countries that have high protection will find program they benefit from domestic reforms that lower (percent) costs of imported inputs, increase domestic Rich Developing competition that spurs productivity growth, and expand exports. Study after study has Agriculture Average 5 10 shown that trade reforms redound first and Maximum 10 15 fastest to the reformer. Negotiations will determine the pace and Manufacturing Average 1 5 details of a final package, but the broad out- Maximum 5 10 lines of a potentially good deal for develop- ment are already evident from this analysis. Realizing that agricultural reform in a time of This illustrative pro-poor program, dis- rapidly rising budget deficits will contribute cussed in detail in chapter 1, if implemented positively to their own economic growth, rich progressively over the five years to 2010 and countries would benefit from reforms in agri- accompanied by a realistic productivity re- culture. Lopping off tariff peaks and phasing sponse, would produce gains for developing out the ATC at the end of 2004 will benefit countries of nearly US$350 billion in addi- developed-country poor who are forced to tional income by 2015. Rich countries would pay more for food and clothing because of benefit, too, with gains on the order of US$170 external protection (Gresser 2002). Further billion. All of this would mean that 8 percent progress on the part of all countries in reduc- fewer people would be living in poverty in ing tariffs in manufactures would benefit de- 2015—140 million fewer people living below veloping countries and stimulate healthy US$2 per day. If greater opening of services, in- South-South trade. For the rich countries, the cluding Mode 4, were to occur, the benefits prospect of greater access to markets in devel- would be substantially greater.2 oping countries—home to 80 percent of the Delivering a Doha deal that spurs develop- world’s population with markets growing two ment will not be easy. Negotiators may well to three times faster than their own—is also a have to transcend the mercantilist mind-set worthy prize. that tends to dominate trade negotiations. All Developing countries, too, have much to segments of the international community must gain. Middle-income countries—continuing a keep their focus on potential gains, not only process begun over the last two decades—may from “winning concessions” from foreign do well to open selected services markets, often partners, but also on the gains from domestic plagued by inefficiency that dampens produc- reforms that “pay for” foreign concessions. tivity of the whole economy, in exchange for Rich-country negotiators will do better, for greater access in agriculture and labor-inten- themselves and for the developing world, if sive goods. Because many countries have al- they keep in mind that their own countries can ready lowered tariffs, the issue is now to bind benefit by directing agricultural subsidies those new lower levels. Finally, low-income away from production subsidies for large countries would benefit if, in relinquishing de- farmers toward income subsidies to relatively mands for exemption from disciplines on their small family farms, delivered in a form that is own tariffs, they succeed in obtaining commit- decoupled from output. Middle-income coun- ments to greater market access in products and try negotiators likewise have to keep in mind services of importance to them, a new commit- that their telecommunications and financial ment to consistency in the administration of services could be much more efficient and less preferences, and development assistance to fa- expensive if more competitors were allowed cilitate trade and implement new WTO rules in to enter well-regulated markets. Low-income accord with domestic capacities and develop- xxix G L O B A L E C O N O M I C P R O S P E C T S ment priorities. Delivering this type of deal Inama, Stefano. 2003. “Trade Preferences and the would go far toward fulfilling the development WTO Negotiations on Market Access.” Mimeo. UNCTAD. promise of the Doha Agenda. International Monetary Fund (IMF). Balance of Pay- ments Statistics Yearbook. Washington, D.C.: IMF. IMF-World Bank. 2002. “Market Access for Develop- Notes ing Country Exports—Selected Issues.” Washing- 1. Global Economic Prospects 2002: Making Trade ton, D.C.: World Bank, September 26. Work for the World’s Poor (World Bank 2001), in ad- Mattoo, Aaditya. 2003. “Services in a Development dition to analyzing agriculture, labor-intensive manu- Round.” Paper presented to the OECD Global factures, and services, dealt with regulatory impedi- Forum on Trade, OECD, Paris, June 5–6, 2003. ments in transportation (chapter 4) and intellectual Mattoo, A., R. Rathindran, and A. Subramanian. property rights and TRIPS (chapter 5). Global Eco- 2001. “Measuring Services Trade Liberalization nomic Prospects 2003: Investing to Unlock Global Op- and Its Impact on Economic Growth: An Illustra- portunities (World Bank 2002a) analyzed two of the tion.” World Bank Policy Research Working Singapore issues—investment and competition policy— Paper 2655. World Bank, Washington, D.C. from a development perspective (chapter 4). Newfarmer, R. 2003. “From Singapore to Cancún: 2. Global Economic Prospects 2002 presents illus- Investment.” Trade Note 2. Washington, D.C.: trations of the gains from services liberalization. While World Bank. May (www.worldbank.org/trade). we do not have firm estimates of relative parameters, Ravallion, Martin. 1997. “Can High-Inequality Coun- several studies have shown that gains are likely to be a tries Escape Absolute Poverty?” Economics Letters multiple of merchandise liberalization. See World Bank 56(1): 51–57. (2001), chapter 6. Rodriguez, Francisco, and Dani Rodrik. 1999. “Trade Policy and Economic Growth: A Skeptic’s Guide to the Cross-National Evidence.” NBER Working References Paper W7081. NBER, Cambridge, Mass. Bernanke, Ben, and Kenneth Rogoff, eds. 2001. Self, R. J., and B. K. Zutshi. 2003. “Mode 4: Negoti- Macroeconomics Annual 2001. Cambridge, Mass.: ating Challenges and Opportunities.” In Aaditya- MIT Press. Mattoo and Antonia Carzaniga, eds., Moving Dollar, David, and Aart Kraay. 2001. “Trade, Growth, People to Deliver Services. Washington, D.C.: and Poverty.” World Bank Policy Research De- Oxford University Press and World Bank, 2003. partment Working Paper 1615. Srinivasan, T.N., and Jadgish Bhagwati. 2000. “Out- Evenett, S. 2002. “The WTO Government Procure- ward-Orientation and Development: Are Revi- ment Agreement: An Assessment of Current Re- sionists Right?” Macroeconomics Annual 2000. search and Options for Reform.” Paper presented Cambridge: MIT Press. at the roundtable “Informing the Doha Process: Walmsley, T. L., and A. Winters. 2003. “Relaxing the New Trade Research for Developing Countries.” Restrictions on the Temporary Movements of Egypt. May 20–21. Natural Persons: A Simulation Analysis.” CEPR Fink, C. 2003. “Implementing the Doha Mandate on Discussion Paper 3719. London: Center for Eco- TRIPS and Public Health.” Trade Note 5. Wash- nomic Policy Research. ington, D.C.: World Bank. May (www.worldbank. World Bank. 1996. The Chinese Economy: Fighting org/trade). Inflation, Deepening Reforms. Washington, D.C.: Gresser, E. 2002. “America’s Hidden Tax on the Poor: World Bank. April. The Case for Reforming U.S. Tariff Policy.” Pro- ———.2001. Global Economic Prospects 2002: Mak- gressive Policy Institute Policy Report. Washing- ing Trade Work for the World’s Poor. Washing- ton, D.C.: Progressive Policy Institute. March. ton, D.C.: World Bank. Hatcher, Mark. 2003. Draft Model Schedule for Mode ———.2002a. Global Economic Prospects 2003: In- 4: A Proposal. In Aaditya Mattoo and Antonia vesting to Unlock Global Opportunities. Wash- Carzaniga, eds., Moving People to Deliver Ser- ington, D.C.: World Bank. vices. Washington, D.C.: Oxford University Press ———.2002b. World Development Report 2003: Sus- and World Bank, 2003. tainable Development in a Dynamic World. Hummels, D. 2001. “Time as a Trade Barrier.” Mimeo. Washington, D.C.: World Bank. Department of Economics, Purdue University, Lafayette, Ind. xxx Abbreviations and Data Notes ACE Automated commercial environment ACP African Caribbean and Pacific states ACPC Association of Coffee Producing Countries AGOA African Growth and Opportunity Act APEC Asia Pacific Economic Cooperation ASAs Air service agreements ASCM Agreement on Subsidies and Countervailing Measures ASEAN Association of Southeast Asian Nations ASECNA Agency for Air Transport Security in Africa ATC Agreement on Textiles and Clothing ATPA Andean Trade Preferences Act ATPDEA Andean Trade Promotion and Drug Eradication Act CACM Central American Common Market CAP Common Agricultural Program CARICOM Caribbean Community CBERA Caribbean Basin Economic Recovery Acts CBI Caribbean Basin Initiative CBTPA Caribbean Basin Trade Partnership Act CEE Central and Eastern Europe CEPR Center for Economic and Policy Research CIS Commonwealth of Independent States COMTRADE U.S. Commodity Trade Statistics CSI Container Security Initiative C-TPAT Customs-Trade Partnership Against Terrorism DFID Department for International Development DRIFE Danish Research Institute of Food Economics DTIS Diagnostic Trade Integration Study ECB European Central Bank xxxi G L O B A L E C O N O M I C P R O S P E C T S EEA European Economic Area EMBI Emerging Markets Bond Index EU European Union FAO Food and Agricultural Organization FDA Food and Drug Administration FDI Foreign direct investment FTA Free Trade Area GATS General Agreement on Trade in Services GATT General Agreement on Tariffs and Trade GCC Gulf Cooperation Council GDP Gross domestic product GSP Generalized System of Preference GTAP Global Trade Analysis Project HITS Harmonized Tariff Schedule HIV/AIDS Human immunodeficiency virus/acquired immune deficiency syndrome HKSC Hong Kong Shippers Council IADB Inter-American Development Bank ICAO International Civil Aviation Organization ICO International Coffee Organization ICTSD International Center for Trade and Sustainable Development IISD International Institute for Sustainable Development ILO International Labour Organization IMF International Monetary Fund IMO International Maritime Organization INS Immigration and Naturalization Service ISM Institute for Supply Management LDCs Least developed countries LME London Metals Exchange LNG Liquid natural gas MFN Most favored nation NAFTA North America Free Trade Agreement NASDAQ National Association of Security Dealers NAV Non-ad-valorem NIEs Newly industrialized economies OECD Organisation for Economic Co-operation and Development OPEC Organization for Petroleum Exporting Countries PECC Pacific Economic Cooperation Council PPI Purchasing Parity Index xxxii A B B R E V I A T I O N S A N D D A T A N O T E S PROCAMPO Programa de Apoyos Directos al Campo (National Program for Agricultural Direct Support) PSRE Program de soutien à la relance economique QUAD Canada, European Union, Japan, and United States RDS Research, development, and statistics SAARC South Asian Association for Regional Cooperation SARS Severe Acute Respiratory Syndrome SDT Special and differential treatment SMEs Small and medium-sized enterprises SOLAS Safety of Life at Sea Convention SOPEMI Continuous Reporting System on Migration STAR Secure Trade in the APEC Region TMNP Temporary movement of natural persons TN Trade NAFTA TRIPS Trade-related aspects of intellectual property rights TRQs Tariff rate quotas UN/ECE United Nations Economic Commission for Europe UNCTAD United Nations Conference for Trade and Development UNDCP United Nations International Drug Control Program UNECE United Nations Economic Commission for Europe USAID United States Agency for International Development USCS U.S. Customs Service USDA U.S. Department of Agriculture USTR U.S. Trade Representative WCO World Customs Organization WFP World Food Program WITS World Integrated Trade Solutions WTO World Trade Organization Data notes The “classification of economies” tables at covers all lower- and middle-income coun- the end of this volume classify economies by tries, including countries with economies in income, region, export category, and indebt- transition. edness. Unless otherwise indicated, the term “developing countries” as used in this volume All dollar figures are U.S. dollars. xxxiii 1 Global Outlook and the Developing Countries The global economy continues to be weak couraging rebound in most countries of the For the third year in a row the global economy region. Activity in South Asia is holding up in 2003 is growing well below potential, at an well. Countries in East Asia lost some growth expected rate of 2 percent. The global slow- momentum due to SARS, but its apparent con- down that began in 2001 with the bursting of tainment has opened the way to a resumption the equity-market bubble evolved into a sub- of rapid growth. Africa continues to underper- dued recovery during 2002. Initially, the sharp form: although the region’s commodity prices downturn in business investment was a critical have firmed, they are still well below long-term factor behind sluggish growth, as corporations trends. War has affected regional performance worldwide redressed the substantial financial in the Middle East and North Africa; while imbalances that had emerged during the boom many countries in Central and Eastern Europe of the late 1990s. The pace of activity faltered are undergoing sluggish growth tied to lacklus- again at end-2002 and early 2003 in response ter conditions in Western Europe, especially in to events that undermined confidence: the Germany. buildup to war in Iraq, transatlantic tensions, persistent concerns about terrorism, and the Macro policy response has been strongly outbreak of Severe Acute Respiratory Syn- supportive, but it is approaching limits drome (SARS). Consumer and business con- Policymakers, particularly in the United States, fidence waned again—and so did spending. reacted to the slowdown in 2001 with signifi- Manufacturing production as well as GDP cant monetary easing and fiscal stimulus. The growth in the rich countries slowed consider- stimulus and the effects of automatic stabilizers ably at the turn of the year. The momentum of prevented a sharper downturn in the global goods production retrenched to negative terri- economy and helped improve the external en- tory, and G-7 GDP growth braked from an vironment for developing-country growth. annualized pace of 2.8 percent during the But the scope for substantial further ma- third quarter of 2002 to 0.8 percent by the croeconomic stimulus is rapidly dissipating. first quarter of 2003. Fiscal deficits threaten to become part of the Developing countries faced a difficult envi- problem instead of part of the solution, espe- ronment in 2002 to mid 2003. Latin America’s cially since a quick reversal of the deficit is GDP contracted in 2002 because of political not anticipated. The U.S. general government problems in Venezuela, investor concerns budget position (including Social Security), for about Brazil in the run-up to elections, and example, shifted dramatically from a surplus fallout from Argentina’s default. Per capita in- of 2.3 percent of GDP in 2000 to a deficit of comes will barely rise this year, despite an en- 3.2 percent as of the first quarter of 2003. The 1 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Congressional Budget Office projects that the countries cannot be ruled out, as investment budget position is unlikely to return to surplus growth dropped sharply during the first quar- until 2012. In Europe, several large countries ter of 2003. Finally, the U.S. current account have breached the 3-percent-of-GDP fiscal deficit is surpassing historic levels. During deficit limits embedded in the Maastricht cri- 2002, U.S. external financing needs claimed teria for the common currency. And Japan has 10.3 percent of the savings of the rest of the limited fiscal scope, given persistent deficits in world—more than double the levels of 1998. the 6–7 percent range. Interest rates have been Moreover, the composition of finance also brought down sharply in the United States as shifted toward short-term flows: net FDI well as in Japan, where they stand at an effec- flows were negative by almost $100 billion; tive rate of zero. Following the recent 50-basis U.S. banks’ overseas lending had ceased; and point cut in rates, Europe still has modest foreign official inflows (most from East Asia) headroom for monetary easing should the Eu- increased to nearly $100 billion, from $5 bil- ropean Central Bank choose to relax its infla- lion in 2001. A sudden reversal in these short- tion target. In fact, downward price trends in term flows could undercut U.S. and world the United States and Europe have triggered growth. The 25 percent fall of the U.S. dollar concerns of possible deflation. against the euro in the last 18 months repre- sents at least a partial adjustment. Activity should build gradually through 2004–05, but risks remain Structural reforms could boost confidence Barring additional shocks, global growth With scope for additional macroeconomic should pick up to 3 percent in 2004, as firms stimulus fading, the focus of policy in the rich in the rich countries make progress in adjust- countries should arguably shift toward struc- ing balance sheets and begin to upgrade capi- tural reforms that help restore business and tal stock and replenish inventories (table 1.1). consumer confidence. These could include ef- The financial headwinds that have con- forts to resolve the nonperforming loan prob- strained investment are apparently diminish- lem in the Japanese banking system and to ing across the OECD centers. Early signs of re- achieve positive inflation rates there; addressing newed economic activity are appearing in the corporate governance and related issues in the United States—including an upturn in orders, United States, and needed labor market reforms production, and exports, as well as firming in Europe. A rekindling of multilateral consen- equity markets. Yet conditions in Europe and sus on economic policy would also contribute Japan remain extremely slack. Improvement to renewed confidence, which had been shaken in confidence will prove the key to a revival in by geopolitical tensions and security concerns. capital spending and growth. Following an advance of 4 percent in 2003, developing Intensified trade underpins strong countries are likely to grow at 4.9 percent in developing country growth in the long run 2004, grounded in a revival of world trade, One important and ongoing program is the the fading of global tensions, and the rekin- Doha Development Agenda, where progress dling of domestic demand. could do much for near-term sentiment and But risks to the outlook remain. First, the eventually for global growth. Intensified trade pace of stabilization in the Middle East re- relations during the 1990s and the increas- mains uncertain. Second, SARS, though now ingly global nature of production and distri- apparently under control, could reemerge next bution have sharply increased productivity in flu season and would present challenges to tradable sectors and drastically changed trade policymakers worldwide, especially in China. patterns, laying the foundation for future Third, and more broadly, a reversal of the in- growth. Productivity growth in manufacturing cipient investment rebound in the industrial sectors that compete in international markets 2 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Table 1.1 Global growth should accelerate, but risks persist Global conditions affecting growth in developing countries and world GDP Current GDF2003 estimate Current forecasts forecasts 2001 2002 2003 2004 2005 2003 2004 Global conditions World trade (volume) –0.7 3.0 4.6 7.9 7.9 6.2 8.1 Inflation (consumer prices) G-7 OECD countries a, b 1.5 1.0 1.4 0.9 1.4 1.4 1.3 United States 2.8 1.6 1.9 1.2 2.3 2.5 2.3 Commodity prices (nominal $) Commodity prices, except oil ($) –9.1 5.1 6.9 1.1 1.5 8.2 2.3 Oil price ($, weighted average), $/bbl 24.4 24.9 26.5 22.0 20.0 26.0 21.0 Oil price (percent change) –13.7 2.4 6.3 –17.0 –9.1 4.3 –19.2 Manufactures export unit value ($) c –4.5 –0.1 4.0 –0.4 1.5 5.6 –0.1 Interest rates LIBOR, 6 months ($, percent) 3.5 1.8 1.0 2.0 3.8 1.7 3.2 EURIBOR, 6 months (Euro, percent) 4.2 3.4 2.1 2.1 3.1 2.4 2.3 GDP (growth) d World 1.3 1.9 2.0 3.0 2.9 2.3 3.2 Memo item: World GDP (PPP) e 2.3 3.0 3.1 3.9 3.8 3.2 4.1 High-income countries 0.9 1.6 1.5 2.5 2.4 1.9 2.9 OECD countries 1.0 1.6 1.5 2.5 2.3 1.8 2.8 United States 0.3 2.4 2.2 3.4 2.8 2.5 3.5 Japan 0.4 0.1 0.8 1.3 1.3 0.6 1.6 Euro Area 1.5 0.8 0.7 1.7 2.1 1.4 2.6 Non-OECD countries –1.1 2.4 2.1 4.1 4.4 3.0 4.3 Developing countries 2.9 3.3 4.0 4.9 4.8 4.0 4.7 East Asia and Pacific 5.5 6.7 6.1 6.7 6.6 6.4 6.6 Europe and Central Asia 2.2 4.6 4.3 4.5 4.1 3.7 3.7 Latin America and the Caribbean 0.3 –0.8 1.8 3.7 3.8 1.7 3.8 Middle East and North Africa 3.2 3.1 3.3 3.9 3.5 3.7 3.9 Oil exporters 2.9 3.2 3.9 3.9 3.3 4.0 3.7 Diversified economies 3.8 2.8 2.4 3.7 3.8 3.1 4.2 South Asia 4.9 4.2 5.4 5.4 5.4 5.3 5.2 Sub-Saharan Africa 3.2 2.8 2.8 3.5 3.8 3.0 3.6 Memorandum item Developing countries: excluding China and India 1.7 2.0 3.1 4.1 4.1 2.9 3.9 a. Canada, France, Germany, Italy, Japan, the United Kingdom, and the United States. b. In local currency, aggregated using 1995 GDP weights. c. Unit value index of manufactures exports from G-5 to developing countries, expressed in U.S. dollars. d. GDP in 1995 constant dollars: 1995 prices and market exchange rates. e. GDP measured at 1995 PPP (international dollar) weights. Sources: Development Prospects Group, baseline, July 2003 and GDF 2003 forecasts of March 2003. is traditionally 1.5 percentage points higher gered by increased competition on a global than economy-wide productivity growth. This scale. Developing countries as a group have differential has increased to 2.5 percentage benefited from the intensification of trade in points during the last decade. Sharp techno- manufactures and associated productivity logical progress in manufacturing was partly gains, as the share of manufactured goods in an autonomous process—driven by advances their exports increased from 20 percent in in computer technology—but was also trig- 1980 to more than 70 percent in 2001. 3 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Under a set of favorable but plausible as- penditure could not continue to grow at the sumptions, developing countries are expected high rates achieved in the early phase of re- to experience an acceleration of per capita in- covery, though deficits continued to widen. come growth through 2015. The East Asia Late in 2002 and through early 2003, U.S. region is an exception because its already high consumption, a major driver of global de- growth of 6 percent annually over the last mand, slowed from an earlier pace of 4 per- decades will be difficult to maintain, as econo- cent to near 2 percent—partly as a reflection mies mature and the gap with high-income of the dramatic drop in consumer confidence countries narrows, even though it is likely to on the eve of the Iraqi conflict and partly in remain the fastest-growing region in the devel- reaction to high oil prices and weakening of oping world. the dollar. By the first quarter of 2003, GDP growth had slowed from generally stronger Poverty remains a challenge, especially first-half 2002 rates to 1.4 percent (saar) in for Africa the United States, to 0.6 percent in Japan, to Broad acceleration of per capita growth 0.2 in the Euro Area (figure 1.1). would translate into a sharp reduction in the Manufacturing output advances slowed incidence of poverty, from 28.3 percent in discernibly at the turn of the year, and intensi- 1990 to a projected 12.5 percent by 2015, fied during the spring. Growth momentum in meeting, on average, the millennium develop- goods production suffered a “double dip,” to ment goal (MDG) of 14.8 percent. However, stand at –1.2 percent for the Euro Area, –2.0 the gap between strong and weak performers percent for Japan, and –2.3 percent for the will remain large. Even if Sub-Saharan Africa United States as of April–June 2003 (figure could turn falling per capita incomes into an- 1.2). The end to combat in the Iraqi campaign nual increases of 1.6 percent—as assumed in helped to boost U.S. consumer confidence the baseline scenario—its rate of growth would from nine-year troughs reached in March; but be less than one-third the rate of growth that is response of consumers in Japan and especially expected in East Asia. The relatively poor per- in Europe was muted, despite an incipient up- formance of Sub-Saharan Africa makes the MDGs for that region especially challenging. For example, under the baseline scenario the percentage of people living on $1 per day or Figure 1.1 Growth in the OECD countries falters less will be only 42.3 percent in 2015 instead Quarter/quarter, percent change, saar of 24 percent as targeted by the MDGs. 3.5 3.0 The industrial countries: Deficits, 2.5 confidence, capital spending, and the dollar 2.0 2002 Q4 2002 H1 ᮡ 1.5 2003 Q1 Confidence is the key to the long-awaited ᮡ breakthrough to growth 1.0 ᮡ The high-income OECD countries have faced 0.5 substantial difficulties in overcoming the lega- cies of the second half of the 1990s, including 0.0 the equity market downturn. The recovery that –0.5 began in early 2002 faltered after the summer United States Japan Euro Area of that year as the rebound in investment Sources: National agencies and Eurostat. showed signs of weakness. Government ex- 4 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Figure 1.2 OECD manufacturing shows a distinct “double dip” Manufacturing IP, 3-month/3-month, percent change, saar Japan 15 ᮡ Euro Area 10 ᮡ 5 0 ᮡ –5 United States –10 –15 –20 Jan. April July Oct. Jan. April July Oct. Jan. April July Oct. Jan. April 2000 2000 2000 2000 2001 2001 2001 2001 2002 2002 2002 2002 2003 2003 Sources: National agencies and Eurostat. turn in equity markets. Rather, focus returned States, a contribution of –1.0 percentage to the set of weak fundamentals underlying points during the first half of 2002 changed sluggish growth in the rich countries—notably into one of +0.9 percentage points in the first substantial debt overhangs in the U.S. corpo- quarter of 2003. The opposite occurred in Eu- rate, household, and, increasingly, government rope and Japan. There, during the first half of sectors. 2002, net exports added respectively 0.9 and The sharp depreciation of the dollar began 1.4 percentage points to GDP growth, but to yield shifts in the contribution of net ex- these rates turned around in the first quarter ports to GDP growth (table 1.2). In the United of 2003, to –1.8 and –0.2 percentage points. Table 1.2 Weak fundamentals underlie sluggish growth in the rich countries Recent developments in GDP and components, United States, Euro Area, and Japan (percent) United States Euro Area Japan Growth H1-02 H2-02 Q4-02 Q1-03 H1-02 H2-02 Q4-02 Q1-03 H1-02 H2-02 Q4-02 Q1-03 GDP 3.5 2.7 1.4 1.4 1.0 1.1 0.3 0.2 0.9 3.0 1.5 0.6 Private consumption 3.5 3.0 1.7 2.0 0.0 1.6 1.4 1.7 2.0 1.5 –0.2 0.8 Fixed investment –2.8 0.7 4.4 –0.2 –3.2 –0.9 0.9 –4.8 –5.3 1.0 3.2 –1.9 Government 5.7 3.0 4.6 0.4 3.3 2.0 1.2 1.4 2.4 1.7 0.4 2.5 Growth contributions Private consumption 2.4 2.1 1.2 1.4 0.0 0.9 0.8 1.0 1.1 0.9 –0.1 0.4 Investment 1.0 0.9 1.0 –0.9 –0.5 –0.2 0.7 0.8 –2.0 1.5 0.1 0.0 Fixed capital –0.5 0.1 0.7 0.0 –0.7 –0.2 0.2 –1.0 –1.4 0.3 0.8 –0.5 Change in stocks 1.5 0.7 0.3 –0.9 0.2 0.0 0.5 1.8 –0.6 1.2 –0.7 0.5 Government 1.0 0.5 0.8 0.1 0.7 0.4 0.2 0.3 0.4 0.3 0.1 0.4 Net exports –1.0 –0.9 –1.9 0.9 0.9 0.0 –1.4 –1.8 1.4 0.4 1.6 –0.2 Note: H=half year; Q=quarter year. Sources: National agencies, OECD, and World Bank data. 5 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Although the depreciation of the dollar and signs of revival in the U.S. economy are now turnaround in U.S. exports was a movement emerging, but the worst may not be over for toward more balanced conditions, weak de- Europe and Japan. mand in external markets now poses special challenges for Europe and Japan. For the lat- Consumer spending has slowed ter, the near-term effects of SARS in East Asia From late 2002 to early 2003, anticipation will likely exact an additional toll on exports. of war in Iraq provided a “non-economic” Recently, the issue of deflation has emerged overlay to the fundamental factors dampening in the United States, and more so in Europe in consumer sentiment. Consumer confidence the wake of 25 percent currency appreciation plunged to near-record lows on both sides of and flat output growth there. Labor market the Atlantic in the months leading up to war conditions continued to deteriorate across the (figure 1.3). Though the estimated sensitivity of OECD centers, with expectations widely held personal spending to changes in sentiment is for a more prolonged period of sub-par growth surprisingly small, the relationship suggests that in the global economy. for the United States, the decline in confidence From this set of initial conditions, a break- since late 2000, tied to economic conditions through to stronger growth will hinge on a and war jitters, yielded a fall-off in consump- restoration of confidence among consumers tion of a cumulative 1.2 percent, or $75 billion and businesses. Under these circumstances, and (box 1.1). Recovery of U.S. sentiment since that in an environment of low interest rates (cou- time, boosted by incipient gains in equity mar- pled with moderate gains in financial markets), kets, has been moderately encouraging. How- the standard factors that boost investment af- ever, the Conference Board index flattened-out ter a recession—increasing obsolescence of the in May and fell in June (to 83.5) to bring the capital stock, improved expectations for de- measure to its level of late Fall 2002. European mand—are likely to yield a gradual upturn in confidence has recovered little, as economic the pace of investment growth, and with it a re- growth slows and developments in Germany in sumption of economic recovery. Indeed, some particular have deteriorated markedly. Figure 1.3 Consumer confidence recovers from pre-war lows Conference Board (U.S.) and EU Commission surveys of consumer confidence 150 2 0 140 –2 European Union (right scale) 130 ᮡ –4 120 –6 –8 110 –10 100 –12 90 –14 ᮡ United States (left scale) –16 80 –18 70 –20 60 –22 January July January July January July January July 2000 2000 2001 2001 2002 2002 2003 2003 Source: Conference Board, European Commission. 6 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Box 1.1 Consumer confidence and U.S. private consumption T he U.S. consumer accounts for more than two- thirds of U.S. and 20 percent of world expendi- ture, and has contributed nearly one-third to total Actual and predicted confidence effects of the build-up to war world GDP growth since the onset of slower growth 120 in 2001. Consumers seemed able to shake off the Predicted depressing effects of lower confidence during this 110 ᮡ period. The bursting of the stock market bubble, with its contractionary effects on household wealth, 100 slowed consumption and growth and revealed sub- 90 stantial overinvestment in telecommunications and other high-technology industries. These develop- ᮡ 80 ments left confidence levels vulnerable to short-term Actual events, which came in the form of the Iraq war. 70 The Conference Board’s index of consumer con- 60 fidence was down 20 percent on average in the first 01 01 .2 1 1 02 02 02 2 03 03 03 0 00 00 eight months of 2001 and dropped another 25 per- 20 20 20 20 20 20 20 20 20 .2 n. ril ly n. ril ly n. ril ly ct ct Ju Ju Ju Ap Ap Ap Ja Ja Ja cent after September 11. Yet in the fourth quarter of O O 2001 consumers flocked back to the malls in re- Source: World Bank, Development Prospects Group. sponse to generous incentives and spending surged by 6 percent (saar). By mid-2002, confidence had recouped most of the losses suffered after September, turned out to be short lived. But, starting in the but as war in Iraq seemed more likely, confidence fourth quarter of 2002 and first quarter of 2003, tumbled again, dropping a further 45 percent by economic factors began yielding a sustained and sub- March 2003, when military action began. stantial over-prediction. The mean squared predic- Economic and noneconomic determinants of tion error is more than five times larger after Septem- confidence. The impact of noneconomic factors such ber 2002 than before—compelling evidence of the as ‘terrorist threats’ and ‘war jitters’ on consumer war jitters story. By March 2003, when the war confidence and spending can be significant. More- began, confidence was only two-thirds of what eco- over, their dynamics are different from traditional nomic conditions alone would have indicated—the economic factors insofar as they can appear and re- biggest discrepancy since the first Gulf War in 1991. verse suddenly. For instance, the Conference Board’s In April, however, confidence rebounded sharply, index soared 32 percent in April 2003 after a quick narrowing the discrepancy to only 9 percent. resolution in Iraq. The impact on consumer spending. Many stud- The decomposition of consumer confidence into ies have found that variations in consumer confi- economic and non-economic components is not dence help to explain aggregate consumer spending, straightforward. By definition, the economic compo- above and beyond what household disposable in- nent should track coincident and leading indicators come and wealth can predict alone, though the ef- such as unemployment, household debt, and stock fects are relatively small. A simple regression carried prices. However, objective proxies for war jitters out for the purposes of this study yields a typical re- or perceptions of terrorist threats are problematic. sult, which implies that a 1 percent increase in confi- Thus, instead of measuring the role of non-economic dence would raise real consumption growth by .023 factors directly, we take the indirect route of regress- percent over the subsequent year. Using this estimate, ing confidence on a set of economic variables and a cumulative 1.2 percent reduction in consumption, interpreting the residual as representing the non- or around $75 billion since late 2000, is attributable economic component.1 The results are illustrated in to the fall in confidence. the figure. Note the large negative residual that ap- peared after September 11, 2001, although this (Box continues on next page) 7 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Box 1.1 (continued) What happens next? The equation predicts that than given up this gain. The reason? Soaring unem- the 30 percent improvement in confidence after the ployment. It was not until 1993 that U.S. consumption war, if sustained, could add as much as an additional again recovered to near its long-run average growth 30*.023 = 0.7 percent, or around $45 billion to con- rate. Moreover, our model attributes a cumulative 1.2 sumption levels over the next year. It is likely that percent reduction in consumption since 2000 to confi- durables such as housing and automobiles would be dence, but 9.2 percent to changes in income and the main beneficiaries. But the experience after the wealth. In short, the future course of employment, in- 1991 Gulf War sounds a note of caution. Then, too, come, and asset prices will be by far the most impor- confidence dipped sharply as the war approached, and tant determinants of consumption spending. surged by 36 percent in March 1991 following the ceasefire. Less than a year later, however, it had more Source: World Bank staff. Since 2000, spending in the rich countries of 2000—mitigated to $4.4 trillion by real es- has been buffeted by divergent trends in net tate appreciation. Moreover, low interest rates worth; most households suffered sharp de- have encouraged widespread refinancing and clines in the value of financial assets. In the equity cash-outs to supplement flows of per- United States and the United Kingdom, how- sonal income and spending. On the other hand, ever, rapid appreciation of housing values has rapid buildup of mortgage and other consumer partially offset the deterioration in financial debt has placed U.S. households in an exposed worth (figure 1.4). For example, U.S. house- position should interest rates begin to rise with holds suffered loss of $7.7 trillion in net finan- eventual recovery. At record levels of $8.9 tril- cial worth since peak levels of the first quarter lion, representing 80 percent of GDP or 111 percent of disposable income as of the first quarter of 2003, burgeoning household debt Figure 1.4 The drop in U.S. household net could hamper vigorous spending responses in worth has been offset by real estate later stages of the anticipated recovery. appreciation Trillions of dollars 40 And the investment rebound relapsed The sluggish pace of economic recovery is also 35 linked to hesitant patterns of business capital 30 spending, common across the industrial coun- tries. After picking up to a 2.5 percent pace in 25 Households’ net financial worth the final quarter of 2002, G-7 fixed investment 20 relapsed to growth of just 0.3 percent during 15 the first quarter of 2003. U.S. business outlays dropped by a disappointing 3.4 percent, capital 10 spending in Japan fell by 2 percent, and Euro 5 Value of real estate holdings Area investment plummeted by 4.8 percent, as 0 that in Germany fell 6.8 percent (figure 1.5). At midpoints of the business cycle, the 95 96 97 98 99 00 01 02 03 prime mover for growth would normally tran- 19 19 19 19 19 20 20 20 20 First quarter sition from consumer spending and inventory Source: Federal Reserve Board. building to more robust advances in fixed cap- ital formation. Unlike in previous business cy- 8 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Figure 1.5 Capital spending has been hesitant in all industrial countries Real fixed investment, quarter/quarter, percent change, Q/Q, saar 4 Euro Area 2 ᮡ 0 –2 –4 –6 –8 –10 ᮡ –12 ᮡ Japan –14 United States –16 2001 Q1 2001 Q2 2001 Q3 2001 Q4 2002 Q1 2002 Q2 2002 Q3 2002 Q4 2003 Q1 Sources: National agencies and Eurostat. cles, however, several factors have combined to growth in capital spending, at least for the to inhibit a vigorous rebound in investment. United States—and possibly for Japan—may Corporate debt legacies of the 1990s’ boom be approaching. continue to curtail investment plans across the Business sentiment is now displaying dis- OECD. For example, U.S. non-financial cor- tinct divergence between the United States and porate debt as a proportion of GDP rose from 38 percent in 1995 to 47 percent in 2002. Re- Figure 1.6 Corporate profits have risen cent data, however, show U.S. debt rates stabi- moderately in the United States and Japan lizing over the last quarters of 2002 and the Adjusted profits in billions of dollars (2Q-ma) and trillions first of 2003—a positive sign that adjustment of yen (3Q-MA) efforts are beginning to yield fruit. Corporate 750 11,000 profits fell 10 percent in the United States in United States 700 [left scale] 2001, while those of Japanese manufacturers 10,000 ᮡ plummeted 48 percent, reflecting onset of re- 650 cession. The tortuous road toward restoration 600 9,000 of profit growth has involved substantial cuts 550 to capital spending and to employment over 8,000 the last years. However, there is evidence that 500 profits are staging recoveries in the United 450 7,000 ᮡ States and Japan, as well as in several sectors Japan of industry in Europe. U.S. profits enjoyed a 15 400 [right scale] 6,000 percent rebound in 2002, but growth eased 350 to a 5 percent pace in the first quarter of 2003. 300 5,000 In Japan, profits of major manufacturers ex- 1 1 3 1 3 1 3 Q perienced a substantial comeback, rising 38 Q Q Q Q Q Q 94 95 97 98 00 01 03 19 19 19 19 20 20 20 percent in recent quarters (y/y)—despite nu- merous challenges (figure 1.6). These develop- Sources: Department of Commerce, Bank of Japan, ESRI. ments offer additional evidence that the corner 9 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Figure 1.7 Business confidence remains poor, but better in the United States than in Europe European and U.S. business confidence, January 2000–June 2003 62.5 62.5 60.0 Euro area manufacturing PMI [right scale] 60.0 ᮡ 57.5 57.5 55.0 55.0 52.5 52.5 50.0 50.0 47.5 47.5 45.0 45.0 42.5 42.5 ᮡ ISM manufacturing index [left scale] 40.0 40.0 Jan. June Nov. April Sept. Feb. July Dec. May 2000 2000 2000 2001 2001 2002 2002 2002 2003 Sources: ISM and Reuters (Euro Area PMI). Europe and Japan, where assessments of con- government’s general budget position shifted ditions have worsened. The ISM survey cover- dramatically from a surplus of 2.3 percent of ing U.S. manufacturing and non-manufacturing GDP in 2000 to a deficit of 3.2 percent during sectors fell below the 50 percent line that di- the first quarter of 2003. In part this reflects vides expansion from contraction during March reduced revenues associated with sluggish ac- and April—before manufacturing rebounded tivity and the operation of automatic stabiliz- to 49.8 in June—and services sharply to 54.5 ers. But the shift to deficit has been more pro- in May. In contrast, the composite PMI for the nounced due to tax reduction and funding of Euro Area entered the “contraction zone” in the Iraqi campaign at the federal level, and by March, and fell further to 48.1 in June, reflect- increasing shortfalls at the state and local lev- ing declines in both services and manufactur- els (figure 1.8). Before the recent additional ing indicators (figure 1.7). It appears that $350 billion tax reduction was enacted, the financial conditions weighing against capital Congressional Budget Office projected that spending are easing across the OECD, and the federal on-budget fiscal position was un- business sentiment is now reviving in the likely to return to surplus until 2012. Persis- United States. Yet a key uncertainty persists: the tent deficits of such magnitude carry the po- robustness of future demand, which is strongly tential to constrain growth in the medium tied to the effects of policy. term and beyond, largely as long-term interest rates rise in response to much-increased sup- After providing substantial stimulus, ply of Treasury securities. economic policy is reaching limits On both fiscal and monetary fronts, policy- Fiscal deterioration and the current account makers in the rich countries injected signifi- The deterioration of the fiscal position has cant stimulus, first to limit the global eco- played a role in the doubling of the U.S. current nomic downturn of 2001, and over the last account deficit from 2.3 percent of GDP in 18 months to foster conditions conducive for 1998 to 4.6 percent in 2002. The public-sector stronger recovery. On the fiscal front, the U.S. financial balance (saving less investment) 10 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S investment), although substantial, was far from Figure 1.8 The U.S. fiscal deficit is sufficient to make up the gap (figure 1.9a). widening quickly Aside from the underlying savings-investment U.S. federal, state, and local current fiscal balances, imbalance that drives the need for foreign cap- 1999–2003 (billions of dollars) ital, the widening of the U.S. external deficit 250 is attributable to several factors. Booming do- 200 mestic conditions during the second half of the 150 1990s attracted imports at a growth rate of 10 100 Federal balance percent annually, while the strong dollar and 50 0 only moderate growth in U.S. trade partners –50 served to constrain export growth to 6 percent. –100 In turn the trade balance deteriorated from a –150 deficit of $175 billion in 1995 to $485 billion –200 in 2002. Foreign capital flows funding the cur- ᮡ –250 rent account deficit have become increasingly State/local balance –300 volatile over the last years, shifting rapidly in –350 composition and in region of origin. The ef- fects of sustaining such capital inflows over an 1 3 1 3 1 3 1 3 1 Q Q Q Q Q Q Q Q Q 99 99 00 00 01 01 02 02 03 extended period remain a question of some 19 19 20 20 20 20 20 20 20 Source: Department of Commerce. concern (box 1.2). The present level of the current account deficit, at 5.1 percent of GDP in the first quar- shifted from balance in 1999 to deficit of 3.6 ter of 2003, is an historic record. Of particu- percent of GDP in 2002, and further to a short- lar concern is the unprecedented level of deficit fall of 4.3 percent in the first quarter of 2003. occurring at an early phase of economic recov- At the same time, improvement in the private- ery, when external positions are normally closer sector balance (largely through compression of to balance given only moderate changes in Figure 1.9 The U.S. current account deficit is at record levels a. The public sector balance is driving the deficit Percent of GDP 2 Public-sector financial balance 1 ᮡ 0 –1 –2 –3 –4 ᮡ –5 Private sector ᮡ financial balance Current account balance –6 2000 Q1 2000 Q3 2001 Q1 2001 Q3 2002 Q1 2002 Q3 2003 Q1 Source: U.S. Department of Commerce. (Figure continues on page 13) 11 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Box 1.2 Financing the U.S. current account deficit: From equity to debt A s the U.S. current account deficit more than dou- bled in dollar terms between 1998 and 2002 to reach $480 billion, U.S. calls on international finan- U.S. external financing requirements and net financial flows Billions of dollars cial markets increased in tandem. The country’s ex- 700 700 ternal financing requirement climbed to 4.6 percent FDI U.S. financing 600 requirements 600 of GDP in 2002, representing some 10.3 percent of ᮡ ᮡ 500 500 the savings of the rest of the world. U.S. financing Treasury bills and 400 other official debt 400 requirements largely have been smoothly met by net Bonds ᮡ inflows from abroad. But the form of financing (that 300 300 is, the composition of the net foreign-asset flow) has 200 ᮡ 200 changed dramatically over just the last three years, 100 100 while the region of origin of the inflow has shifted as 0 0 well. The underlying nature of the capital inflow can, ᮡ –100 –100 ᮡ Banking Equities as demonstrated aptly in recent years, lead to ques- –200 –200 tions of medium-term sustainability of the external 1996 1998 2000 2001 2002 deficit; while shifts in origin of flows can influence Source: U.S. Department of Commerce. market expectations for adjustment in the value of the dollar against its major partner currencies. During 2000, the final year of the boom, the The source of inflows to U.S. equity and “corpo- United States required net inflows of some $457 bil- rate” bond markets shifted over recent years (second lion to cover its current account deficit. Financing figure). Although the United Kingdom’s status as a was readily available, through increased mergers-and- major financial hub for the European region domi- acquisitions activity that yielded net purchases of cor- nates flows, a discernible shift is clear: from “other porate bonds of $281 billion; complemented by Western Europe” (Euro Area) and Japan as centers strong $160 billion in FDI flows and some $85 bil- of demand for U.S. securities to “other countries” lion attracted to 45 percent gains in NASDAQ (see (East Asia and Latin America). During 2000, the first figure). Together the equity component of flows Euro Area accounted for 25 percent of total net (FDI and stocks) accounted for more than 50 percent of requirements—a strong fundamental position. The onset of recession in the United States during 2001 Net flows into U.S. private securities by was clearly a transition year for investor perceptions. region of origin Billions of dollars The fall of equity markets (NASDAQ down 50 per- 450 cent) forced a compression of the highly diversified 400 Other countries flows of 2000, as equity and FDI dropped to negligi- Japan 350 ble amounts and investors flocked into debt instru- Shares % 300 ments. Long-term debt-related securities almost cov- Other 250 Western Europe ered the U.S. requirement of some $420 billion in the 35% 200 year. The transition in composition of inflows 150 17 evolved more fully in the difficult environment of 5 100 2002. FDI recorded net outflow of almost $100 bil- United Kingdom 50 43 lion; purchases of Treasuries by risk-averse investors 0 increased by $100 billion; a virtual cessation in over- 1996 1998 2000 2001 2002 seas lending by U.S. banks yielded a net buildup in Source: U.S. Department of Commerce. bank liabilities of $70 billion; and foreign official as- sets (largely East Asian) increased by $90 billion from an inflow of $5 billion during the previous year. (Box continues on next page) 12 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Box 1.2 (continued) foreign purchases, falling to 5 percent in 2002—a virtuous” circle may result, as sources of finance be- large decline in relative demand for dollar-based as- come less stable. Greater volatility may be imparted sets, tending to boost the value of the euro; in con- to the dollar, leading in turn to higher and more trast Japan’s share increased from 9 to 17 percent, volatile interest rates. When combined with percep- while East Asia and Latin America almost doubled tions of a persistently growing current account their shares in purchases from 19 to 35 percent by deficit, the confidence of foreign investors in dollar- 2002. Such a shift in the origin of funds may set the denominated assets may be adversely affected. stage for future currency movements. Implications for the availability of finance for In looking ahead, the downshift in equity, per- developing countries also arise from the increasing sistence of large debt flows and dependence on U.S. share of international capital flows. Although volatile official reserve assets for some 20 percent of “crowding out” has yet to be in clear evidence, the coverage, and the fall of the U.S. dollar against the potential for it exists should deficits continue to rise. euro, augur poorly for the inherent stability of fi- The U.S. current account appears to have increased nancing in the short to medium terms. A “less than as a source of financial risk for the global outlook. demand (figure 1.9b). In looking forward, with economic activity, an expected improvement in continued requirements for massive net inflows the goods balance should build, bringing about of foreign capital, the U.S. income-account medium-term stabilization of the deficit in dol- deficit is likely to widen, while terrorism-related lar terms, declining as a proportion to GDP. issues may continue for an extended period, Member states of the Euro Area have strug- pressuring U.S. services and net transfer posi- gled to cope with the requirements of the tions. Yet assuming a gradual buildup of global Stability and Growth Pact of the Maastricht Figure 1.9 (continued) b. The magnitude of the current account deficit is unprecedented for an early recovery phase of the cycle Billions of dollars 100 Services/Income/Transfers ᮡ 0 —100 ᮡ ᮡ —200 Merchandise Current account balance —300 —400 —500 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 Source: U.S. Department of Commerce. 13 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Figure 1.10 Market interest rates have dropped Benchmark U.S. and European interest rates, 2001–2003 (percent) 7 6 U.S. 10-year ᮡ 5 4 ᮡ 3 EURIBOR ᮡ 2 U.S. LIBOR 1 00 00 00 00 00 0 00 01 01 01 01 00 1 01 02 02 02 02 00 2 02 03 03 03 03 . 20 20 20 20 .2 . 20 . 20 20 20 20 .2 . 20 . 20 20 20 20 .2 . 20 . 20 20 20 20 an rc h ay ul y ep t ov an rc h ay ul y ep t ov an rc h ay ul y ep t ov an rc h ay ul y J a M J S N J a M J S N J a M J S N J a M J M M M M Source: Datastream. Treaty, which, to set a foundation for the sin- for those falling into unemployment. Japan’s gle currency, limits fiscal deficits to 3 percent public-debt burden, however, at some 150 of GDP and outstanding government debt lev- percent of GDP, against the background of a els to about two-thirds of GDP. During 2004 rapidly aging demographic profile, suggests Italy is anticipated to join France, Germany, that a required movement toward balance in and Portugal in the group of countries that current budget terms will present considerable have breached the 3 percent limit. The consol- challenges to policymakers. idated balance for the Euro Area has deterio- In monetary policy, authorities in the in- rated by 1.5 points of GDP over the period dustrial countries have been aggressive—more since early 2000, while that for Germany has so in the United States, less in Europe (figure shifted markedly from a 1.1 percent surplus to 1.10). While operating with effective policy a 3.8 percent deficit. For the latter, especially, interest rates at zero, the Bank of Japan (BOJ) the tightening of policy to return to within the has proposed a wide range of policy measures limits by 2005–06 is inopportune because eco- aimed at easing the deflationary state of the nomic conditions have deteriorated sharply; economy. But evidence to date suggests that fiscal “austerity” only adds to these tendencies. little has been achieved in this area. Aggressive Japan’s fiscal stance continues to deterio- easing of policy rates by the Federal Reserve, rate, moving from a deficit of 6 percent of from 6.5 percent at end-2000 to 1 percent GDP in 2000 to 7 percent during 2002. Al- at present, clearly supported consumer and though public-sector investment outlays asso- housing-related activity, though with little ciated with public works and other traditional apparent effect on business capital formation. supplementary budget measures have been cur- Consumption of durable goods, especially autos, tailed in recent years (falling by a cumulative has found support from record low interest 4.5 percent since 2000), more public monies rates. And mortgage refinancings have been have been allocated to shoring up the banking spurred to new highs, placing extra disposable system and underpinning social safety nets income into consumers’ pockets. 14 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S The European Central Bank (ECB) has adopted a more conservative approach to pol- Figure 1.11 Is deflation a danger for icy ease, given its central mandate of control- Europe and the United States? ling inflation for the group of EMU countries. Core CPI measures, percent change, year-on-year The ECB repurchase rate has been lowered in 3.0 United States ᮡ steps from 4.75 percent in late 2000 to 2 per- 2.5 cent at present. The rapid appreciation of the Euro Area 2.0 euro has placed downward pressure on import ᮡ prices, which eventually will compress prices 1.5 at the consumer level, thereby creating an op- 1.0 portunity for the ECB to lower interest rates ᮡ somewhat further. Policy in Japan has been 0.5 Germany geared, first, to lifting the economy from its 0.0 deflationary state—which began in 1998 in the –0.5 Japan aftermath of the East Asian financial crisis and ᮡ has intensified since—and to redressing the –1.0 tenuous conditions in the commercial banking –1.5 sector. Core consumer price inflation in Japan Jan. July Jan. July Jan. July Jan. July 2000 2000 2001 2001 2001 2002 2003 2003 has ranged between –0.5 and –1.5 percent over 2000–03, defying the BOJ’s attempts to sup- Source: Datastream. port higher price levels through various mea- sures. Broader money remains stagnant, as the decline in commercial bank lending effectively deflation can be dangerous when it reflects de- cancels intermediation through the economy. mand shocks, such as the bursting of an asset By intervening in the non-performing loan price bubble. These have the potential to set in market (NPL), the BOJ now intends to expand motion a downward spiral of falling prices, de- the number of channels available to inject li- layed consumption, and declining profits and quidity into the system, while supporting work- production. outs of bad loans. The challenges are daunting. Monetary authorities should shift their fo- cus more decisively, if they have not done so Deflation risk? After two decades of disinfla- already, from avoiding inflation to maintain- tion in high-income countries, set off in the ing low inflation, fighting deflation as aggres- early 1980s by monetary tightening to curb sively as they would high inflation. Asymmet- self-enforcing inflationary pressures, a spread ric price targets and careful monitoring of of deflation has become a real risk. While output gaps are important ingredients in such Japan is experiencing its fourth consecutive a policy. Output gaps for the rich countries year of falling prices, core-CPI inflation fell to turned negative as of 2001 and have widened 1.7 percent in the Euro Area and eased to 1.5 precipitously over the last two years (figure percent in the United States in June (figure 1.12). OECD estimates place the U.S. output 1.11). As these measures of inflation probably gap at 1.5 percent in 2002, rising to 2.1 per- overstate “true” inflation by 0.5–1 percentage cent during 2003. And the gap in Germany points, the world’s major economies could be is particularly severe, shifting by a full point balanced on the edge of deflation. from 1.3 to 2.3. Deflation need not be damaging if it reflects However, as fiscal stimulus becomes gains from deregulation, technological ad- quickly ineffective in the fight against stub- vances and rapid productivity growth. Today, born underperformance, so, after time, can the computer and telecoms revolution, along monetary policy. Both fiscal and monetary with globalization, is pushing down prices. Yet stimulus may at some point become part of 15 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 The dollar’s fall means near-term pressure Figure 1.12 Output gaps are widening, on growth, medium-term benefits bringing deflationary pressures to bear Recent developments in foreign exchange Deviation of actual from potential GDP as proportion markets, reflecting an intensification of the of potential dollar’s decline from early 2002 peaks, have 0.5 2002 2001 2003 come to influence trends in OECD trade growth and the pace of economic activity. ᮡ ᮡ ᮡ 0.0 Over the period, the dollar has fallen by 12.5 –0.5 percent on a real trade-weighted basis. But the 25 percent fall of the dollar vis-à-vis the –1.0 euro—linked in part to diminishing European demand for dollar-denominated financial as- –1.5 sets—is especially of concern to the future pace of growth in Europe (figure 1.13). –2.0 Euro-Area export prices expressed in dollar terms rose by 2.5 percent during 2002, but –2.5 with the dollar’s rapid fall they are anticipated United Euro Germany Japan States Area to rise by nearly 12 percent in 2003. Growth of German and French exports fell sharply Source: OECD data and projections. to negative territory (10 percent decline) as of early 2003. While diminished impetus from exports will detract from GDP advances, the the problem instead of the solution. High fis- stronger euro should imply a rise in real in- cal debt could curb a recovery, and low in- comes, potentially supporting domestic de- terest rates for too long could encourage the mand. The 11 percent depreciation of the yen- accumulation of excessive debt that could in- crease debt service burdens once the economy Figure 1.13 The dollar has fallen sharply starts recovering again. To the extent that mar- since early 2002 ket participants position their balance sheets Euro and yen per USD and REER, around the expectation of low rates for an ex- January 2002 = 100 tended period, then the risk of sudden, violent 105 movements in bond (and possibly currency) REER markets increases, not unlike what happened 100 in 1993–94. To avoid that underperformance, policies 95 that address the structural foundations of economies become increasingly more impor- 90 ᮡ tant than macro stimulus. Such policies can Yen per dollar not only address specific inefficiencies, but 85 also provide a boost to confidence. A break- ᮡ 80 through in the Doha trade negotiations (see Euro cents per dollar the end of this chapter) is an example of such 75 structural improvements. Other examples are 02 2 02 02 2 2 03 3 03 00 00 00 00 the strengthening of corporative governance, 20 20 20 20 20 .2 .2 .2 .2 n. ay ly n. ay ar pt ov ar Ju Ja Ja M M Se especially in the United States; labor and prod- M M N uct market reforms in Europe; and decisive Source: JP Morgan Chase, Datastream. elimination of bad loans in Japan. 16 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S dollar rate poses complications, in a manner The risk of weaker growth is substantial, how- similar to Europe, for sustaining economic ac- ever, given deterioration of conditions in Ger- tivity in Japan. Momentum underlying Japanese many. Short-to-medium-term growth in Japan exports plummeted from 35 percent growth will be conditioned to a large degree by de- during spring 2002 to decline in early 2003, velopments in export markets, especially in tied to a sharp 35 percent fall-off in shipments broader East Asia; the competitiveness of the to the United States and slowing exports to yen; and the success or failure of policy in mit- Asia, from 45 to 3.5 percent rates. Authorities igating the deep-seated problems in the bank- have been active in market intervention to slow ing system. the pace of dollar depreciation. Some positive developments have become Ongoing realignment of currency values, apparent in the U.S. landscape. If sustained, however, will carry medium-term benefits. Re- these developments may augur well for gradual cent moves are in a balancing direction, as the recovery. Since September 2002, consumer con- strength of the dollar over most of the 1990s fidence has recovered moderately, while the raised the question of overvaluation against a Dow Jones index has recouped some 20 per- longer-term equilibrium rate for the currency cent of its value (figure 1.14a). Manufacturing based on fundamental factors. The strong dol- sentiment has improved and production risen, lar contributed to the massive U.S. external while sentiment for services has returned to its imbalance, but at the same time maintained early fall levels. A possible upturn in the high- downward pressures on inflation, allowing tech/semiconductor cycle has been underpinned room for substantial interest rate reductions. by near 30 percent gains in orders for comput- A medium term boost to U.S. competitiveness ers and communications equipment. And the may be expected, provided that the dollar sta- effects of dollar depreciation may be coming bilizes or moves only gradually. Still, prospects into evidence as export momentum breached for stabilization of the current account deficit positive territory with a 3 percent advance from remain at some risk. April to May. Yet the pace of consumer demand remains cloudy, and evidence for incipient re- Expansion is likely by 2004 covery in capital spending is sketchy. As additional impetus to growth from policy Against this background, the projections action is anticipated to be small, recovery posit a subdued 1.5 percent growth outturn in among the rich countries will hinge on the ef- 2003 for the high-income OECD countries, fects of policy actions taken to date, and im- marking a 0.1 point deterioration of growth portantly, a restoration of confidence. The mo- from 2002, largely because of sluggish con- mentum of economic activity will likely build ditions in the United States and particularly over 2003, reaching more robust proportions in Europe during the first half of the year. The by 2004. Recovery should be paced by ad- advance in aggregate GDP is anticipated to rise vances in the United States and by a return to to 2.5 percent by 2004—closer to long-term stronger domestic demand in the emerging and averages—as momentum builds in U.S. invest- high-income economies of East Asia. Revival ment and output growth achieves 3.4 percent of growth in the Euro Area will be tortuous, rates there. A revival in world trade toward 8 lagging that of the United States, and may find percent growth should help buoy performance its focus in domestic rather than external de- in Europe and Japan. Interest rates will begin mand. A lower dollar will help European con- to respond to improved economic prospects, sumers—as real incomes rise with lower infla- leading U.S. growth to taper off moderately tion, and a virtuous cycle of stronger business into 2005, while Europe and Japan continue to output and eventually more substantial in- register improvement, given the later start of creases in capital expenditure could emerge. their recoveries (figure 1.14b). 17 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Figure 1.14 OECD recovery begins in the United States a. Signs of revival in the U.S. economy Index: Sept. 2002 = 100 130 Latest ᮡ 120 March 2003 110 Sept. 2002 ᮡ ᮡ 100 90 80 70 60 Consumer confidence Dow Jones ISM Computers and manufacturing communications orders Sources: Conference Board, Dow Jones, Institute for Supply Management, and U.S. Department of Commerce. b. OECD recovery to build through 2004 Percent 3.5 3.0 2.5 2005 ᮡ 2004 2.0 ᮡ 1.5 2002 2003 1.0 ᮡ ᮡ 0.5 0.0 OECD United States Japan Euro Area Source: World Bank, DECPG. But downside risks have not disappeared will continue to present challenges to policy- Some of the global uncertainties related to makers worldwide (box 1.4). However, these noneconomic events—the war in Iraq and uncertainties do not constitute the most seri- SARS—have eased, but have not completely ous risk to the global recovery. More impor- disappeared. Prospects for eventual stabiliza- tant risks are to be found in the high-income tion and resolution of developments in the countries. Middle East remain uncertain, and the poten- A noteworthy feature of current global tial for adverse events remains, while SARS prospects is that many of the downside risks 18 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S originate in high-income countries, and that a agement decisions in emerging markets. Over- strong recovery is not predicted for any of all, the environment of the last few years has these countries in the near term. The rebound been a difficult one, but conditions should im- in the investment cycle, which is crucial to the prove gradually through 2005 (table 1.3). overall recovery, remains uncertain. Although Although export market growth for devel- economic and financial conditions—low inter- oping countries chalked-up a meager 2.5 per- est rates, rising equity prices, improved corpo- cent advance during 2002, export volumes for rate balance sheets, and increased profits—are the group accelerated from nil in 2001 to 7.7 favorable for a recovery in investment, a lack percent in the year, implying substantial gains of confidence could delay or even abort the re- in market share. As market growth rises toward bound. Other risks are of a financial nature. a 7–8 percent range by 2003–05, emerging- The weakening of the dollar has, until now, market exports should continue to gain share been a movement toward equilibrium. But the and expand at rates around 10 percent. Trade dollar’s fall is likely not enough to sustain prices bode well for improvement in export demand for dollar-denominated assets, as the revenues during 2003, with terms of trade an- burgeoning U.S. fiscal and current account ticipated to rise by 0.4 percent, but modera- deficits are likely to call for an additional $1.5 tion in non-oil commodity prices and in oil trillion in net foreign inflows over the period markets by 2004–05 should serve to dampen through 2005. A further sharp fall in the initial gains. dollar could destabilize the recovery in the Advanced market interest rates are ex- high-income countries. Additional risk may pected to trough during 2003, to rise moder- arise if vulnerabilities in the banking sector ately in 2004, and return to long-term average of the main industrialized economies intensify rates by 2005, a net increase of some 280 basis in response to deflationary pressures. Risks points from current 1 percent levels of US$- in the high-income countries call for decisive LIBOR, and 100 points for EURIBOR, as Eu- policy initiatives to improve the structure of ropean growth fundamentals remain subdued economies. Moreover, uncertainties surround- for a longer period. Meanwhile, there has been ing the recovery in high-income countries sharp compression in average emerging-mar- transmit to the international environment and ket spreads, and a continuation of that trend to the medium-term outlook for developing would help keep the cost of capital for devel- countries, implying that expected improvement oping countries at moderate levels. in developing-country growth should not be The near-term outlook for capital flows to taken for granted. developing countries is mixed, however, as con- tinued restraint on the part of commercial banks in extending new loans is offset by much The external environment for stronger trends in bond issuance. Gross capital developing countries: Gradual market flows are anticipated to rise by some 9 improvement, but a bumpy percent in 2003. But FDI flows in the current road ahead economic and geopolitical environment, poten- T he external environment facing develop- ing countries has been shaped by the set of recent developments in the rich countries as tially exacerbated by the SARS epidemic in China (the major recipient country), are antici- pated to experience no growth during 2003— well as expectations for OECD growth and an assessment that may prove somewhat opti- import demand. The evolution of trade prices, mistic. On balance, the outlook envisions an interest rates, and financial flows will play a environment increasingly conducive for growth, major role in determining external balances, but a potentially bumpy road for policymakers while influencing reserves and liability man- to navigate. 19 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Table 1.3 The difficult environment for developing-country growth should improve External factors affecting developing-country growth, 2001–05 (percent change, except as noted) 2001 2002 2003 2004 2005 OECD import demand –0.7 1.7 4.1 6.3 6.7 Export market growth a –0.2 2.5 6.8 8.0 7.7 Developing export volume 1.1 7.7 10.0 10.9 9.7 Non-oil commodity prices –9.1 5.1 6.9 1.1 1.5 Oil price $/bbl 24.4 24.9 26.5 22.0 20.0 MUV b –4.5 –0.1 4.0 –0.4 1.5 Terms of trade –0.7 –1.1 0.4 –1.5 –2.7 US LIBOR 3.5 1.8 1.0 2.0 3.8 EURIBOR 4.2 3.4 2.1 2.1 3.1 Emerging market spread (bp) 797 728 610 — — Financial flows ($ billion) 326 293 309 — — FDI inflow 171 143 145 — — Gross capital market flows 155 150 164 — — Equity placement 7 11 12 — — Bond financing 62 55 70 — — Bank lending 86 84 82 — — — Not available. a. Import demand in partner markets. b. Manufactures Unit Value index, expressed in U.S. dollars. Source: World Bank data and projections. World trade is growing more slowly than A gradual advance in OECD industrial expected in 2003, but fundamentals production and investment, combined with suggest a buildup of momentum into 2004 somewhat diminished strength in East Asian A decidedly soft patch in trade momentum de- intraregional trade, underpin the anticipated veloped over the second half of 2002 and recovery of trade volumes over the second half through early 2003. Demand was bogged of 2003 and into 2004. And the medium-term down by widespread weakness in the automo- outlook appears generally positive, as domes- tive, air travel, and technology sectors. And tic demand in the rich countries, as well as slackening of domestic demand conditions in low-and-middle income countries, should even- the OECD countries compounded these trends. tually rise in response to substantial policy European import demand went through sharp stimulus. In contrast, the view for a revival of compression: trade among EU member states trade related to the global semiconductor especially affected by recession-like conditions cycle is uncertain at present. Sales may have on the Continent, and demand from outside the reached a local peak, as unit shipments are Union falling rapidly as well. After a pick-up to now close to 2000 levels, while prices are 12 percent rates at year-end, U.S. import vol- likely to remain soft, reflecting the recent shift umes retrenched to a decline of 5 percent by of production toward developing East Asia. April 2003—a pattern echoed by Japanese for- East Asia has been generating strong trade eign demand—before rising again in May (fig- momentum on its own account, led by China. ure 1.15). Against this background, following Chinese nominal imports (excluding oil) have an advance of 3 percent in 2002, estimates for grown at a compound rate of 12.2 percent world trade growth in the current year have since 1995, with much of the demand being been marked down from 6.2 percent antici- met by regional producers. China’s share in the pated in spring 2003 projections to 4.6 percent. merchandise exports of East Asia (Indonesia, 20 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Figure 1.15 OECD-area imports have declined sharply since April 2000 U.S. and EU growth of merchandise import values, 3 month moving average, saar 25 25 Japan 20 EU internal United States 20 ᮡ ᮡ ᮡ 15 15 10 10 5 5 0 0 ᮡ –5 –5 EU external –10 –10 –15 –15 Jan. April July Oct. Jan. April July Oct. Jan. April July Oct. Jan. April 2000 2000 2000 2000 2001 2001 2001 2001 2002 2002 2002 2002 2003 2003 Note: EU weighted intra- and outside-EU imports. Sources: National agencies. Republic of Korea, Taiwan (China), Malaysia, Thailand) has doubled over the last two years Figure 1.16 China’s share of East Asian and quadrupled over the last ten (figure 1.16). exports keeps rising Share in percent The dollar’s weakness may improve 20 merchandise trade balances for developing countries For exporters who sell at a markup to local 15 currency costs, dollar weakness implies higher dollar revenues per unit sold. Over the last two years developing countries’ export prices 10 have tended to grow at higher rates than those of high-income countries, yielding a cumula- tive improvement in the terms of trade. Yet, this development must be interpreted as a 5 double-edged sword, with much of the im- provement explained by higher oil prices. As a group, developing countries are net 0 energy exporters; but within the group there 90 91 92 93 94 95 96 97 98 99 00 01 02 03 are significant net importers, many of which 19 19 19 19 19 19 19 19 19 19 20 20 20 20 have been experiencing deterioration in their Note: Chart shows Mainland China’s 3-month moving average imports from Korea, Taiwan (China), Malaysia, terms of trade. Indeed, two-thirds of develop- Thailand, and Indonesia as a percent of total exports of ing countries, with GDP of $2.9 trillion (56 goods from these countries to all destinations. All data in U.S. dollars for January of the respective year. percent of total developing-country GDP) and Source: Global Trends Team, national sources via population of 3.5 billion (75 percent of the de- Datastream. veloping world) are net energy importers. To a 21 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Table 1.4 Developing countries’ exports will grow faster than those of the high-income countries Merchandise export volumes, 2001–05 (percent change) 2001 2002 2003 2004 2005 World –0.7 3.2 4.6 7.9 7.9 High-income countries –1.2 2.1 3.2 7.1 7.4 OECD –0.5 1.5 2.6 7.0 7.4 United States –5.9 –4.2 0.0 7.6 9.7 Japan –8.3 7.3 6.7 7.0 6.5 Euro Area 3.2 1.1 2.1 5.7 7.1 Other high income –6.8 7.1 7.7 8.1 7.3 Developing countries 1.1 7.7 10.0 10.9 9.7 East Asia and Pacific –0.6 15.5 14.5 13.7 11.4 Latin America and Caribbean –1.2 2.6 7.9 11.4 10.4 Europe and Central Asia 4.9 6.3 8.5 8.7 8.5 South Asia 2.1 –1.8 5.4 7.7 8.2 Middle East and North Africa 5.0 –0.2 4.2 5.5 5.2 Sub-Saharan Africa 2.5 1.2 2.7 5.4 5.7 Source: World Bank data and projections. degree, the decline in terms of trade for net other OPEC countries increased significantly. energy importers may have been moderated Saudi Arabia raised output to more than by price increases for non-oil commodities, 9mb/d to help offset the prevailing and poten- but much of the improvement in price has tial shortfall, and other OPEC members out- been limited to a few products—among them side Venezuela added more than 1mb/d. The cocoa, rubber, and coarse grains. True terms- global oil price began falling shortly before the of-trade gains for this group are likely to wait Iraq war began, as traders did not want to be further softening in oil price and a firming caught short as the conflict commenced, par- in non-oil prices associated with global eco- ticularly if a release of strategic stocks were to nomic recovery. cause prices to crash as they did at the start of The medium-term outlook for developing the Gulf War in 1991. Iraq’s oil infrastructure country exports is encouraging, given expected was only minimally damaged during the war, growth in export markets, combined with and prices declined below $25/bbl during continuing pickup in market shares. East Asia April. But Iraqi exports did not resume shortly is likely to dominate developing country trade after the war’s end, as widespread looting of oil flows over the forecast period, but better field equipment and of service industries that performance is anticipated for Latin America, supply the oil sector, such as electricity and Europe and Central Asia, and South Asia water, significantly delayed restart of produc- (table 1.4). tion. Global crude oil inventories have re- mained persistently low, particularly in the The eventual return of Iraqi oil exports United States, as OPEC has well-managed sales will pressure OPEC and the oil price . . . of its surplus production into the market. Oil prices averaged $31.3/bbl during the first While U.S. imports have risen, continued tight- quarter of 2003. Fears of loss of oil supplies ness in physical markets has caused prices to from war in Iraq, low stocks, strong weather- rebound to around $27/bbl as of mid-June related demand, and sharply reduced exports 2003 (figure 1.17). from Venezuela all contributed to the price Conditions surrounding Iraqi oil exports hike. In response to the rise, supplies from will likely be the most important drivers of oil 22 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S tions to double or triple capacity, but such an Figure 1.17 The price of oil fell sharply endeavor would take several years to achieve before the war in Iraq and require substantial investment. Neverthe- Brent oil price $/bbl less higher capacity in Iraq, along with growing March 20 April 9 capacity within other OPEC members and sig- 35 nificant expansion outside OPEC, is expected yield lower prices over the forecast period (box 33 1.3). Large increases in non-OPEC supplies that will capture much of the expected moderate 31 growth in world oil demand are expected in the next several years. A “high-side” risk to the 29 medium-term outlook for the oil price is that 27 OPEC might exert stronger discipline to sustain higher prices. Such a policy, however, would 25 raise incentives for non-OPEC producers to bring additional supplies to market. 23 But the dollar’s decline and supply-side 02 3 03 03 03 03 /0 1/ 3/ 2/ 2/ 3/ conditions outweigh the effects of the Iraq 30 /3 3/ 4/ 5/ 6/ 1/ 12 Source: DECPG Commodities Group. war on non-oil prices Non-oil commodity prices have patently been affected by the war in Iraq—through increased freight costs, a buildup of inventories prior to prices in the near term. Iraq’s production for the war, and the continuing reduction of stocks domestic needs resumed quickly after the war, in the wake of military engagement. In addi- but export volumes were limited to pre-war tion, the war generally dampened consumer storage that was shipped in late June. Oil ex- and industrial demand. Other factors, how- ports are scheduled to resume in the third ever, have been more important in influencing quarter of the year, but technical difficulties as recent commodity price trends—among them well as recent pipeline explosions will initially the weakening of the dollar, supply conditions constrain shipments to about 1mb/d. However, in individual commodities, and extremely de- the risk of further disruption is high, and it is pressed levels of some commodity prices be- unlikely that Iraq will achieve pre-war produc- fore the recent recovery. The index of non-oil tion levels of some 2.5mb/d during 2003. A commodity prices rose 5.1 percent in 2002 gradual return of exports should enable OPEC after having declined by a cumulative 33 per- to curtail production in a timely manner and cent from 1997 to 2001. Prices are anticipated keep the oil price within its target band of to increase an additional 7 percent during 2003 $22–28/bbl during 2003, particularly given the and advance at a modest 1 percent per year for extremely low level of global inventories. Dif- the following three years. But even after recov- ficulties for OPEC could deepen in 2004 when ery in 2003, nominal prices will remain well Iraq’s exports are anticipated to reach pre-war below their peaks of the mid-1990s. levels. Initial investment in Iraq’s oil sector will fo- The recovery in agriculture prices cus on refurbishing and improving efficiency of is expected to be modest the present infrastructure. Limited growth in The index of nominal agricultural prices rose capacity could occur in 2004, but production is 8.5 percent in 2002. It is projected to increase not expected to rise significantly above pre-war by 7 percent in 2003 and by a more modest 1 levels until 2005 and beyond. There are sugges- percent per year for the next three years. The 23 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Box 1.3 OPEC struggles to achieve higher prices amid growing supply competition competition and below-trend demand growth, and to O il markets suffered minimal disruption as a re- sult of the Iraqi conflict, as other OPEC produc- ers, particularly Saudi Arabia, raised production in fall below $20/bbl by mid-decade. By 2006–07, sig- nificant new supplies from West Africa, the Caspian, anticipation of reduced Iraqi output and to replace Russia, deepwater-areas of the Atlantic basin, and Venezuelan exports. Oil prices peaked at $34.2/bbl elsewhere are expected to come on stream and, cou- just before the conflict commenced but quickly fell pled with rising capacity within OPEC, exert severe below $25/bbl. Oil stocks have stayed persistently downward pressure on prices. A risk to the forecast low and prices have since recovered. is that OPEC could maintain strong production disci- Oil prices have been extremely volatile during pline over the next few years to keep prices at or the last few years, largely because of fluctuations in above $25/bbl. To achieve this goal, however, the OPEC production. Beginning in 2000, OPEC has targeted a price band of $22–$28/bbl for its basket of crude. By Before and during war in Iraq, OPEC raised production and large the organization has been suc- and quotas cessful. With the exception of the post- Millions of barrels per day, January 1996–July 2003 September 11th slump, oil prices have 30 averaged about $27/bbl since late 1999. 29 OPEC-10 OPEC including Iraq This compares with an average price of 28 ᮡ about $17.6/bbl over the 1986–99 period 27 when OPEC was mainly concerned with ᮡ 26 regaining market share. ᮡ 25 OPEC’s new strategy is to keep in- OPEC-10 Quota ventories low and prices high, but doing 24 so requires it to both raise and lower 23 production during the year because of 22 the seasonal pattern of oil consumption. 21 Typically the industry builds stocks in the Jan. 1996 Jan. 1997 Jan. 1998 Jan. 1999 Jan. 2000 Jan. 2001 Jan. 2002 Jan. 2003 summer for use during the peak-demand Sources: IEA and DECPG Commodities Group. winter season. However, it is difficult to precisely anticipate demand and supply three to six months forward, and this can result in large imbalances and volatile prices. cartel would have to yield market share for higher Achieving higher prices has been counter- prices. High prices would add to the growing pres- balanced by loss of OPEC’s market share. In recent sures on world demand and competing supplies. In years virtually all of the modest growth in world oil the end, it is probable that prices would still fall demand (less than 1 percent annually from 1997 to below $20/bbl within a few years. 2002) has been captured by non-OPEC producers— In the longer term, demand growth will be mod- and much by the former Soviet Union. For OPEC 10 erate, as it has been for the past 20 years, but new (excluding Iraq), its crude oil production as a share technologies, environmental pressures, and govern- of total world supply fell from 35 percent in 1996–97 ment policies could further reduce growth. Prices to 30 percent in 2002. Continued high oil prices will below $20/bbl are sufficiently high to generate ade- dampen future demand growth and, more importantly, quate development of conventional and unconven- stimulate development of non-OPEC supplies. Rising tional oil supplies, and there are no apparent re- capacity within OPEC, desires for higher quotas (Alge- source constraints far into the future. In addition, ria and Nigeria), and a recovery of Iraq’s exports could new areas continue to be developed (for example, strain OPEC’s efforts to support higher prices. deep-water offshore sites and the Caspian Sea), and Oil prices are expected to decline from $26.5/bbl development costs continue to fall because of new in 2003 to $22/bbl in 2004 because of rising supply technologies, shifting supply curves outward. 24 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Figure 1.18 Agricultural prices have begun to decline as crop prospects improve Prices in US$, indices May 2001 = 100 130.0 Beverages 122.5 ᮡ 115.0 ᮡ 107.5 Food 100.0 ᮡ 92.5 Raw materials 85.0 May 2001 Aug. 2001 Nov. 2001 Feb. 2002 May 2002 Aug. 2002 Nov. 2002 Feb. 2003 May 2003 Source: DECPG Commodities Group. advance in 2003 is driven primarily by the re- especially in China—and prices have risen covery of coffee, rubber, and vegetable oil quite sharply. Gold also posted higher prices prices after severe declines from 1997 to 2001. tied to reduced producer hedging and concerns Most other agricultural commodity prices are about U.S. equity markets and the dollar. Al- expected to show only modest gains, however, though production cutbacks have been made because of high stock levels, excess production for most metals, idle capacity will haunt the capacity, and slow demand growth associated market even after the anticipated recovery in with the modest recovery of global economic demand is finally underway. Metals prices are activity. Sharp increases in wheat prices at mid- expected to increase by about 6 percent per 2002 were quickly reversed as supply condi- year for the next two years, after decreasing by tions in the United States improved and non- about 3 percent in 2002. Copper, the most bal- traditional exporters—Ukraine and Russia— anced market among base metals, could tilt increased international shipments. Other com- into deficit next year along with nickel. Other modities, such as coffee and sugar, are bur- metals are expected to remain in surplus for a dened by large stock overhang; price increases longer period, unless further production cuts are not anticipated despite historically low lev- occur, and prices are likely to lag behind that of els (figure 1.18). copper during the recovery. Metals prices have undergone rallies over the past years but have faltered—mainly be- Financial markets have shown cause of renewed expectations of weak de- some surprises mand. Recently, worries about the impact of The divergence between the sharp rally in de- the SARS virus on East Asian demand has also veloping countries’ fixed-income asset prices taken the luster out of metals prices. Invento- and the fairly tepid capital flows in response to ries remain high for most metals and most mar- those prices has been the hallmark of market- kets are expected to remain in surplus this year. based financing during early 2003. This devel- Nickel is the exception; stocks are low, there opment was not expected at the start of the has been strong demand for stainless steel— year (see World Bank 2003). While asset prices 25 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 have evolved along expected trends—though the lowest level since 1998 (figure 1.19). The more sharply and swiftly than anticipated— decline in spreads for several countries that had advances in gross capital flows have been attracted investor concern was more spectacu- lower than expected—and lower than war- lar. Benchmark spreads for Brazil, Argentina, ranted by the spectacular pickup in market and Uruguay declined by 1,700, 1,300, and prices. 700 basis points, respectively over the period. The rally in developing-country debt prices Spreads on Turkish debt, which had escalated that commenced in fall 2002, once concerns to 1,100 basis points by end-March 2003, over elections in Brazil and Turkey had faded, dropped close to 700 basis points by mid-May. gained momentum through mid-2003. As in- Gains in fixed-income prices stand out more vestors pursued yields higher than those in in- clearly when benchmarked against the volatil- dustrialized countries, the availability of capi- ity and performance of global equity markets, tal for investment in developing-country debt which returned, on average, barely any gains increased swiftly, outpacing demand for debt between October and mid-May, even when tak- financing from developing-country borrowers. ing into account the fall of the U.S. dollar. The outcome was not new volumes of financ- Despite the narrowing of benchmark ing but a sharp adjustment in benchmark spreads, gross market-based capital flows to prices. These dynamics played out differently developing countries lost some ground during than they have since the late 1990s, when low the early months of 2003 after mustering industrial-country yields failed to promote a strength late in 2002 (figure 1.20). During the surge in availability of capital for developing first quarter, capital flows slumped to $34 bil- countries (because investors were focusing more lion, down about 14 percent over fourth- on credit risk than returns). quarter levels. After a strong start to the year, Benchmark spreads on emerging markets flows eased in February and dropped further in dropped from a peak of 869 basis points in Sep- March. But a revival in both bond issuance tember 2002 to 500 basis points by mid-year, and banking lending during April–May brought Figure 1.19 Emerging-market spreads rallied sharply after late 2002 Basis points 1,400 Brazil [right scale] 2,500 ᮡ 1,200 2,000 1,000 1,500 800 ᮡ All emerging markets [left scale] 1,000 600 400 500 1/ 02 2/ 2 3/ 2 3/ 02 4/ 2 5/ 2 5/ 02 6/ 02 7/ 2 7/ 2 8/ 2 8/ 02 9/ 02 10 02 10 /02 11 /02 12 /02 12 /02 1/ 02 3 2/ 3 3/ 03 4/ 3 4/ 3 3 /0 /0 /0 /0 /0 0 /0 /0 0 /0 0 /0 2/ 3/ 2/ / 1/ / / / 8/ 6/ / 7/ 22 11 23 12 22 11 21 10 30 19 /9 9 8 /8 17 26 18 27 1/ /2 /1 /2 2/ Source: World Bank, DECPG Finance Team. 26 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S creased volatility in financial markets during Figure 1.20 Bond issuance dominates the early months of 2003. Many borrowers capital market flows in 2003 avoided or postponed borrowing. Billions of dollars, monthly averages The demand for external debt financing by 14 Bonds Banking developing countries remained limited, as Equities many (especially in East Asia) have learned to 12 ᮡ ᮡ live with less debt, while others are being kept 10 in check even though financing conditions have eased relative to 2002. Capital flows 8 could strengthen during the remainder of 2003. Borrowers may (a) take advantage of favorable 6 financing conditions and move up their financ- 4 ing targets; (b) establish benchmarks for the ᮡ purpose of setting favorable pricing in the fu- 2 ture, or (c) undertake liability management, given the improvement in terms of borrow- 0 2002 Q1 Q2 Q3 Q4 2003 Q1 April- ing over terms on previously contracted debt. Source: World Bank data. May Novel aspects of financing are already being encouraged. Mexico arranged a $2 billion bank loan to retire some of its Brady debt. And bond issues with collective action clauses flows to a rate of $40 billion, comparable have swiftly become more acceptable with rel- to the patterns established during the second atively little or no premium. Mexico, Brazil, quarter of 2002. The ups and downs in over- and South Africa issued bonds with such all capital flows were led by bond financing, clauses between March and April. which is sensitive to short-run developments. A The stimulus to developing-country capital noteworthy feature in bond financing was the flows from low industrial-country yields could easing of access for Latin American borrowers persist over the short-to-medium terms, as compared with 2002. Brazil was able to come yields in the G-3 countries are unlikely to rise to the market with a $1 billion issue after being sharply and swiftly even if they do turn the absent for a year. However, the major portion corner. Barring major fallout, investors may of bond issuance from Latin America has been continue to grow less risk averse. Most of the accounted for by investment-grade-rated coun- vulnerabilities in emerging-market economies tries. Contrasted with 2002 outturns, bank lend- and in high-yield industrial-country markets ing to developing countries has remained sub- have been exposed and have evolved through dued, with monthly lending standing some 20 changes in market prices. Substantial fallout percent below comparable 2002 levels, as banks from the troubled Latin American borrowers of almost all domiciles have been slow in ex- was largely avoided. Foreign exchange reserve tending credit to developing countries. positions of developing countries in general Market-based capital flows to developing are holding steady, and their dependency on countries have so far performed below the es- external debt is falling. Vulnerability to ad- timates set out for 2003 earlier this year, by verse external shocks appears to have been re- about $2 billion a month on average. The sub- duced from the high levels of the last two dued response of capital flows to the strong years, and flexible exchange rates have helped performance of asset prices can be traced to in this area. And for several countries, the several factors: (a) uncertainty in the runup to macroeconomic implications of accession to the Iraq war, (b) renewed concerns over the the European Union have vastly improved strength of the economic recovery, and (c) in- credit risk perceptions. 27 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 The developing countries: Back on gions, to a large degree, integrated with on- track toward growth? going global developments. For developing countries as a group, output has recovered to 7 D eveloping countries appear poised for growth to resume at more rapid rates. De- spite continuing geopolitical uncertainties and percent year-over-year rates from troughs expe- rienced in early 2002 (figure 1.21a). The subdued conditions in the rich countries at MENA region, which accounts for some 9 per- present, developments in several areas are con- cent of developing-country value added, has en- verging to support the expectation of a step-up joyed a 15 percent surge in crude petroleum in output growth. Should conditions in the and other energy production, as OPEC mem- external environment improve as expected over bers have exceeded quota levels in expectation the next years, countries should have the op- of the loss of Iraqi output. In contrast, some un- portunity to advance export-market shares derlying influences of SARS may be discerned and grow export volumes at more robust rates. in drooping production figures for the East International financial markets have become Asia and Pacific region (EAP), though other more accessible to developing countries, and factors, including the slowdown of world trade compression of spreads has reduced the cost of and poor conditions in high-tech markets, are raising capital. International reserves have likely the more important elements at work. risen to record proportions, especially in devel- Latin America, meanwhile, is building on revi- oping East Asia, providing a degree of cush- talization of output in Argentina, and earlier ioning against potential vagaries in global fi- in Brazil and Mexico, but activity in the latter nancial markets. countries appears to have eased recently. Eu- rope and Central Asia (ECA) has benefited Developing-country fundamentals from Turkey’s dramatic recovery from its fi- are improving nancial crisis of 2001, while hydrocarbon ex- On the domestic front, fiscal positions have porters in the Commonwealth of Independent generally improved across developing regions, States (CIS), including Russia, have ramped up allowing a degree of stimulus to be undertaken oil and gas production to fill part of the gap left to bolster domestic demand, and prospectively, by loss of Iraqi crude oil. As some of the near- more can be achieved here. While industrial term distortions to production dissipate—the production supporting both external and inter- recovery now underway for low- and middle- nal demand growth is currently mixed across income countries should firm and broaden over regions, the seeds of firmer recovery appear to the coming quarters and into 2004. be fairly widespread. Inflation is moderating, Inflation has moderated for developing and this trend should be enhanced as the oil countries as a group, from recent peaks of 5.5 price moves downward over the next years— percent as of mid-2001 toward a 3 percent and for several countries, currency apprecia- year-over-year pace at present (figure 1.21b). tion will abet improved inflation performance. Improvement has been spearheaded by devel- Finally, recent political and financial difficul- opments in the EAP region, where tendencies ties in Latin America appear to have moder- toward deflation in China (largely productiv- ated, though uncertainties persist. In contrast, ity driven), and easing of price pressures in developments in the Middle East and North most ASEAN countries have yielded a reduc- Africa region (MENA) remain unsettled in the tion of 3 percentage points in annual inflation, aftermath of conflict, and heightened uncer- from 5 toward 2 percent rates. And develop- tainty and risk aversion among exporters, in- ments in the ECA region have been quite vestors, and tourists is likely to affect growth promising as well, with a 4 point fall in CPI for the countries of the region for some time. over the like period, from 6 to 2 percent, re- Recent trends in industrial production high- flecting improvement in Turkey, Russia, and light the current mixed picture across key re- many of the countries of Central Europe. In- 28 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Figure 1.21a Regional trends in industrial production are mixed IP, 3-month/3-month, percent change, year/year 16.0 East Asia ᮡ 12.0 Europe and Central Asia 8.0 ᮡ 4.0 0 ᮡ Latin America All developing ᮡ ᮡ –4.0 Middle East and North Africa –8.0 Jan. April July Oct. Jan. April July Oct. Jan. April July Oct. Jan. April 2000 2000 2000 2000 2001 2001 2001 2001 2002 2002 2002 2002 2003 2003 Sources: National agencies. Figure 1.21b Inflation is moderating in the developing world Consumer price index, percent change, year/year 5.5 4.5 3.5 2.5 1.5 Jan. April July Oct. Jan. April July Oct. Jan. April July Oct. Jan. 2000 2000 2000 2000 2001 2001 2001 2001 2002 2002 2002 2002 2003 Note: GDP weighted growth. Source: National Agencies. flation remains at moderate 3 percent rates The fall of the dollar vis-à-vis partner cur- in South Asia (SA), but has re-accelerated in rencies has affected more than the cross-rates Latin America and Sub-Saharan Africa, re- for the euro, yen, or sterling. Since October flecting earlier currency devaluations in the 2002, the brunt of dollar depreciation has former region, and adverse fiscal conditions fallen upon several Latin American currencies, and large-scale currency decline in several importantly the Brazilian real and Argentine large African states. peso (20–25 percent appreciation), reflecting 29 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 and transformations also have induced sharp Figure 1.21c Major currencies in Latin variations in aggregate growth—among them America and East Asia are firming up the Latin debt crisis of the early 1980s, the Dollar depreciation against major currencies since period of recession associated with transition October 2002, percent change in the former COMECON countries during –0 the early 1990s, the East Asian financial cri- sis of 1997–98, and the Russian devaluation –5 of 1999 (figure 1.22). The global downturn of 2001 exacted a toll on developing coun- –10 try growth, but the deceleration of output has been much less dramatic than in earlier episodes. –15 In the base-case projections, developing- country growth is expected to reaccelerate in –20 all regions from the relatively sluggish perfor- mance of 2001–02 to an average GDP ad- –25 vance of 4 percent during 2003, reaching about 5 percent growth by 2004–05, consis- Ar al so ro St $ g n Ro t le S$ on $ h C lin Ye NT Re ub Ba Eu Pe W tent with previous peaks in 2000 and 1996–97 er Br K Source: Datastream. (figure 1.22). Latin America is expected to see the most substantial gain, as the region con- solidates following the Argentine crisis. A re- turn to stronger growth in India powers the in part strong rallies in financial markets in South Asia region, while more moderate gains those countries, along with sharply narrowing are achieved in Europe and Central Asia— spreads on benchmark bonds and fixed- tracking the slower pace of revival in EU ac- income funds (figure 1.21c). Several East tivity. The pickup in growth will be lower Asian currencies have also appreciated more in the Middle East and North Africa, where moderately against the U.S. unit—notably the uncertainty regarding the regional political Thai baht and Korean won, while the Russian and economic situation is likely to persist, and rouble has risen in like proportion. Currency in Sub-Saharan Africa, where only moderate appreciation in Latin America is likely to de- gains in commodity prices and sluggish Euro- liver mixed results: while serving to dampen pean growth play a role (figure 1.23). export prospects—an important vehicle for economic recovery—appreciation will help to SARS carries some adverse economic reduce inflationary tendencies in both Ar- affects for the East Asian economy gentina and Brazil, providing room for needed The East Asia and Pacific region was able to interest rate reductions and a stronger founda- overcome a retrenchment in world trade, a tion for domestic growth. sluggish high-tech sector, geopolitical uncer- tainty, and soaring oil prices in 2002 to register Growth should pick up across regions solid 6.7 percent growth (figure 1.24). China during 2004–05 and Vietnam, at 8 percent and 6 percent re- Over the last years, GDP growth for the de- spectively, were the engines of the regional veloping and transition countries has under- economy, as exports spilled over into domestic gone dramatic shifts and suffered volatility demand and generated strong demands for in- associated with global business-cycle develop- termediate imports from other countries in the ments, as during the recessions of the early region, including Korea, Thailand, Malaysia, 1980s and 1990s. Several region-specific crises and the Philippines. The main exceptions to the 30 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Figure 1.22 Developing countries are on track toward long-term growth GDP growth, percent per annum, 1981–2005 5 4 Early 1980s Trend debt crisis 3 ᮡ 2001 global downturn 2 East Asia financial crisis 1 1990s recession transition countries 0 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 Source: World Bank data and projections. 2003, in part because of the outbreak of SARS Figure 1.23 Growth rates in developing centered in China and Hong Kong. Accord- countries will rise through 2005 ing to favorable World Bank estimates, SARS GDP growth, 2001–02 and 2003–05 (percent) may subtract about 0.5 percent from regional growth during 2003 through its effects on All Developing travel, tourism, and domestic consumption and East Asia investment (box 1.4). Some countries are likely to be more affected than others—because of Latin America and the Caribbean high tourism shares in GDP or their impor- Europe and tance as hubs of international trade. In Hong Central Asia Kong (China) and Singapore, growth may drop South Asia by several percentage points from 2002 results. 2001–02 Meanwhile, growth in the Philippines is ex- ᮡ Middle East and North Africa pected to ease because of consolidation of gov- ᮡ 2003–05 ernment finance, and Indonesia is vulnerable to Africa a fall in confidence, especially in conjunction –1 0 1 2 3 4 5 6 7 with high levels of household debt. Though near-term prospects are muted, Source: World Bank Development Prospects Group. longer-term performance is expected to re- main strong. In 2004, growth is expected to reach 6.7 percent based on a quick contain- buoyant performance were Hong Kong (China) ment of the SARS epidemic, together with re- and Singapore, which failed to capitalize on covery in industrial countries and an upturn in burgeoning intraregional trade, and Indonesia. the high-tech cycle. By 2005, the EAP region is Growth in East Asia is expected to slow by expected to return to a long-term trend rate of more than half a percent to 6.1 percent in output growth of around 6.5 percent. 31 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 parallel with improving investor sentiment re- Figure 1.24 Before the SARS outbreak, garding local equity and fixed-income markets. East Asian* GDP was growing robustly Though still dependent on an accommodating GDP growth, percent change, Q/Q saar external environment, threats of financial 9.0 crises and attendant spillovers appear to have receded. In Brazil, President Lula da Silva’s commit- ment to balanced macroeconomic policies has 6.0 helped to restore confidence, and has already facilitated access to international finance, while additional support from the International Monetary Fund should be forthcoming. Mean- 3.0 while, rising consumer confidence and receding unemployment in Argentina indicate a growing consensus that the worst may be over. Yet in- 0.0 flation risks persist, and the credibility of eco- nomic policy will need to be established by the 1 2 3 4 1 2 3 4 Q Q Q Q Q Q Q Q new president. Both countries continue to face 01 01 01 01 02 02 02 02 20 20 20 20 20 20 20 20 the challenges of supporting growth in the con- *Indonesia, Malaysia, Philippines, and Thailand. Source: Aggregation from national agencies. text of tight constraints to fiscal and monetary policy. In contrast, the Republica Bolivariana de Venezuela restarted its oil exports following Latin America shows signs a strike-induced shut-in lasting some 60 days. of renewed recovery But, there the political crisis appears to be Latin America showed increasing evidence of unresolved and macroeconomic performance a recovery during 2003, after growth fell to continues to suffer, with GDP declining by 9 –0.8 percent in 2002. In fact, the negative percent in 2002 and forecasted to fall 14 per- growth performance for 2002 results exclu- cent for 2003. Peru’s performance has been re- sively from severe contractions in Argentina, markable for 2002 and prospects for 2003 are Uruguay, and Venezuela; and a regional aver- still favorable, but recent political turmoil will age excluding these countries registers positive potentially be a drag on growth and raise pub- growth of 1.6 percent, underscoring that con- lic deficits. Recent political unrest in Bolivia tagion had been limited. Industrial production also had damaging repercussions, and in the returned to positive growth during the fourth Dominican Republic a massive bank fraud trig- quarter of 2002, up 5.6 percent, underpinned gered the collapse of the second-largest finan- by gains in Argentina, Chile, and Brazil (figure cial institution, with potentially heavy financial 1.25). Falling yield spreads confirmed the im- and economic costs. pression of an improvement in financial health. Given the extent of recent difficulties and In Brazil, for example, spreads declined from a remaining uncertainties, near-term growth peak of 2,067 basis points in October 2002 prospects remain subdued. However, the to 754 basis points by April 2003, as risk aver- longer-term outlook is more optimistic. Multi- sion eased among global investors. Though lateral trade agreements and regional market- capital flows, especially bank lending, to Latin opening initiatives in the European Union (EU) America and the Caribbean remain weak, the and U.S. markets bode well for external per- strengthening of several local currencies against formance. Exports are diversifying into areas the dollar (Argentine peso, Brazilian real, such as tourism and other services. Macro- Chilean and Mexican pesos) has occurred in economic stability and commitment to sound 32 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Box 1.4 Economic effects of Severe Acute Respiratory Syndrome (SARS) to areas where there were SARS outbreaks. The air- W hile the threat of Severe Acute Respiratory Syndrome (SARS)—a virus that spreads easily and has no known antidote—has shaken China, line industry was hit especially hard, as the SARS virus came at an already difficult time, following the Canada, and the world at large, the spread of the September 11 attack and the war in Iraq, which al- outbreak has stopped and public panic has largely ready diminished vacation travel, in particular, but subsided. The outbreak seems to have been brought also corporate travel, given sluggish world growth. under control as a result of stringent public health Manufacturing trade was also affected. International measures in the affected countries, though there may firms with production lines in affected countries suf- also be a seasonal component to the disease. The eco- fered some interruptions to production because of nomic impact should thus be concentrated in the sec- restrictions on travel. Some (temporary) substitutions ond quarter of 2003, and is expected to be largely may have been made; however, sustained losses in limited to the East Asia region, with a relatively small market share are unlikely. Finally, the SARS outbreak impact on global growth. As of end-June 2003, it is temporarily depressed consumer and business confi- estimated that SARS had infected 8,450 people in 32 dence in the countries most affected, but contain- countries and to have killed over 800 people. ment of the outbreak should result in a quick return Short-term economic activity—particularly in to pre-SARS confidence levels. The macroeconomic ‘face-to-face’ exchanges such as in the retail sector— policy response to the outbreak is expected to be slowed in the most affected countries. Overall, the most evident in increased fiscal outlays through economic impact was largely transmitted through ag- transfers and tax breaks, in addition to increased gregate demand, but also to a lesser extent through health care expenditures to respond to the outbreak. supply mechanisms. Tourism and trade-related ser- vices declined as people became less willing to travel Source: World Bank staff. Figure 1.25 Argentina, Brazil, and Chile see strong upturn in production IP, 3-month/3-month, percentage change, year/year 15 15 10 Argentina [right scale] ᮡ 10 Brazil 5 ᮡ Chile 0 ᮡ 5 –5 –10 0 –15 ᮡ Latin America ᮡ –20 Venezuela [right scale] –5 –25 2000 Q2 2000 Q4 2001 Q2 2001 Q4 2002 Q2 2002 Q4 2003 Q2 Sources: National agencies. 33 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 fiscal and monetary policies are becoming the norm and already yielding benefits in the form Figure 1.26 Growth will cool in CIS while of less costly international borrowing and more picking up in Central and Eastern Europe robust domestic financial markets. GDP growth, percent per annum 10.0 Central Asia The outlook for Europe and Central Asia ᮡ 7.5 Central/Eastern is mixed: greater EU demand, but flagging Europe oil prices ᮡ 5.0 Output expanded by 4.6 percent in the ECA re- gion during 2002, primarily resulting from the 2.5 strength of domestic demand, which more than offset lackluster growth in the region’s main 0.0 ᮡ export markets. A number of economies en- Europe and Central Asia joyed a pickup in growth during the year –2.5 (Croatia, Estonia, Lithuania, Poland, Slovak ᮡ Commonwealth of Republic, Turkey, Armenia, Azerbaijan, Be- –5.0 Independent States larus, Georgia), though excluding Turkey ac- tivity was marked down about half a point to –7.5 3.9 percent. Output in the Commonwealth of 95 96 97 98 99 00 01 02 03 04 05 19 19 19 19 19 20 20 20 20 20 20 Independent States (CIS) eased to 4.7 percent Source: World Bank data and projections. growth in 2002, from robust 5.8 percent out- turns of 2001, and from the spike in growth of 8.4 percent posted in 2000. The critical factor in this development was erosion of stimulus to excluding Turkey, growth in 2003 is projected the Russian economy stemming from the rou- to ramp-up moderately (by 0.5 points) as a re- ble devaluation of 1998 and the rents from sult of three factors: a gradual recovery to- strong energy prices. In turn, diminished im- ward year-end in the EU, the group’s main ex- port demand from Russia—representing an im- port market; notable acceleration of growth in portant export market for the remaining CIS Poland (representing about 13 percent of the countries—contributed to the slowdown for region’s GDP); and expected improvement in the rest of the group. Activity in the Central growth performance in the Czech Republic, and Eastern European Countries (CEECs), ex- Slovenia, and Albania. An expected boost cluding Turkey, was unchanged in 2002 rela- to consumer confidence is likely because of tive to 2001, at 2.9 percent. Including Turkey, progress in the EU accession process.2 Growth growth averaged 4.5 percent for the group, a is projected to strengthen slightly among the sharp upswing from contraction of 0.8 percent CIS countries in 2003, as domestic demand in 2001, reflecting a 7.8 percent recovery en- has begun to firm in Russia, which in turn joyed by Turkey in 2002. For the CEECs, do- should support growth in other CIS countries mestic demand was spurred by fiscal policy dependent on Russia’s import demand. ECA (Hungary, Czech Republic, Poland, Slovenia, regional growth is expected to accelerate to Slovakia) and/or easing of monetary policy 4.5 percent in 2004 and then to slow to 4.1 (Czech Republic, Latvia, Lithuania, Romania). percent in 2005, reflecting divergent trends at Aggregate growth for the region is antici- the sub-regional level. Growth in the CEEC pated to slow moderately to 4.3 percent in sub-region (including Turkey) is likely to ac- 2003, as a return to more modest advances in celerate from 3.5 in 2003 to 4.3 and 4.7 per- Turkey, under the burden of required fiscal cent in 2004 and 2005, respectively, in part consolidation and related issues, will carry because of firming of external demand and some weight (figure 1.26). Among the CEECs, significant inflows of FDI to new EU mem- 34 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S bers, in addition to EU transfers. Growth is sequences were particularly adverse. For Egypt, likely to slow in the CIS from 5.3 percent lower non-oil trade also affected revenues from in 2003 to 4.6 and 3.4 percent in 2004 and the Suez Canal. 2005, respectively, assuming a substantial fall Short-term prospects for the region will be in the oil price in both 2004 and 2005 and a conditioned by political resolution in post- decline in growth impetus through fiscal link- conflict Iraq. Uncertainties over the future of ages, especially in Russia. the country, with respect to governance, aid flows, and reconstruction, will continue to af- The war in Iraq and its aftermath fect the region for some time. Nevertheless, are the key factors for the Middle East growth should accelerate somewhat during and North Africa region 2003. The oil-exporting countries are expected The buildup to the war in Iraq and its after- to grow more quickly in 2003 as a result of math have dominated events in the Middle fiscal pump priming and increased oil pro- East and North Africa region over the past duction quotas. For the diversified exporters, year. The developments were, on balance, pos- particularly Jordan and Egypt, a gradual re- itive for oil exporters. The oil price surged, covery in tourism and other sectors affected by peaking at $38/bbl, and production quotas the conflict could unfold in the second half were raised in early 2003 (figure 1.27). Buoy- of 2003, but such recovery would be fragile. ant oil revenues boosted economic perfor- Other factors, not directly associated with re- mance by supporting higher fiscal expendi- cent developments in Iraq, will shape the near- tures. Elsewhere in the region, however, term outlook. Egypt is suffering from a period tourism and trade, not fully recovered from the of extended weakness in the domestic sector, effects of September 11, 2001, were further and despite reforms to the exchange rate re- battered by the prospect of conflict in Iraq. The gime earlier in 2003, growth expectations for most affected countries were those closest to the year have dimmed as private investment the conflict. Tourism in Jordan and Egypt was remains subdued. Moroccan agriculture will seriously affected; for Jordan, where tourism provide a substantial boost to output in 2003 accounts for some 10 percent of GDP, the con- following the severe drought in that country. A Figure 1.27 Middle East oil production has increased to prevent shortages Oil prices, and production three-month moving average, percent change, year/year 30 Oil price, 3mma [right scale] 32 ᮡ 25 30 20 28 15 10 26 ᮡ 5 24 0 Algeria production 22 –5 ᮡ [left scale] Saudi Arabia production 20 –10 [left scale] –15 18 Jan. 2000 June 2000 Nov. 2000 April 2001 Sept. 2001 Feb. 2002 July 2002 Dec. 2002 May 2003 Source: World Bank data. 35 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 similar situation exists in Tunisia, where agri- situation and progress in macroeconomic sta- cultural output fell by an estimated 11 percent bilization in both Pakistan and Sri Lanka. during 2002. Growth is anticipated to accelerate through- The economic consequences of the conflict out the region in 2003, to an average of 5.4 in Iraq will play out through its impacts on percent, assuming a return to trend agricultural confidence and investment spending. A pro- production, a recovery in external demand, tracted process of reconstruction could exac- continued improvement in political stability erbate these problems. A downward trend in and regional security, and a firming of domes- the oil price presents a further risk. With slug- tic demand, especially in India (figure 1.28). In gish growth in world demand, oil prices could the medium term South Asian growth is likely trend lower than anticipated, cutting ex- to be sustained near 5.4 percent, assuming a porters’ incomes and putting fiscal expendi- continued recovery in external demand and es- ture programs at risk. Moreover, further polit- tablishment of normal trends in agricultural ical instability in the tense environs of the output. Bangladesh and India should benefit region cannot be ruled out, a development from an ongoing recovery in domestic demand. that would hamper investment and growth for Both Pakistan and Sri Lanka are projected to an extended period of time. enjoy continued macroeconomic stability and an associated acceleration of growth. Similarly, A stronger external environment, upswing Nepal is anticipated to experience a pick-up in agricultural cycle, should boost South in growth, assuming continued improvement Asian growth in the security situation there, with a recovery in Growth in the South Asia region slowed to 4.2 domestic demand and in tourism receipts. percent during 2002 from 4.9 percent in 2001, Furthermore, recently improved relations marking a downward revision from previous between India and Pakistan are hoped to lead estimates, largely because of adverse weather to greater stability in the region, paving the conditions and declines in agricultural output way for increased business confidence and sta- in India, Bangladesh, and Nepal. Nepal experi- bility. The fiscal positions of the South Asian enced a plunge in tourism receipts and a sharp economies are forecast to improve moderately, fall in manufacturing output, as domestic in- assuming some progress in raising budget rev- surgency intensified. Pakistan and Sri Lanka enues (as a share of GDP) and improvement in both enjoyed a pickup in growth during the the management of government expenditure. year linked to strong government spending in Inflation is projected to increase somewhat, al- Pakistan; and for Sri Lanka, a recovery in the beit still at moderate levels, because of stronger services sector and improved political stability growth and assumptions of a more accom- tied to progress in peace talks and implemen- modative monetary stance in many countries. tation of a year-long cease-fire. Current ac- Falling oil prices are expected to provide some count balances for the two largest economies, offset to these domestic factors. India and Pakistan, posted surpluses and the re- gion’s aggregate external balance strengthened. Sub-Saharan Africa maintains positive A number of economies experienced a signifi- per capita growth in spite of a difficult cant increase in remittances during 2002. external environment These were driven largely by: incentives intro- A subdued world economy together with famil- duced by the Bangladeshi government to chan- iar problems of drought and civil strife held nel remittances through official sources; high growth in Sub-Saharan Africa (SSA) to 2.8 per- interest rate differentials in India—reflecting cent in 2002. Import demand from Europe, the significant government borrowing require- region’s main trading partner, was particularly ments there; and improvements in the security weak. Though most (dollar denominated) com- 36 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Figure 1.28 Indian production of food and automobiles recovered sharply in early 2003 Industrial production, 3-month/3-month, percent change, year/year 24 20 16 Food ᮡ Transport equip 12 ᮡ 8 4 Total production 0 –4 –8 Jan. April July Oct. Jan. April July Oct. Jan. April July Oct. Jan. April 2000 2000 2000 2000 2001 2001 2001 2001 2002 2002 2002 2002 2003 2003 Sources: Feri and national sources. modity prices have rebounded from recent to preliminary estimates 12 countries in the re- lows, terms of trade for non-oil exporters have gion achieved growth of 4 percent or better and recovered little of their losses of the past few average per capita income rose for a fourth suc- years. At the same time, travel and tourism suf- cessive year—the longest sustained rise in over fered not only from slower world income two decades. growth, but also from terrorist fears and the The region’s largest economy, South Africa, buildup to war in Iraq. As a result, foreign trade registered relatively sound performances dur- made a negative contribution to the region’s ing 2002. Growth slowed toward the end of growth. Domestic economies also slowed as the year, but remained in positive territory as poor weather or civil disorder disrupted agri- it has since 1999. Because of the increasing cultural production in countries containing strength of the rand, foreign trade contributed over half the region’s population, depressing in- negatively to growth, but domestic absorption comes and demand. Notably, though, invest- was strong enough to offset that impact and ment spending was relatively resilient. There growth overall reached 3.0 percent. Invest- were pluses to note as well. South Africa over- ment was particularly strong, up 6.5 percent in came the depressed tourism market to become spite of high real interest rates. In Nigeria, the the world’s fastest growing tourist destination picture was mixed. The successful presidential in 2002, with arrivals up over 20 percent. Non- election helped cement the fledgling demo- traditional exports covered by AGOA prefer- cratic process; however, progress on fiscal and ences—transportation equipment, textiles and economic reforms continues to be frustratingly apparel, and agricultural products—registered slow. Despite high oil prices, budget grid- strong growth despite the slowdown in the U.S. lock and a reduced OPEC quota held growth economy, indicating that with the right incen- to only 1.9 percent. Progressively weaker oil tives and opportunities, SSA countries can be prices over the next few years will put pressure competitive. Most encouraging of all, according on fiscal and external accounts, though expan- 37 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 sion of the energy sector, especially of liquid natural gas (LNG), should sustain moderate Figure 1.29 Growth in Africa is expected real growth in the medium term. to improve modestly In the medium term, the economic perfor- GDP growth, 2001–02 and 2003–05 mance of Sub-Saharan Africa should benefit as Sub-Saharan the global recovery consolidates. Yet, with ex- Africa pectations for Europe at best moderate, the ex- East/South RSA ternal impetus to growth will remain weak. For the region as a whole, growth is expected to re- West Nigeria main unchanged at 2.8 percent in 2003, then CFA 2001–02 rise to 3.5 percent in 2004. Both oil and non-oil countries 2003–05 producers will share in the acceleration. For oil Oil producers—including additions to the list such exporters as Côte d’Ivoire—rising capacity presages sub- RSA and stantial growth in medium-term production Nigeria and exports, even though prices and terms of 0 1 2 3 4 trade are expected to fall sharply. Major energy- Note: RSA is Republic of South Africa. related infrastructure projects will further sup- Source: DECPG. port demand. For the rest of the region, the re- cent rebound in commodity prices has largely run its course, but at least the expectation is for Trade, growth, and poverty a measure of stability in key export markets at in developing countries levels where exporters in Sub-Saharan Africa can continue to compete. With luck, better weather conditions will stimulate domestic in- W orld trade growth reached unprece- dented levels in the 1990s, accompanied by accelerating flows of foreign direct invest- comes and expenditure as well (figure 1.29). ment (FDI). The sectoral and regional compo- In the longer term, per capita growth is ex- sition of trade changed dramatically with the pected to average 1.6 percent—a substantial spectacular growth of export volumes. While improvement on long-run historical trends, natural resource–based commodities—agricul- though barely half what would be needed to ture, oil and gas, minerals—were driving fac- achieve the MDGs. The region continues to tors in developing countries’ export growth in face immense development challenges from the past, more recent trade growth has been HIV/AIDS, to low savings and investment, poor driven largely by manufactured goods—high- infrastructure, shortages of human capital, and tech products as well as low-skill-intensive negative perceptions of international investors goods. These trends have led to stronger eco- (box 1.5). Nor does the forecast anticipate nomic growth in many countries and signifi- any significant help from a reversal of recent cant reductions in global poverty. terms of trade losses. But the region’s most crit- Can these trends be sustained over the next ical need is to re-establish civil order, where 10–15 years? And can they be broadened to in- lacking, and to raise standards of governance clude countries that have not benefited from and policymaking, for these are the most pow- trade growth, but have very large proportions erful predictors of economic performance. of poor people? The answer to both questions is Here there is encouraging progress to report, “probably.” The long-term forecast anticipates with signs of institutional strengthening at that the MDG of halving the number of the both the country and regional levels. On bal- world’s people living in extreme poverty will be ance, assuming a continuation of this trend, reached by 2015. Nonetheless, significant pock- the forecast of positive, albeit moderate ets of poverty will persist, and the goal will not growth for the region should be achievable. be achieved in all developing regions. 38 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Box 1.5 AIDS is taking a rising toll in Sub-Saharan Africa • Slower labor-force growth and a higher proportion S ub-Saharan Africa continues to be the epicenter of the AIDS epidemic. According to UNAIDS, 28.5 million Africans were infected in 2001 and 2.2 mil- of younger, less-skilled, and less productive workers lion died, lowering population growth by one-third • Lower productivity because of illness or worry on of a percentage point. Given social and financial hur- the job, or more time off work dles, treatment and care programs are likely to have at best a modest impact on the course of the epi- • Higher costs to governments and employers of demic. UNAIDS predicts 55 million deaths attribut- health care, training, and sick pay able to AIDS in Africa between 2000 and 2015 (UNAIDS 2002). • Reduced household savings after payments for Although effective antiretroviral therapies have treatment or funerals, and, simultaneously, less been developed, they are not in widespread use. Part public and private investment because of financing of the explanation is cost. Providing these drugs to constraints, uncertainty, and lower expected the entire infected population of Sub-Saharan Africa profits. would cost nearly $9 billion, about 70 percent of current official development assistance to the region. Quantifying these channels is not straight- Nevertheless, it can be argued that such an expendi- forward, but the preponderance of results suggests ture, equivalent to just 0.04 percent of OECD GDP, an overall reduction of per capita growth somewhere would be cost effective from a development stand- between 0.5 and 1 percentage point. This reduction point, by alleviating the burden on health-care sys- is a significant cost to a region where long-term tems and raising productivity (Moatti and others growth lies in the 0.8 to 1.6 percent per capita 2002). While more money is becoming available, it range, and it underscores Greener’s point (2002) remains only a fraction of what is needed. that economic policy is important so that better In addition to the financial obstacles to treat- economic performance can offset the enormous ment, there is a deadly culture of denial to be over- devastation. come. In Botswana, the most afflicted country in the HIV/AIDS has major implications for public region, the incidence rate in the adult population is finance and the provision of health services. Even near 40 percent and life expectancy has dropped without the epidemic, African public health bud- from more than 65 to less than 40. Yet a $100 mil- gets, which average $50 per capita, would be lion partnership between the Bill and Melinda Gates woefully inadequate. In addition, the disease may Foundation and Merck to provide free antiretrovirals have major, though not well studied, implications to all who need them has, so far, achieved only lim- for income distribution. An individual household is ited success. Less than 5 percent of Botswanans have either affected or not; and for those vulnerable to been tested for HIV, and fewer than 0.1 percent of being tipped into poverty by the loss of one or more those thought to be infected are enrolled for free breadwinners, the effect can be tragic. The threat treatment. posed by large numbers of homeless, uneducated, Many researchers have explored the economics angry youths with no parents and no prospects may, of HIV/AIDS using macroeconometric and CGE in the long run, turn out to be the greatest cost of models. The magnitudes of impact appear surpris- the epidemic. ingly small, seemingly out of proportion to the human tragedy. From a macroeconomic standpoint, impacts of HIV/AIDS arise from: Sources: Greener (2002), Moatti et al. (2002). 39 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Trade performance over the 1990s Within a decade export revenues in develop- was unprecedented ing countries rose from less than 15 percent of From several points of view, trade perfor- GDP to almost 25 percent (figure 1.31). mance during the 1990s was unprecedented. The overall volume of trade accelerated rela- The change in the composition of exports tive to output, growing nearly 2.5 times faster was a major factor underpinning growth than GDP, compared to an historical average More remarkable than the overall growth of of 1.5. Such increase in income elasticity was trade was the transformation in the product a global phenomenon, although it was clearly mix of exports. Developing countries now rely more pronounced in developing countries, less on shipments of primary commodities than which had experienced a sharp fall in trade on manufactured goods. Whereas two decades during the debt crises of the 1980s, and a ago developing countries derived 70 percent of sharp boom just before the financial crises of merchandise export revenue from sales of pri- 1997 and 1998 (figure 1.30). mary commodities—agriculture and energy— The robustness of the recent trade expan- the situation is now completely reversed, with sion was highlighted following the East Asian 80 percent of revenue coming from exports of financial collapse. Trade flows recovered from manufactures. Even exports from Sub-Saharan that crisis much more quickly than they did Africa are no longer primarily resource-based, after the Latin American debt crises of the as the share of manufactures in African exports early 1980s. During the 1990s, developing has risen from 25 percent during the late 1970s countries’ merchandise exports increased at to 56 percent today. Almost all of the increase an annual rate of 8.5 percent, up from growth was realized during the last decade. trends of less than 2 percent during the 1980s. With the rising share of manufactures in Despite the financial crisis of 1997, exports total exports, underlying high growth rates in from East Asia increased on average by 13.4 most manufacturing sectors have an increas- percent per year during the decade, almost ingly large impact on overall export growth. doubling the strong performance of the 1980s. The shift toward manufacturing clarifies some Figure 1.30 Income elasticity has risen globally, but particularly in the developing world Trade to GDP elasticity by region, 1967–2001 4 3 Industrial countries 2 ᮡ 1 ᮡ World total 0 ᮡ –1 Developing countries –2 67 70 73 76 79 82 85 88 91 94 97 00 19 19 19 19 19 19 19 19 19 19 19 20 Source: World Bank data. 40 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S regions was due to differences in sectoral Figure 1.31 Export-to-GDP ratios have growth rates, the other half to the larger share risen sharply in developing countries of manufactures in East Asia. Merchandise exports, percent of GDP, 1980–2002 Such compositional effect will continue to 25.0 influence overall growth numbers during the coming decade as the share of manufactured 22.5 Developing countries products rises further. Should all developing ᮡ countries achieve the same growth of export 20.0 revenue at a sector level as during the 1990s, overall revenue growth would rise to 20 per- 17.5 cent per year, instead of the 11 percent real- ᮡ ized annually during the 1990s. The composi- 15.0 tion effect of 9 percent a year applies to most High-income countries regions, with regional effects ranging from 5.1 12.5 percent in Latin America to 10.6 percent in Europe and Central Asia. 10.0 80 82 84 86 88 90 92 94 96 98 00 02 Changes in the composition of trade can 19 19 19 19 19 19 19 19 19 19 20 20 Source: World Bank data. be traced to several factors What accounts for the growth in manufac- tured exports? Will trade continue to grow at of the regional differences in overall trade per- such robust rates? The driving forces are a formance—growth has been fastest where the combination of policy reforms, structural share of manufactured products in total ex- change in global production processes, and ports was already large—and suggests that the general economic trends related to continuous acceleration of overall trade growth was not a increases in real per capita incomes. The con- temporary phenomenon, as the share of man- tribution of these various forces cannot be de- ufactured products is likely to increase further. composed in a linear fashion, because some are To illustrate the importance of the export tightly linked with others. However, because mix, annual growth of export revenue during the strong links work in a virtuous circle, it is the 1990s for East Asia and Sub-Saharan likely that the combined effects evident in cur- Africa are calculated with different sectoral rent trends will continue into the coming years. weights, while using actual growth rates at a Policy reforms began in the 1970s in East sectoral level. Merchandise export revenue in Asia. They were later initiated in other regions, Sub-Saharan Africa advanced at an annual culminating in a rapid acceleration of reforms rate of 5.4 percent during the 1990s. How- during the 1990s. A key element of the policy ever, if the region had already reached the ma- change was lowering of trade barriers in man- turity of East Asia, with a significantly larger ufacturing—unilaterally, regionally, and multi- share of manufactures in total exports, the laterally. But in all successful cases, change was same sectoral growth rates—double digit for embedded in broader domestic institutional manufacturing and small for natural re- reforms. Technological progress lowered trans- sources—would have led to 10.6 percent over- portation costs, improved communications and all revenue growth. Alternatively, applying business practices, and made it possible to East Asian growth rates to Sub-Saharan shares build global production networks. The last would have led to 10 percent growth, instead change radically altered geographic specializa- of the 17 percent annual growth actually real- tion patterns and intensified trade in interme- ized. This suggests that roughly half of the diate products. Income growth triggered con- difference in overall growth between the two sumers’ desire for more and newer varieties of 41 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 goods, creating markets for foreign products. economies of poor countries, it is a necessary These factors reinforced one another: Lower condition to be able to justify indispensable in- trade barriers triggered a new, global organi- vestments—in public and private infrastructure zation of production to take advantage of di- and education—to enable these economies to versity in comparative advantage across the take off. The Doha Round will be a key com- world. Desire for new products and a search plement to other more limited efforts to reduce for new markets provided a strong incentive trade barriers, for example, regional free trade for lower trade barriers. And importantly, tech- agreements and unilateral reform efforts. nological progress and income growth were spurred by increased global competition and Greater trade will build on ongoing efficiency gains through global networks. reforms to spur per capita growth The decline in manufactures trade prices in all regions relative to domestic price deflators is a clear Intensified trade relations have laid the foun- indicator of strong productivity growth in dations for continuation of a virtuous circle in sectors operating on global markets. Prices of which access to new markets, increased com- merchandise exports from high-income coun- petition, and productivity growth reinforce tries fell by almost 2 percent per year relative each other. Such an upward spiral would ac- to domestic prices. This marked a significant celerate per capita income growth in many de- acceleration of the price differential compared veloping countries, especially those in which to the 1980s, when relative export prices fell the effect of reforms is visible in certain sec- 1 percent per year. A similar indication of ac- tors—even if not yet on an aggregate level. As celerating productivity growth in sectors pro- competitive manufacturing sectors grow and ducing for global markets was observable in reforms spread further, the results will become East Asia, where exports were already heavily evident in economic figures over the next 10 concentrated in manufactures. In that region years (table 1.5). The growing reliance on the price differential changed almost 2 per- manufacturing will also reduce vulnerability centage points during the 1990s. Though for to sharp and disruptive commodity cycles. other developing regions price trends are The impact of industrialization on produc- mixed—partly because of different specializa- tivity is reflected, for example, in the sharp tion patterns, partly because of imperfections acceleration of per capita income growth in in domestic-factor markets—an acceleration Sub-Saharan Africa, from annual decline of 0.2 of productivity growth in manufacturing sec- percent during the 1990s to an increase of 1.6 tors that compete in global markets appears to percent per year during the coming years. On be a worldwide phenomenon. the other hand, countries that, over the last While there are several key factors driving decades, have experienced a rapid catching-up the changes in trade, there is no doubt that the in productivity as a result of integration in sustained dismantling of trade barriers has been global markets are bound to experience some a primary driver. For example, the growth of slowdown, as the gap with the technological production networks and their association with frontier narrows in some sectors. However, trade growth would not have been possible if East Asia is still anticipated to outperform trade barriers had remained high. Expanding other regions, with average per capita income the benefits of trade to a broader range of coun- growth of 5.4 percent—lower than the 6.4 per- tries will require significant further decline in cent growth achieved during the 1990s. trade barriers, particularly for those commodi- As these fundamental structural shifts pro- ties in which poor countries have a comparative mote development, other conditions across advantage—agriculture and low-skill-intensive countries are improving. Reduced imbalances manufacturing. While expanding market access in the external and internal accounts of devel- is not a sufficient condition to catalyze the oping countries have lowered their vulnerabil- 42 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Table 1.5 GDP per capita will grow faster in the developing world than in the OECD area Real GDP per capita, annual average percentage change, 1980s, 1990s, and forecasts Forecast Medium term Long term 1980s 1990s 2001–05 2006–15 World total 1.3 1.2 1.0 2.2 High-income countries 2.5 1.8 1.4 2.5 OECD 2.5 1.8 1.4 2.4 United States 2.2 2.2 1.5 2.5 Japan 3.5 1.2 0.7 1.9 European Union 2.1 1.7 1.5 2.3 Non-OECD countries 3.1 3.8 1.1 4.2 Developing countries 0.7 1.7 2.7 3.4 East Asia and the Pacific 5.6 6.4 5.4 5.4 Europe and Central Asia 0.6 –1.8 3.8 3.3 Latin America and the Caribbean –0.9 1.7 0.3 2.5 Middle East and North Africa –0.6 1.2 1.4 2.5 South Asia 3.6 3.3 3.4 4.1 Sub-Saharan Africa –1.1 –0.2 1.0 1.6 Note: Aggregations are moving averages, reweighted annually after calculations of growth in constant prices. Source: World Bank. ity to swings in international interest rates of the 1990s. Acceleration in the developing and exchange rates and provided governments countries will be more dramatic, with a projec- with some room to manage economic down- tion of per capita growth of 3.4 percent for turns. And macroeconomic conditions have 2006–15, driven partly by a turnaround in Eu- improved—for example, lower inflation and rope and Central Asia—an improvement al- interest rates are providing an improved envi- ready under way in the late 1990s. With adhe- ronment for long-term investment by both do- sion to the European Union only months away, mestic and foreign entities. the accession countries can anticipate the type Apart from acceleration in growth, the rel- of growth experienced by Portugal and Spain ative importance of growth-supporting factors upon their accession—built on solid investment is likely to change. Productivity will likely in- flows, improved market access, and financial crease in importance relative to population assistance from Brussels. The other countries in growth and capital accumulation (figure the region will benefit through trade linkages; 1.32). This is especially true for countries in they, too, will continue to consolidate the ben- Latin America, where population growth dur- efits from reforms initiated during their transi- ing the coming 10 years is expected to be 0.5 tion from planned economies. percentage points lower than during the Somewhat more tentatively, the scenario 1990s, capital accumulation will slow, and also presumes improved economic perfor- countries will be less able to rely on continued mance in Sub-Saharan Africa, which has wit- large current-account deficits. Reforms will nessed two decades of negative per capita have created the right environment to absorb income growth. Despite the nearly 2 percent- technological innovations. age point turnaround in per capita growth, a For the high-income countries, the scenario rate of 1.6 percent per capita if achieved, suggests per capita growth of around 2.5 per- would still leave Sub-Saharan Africa at the cent over 2006–15, an acceleration of 0.7 per- low end of the developing-country growth centage points from the average growth rate spectrum, inadequate to make much of a dent 43 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Figure 1.32 Productivity will contribute more to GDP growth through 2015 than will capital or labor a. Decomposition of GDP growth, 1990–2000 Average percent per annum 8 6 Productivity 4 ᮡ Capital 2 Labor ᮡ ᮡ 0 –2 –4 High-income East Asia Europe and Latin America Middle East South Sub-Saharan countries and Pacific Central Asia and the and Asia Africa Caribbean North Africa b. Decomposition of GDP growth, 2005–2015 Average percent per annum 8 6 Productivity 4 ᮡ Capital 2 Labor ᮡ ᮡ 0 –2 –4 High-income East Asia Europe and Latin America Middle East South Sub-Saharan countries and Pacific Central Asia and the and Asia Africa Caribbean North Africa Source: DECPG staff estimates. in poverty and other MDGs. Nonetheless North Africa. Recent and prospective policy there are positive signs emerging from parts changes suggest that the region could attain a of Sub-Saharan Africa, with per capita income per capita growth rate of 2.5 percent over the growth positive for the fourth consecutive next 10 years. Such an achievement would year, and a dozen countries achieving growth facilitate effort—particularly in private-sector rates in excess of 4 percent per annum. development—to absorb the rapid increase in The main difference with the long-term the labor force. scenarios contained in Global Economic Growth rates in East Asia and Pacific are Prospects 2003 resides in the Middle East and expected to remain strong through the long 44 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S term. At official exchange rates, incomes in And growth will greatly reduce poverty rich countries remain more than 20 times Strong economic growth—particularly in higher than the average in East Asia and Pa- China and India, with yet-high concentrations cific, and even at purchasing-power exchange of poor—will lead to substantial reductions in rates, incomes are six times higher. the incidence of poverty through 2015, with After the Middle East and North Africa, the MDG of halving extreme poverty being South Asia is the region least integrated into achieved on a global level—if not in each the global economy. Growth over the last country or region. Sub-Saharan Africa—un- decade can be attributed partly to trade policy less dramatic changes occur—almost surely reforms. Those reforms are expected to be will not reach the goal. Even under a some- pursued by local governments over the next what optimistic economic forecast, the per- decade, giving rise to sustained growth at rates centage of the African population living on approaching those of East Asia. $1/day or less will remain above 42 percent Latin America, like many of the other re- through 2015, far from the goal of 23.7 per- gions, has significantly diversified its exports cent. Sub-Saharan Africa will have nearly over the last years and gained a strong three times the incidence of poverty as the foothold in global production networks. It is next poorest region (table 1.6). still, on average, relatively closed compared to South Asia, having approximately the same some other developing regions and therefore initial level of poverty as Sub-Saharan Africa will benefit from pursuing trade reforms in 1990, had already achieved impressive within the hemisphere and beyond. Improved gains by 2000 (from 42 percent to 32 percent macroeconomic conditions—lower inflation- according to the latest figures) and will likely ary expectations and more flexible exchange almost halve that level by 2015.3 rates—will provide better starting conditions According to current projections, Latin to achieve higher potential growth of around America will reach 2015 some 38 percent 2.5 percent in per capita terms. above the target, and Europe and Central Asia, The risks to the long-term forecast are nev- 90 percent above—the latter after recovering ertheless not to be minimized. Many countries sharply from the large rise in poverty during are still grappling with relatively low revenue- the transitional 1990s. collection rates, which could become critical as This year’s poverty forecast contains some countries attempt to make up for lost tariff rev- noteworthy changes from last year. The changes enues. The limits on government revenues are can be attributed to four factors—changes in straining the capacity for required investment the initial estimation of the number of poor,4 in public infrastructure—roads, ports, educa- changes in the macroeconomic relation be- tion, and other social services. The private sec- tween per capita consumption growth and tor can (and will) fill in some of the gaps, pro- poverty reduction, changes in the economic vided the right structures are in place. Conflict forecast, and changes in the population fore- could affect growth prospects, particularly at cast. The first two factors are derived from the regional level, as we have seen in some updated surveys and National Account data. parts of Sub-Saharan Africa in recent years or The third factor—changes to the economic in the Middle East and Central Asia. Finally, a forecast—are mainly attributable to changes lack of progress in trade reforms or, worse yet, in the short- and medium-term forecast with a return to protectionist tendencies could curb little or no change to the long-term forecast the expansion of trade and the ensuing bene- (with one exception, noted below). The poverty fits. The final section of this chapter assesses forecast also incorporates the World Bank’s some of the economic benefits to be derived most recent population projections. Growth from further trade reform, above and beyond in population has been revised downward in the baseline forecast presented here. many regions. If per capita consumption 45 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Table 1.6 Global poverty will decrease significantly, but not uniformly across regions Regional breakdown of poverty estimates in developing countries, various measures GEP 2003 GEP 2004 Region 1990 1999 2015 1990 2000 2015 Number of people living on less than $1 per day (millions) East Asia and Pacific 486 279 80 470 261 44 China 377 223 73 361 204 41 Rest of East Asia and Pacific 110 57 7 110 57 3 Europe and Central Asia 6 24 7 6 20 6 Latin America and the Caribbean 48 57 47 48 56 46 Middle East and North Africa 5 6 8 5 8 4 South Asia 506 487 264 466 432 268 Sub-Saharan Africa 241 315 404 241 323 366 Total 1,293 1,168 809 1,237 1,100 734 Excluding China 917 945 735 877 896 692 $1 per day headcount index (percent) East Asia and Pacific 30.5 15.6 3.9 29.4 14.5 2.3 China 32.9 17.8 5.3 31.5 16.1 3.0 Rest of East Asia and Pacific 24.2 10.6 1.1 24.1 10.6 0.5 Europe and Central Asia 1.4 5.1 1.4 1.4 4.2 1.3 Latin America and the Caribbean 11.0 11.1 7.5 11.0 10.8 7.6 Middle East and North Africa 2.1 2.2 2.1 2.1 2.8 1.2 South Asia 45.0 36.6 15.7 41.5 31.9 16.4 Sub-Saharan Africa 47.4 49.0 46.0 47.4 49.0 42.3 Total 29.6 23.2 13.3 28.3 21.6 12.5 Excluding China 28.5 25.0 15.7 27.2 23.3 15.4 Number of people living on less than $2 per day (millions) East Asia and Pacific 1,114 897 339 1,094 873 354 China 819 627 219 800 599 256 Rest of East Asia and Pacific 295 269 120 295 273 98 Europe and Central Asia 31 96 45 31 101 48 Latin America and the Caribbean 121 132 117 121 136 124 Middle East and North Africa 50 68 62 50 72 38 South Asia 1,010 1,128 1,139 971 1,052 968 Sub-Saharan Africa 386 480 618 386 504 612 Total 2,711 2,801 2,320 2,653 2,737 2,144 Excluding China 1,892 2,173 2,101 1,854 2,138 1,888 $2 per day head count index (percent) East Asia and Pacific 69.7 50.1 16.6 68.5 48.3 18.2 China 71.6 50.1 15.7 69.9 47.3 18.4 Rest of East Asia and Pacific 64.9 50.2 18.4 64.9 50.8 17.6 Europe and Central Asia 6.8 20.3 9.3 6.8 21.3 10.3 Latin America and the Caribbean 27.6 26.0 18.9 27.6 26.3 20.5 Middle East and North Africa 21.0 23.3 16.0 21.0 24.4 10.2 South Asia 89.8 84.8 68.0 86.3 77.7 59.2 Sub-Saharan Africa 76.0 74.7 70.4 76.0 76.5 70.7 Total 62.1 55.6 38.1 60.8 53.6 36.4 Excluding China 58.7 57.5 44.7 57.5 55.7 42.0 Source: World Bank staff estimates. 46 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S growth rates remain unchanged, the revised Looking forward to 2015, there are some rates of population growth will have no effect rather sharp changes in the poverty forecast, on the poverty headcount index (all else focusing on the $1/day indicator. At a global equal), but it would lower the absolute num- level, the new forecast for 2015 for the head- ber of poor. count index drops to 12.5 percent, compared One of the changes in table 1.6 is the use with 13.3 percent in last year’s report—a 5.9 of the year 2000, rather than 1999, as the percent decline. The projected number of poor base year for the forecast. Because developing in 2015 declines to 734 million, from 809 mil- countries were booming in 2000, the head- lion last year. Part of the decline in the number count index—assuming distribution neutral- of poor can be attributed to a decline in the ity—would have declined substantially.5 Most population growth rate through 2015. regions indeed witnessed a decline in the head- Compared with last year’s forecast, the count index in 2000, compared with the 1999 headcount index for East Asia declines by headcount index reported last year, except for 44 percent, from 3.9 percent to 2.3 percent. By the Middle East and North Africa, and Sub- and large, the decline is the result of a new Saharan Africa.6 At the global level, the esti- estimate of the headcount elasticity, since the mate of the headcount index in 2000 is 21.6 long-term GDP growth rates remain largely percent, compared with the 23.2 percent esti- unchanged. The other key difference is for the mate for 1999 in last year’s report—represent- Middle East and North Africa region. The in- ing a decline of 6.9 percent. cidence of poverty starts from a low level (be- However, not all of the changes can be at- tween 2 and 3 percent). With growth prospects tributed to the change in the base year. This revised upward, the headcount index is ex- year’s estimates are also based on new surveys pected to drop to 1.2 percent, compared with and methodology affecting, among others, the last year’s 2.1 percent, despite the upward re- two largest developing countries—China and vision in the base number of poor. In South India. The estimate of the number of poor in Asia, the headcount index in 2015 has been re- China has been revised downward for two vised upward from 15.7 percent to 16.4 per- reasons. The first is that, for the first time, cent, despite the decrease in the initial level. consumption-based data—deemed more ap- The new survey information suggests that the propriate for poverty analysis—are available estimate of the headcount elasticity has de- on a time-series basis. The second is that rural/ clined, as the shape of the income distribution urban population shares have been revised, curve has shifted the poor toward the left end with the urban share higher than previously of the distribution tail, away from the poverty estimated. The lower incidence of poverty in line. For Sub-Saharan Africa, the 2015 head- urban areas accounts for the downward re- count index is forecast at 42.3 percent, a drop vision in the overall poverty level for China. from last year’s 46 percent. This improvement The new survey used in India has also led to incorporates some new survey information, a downward revision in the estimate of the but positive trends in the short- and medium- poverty level. A new household survey pro- term forecast, particularly for the oil exporters, vided the basis for our revision of Pakistan’s account for most of the change. poverty profile. The Middle East and North Africa regime benefits from new surveys in Morocco and Tunisia. The estimated level of Looking ahead to the initial poverty in Yemen has raised the regional Doha Round level of poverty, but from a very low base.7 There are eleven new surveys for Latin Ameri- can countries, including Brazil and Mexico. A successful outcome of the multilateral trade negotiations known as the Doha De- velopment Agenda, or the Doha Round, would 47 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 greatly improve the growth and poverty out- subsidies, and the elimination of the use of comes discussed above. This section analyzes specific tariffs, tariff rate quotas (TRQs), and an illustrative pro-poor scenario of multilateral antidumping duties and sanctions. Specific tar- trade reform, using the World Bank’s global iffs and TRQs have unpredictable and perhaps trade model. unintended consequences for exporters8 and consumers, and—in the case of TRQs—have An illustrative pro-poor scenario proven difficult to administer. Antidumping Several proposals to improve market access for measures, in increasing use since 1995, have a merchandise trade have been tabled in advance chilling effect on many exporters by presenting of the upcoming WTO ministerial meetings them with the possibility of losing market ac- in Cancun in September 2003 (see chapter 2). cess on short notice. Most observers concur that any agreement Analyzing the impact of the pro-poor sce- reached in Cancun should meet at least three nario is not a trivial exercise, partly because of criteria—it should be simple, it should address data issues—the difficulty in quantifying spe- tariff peaks, and it should be development- cific tariffs and TRQs—and partly because of friendly. An illustrative pro-poor scenario and methodological issues—modeling TRQs and its potential impact on incomes, trade, and the impact of antidumping measures raises a poverty are discussed below. number of thorny problems. Nonetheless, Rich countries would be subject to a maxi- drawing on the work published in Global Eco- mum tariff in agriculture of 10 percent—that nomic Prospects 2002, the next section de- is, all peaks would be cut back to a maximum scribes the impact of the foregoing scenario, of 10 percent (table 1.7). The average tariff using the World Bank’s trade model (with target would be 5 percent. On the manufactur- modifications). ing side, tariff peaks would be scaled back to 5 percent, with a 1 percent target on the aver- Assessing the impact of the pro-poor age manufacturing tariff rate, which is already scenario is complex low. Developing countries would be subject to The starting point is the GTAP database—used a maximum agricultural tariff of 15 percent, by trade analysts worldwide for trade assess- with a targeted average of 10 percent (double ments. The base year is 1997, with more recent the rich-country average). In manufacturing, agricultural protection data, but excluding the peak would be capped at 10 percent; the other developments—such as recently signed targeted average would be 5 percent. regional trade agreements, China’s WTO ac- The pro-poor scenario includes elimination cession offer, and future changes to farm pro- of export subsidies, decoupling of all domestic grams (such as the recently agreed reform of the EU’s Common Agricultural Program). Tak- ing these points into consideration, together Table 1.7 Tariffs could be cut clearly with other features of the GTAP dataset, the and simply Pro-poor tariff targets by type of country and sector (percent) impact assessment likely will overstate some of the benefits of the pro-poor scenario. It is also Industrial Developing possible, however, that other factors, such as Agriculture the pro-competitive and dynamic effects of Average 5 10 trade reform, could prove our impact assess- Maximum 10 15 ment to be an underestimate. Manufacturing Summary indications of the scenario’s im- Average 1 5 pact on tariffs using the GTAP level of pro- Maximum 5 10 tection are presented in table 1.8. The table is Source: World Bank. divided into two broad sectors—agriculture (in- 48 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Table 1.8 The pro-poor tariff scenario would significantly lower protection Analysis of initial and final average tariffs in agriculture and manufacturing under the World Bank’s tariff scenario, using World Bank trade model (percent) Agriculture Manufacturing Percent Percent Average tariff reduction Average tariff reduction in non-peak in non-peak Initial Final tariffs Initial Final tariffs OECD Cairns countries 15.9 6.0 100.0 2.2 1.0 56.3 European Union with EFTA 22.4 7.4 100.0 4.2 1.2 100.0 United States 10.8 5.0 60.6 2.4 1.0 57.7 Japan 50.3 10.0 100.0 1.6 0.9 0.0 Korea, Rep. of, and Taiwan (China) 49.4 8.2 100.0 5.7 3.1 100.0 Hong Kong (China) and Singapore 1.6 1.6 0.0 0.0 0.0 0.0 Brazil 12.0 10.0 21.3 14.6 6.9 100.0 China 38.8 11.0 100.0 13.8 8.6 100.0 India 25.9 13.2 100.0 21.7 8.8 100.0 Indonesia 9.0 7.4 0.0 8.6 5.0 29.0 Russian Federation 11.8 10.0 12.3 12.3 5.6 100.0 Mexico 23.6 10.0 61.5 2.7 2.3 0.0 Southern African Customs Union 45.7 12.2 100.0 9.4 5.0 32.4 Vietnam 32.5 11.4 100.0 16.9 5.2 100.0 Rest of South Asia 23.6 10.3 100.0 31.3 9.9 100.0 Rest of East Asia 22.9 11.9 100.0 7.9 5.0 29.6 Rest of Latin America and Caribbean 14.5 10.0 77.9 12.0 6.3 100.0 EU accession countries 24.9 12.8 100.0 7.5 5.0 35.5 Rest of Europe and Central Asia 22.4 10.0 92.5 5.0 4.7 0.0 Middle East 76.4 12.8 100.0 8.4 5.0 49.4 North Africa 35.5 10.2 100.0 18.6 9.1 100.0 Rest of Sub-Saharan Africa 20.6 10.5 100.0 16.0 8.8 100.0 Rest of WORLD 8.1 6.0 0.0 7.5 5.0 15.6 Note: Agricultural tariff peaks are cut to 10 and 15 percent respectively in high-income and developing countries. Manufacturing tariff peaks are cut to 5 and 10 percent respectively in high-income and developing countries. The targets for average agricultural tariffs are respectively 5 and 10 percent, respectively, for high-income and developing countries. Targets for the average manufac- turing tariff are 1 and 5 percent, respectively, for high-income and developing countries. Sources: GTAP release 5.3 and World Bank staff calculations. cluding processed foods) and manufacturing. cases cutting the peaks is sufficient to achieve The first column in each sector provides the (im- or surpass the targeted tariff level. This is the port weighted) average tariff.9 Thus for Japan, case, for example, for Indonesian agriculture, the average tariff is 50 percent in agriculture but where the average tariff is already below the less than 2 percent in manufacturing. target and reducing the peaks simply drops the The second column shows the targeted average. average tariff rate. The rich-country agricul- The third column provides the level of re- tural target of 5 percent is achieved in the duction in non-peak tariffs. In the United United States, for example. In Japan and some States, for example, non-peak agricultural tar- regions, it is not achieved. In Japan, for exam- iffs would be reduced by an average of 60 per- ple, the scenario would lead to an average rate cent to achieve the 5 percent average target. of 10 percent, twice the target rate, because Rich countries would see a significant drop Japan’s peaks are high and cover a wide per- in the average agricultural tariff. In the ad- centage of agricultural trade. Initial tariffs vanced countries of Asia, in particular, average below the peaks would have to become nega- agricultural tariffs would fall from near 50 per- tive for the target to be achieved.10 In some cent to somewhere around 10 percent—still 49 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 above the target level. The United States would achieve the target, but from a more modest 11 Figure 1.33 The pro-poor reform scenario percent initial level, whereas the European promises substantial income gains Union would see a drop in the average (extra- Change in real income in 2015 relative to baseline ($1997 billion) EU) tariff from 22 percent to just above 7 per- 350 cent. Developing countries would also see sig- nificant reductions, higher on average than in 300 rich countries. Many regions would not meet the 10 percent target because of the many tar- 250 iffs higher than the 15 percent cutoff point. Static gains Dynamic gains 200 In manufacturing, the average tariff rate in ᮡ rich countries would decline slightly following 150 ᮡ the removal of peaks. In developing countries, the 5 percent target could be met in many 100 regions, with Mexico and the EU-accession 50 countries already below the target with their prevailing initial tariff rates.11 0 The “decoupling” part of the scenario is High-income countries Developing countries achieved by removing all domestic support in Source: World Bank staff simulations. agriculture—input and output subsidies and payments to land and capital. These would be replaced by direct payments to farm house- and rich countries, could be much higher if dy- holds. Such lump-sum transfers, not modeled namic effects—such as increases in productiv- explicitly, are important in determining in- ity14 and increasing FDI—are taken into con- come distribution within each region, but they sideration. Varying the trade model to link have no direct effects on the distribution of sectoral productivity to the export/output ratio production, on income across regions, or on shows that dynamic effects can indeed be sub- the results described below. Export subsidies, stantial. In developing countries, the dynamic which have played a relatively small role re- gains from the trade reform scenario are some cently, are also removed. The impacts of spe- 120 percent higher than the static gains. The cific tariffs and TRQs in agriculture are cap- dynamic gain for rich countries is less dra- tured in the base protection data and are matic, because of the low GDP weight of agri- modeled as part of the ad valorem equivalent culture, for which protection is strongest. tariff. The full ad valorem tariff is subject to The reduction of trade barriers in agricul- the reductions described in table 1.8. ture and food yield $193 billion (in 2015), two- thirds of the total static gains from merchandise The scenario would yield significant gains trade reform of $291 billion (table 1.9). More The scenario described and analyzed above than 50 percent of these gains in agriculture would generate $291 billion in global eco- and food, $101 billion, are reaped by develop- nomic gains—nearly 75 percent of the total ing countries, of which 80 percent is the result potential gains from full merchandise trade re- of own-reform in these two sectors. In other form.12 Measured in static terms, some words, reform of agriculture and food in rich $159 billion13 in additional income would be countries would lead to a gain of some $20 bil- reaped by developing countries in 2015 (com- lion for developing countries as a whole. Man- pared to the baseline); rich countries would ufacturing liberalization by rich countries gain around $132 billion (figure 1.33). The would lead to gains of $25 billion to develop- gains, which would raise income levels by 1.5 ing countries and could potentially even lead to and 0.5 percent, respectively, for developing a (small) loss to rich countries as increased mar- 50 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Table 1.9 A large share of real income gains comes from lowering of barriers in agriculture and food (real income gains in 2015 relative to the baseline in $1997 billion) Liberalizing region Low- and middle- High-income income countries countries All countries Decomposition of static impacts Gains to low- and middle-income countries Agriculture and food 80 20 101 Manufacturing 33 25 58 All merchandise trade 114 44 159 Gains to high-income countries Agriculture and food 23 64 91 Manufacturing 44 –3 41 All merchandise trade 67 63 132 Global gains Agriculture and food 103 84 193 Manufacturing 77 22 98 All merchandise trade 181 107 291 Decomposition of dynamic impacts Gains to low- and middle-income countries Agriculture and food 167 75 240 Manufacturing 95 9 108 All merchandise trade 265 85 349 Gains to high-income countries Agriculture and food 19 100 117 Manufacturing 36 13 48 All merchandise trade 55 115 169 Global gains Agriculture and food 185 174 358 Manufacturing 131 22 156 All merchandise trade 321 199 518 Source: World Bank staff simulations. ket access by developing countries generates 1.34). Globally, merchandise trade would in- terms-of-trade losses for rich countries—losses crease by about 10 percent (more than not fully compensated by the gains in allocative $800 billion), but exports from developing efficiency. There is a degree of asymmetry in the countries would rise by 20 percent (nearly impacts of the reforms on rich and on develop- $540 billion). The largest percentage increase ing countries. The former gain significantly in trade (nearly 50 percent) would occur in more from their own reforms ($114 billion of processed foods. Agricultural trade would rise the $159 billion total), whereas the impacts by 32 percent. Developing countries should for rich countries are more or less evenly spread see an increase in their exports of textiles, between own-reform and developing-country clothing, and footwear, although its magni- reforms. tude would depend on the final implementa- Trade would increase sharply under the tion of the Uruguay Round. scenario—particularly in the most severely The number of poor would decline sub- protected sectors: agriculture and food (figure stantially under the partial reform scenario de- 51 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Figure 1.34 Exports should rise sharply Change in export volumes in 2015 relative to baseline ($1997 billion) 200 150 Developing countries ᮡ 100 High-income countries ᮡ 50 0 Agriculture Energy Processed Textile, clothing Other foods and footwear manufacturing Source: World Bank staff simulations. scribed here. At the world level, the number of in absolute terms would come in Sub-Saharan persons living on $1/day or less would decline Africa. The region’s unskilled workers would by 61 million, or 8 percent of the current fore- see the largest percentage increase in nominal cast for 2015 of 734 million (figure 1.35).15 wages and decreases in the cost of living. The The number living on $2/day or less would de- largest percentage fall would occur in the cline by 144 million. The greatest reduction Middle East. This region has the highest over- Figure 1.35 Millions of people would be moved out of poverty Changes in number of poor in 2015 relative to base (millions) 70 60 50 40 30 $2 per day 20 ᮡ 10 $1 per day ᮡ 0 East Asia Europe and Latin America Middle East South Asia Sub-Saharan and Pacific Central Asia and the and Africa Caribbean North Africa Source: World Bank staff simulations. 52 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Figure 1.36 Gains for most, but adjustment costs for some Percent change in rural value added in 2015 relative to the baseline Latin America Sub-Saharan Africa Europe and Central Asia Middle East and North Africa South Asia Rest of East Asia China NIEs Other OECD United States Japan Europe –30 –20 –10 0 10 20 30 40 50 60 Percent Note: The negative impacts on Chinese farmers could be overstated since the baseline simulation does not incorporate the impacts of WTO accession. Source: World Bank staff simulations. all barriers to imports and a substantial tax transitional impacts will be mitigated to the ex- on consumers. However, the region has a rela- tent that rural economies are diversified. tively low level of poverty, particularly com- As can be seen in Figure 1.36, rural value pared to Sub-Saharan Africa and South Asia. added in Europe and in Japan could decline The positive impact on overall growth, ac- by more than 20 percent over the long term. companied by a sharp boost in trade and a And within agriculture, the distribution of the poverty outlook improvement leaving all re- impacts is likely to be highly differentiated— gions better off in aggregate, does not signify across sectors within agriculture, as well as by that the reforms are without adjustment costs, factor ownership. For example, tenant farmers even over the long term.16 For example, given could be better off than landowners because the the levels of protection in the agricultural sec- price of land is expected to fall in most OECD tors, particularly in the OECD countries, farm- regions with the removal of protection. Farm- ers stand to lose the most from reductions in ers in some developing regions could also wit- protection. The change in agricultural incomes ness a decline in overall agricultural income— needs to be put in context. First, the adjust- particularly in China and the Middle East and ment will occur over a 10-year period, allow- North Africa. On the other hand, farmers in ing for a gradual adjustment. Second, the ad- Canada, Australia, and New Zealand will reap justment in most countries will be limited to a significant rewards from this reform, as will small share of GDP. In high-income countries, farmers in the rest of East Asia (for example in agricultural output is less than 3 percent of new market access for rice and vegetable oils), total output on average. For developing re- Latin America (grains, livestock, and sugar), gions, agricultural output varies from a low of and in Sub-Saharan Africa (sugar). 7 percent for upper middle-income countries These adjustments will be accompanied by to 24 percent for low-income countries. Addi- structural shifts in world agricultural and food tionally, manufacturing will expand, and the output, following closely the patterns of changes 53 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Figure 1.37 Significant shifts in global output patterns Percent change in output in 2015 relative to the baseline Latin America Sub-Saharan Africa Europe and Central Asia Middle East and North Africa South Asia Rest of East Asia Processed foods China ᮡ NIEs Other OECD ᮡ United States Agriculture Japan Europe –30 –20 –10 0 10 20 30 40 Percent Source: World Bank staff simulations. in rural value added (figure 1.37). The changes for both agriculture and processed foods. The in agricultural output tend to dominate those beneficiaries include Canada, Australia, and in processed foods, despite relatively similar New Zealand among rich countries; Asia out- levels of protection, in part because of the side of China; Sub-Saharan Africa; and Latin lower costs of inputs—that is, raw agricultural America. The NIEs—particularly Korea and commodities—for the processing sectors. The Taiwan (China)—show that they could be main reductions in output occur in Europe, competitive in processed foods were they to re- Japan, and the Middle East and North Africa move agricultural protection. 54 Annex 1 Historical trade dynamics for developing countries Table 1.A1 Sectoral export decomposition for developing countriesa Percent Share Average growth Contribution to (percent) (percent per annum) growth (percent) 1977 1987 1997 1977–87 1987–97 1977–87 1987–97 High income Asia excluding Japan b Agriculture 5.1 1.8 0.3 4.6 –5.9 0.9 –0.4 Oil and gas 0.6 0.1 0.1 –6.5 12.6 –0.1 0.1 Other natural resources 0.9 0.4 0.2 5.6 5.5 0.2 0.1 Processed foods 8.7 5.1 2.3 9.7 3.1 4.0 0.9 Textiles, apparel, and leather goods 30.0 25.1 13.5 13.8 4.9 23.7 7.7 Motor vehicles and parts 0.4 1.8 3.4 33.6 19.0 2.2 4.2 Electronic equipment 10.9 17.1 31.7 21.2 18.7 19.0 39.0 Other machinery 12.6 17.3 22.4 19.6 14.5 18.7 24.9 Other manufacturing 30.9 31.4 26.1 16.0 9.6 31.5 23.4 Total 100.0 100.0 100.0 15.9 11.6 100.0 100.0 East Asia and Pacific Agriculture 19.1 13.1 2.2 7.0 –1.8 9.9 –0.6 Oil and gas 25.3 20.7 3.6 8.9 –1.4 18.3 –0.7 Other natural resources 11.3 5.5 2.1 3.5 6.3 2.4 1.2 Processed foods 20.3 15.2 6.9 8.0 8.5 12.5 4.8 Textiles, apparel, and leather goods 5.1 14.0 22.4 23.0 23.0 18.8 24.5 Motor vehicles and parts 0.1 0.7 0.7 36.9 17.3 1.0 0.7 Electronic equipment 1.8 7.2 20.9 27.7 30.7 10.0 24.4 Other machinery 2.8 4.4 13.8 16.2 31.5 5.3 16.1 Other manufacturing 14.3 19.1 27.4 14.4 21.7 21.7 29.5 Total 100.0 100.0 100.0 11.1 17.4 100.0 100.0 South Asia Agriculture 27.8 22.1 5.9 5.3 –1.1 16.9 –1.0 Oil and gas 0.3 1.7 0.1 29.4 –15.9 2.9 –0.6 Other natural resources 13.4 16.9 5.2 10.3 0.4 20.1 0.3 Processed foods 19.8 12.8 10.3 3.2 10.5 6.5 9.3 Textiles, apparel, and leather goods 14.0 27.9 46.5 15.4 18.8 40.2 54.4 Motor vehicles and parts 0.9 0.6 1.2 3.9 20.6 0.4 1.4 Electronic equipment 0.3 0.4 1.2 10.5 25.0 0.5 1.6 Other machinery 5.3 4.7 4.5 6.4 12.5 4.1 4.5 Other manufacturing 18.2 13.0 25.0 4.2 20.5 8.4 30.1 Total 100.0 100.0 100.0 7.8 12.8 100.0 100.0 Latin America and the Caribbean Agriculture 26.0 17.8 10.0 4.3 2.8 11.1 4.1 Oil and gas 14.2 16.3 6.9 9.7 0.0 17.9 0.0 Other natural resources 7.1 5.8 4.4 6.0 6.0 4.7 3.4 Processed foods 20.1 14.1 12.0 4.5 7.1 9.1 10.4 Textiles, apparel, and leather goods 2.7 3.8 8.5 12.2 18.1 4.7 12.0 Motor vehicles and parts 1.5 3.6 8.4 18.5 18.6 5.3 12.0 Electronic equipment 1.7 3.1 5.9 14.9 16.2 4.2 7.9 Other machinery 3.6 7.6 13.3 16.7 15.1 11.0 17.5 Other manufacturing 23.1 27.9 30.6 10.4 9.9 31.9 32.6 Total 100.0 100.0 100.0 8.3 8.9 100.0 100.0 (Table continues on next page) 55 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Table 1.A1 (continued) Percent Share Average growth Contribution to (percent) (percent per annum) growth (percent) 1977 1987 1997 1977–87 1987–97 1977–87 1987–97 Europe and Central Asia Agriculture 22.8 14.7 5.2 5.2 10.6 9.6 3.8 Oil and gas 0.2 2.6 11.0 45.1 41.7 4.2 12.2 Other natural resources 18.6 9.7 4.7 3.1 14.1 4.2 3.9 Processed foods 15.0 13.4 6.3 8.7 13.8 12.4 5.2 Textiles, apparel, and leather goods 9.9 18.0 12.6 16.8 18.4 23.2 11.8 Motor vehicles and parts 1.3 1.5 3.0 11.7 31.4 1.7 3.2 Electronic equipment 0.6 0.6 2.1 10.3 38.8 0.6 2.3 Other machinery 10.1 8.5 9.3 8.1 24.0 7.5 9.5 Other manufacturing 21.6 30.9 45.9 14.0 27.7 36.8 48.1 Total 100.0 100.0 100.0 9.9 22.8 100.0 100.0 Middle East and North Africa Agriculture 1.8 2.4 1.9 2.5 2.1 –15.9 1.0 Oil and gas 86.2 69.4 42.4 –2.5 –0.4 586.9 –5.0 Other natural resources 1.0 2.3 2.0 8.1 3.0 –37.8 1.4 Processed foods 0.9 1.5 1.4 4.7 4.6 –16.0 1.4 Textiles, apparel, and leather goods 1.0 2.1 5.8 7.1 15.6 –31.6 12.2 Motor vehicles and parts 0.5 0.3 0.5 –5.9 11.2 6.6 0.9 Electronic equipment 0.2 0.6 2.0 13.6 18.0 –12.6 4.4 Other machinery 0.8 2.4 4.1 11.4 10.1 –47.7 6.9 Other manufacturing 7.6 19.0 39.9 9.3 12.7 –332.0 76.7 Total 100.0 100.0 100.0 –0.3 4.6 100.0 100.0 Sub-Saharan Africa Agriculture 23.9 20.4 13.2 1.1 0.9 8.8 2.8 Oil and gas 37.9 38.2 18.3 2.8 –2.1 39.2 –10.8 Other natural resources 13.3 12.7 12.2 2.2 4.9 10.6 11.4 Processed foods 8.9 7.0 6.5 0.3 4.5 0.9 5.6 Textiles, apparel, and leather goods 0.3 1.2 4.7 16.5 20.8 4.0 9.7 Motor vehicles and parts 0.1 0.2 0.9 9.2 25.2 0.4 2.0 Electronic equipment 0.1 0.2 0.4 12.7 13.1 0.6 0.7 Other machinery 1.0 1.7 4.1 8.6 15.1 4.0 7.6 Other manufacturing 14.4 18.4 39.8 5.3 13.8 31.5 70.9 Total 100.0 100.0 100.0 2.7 5.4 100.0 100.0 Low- and middle-income countries Agriculture 13.4 12.6 5.2 4.1 1.9 11.2 1.4 Oil and gas 51.6 33.5 12.0 0.4 0.5 3.3 0.8 Other natural resources 6.3 6.3 3.8 4.8 5.9 6.3 2.5 Processed foods 9.5 10.0 7.4 5.4 8.0 10.9 6.0 Textiles, apparel, and leather goods 2.6 7.1 15.5 16.1 20.3 14.8 19.8 Motor vehicles and parts 0.6 1.4 2.8 13.6 19.4 2.6 3.5 Electronic equipment 0.7 2.7 9.8 20.4 26.6 6.1 13.5 Other machinery 2.1 4.7 10.6 13.8 20.8 9.1 13.6 Other manufacturing 13.4 21.7 33.0 10.0 16.1 35.7 38.9 Total 100.0 100.0 100.0 4.8 11.3 100.0 100.0 a. The years represent three-year averages to remove some of the volatility from the data. Thus 1977 represents the average of 1975–77, 1987 is the average of 1985–87, and 1997 is the average of 1995–97. Relative caution is advised regarding Europe and Central Asia, where data prior to 1990 are not always reliable. This would have only a small impact on the total because of the region’s relatively small weight. b. High-income Asia is excluded from the low- and middle-income region in the totals. It is provided for information and includes Hong Kong (China), Republic of Korea, Singapore, and Taiwan (China). Source: GTAP release 5.0. 56 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Table 1.A2 Regional export decomposition for developing countriesa Percent Share Average growth Contribution to (percent) (percent per annum) growth (percent) 1977 1987 1997 1977–87 1987–97 1977–87 1987–97 High income Asia excluding Japanb Quad countries 65.6 69.4 47.0 16.5 7.4 70.5 35.9 High income Asia excluding Japan 7.8 8.6 15.1 16.9 18.1 8.8 18.3 East Asia and Pacific 8.5 9.2 25.0 16.9 23.3 9.5 32.8 South Asia 0.7 2.1 1.6 28.7 8.5 2.5 1.3 Latin America and the Caribbean 1.3 1.1 2.8 13.5 22.6 1.0 3.6 Europe and Central Asia 0.2 0.2 1.6 13.2 39.3 0.2 2.3 Middle East and North Africa 7.1 3.9 2.1 9.1 5.2 2.9 1.3 Sub-Saharan Africa 2.6 1.2 1.2 7.1 11.8 0.8 1.2 Rest of the world 6.1 4.4 3.6 12.0 9.5 3.9 3.2 World total 100.0 100.0 100.0 15.9 11.6 100.0 100.0 East Asia and Pacific Quad countries 66.1 58.6 54.8 9.8 16.6 54.6 53.8 High income Asia excluding Japan 18.7 26.2 26.2 15.0 17.4 30.3 26.2 East Asia and Pacific 4.8 4.9 7.1 11.4 21.7 5.0 7.6 South Asia 1.3 2.1 1.5 16.4 13.2 2.6 1.3 Latin America and the Caribbean 2.4 1.3 2.0 4.8 22.2 0.7 2.1 Europe and Central Asia 0.3 0.7 1.4 21.3 26.6 0.9 1.6 Middle East and North Africa 2.9 2.6 2.6 10.1 17.5 2.5 2.7 Sub-Saharan Africa 1.4 0.7 1.0 3.6 21.5 0.3 1.1 Rest of the world 2.1 2.8 3.4 14.2 19.8 3.1 3.6 World total 100.0 100.0 100.0 11.1 17.4 100.0 100.0 South Asia Quad countries 52.7 64.8 62.3 10.0 12.4 75.6 61.2 High income Asia excluding Japan 4.4 7.3 9.5 13.4 16.0 9.8 10.5 East Asia and Pacific 4.6 2.8 7.3 2.7 24.0 1.3 9.1 South Asia 4.7 4.8 3.5 8.1 9.1 5.0 2.9 Latin America and the Caribbean 0.6 0.6 1.6 7.8 24.2 0.6 2.0 Europe and Central Asia 2.0 1.2 2.9 2.7 22.9 0.5 3.5 Middle East and North Africa 22.2 12.8 7.4 2.0 6.8 4.3 5.1 Sub-Saharan Africa 4.1 2.2 2.8 1.3 15.9 0.5 3.1 Rest of the world 4.7 3.5 2.8 4.5 10.5 2.4 2.5 World total 100.0 100.0 100.0 7.8 12.8 100.0 100.0 Latin America and the Caribbean Quad countries 75.0 77.2 68.9 8.6 7.7 79.1 62.8 High income Asia excluding Japan 0.7 1.9 2.9 20.1 13.6 2.9 3.7 East Asia and Pacific 1.0 1.6 2.8 13.6 15.0 2.1 3.7 South Asia 0.3 0.6 0.4 15.2 5.9 0.8 0.3 Latin America and the Caribbean 15.9 12.0 20.2 5.3 14.7 8.8 26.2 Europe and Central Asia 1.0 0.7 1.0 5.2 12.5 0.5 1.2 Middle East and North Africa 3.3 2.9 1.8 6.9 4.1 2.6 1.1 Sub-Saharan Africa 1.0 1.4 0.8 11.4 2.8 1.6 0.3 Rest of the world 1.8 1.6 1.1 7.1 4.6 1.5 0.7 World total 100.0 100.0 100.0 8.3 8.9 100.0 100.0 (Table continues on next page) 57 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Table 1.A2 (continued) Percent Share Average growth Contribution to (percent) (percent per annum) growth (percent) 1977 1987 1997 1977–87 1987–97 1977–87 1987–97 Europe and Central Asia Quad countries 75.7 68.3 57.9 8.8 20.8 63.6 56.4 High income Asia excluding Japan 0.9 0.5 3.2 4.1 46.2 0.3 3.5 East Asia and Pacific 0.6 2.3 5.5 26.0 33.7 3.4 5.9 South Asia 1.1 1.5 0.6 13.1 11.5 1.7 0.4 Latin America and the Caribbean 2.6 1.5 1.3 4.0 21.2 0.8 1.3 Europe and Central Asia 0.9 1.9 23.5 18.7 57.5 2.6 26.7 Middle East and North Africa 10.9 19.4 3.9 16.4 4.6 24.8 1.6 Sub-Saharan Africa 1.5 0.4 0.3 –4.2 20.7 –0.3 0.3 Rest of the world 5.7 4.2 3.9 6.4 22.0 3.1 3.9 World total 100.0 100.0 100.0 9.9 22.8 100.0 100.0 Middle East and North Africa Quad countries 77.3 74.6 66.6 –0.7 3.4 158.5 52.4 High income Asia excluding Japan 5.1 6.5 11.3 2.2 10.6 –37.8 19.8 East Asia and Pacific 1.9 2.1 6.3 0.5 16.7 –3.1 13.6 South Asia 1.0 3.5 3.6 13.3 4.6 –75.5 3.6 Latin America and the Caribbean 6.0 3.2 1.9 –6.4 –0.7 89.1 –0.4 Europe and Central Asia 1.1 2.7 2.8 9.1 5.2 –46.0 3.1 Middle East and North Africa 4.6 4.8 4.0 0.0 2.8 0.4 2.6 Sub-Saharan Africa 0.9 0.7 1.4 –2.7 11.6 6.8 2.5 Rest of the world 2.1 1.9 2.2 –1.2 6.1 7.6 2.8 World total 100.0 100.0 100.0 –0.3 4.6 100.0 100.0 Sub Saharan Africa Quad countries 83.7 84.4 70.8 2.8 3.5 86.6 50.9 High income Asia excluding Japan 1.1 2.3 6.5 10.6 16.6 6.3 12.5 East Asia and Pacific 0.6 0.8 4.8 4.6 26.5 1.2 10.6 South Asia 0.5 1.0 1.9 8.7 12.9 2.3 3.3 Latin America and the Caribbean 3.3 3.9 2.9 4.4 2.5 5.7 1.6 Europe and Central Asia 0.3 0.6 1.3 11.1 14.1 1.6 2.3 Middle East and North Africa 2.0 1.4 1.7 –0.7 7.4 –0.5 2.1 Sub-Saharan Africa 3.9 4.0 7.2 3.1 11.7 4.5 11.9 Rest of the world 4.6 1.7 2.9 –7.1 11.4 –7.8 4.7 World total 100.0 100.0 100.0 2.7 5.4 100.0 100.0 Low and middle income Quad countries 75.5 72.1 61.7 4.3 9.6 66.5 56.2 High income Asia excluding Japan 5.1 8.9 13.5 10.8 16.0 15.3 15.9 East Asia and Pacific 2.1 2.6 5.6 7.2 20.3 3.5 7.2 South Asia 1.0 2.0 1.5 12.8 8.5 3.8 1.3 Latin America and the Caribbean 6.9 5.1 6.1 1.8 13.2 2.2 6.6 Europe and Central Asia 0.9 1.3 4.4 9.3 25.7 2.1 6.0 Middle East and North Africa 4.5 4.4 2.9 4.5 7.0 4.2 2.2 Sub-Saharan Africa 1.5 1.3 1.4 3.0 12.2 0.9 1.4 Rest of the world 2.6 2.2 2.9 3.3 14.2 1.6 3.2 World total 100.0 100.0 100.0 4.8 11.3 100.0 100.0 a. The years represent three-year averages to remove some of the volatility from the data. Thus 1977 represents the average of 1975–77, 1987 is the average of 1985–87, and 1997 is the average of 1995–97. Relative caution is advised regarding Europe and Central Asia, where data prior to 1990 are not always reliable. This would have only a small impact on the total because of the region’s relatively small weight. b. High-income Asia is excluded from the low- and middle-income region in the totals. It is provided for information and includes Hong Kong (China), Republic of Korea, Singapore, and Taiwan (China). Source: GTAP release 5.0. 58 G L O B A L O U T L O O K A N D T H E D E V E L O P I N G C O U N T R I E S Notes 10. The formula for cutting the non-peak tariffs is 1. Technically, the non-economic component of given by the following expression: confidence is modeled as an unobserved state variable τ1 i = χ ⋅ τi for i ∈ {Agric, Manu} where 0 and estimated using a Kalman filter/smoother, with data pertaining to economic conditions as exogenous τ 1 a ∑ Mi − ∑ τ1 p Mi i ∈{Peak} χ= i controls. The model was fitted on monthly data over the period 1984–2001 (September). Then the estimate ∑ τ M0 i i i ∈{Peak} was used together with actual, observed economic data to predict confidence out of sample over the period Oc- The target average is τa1. The import levels are given by tober 2001–April 2003. M. All tariff peaks are reduced to τp1. All other tariffs 2. It was decided at the December 2002 Copen- are reduced by the factor χ, given by the formula hagen Summit to invite eight transition countries to above. There is nothing preventing the adjustment fac- join the EU in May 2004. tor from being above 1 (that is, initial tariffs could in- 3. The differences between Sub-Saharan Africa and crease to achieve the target reduction) or below 0 (that South Asia are not limited to the higher growth rate in is, non-peak tariffs would have to become negative to the latter over the last decade. The poverty gap indi- achieve the target). In the illustrative pro-poor sce- cator—a measure of the average distance to the poverty nario, the reduction factor is capped above by 1—in line for the poor—has been much higher in Sub- which case the average will be below the target, and Saharan Africa. This implies that even at identical growth bound below by 0 in which case the average will be rates, poverty would have decreased more rapidly in above the target. This is the case of Japan in agricul- South Asia. ture. An alternative would be to further reduce the 4. The initial poverty estimate will also be subject peaks so that the target is achieved. to revisions in the national income and product ac- 11. Setting the level of the maximum tariff can have counts because adjustments to the survey-based con- some unintended consequences. In the case of develop- sumption levels will follow adjustments to consump- ing country manufacturing tariffs, for example, setting tion derived from the national accounts. the maximum tariff to 20 percent instead of 15 percent 5. A numerical example may help clarify the proce- could actually reduce the overall average. The reason is dure. Say population is 1,000 in 1999 and per capita that all tariffs in the 15 to 20 percent band in the lat- consumption growth in 2000 is 5 percent and popula- ter (15 percent) case are bound to below 15 percent. In tion growth is 1 percent. If the headcount index is the former (20 percent) case those tariffs would be re- 40 percent in 1999, then the number of poor is 400. duced by a certain percentage amount—perhaps even With a headcount elasticity of 1.5, the headcount index 100 percent. The difference in the average tariff will would improve from 40 to 37 percent (40*(1–1.5 depend on the number of tariffs within the 15 to 20 *0.05)), assuming there is no change in the distribution percent band and the relative import weights. of income. Total population in 2000 reaches 1,010, 12. The model is similar to that used to produce the thus the number of poor is 374, a decline of 6.5 per- gains from trade reform in Global Economic Prospects cent, whereas the headcount index improves by 7.5 per- 2002. More information regarding the nature of cent. The difference is the population growth rate. the simulations and a more detailed description of the 6. These comparisons are not strictly exact because results is given in that report. The model documen- last year’s 1999 levels do not incorporate the new sur- vey information nor other adjustments to the historical tation is available at //http://www.worldbank.org/ data. prospects/pubs/techref.pdf. Global Economic Prospects 7. The survey coverage in the Middle East and 2002 reported global merchandise trade reform gains North Africa region is particularly sparse compared to of $355 billion (in the static exercise). The gains have most other developing regions. Per capita GDP across risen modestly in this report—essentially for two rea- the region tends to be relatively high and therefore sons. We have upgraded the database from release 5.0 poverty rates low, so small changes to poverty levels in of the GTAP data set to release 5.3, and we have a new one or two sizeable countries with a fair number of baseline reflecting two years of observed changes in poor—for example Egypt or Yemen—can have a dis- economic performance and a (minor) reevaluation of proportionate impact on the regional poverty level. long-term growth prospects. The current model is also 8. For example specific tariffs penalize more com- based on a different regional and sectoral aggregation, petitive exporters with relatively lower prices. which can affect the impacts of trade reform. (There 9. In the case of the European Union, intra-EU are more regions and sectors and, all else being equal, trade is excluded from the average tariff calculation one would expect this to raise the real income gains by and the reduction. removing some of the aggregation bias of trade policy 59 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 instruments). There have also been a few minor changes 16. The positive income gains—identified in virtu- to model specification and parameters. ally all of the model’s regions—imply that transitional 13. All figures, unless otherwise stated, refer to mechanisms can be implemented, leaving everyone bet- changes in 2015 compared with the baseline level. Dol- ter off. Whether these mechanisms are designed and lar amounts are measured in real terms and are based implemented is an important issue, but typically the de- on 1997 dollars. Figures can be converted to 2003 cision of local governments. terms by adjusting for economic growth and inflation. The former would involve dividing a figure by References (1.03)^(2015–2003), where it is assumed that the Greener, Robert. 2002. “AIDS and Its Macroeconomic global economy grows at 3 percent per annum in real Impact.” In Steven Forsythe, ed., State of the Art: terms between 2003 and 2015. The inflation adjust- AIDS and Economics. International AIDS and ment involves multiplying a figure by (1.025)^ Economics Network, 49–55. Available at http:// (2003–1997) where it is assumed that the rate of infla- www.iaen.org/library/statepidemic/chapter7.pdf. tion is 2.5 percent per annum between 1997 and 2003. Moatti, J. P., I. N’Doye, S. M. Hammer, P. Hale, and M. (Both growth and inflation rates are approximations.) D. Kazatchkine. 2002. “Antiretroviral Treatment The total adjustment factor is 81 percent, so that the for HIV-Infected Adults and Children in Develop- global gain of $295 billion in 2015 is more or less ing Countries: Some Evidence in Favor of Ex- equivalent to $235 billion in 2003 global GDP and panded Diffusion.” In Steven Forsythe, ed., State prices. of the Art: AIDS and Economics. International 14. Few dispute that trade openness will improve AIDS and Economics Network, 96–117. Avail- productivity. There is nonetheless great incertitude able at http://www.iaen.org/library/statepidemic/ about the channels—greater domestic competitiveness, chapter12.pdf. imports of technology-laden goods, FDI, export-driven UNAIDS. 2002. Fact Sheet: Regional Roundup: Sub- competitiveness—and the magnitude. The results re- Saharan Africa. July. Available at http://www. ported herein are intended to illustrate the potential unaids.org/barcelona/presskit/factsheets/FSssafrica_ magnitudes. en.pdf 15. This compares with a reduction of 114 million World Bank. 2003. Global Development Finance 2003. persons in the case of full merchandise trade reform. Washington, D.C.: World Bank. 60 2 Trade Patterns and Policies: Doha Options to Promote Development Developing countries have become major Central Asia, and South Asia, but only 60 per- players in the global economy cent in Latin America. Sub-Saharan Africa and Over the past two decades, developing coun- the Middle East and North Africa have yet to tries have increased their share of global trade reach the 30 percent mark, and many coun- from about one-quarter to one-third. As a tries—particularly poor countries—remain de- group, they have moved beyond their tradi- pendent on exports of agricultural and resource tional specialization in agricultural and re- commodities. source exports into manufactures. Countries The rising tide of exports did not lift all that were low income in 1980 managed to raise boats. Forty three countries had no increase on exports of manufactures from roughly 20 per- average in their merchandise exports for the 20 cent of their total exports to more than 80 per- years after 1980. Of this group, 20 countries re- cent, and many have entered the ranks of mained strongly dependent on oil or other nat- today’s middle-income countries. The middle- ural resources, such as phosphates for Nauru or income group of 1980 also increased its manu- copper for Zambia. Severe conflicts undercut factured share, but somewhat less rapidly, to the performance of another 18 countries, in- reach nearly 70 percent. This dramatic change cluding Rwanda and Timor Leste. Trade em- in trade volume and composition has given de- bargos stifled the export performance of five veloping countries a new interest—and a pow- other countries, including Libya and Sudan. In erful voice—in the ongoing Doha round. almost all these countries, the investment cli- These changes are not just due to declines in mate was not sufficiently favorable—for a the prices of agricultural and resource com- range of reasons, sometimes resource depletion, modities relative to manufactures—the strong sometimes poor economic management—to at- shift in the composition of exports shows up tract the investments needed to transform the even when price changes are removed. Further, pattern of exports. it is not just an artifact of a few large, high- growth exporters such as China and India. The Developing countries are moving into share of manufactures in the exports of devel- high-value-added products oping countries other than China and India Growth in traditional labor-intensive manufac- rose from one-tenth in 1980 to almost two- tures accounts for only part of the gain in ex- thirds in 2001. It increased sharply, but not ports of manufactures. Exports of textiles and equally, in all regions. The share of manufac- clothing from low-income countries grew at tures in merchandise exports is now between 14 percent per year between 1981 and 2001, 80 and 90 percent in East Asia, Europe and but other commodity groups grew even faster. 63 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Exports of electronic products, many of which been progressively emulated by developing did not exist in 1980, grew at 21 percent per countries over the last two decades, with the re- year—fast enough to double every few years. sult that developing countries’ agricultural ex- Further, developing countries expanded their ports grew more slowly than if agriculture had range of markets, with the share of developing- taken the liberalizing path of manufactures. country markets growing from 15 to 35 per- cent over the period. The continual move to Now comes the hard part: reducing new products and new markets helped high- protection of sensitive sectors growth exporters like India and China to avoid Developing-country exports now face obsta- sharp declines in their terms of trade, which, cles in the most sensitive sectors. Industrial- given the rapidity of their export growth, might country tariffs on manufactures from develop- otherwise have been expected. ing countries are five times higher than they Between 1991 and 2001, all regions im- are on manufactures from other industrial proved their competitiveness in the global countries. The barriers imposed by developing marketplace as measured by market share. Eu- countries on other developing countries, how- rope and Central Asia, Latin America, and ever, are even higher. Of course, protection South Asia outperformed the other regions, takes forms other than ad valorem tariffs— but all gained market share at the expense of among them quotas, specific duties, and an- the rich countries. This was not true in the tidumping duties. As with tariffs, these mea- preceding decade, when several regions lost sures tend to be used more frequently against market share, notably Europe and Central the labor-intensive products from developing Asia, the Middle East and North Africa, and countries. Average antidumping duties are ten Sub-Saharan Africa. times higher than tariffs in industrial coun- tries, and around five times higher than in de- Why did such rapid and fundamental veloping countries. In short, both groups of changes in trade patterns occur? countries impose substantially higher barriers Investments in people and in factories both on exports from developing countries. played a role. Average educational levels and capital stock per worker rose sharply through- The way in which protection is reduced will out the developing world. Also, improvements make a difference to developing countries in transport and communications, in conjunc- Several approaches—modalities—for negotia- tion with developing-country reforms, allowed tions have been proposed for reducing tariffs the production chain to be broken up into on agricultural and nonagricultural goods. The components, with developing countries play- 146 World Trade Organization (WTO) mem- ing a key role in global production sharing. bers are now discussing formulas that provide Policy was no less important. The dramatic enough discipline to bring about liberalization liberalization of tariff and nontariff barriers in and address tariff peaks and escalation, while developing countries after the mid-1980s in- offering enough flexibility to accommodate the creased developing countries’ competitiveness. constraints of all members. Besides tariffs, re- The negative impacts of protection on all ex- form of the rules on antidumping measures is a port activities declined, but more so for manu- critical priority. Antidumping measures, origi- factures and processed primary products than nally intended as a response to anticompetitive for agriculture and natural resources. Although behavior, are now widely regarded as being successive multilateral trade rounds liberalized used to facilitate market cartelization and are global manufactures, rich countries continued increasingly a source of nontransparent and to protect their agriculture. That pattern has costly protection. 64 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T Changing patterns in developing- products (figure 2.1). Manufacturing exports country exports have risen in importance in high-income coun- tries, with their share in total exports rising H istorically, developing countries have been regarded as exporters of primary commodities and importers of manufactures, from around 70 percent to more than 80 per- cent in the 20 years preceding 2001. This shift a theme repeated even in recent textbooks on was much more marked in the middle- and development (Todaro 1994). The situation has low-income countries. In the middle-income changed drastically since the beginning of the group, the share of manufactures in total ex- 1980s, however, as developing countries have ports rose from 20 percent to almost 70 per- become important exporters of manufactured cent over the period. In low-income countries, Figure 2.1 Developing countries have become important exporters of manufactured products a. Manufactured products now make up approximately b. When China and India are excluded, manufactures 80 percent of exports from low-income countries still make up more than 60 percent of exports Share of exports by sector, 1981–2001 (percent) Share of exports by sector, 1981–2001 (percent) 100 100 90 90 Manufacturing exports 80 80 ᮡ 70 70 Resources exports Resources exports Manufacturing exports ᮡ 60 60 ᮡ ᮡ 50 50 40 40 30 Agricultural exports 30 ᮡ 20 ᮡ 20 Agricultural exports 10 10 0 0 81 83 85 87 89 91 93 95 97 99 01 81 83 85 87 89 91 93 95 97 99 01 19 19 19 19 19 19 19 19 19 19 20 19 19 19 19 19 19 19 19 19 19 20 c. Manufactures make up nearly 70 percent of exports d. During the same period, export patterns of from middle-income countries high-income countries have remained stable Share of exports by sector, 1981–2001 (percent) Share of exports by sector, 1981–2001 (percent) 100 100 90 90 80 Manufacturing exports 80 ᮡ 70 ᮡ 70 Manufacturing exports 60 60 50 50 40 Resources exports 40 30 ᮡ 30 Agricultural exports 20 20 ᮡ ᮡ 10 10 ᮡ Agricultural exports Resources exports 0 0 81 83 85 87 89 91 93 95 97 99 01 81 83 85 87 89 91 93 95 97 99 01 19 19 19 19 19 19 19 19 19 19 20 19 19 19 19 19 19 19 19 19 19 20 65 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 the share of manufactures rose from 20 per- velopment must consider the impact of poli- cent to more than 80 percent. cies on trade in manufactured products.1 Nor are China and India the only countries driving these changes. Even when China and Developing countries are moving up the India are excluded, the rise in the share of man- value-added ladder ufactures is from 10 percent to more than 60 A decomposition of the growth rates of ex- percent of total exports. Clearly, China and ports from each group of countries by level of India are important, but much broader changes technology indicates that developing countries in the composition of developing-country ex- are gaining ground in higher-technology ex- ports are under way. If we eliminate the dispro- ports (table 2.1). The low-income group had portionate effects of large exporters altogether, by far the highest growth rate of total exports, by considering simple average export shares, at 14 percent per year—a rate sufficient to the average share of manufacturing exports cause exports to expand 14-fold over the 20- rose from 25 percent to 50 percent in the un- year period considered.2 The middle-income weighted low-income country group, and from countries also experienced substantially higher 28 to 48 percent in the middle-income group. growth rates than the high-income countries, The share of manufacturing exports in total suggesting that developing countries have been exports has risen sharply in all regions (figure catching up with developed countries in their 2.2). In East Asia and the Pacific, the increase trade patterns—in strong contrast with the in- began from a high base—over 50 percent—but dications of divergence observed in many anal- then increased to almost 90 percent by 2001. In yses of economic growth (Pritchett 1997). Europe and Central Asia, the manufactures At the product level, growth in exports of share began at an even higher level, over 60 raw primary products was relatively low, at percent, and rose to almost 90 percent by 2 percent per year globally. In processed agri- 2001. Because of Latin America’s strong nat- cultural products, such as meats, processed ural resource endowments (de Ferranti, Perry, foods, alcoholic beverages, tobacco products, Lederman, and Maloney 2002) the situation and processed woods, growth rates were sub- there was quite different initially, with manu- stantially higher, at 6 percent globally, 7 per- factures contributing only 20 percent of total cent for the low-income country group ex- exports in 1981. That share had almost tripled cluding China and India, and 12 percent for by 2001—to more than 60 percent. In the Mid- China and India. dle East and North Africa, resource exports, Trade in low-technology manufactures particularly oil, remain dominant, although (such as textiles and clothing), simple manu- their share fell from more than 90 percent to factures (toys, sporting goods), and iron and around 60 percent during the period under steel products grew at rates substantially above scrutiny, while the importance of manufactur- the world average rate. Exports of these prod- ing exports rose from close to zero to around ucts from the low-income country group grew 30 percent. In South Asia, manufacturing ex- at much higher rates than from other coun- ports rose from around half of total exports to try groups, with export growth rates of 14 per- more than 80 percent. Resource-based exports cent for textiles and 16 percent for other low- and agricultural exports remained important technology products. in Sub-Saharan Africa, although the share of In medium-technology manufactures, a sim- manufactures rose from 10 to 27 percent. ilar pattern emerges, with global growth rates Clearly, agricultural and resource exports above the world average, and growth rates of remain important for many countries and re- exports from low- and middle-income coun- gions, particularly in Africa. However, the tries greatly outstripping rates from the high- broad-based nature of the shift into manufac- income countries. Exports of automobiles and turing exports means that developing-country components from low- and middle-income policymakers and others concerned about de- countries grew particularly rapidly, at more 66 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T Figure 2.2 Manufactures account for a growing share of exports in all regions a. Manufactures now make up almost 90 percent of b. The same is true of the developing countries of Europe exports from East Asian developing countries and Central Asia Share of exports by sector, East Asia and Pacific, 1981–2001 Share of exports by sector, Europe and Central Asia, (percent) 1981–2001 (percent) 100 100 Manufacturing exports 90 90 Manufacturing exports ᮡ 80 80 ᮡ 70 70 60 60 50 50 40 Resources exports 40 Resources exports 30 ᮡ 30 ᮡ Agricultural exports 20 Agricultural exports 20 ᮡ ᮡ 10 10 0 0 81 83 85 87 89 91 93 95 97 99 01 81 83 85 87 89 91 93 95 97 99 01 19 19 19 19 19 19 19 19 19 19 20 19 19 19 19 19 19 19 19 19 19 20 c. The share of manufactures in exports from Latin d. Manufactures grew from insignificance in exports from America and the Caribbean tripled in the last two decades the Middle East and North Africa Share of exports by sector, Latin American and the Share of exports by sector, Middle East and North Africa, Caribbean, 1981–2001 (percent) 1981–2001 (percent) 100 100 90 90 Resources exports 80 80 ᮡ 70 70 Manufacturing exports Resources exports 60 60 ᮡ ᮡ 50 50 40 Agricultural exports 40 Manufacturing exports 30 ᮡ 30 ᮡ 20 20 Agricultural exports 10 10 ᮡ 0 0 81 83 85 87 89 91 93 95 97 99 01 81 83 85 87 89 91 93 95 97 99 01 19 19 19 19 19 19 19 19 19 19 20 19 19 19 19 19 19 19 19 19 19 20 e. Manufactures grew to almost 80 percent of exports f. The share of manufactures in exports from Sub-Saharan from South Asia Africa nearly tripled, but from a low baseline Share of exports by sector, South Asia, 1981–2001 Share of exports by sector, Sub-Saharan Africa, 1981–2001 (percent) (percent) 100 100 Manufacturing exports 90 90 ᮡ 80 80 70 70 Resources exports 60 60 ᮡ 50 50 Agricultural exports Agricultural exports 40 40 30 ᮡ 30 ᮡ 20 20 ᮡ ᮡ 10 10 Manufacturing exports Resources exports 0 0 81 83 85 87 89 91 93 95 97 99 01 81 83 5 87 89 91 93 95 97 99 01 8 19 19 19 19 19 19 19 19 19 19 20 19 19 19 19 19 19 19 19 19 19 20 Source: COMTRADE. 67 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Table 2.1 Developing countries are becoming exporters of high-value products Annual growth rates (percent) Low income, less China Low China and Middle High and India income India income income World Primary products 1 2 5 1 4 2 Resource-based manufactures Agricultural 7 8 12 6 6 6 Other 4 7 10 5 5 5 Low-technology manufactures Textiles 14 15 15 7 5 8 Other 16 19 20 10 6 8 Medium-technology manufactures Automotive and components 22 20 19 19 7 8 Process industry products 14 13 12 11 6 7 Engineering products 21 23 24 12 7 8 High-technology manufactures Electronic 21 26 36 17 10 13 Other 10 16 20 12 9 9 Total 13 15 17 10 6 7 Note: Table 1 presents the annual growth rates by product group and by country groups assigned on the basis of 1981 income levels to avoid the selection bias that results when end-of-period attributes are used as the basis for selection. Product definitions are supplied by the WTO. Data analysis undertaken in World Integrated Trade Solutions (WITS) using “mirror” data from UN COMTRADE. Country groups defined by income status in 1981.While the results from this approach must be treated with some caution, because the level of technology of the process involved is frequently more important than the technology level of the product, examining the nature of the products being traded is clearly of interest. Source: COMTRADE, WITS, WTO. than 22 percent per year. Exports of engineer- between 1981 and 2001 (figure 2.3). In 1981, ing products such as engines, pumps, and these countries depended on resource-based instruments from low-income countries grew products for 87 percent of their exports, a at close to 21 percent per year. The highest share that had fallen to 25 percent by 2001. growth rates of all occurred in high-technology The share of low-technology manufactures products—particularly electronic goods, such rose substantially, from 13 to 38 percent, as computers, televisions, and components. while that of medium-technology exports World trade in these goods grew more than went from 3 to 15 percent. High-technology twice as fast as overall world trade. Export exports exploded from 2 to 21 percent. growth from low- and middle-income coun- The middle-income countries in 1981 were tries was much faster again, with exports of originally much more dependent than other electronic products from low-income countries countries on resource-based products—an im- growing at the extraordinary rate of more than portant contributor to their economic success 21 percent per year—enough to expand ex- up to that point (figure 2.3b). In 1981, re- ports almost 50-fold over the period. source-based products accounted for 81 per- cent of their exports, a share that fell to a still- Low-income countries are less reliant than substantial 39 percent in 2001. The share of before on resource-based exports low-technology manufactures rose from 9 to The importance of the growth rates noted 18 percent in the same countries, while the above depends greatly on the share of each share of medium-technology exports more broad product type—resource-based, low- than tripled from 6 to 27 percent, and high- technology, medium-technology, and high- technology manufactures jumped from 3 to 24 technology—in total exports. percent. Low-income countries showed the most The transformation of high-income coun- dramatic transformation of export patterns tries’ exports was much less dramatic than for 68 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T Box 2.1 Poor export performance in 43 countries N ot all countries participated in the otherwise positive trends for developing countries. Forty- nine countries experienced negative real growth rates modities. The oil boom led to a significant increase in public spending and a top-heavy civil service, which makes it difficult to respond swiftly to de- over the 20-year period for merchandise exports. creases in the price of oil. To make matters worse, Six of the 49 were tourism-based economies that the lack of a clear agricultural policy transformed the did poorly in merchandise trade but in fact experi- country into a net importer of food, accelerating the enced rising national incomes associated with already deteriorating trade balance. tourism exports. Eighteen countries experienced severe conflicts Of the 43 export-contracting countries, poor or war—among them Comoros, Rwanda, and Timor performance was attributable to combinations of Leste. Another five countries, including Libya and excessive dependence on one or two primary prod- Sudan, experienced trade embargoes. ucts, civil conflict, and politically motivated trade A more felicitous tale is that of Barbados, one of embargoes—often complicated by inept governance. the tourism-based economies. In 1981 the country In 1981, these countries derived an average of 85 was highly dependent on sugar. But progressive and percent of their export earnings from primary prod- stable political leadership, investments in education, ucts; 20 years later the average was 75 percent. Of and public investments in infrastructure to support the 43 countries, 20 were less-developed countries. tourism diversified and transformed the economy. Twenty cases were heavily dependent on one or Barbados once had the same per capita income as two primary products, such as oil (Cameroon), phos- Jamaica; today it is one of the most prosperous phates (Nauru), or copper (Zambia), and failed to countries in the Caribbean, with a per capita GDP diversify over the next two decades. Cameroon, for of $9,700 in 2000. example, despite its richness in natural resources, relies on oil for about one-third of export revenues, and timber or cocoa for much of the rest, leaving it vulnerable to fluctuation in the prices of these com- Source: World Bank staff. the developing countries. The share of resource- the countries best suited to it. Labor-intensive based exports fell from 37 to 5 percent, while stages of production, for example, are typi- the share of low-technology exports remained cally done in labor-abundant countries. Poten- close to 13 percent (figure 2.3c). The export tially, production sharing can greatly expand share of medium-technology manufactures the range of activities in which developing rose from 36 to 38 percent, while that of high- countries can participate—holding out the technology exports increased sharply—from promise of increasing employment and reduc- 13 to 24 percent of total exports. ing poverty. Of course, breaking the once-rigid linkages Global production sharing is creating among stages in the production process makes new opportunities it more difficult to interpret the implications of Much of the change in developing-country the shift to manufactures—particularly high- export patterns, and particularly the rise in technology products. In many cases, developing high-technology exports, is associated with countries undertake only those production ac- the phenomenon of global production sharing tivities that require low-skilled labor—a low- (Deardorff 2001; Hummels, Ishii, and Yi tech part of the production of high-tech com- 2001). Production sharing benefits rich and modities. However, the buoyant demand for poor countries by allowing production to be such commodities helps offset the relatively broken into discrete stages, each performed in stagnant demand for some traditional agricul- 69 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 tural commodities and can create important Figure 2.3 Technology-laden manufactures productivity gains through learning-by-doing have increased as a share of exports from and the expansion of productive firms. each group of countries, while the share of The move to global production sharing resource-based exports has diminished heightens the importance of timely, efficient, a. Low-income countries are moving out of resource- and low-cost transportation. Even quite small based industries into low-technology exports differences in transport costs and the timeliness Share of exports by sector, low-income countries, 1981–2001 (percent) of transportation services can have quite dra- 100 matic consequences for national incomes. 1981 90 Hummels (2001) estimates that an increase of ᮡ 80 one day in the time taken to deliver a good is 70 equivalent to an increase of 0.8 percent in the 60 50 cost, not just of transportation, but of the good 40 2001 itself. Redding and Venables (2001) conclude 30 that differences in transport costs in a world of ᮡ 20 global production sharing may account for a 10 large proportion of the observed differences in 0 Resource- Low tech Medium High tech incomes among countries. In this mode of pro- based tech duction, countries must pay transport costs to b. Middle-income countries are increasing the level of obtain their inputs and to consign their out- technology in their exports puts. If value added is a small share of output Share of exports by sector, middle-income countries, 1981–2001 (percent) value, as is frequently the case, then transport 100 costs have enormous leverage on the residual 90 returns available to pay workers and owners 80 of capital. If value added is 20 percent of the 70 60 gross output value in the absence of trade 50 costs, for example, then a transport cost of 10 40 percent of output to ship products out, and an 30 equal cost to bring in components, could wipe 20 out returns to productive factors. 10 To gain an idea of the potential importance 0 Resource- based Low tech Medium tech High tech of global production sharing in developing countries, we have calculated indexes of verti- c. In high-income countries, the share of high- technology exports has risen rapidly cal specialization of the type developed by Share of exports by sector, high-income countries, Hummels, Ishii, and Yi (2001) for several de- 1981–2001 (percent) veloping countries. These indexes expose the 100 share of imported inputs embodied in each unit 90 of goods exported—either directly or after indi- 80 70 rect use of imported inputs is taken into ac- 60 count. Although imperfect—they do not allow 50 for differences between export- and domesti- 40 cally oriented sectors in their use of inter- 30 mediate inputs—these measures provide a 20 structured assessment of the extent and changes 10 0 in production sharing.3 Two sets of results are Resource- Low tech Medium High tech based tech presented in figure 2.4. The lower bars estimate Source: COMTRADE. the share of export value accounted for by di- rect use of imported intermediates, whereas the 70 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T Figure 2.4 Global production sharing is increasingly important for China and India Share of imported inputs in a unit of exports, India, 1980–1998 (percent) 25 20 Direct and ᮡ indirect use Direct and Direct use (India) indirect use (India) Direct use (China) 15 (China) ᮡ 10 ᮡ ᮡ 5 0 1980 1985 1990 1995 1998 Source: World Bank staff. higher bars represent direct plus indirect use of index rose from 6.4 to 7.9 percent—not much imported intermediates. more than a tenth of Singapore’s. Although The importance of global production shar- Colombia’s larger economy would be expected ing in India has more than doubled since to show less vertical specialization than Singa- 1980. In China, even though production shar- pore’s, the fact that it is so much less integrated ing began from a considerably higher level than China’s or India’s suggests that con- than in India, it almost doubled over the pe- straints on transport and communications may riod, to 22 percent. Even so, the estimates un- be inhibiting Colombia’s participation in derstate the importance of global production global production sharing.4 sharing in China, where policy has strongly fa- China and India have tightened their inte- vored the use of imported inputs in labor- gration with the world economy since 1980. intensive production of manufactures (Iancho- Their experience suggests that successful ex- vichina 2003), and where exports based on porters of manufactures can avoid the prob- the processing of imported intermediates ac- lems of declining terms of trade that preoccu- count for about half of total exports. How- pied many thinkers in the 1950s and 1960s ever, the graph highlights the substantial in- (Bloch and Sapsford 2000) and that remain im- crease in the importance of the phenomenon plicit in many current models of world trade in China over the period—particularly since and growth. A striking feature of the expansion 1987, when duty-free access was extended to of exports from developing countries is that the a wide range of imported intermediates used terms of trade of countries whose exports have in the production of exports. risen extremely rapidly have not deteriorated to To take several other examples, Singapore’s the extent that one might predict using conven- economy is much more integrated into the tional economic models. Most of the models world economy than is middle-income Colom- used by economists would predict that large in- bia. Singapore’s total vertical specialization creases in exports would be followed by sub- index hovered around 60 percent of the value stantial declines in export prices, as countries of its exports over the past two decades, with exported more and more of the same products. a direct specialization index of more than 50 Declines in the terms of trade of China, percent. By contrast, in Colombia, the total India, and other high-export-growth countries, 71 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Figure 2.5 Soaring exports from China and India had only a moderate effect on China’s and India’s terms of trade a. China Terms of trade and exports of goods and services, 1979–2001 (billions of dollars) 140 350 120 Terms of trade [left scale] 300 ᮡ 100 250 80 200 Exports of goods and services [right scale] 60 150 ᮡ 40 100 20 50 0 0 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 b. India Terms of trade and exports of goods and services, 1979–2001 (billions of dollars) 160 80 140 Terms of trade [left scale] 70 ᮡ 120 60 100 50 Exports of goods and services [right scale] 40 80 ᮡ 60 30 40 20 20 10 0 0 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 Source: World Bank staff. however, have been much more modest than considerably greater had the terms of trade would be expected given their high rates of ex- not deteriorated in this way, but the gains in port expansion. China’s export revenues grew growth of export value, and its purchasing almost thirty-fold (3,000 percent) in value power, were clearly enormous. While reaping terms between 1979 and 2001 (figure 2.5). immense benefits from its burgeoning export Over the same period, the ratio of China’s trade, China essentially shared some of those export prices to import prices—its terms of benefits with its trading partners in the form of trade—declined by nearly 30 percent. Clearly, improvements in their terms of trade. India’s China’s gains in export value would have been exports grew sevenfold during the same period, 72 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T while its terms of trade deteriorated by perhaps ments in competitiveness resulting from re- 4 percent. forms in policies, or on growth in investment A key to the apparent discrepancy between and productivity? the predictions of economic models and actual Differences between export growth in a outcomes is that export growth in some devel- given region and average world growth rates oping countries appears to have been accom- can be ascribed to two key factors: (1) growth panied by vigorous expansion in the range in world demand for the region’s products and of products exported and in the markets in (2) increases in competitiveness because of which those exports were sold (Hummels and lower output prices, improvements in quality, Klenow 2002; Kehoe and Ruhl 2002; Evenett or shifts in the pattern of exports to products and Venables 2003), and by increases in the in greater demand. quality of the goods exported (Schott 2002). From 1981 to 1991, the developing coun- These important developments mean that pol- tries of East Asia experienced export growth icymakers in developing countries can worry of 232 percent, compared with the global aver- much less about declining terms of trade if age growth rate of 115 percent. The demand for they can focus on reform of policies—both at the products exported from East Asia grew the border and behind it—and on competing slightly faster than the world average, at 124 successfully in new products and markets. percent, but East Asian export performance was Clearly, the dramatic changes seen in devel- outstanding primarily because of an increase oping countries’ export patterns can be ex- in competitiveness that raised East Asia’s ex- pected to have a major impact on their inter- ports by 109 percentage points relative to over- ests in the current WTO negotiations. In the all market growth. Europe and Central Asia lost early 1980s, before the Uruguay Round, de- competitiveness over the same period, causing veloping countries relied heavily on exports of their exports to grow by 94 percent relative to resource-based products and had relatively growth in the market for their exports of 124 limited interest exports of manufactured prod- percent and to world export growth of 115 per- ucts, which until then had been the focus of cent. The commodity-dependent exporters of WTO negotiations on market access. Since the Middle East and North Africa suffered that time, however, the interests of the middle- heavily from contractions in the demand for income countries and the many poor countries the products they produced; the world market that now export high-technology products for their exports shrank by 21 percent. In addi- have broadened dramatically. Further, given tion, they lost competitiveness within their own the dramatic increase in vertical specialization product markets, with the result that their ex- in production, all countries are much more de- ports fell by 24 percent over the period. By con- pendent on the availability of the services trast, the product mix of South Asian exporters needed to support decentralized production— was helpful; the markets for their products grew giving developing countries a much greater by 129 percent over the period. South Asian stake in negotiations under the General Agree- countries also experienced substantial improve- ment on Trade in Services (GATS). ments in competitiveness, which accounted for an additional 70 percentage points of export growth, bringing their total export growth to Behind the patterns: Economic just under 200 percent. The market for the and policy determinants products exported by Latin America and the W hat caused the transformation in world trade patterns described in the previous section? Did exports grow “passively” in re- Caribbean expanded by 54 percent—less than half the average for the world as a whole—but these countries managed to gain an additional sponse to expansion in world markets? Or 21 percent increase in their exports through in- did the observed growth depend on improve- creases in competitiveness. 73 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Table 2.2 Developing countries’ exports became more competitive in the 1990s Source of export growth relative to world average growth, 1981–2001 (percent) 1981–91 1991–2001 Total Demand Competitiveness Total Demand Competitiveness Industrial countries 133 148 –16 48 70 –22 Europe and Central Asia 94 122 –28 255 48 206 East Asia and Pacific 232 124 109 139 75 64 Latin America and Caribbean 75 54 21 137 50 87 Middle East and North Africa –24 –21 –3 60 58 3 South Asia 199 129 70 113 36 77 Sub-Saharan Africa 10 20 –10 68 35 33 World 115 — — 68 — — Source: COMTRADE. For 1991 to 2001, export growth in all de- than closed economies, where growth in any veloping country regions outstripped that of factor can be expected to depress its price, as the industrial countries (table 2.2). While high the domestic demand for the goods in which in East Asia and the Pacific, it was even higher it is used intensively becomes saturated. Of in the developing countries of Europe and Cen- course, world markets, too, are finite, and tral Asia, where market share responded to rapid increases in supply can lead to declines a dramatic improvement in competitiveness.5 in world prices—as appears to have occurred Of the developing country regions, only East in coffee markets in recent years. But world Asia and the Pacific benefited from above- markets are much larger than those of indi- average growth in demand. But all regions ex- vidual countries. The problem of saturation is cept the Middle East and North Africa grew at much less likely to become serious for trade or above world average growth rates through in manufactures, because there is much more increases in competitiveness. two-way trade among developing countries in these goods. For this reason, Martin (1993) What explains improvements found that each developing country was likely in competitiveness? to be better off if all developing countries ben- Changes in production factors used by devel- efited from increases in manufacturing pro- oping countries probably improved their com- ductivity than if it alone benefited. petitiveness. One of the most misunderstood Other likely influences on the structure of predictions of economics is that changes in outputs and exports include changes in trade the factors employed in open economies will and investment policies; changes in the market change the mix of goods produced and ex- opportunities facing developing countries; and ported, rather than the prices of the input fac- the development of new market opportuni- tors. Increases in the amount of capital per ties in which developing countries already worker in an open economy can, for instance, have, or can develop, a comparative advantage. be expected to increase the share of output Clearly, these influences are related—increases from capital-intensive sectors, rather than de- in market opportunities and improvements in press the return on capital. Similarly, increases trade and investment policies are likely to stim- in the amount of education per worker can be ulate investment in physical and human capital. expected to increase the share of output from Increases in the importance of foreign di- knowledge-intensive activities, rather than de- rect investment are another contributing fac- clines in returns to education and increases in tor to the changes in developing countries’ unemployment of skilled workers. In this re- participation in international trade. As docu- spect, open economies are much better placed mented in World Bank (2002), foreign direct 74 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T Table 2.3 Investment in people and in capital grew rapidly Percent annual changes in factor endowments, 1960–90 Secondary Tertiary Capital Education education education per worker per worker per worker per worker Industrial 3.7 0.3 2.2 4.9 Developing East Asia and Pacific 5.1 4.2 9.2 3.4 Latin America and Caribbean 2.4 2.0 5.3 6.7 Middle East and North Africa 3.4 2.3 1.9 6.3 South Asia 3.2 3.3 4.3 6.4 Sub-Saharan Africa 2.1 4.2 9.7 12.6 Source: Nehru and Dhareshwar (1993); Nehru, Swanson, and Dubey (1995). investment grew dramatically during the 1990s. sures the growth of capital per worker, while Not only did it bring capital to developing subsequent columns measure years of educa- countries, augmenting the total supply of cap- tion and average years of secondary and ter- ital per worker, but it brought know-how, and tiary education per worker. While these are connections with other elements in the net- only crude measures of the growth of these in- work of global production sharing. puts per worker, they do represent an indica- One likely contributor to the observed tion of the efforts that have been made in de- change in the mix of developing-country ex- veloping countries to increase the capital and ports is the rising amount of capital per worker skills available per worker. available in some developing economies. In The relationship between accumulation of East Asian economies, the annual growth rates factors of production and the export mix is of capital per worker have been almost one likely to be quite complex, with countries first and a half times those in the advanced indus- expanding their output of labor-intensive man- trial countries (Nehru and Dhareshwar 1993; ufactures and then, beyond a certain level of Nehru, Swanson and Dubey 1995). In other capital and skills, moving into a different range regions, the average rate of growth in capital of products (Schott 2003). Further, questions per worker has been lower than in the indus- have arisen regarding how effectively many trial countries, even though some developing countries have been able to use the additional countries outside East Asia have rates of saving capital and human skills (Pritchett 2000, and investment that match those found in East 2001). It seems highly likely, however, that the Asia. Increases in the amount of secondary and observed rapid increases in capital and skills tertiary education per worker have been much per worker have been important in many cases higher for most developing country regions of successful development and that they are than in the industrialized world—albeit fre- vital to long-term progress. Without large in- quently from a low level. creases in the availability of skilled labor, it To the extent that these resources have been would be difficult to explain the rapid in- effectively employed, this deepening of finan- creases in the exports of high-technology prod- cial and human capital per worker can be ex- ucts from developing countries—especially pected to encourage a shift away from labor- from the low-income countries. Even where intensive activities toward activities that use high-technology exports involve routine opera- more capital and skills. Broad estimates of the tions performed on sophisticated imported in- growth rates of financial and educational cap- puts, advanced organizational and technical ital are presented in table 2.3 for each devel- skills are needed to ensure consistent and reli- oping country region. The first column mea- able supplies of high-quality exports. 75 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Lowering protection throughout the provement in market access by weighting the developing world created new price change due to tariff cuts in industrial opportunities— countries by two-thirds and the reduction in Since the mid-1980s, the large-scale liberaliza- developing countries by one-third, yielding an tion of trade policies in developing countries average price reduction for developing country has widened market access and lowered the exports of 6.4 percent, almost exactly half the implicit taxation on exports that import tariffs stimulus that comes from developing countries’ entail. Average tariffs in developing countries own exports. This suggests that, in aggregate, fell to around 12 percent by 2000—about developing countries’ own liberalization has one-third of their level in 1983. This large re- been the primary channel through which trade duction was accompanied by even larger re- reform has expanded developing countries’ ductions in nontariff barriers and exchange- export growth. Because reform in any one de- rate overvaluation—both of which strongly veloping country benefits other developing exacerbated the protectionist effects of tariffs countries as well, the total contribution of de- in the 1980s (World Bank 2000). veloping country reform can be captured by Absolute reductions in protection were even combining the “own-liberalization” effect with larger in individual countries. India, for exam- the market-access benefits provided by other ple, reduced its average tariff from 100 percent developing countries. When we do this, we find in 1986 to 32 percent in 1999. While some re- that 88 percent of the stimulus to developing- ductions in protection have occurred in indus- country exports following tariff liberalization trial countries—through tariff reductions and derives from developing-country liberalization. through abolition of nontariff barriers—the Such large reductions in protection can be changes have been quite small relative to those expected to have marked effects on the pattern in developing countries. Between 1980 and of exports, as well as their level. Protection 2001, the average tariff in industrial countries raises the costs of all domestic industries by in- fell from 9.8 to 3.7 percent—a significant fall, creasing the costs of their inputs—including but much smaller than that observed in devel- both intermediate inputs and factors.6 How- oping countries, where the average tariff fell ever, this effect varies among sectors. Typically, from 30 percent to 12.7 percent over the same manufactures are much more vulnerable to period. the adverse effects of protection because they These figures, and some standard assump- are more dependent than agricultural and tions, allow us to divide up the contribution of resource-based activities on imported inter- trade reform to developing country export mediate inputs. Further, this vulnerability has growth into a component due to countries’ grown over time as production has moved own liberalization and one due to improved from regionally integrated production—the market access and export demand. The tariff original approach taken by firms such as Ford reductions in developing countries reduced the and the large integrated steel mills in an earlier price of imports to domestic consumers by an era of industrialization—to internationally in- average of 12 percent, while import prices in tegrated production networks involving many the industrial countries were reduced by 3.4 firms and countries. percent. The increase in the demand for ex- Protection regimes are often erected to pro- ports from developing countries is determined mote industrialization without thorough con- by the reductions in import prices in their mar- sideration of their impact on the production of kets—both in industrial countries and in other manufactured goods and the structure of ex- developing countries. Over the period from ports. Tariffs and other trade barriers affect 1986 to 2001, the industrial countries ab- exports primarily by raising the costs of pro- sorbed two-thirds, on average, of developing- duction inputs. Because protection policies country exports. Therefore, we estimate the im- rarely improve the returns small developing 76 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T countries can obtain from sales of their ex- ing and in capital- or labor-intensive manufac- ports, their impact on exports can be judged tures were taxed at essentially prohibitive by considering their effect on the costs of in- levels. In India, the taxes directly imposed termediate inputs—and hence on the returns by protection on agricultural processing and available for payment to factors. This can be capital-intensive manufacturing averaged more done simply by applying the concept of the ef- than 60 percent. (Nontariff measures, domes- fective rate of protection to measure the effect tic licensing requirements, and exchange-rate of protection on the value added in export distortions, if computed, would have further production. While this approach underesti- increased the effective tax.) In Brazil, the esti- mates the adverse effects of protection by ig- mated impact of tariffs on returns from ex- noring indirect cost-increasing effects, it pro- porting manufactured products and processed vides a simple and transparent indication of agricultural goods was even more sharply neg- the direct effects. ative—around 70 percent. In China, the direct impacts of protection appear to have been on —by reducing the implicit taxes the same order of magnitude, with agricultural on exports processing facing taxes of more than 70 per- The burden on exports of tariffs on intermedi- cent and labor-intensive manufactures close to ate input costs7 is illustrated by the cases of 60 percent. These problems were compounded Brazil, China, India, and Malawi in 1986, by strong policy-driven obstacles to the expan- when estimated rates of average protection sion of state-run firms, which eventually were were first available for each country, and mitigated by the emergence of an entirely new 1997, following large reductions in protection class of firms—the township and village enter- (table 2.4). The impact of protection on ex- prises—not subject to the constraints of the ports differs considerably from country to state-run firms. The export tax rates in Malawi country, but two key features are evident. appear to have been much lower than in China First, agricultural processing and manufactur- and India, even before the reforms, perhaps be- ing (whether labor or capital intensive) for ex- cause such a small and trade-dependent econ- port are much more heavily taxed than are omy simply could not maintain the types of agricultural and resource commodities. Sec- trade barriers found in the bigger countries. ond, the rate of taxation has generally de- Although a very few agricultural processing clined substantially since the mid-1980s, while and manufacturing activities that depended remaining substantial for industrial products. less on intermediate inputs might have been At the levels of protection prevailing in able to survive at average tariff rates of 100 1986, export activities in agricultural process- percent (as in India), it seems highly likely that Table 2.4 Tariffs hurt exports—but less so in the 1990s than in the 1980s Cost penalties on exports associated with import tariffs (percent) Brazil China India Malawi 1986 1997 1986 1997 1986 1997 1986 1997 Agriculture –43 –5 –28 –15 –14 –5 –9 –7 Agricultural processing –83 –28 –72 –54 –64 –39 –20 –16 Resources –45 –6 –14 –7 –9 –3 –6 –5 Labor-intensive manufacturing –72 –17 –54 –35 –45 –23 –18 –15 Capital-intensive manufacturing –79 –22 –46 –28 –60 –35 –11 –9 Services –31 –3 –26 –14 –16 –6 –5 –4 Note: Effective rate of protection applying to exporters is the proportional change in returns to value-adding factors resulting from tariff protection. Source: World Bank data. 77 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 reductions in tariffs—and nontariff barriers— duces the pressure for more general reductions of the type observed around the world between in protection, since exporters—a potentially 1986 and 2001 (World Bank 2000) must have powerful source of pressure for reductions in contributed to the great expansion of develop- tariffs—no longer suffer the direct impact of ing countries’ manufacturing exports. protection (Cadot, de Melo, and Olarreaga Reductions in average tariffs were comple- 2002). mented by the introduction of duty-exemption Redressment of behind-the-border costs im- or drawback arrangements under which export posed by inadequate infrastructure and ex- producers obtained access to duty-free inputs cessive, inappropriate regulation has also for use in export production. These arrange- helped developing-country exports (Dollar, ments offer one way, legal under GATT, to re- Hallward-Driemeier, and Mengistae 2003). duce the burdens imposed by import duties. Other behind-the-border problems include Some exporters, such as China, have used them those associated with clearance through cus- successfully to develop labor-intensive exports toms—excessive or arbitrary inspections or re- (Ianchovichina 2003; Ianchovichina and Mar- quests for documentation, demands for bribes tin 2003). or other informal payments, and so on. Sev- However, such policies are an imperfect so- eral of these problems are dealt with in greater lution to the problems created by protection. detail in chapters 5 and 6. Whether introduced throughout the economy The dramatic changes in the nature of their or in specific free-trade zones, such arrange- participation in world trade have greatly ments are administratively demanding. In many changed the incentives of developing countries cases, particularly in Africa, they have failed to participate in the world trading system. to operate successfully (Madani 1999). Further, When developing countries exported goods— they tend to encourage firms to concentrate on cocoa, rubber, coffee—that did not compete di- production activities that add a small amount of rectly with those produced in developed coun- value to imported inputs, rather than on activi- tries, they had little incentive to participate in ties more closely integrated with domestic pro- politically difficult exchanges of market-access duction. Ianchovichina (2003) found that ex- concessions that characterized the multilateral porting activities had become much more trading system. At the same time, the effects on import-intensive than other industries as a result exports of their protection regimes were rela- of the incentives created by duty exemptions. tively subdued, since their primary exports—as Since one of the key lessons of the new eco- we have seen—required fewer intermediate in- nomic geography is that there may be substan- puts. The shift to manufactures increases the tial gains from activities that encourage importance of access to markets in which there the development of backward—as well as for- is likely to be strong domestic competition. ward—linkages (Amiti 2003), incentives to- And the prices of export-oriented manufac- ward shallow processing activities may cause tures are, of course, very sensitive to the costs highly protected economies to miss many op- of intermediate inputs, since exporters are un- portunities for growth. Reductions in overall able to pass these costs on without pricing tariffs are a much better alternative than duty themselves out of the market. exemption. Not only do they remove the incen- tive for unnecessarily shallow specialization, but they also reduce the price of nontraded goods Market access for development: and factor inputs (Corden 1997), and further The agenda increase the stimulus to production for export. Another problem with relying on duty ex- emptions rather than relatively low and uni- R eciprocal exchanges of tariff reductions, the key element of all previous WTO ne- gotiations, will be a critical element in the cur- form tariff rates is that their introduction re- rent negotiations. Tariffs, however, are not the 78 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T only issue. Two additional topics central to Korea, for which clothing and textiles are market access for developing countries are: industries that likely will be allowed to decline and “sunset” • The phasing out of textile and clothing • Exporters such as Mauritius and Cambo- quotas, and dia, whose exports have been less tightly • Frequent recourse to antidumping mea- restricted by quotas and which have had sures. preferential tariff access for at least some commodities Phasing out quotas on textiles and • Countries such as Mexico and Turkey clothing is crucial that face neither quota constraints nor The commitment to phase out quotas on tex- tariff barriers in their major markets. tiles and clothing was made in 1994 as part of the Uruguay Round agreement. That commit- Abolition of quotas will remove much of ment took the form of an Agreement on Tex- the incentive for continued production in the tiles and Clothing, under which quotas were first group of exporting countries and reduce to be phased out in three tranches. Products the margin of preference enjoyed by the free- accounting for 16 percent of 1990 imports access countries. (Their preference will drop were to return to GATT disciplines immedi- from the margin provided by tariffs plus the ately, with an additional 17 percent returning in export-tax equivalent of other countries’ quo- 1998, and 18 percent in 2002. However, be- tas, to just the margin provided by tariffs.) cause the imports used as the baseline included Results provided by simulation models sug- products typically traded only by the industri- gest that countries such as China and India, alized countries, importing countries were able which have relatively low production costs, to meet their commitments without abolishing are likely to make substantial gains in market any significant quotas until the third phase of share following abolition of the quotas (Yang, integration, beginning January 1, 2002. The Martin, and Yanagishima 1997; François and delay in the abolition of quotas has meant that Spinanger 2002). These results are condi- perhaps 85 percent of the effective quotas tioned on the assumptions of the models, and against developing countries remain in effect— particularly on differentials in the extent to including the most restrictive. Unless the indus- which the quotas restrict the exports of differ- trial countries go back on their solemn com- ent countries. While some countries, such as mitments, often reaffirmed, all of the remaining China, provide high-quality data on the extent quotas will be abolished on January 1, 2005. to which the quotas restrict their exports, data It is difficult to predict the impact of quota for many other exporters are much less widely abolition, since the textile industry is so heavily available. Another indicator of the underlying distorted by quota and tariff protection in both competitiveness of individual exporters is the industrial and developing countries. What is share of their exports shipped to nonquota clear is that some countries, such as China and markets. The more efficient the supplier and India, with strong underlying comparative ad- the more restrictive the quotas it faces, the vantage in the production of these goods, have more of its exports it will tend to ship to less had their exports sharply restricted by the pres- lucrative nonquota markets (figure 2.6). ence of the quotas. Other suppliers, much less While the share of clothing exports (using severely restricted by quotas and/or tariffs, have WTO categories) exported to nonquota mar- been able to expand their exports considerably. kets is a crude index of the extent to which ex- This group includes three distinct categories: ports are restricted by quotas, some interesting patterns appear. The first is that some coun- • Exporters such as Hong Kong, China; tries—Albania, Costa Rica, Mexico, Morocco, Taiwan, China; and the Republic of Pakistan, and Tunisia—directed almost all of 79 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Figure 2.6 Many developing countries face an adjustment when quotas are lifted Share of clothing exports to nonquota markets by developing-country exporters, 2001 (percent) Albania Morocco Mauritius Tunisia Costa Rica Mexico Pakistan India Czech Republic Malaysia Taiwan, China South Africa Thailand Indonesia Hong Kong, China Korea, Republic of Colombia China 0 10 20 30 40 50 60 70 80 90 Source: COMTRADE. their clothing exports to the quota-restricted these quotas has been rapid growth in exports countries, suggesting that their quotas were of these products, particularly from China. large enough to let them focus on these mar- Whether current supplying countries main- kets, or that their competitiveness did not allow tain or lose market share following quota abo- them to export to less-lucrative nonquota mar- lition will depend on whether they undertake kets. Another group of countries, such as the reforms in advance to maintain their com- Czech Republic, India, Indonesia, and South petitiveness. The current system contains, for Africa, exported more than 10 percent of their many countries, disincentives for policy re- exports to nonquota markets—suggesting both forms that lower costs, improve efficiency, and restrictive quotas and an ability to compete at increase supply. Increases in the supply of ex- currently depressed world prices for clothing. ports from a country that has filled its quotas Finally, Colombia and China exported 50 and must be shipped to a limited range of markets 79 percent, respectively, of their exports to non- not constrained by quotas, where prices are quota markets. Countries in this category ap- likely to decline if significant additional quanti- pear likely to be highly restricted and to have ties are exported (Elbehri, Hertel, and Martin strong potential for expanding their exports 2003). Once the quotas are abolished in the following abolition of the quotas. world’s largest markets, however, the gains The abolition of some restrictive quotas in from reforms that reduce costs are likely to be January 2002 provides another source of in- much greater. If countries use the greater com- sight into the implications of quota abolition. petition that follows the abolition of quotas as Because the abolished quotas covered only a a stimulus for reforms that increase productiv- small fraction of total textile and clothing trade, ity, they stand to gain much more than they one would expect a disproportionate response could have hoped to gain in the past. Bhard- to their disappearance, since additional re- waj, Kathuria, and Martin (2001) point to sources waiting to be channeled into textiles areas in India, for example, where such reforms and clothing could, for the moment, be redi- are needed to allow the industry to become rected only into the products liberalized. In fact, more competitive. Needed reforms include: a key feature of the adjustment to abolition of eliminating policies that create disincentives for 80 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T Table 2.5 Quota abolition in China will manufactured goods, it is more likely to suffer move resources from other activities to from increased competition following aboli- textiles and clothing tion of the textile and clothing quotas than if Percent change in export volumes it had a more balanced export pattern. Thus, Sector Anticipated change policy should not only seek to improve pro- ductivity in the textiles and garment sector, Apparel 125.7 Automobiles and parts –22.8 but also to improve productivity in other sec- Cotton –8.6 tors, where competition may be less intense Electronics –10.6 following abolition of the quotas. Leather and shoes –5.0 Metal products –11.9 Textiles 41.9 How tariffs are reduced will affect the Other manufactures –14.1 development promise of Doha Source: Staff results from model of Ianchovichina and Martin Given the mercantilist nature of international (2001, 2003). trade negotiations, developing-country policy- makers contemplating the Doha Development Agenda will want to identify the export sec- factory production, eliminating reservation of tors in which they face the most significant particular activities for handloom production, trade barriers. The average tariff barriers fac- and improving duty exemption arrangements. ing exporters from each region are shown in The specific needs for policy reform will, of table 2.6. Because separate negotiations on course, vary across countries. market access are being conducted for agricul- How will the phase-out of quotas on tex- tural goods and nonagricultural goods, the tiles and clothing affect other sectors? By in- table is divided into two sections. ducing highly competitive producers to shift Tariffs imposed by the industrial countries from other activities into textiles and clothing, on imports from developing countries are typ- the abolition of quotas is likely to reduce sup- ically much higher than those they levy on plies of other goods, creating opportunities for other industrial countries. In agriculture, the other exporters. The likely response of Chi- industrial countries impose an average 15 per- nese industry, for example, to the abolition of cent tariff on imports from other industrial current quotas on textiles and clothing is a countries, whereas the rates on imports from shift in resources away from other goods developing countries range from 20 percent (table 2.5). The specific results presented in (Latin America) to 35 percent (Europe and table 2.5, produced using a model by Ian- Central Asia). Outside of agriculture, the dis- chovichina and Martin (2001, 2003), should crepancy is even more striking. Tariffs on im- not be seen as predictions of outcomes, how- ports from other industrial countries average 1 ever, since the phase-in of China’s liberaliza- percent, while those from developing countries tion commitments under its WTO accession, face tariff averages ranging from 2.1 percent and the continuing high rates of investment in (Latin America) to 8.1 percent (South Asia). physical and human capital in China, tend to The differences in tariff averages reflect in stimulate the output of many activities, in- part the presence of major trading blocs such cluding some of those mentioned in the table. as the European Union and the North Ameri- But the changes anticipated by the model can Free Trade Agreement (NAFTA), which do suggest the importance of examining the include key industrial-country trade partners. disincentives for production of goods other In part, also, they reflect differences in the pat- than textiles and clothing. If a country has, for tern of exports and the broad profile of tariffs. example, a duty-exemption arrangement cov- In the GATT trade rounds during which the ering the needs of the textile and clothing sec- greatest strides toward liberalization were tors and has not developed exports of other made (the Kennedy and Tokyo Rounds of the 81 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Table 2.6 Industrial countries levy higher tariffs on imports from developing countries than from other industrial countries—and some regions have high tariff walls Protection rates facing exporters in each region, 1997 (percent) Importing Region Europe and Sub- East Central Latin Middle South Saharan Exporting Region Asia Asia America East Asia Africa Industrial Agriculture East Asia 31.0 30.3 15.5 45.3 38.4 19.0 30.5 Europe and Central Asia 24.2 36.4 23.8 55.3 34.2 12.7 35.1 Latin America and Caribbean 42.1 36.0 14.8 50.3 29.7 24.7 20.4 Middle East 23.0 43.4 14.9 76.4 31.8 18.9 23.4 South Asia 16.6 34.6 13.7 41.1 27.7 11.0 25.8 Sub-Saharan Africa 26.7 20.3 14.4 39.1 30.9 33.6 23.6 Industrial 33.3 43.7 20.1 65.4 16.4 24.0 15.3 Nonagriculture East Asia 8.2 13.8 15.1 12.2 28.1 14.5 5.1 Europe and Central Asia 6.4 6.4 11.4 8.6 25.8 12.8 5.9 Latin America and Caribbean 4.3 6.7 15.4 8.9 19.4 11.9 2.1 Middle East 5.4 11.5 8.8 11.4 33.6 11.7 6.0 South Asia 7.1 11.0 13.6 10.2 19.0 17.4 8.1 Sub-Saharan Africa 4.4 6.1 11.7 6.1 27.6 20.6 4.2 Industrial 7.4 9.6 8.5 10.4 25.2 12.2 1.0 Source: Weighted averages calculated using GTAP Version 5 Database (www.gtap.org). Most-favored-nation rate except for major free-trade blocs such as the European Union and the North American Free Trade Agreement. Does not include other preference schemes. 1960s and 1970s), developing countries were other developing countries under the General- not active participants in the trading of recip- ized System of Preferences. rocal market-access concessions. Under the cir- Recent research suggests, however, that con- cumstances, it was more likely that their ditions such as rule-of-origin requirements re- products would be omitted from the sharp re- duce the benefits provided by such agreements ductions in tariffs made in those rounds.8 substantially below the gains implied by the A reasonable objection to this interpreta- nominal preferences (Brenton 2003). Further- tion is that some developing countries face more, many of these countries suffer from re- substantially lower tariff rates than are pre- strictions on their access to markets for a wide sented in table 2.6 and may even benefit from range of other products. And other countries access to industrial country markets at prices with a large fraction of the world’s poor receive above world market levels. This is true for no benefit at all from these preferences—in fact many countries and groups of countries. The they are harmed by diversion of their exports countries of the African, Caribbean, and Pa- to preference-receiving countries. China and cific group enjoy preferences on many of their India alone contain well over 500 million peo- exports to the European Union. The least- ple living on $1 per day or less (World Bank developed countries receive preferences under 2003). These countries receive only minimal the Union’s Everything But Arms Agreement. benefit from these preferential arrangements. Other countries receive preferences as mem- Developing countries tend to levy higher bers of Euro-Mediterranean agreements, the tariffs on imports from other developing coun- U.S.–Caribbean Basin Initiative, the U.S Africa tries than do the industrial countries (table Growth and Opportunity Act, and preferences 2.7). This is particularly striking in the case of provided to least developed countries and agricultural products, where the tariffs levied 82 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T Table 2.7 Developing countries pay large amounts in tariffs to their neighbors The share of the burden on each region’s exports imposed by region of destination (percentage of total barriers faced by exporting region) Importing Region Europe and Sub- East Central Latin Middle South Saharan Exporting Region Asia Asia America East Asia Africa Industrial Total Agriculture East Asia 32.8 0.7 0.6 5.6 4.7 2.1 53.6 100 Europe and Central Asia 1.7 19.8 0.6 13.9 0.4 1.4 62.3 100 Latin America (LAC) and Caribbean 13.7 1.6 10.8 14.8 1.4 1.9 55.6 100 Middle East and North Africa 2.1 3.0 0.6 44.4 1.6 1.0 47.3 100 South Asia 12.6 2.0 0.7 28.2 7.5 1.8 47.4 100 Sub-Saharan Africa 7.1 2.3 0.4 4.8 4.0 10.0 71.4 100 Industrial 16.1 3.2 5.0 19.1 0.6 2.7 53.3 100 Nonagriculture East Asia 37.4 1.5 6.3 4.8 5.9 4.2 39.9 100 Europe and Central Asia 2.8 13.5 2.1 5.8 2.0 2.9 71.0 100 Latin America and Caribbean 3.7 0.3 63.8 1.7 1.2 1.5 27.7 100 Middle East and North Africa 8.2 1.8 2.4 12.4 28.6 3.1 43.4 100 South Asia 8.8 0.6 3.1 9.7 6.8 7.9 63.0 100 Sub-Saharan Africa 9.0 0.7 4.2 1.7 7.7 40.4 36.3 100 Industrial 31.1 6.8 15.3 14.3 6.5 6.1 19.9 100 Source: Weighted averages calculated using GTAP Version 5 Database (www.gtap.org). Most-favored-nation rate except for major free-trade blocs such as the European Union and the North American Free Trade Agreement. on developing-country exports are frequently cases where trade flows are strongly inhibited twice as high as the already high rates levied by by high tariff rates or where much trade takes the industrial countries. For nonagricultural place at preferential rates, the measure does products, the differences are even greater in provide at least a crude adjustment for the rel- proportional terms. The tariffs imposed by ative importance of different markets.9 South Asia on imports from developing coun- In agriculture, all developing regions face tries, for instance, are frequently five times as their most significant barriers in the industrial high as the rates imposed by industrial coun- countries. Although the burden is greatest in tries. It is also notable that countries levy high Sub-Saharan Africa, where over 70 percent of tariffs on imports from other countries in their the barriers faced are imposed by the indus- region, particularly in Sub-Saharan Africa, trial countries, the industrial countries ac- where the average tariff levied is higher on count for more than 50 percent of the barriers African products than on imports from any facing all developing regions except South other region. Asia and the Middle East and North Africa. How important are the effects of the tariff Even there, industrial-country barriers are rates discussed above? The answer depends on substantial, accounting for almost half of the the size of the trade volumes to which they total direct burden imposed on their exports. apply. One way to get a rough indication of the In only one region, the Middle East and North importance of tariffs in particular markets is to Africa—where barriers imposed by other examine the value of the duty charged on ex- countries of the region loom particularly ports to that market. This measure effectively large—do the barriers erected by the develop- weights each tariff rate by the value of the trade ing world approach the effect of those im- to which it applies. Although clearly flawed in posed by developed countries. After the devel- 83 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 oped countries, however, the most important members to accommodate and at the same barriers are often found in neighboring coun- time isolate a powerful interest that might oth- tries, as in the case of East Asia, where neigh- erwise set back an entire liberalization pro- bors’ barriers account for almost a third of the gram.10 Since the Uruguay Round these trade barriers facing exporters, and the Middle East, rules—particularly antidumping—have been where they account for close to 50 percent. invoked with increasing frequency—particu- In nonagricultural trade, industrial-country larly by developing economies. Moreover, barriers are clearly much more important than WTO members increasingly treat their use as would be implied by the relatively low tariff a reserved right of unilateral protection simi- rates shown in table 2.7. Their importance to lar to the national-security exception, rather exporters, however, varies substantially by re- than as an instrument to manage an ongoing gion. For developing countries in Europe and and multilateral process of liberalization. (See Central Asia, they are particularly important, box 2.2.) accounting for over 70 percent of total levies Recent data on the incidence of use of an- on exports. Industrial-country barriers are tidumping rules by broad country groups re- also particularly important in South Asia, veal several ominous patterns. The first is a where they account for close to two-thirds of tendency for both developed and developing the tariff burden. By contrast, the industrial- countries to resort to antidumping measures— country share is closer to 40 percent in East 1,979 between 1995 and June 2002 (table Asia, the Middle East and North Africa, and 2.8). More antidumping actions were initiated Sub-Saharan Africa, and under 30 percent in by developing countries against other develop- Latin America and the Caribbean. For the in- ing countries than by or against industrial dustrial countries themselves, barriers in other countries. industrial countries are now relatively small, The emergence of developing countries as at under 20 percent of the total, a fact that un- major users of antidumping measures is a re- doubtedly contributes to developed countries’ cent phenomenon. The use of this form of pro- interest in securing trade liberalization in de- tection, which requires complex and expen- veloping countries. These generalizations not- sive administrative processes, has traditionally withstanding, the differences mask country- been eschewed by developing countries. How- level patterns, obliging policymakers in each ever, perhaps in part because WTO has be- country to do their own analysis. come less tolerant of some other avenues for introducing discretionary protection, such as Use of antidumping actions measures for balance-of-payment purposes, to generate protection has reached many developing countries have begun to an advanced stage make use of antidumping measures. In 1996, GATT and the WTO acknowledge the sover- 767 antidumping actions were pending, of eign rights of member countries to impose cer- which 581 had been introduced by industrial tain new trade restrictions or to replace old countries. By June 2002, the number of pend- ones. A number of these are “exceptions” to ing actions had grown to 1,189, of which 636 the general intention of providing an open in- had been initiated by industrial countries and ternational trading system, such as import re- 553 by developing and transition economies. strictions that relate to national security. Oth- All groups of countries show a striking ten- ers are part of the management of the trading dency to impose antidumping measures dis- system. These are usually described as being al- proportionately against the exports of devel- lowed under “GATT/WTO rules” rather than oping and transition economies. But when the as “exceptions.” actions shown in table 2.8 are divided by Within the GATT/WTO system the general the dollar value of imports from each group, justification for such rules is that they allow the result is that industrial countries impose 84 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T Box 2.2 Swimming upstream: The case of Vietnamese catfish On July 22, 2003, the New York Times wrote the ‘tra’. . . . Catfish Farmers of America, ran adver- following editorial that illustrates well the vicissi- tisements warning of a ‘slippery catfish wannabe,’ tudes of exporters in antidumping waters: saying such fish were ‘probably not even sporting real whiskers’ and ‘float around in Third World . . . After embracing decidedly un-Marxist re- rivers nibbling on who knows what.’ Not satisfied forms, Vietnam became one of globalization’s with its labeling triumph—an old trade-war trick brightest stories in the 1990’s. The nation, a one- perfected by the Europeans—the American group time rice importer, transformed itself into the initiated an antidumping case against Vietnamese world’s second largest rice exporter and a player catfish. And for the purposes of this proceeding, in the global coffee trade. The rural poverty rate Congressional taxonomy notwithstanding, the fish was slashed to 30 percent from 70 percent. The in question were once again regarded as catfish, normalization of ties between Hanoi and Washing- not basa or tra. . . . ton brought American trade missions bent on ex- panding Vietnamese free enterprise. One of these Antidumping cases involve allegations that im- delegations saw in the Mekong Delta’s catfish a ports are being sold more cheaply than they are golden export opportunity, with the region’s nat- back home or below cost, practices rightly banned ural conditions and cheap labor affording Vietnam by trade laws. . . . In this case, the Commerce De- a competitive advantage. Sure enough, within a partment had no evidence that the imported fish few years, an estimated half-million Vietnamese were being sold in America more cheaply than in were living off a catfish trade nurtured by private Vietnam, or below their cost of production. But entrepreneurs. Vietnam captured 20 percent of rather than abandoning the Mississippi catfish the frozen catfish-fillet market in the United farmers to the forces of open competition, the de- States, driving down prices. To the dismay of the partment simply declared Vietnam a “nonmarket” Mississippi Farm Bureau, even some restaurants economy. The designation allowed it simply to in that state—the center of the American catfish stipulate that there must be something suspect industry—were serving the Vietnamese species. going on somewhere—that Vietnamese farmers must not be covering all the costs they would in a Soon . . . Vietnamese farmers were caught in a functioning market economy. Tariffs ranging from nasty two-front war being waged by the Catfish 37 percent to 64 percent have been slapped by the Farmers of America, the trade group representing department on Vietnamese catfish. . . . Mississippi Delta catfish farmers. The Mississippi catfish farmers are generally not huge agribusi- Prices along the Mekong crashed, as the expor- nesses, and many of them struggle to make ends ters who buy his fish moved to protect their mar- meet. But that still does not explain how the gins. . . . Faced with the prospect of losing their United States, the international champion of free investment, [farmers] might be shocked to learn market competition, could decide to rig the cat- that [the] Commerce Department says they do fish game to cut out the very Vietnamese farmers not operate in a free market. . . whose enterprise it had originally encouraged. The United States International Trade Commission, Last year . . . the American catfish farmers man- an administrative agency in Washington, provided its aged to persuade Congress to overturn science. final verdict on July 23. The verdict stated that the An amendment, improbably attached to an ap- American catfish industry was hurt by unfair compe- propriations bill, declared that out of 2,000 cat- tition due to dumping by Vietnam—making the tar- fish types, only the American-born family— iffs permanent. named Ictaluridae—could be called ‘catfish.’ So the Vietnamese could market their fish in America only by using the Vietnamese terms “basa” and Source: New York Times, July 22, 2003. 85 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Table 2.8 Most antidumping actions are filed by developing countries against other developing countries Antidumping actions initiated between 1995 and June 2002 Initiated against Industrial Developing Transition All Initiated by economies economies economies economies Industrial economies a 198 494 127 819 Developing economiesb 357 649 138 1,144 Transition economies c 4 6 6 16 All economies 559 1,149 271 1,979 a. Australia, Canada, 15 European Union members, Iceland, Japan, New Zealand, Norway, Switzerland, and the United States. b. All other economies excluding industrial economies and transition economies. China is included in the totals for developing economies c. 27 transition economies, as defined by World Development Report 1996 (Albania, Armenia, Azerbaijan, Belarus, Bosnia- Herzegovina, Bulgaria, Croatia, Czech Republic, Estonia, Macedonia (FYR), Georgia, Hungary, Kazakhstan, Kyrgyzstan, Latvia, Lithuania, Moldova, Poland, Romania, Russian Federation). Source: WTO Antidumping Committee Reports. antidumping measures on developing coun- now average only 4 percent, their antidumping tries more than twice as frequently as the de- rates have been seven to ten times higher. This veloping countries’ volume of exports would is new protection, not a temporary return to imply. Against transition countries, the statis- rates of protection previously negotiated down tics are even worse, with measures being ap- under WTO auspices. The antidumping duties plied four times as frequently as their volume imposed by developing countries also are much of trade would imply. Developing countries higher than their tariff rates, with provisional are only marginally better in their treatment of measures ranging from 84 to 126 percent. developing countries, and impose antidump- And antidumping rates are discriminatory. ing measures against other developing coun- Antidumping decisions often apply different tries 50 percent more frequently than would be rates to imports from different sources (table suggested by the volume of imports. Industrial 2.9).11 The same biases are found in anti- countries are treated relatively lightly by both dumping rates as have been reported for tar- industrial and developing countries, with only iffs and various nontariff forms of protection: 43 percent as many actions imposed on them rates applied by developing economies are as would be implied by their value of imports. higher, and the bias against imports from de- Rates of protection being applied through veloping economies is even more pronounced antidumping measures are astonishingly high in rates applied by developing economies than (table 2.9). Though industrial economy tariffs in those applied by industrial economies. Table 2.9 Antidumping rates are much higher than tariff rates Averages of highest and lowest rates applied in antidumping cases (percent) Average antidumping margins Provisional measures Definitive duties Average Low High Low High tariff rates Industrial economies 28 41 31 48 4 Developing economies 84 126 58 83 13 All economies 50 75 43 64 5 Note: Post-Uruguay Round applied tariff rates; antidumping measures in place as of December 31, 2002; ad valorem rates. Source: Calculations based on countries’ notifications to the WTO. 86 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T The scale of the antidumping duties levied rules are frequently applied to much broader varies a great deal by country within groups, groups of products in developing countries with some countries imposing extraordinarily than has been the case in the past. high duties and other countries relatively mod- The large differences among countries in erate rates—although virtually all such duties antidumping duty rates highlight the flexi- are very high relative to currently prevailing bility in the WTO rules on antidumping, flex- tariff rates. While some countries, such as Aus- ibility that allows countries to “find” high tralia, impose duties in the same order of mag- margins of dumping, and to allow countries nitude as current tariffs, some countries are im- to impose high duties against them. It strains posing duties at truly astronomical rates, and credulity to believe that exporters to Mexico rates that seem likely to be prohibitive under and Argentina are 10 times as prone to dump- almost all circumstances. A number of Latin ing as exporters to the United States. Different American countries have measures above 100 interpretations of the same WTO rules, there- percent, and frequently substantially above this fore, are leading to widely different outcomes, level—the rates in Argentina, Mexico, and Peru raising serious questions about the objectivity exceed 300 percent (table 2.10). Further, these of the process. Table 2.10 Antidumping duties are high Average dumping margins for measures currently in place, by country (percent) Average antidumping margins Against all economies Provisional measures Definitive duty By country Low High Low High Argentina 163 328 62 63 Australia 6 16 20 43 Brazil 54 64 38 47 Canada 40 41 42 42 China 28 50 27 50 Colombia 34 40 53 67 Czech Republic n/a n/a 29 73 Egypt n/a n/a 22 55 European Union (15) 33 45 32 46 India 64 91 69 105 Israel 4 10 11 26 Jamaica 256 256 104 104 Korea 47 56 31 41 Malaysia 28 65 10 25 Mexico 269 345 51 65 New Zealand 42 42 28 62 Peru 75 330 73 246 Poland n/a n/a 23 23 South Africa 40 63 38 53 Taiwan, China n/a n/a 76 116 Thailand 36 47 32 36 Trinidad and Tobago 119 119 135 135 Turkey n/a n/a 93 101 United States 30 49 29 50 Venezuela 116 119 123 123 n/a not applicable. Note: Dumping margins associated with definitive measures in place as of December 31, 2002. Where applicable, numbers reflect dumping margins determined during the latest review of each case. For several countries (China, India, Peru, and Thailand), the table illustrates dumping margins related to antidumping measures in place as of June 30, 2002. Source: Semiannual reports under Article 16.4 of Antidumping Agreement, submitted by individual WTO members to Committee on Antidumping Practices. The list of countries currently maintaining antidumping measures has been extracted from the Report (2002) of the Committee on Antidumping Practices, and from the above-mentioned Semiannual Reports of WTO members. No data were available for Indonesia or the Philippines. 87 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 on their export opportunities. The Ministerial Figure 2.7 Antidumping barriers by Declaration launching the Doha Development sector and by country group Agenda includes consideration of antidump- Dumping margins in place, as of December 2002: ing measures, and provides an opportunity for target countries and sectors beneficial reform. Although reform will be dif- High Technology ficult given the strong political support for this Manufactures type of protection from industries in both de- veloped and developing countries, the rapidly Medium Technology Manufactures growing economic damage to all economies— those imposing these duties and those suffer- Low Technology ing from them—makes reform a high priority. Manufactures Transition economies ᮡ Resource Based Developing From Doha to Cancún and ᮡ Manufactures economies ᮡ Industrial beyond: How should protection Primary economies be reduced? M Products ost of the protection in world markets 0 20 40 60 80 100 120 140 is imposed through barriers to market Average dumping margins (%) access—an essential pillar of other elements of Source: Calculations based on countries’ notifications to protection regimes, such as export subsidies in the WTO. agriculture. Without a supporting tariff to pre- clude imports, an export subsidy will become merely a subsidy on return—exports will flow out in order to collect the subsidy, but the The intensity of use of antidumping mea- goods exported will be replaced by imports, sures depends heavily upon the sectors in which with no significant impact on domestic pro- it is applied. Figure 2.7 shows the height of the ducer prices. Reducing tariffs and other bar- barriers imposed against developing countries riers to market access, therefore, is the central by the level of technology in their exports. It re- issue in removing the protection that distorts veals that antidumping barriers are dispropor- world markets. tionately high against their exports of primary WTO members are negotiating, in accor- products, resource-based manufactures, and dance with the agenda agreed at Doha, on ap- low-technology manufactured exports. For pri- proaches to reduce protection on agricultural mary products, the average antidumping duty and nonagricultural goods. For nonagricultural against developing countries is three times as goods, in particular, the negotiations are to in- high, at around 60 percent, as against the in- clude reduction or elimination of tariff peaks dustrial countries. This means that antidump- and tariff escalation, and to emphasize prod- ing barriers are relatively lower on the most ucts of interest to developing countries (WTO dynamic exports of developing countries, high- 2001). An emphasis on reducing tariff peaks is technology exports. However, the antidumping important both because such peaks are very barriers against these exports remain very high costly to the countries imposing such tariffs, in absolute terms, at 65 percent. and because they are frequently on products The use of antidumping actions to generate of particular interest to developing countries. protection has reached an advanced stage. It Francois, Martin, and Manole (2003) find that now threatens developing countries’ trade approaches to trade reform that most sharply both through the damaging effects on their reduce peak tariffs result in larger reductions in own economies, and through adverse impacts the average tariffs facing low-income countries. 88 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T Box 2.3 The scourge of the specific F or most products, most coun- tries impose protection through ad valorem tariffs. This Ad valorem equivalents of specific tariffs are usually higher on Ad valorem equivalents of specific tariffs are usually higher on imports from imports developing from developing countries countries than from high-income than from high-income countries countries form of protection is more trans- a. Ad valorem equivalent (AVE) protection on overcoat exports to the United States parent than other types of tariffs, Ad valorem equivalents of U.S. specific tariffs on overcoats (percent) such as specific or compound tar- 30 Ecuador Vietnam iffs. A 10 percent tariff rate raises Pakistan AVE 25 the price of an imported good by ᮡ Venezuela France the same 10 percent whether the 20 India China Australia Ireland United Kingdom Germany Spain Japan good is a car worth $20,000 or a Thailand Uruguay Italy Canada Belgium 15 Morocco Turkey ᮡ bicycle valued at $50. A specific Bolivia New Zealand Linear (AVE) tax of $50, by contrast, has an 10 enormously different impact on 5 the bicycle and the car. Further, 0 the impact of the specific tariff 400 900 1,200 4,800 14,900 21,300 23,300 36,000 depends on market conditions. If GNI($)/capita bicycle prices tumble to $20, then b. AVE tariffs on exports of leather shoes from Pakistan to Japan the specific tariff of $50 will raise Ad valorem equivalents of EU specific tariffs on leather shoes (percent) prices by 250 percent instead of 800 100 percent. PAK 700 Specific tariffs, which are 600 AVE very common on agricultural ᮡ 500 VNM products, also are important on a 400 number of industrial products of IND LKA THA Linear (AVE) 300 interest to developing countries, IDN ᮡ 200 CHNMARROM TWN particularly textiles, clothing, and TUR BRAARG HKG PRTGRC ESP DEUAUTNLDGBR 100 ITA FRA USA footwear. Recent research by Schott (2001) has revealed a 0 striking tendency for poorer –100 countries to export versions of 0 0 0 0 0 1, 0 0 2, 0 0 3, 0 0 10 00 11 0 13 00 14 0 19 00 22 0 23 00 23 0 24 00 24 0 25 00 34 0 00 41 42 50 70 80 90 20 70 00 50 00 0 0 0 0 0 0 0 ,7 ,8 ,4 ,9 ,5 ,7 ,7 ,9 ,0 ,2 ,3 ,9 1, 2, 7, many manufactured products GNI/cap with lower unit values than those exported by higher-income coun- c. Ad valorem equivalent (AVE) protection on starch exports to the European Union tries. As countries develop, they Ad valorem equivalents of EU specific tariffs on starch (percent) frequently upgrade their prod- 300 Czech Republic ucts, using more capital and skill 250 AVE (compound) to produce a better version of the 200 Thailand ᮡ same product. Schott finds that ᮡ 150 the unit value of a shirt from Linear AVE (compound) 100 Japan, for instance, is 20 times Australia Canada that of the same good from the 50 Japan Philippines. In this situation, spe- 0 1,970 5,270 19,770 21,340 35,990 cific tariffs are likely to be much GNI($)/capita more of a burden on poorer de- veloping countries than on indus- Source: Stawowy (2001) and unit value data from COMTRADE. trial countries. (Box continues on next page) 89 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Box 2.3 (continued) We draw on a detailed analysis of ad valorem Also in the EU market, Japan pays less than 50 per- equivalents undertaken at UNCTAD (Stawowy cent, while Thailand and the Czech Republic pay 2001) to compare these burdens across countries. over 100 percent on their exports of starch—surely We find many cases where the ad valorem enough to stiffen resistance to this form of protection equivalents of specific tariffs are sharply higher on once the facts are known. poor than on rich countries. The top figure shows, The nontransparency of specific tariffs, and the for instance, that Vietnam and Ecuador face tariffs fact that they can discriminate so strongly against on overcoats twice as high as Japan does—a type of lower-income countries, are two good reasons for discrimination covered up by the specific tariff. Ex- supporting proposals (WTO 2003a, Annex I) to ports of leather shoes from Pakistan to Japan feel the eliminate specific tariffs under the Doha Develop- boot strongly—they are subject to tariffs of around ment Agenda. The proposals for the elimination of 700 percent, while exports from the advanced indus- specific tariffs clearly deserve strong support from trial countries are taxed at less than 100 percent. everyone concerned about the welfare of the poor. Using a formula is important oping countries were excluded, partly because A key choice in multilateral market-access developing countries were not active deman- negotiations is whether to proceed using a deurs in these negotiations. “request-and-offer” approach or a formula ap- The Uruguay Round (1986–94) used a sim- proach. Under a request-and-offer approach, pler approach for nonagricultural tariffs that countries with major supply-and-demand in- involved setting broad tariff-reduction goals, terests in a particular area agree on bilateral such as a 30 percent reduction on industrial tariff “concessions” that are then generalized products, but left the distribution of the cut to all other members on a most-favored-na- across sectors up to negotiations between trad- tion basis. As noted by Baldwin (1987), this ing partners. This approach was successful in approach was successful in achieving substan- achieving substantial tariff reductions. It was tial reductions in protection under the GATT not, however, successful in achieving higher only in the initial Geneva Round of negotia- proportional cuts in higher tariff rates and tions in 1947. The request-and-offer approach thereby in reducing tariff escalation. Abreu made disappointingly slow progress in the (1996) observes that manufactured goods with four following rounds of negotiations.12 Only higher tariff rates typically had smaller pro- with the introduction of a comprehensive for- portional tariff cuts. mula approach during the Kennedy Round In the Doha negotiations, there appears to (1963–67) was it again possible to cut protec- be a broad consensus that some sort of for- tion substantially—35 percent versus an aver- mula will be required to obtain reductions in age of 2.5 percent in the previous four negoti- protection sufficiently broad-based to achieve ations. The next round, the Tokyo Round increases in market access. The increase in the (1974–79), used a more sophisticated for- number of active participants seeking to use mula, the so-called Swiss formula. It achieved the WTO to achieve increases in market a 30 percent reduction in average tariffs and access—and to lock in reductions in tariffs— brought down the higher tariffs by much more would make the one-on-one, request-and-offer than the lower ones. Unfortunately, however, procedures even less likely to succeed than they many products of particular interest to devel- have in past GATT negotiations. 90 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T Some formulas are more prodevelopment large distortions in production and consump- than others tion patterns. The economic costs of these dis- Even with a broad agreement to employ a for- tortions rise with the square of the tariff, so the mula, much needs to be done to bridge the gap costs of a tariff twice as high as the average are among alternative approaches. Many different four times those of an average tariff. Further, approaches have been proposed, and not all high tariffs are inefficient in raising tariff rev- can be considered in this study, but examina- enues because the volume of imports across tion of a few key proposals is instructive. In these tariff barriers is likely to be small. agriculture, the United States and the Cairns From the point of view of developing-coun- Group have proposed a Swiss-formula ap- try exporters, approaches that allow countries proach that would reduce all tariffs below 25 to retain tariff peaks—and especially approaches percent, regardless of their initial level. At the that would allow the industrial countries to re- other end of the spectrum, Europe and Japan tain high peaks—are likely to be problematic have proposed an approach involving a “head- for another reason. The tariff structures of most line” reduction of 36 percent that could easily industrial countries involve quite low average be evaded, and only a 15 percent required re- tariffs as a result of the eight previous rounds of duction in individual tariffs. This would allow GATT/WTO negotiations. But they continue to protection for Japanese rice, for example, to contain many high peaks on products of partic- remain at close to 600 percent. ular interest to developing countries (Hoekman In nonagricultural market access, China, and Olarreaga 2002). The coefficient variation India, and Korea all have made proposals for of industrial country tariffs—the variation rela- formulas that would sharply reduce tariffs and tive to the average level—is now much higher in reduce high tariffs relative to average tariffs. these countries than it is in developing coun- The European Union has advocated an ap- tries, where the wave of tariff reductions during proach involving different tiers for tariffs, an the 1980s and 1990s cut higher tariffs in line approach that would facilitate larger reduc- with average tariffs. tions in high tariffs relative to lower rates. There is a particular concern with ap- Japan has proposed a formula that would give proaches that allow the retention of high countries flexibility with individual tariffs but peaks—such as the average-cut approach pro- that would require larger reductions in aver- posed by a number of high-protection coun- age tariffs on high-tariff goods. tries in the negotiations on agricultural market Adopting a formula that limits tariff peaks access.13 Such an approach, based on a mean- relative to average tariffs is important for de- ingless measure, provides scope for evasion of veloping countries. Some countries have pro- countries’ commitments in the Doha Develop- posed “flexible” approaches allowing reten- ment Agenda to reduce protection and to pro- tion of peak tariffs, while others have proposed vide special and differential treatment in favor approaches that sharply reduce peaks relative of developing countries. to other tariffs. Politically, it is frequently at- A wide range of formulas is available, the tractive to retain high tariffs on sensitive prod- effects of which depend upon a combination of ucts where political support for protection is factors, including the extent to which the for- strong. This political convenience needs to be mulas would reduce average rates of protec- weighed against the adverse implications for tion, as well as the variability of tariffs around the country itself, and for its trading partners, that average (see box 2.4). An important issue if high protection is retained in some sectors. to understand is the extent to which changing Retaining high levels of protection in some sec- the dispersion of overall protection, for a given tors is likely to be costly to the importing econ- average cut in tariffs, affects the average re- omy because high rates of protection mean duction in protection facing developing coun- 91 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Box 2.4 “Average cuts,” the cut you have when you’re not having a cut Y ears ago, as concern grew about the impacts of drunk driving, the venerable Clayton’s Nonalco- holic Tonic repositioned its advertising with the slo- alization. It can be totally deceptive in its suggestion that sizable reductions in protection are required. Even when it is built into a tiered reduction in pro- gan “Clayton’s, the drink you have when you’re not tection, such as the three-band system of tariff having a drink.” Macho individuals could still exhibit cuts with higher cuts in higher tariffs proposed in the bravado of having one or two “for the road” the Harbinson draft (WTO 2003c), the use of the while avoiding the sharply increased penalties associ- average-cut approach makes the average cuts speci- ated with drunk driving. Today, as concern builds fied in each group almost meaningless, by providing about the adverse impacts of agricultural policy dis- an incentive to reduce the highest tariffs in each band tortions for world trade, and particularly for develop- by the smallest possible amount. ing country farmers, the notion of an average cut in The proponents of the average-cut approach tariffs takes on a similar tone. With this device, poli- argue that more specific formulas are not acceptable cymakers can commit to sharp cuts in agricultural because they do not provide the flexibility needed to protection without actually doing anything at all. reduce their agricultural tariffs. But, as we have seen, To see how this paradox arises, consider a puta- the average-cut approach provides complete flexi- tive agreement to cut agricultural protection by 50 bility to do nothing (or anything), and none of the percent in a country with just two agricultural tar- discipline that is the sine qua non of world trade iffs—one of 1 percent and one of 100 percent. A cut rules. To provide some discipline, proponents of the of 100 percent in the 1 percent tariff, and of zero in average-cut approach allow for a minimum cut in the 100 percent tariff, yields the necessary 50 percent each tariff line. But this approach is extremely rigid average cut in tariffs—great for the headlines. unless the minimum cut is very low, in which case But in reality, virtually nothing has been done. there is no discipline. The average tariff has fallen by half of 1 percent, If the desire for flexibility made by the propo- from 50.5 percent to 50 percent. Of course, policy- nents of average cuts were accepted, what could be makers could load the cuts onto the high tariff, tak- done to provide some discipline? Fortunately, there is ing the average tariff from 50.5 percent to 0.5 per- a simple and nondeceptive alternative. A rule that cent. It is unlikely that they would do this, however. specifies a “cut in average tariffs” would allow for The 100 percent tariff is high because it is supported flexibility while ensuring that protection was reduced by strong interest groups, which would surely op- on average, while providing some reward for cutting pose their industry being “sacrificed.” The funda- high tariffs, rather than low tariffs. The simple, but mental problem with the average-cut approach is not purely semantic, move to a requirement for a cut that it provides no reward for cutting a high tariff in the average—rather than an average-cut—would rather than a low one, and hence allows policymak- preserve flexibility, while introducing some discipline. ers to avoid the agreed goal of achieving substantial improvements in market access. As we have seen, the “average-cut” approach provides little or no guidance on the effects of liber- Source: World Bank staff. tries. Approaches that attack tariff peaks more average tariffs. Francois, Martin, and Manole aggressively will be more beneficial to develop- (2003) show that approaches that are more ag- ing countries, since developing countries face gressive in reducing high rates of protection disproportionately high tariffs. Francois and will result in larger cuts in average rates of pro- Martin (2003) derive a flexible Swiss formula tection faced by developing countries. Such ap- that allows higher tariffs to be addressed more proaches also reduce variation in tariffs, which or less aggressively for any given reduction in is a major source of cost to the importing 92 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T Box 2.5 The implications of five tariff-cutting proposals A s is evident from the table, the proposal by China uses a Swiss-formula approach, but with a ceiling for each country based on the average level mula. Bound tariff rates are to be used as the base rates, with unbound tariffs for each product to be re- placed initially with the higher of the highest bound of its own tariffs (see Annex table 2A.1). After the rate in its schedule, or the applied rate for that prod- application of the formula, all tariffs will be below uct. For illustrative purposes, we set the proportional the original average, and countries with higher aver- cut at 50 percent. age tariffs have tariff reductions that are, in percent- The U.S. approach involves reducing to zero all age points, larger than those in countries with lower tariffs of 5 percent or less. Then, a Swiss formula tariffs. Countries with more variable tariffs will, in with a ceiling parameter of 0.08 is applied to all general, face larger percentage reductions in tariffs, other tariffs. Applied rates, rather than bound rates, because the tariff peaks that give rise to variability are generally used as the base. This approach is to be are reduced very sharply. In industrial countries, the implemented by 2010, with all tariffs being reduced formula is applied to countries’ applied tariff rates. to zero by 2015. In developing countries, the formula is generally ap- The proposal by the chair of the Market Access plied to the average of applied and bound tariffs. Committee involves use of a Swiss formula with the The proposal by the European Union seeks to ceiling parameter equal to each country’s average compress the distribution of tariffs by dividing tariffs tariff, scaled up or down by a parameter, B, to be up into bands and applying higher proportional cuts negotiated. The base rate is to be the bound rate, un- on tariffs in higher bands. Tariffs below some thresh- less the tariff line is unbound, in which case it is to old to be negotiated, such as 2 percent, are to be set be twice the applied rate. In addition, the proposal to zero. A specific proposal to set the bands is of- calls for reductions of tariffs to zero in a range of fered in an addendum to the proposal, and it is that sectors, including electronics, fish, footwear, leather specific proposal that we analyzed. goods, motor vehicle parts, and textiles and clothing. The proposal by India involves making a pro- In this initial assessment of the effects of the propos- portional cut, of a magnitude to be negotiated, in all als, these elements—whose inclusion is still to be tariffs. The proportional cut in developing-country negotiated—have not been included. tariffs would be two-thirds of that in developed countries. After application of the formula, tariffs above three times the average would be reduced to three times the average after application of the for- Source: World Bank staff. country (Martin, Van der Mensbrugghe, and base tariff rates to be cut, the approach taken Manole 2003). to cutting, the treatment of tariffs that are ini- A much wider range of approaches has been tially unbound, and whether any supplemen- offered in the negotiations on nonagricultural tary provisions call for eliminating tariffs on market access than in agriculture (see chapter particular groups of products. To gain some 3). We examine five proposals on market ac- idea of the consequences of these measures, we cess for nonagricultural products that are suf- examine their implications for tariffs facing ficiently specific to allow analysis of their im- low-income countries in four major markets— plications for tariffs—those from China, the Brazil, the European Union, India, and the European Union, India, the United States, and United States. from the chair of the WTO’s Market Access One key issue in any of the formulas is the Committee. The implications of any of these selection of the base tariff rates. Traditionally, formulas (box 2.5) depend heavily upon the the GATT/WTO system has used bound rates 93 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 as the basis for subsequent negotiations. As The situation in developing-country markets well as having the advantage of familiarity, is considerably different, however. There, tariff this approach has an important dynamic ad- reductions are much smaller as a proportion of vantage. If countries believe that bound rates the initial tariff rates—although not necessarily will be the basis for future negotiations, they in percentage points of tariff, since the initial will feel free to reduce applied rates when they tariff rates are much higher. The smaller reduc- are convinced that this is in their economic in- tions as a percentage of initial tariffs are a re- terests. Negotiations based on bound rates flection in large measure of the “tariff binding embody a pure, one-for-one form of credit for overhang”—the situation where tariff bindings such autonomous liberalization.14 If a country are substantially above applied rates. They also chooses to reduce its applied rates below the reflect the smaller percentage reductions for de- bound level, then reductions in applied rates veloping countries inherent in India’s formula, flowing from any future agreements to cut as well as India’s proposal to allow unbound bound rates by a particular percentage will be tariffs to be set at the highest binding in the smaller since any given cut in bound rates will schedule, or at the applied rate for the com- require a correspondingly smaller reduction in modity being liberalized, whichever is higher. applied rates. Use of applied rates as the base The U.S. formula would lead to substantial re- for tariff reductions creates potential incen- ductions in applied tariffs in India and Brazil, tives to increase applied rates and is certainly with tariffs facing low-income exporters falling likely to have a chilling effect on future au- to 6 percent in India and 5 percent in Brazil. tonomous liberalization. The EU formula also would result in large cuts In examining the implications of the differ- in applied rates in developing countries, with ent formulas, we build on the analysis of WTO tariffs facing low-income exporters falling to (2003b), which examines the implications of 10.8 in India and just over 8 percent in Brazil, tariff reduction formulas for a hypothetical a striking finding given that this approach is schedule of tariff bindings. We apply the tariff based on bound tariff rates. reduction formulas to the specified base tariffs A key issue is whether the reductions in tar- (bound, applied, or a combination), and then iffs are to be measured using percentage cuts examine the implications for applied rates by in tariffs or alternative measures, such as the comparing the resulting tariff bindings with percentage-point reduction in tariffs or the applied rates prior to the analysis. If the new percentage-point reduction in the price of im- binding is below the prior applied rate, then ports brought about by the tariff reduction.15 the new applied rate is reduced to the new As noted above, only the U.S. formula and, to bound rate. Table 2.11 presents weighted av- a lesser extent, the European formula, bring erage tariffs for each of the four focus econo- about large percentage-point reductions in mies under each of the five proposals. tariffs in India or Brazil. If however, progress A striking feature of the results is just how is measured in terms of the percentage-point sharply all five formula approaches reduce cut in tariffs or the percentage-point cut in im- tariffs in the industrial-country markets con- port prices, then the reductions in India and sidered. Average tariffs facing the low-income Brazil look much more substantial, even under countries in the United States, for example, the Indian formula or the Chinese formula, are reduced from 4.3 percent to 1.4 percent by where the percentage cut in tariffs is relatively the Chinese formula and 1.3 percent by the small. Clearly, much patient analysis will be U.S. formula. All of the formulas also would required to assess the consequences of the pro- substantially reduce the standard deviation of posed reforms for individual countries—and a tariffs in the industrial countries—indicating lot of hard bargaining before WTO members that they would reduce tariff peaks sharply. are likely to agree on a specific package of re- 94 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T Table 2.11 Competing formulas make a big difference for tariffs Weighted average on tariffs (percent) for Brazil, European Union, India, and the United States under various proposed formulas Tariffs facing exporters Applied tariffs, Applied tariffs, weighted average, weighted average, Applied tariff, low-income middle-income Standard Affected area Formula average countries countries deviation Brazil Initial tariff 15.95 14.55 14.00 5.96 Chinese formula 12.34 11.47 10.17 3.49 European formula 10.11 9.66 8.14 2.51 Indian formula 15.50 14.20 12.28 5.83 U.S. formula 5.04 4.71 4.00 1.20 Chairman’s formula B=2 15.60 14.20 12.37 5.68 Chairman’s formula B=1 13.33 12.17 10.64 4.08 European Union Initial tariff 4.18 5.28 3.76 3.72 Chinese formula 1.85 1.99 1.45 1.14 European formula 2.06 2.67 1.87 2.02 Indian formula 1.94 2.49 1.78 1.77 U.S. formula 1.29 1.73 1.23 1.87 Chairman’s formula B=2 2.18 2.52 1.80 1.64 Chairman’s formula B=1 1.56 1.74 1.25 1.08 India Initial tariff 32.99 28.12 26.71 8.57 Chinese formula 28.19 24.09 21.98 8.48 European formula 11.98 11.24 10.82 1.90 Indian formula 28.30 24.23 21.96 8.67 U.S. formula 6.31 5.93 5.71 0.73 Chairman’s formula B=2 29.40 25.15 22.96 8.14 Chairman’s formula B=1 23.78 20.99 19.36 5.82 United States Initial tariff 3.70 4.26 3.23 4.53 Chinese formula 1.40 1.35 1.11 1.14 European formula 1.76 2.10 1.56 2.20 Indian formula 1.50 1.69 1.25 1.67 U.S. formula 1.00 1.31 0.82 1.80 Chairman’s formula B=2 1.63 1.75 1.37 1.59 Chairman’s formula B=1 1.13 1.17 0.95 1.03 Source: World Bank staff. forms. We hope that the analysis presented areas—textiles, clothing, and other labor- here will make an important contribution to intensive manufactures, as well as agriculture— clarifying the issues and tradeoffs involved in in both rich and poor countries. In nonagri- these choices. cultural goods, three efforts are particularly crucial: Putting development into the Doha Development Agenda requires serious • First, the progressive phase-out of quotas liberalization under the Agreement on Textiles and Although trade policy reform has contributed Clothing, now lagging, is critical to pro- to an unprecedented shift in export composi- viding market access for developing tion and trade growth, the next steps will be countries. Reforms in current quota- difficult. Today’s protection remains heavily holding countries may be necessary to concentrated in the most politically sensitive ensure a smooth adjustment. 95 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 • Second, efforts to cut back antidumping mula, and of its coefficients of reduction, measures that create a patchwork of ad is important. hoc protection are essential if market ac- cess granted by the right hand, through Much must be done behind the border to quota elimination and tariff reductions, is ensure that countries that have missed out dur- not to be withdrawn by the left hand of ing the current wave of international integra- new access-restricting antidumping suits. tion will be able to take advantage of the op- • Third, moving forward in nonfarm trade portunities provided by a more streamlined requires a Swiss-formula, or related top- and development-friendly trading system. If down approach, that will require dis- these measures are adopted—along with oth- proportionately greater reductions in ers associated with agricultural trade, labor high tariffs so as to mitigate antidevelop- mobility, and the special treatment some devel- ment policy embedded in trade regimes oping countries enjoy—then the promise of the around the world. The choice of the for- Doha Development Agenda may be realized. 96 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T Table 2A.1 The various liberalization proposals have very different features Features of trade liberalization proposals advanced by China, the European Union, India, the United States, and the Chair of the WTO Market Access Committee Proposal Base rate (T0) Unbound tariff Sectoral tariff elimination Formula China Developed:  (A + B × P) × T  applied rate T1 =  0   (A + P ) + T0  2 Developing: simple average of applied and A: Simple average of T0 bound rate P=T0/A B=3 for 2010, B=1 for 2015 European Bound rate Applied duty,  BU − BL  Union November 14, T1 = B1 L + (T0 − B0 L )×  1 1    B0 − B0  U L 2001  BL U 0 and B 0 are lower and upper limits in the base bracket BL U 1 and B 1 in the new bracket India Bound rate Higher of Step 1. maximum bound T1 = (1 − (A × Y / 100)) × T0 rate and applied rate on cut-off date A=1 for developed countries A=0.67 for developing countries Step 2. Minimum of T1 and three times the average of results from Step 1 United States Lesser of applied Wood products; Step 1. Period to 2010. rate and bound non-ferrous metals; T1 = 0 if T0 ≤ 5% rate bicycle parts; soda ash; 8 × T0 photographic film; T1 = if T0 > 5% electronics; fish and 8 + T0 fishery products; scientific equipment; environmental Step 2. From 2015, all tariffs goods; Information go to zero. Technology Agreement (ITA) products and goods covered by the Agreement on Trade in Civil Aircraft, Uruguay Round zero- for-zero sectors B × Ta × T0 Chair, WTO Bound rate Two times the Electronics and electrical T1 = B × Ta + T0 Market Access MFN applied goods; fish and fish Committee rate (2001). products; footwear; leather Ta is the initial average tariff If MFN applied goods; motor vehicle parts B is a coefficient to be rate is zero, then and components; stones, determined 5 percent. gems, and precious metals; and textiles and clothing Source: WTO. 97 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Notes 9. Details on the barriers faced by individual coun- 1. One potential concern with any measure of the tries on individual products in particular markets can share of particular exports is the impact of price be obtained using the WITS software (see www.wits. changes. If, for instance, changes in the observed shares worldbank.org). of exports reflect primarily changes in prices over the 10. Another justification might be that such rules sample, then they might be reversed by subsequent price guide members toward good policy actions—actions that advance the national economic interest of the changes. It is difficult to fully adjust for price changes, members that apply them. This historically is not the but at least a crude indication was obtained by deflat- rationale for their inclusion in the GATT/WTO. Em- ing each component of exports by a suitable deflator pirically, extensive research has shown that antidump- and comparing the result with undeflated export shares. ing actions are not “good” protection, they are ordi- It appears that the changes in shares have primarily nary protection with a good public relations program. been the result of changes in the volume of exports. 11. The average of rates actually applied is thus at This implies that they are much more likely to be sus- least as high as the lower figure, no higher than the tained than changes resulting only from changes in the higher figure. prices of particular export goods. Increases in the prices 12. A simple formula was first tried out in the Dil- of resources—and particularly oil—have caused in- lon Round (Ernest Preeg, personal communication). creases in the importance of exports of these goods in 13. This approach focuses on achieving specified the past, but these have been reversed by subsequent average tariff cuts, rather than cuts in the average. Un- price declines. fortunately, a given average cut can likely be achieved 2. The use of categories defined by income status at with minimum political impact by imposing large cuts the beginning of the sample period makes an enormous in the lowest tariffs, rather than the large cuts in the difference. If we define our low-income sample by in- highest tariffs that are economically desirable. come status at the end of the period, the trade growth 14. Approaches that use historical applied rates, rate of the low-income group is below average. such as applied rates at the end of the Uruguay Round, 3. This analysis was undertaken using the 1997 would also have this feature of providing credit for input-output data from the Global Trade Analysis Proj- subsequent, autonomous liberalization. ect (GTAP), input-output information for 1997, and 15. This measure, which involves measuring the the GTAP time series database of trade and trade pat- change in the tariff rate divided by one plus the tariff terns. See Hummels, Ishii, and Yi (2001) for precise de- rate, is widely used in analytical work, and reported finitions of these measures. extensively in Finger, Ingco, and Reincke (1996). 4. While the resistances to integration graphically captured in the book “Why the Emperor’s New Clothes Are Not Made in Colombia” (Morawetz 1981) have References certainly abated, the continuing low level of integration Abreu, M. 1996. “Trade in Manufactures, the Out- suggests the continued existence of considerable resis- come of the Uruguay Round, and Developing tance to the level of integration associated with global Country Interests.” In Will Martin and L. Alan production sharing. Winters, eds., The Uruguay Round and the De- 5. The growth performance of this region is biased veloping Countries. Cambridge: Cambridge Uni- upward because of large-scale under-reporting of trade versity Press. prior to 1990. Amiti, M. 2003. “How the Sequence of Trade Liberal- 6. The rise in the price of factor inputs and the ization Affects Industrial Location.” Mimeo, Jan- prices of nontraded goods is frequently identified as the uary 17. World Bank, Washington, D.C. real exchange rate appreciation associated with protec- Baldwin, R. E. 1987. “Multilateral Liberalization.” In tion policies. J. Michael Finger and A. Olechowski, eds., The 7. This is an underestimate of the costs imposed by Uruguay Round: a Handbook for the Multilateral protection on exporters. In addition, tariffs raise costs Trade Negotiations. World Bank, Washington by raising the costs of nontraded goods and nontraded D.C. factors of protection. This adverse impact, the so-called Bhardwaj, A., S. Kathuria, and Will Martin. 2001. real exchange rate effect, should also be taken into “Implications for South Asian Countries of Abol- account. ishing the Multifibre Arrangement.” Policy Re- 8. Baldwin (1987) estimates that the average tariff search Working Paper 2721. World Bank, Wash- reductions by the industrial countries were 35 percent ington, D.C. in the Kennedy Round and 30 percent in the Tokyo Bloch, H., and D. Sapsford. 2000. “Whither the Terms Round. of Trade: an Elaboration of the Prebisch/Singer 98 T R A D E P A T T E R N S A N D P O L I C I E S : D O H A O P T I O N S T O P R O M O T E D E V E L O P M E N T Hypothesis.” Cambridge Journal of Economics Hummels, D., and P. Klenow. 2002. “The Variety and 24: 461–81. Quality of a Nation’s Trade.” Working Paper Brenton, Paul. 2003. “Integrating the Least Developed 8712. National Bureau of Economic Research, Countries into the World Trading System: The Cambridge, Mass. Current Impact of EU Preferences under Every- Hummels, D., J. Ishii, and K.M. Yi. 2001. “The Nature thing But Arms.” Policy Research Working Paper and Growth of Vertical Specialization in World 3018. World Bank, Washington, D.C. Trade.” Journal of International Economics Cadot, O., J. de Melo, and Marcelo Olarreaga. 2002. 54:75–96. “The Protectionist Bias of Duty Drawbacks: Evi- Ianchovichina, E. 2003. “GTAP-DD: A Model for An- dence from Mercosur.” Journal of Development alyzing Trade Reforms in the Presence of Duty Economics 59 (1): 161–82. Drawbacks.” GTAP Technical Paper 21, Center Corden, M. 1997. Trade Policy and Economic Welfare, for Global Trade Analysis, Department of Agri- 2nd edition. Oxford: Oxford University Press. cultural Economics, Purdue University, Lay- Deardorff, A. 2001. “International Provision of Trade fayette, Ind. [www.gtap.org.] Services, Trade and Fragmentation.” Review of Ianchovichina, E., and Will Martin. 2001. “Trade Lib- International Economics 9 (2): 233–48. eralization in China’s Accession to the WTO.” de Ferranti, D., G. Perry, D. Lederman, and W. Mal- Journal of Economic Integration 16 (4): 421–45. oney. 2002. From Natural Resources to the _______. 2003. “Economic Impacts of China’s Acces- Knowledge Economy: Trade and Job Quality. sion to the WTO.” World Bank, Washington, Latin America and Caribbean Studies, World D.C. [www.worldbank.org/trade] Bank, Washington, D.C. Kehoe, T., and K. Ruhl. 2002. “How Important is the Dollar, D., M. Hallward-Driemeier, and T. Mengistae. New Goods Margin in International Trade?” 2003. “Investment Climate and Firm Perfor- Mimeo. University of Minnesota, Minn. mance in Developing Economies.” Mimeo, Madani, D. 1999. “A Review of the Role and Impact March 5. World Bank, Washington, D.C. of Export Processing Zones.” Policy Research Elbehri, A., T. Hertel, and Will Martin. 2003. “Esti- Working Paper 2238. World Bank, Washington, mating the Impact of WTO and Domestic Re- D.C. forms on the Indian Cotton Industry.” Review of Martin, Will. 1993. “The Fallacy of Composition and Development Economics 7 (3):343–59. Developing Country Exports of Manufactures.” Evenett, Simon J., and A. Venables. 2003. “Export Growth by Developing Countries: Market Entry World Economy 16 (2): 159–72. and Bilateral Trade.” Mimeo. World Trade Insti- Martin, W., van der Mensbrugghe D., and V. Manole. tute, Berne, Switzerland. 2003. “Is the devil in the details?: Assessing the Finger, J. Michael, Ingco M.D., and U. Reincke. 1996. Welfare Implications of Agricultural and Non- “The Uruguay Round: Statistics on Tariff Con- agricultural Trade Reform.” World Bank Work- cessions Given and Received.” World Bank, ing Paper forthcoming, World Bank, Washington, Washington, D.C. D.C. François, J., and Will Martin. 2003. “Formula Ap- Morawetz, D. 1981. Why the Emperor’s New Clothes proaches for Market Access Negotiations.” are not Made in Colombia: a Case Study in Latin World Economy 26 (1):1–28. American and East Asian Exports. Washington, François, J., Will Martin, and V. Manole. 2003. D.C.: Oxford University Press for the World “Choosing Formulas for Market Access Negotia- Bank. tions.” Revised version of paper prepared for Nehru, V., and A. Dhareshwar. 1993. “A New Data- Festschrift conference for Professor Peter Lloyd, base on Physical Capital Stock: Sources, Method- Melbourne. ology, and Results.” Revisita de Análisís Econó- François, J., and D. Spinanger. 2002. “China Acces- mico 8 (1): 37–59. sion, ATC Phaseout, and Market Access in Tex- Nehru, V., E. Swanson, and A. Dubey. 1995. “A New tiles and Clothing.” Mimeo. Erasmus University, Database on Human Capital Stock in Developing Rotterdam. and Industrial Countries: Sources, Methodology, Hoekman, B. and M. Olarreaga. 2002. “Une proposi- and Results.” Journal of Development Economics tion pour l’OMC: La « Super » clause de Nation 46: 379–401. Plus Favorisée.” Reflets et Perspectives de la Vie OECD (Organisation for Economic Co-operation and Economique, XLI, 2. Development). 2002. Monitoring and Outlook Hummels, D. 2001. “Time as a Trade Barrier.” Mimeo. Report. Paris. Krannert School of Management, Purdue Univer- Pritchett, L. 1997. “Divergence, Big Time.” Journal of sity, West Lafayette, Ind. Economic Perspectives 11 (3): 3–17. 99 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 _____. 2000. “The Tyranny of Concepts: CUDIE (Cu- World Bank. 2000. Global Economic Prospects 2001. mulated, Depreciated Investment Effort) Is Not Washington, D.C. Capital.” Journal of Economic Growth 5: 361–84. World Bank. 2002. Global Economic Prospects 2003. ______. 2001. “Where Has All the Education Gone?” Washington, D.C. World Bank Economic Review 15 (3): 367–91. World Bank. 2003. World Development Indicators. Redding, Stephen, and Anthony Venables. 2001. “Eco- Washington, D.C. nomic Geography and International Inequality.” WTO. 2001. “Ministerial Declaration.” Document WT/ CEP Working Paper 495. London School of Eco- MIN(01)/DEC/1, November 20. nomics, London. WTO. 2003a. “Draft Elements of Modalities for Ne- Schott, P. 2001. “Do Rich and Poor Countries Special- gotiations on Non-Agricultural Products.” World ize in a Different Mix of Goods? Evidence from Trade Organization, Negotiating Group on Mar- Product-Level U.S. Trade Data.” Working Paper ket Access, TN/MA/W/35, Geneva. 8492. National Bureau of Economic Research, WTO. 2003b. Formula Approaches to Tariff Nego- Cambridge, Mass. tiations: Note by the Secretariat.” Document TN/ _____. 2002. “Across-Product Versions Within-Product MA/S/Rev 2, April 11. Specialization in International Trade.” Mimeo. WTO. 2003c. “Negotiations on Agriculture: First Yale School of Management, New Haven, Conn. Draft of Modalities for the Further Comments.” _____. 2003. “One Size Fits All? Heckscher-Ohlin Spe- Document TN/AG/W/1/Rev. 1, March 18. cialization in Global Production.” American Eco- Yang, Y., W. Martin, and K. Yanagishima. 1997. nomic Review 93(3): 686–708. “Evaluating the Benefits of Abolishing the MFA Stawowy, Wojciech. 2001. “Calculation of Ad Valorem in the Uruguay Round Package.” In T. Herteb, Equivalents of Non-Ad Valorem Tariffs.” Mimeo, ed., Global Trade Analysis: Modeling and Appli- UNCTAD, October. cations. Cambridge, Mass.: Cambridge University Todaro, M. 1994. Economic Development. New York: Press. Longman. 100 3 Agricultural Policies and Trade Trade in agriculture is important to the the industrial countries. In products for which world’s poor— they competed with industrial countries, such Agriculture is the largest employer in low- as sugar and beef, some countries could export income countries, accounting for about 60 per- limited amounts under preferential-access pro- cent of the labor force and producing about 25 grams. In an effort to generate public revenues percent of GDP. Even in middle-income coun- from commercialized export activities, govern- tries, where agriculture’s share of GDP is only ments levied export taxes on agricultural prod- about 15 percent, the sector still accounts for ucts while protecting manufacturing through more than 25 percent of employment. When high import tariffs and other import restric- coupled with agro-related industries and food- tions. Even for agricultural products that were related services, its share, even among middle- not exported, price controls, exchange rate income countries, is typically 25 to 40 percent policies, and other restrictions kept prices low of GDP. About 73 percent of the poor in de- for urban consumption. veloping countries live in rural areas. Rural In the last decade, developing countries development, therefore, is central to alleviating shifted from taxing agriculture to protecting it. poverty. Import restrictions on manufactured products Government policy has heavily distorted have declined dramatically, exchange rates have agricultural performance in both developing been devalued, multiple-exchange-rate systems and developed countries. Until the 1990s, in- penalizing agriculture have been abandoned, dustrial countries generally protected agricul- and export taxes have effectively disappeared ture, whereas developing countries generally (World Bank 2000; Jansen, Robinson, and Tarp taxed it (Schiff and Valdes 1992). Industrial 2002; Quiroz and Opazo 2000). Meanwhile, countries supported their agricultural sectors reforms in most industrial countries, including through subsidies to producers, high tariffs, many of the successful middle-income coun- and other nontariff measures such as import tries, have been modest—despite the inclusion restrictions and quotas. of agriculture under the World Trade Organi- zation (WTO) in the Uruguay Round of inter- —but agricultural policies have often national trade negotiations. The result of these worked to the detriment of the poor policies has been overproduction and price de- Most of the developing countries generated clines in many commodities, reducing opportu- the bulk of their agricultural GDP in lower- nities for many developing countries to expand efficiency production for the domestic market, exports and penalizing the world’s poor. supplying the world market with tropical com- Consequently, although developing coun- modities that could not easily be produced in tries have almost doubled their share of world 103 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 trade in manufactures over the last two countries covered by the Generalized System of decades, their share in agricultural trade has Preferences (GSP) were eligible for preferences, been stuck at around 30 percent. During the and 26 percent of imports received them. De- 1990s, the growth of developing-country agri- veloping countries, too, have maintained high cultural exports to industrial countries slowed border protection and, on average, have higher as exports to other developing countries ac- agricultural tariffs than industrial countries. celerated. During this period, 56 percent of However, direct comparisons are difficult be- the growth of developing-country agricultural cause of the complex nature of protection in trade was accounted for by sales to other de- industrial countries. veloping countries and 44 percent by sales to Within OECD countries, budget subsidies industrial countries. The middle-income coun- and transfers from consumers (from high tar- tries have managed to increase global market iffs and quantitative restrictions on domestic share, principally by entering into other devel- production of selected commodities) amounted oping countries’ markets and by aggressively to about $250 billion in 1999–2001. This pro- diversifying into nontraditional exports, such tection decreased from 62 percent of farm as seafood products, fruits, vegetables, cut revenues in 1986–88 to 49 percent in 1999– flowers, and processed foods. Growth of these 2001—still a very high percentage. Of this sup- nontraditional exports has outpaced growth port, 70 percent came from consumers via of traditional commodities by three to one. higher prices associated with border protection Meanwhile, many low-income countries, ex- and 30 percent from direct subsidies. In devel- cept for China, have had less success—their oping countries, almost all support is gener- share of world agricultural trade has declined. ated by border barriers. A silver lining to this dark cloud is that some developed-country High border protection in rich countries subsidies have been at least partially delinked frustrates development from levels of production, lowering the incen- These patterns reflect—among other things— tive to overproduce. These partially decoupled the structure of global protection. Border pro- subsidies increased from 9 percent of total sub- tection in rich countries continues to be high, sidies in 1996–98 to more than 20 percent in nontransparent, and antidevelopment. Average 1999–01. agricultural tariffs in industrial countries, Although official export subsidies may be when they can be measured, are two to four small and shrinking, effective export subsidies times higher than manufacturing tariffs. In ad- created by domestic support are increasing, dition, about 28 percent of domestic produc- lending unfair advantage to industrial coun- tion in countries belonging to the Organisation try producers. Currently, cotton is not classi- for Economic Co-operation and Development fied as receiving export subsidies. Its domes- (OECD) is protected by tariff rate quotas tic and export prices in the United States and (TRQs). More than 40 percent of the tariff the European Union are the same—and those lines in the European Union (EU) and United prices are less than half the cost of production. States contain specific duties, which make it Similar differences exist in many other prod- difficult to calculate average tariffs and ob- ucts, a gap that will increase as industrial scure actual levels of protection. Tariff peaks countries move from protection through bor- as high as 500 percent confront imports from der barriers to support through coupled or developing countries. Tariffs also increase by partially decoupled subsidies. degree of processing, creating a highly esca- lating tariff structure that limits access for Success in the Doha Round requires processed foods. Preferences do not compen- reductions in agricultural protection sate for these high levels. In the United States, To be meaningful for the world’s poor, the only 34 percent of agricultural imports from Doha Round must bring reductions in agricul- 104 A G R I C U L T U R A L P O L I C I E S A N D T R A D E tural protection around the world. The benefits culture without support can be more dynamic of global liberalization in agriculture—elimina- and efficient. tion of all border barriers and subsidies—are Finally, along with greater market access, estimated to be very large for industrial and de- low-income countries need help in eliminating veloping countries alike, topping $350 billion behind-the-border barriers, especially the seg- for the world. With liberalization, agricultural mentation of their rural markets. Those mar- production would marginally shift from North kets should be linked to wider markets at to South, and the highly depressed world prices home and abroad (box 3.1). for many commodities would increase: 10–20 percent for cotton, 20–40 percent for dairy products, 10–20 percent for groundnuts, 33–90 Poverty, rural households, percent for rice, and 20–40 percent for sugar and trade in agriculture (Beghin and Aksoy 2003). The impact of these price changes on low-income net importers Agriculture is the livelihood of the would be small and manageable. To date, how- world’s poor ever, many of the proposals designed to elicit Growth in agriculture has a disproportionate consensus on agricultural reform are modest. effect on poverty because more than half of The average applied tariffs in the Quad coun- the population in developing countries resides tries would be halved at best under such pro- in rural areas.1 Some 57 percent of the devel- posals. Tariff peaks would remain above 100 oping world’s rural population live in lower- percent for many countries. The outcomes for middle-income countries; 15 percent in the developing countries are even less significant. least developed countries (LDCs).2 Although For most of them, the cuts required by one most of the world’s poor countries are in Sub- prominent proposal would leave their bound Saharan Africa, they account for about only tariffs above their current applied rates, and tar- 12 percent of developing world’s rural popu- iff escalation and peaks would still be very high. lation, whereas Asia accounts for 65 percent. A serious agreement to reduce border pro- Using the $1-a-day measure of poverty, tections would produce benefits for the most of the world’s poor live in India, China, world’s poor that far exceed those that can be and other lower-middle-income countries anticipated from present levels of develop- (table 3.1). National poverty data—which ment assistance. A first order of business is to allow separation of rural and urban household create a more transparent and simpler trade information but are not available for all coun- regime in all countries by converting specific tries—yield results that are very similar to tariffs to ad valorem tariffs, eliminating mini- those obtained using the $1-a-day measure. mum price regulations, cutting peak tariffs, They show that four countries—India, Bangla- changing the structure of TRQs so they in- desh, China, and Indonesia—account for 75 crease over time, and introducing a transpar- percent of the world’s rural poor. It is in Asia, ent system of reallocation to more efficient therefore, that rural income growth will have producers. Rich countries should phase out the greatest impact on rural poverty. export subsidies and subsidies that encourage overproduction, both of which are directly Poverty is more common in rural areas prejudicial to poor farmers around the world. In countries for which separate rural and These reforms would also make the agri- urban income data are available, 63 percent culture in industrial countries more efficient, of the population, and 73 percent of the poor, environmentally sustainable, and more sup- live in rural areas. This is true for all regions. portive of the small family farms. The experi- A high incidence of rural poverty is found in ence of New Zealand, the only OECD country all developing countries, whatever their level of to reform fully, clearly demonstrates that agri- income. More of the population is poor in low- 105 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Box 3.1 The impact of national trade integration and reform on poverty P overty in rural areas of low-income countries is closely correlated with distance to local and na- tional markets. In addition to geographic distance, only 55 percent would benefit without these condi- tions. Without improvements in the functioning of local and national markets, economic gains for the the concept of distance to market includes various poor may reach only one-fourth of their potential. costs of moving goods to and from markets. A case study in Madagascar illustrates that Case studies in Armenia, Malawi, and Nepal improvements in trade policies may not be sufficient show that reductions in transportation costs bring to restore sustained growth in the agricultural sector strong gains in household welfare for individual without better transport infrastructure and other re- farmers. Among these households, the poorer ones forms. In Madagascar, where poverty is closely related benefit disproportionately because transportation to remoteness, defined to include lack of infrastructure costs make up a larger percentage of their household and access to basic services, integrating the poor into expenditures. regional markets and the national economy will make Case studies in Ethiopia and Guinea reveal that a real contribution to increasing their incomes. In the many of the poor will be left behind by trade reform absence of integration, economic growth will tend to if no improvements are made in domestic markets. benefit those who are already favored. In Ethiopia, for example, 80 percent of the poor would benefit from freer trade under conditions of full market participation and price transmission, but Source: Kudat, Ajwad, and Sivri (2003). income countries, however, and in the LDCs Most poor countries are very dependent the poverty rate for rural households reaches on agriculture for household income. In Ethio- almost 82 percent (table 3.2). The rural share pia and Malawi, for example, about three- of the total number of poor households is de- quarters of household income is derived from clining with urbanization. Still, with current agricultural activities, mainly subsistence farm- trends, the rural share of the global number of ing. But cash income is also crucial (table 3.3). poor will not fall below 50 percent before Whether derived from cash (export) crops or 2035 (Ravallion 2000). other sources, cash income allows farmers to Table 3.1 Most of the world’s poor live in rural areas outside the least developed countries Distribution of poor in developing countries (1999) Percentage Poverty headcount, Population in millions (2001) of world’s under $1/day in 1999 Percent rural National Rural Urban rural population (percent) (millions) Least developed countries 596 443 153 74 15 49 292 Other low income 839 501 338 60 17 26 218 Middle income a 1,435 478 957 33 16 8 114 China 1,272 805 467 63 27 18 226 India 1,032 745 288 72 25 35 358 Total 5,175 2,972 2,203 57 100 23 1,209 a. Excluding China and India. Source: World Bank data. 106 A G R I C U L T U R A L P O L I C I E S A N D T R A D E Table 3.2 Rural poverty is higher in poorer sources of income include salaries and wages countries from other activities; investment income such Share of national population and of poor living in rural areas (percent) as interest, dividends, and rents; and social transfers from health, pension, unemployment, Samplea All developing countries and child-allowance schemes. Share of rural Rural dwellers who Rural dwellers are poor dwellers Farmers in industrial countries earn above-average incomes Upper middle In many industrial countries, the average in- income 19 37 22 Lower middle comes of farmers are higher than the national income 64 72 61 average, reaching almost 250 percent of aver- Low income 65 74 60 age income for the Netherlands, 175 percent Least developed 76 82 68 All developing for Denmark, 160 percent for France, and 110 countries 63 73 56 percent for the United States and Japan. In most other countries, the level of income is a. Sample consists of 52 countries for which separate rural and urban income data are available. either equal to or marginally lower than the Source: World Bank data. average income (OECD 2002d). In lower in- come OECD countries such as Greece, Korea, and Turkey, rural incomes are lower—around Table 3.3 Even in subsistence economies, cash is important 75–80 percent of urban incomes. Percentage of total household income derived from various As countries become wealthier, the share of sources in rural areas, 1990s rural household income from nonfarm sources Ethiopia Malawi Mexico rises. Off-farm income for major field crops in the United States, for example, is more than ten Total agricultural income 77 76 24 times greater than farm income and eight times Agricultural cash income 18 16 21 Subsistence farming 59 60 3 greater than government payments (table 3.4) Transfers 16 7 13 Government payments exceed what U.S. farm- Wages 3 8 42 ers make from the market in farming. In fact, Other 4 9 21 Total 100 100 100 most farms lose money from farming alone.3 Source: World Bank household data. Of agricultural subsidies, only half reaches farmers, and most goes to the richest buy inputs—such as fertilizers—that increase Agricultural protection in industrial countries food-crop yields, lowering the incidence of helps the relatively better-off rural house- poverty and malnutrition. holds—and it does so very inefficiently.4 Ac- The share of nonfarm income in rural households increases with a country’s level of development. In Mexico, for example, the Table 3.4 U.S. farmers earn less from share of farm income in total rural income is farming than from other sources much lower than in Ethiopia and Malawi. In- Shares of U.S. farmers’ income from various sources (billions of dollars) comes from farming are complemented by other sources, so that the direct impact of agri- Income source Value cultural price and output variations have a Farming 11.6 much smaller impact on rural households. In Government payments 14.7 industrial countries, when a broad definition Off-farm activities 122.7 of farm households is adopted, the share of Source: USDA, “Agricultural Income and Finance Outlook,” farm income declines even further. Other September 26, 2002. 107 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 cording to OECD estimates, agricultural sup- worth. In Japan and Canada, the largest 25 port policies deliver additional income to farm percent of farms receive 68 percent and 70 per- households at a rate of 50 percent or less of the cent of support payments, respectively. amounts transferred from consumers and tax- In short, the subsidy programs prominent in payers for support purposes (OECD 2002e). current food and agriculture policy are not tar- In the case of market price support and defi- geted to keeping small, struggling family farms ciency payments, the share is one-fourth or in business but instead provide hefty rents less; for input subsidies, less than one-fifth. to large farmers. Nor are current production- Only one-quarter of every dollar of producer based policies effective in achieving their vari- support actually finds its way into the pro- ous other objectives (such as environmental ducer’s pocket—the rest goes to input suppliers sustainability and rural development). By in- and owners of other factors of production creasing land prices they also lead to the (OECD 1999, De Gorter 2003). The most im- creation of larger farms and the elimination of portant outcome of these programs is that they small family farms. Meanwhile, their unin- lead to much higher land prices. tended spillover effects on global markets, and The largest farm operations, which gen- on other countries, are large and negative. erally are also the most profitable and the At the most general level, it is probable that wealthiest, receive most of the benefits of sup- agricultural protection in rich countries wors- port systems. In the United States, the largest ens global income distribution. First, farmers 25 percent of farms have average gross farm re- in the North earn more on average than their ceipts of more than $275,000 and average farm own national averages. Second, the lion’s share net worth of more than $780,000. They receive of farm aid goes to the largest and wealthiest 89 percent of all support—in part because they farmers. At the other end of the global distribu- produce a similar share of output. The remain- tion spectrum, more of the poor tend to live in ing 1.6 million U.S. farms on average receive rural areas, and protection in rich countries little support. Through the lens of household tends to depress prices and demand for their income surveys, the story is similar: At one ex- goods. treme, farm households with an average in- come of $275,000 received payments averaging International markets are important to $32,000. At the other end of the spectrum, sustained income growth in developing farm households with incomes averaging countries $13,000 received $2,200 in program payments. When subsidies depress prices the impacts in In the European Union, where farm num- poor countries can be severe. To illustrate the bers and structures differ somewhat, the distri- impact of commodity price changes, Minot bution of support is not markedly different. and Daniels (2002) used household income The largest 25 percent of farms have average data to estimate the potential impact of cotton gross farm receipts of more than €180,000 and price declines in Benin and tobacco price de- average farm net worth of almost €500,000. clines in Malawi, the major export crops of They produce 73 percent of farm output and those two countries. Cotton prices have de- receive 70 percent of support. Farms of the clined by almost 40 percent over the last few next largest size have much smaller gross farm years. In Benin, a poor country, the impact of receipts, averaging just over €43,000, and av- this decline in world cotton prices, if it were erage farm net worth of about €230,000. They fully passed on to farmers, would reduce over- produce 17 percent of output and receive 19 all welfare in rural areas by 6–7 percent and percent of support payments. The remaining 2 that of cotton farmers by about 19 percent. million EU farms produce little, receive little The richest quintile of households, meanwhile, support, but have a sizeable average farm net would experience a decline in income of 4 per- 108 A G R I C U L T U R A L P O L I C I E S A N D T R A D E cent. Thus this price change alone would in- crease the poverty rate in Benin by up to 8 per- Figure 3.1 Countries that produce more centage points (depending on the simulations), cash crops also produce more food from 40 percent to 48 percent. Annual growth rates of food and cash crop production in Tobacco constitutes about 80 percent of 25 countries having agricultural output equal to at least 15 percent of GNP, 1980–2001 (percent) Malawi’s exports. A 30 percent decrease in world tobacco prices over the last few years .06 nga has reduced the income of small growers by an gha ben average of 8 percent. The poorest quintile has pak lost about 13 percent, the richest 7 percent. bfa .04 For a typical farmer, the annual net returns civ chd npl ida from tobacco, the country’s most profitable ind tgo car crop, declined from $108 to $26 (Integrated sen cam mlw bng myr ngr kny mli Framework 2003). These rough estimates un- .02 zmb derstate the overall impact of the price de- mdg tnz zmb clines, however, because cash incomes allow farmers to purchase inputs, such as fertilizer gmb and pesticides, that increase the yields for their 0 subsistence crops and have a significant impact –.05 0 .05 .1 .15 on their levels of poverty and malnutrition. Cash crop production growth rate The importance of the global market goes beyond price changes. For countries with a rel- atively small urban population, agricultural ex- ports can produce faster growth than can do- mestic market demand—however fast domestic Trade and export growth demand might be growing. In such cases, the in agriculture international market provides growth opportu- nities without the constraint of sharply lower prices, which often accompany an increase in T he last two decades were periods of very rapid growth in exports from developing countries to other developing countries and to agricultural production. Although food pro- the industrialized world (table 3.5). Growth in duction for home consumption and the domes- the world economy accounts for some of this tic market accounts for most agricultural pro- export growth, but lower trade barriers, im- duction in the developing world, agricultural proved supply capabilities, and increases in exports and domestic food production are specialization are more important. The rapid closely related. Export growth contributes sig- growth in exports was true both in manufac- nificantly to the growth of nonexport agricul- turing, where levels of protection have been ture by providing cash income that can be used reduced significantly, and in agriculture, to modernize farming practices. For those leav- where significant protection remains. Never- ing the farm, growth and modernization of theless, manufacturing export growth rates agriculture create jobs in agricultural process- were much higher. ing and marketing. On balance, cash-crop income complements Agricultural trade makes up a growing and enhances food production, particularly in share of trade among developing poorer countries where opportunities to earn countries, but agricultural export shares nonfarm income are more limited (figure 3.1) to rich countries are stable (Watkins 2003; Von Braun and Kennedy 1994; Although developing countries’ exports accel- Minot and others 2000). erated during the 1990s, agricultural exports 109 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Table 3.5 Manufacturing exports grew much faster than agricultural exports Export growth rates (percent) Developing countries’ export growth rates World export growth rates Total Developing to developing Developing to industrialized 1980–81 to 1990–91 to 1980–81 to 1990–91 to 1980–81 to 1990–91 to 1980–81 to 1990–91 to 1990–91 2000–01 1990–91 2000–01 1990–91 2000–01 1990–91 2000–01 Agriculture 4.3 3.6 3.4 4.8 3.6 7.8 3.4 3.3 Manufacturing 5.9 4.8 7.6 8.9 7.3 10.0 7.8 8.3 Note: Manufacturing exports are deflated by the U.S. purchasing parity index (PPI) for finished goods less food and energy. Agri- culture exports are deflated by the U.S. PPI for farm products. Source: COMTRADE. Table 3.6 South-South exports in agriculture are rising as South-North export shares fall Share of global agricultural and manufacturing exports by source and destination, 1980–2001 (percent) Developing countries Industrialized countries 1980–81 1990–91 2000–01 1980–81 1990–91 2000–01 Agriculture exports 35.9 32.9 36.9 64.1 67.1 63.1 To developing 9.9 9.2 13.7 15.3 11.9 14.7 To industrialized 26.0 23.7 23.2 48.8 55.3 48.4 Manufacturing exports 19.3 22.7 33.4 80.7 77.3 66.6 To developing 6.6 7.5 12.3 21.7 15.2 19.0 To industrialized 12.7 15.2 21.1 59.0 62.1 47.6 Source: COMTRADE. did not keep pace with manufactured exports, dustrial countries, from 5.4 percent annually largely because agricultural export growth ac- during the 1980s to 1.9 percent in the 1990s, celerated only to the other developing coun- was offset by the increase in import growth in tries (table 3.6).5 developing countries, which increased from 3 Developing countries increased their share percent annually to 6 percent. of global manufacturing exports from 19 per- cent in 1980–81 to 33 percent in 2000–01. Ex- Product trends differ panding trade among developing countries What accounts for the shift in markets for the contributed to the gain in share, but higher agricultural exports of developing countries? exports to industrial countries also played a Price changes alone do not appear to explain significant part. In agriculture, by contrast, the it (box 3.2). Static markets in industrial coun- developing countries maintained, but did not tries for traditional developing-country prod- expand, their one-third share of world agri- ucts such as coffee and tea probably contri- cultural trade over the last two decades. The buted to declining import growth rates, as did steady decline in the developing countries’ share the decline in GDP growth rates, combined of agricultural exports to industrial countries with low elasticity of demand.6 over the period was counterbalanced by an in- To explore the phenomenon further, we sep- crease in their share of exports to other devel- arated agricultural exports into four sub- oping countries. In other words, the significant groups. The first consists of mostly tropical, deceleration of nominal import growth in in- developing-country products such as coffee, 110 A G R I C U L T U R A L P O L I C I E S A N D T R A D E Box 3.2 Did agricultural exports slow down solely because of falling prices? I n nominal terms, export growth in agricultural products decelerated significantly during the 1990s. Can the slowdown be attributed to the price the authors tried several alternatives to compensate for these limitations. The unit-value indices from trade data gave inconsistent results and were elimi- declines observed in the late 1990s? The existing nated, leaving three series, one from the U.S. pur- price series for agricultural commodities have certain chasing parity index (PPI) series for farm products, limitations. Most of the standard series are based on which includes all products, and two from raw com- raw commodities that constitute a much smaller per- modity indices. One of the latter uses world trade centage of the global trade flows. In most cases they weights; the other, developing-country export exclude seafood, fruits, and vegetables—now the weights. The behavior of the three indices over the largest trade items. For the purposes of this chapter last two decades is shown in the table below. 1980–81 to 1990–91 1990–91 to 2000–01 U.S. farm products PPI 4.7 –6.8 Raw commodities (world trade weights) –8.3 –6.6 Raw commodities (developing countries’ weights) –22.7 –15.2 If the U.S. PPI is used, a small fraction of the greater in the 1980s, and if they are used to deflate nominal changes in trade flows in the 1990s can be the nominal exports, the deceleration would be ac- attributed to price declines in the 1990s. Raw com- centuated. For that reason, the U.S. food products modity indices show that the price declines were PPI was used to deflate aggregate exports. cocoa, tea, nuts, spices, textile fibers, and sugar gains in export growth. When growth in ex- and confectionary products. The second is ports of manufactures (including processed made up of temperate products highly pro- food) to industrial countries is decomposed be- tected in industrial countries—meats, milk and tween demand and market share, only 21 per- products, grains, animal feed, and edible oil cent of developing countries’ export growth and oilseeds. The third category is the dynamic appears to have been caused by demand in- nontraditional products: seafood, fruits, veg- creases. The other 79 percent was caused by etables, and cut flowers. The last category in- changes in market share (box 3.3). Limited cludes other processed agricultural products, raw-commodity information collected by such as tobacco and cigarettes, beverages, and OECD does not show any significant change in other processed foods. import-penetration ratios in OECD countries Import growth rates in industrial countries over the last decade (OECD 2001). Mean- declined across all groups, while the opposite oc- while, the developing countries gained market curred in developing countries (figure 3.2). But share in every manufacturing subsector—ex- changes in demand are only part of the picture. cept food processing. The protection rates for In attributing causes to differential growth food processing in industrial countries are ex- rates, it is important to consider the relative tremely high—far above those of any other roles of demand growth and market-share manufacturing subsector. 111 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Figure 3.2 Import growth rates of nontraditional export commodities decreased in industrial countries but increased in developing countries a. Industrial countries b. Developing countries Import growth rates (nominal USD, percent per annum) Import growth rates (nominal USD, percent per annum) 10 10 1991–2001 9 9 ᮡ 8 8 7 7 1981–1991 1991–2001 6 6 ᮡ 1981–1991 5 ᮡ 5 ᮡ 4 4 3 3 2 2 1 1 0 0 Tropical Temperate Seafood, Other Total Tropical Temperate Seafood, Other Total products products fruits, processed products products fruits, processed vegetables, products vegetables, products and flowers and flowers Source: COMTRADE. The evolving structure of trade: toward ing countries also have marginally expanded nontraditional products with lower rates their exports of temperate products (grains, of protection meats, and milk)—but mostly to other devel- World trade has moved away from traditional oping rather than industrial countries. These export commodities to other categories of important developments will require changes goods. This is true of both developing and in- in how developing countries’ agricultural trade dustrial countries. The product groups that is conceived and analyzed (figure 3.3). gained significantly between 1980–81 and Their trade gains have brought more devel- 2000–01 are fruits, vegetables, and cut flowers oping countries up against rising food safety (19 percent); fish and seafood (12.4 percent); standards in the developed world. Meeting and alcoholic and nonalcoholic drinks (8.7 such standards has a cost—not just in compli- percent). Although products in these categories ance, but also in documenting that compliance. tend to have high income elasticities, they also This cost can be repaid in the form of higher enjoy lower rates of protection in industrial trade. Various mechanisms exist to help devel- and large developing countries. Product groups oping countries rise to the standards (box 3.4). that showed significant declines during the pe- Industrial-country export structures also riod were grains (14.3 to 9.5 percent); coffee, have changed. Exports of protected products cocoa, and tea; sugar and sugar products; and have declined, whereas those of beverages, textile fibers—all of which are among the tra- fruits, and vegetables have grown. These ditional exports of developing countries. The changes are discernible despite the fact that declines were caused by a combination of price intra-EU trade is included in the global export declines, low demand elasticities, and—in the data. One cause of the change is that greater case of sugar, grains, meats, and milk—high domestic production of protected products rates of protection and expanded production has made many industrial countries more self- in industrial countries. sufficient in those products, reducing trade. While moving away from traditional ex- As a group, developing countries lost ex- ports and into expanding subsectors, develop- port market share during the 1980s, but 112 A G R I C U L T U R A L P O L I C I E S A N D T R A D E Box 3.3 Decomposing export growth in manufacturing Gross production data in the three non-U.S. coun- M ost market-share analysis has not looked into the shares of exports from developing coun- tries in the consumption of industrial countries. tries have been converted to U.S. dollars at current exchange rates. Because the U.S. dollar appreciated Below are estimates of developing-country exports in significantly against the currencies of the other three the domestic consumption and production of countries in the late 1990s, this conversion underesti- Canada, Germany, Japan, and the United States, mates domestic production and demand growth. It which together absorb about 70 percent of develop- also overestimates the share of imports, which are ing countries’ manufactured exports to industrial denominated in U.S. dollars. countries. Demand change is estimated assuming a constant The table below shows the shares of exports share of exports in domestic demand between the two from developing countries in the four countries’ time periods; that is, market shares do not change. total absorption (demand) and the growth of exports The market share changes are then estimated as the from developing countries. Absorption is estimated difference between the actual export growth and the as gross production minus exports, plus imports. export growth under a constant market share. Developing countries increased their share of industrial countries’ manufacturing imports— largely by increasing their market share, 1991–99 (percent) Share of developing countries’ Export growth due to exports in domestic demand Growth in exports Change in Change in 1991 1999 from developing countries demand market share Canada 4.51 7.64 117.25 28.16 89.08 Japan 2.24 4.38 95.04 –0.25 95.29 United States 5.10 9.04 169.42 51.99 117.43 Germany 7.44 8.91 18.31 –1.22 19.53 Total 4.46 7.63 110.90 23.38 87.52 Sources: UNIDO, COMTRADE. Using UNIDO and COMTRADE data, UNCTAD estimated these ratios until 1995. UNIDO’s coverage in terms of gross production has become more limited since 1995. The relationship between domestic demand The same conclusion holds true for the 15 growth in industrial countries and export growth three-digit ISIC subsectors that range from very cap- from developing countries is relatively weak. Market ital intensive (rubber and glass) to very labor inten- share gains caused by the restructuring of global sive (garments and footwear). production are a much more powerful factor. The only subsector in which demand growth Between 1991 and 1999, exports of manufac- was greater than the market share gains, and in tures from developing countries to these four coun- which the developing countries lost market share, tries increased by about 139 percent, compared to was food processing. In that subsector, the market about 60 percent for world trade, while the total share of developing countries declined from 2.42 increase in domestic demand was only 29 percent. percent in 1991 to 2.40 percent in 1999. Why? The rest of the export growth was a result of the Food processing enjoyed the greatest protection of increases in market shares of developing country any subsector, and protection did not decline over exports in industrial-country markets. A change of the last decade. Because a large portion of agricul- one percentage point in absorption shares during tural exports are classified under food processing, the decade would increase exports from developing protection of the subsector explains part of the de- countries by approximately 28 percentage points, celeration of agricultural exports from developing to equal to the total absorption growth over the decade. industrial countries during the 1990s. Source: Aksoy, Ersel, and Sivri (2003). 113 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Figure 3.3 Developing countries’ exports of nontraditional products have surged, but industrial countries’ exports have changed little a. Developing countries b. Industrial countries Percent of developing country exports Percent of developing country exports 60 60 50 50 1981 ᮡ 40 40 1991 30 ᮡ 30 2001 1991 ᮡ 20 20 1981 2001 ᮡ ᮡ ᮡ 10 10 0 0 Tropical Temperate Seafood, Other Tropical Temperate Seafood, Other products products fruits, processed products products fruits, processed vegetables, products vegetables, products and flowers and flowers Source: COMTRADE. reversed that trend in the 1990s (table 3.7). Global agricultural protection: Modest expansion in the 1990s brought them The bias against development back to where they had been in the early 1980s. Global gains were made by middle- and Progress in the Uruguay Round was more low-income countries, mostly to other devel- formal than real oping countries. China is an exception to this Since the 1980s, two important developments trend, having increased its export shares in all have occurred in agricultural trade policy. markets. Even in the 1990s low-income coun- First, most developing and a few industrial tries continued to lose market share in their countries have made major reforms in their exports to industrial countries, making up protection regimes involving unilateral and re- the loss by expanding their export shares in gional reductions in tariffs and quotas. For developing-country markets. In tropical prod- example, unilateral reforms in the 1990s ef- ucts, where global shares declined, low-income fectively eliminated export taxation in most countries increased their shares to the other developing countries. Average tariffs have developing countries. declined rapidly, while other import restric- The LDCs lost export market share in both tions, such as foreign exchange allocations for markets during both decades. Unlike other de- imports, have effectively disappeared (World veloping countries, they have not been able to Bank 2001). Manufacturing tariffs dropped make up their market-share losses in tropical more than agricultural tariffs. In at least one products by expanding their shares in the way, agricultural protection expanded: Many growing subsectors: seafood and fruits and middle-income countries began subsidizing vegetables. Their only gains have come in sea- their agricultural products. food, and much of the expansion has come Second, the Uruguay Round Agreement from industrial-country vessels fishing in their on Agriculture brought agricultural trade into waters. In highly protected products, South- WTO disciplines. Before Uruguay, agricultural South trade has expanded, possibly as a result products had no bound tariffs, and tariffs often of regional trading arrangements. were supplemented by nontariff measures such 114 A G R I C U L T U R A L P O L I C I E S A N D T R A D E Box 3.4 Food safety standards: From barriers to opportunities A gricultural trade is shifting toward high-value, perishable commodities such as fresh fruits, veg- etables, meats, and fish. With this change have come Nearly all of the cases of allegedly protectionist use of food safety measures brought before the WTO have involved trade between developed countries consumer concerns over food safety. In response, over issues such as hormone residues in meat and ge- governments and private companies have developed netically modified foods. Although some food-import a growing array of rules, regulations, and standards. bans have been heavily publicized, their application Some fear that these standards will be used by high- against developing countries is quite rare and typi- income countries as a tool of trade protection. cally has involved complementary rather than com- Some developing countries have risen to the petitive products. However, some evidence suggests higher standards. Kenya’s exporters send fresh veg- that developing countries employ safety regulations etables and salad greens by air freight to major as a protectionist measure against other developing European supermarket chains. In that industry, food countries. safety standards have accelerated the adoption of The available evidence suggests that most food- modern supply-management techniques and stimu- safety-related problems that developing-country ex- lated public-private collaboration (Jaffee 2003). porters encounter are well within their capacities to Many developing-country suppliers, however, will resolve. According to data from the U.S. Food and not be able to meet the more stringent standards Drug Administration, most detentions of developing- without technical advice, upgraded production and country food products involve labeling violations or processing facilities, better enforcement of standards, very basic problems of food hygiene—and thus of and closer working relationships with importers in quality assurance (see table). No firm can operate high-income countries. long without addressing such problems. Detentions by U.S. Food and Drug Administration of imports from developing countries 1997 and 2001 (percent) Latin America and Asia India Reasons for contravention the Caribbean 1996–97 1996–97 2001 Food additives 1.4 7.4 7.4 Pesticide residues 20.6 0.4 1.9 Heavy metals 10.7 1.5 0.6 Mold 11.9 0.8 0.4 Microbiological contamination 6.2 15.5 15.3 Decomposition 5.2 11.5 0.3 Filth 31.4 35.2 26.4 Low acid canned food 3.6 14.3 4.1 Labeling 5.0 10.8 15.7 Other 1.7 2.6 27.8 Total 100.0 100.0 100.0 Total number 3,985 5,784 2,148 Source: USFDA. Even for more complex food safety issues, de- long demanded by overseas customers and con- veloping countries have room to maneuver. An array sumers are now being demanded by discerning of strategies exists to help them meet product and domestic consumers as well. They are well within process standards for international markets. Espe- producers’ reach. cially in middle-income countries, the good manu- The European Union lays down harmonized facturing practices and good agricultural practices hygiene requirements governing the catching, pro- (Continues on next page) 115 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Box 3.4 (continued) cessing, transportation, and storage of fish and fish- cess to credit and vaccinated their animals to ensure a ery products. Processing facilities must be inspected supply of healthy milk. The camel cheese, after trans- and approved by a specified authority in the country port and production costs, was priced at $10 per of origin. Countries whose local requirements have kilogram in the EU. After winning a prize at a trade been found by the Commission to be at least as strin- fair, the cheese soon found its way into elite stores gent as those in the European Union and for which like Harrods in London and Fauchon in Paris. How- specific import requirements have been established ever, it proved to be difficult to find the correct tar- are placed on “List I” and enjoy reduced physical in- iff line for the product, and grouping it with “other spection at the border. dairy, cheese” exposed it to a much higher tariff than Between 1997 and October 2002, the number of regular cheese. To make matters worse, the EU soon countries achieving List 1 status increased from 27 to decided to abolish imports of camel cheese from 72. More than half are low-income or lower middle- Mauritania, arguing that the presence of “hoof and income countries; half of these are low-income mouth” disease in Mauritania could be transmitted African countries. Another 35 countries are on List from camels to other livestock, even though there is II, including the United States (Henson and Mitullah no real evidence that camels are capable of spreading 2003). the disease. The EU then imposed another restriction: Food safety compliance costs can include the camel cheese could indeed be imported—but only if cost of adjusting production and processing facilities; mechanical methods were used to obtain milk used the recurrent costs to implement food safety manage- in its production—an unworkable proposal for the ment systems; and the costs of certification, monitor- low-income milk producers who were located miles ing, and enforcement. Relatively few estimates are away from major ports. Mauritania did not dispute available on the magnitude of these costs. When a this case at the WTO because of the sheer costs in- country is already exporting high-valued foods, com- volved—costs that were not justified for exports of pliance may require only incremental production $3 million to $5 million worth of cheese per year. changes and public-sector oversight. However, for Catfish producers in Vietnam have had similar other suppliers the costs of reaching internationally difficulties accessing the American market, initially competitive levels may be high. The Bangladeshi because of labeling rulings (and then later because shrimp industry invested an estimated $18 million of anti-dumping judgments; see box in Chapter 2). in the latter half of the 1990s to upgrade fish- The emerging set of international and developed- processing facilities and product-testing laboratories, country food safety standards present challenges for and to make other changes in response to repeated many exporters in developing countries. Concerted ef- quality and safety detentions on exports to the Euro- forts to address basic hygiene and quality-assurance pean Union and the United States. However, these requirements and to provide relatively simple train- expenditures have been rewarded with rapidly in- ing for farmers could go a long way in ensuring com- creasing (and better priced) shrimp exports—which pliance with most official food safety standards. In totaled $296 million in 2000 (Cato and others circumstances where compliance requires greater in- 2000). vestment—both by the public and private sectors— Standards can also be a barrier to trade. Con- partnerships between developed and developing coun- sider the case of camel milk cheese exports to the EU. tries and among developing countries may fill the bill. Tiviski SARL, a dairy processor in Nouakchott, Mau- Beyond this, the public has to remain vigilant that ritania, developed a technology to produce “pate standards do not become misused as instruments of molle” cheese from camel milk. It obtained the milk protection. from nomad milk producers who were very poor. In return, Tiviski provided the producers with cheap ac- Source: World Bank staff. 116 A G R I C U L T U R A L P O L I C I E S A N D T R A D E Table 3.7 Developing countries have shared unequally in export market gains Export shares of food and agricultural products by income level (as percentage of total world trade) Exports to industrial countries Exports to developing countries Total exports Income level 1980–81 1990–91 2000–01 1980–81 1990–91 2000–01 1980–81 1990–91 2000–01 Industrial 48.8 55.3 48.4 15.3 11.9 14.7 64.1 67.1 63.1 Middle-income* 19.6 18.4 17.0 7.3 6.4 9.8 26.9 24.8 26.8 Low-income 5.2 3.4 3.4 1.4 1.3 2.0 6.5 4.8 5.4 of which LDCs 1.6 0.8 0.7 0.7 0.4 0.5 2.3 1.3 1.1 China 0.7 1.3 2.1 0.9 1.2 1.4 1.7 2.5 3.5 India 0.5 0.5 0.6 0.3 0.3 0.5 0.8 0.8 1.1 Total 74.9 78.9 71.6 25.1 21.1 28.4 100.0 100.0 100.0 * Excluding India and China. Source: COMTRADE. as import quotas or bans, quantitative restric- 50 percent reduction by dropping a tariff from tions, variable levies, and monopoly purchas- 2 percent to 1 percent—or in areas of low sen- ing by state-owned or other companies. Import sitivity, while making only the minimum re- barriers were coupled with the widespread use duction in sensitive product areas. The Round of production-related subsidies, such as price offered limited opportunities to make mini- supports, which often led (and still leads) to mum import commitments for certain prod- increases in production above the level of mar- ucts instead of adopting tariffs on them. The ket equilibrium. Excess production had to be minimum import option was taken by Japan, stockpiled or exported, sometimes with the Korea, and the Philippines for rice, and by Is- help of further subsidies. With the intention of rael for certain sheep and dairy products. (Ja- aligning agricultural trade rules with those ap- pan has since tariffied rice imports.) plying to trade in other goods, the Uruguay Once a tariff was established, bindings and Round negotiators agreed that all import barri- reductions were negotiated. In cases where ers, other than those in place for health and tariffs were high, or where quotas had been al- safety reasons, should take the form of transpar- lowed in some imports, minimum and current ent tariffs. Before agreeing on tariff reductions, market-access opportunities were also negoti- all border measures had to be converted into ated. The typical result was the establishment their tariff equivalents—a process known as of a minimal tariff rate for a limited volume of “tariffication.” imports—called a tariff rate quota (TRQ). The conversion of nontariff measures into With the removal of nontariff measures, tariffs was generally done using the price-gap some countries worried that they would not be method—the gap being the difference between able to prevent surges in import volumes or domestic and world market prices. After es- falling import prices. To allay these concerns, tablishing the tariff equivalent of an import re- negotiators agreed that a special agricultural striction, reductions were applied from bound safeguard could be applied to certain products. tariffs. Developed countries reduced their tar- The Uruguay Round yielded no meaningful iffs by an average of 36 percent and a mini- reduction in protection in industrial countries. mum of 15 percent over six years; developing In many cases, in fact, protection may have in- countries by an average of 20 percent and a creased as a result of so-called dirty tariffica- minimum of 10 percent over ten years. The tion (Nogues 2002, Ingco 1997). Continued agreed reductions were simple averages, not protection has led to greater import substitu- weighted for the volume of trade, so some tion, while the geographical restructuring of countries made large reductions in tariffs that production that occurred in manufacturing were already low—for example, achieving a did not occur—at least not to the same de- 117 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 gree—in agriculture. Review of the experience ased against developing countries—and to date with the new rules on market access, against the world’s poor. export subsidies, and domestic support indi- cates that the effects of implementation of the Import barriers are the most important Uruguay Round Agricultural Agreement have instrument of protection been modest. The reasons include weaknesses Although the conversion of nontariff barri- in specific aspects of the agreement, such as ers to tariffs during the Uruguay Round was high baseline support levels from which re- an important step forward, average agricul- ductions were made. In some countries, in- tural tariffs in most industrial and developing cluding the United States, reforms undertaken countries were and remain much higher than before the negotiations were adequate to ful- tariffs for nonagricultural products. fill the new rules on reducing domestic sup- This section evaluates the agricultural trade port (OECD 2001). regimes of the Quad countries (Canada, Euro- Today, protection in agriculture takes dif- pean Union, Japan, United States) and 25 ferent forms—tariff protection, subsidies, tar- developing countries in light of the Uruguay iff peaks, TRQs, tariff escalation, and opaque Round’s objectives. Eight of the developing tariffs. In reviewing these forms, the following countries in the sample are large middle- section makes two fundamental points: income countries with significant agricultural sectors. Eight more middle-income countries • First, the various forms of protection are are included to ensure regional balance. Eight often linked. For example, goods pro- lower-income countries round out the sample. duced behind high tariff walls and with Emphasis has been placed on the nature of production subsidies often require export tariffs because a key objective of the Uruguay subsidies to be sold in the world market. Round was to lower tariffs and make them That said, border barriers are more im- more transparent. portant than subsidies. The tariff data in table 3.8 underestimate • Second, virtually the entire interlinked actual border protection. First, specific duties, system of protection, even when used by which generally are higher than ad valorem other developing countries, is heavily bi- rates, are not fully reflected in the simple av- Table 3.8 Agricultural tariffs are higher than manufacturing tariffs in both rich and poor countries Most-favored-nation, applied, ad valorem, out-of-quota duties (percent) Percentage of lines Agriculture Manufacturing covered in agriculture Quad countries 10.7 4.0 86.7 Canada (2001) 3.8 3.6 76.0 European Union (1999) 19.0 4.2 85.9 Japan (2001) 10.3 3.7 85.5 United States (2001) 9.5 4.6 99.3 Large middle-income countries a 26.6 13.1 91.3 Other middle-income countries b 35.4 12.7 97.7 Lower-income countries c 16.6 13.2 99.8 a. Brazil (2001), China (2001), India (2000), Korea (2001), Mexico (2001), Russian Federation (2001), South Africa (2001), and Turkey (2001). b. Bulgaria (2001), Costa Rica (2001), Hungary (2001), Jordan (2000), Malaysia (2001), Morocco (1997), Philippines (2001), and Romania (1999). c. Bangladesh (1999), Guatemala (1999), Indonesia (1999), Kenya (2001), Malawi (2000), Togo (2001), Uganda (2001), and Zimbabwe (2001). Source: WTO Integrated Database. 118 A G R I C U L T U R A L P O L I C I E S A N D T R A D E erages. Second, many products are subject to nontariff restrictions. Figure 3.4 Developing countries lowered Because ad valorem equivalents of specific tariffs on manufactured products more than on agricultural products and other duties, where available, are much Average applied tariffs for agricultural and manufacturing higher than the ad valorem rates, and assum- products in developing countries, 1990, 1995, and 2000 ing that the same tariff structure applies to (percent) Canada and Japan, which use non–ad valorem 35 (NAV) rates on 25 percent and 15 percent of 1990 30 their tariff lines, the average tariffs for the two ᮡ countries are seriously underestimated, lower- 25 ing the Quad average. To show the degree of 1995 bias, the third column in tables 3.8 and 3.9 20 ᮡ shows the proportion of tariff lines to which 2000 15 the averages apply.7 ᮡ Excluding Canada, which has a large pro- 10 portion of agricultural NAV tariffs without equivalents, average tariffs in agriculture are 5 much higher than in manufacturing. The differ- 0 ence is especially pronounced in the European Agricultural Manufacturing products products Union—19 percent in agriculture versus only 4.2 percent in manufacturing. Among the devel- Source: TRAINS. oping countries, the results are very similar, with a few exceptions, such as Brazil and Malaysia, where manufacturing tariffs are higher. the lowest. Again, the average tariffs of coun- The developing countries in the sample have tries that have a high percentage of NAV lines higher tariffs than the industrial countries, the (Bulgaria, Russian Federation, South Africa, highest being Morocco (64 percent), Korea (42 and Turkey) are seriously underestimated. percent), and Turkey (49.5 percent). Indonesia Tariffs are widely dispersed and have very (8.5 percent) and Malaysia (2.8 percent) have high peaks. Industrial-country tariffs, although lower on average than those of developing countries, show significant tariff peaks, indi- Table 3.9 Agricultural tariffs: High peaks cating high protection for specific products. and deep valleys The peaks reach almost 1,000 percent in the Tariff peaks and variance in selected countries; MFN, out of quota, applied duties (percent and standard deviation) Republic of Korea, 506 percent in the Euro- pean Union, and 350 percent in the United Percentage Average Maximum Standard of lines States.8 Tariffs in many low-income countries tariff tariff deviation covered have lower peaks and show less variance than those in many of the middle-income countries. Canada 3.8 238.0 12.9 76.0 European Union 19.0 506.3 27.3 85.9 Compared to the slow reform in OECD Japan 10.3 50.0 10.0 85.5 countries, the changes in protection in devel- United States 9.5 350.0 26.2 99.3 oping countries were significant in the 1990s Korea, Rep. of 42.2 917.0 119.2 98.0 Brazil 12.4 55.0 5.9 100.0 (figure 3.4). The average agricultural tariff Costa Rica 13.2 154.0 17.4 100.0 declined from almost 30 percent in 1990 to Indonesia 8.5 170.0 24.1 100.0 about 18 percent in 2000, a decline of 35 per- Malawi 15.3 25.0 9.1 100.0 Morocco 63.9 376.5 68.2 100.0 cent. (The rates shown in the figure are simple Togo 14.7 20.0 6.5 99.9 averages of the average tariffs of about 50 Uganda 12.9 15.0 3.7 100.0 developing countries.) Those reductions were Source: WTO Integrated Database. complemented by the elimination of most ex- 119 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 port taxes as well as import licensing and prices in domestic markets, account for many other quantitative restrictions (World about 70 percent of total protection in Bank 2001). Average tariffs in agriculture re- the OECD countries. main much higher than those in manufactur- • Production-related subsidies given to ing, however, indicating that the general ten- farmers under different schemes, called dency in the 1980s—to protect the industrial “direct support,” usually take the form sector—no longer holds. In their study of 15 of direct budget transfers. developing countries, Jensen, Robinson, and • General support for agriculture—through Tarp (2002) concluded that the bias against research, training, marketing, and infra- agriculture in the 1980s no longer exists. The structure programs—usually is not included economy-wide system of indirect taxes, in- in the estimates of producer supports. cluding tariffs and export taxes, significantly discriminated against agriculture in only one In addition, many countries have subsidies country. It was largely neutral in five, pro- for their consumers, but generally these do not vided a moderate subsidy to agriculture in affect production and thus are not included in producer-support estimates. four, and strongly favored agriculture in five. The support accorded to OECD-country Subsidies underpin the system producers through higher domestic prices and of border protection direct production subsidies was $248 billion in An extensive network of subsidies has evolved 1999–2001 (table 3.10). Some two-thirds of to support agriculture, particularly in the rich the total—$160 billion—came from the bor- countries. Protection takes three major forms. der barriers described above or from market price support mechanisms. The remainder • Border barriers such as tariffs and quan- came in the form of direct subsidies to farmers. titative restrictions, designed to support Another $80 billion in subsidies came from Table 3.10 Most subsidies go to producers—and come from border protection Agricultural support in the OECD countries, 1999–2001 (billions of dollars) European Union Other United European Emerging accession OECD Total States Union Japan supportersa countriesb countries OECD Where total support goes Consumers 21.4 3.8 0.1 0.7 0.0 0.2 26.2 General services 22.8 9.6 12.7 7.1 0.6 2.3 55.1 Producers 51.3 99.3 52.0 30.4 3.0 12.3 248.3 Total 95.5 112.7 64.8 38.2 3.6 14.9 329.6 Where producer support goes Corn 8.3 2.7 Nc 1.7 –0.1 0.2 12.9 Meat c 2.6 34.0 4.1 3.4 0.5 2.8 47.3 Milk 12.4 16.7 4.9 2.7 0.7 4.7 42.1 Rice 0.7 0.2 18.0 7.6 Nc –0.2 26.4 Wheat 4.9 9.5 0.8 0.9 0.3 0.9 17.3 Other 22.3 36.2 24.1 14.1 1.9 3.6 102.2 Where producer support comes from Domestic measures d 32.6 38.5 5.0 4.4 1.4 6.3 88.2 Border measures e 18.7 60.9 47.0 26.0 2.0 5.7 160.1 a. Includes Korea, Turkey, and Mexico. b. Includes Czech Republic, Hungary, Poland, and Slovak Republic. c. Beef and pork. d. Direct payments to producers. e. Tariffs and tariff equivalents of other border measures. Sources: OECD (2002) and authors’ calculations. 120 A G R I C U L T U R A L P O L I C I E S A N D T R A D E Table 3.11 Subsidies account for a large share of farmers’ revenues Percentage of farm-gate prices attributable to border protection and direct subsidies, 1986–2001 Market price support (border protection)a Direct subsidiesa Total producer support (estimate)a Area 1986–88 1995–97 1999–2001 1986–88 1995–97 1999–2001 1986–88 1995–97 1999–2001 OECD 48.2 28.2 31.3 14.3 13.3 17.2 62.5 41.5 48.5 European Union 65.3 28.3 34.3 10.5 20.4 21.7 75.8 48.8 56.0 Japan 145.4 131.7 138.1 16.8 13.0 14.7 162.1 144.7 152.9 United States 16.0 7.5 10.8 18.3 7.4 18.8 34.3 14.9 29.6 Eastern Europe 45.2 8.7 10.4 18.3 4.8 7.5 63.6 13.5 17.9 Australia and New Zealand 4.2 2.8 0.6 6.4 3.9 3.4 10.6 6.8 4.0 Other countries 53.1 42.6 46.3 11.1 12.8 12.2 64.2 55.4 58.5 Other industrial b 165.9 108.1 113.0 72.2 81.9 106.7 238.1 190.0 219.7 Other developing c 31.4 38.1 42.9 6.4 8.0 7.3 37.8 46.1 50.2 a. The denominator is total value of production at farm gate less market price support (both estimated at world prices). b. Includes Norway, Switzerland, and Iceland. c. Includes Korea, Turkey, and Mexico. Source: OECD. programs (such as food stamps) that directly ments into the global market and accentuating benefit consumers ($26 billion) and from gen- price drops. The countercyclical movement eral services to agriculture ($55 billion), such of protection reflects the specific duties and as public investments in agricultural research TRQs that are triggered when prices fall. and extension. The European Union and United States Of the subsidies, the share linked to income have reduced their overall levels of agricultural rather than production (known as “partially support. For example, in the European Union decoupled subsidies”) increased from approxi- farmers’ prices were 65 percent higher than mately 9 percent of total protection in 1986–88 international prices in 1986–88; this ratio to 21 percent in 2001. Major products that ac- decreased to 34 percent in 1999–01. During count for the bulk of support are grains, meats, the same period, however, direct production- milk, and sugar. related payments to farmers increased from Protection rates for producers in the OECD 10.5 percent to 21.7 percent, partially com- decreased from 62.5 percent in 1986–88 to 49 pensating for the decline in border barriers. percent in 1999–01, measured as a percentage Similarly, in the United States, domestic prices, of gross agricultural output at world prices. The relative to international prices, declined from contribution of border barriers to total protec- 16 percent to 10.8 percent. tion fell from 77 percent in 1986–88 to about Aggregate support levels vary significantly 65 percent in 1999–01. After decreasing rapidly among the OECD countries. Some (Iceland, from 1986, overall protection rose again after Norway, and Switzerland) have very high lev- 1997 in response to declines in world agricul- els of support. Australia and New Zealand tural prices. Support to agricultural producers have very low support levels. The European from border protection and direct subsidies in- Union (on the high end) and Canada (on the creased farm-gate revenues in the OECD coun- low end) fall between these extremes. tries by almost 50 percent in 1999–2001 (table The Eastern European countries made the 3.11). But the persistence of high tariffs reduces most significant reductions in protection be- the incentives to eliminate production subsidies tween 1986 and 2001—from 63.6 percent to and various inefficiencies globally. 17.9 percent. Korea’s protection levels have re- Agricultural support tends to be counter- mained very high, with small variations. Mex- cyclical in rich countries, pushing price adjust- ico and Turkey, which started with low pro- 121 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 tection, increased it over this period, mainly through higher border protection. Figure 3.5 Rich countries use non–ad The high domestic price differentials in valorem tariffs more often than do table 3.11 indicate that domestic production is developing countries Tariff lines containing specific, compound, or mixed duties, protected much more significantly than the un- for agriculture and manufacturing by class of country weighted average tariff rates shown in table (as percentage of all lines) 3.8 would imply. In Japan, for example, border 35 protection raises market prices by some 138 percent, whereas the average tariff is just 10 30 percent and the maximum ad valorem tariff is 25 only 50 percent. The difference can only be at- tributed to specific duties and TRQs, which are 20 not included in the data set. For the European Union, the situation is similar. Border protec- 15 Agriculture tion raises prices by more than 34 percent, well 10 ᮡ Manufacturing above the average tariff of 19 percent. In both ᮡ areas, tariffs on many local specialties are very 5 high. For example, in the European Union the average tariffs for grains, meats, and milk and 0 Quad Large Other Lower milk products are 34.6 percent, 32.5 percent, middle middle income and 54.6 percent respectively. income income Source: WTO IDB. Specific duties produce hidden tariff increases in downturns The Uruguay Round objective of providing 1 percent of their tariff lines, and 4 in fewer greater transparency of protection levels than 5 percent of tariff lines. Only 4 countries, through tariffication has not been fully realized, all middle income, have a higher proportion of especially in the key industrial and some tariff lines with NAV rates. Within the Quad, middle-income countries. First, many agricul- Japan has specific, compound, or mixed rates in tural tariffs are still specific, compound, or 15 percent of its tariff lines; Canada in 24 per- mixed. In such cases it is almost impossible to cent; the United States in 40 percent; and the Eu- estimate the real level of protection because it ropean Union in 44 percent. The United States may change over time and with the relative and European Union also have duties that vary price of imports. Even more important are the according to the content of the products in 1 cyclical implications of such tariff structures: percent and 4 percent, respectively, of their tar- protection from specific duties rises as prices de- iff lines. Thus the difference in the transparency cline in the world markets; protection will be of tariff rates is consistent for most developing higher for lower-priced products from the de- and industrial countries, and the biggest prob- veloping countries.9 lem with nontransparency lies with the indus- The proportion of agricultural tariff lines trial and a few middle-income countries.11 that carry specific, compound, and mixed duties Within the Quad, tariff structures show is much higher in rich countries than in de- some differences. In the United States, almost all veloping countries (figure 3.5).10 This means, categories of products have NAV rates between among other things, that the transparency of 30 and 60 percent. In the European Union, cer- agricultural tariffs in developing countries is tain product groups—such as beverages, grains, higher than in industrial countries—and signifi- milk and milk products, and sugar and sugar cantly higher than in manufacturing. Of the 24 products—have more than 90 percent of tariff developing countries included in this sample, 11 lines under NAV. In many developing countries, have no NAV rates, 5 have them in fewer than NAV rates are clustered within a few product 122 A G R I C U L T U R A L P O L I C I E S A N D T R A D E Table 3.12 Specific tariffs are higher than ad valorem rates Average applied, out-of quota, ad valorem and ad valorem equivalents of non–ad valorem tariffs in areas for which equivalents are reported (percent) Average ad valorem Percentage of lines Average ad valorem tariff equivalent of NAV rates containing NAV rates Australia 1.2 5.0 0.9 European Union 10.6 35.2 43.6 Jordan 21.6 58.0 0.8 United States 8.1 11.7 40.4 Source: WTO Integrated Database (IDB). groups. For example, in Malaysia NAV rates as shown in table 3.12, the ad valorem equiva- apply to tobacco and alcohol products; in Mex- lents of specific duties, where known, are higher ico on chocolate and confectionary products; than the ad valorem rates. Second, the propor- and in Korea on nuts, spices, and sugar. tion of specific duties increases with the degree Only four countries in the sample report of processing (figure 3.6). They are found most the ad valorem equivalents of their NAV rates frequently in lines covering final products— (table 3.12). For those four, the average equiv- those classified under food processing. alents are much higher than the average ad valorem rates, suggesting that average duties Tariff escalation is particularly harmful for countries with a large proportion of NAV to development duties are seriously underestimated. Tariff codes that apply higher tariffs to semi- The specific duties are being used primarily processed and fully processed raw materials as an instrument of disguised protection. First, are strikingly antidevelopment. By hindering Figure 3.6 Throughout the world, tariff rates escalate with degree of processing Tariff rates by area and stage of processing (percent) 60 50 40 30 Final Intermediate 20 ᮡ ᮡ Raw 10 ᮡ 0 Quad Canada Japan United States European Lower Large middle Other middle Union incomea incomeb incomec Note: a. Bangladesh (1999), Guatemala (1999), Indonesia (1999), Kenya (2001), Malawi (2000), Togo (2001), Uganda (2001), and Zimbabwe (2001). b. Brazil (2001), China (2001), India (2000), Korea (2001), Mexico (2001), Russian Federation (2001), South Africa (2001), and Turkey (2001). c. Bulgaria (2001), Costa Rica (2001), Hungary (2001), Jordan (2000), Malaysia (2001), Morocco (1997), Philippines (2001), and Romania (1999). Source: WTO Integrated Database (IDB). 123 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Table 3.13 Tariffs rise with level of processing Tariff escalations in selected product groups (percent) European Union United States Korea Japan Tropical products Coffee Raw 7.3 0.1 5.2 6.0 Final 12.1 10.1 8.0 18.8 Cocoa Raw 0.5 0.0 5.0 0.0 Intermediate 9.7 0.2 5.0 7.0 Final 30.6 15.3 12.3 21.7 Sugar Raw 18.9 2.0 a 25.5 Intermediate 30.4 13.8 19.3 11.6b Final 36.4 20.1 50.0 a Expanding commodities Fruits Raw 9.2 4.6 49.6 8.7 Intermediate 13.3 5.5 30.0 13.2 Final 22.5 10.2 41.9 16.7 Vegetables Raw 9.9 4.4 135.4 5.0 Intermediate 18.5 4.4 52.2 10.6 Final 18.0 6.5 34.1 11.6 Seafood Raw 11.5 0.6 15.6 4.9 Intermediate 5.1 3.2 5.8 4.3 Final 16.2 3.5 20.0 9.1 a. All lines are specific. b. 56 percent of lines are specific. Source: WTO Integrated Database. diversification into value-added and processed processing, suggesting potentially large gains if products, areas in which trade is expanding escalation were removed by developing econo- rapidly, such escalation directly penalizes in- mies (Rae and Josling 2003). vestors in developing countries who seek to Tariff escalation is common in both tradi- add value to production for export. tional and new products. For traditional prod- Tariff escalation has long been a feature of ucts (except sugar), raw stages are accorded agricultural and food-processing trade and con- extremely low tariffs, whereas extremely high tinues to be so (Golub and Finger 1979, Lind- tariffs apply to the final stages. A similar pat- land 1997, and Gallezot 2003). Protection es- tern appears in fruits and vegetables, for which calates with the level of processing in almost all the developing countries have found expanding countries and across all products (table 3.13). markets and trade barriers are generally lower. Almost all groups of countries have highly es- The averages reported in table 3.13 mask very calating tariffs (see figure 3.6), and the manu- high peaks on individual products. In the facturing component of agriculture and food United States, for example, the maximum tariff processing has very high protection, explaining on final fruit products is 136 percent; on cocoa the developing countries’ lack of penetration in products it is 186 percent. In the European food processing in industrial countries. Devel- Union the maximum rates on processed fruits oping economies also apply systematic tariff es- and vegetables are 98 percent and 146 percent; calation and high tariffs to the final stage of on cocoa products, 63 percent. 124 A G R I C U L T U R A L P O L I C I E S A N D T R A D E Figure 3.7 The proportion of tariff lines containing non–ad valorem duties increases with degree of processing Tariff lines containing specific, compound, or mixed duties, by stage of processing (as percentage of all lines) 70 60 Intermediate 50 Raw Final ᮡ ᮡ ᮡ 40 30 20 10 [zero] 0 Norway European Union United States Canada Russia Turkey Source: WTO IDB. Specific duties are applied more frequently Export subsidies directly depress to goods with higher degrees of processing. For global prices example, in Canada and the European Union, International trade rules have prohibited ex- the share of specific duties is 17 percent and 22 port subsidies on nonagricultural products percent for raw materials but 30 percent and 58 since 1955. Export subsidies are still allowed percent, respectively, for final products. Among in agricultural products, although these subsi- developing countries, the Russian Federation dies were capped and subjected to reduction applies specific duties in 12 percent of its tariff commitments in the Uruguay Round. During lines for raw materials versus 53 percent of 1995–98, WTO members used 42 percent of lines covering final products (figure 3.7). the budgetary expenditure and 64 percent of the volume allowed for export subsidies, with Tariff rate quotas allow a little in— the European Union accounting for 90 percent and then add a tariff bite of all OECD export subsidies. TRQs, designed to maintain some market ac- Although their use has been reduced, ex- cess, have resulted in more complex tariff port subsidies continue to distort world mar- regimes. Although the number of tariff lines kets.12 The Uruguay Round placed limits on under TRQs is small, TRQs cover some of export subsidies for individual commodities, the main commodities produced in the OECD but allowed some flexibility. Early in the im- countries (figure 3.8). According to OECD data, plementation period, when world prices were almost 28 percent of domestic agricultural high, usage was low and several countries production is protected by TRQs. Rates range carried forward their unused export subsidy from a high of 68 percent in Hungary to 0 credits to be used at a later date. At the same percent in Australia and New Zealand. The time, lower tariffs and the move toward di- European Union and United States have 38 rect production subsidies has and will con- percent and 26 percent of their production tinue to reduce the need for official export protected by the TRQs. subsidies. 125 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Figure 3.8 Tariff rate quotas protect a substantial portion of output in many industrial countries Commodities covered by tariff rate quotas, expressed as a percentage of output 60 50 40 30 20 10 [zero] 0 OECD European United Japan Eastern Australia, Other Other average Union States Europe New Zealand industrial developing Source: OECD, Agriculture Market Access Database (AMAD). Even if tariffs were eliminated altogether, The development tale of five commodities: current production subsidies for agriculture sugar, wheat, cotton, peanuts, and rice would cause the domestic and export price of The development consequences of high protec- many commodities to remain lower than their tion in industrial countries can be traced costs of production in industrial countries. By through the story of key commodities. Although lowering production costs, production sub- the stories are different, they share common sidies favor industrial-country farmers over plots: high protection, regressive subsidies, and developing-country producers, who do not re- low prices that hurt poor producers all over ceive direct subsidies. Cotton subsidies in the the world (Beghin and Aksoy 2003). European Union and the United States are a Sugar is one of the most policy-distorted clear case in point. Tariffs are zero, and do- commodities in the world. The European mestic prices are the same as world or export Union, Japan, and the United States account prices (Baffes 2003, Watkins 2003). Yet in the for the bulk of OECD-zone support to sugar United States in 2001 production subsidies ef- producers, which, at $6.4 billion, is approxi- fectively increased the prices farmers received mately equal to developing-country exports. (or reduced their costs of production) by 51 But other countries (Mexico, Turkey, Poland, percent, leading to increased production and and all almost all temperate-zone sugar beet depressing the global price. At the same time, producers) also provide significant support to export prices for U.S. wheat, corn, and rice their producers. High border barriers in com- were 58, 67, and 77 percent of their costs of bination with the subsidies keep domestic production (Watkins 2003). prices in the United States and the European Decoupling subsidies from production Union about twice as high as the world mar- would reduce such distortions. To fully decou- ket price. ple subsidy payments, the definition of decou- High domestic sugar prices in the European pling must make it clear that the payments are Union, Japan, and the United States have en- independent of production decisions (box 3.5). couraged high-cost, inefficient domestic pro- 126 A G R I C U L T U R A L P O L I C I E S A N D T R A D E Box 3.5 Decoupling agricultural support from production decisions D ecoupling subsidies from production is designed to support producers not on the basis of current output, input use, or prices, but on historical mea- No constraints on input use. Support to specific sectors should be in the form of taxpayer-funded payments and should not require production. Neither sures, thereby limiting distortion to production and land, labor, nor any other input should be required trade. Debated since 1945, decoupling became a seri- to be in “agricultural use.” ous option with the passage of the U.S. Food Secu- Government credibility. Eligibility rules should rity Act of 1985, which reduced set-asides of farm be clearly defined and not allowed to change. The land, public stockholding, and yield payments. The time period on which payments are based should not European Union restructured its Common Agricul- change. Payments should not be increased. New sec- tural Policy in 1992, replacing some price supports tors should not be added to the program. Updating with direct payments. Mexico reformed its price- baselines and adding crops results in a government support policies along similar lines with the intro- credibility problem, making the decoupling policy in- duction of the Programa de Apoyos Directos al consistent over time. As market conditions change, Campo (PROCAMPO) in 1994. The United States governments have discretion to change eligibility then went a step further in the 1996 Farm Bill, re- criteria and payment levels, leaving them unable to placing “deficiency payments” with decoupled sup- make and hold to a binding commitment. As farmers port based on historical data. Turkey introduced a change their production decisions and apply pressure direct income-support program in 2001, aided in for changes in supposedly decoupled support pro- part by a World Bank adjustment lending operation. grams, decoupling is in effect preempted. Following a sharp decline in commodity prices Other programs. Every decoupling program in- in the late 1990s, the United States reintroduced stituted to date has left other support programs in place. Coupled programs tend to interact with the de- deficiency payments in 1999—initially as emergency coupled program, adding incentives to overproduce. assistance and subsequently as countercyclical pay- Time limit. Payments must not extend beyond a ments legitimized in the 2002 Farm Bill. Responding maximum number of years. The European Union to the U.S. reversal, Mexico reintroduced price sup- and Turkey have no limit; the United States had one ports in 2002 by setting target prices similar to those (at least implicitly) but violated it; Mexico’s remains in the United States. in effect. A time limit ensures that payments are tran- The move to decoupled support is a step in the sitory and for adjustment purposes only. right direction. However, if governments wish to Reform within WTO. The level of payments in help farmers adjust to free markets—the avowed aggregate and per farm, and the terms described purpose of decoupling—a simple and minimally dis- above, should be bound in the WTO so that gov- torting way to do that would be to make a one-time ernments can make credible commitments without unconditional payment to everyone engaged in farm- backsliding. ing or deemed in need of compensation. Short of that, decoupling mechanisms should exhibit the fol- lowing characteristics: Source: Baffes and de Gorter (2003). duction of sugar and sugar substitutes. At the creased to displace 10 million tons of sugar same time, they have reduced overall consump- consumption—equivalent to one-third of world tion and gradually transformed these countries exports—since 1970 (Mitchell 2003). Conse- from net buyers of about half of the world’s ex- quently, the world prices of sugar today are ports during the 1970s into net sellers in inter- below the costs of production of some of the national markets in the 1990s. Meanwhile, the most efficient producers. Many producers man- production and consumption of sugar substi- age to keep exporting, either because they enjoy tutes (such as high-fructose corn syrups) in- limited preferential access at high prices in 127 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Figure 3.9 High protection of sugar and wheat has increased domestic production and reduced net imports a. Production and net imports of sugar in the European b. Production and net imports of wheat in the European Union, Japan, and the United States Union Millions of tons Millions of tons 35 120 30 100 Production 25 80 ᮡ ᮡ 20 60 Production 15 40 Net imports Net imports 10 20 ᮡ ᮡ 5 0 0 —20 —5 —40 65 69 73 77 81 85 89 93 97 01 61 65 69 73 77 81 85 89 93 97 01 19 19 19 19 19 19 19 19 19 19 20 19 19 19 19 19 19 19 19 19 20 Source: FAO. industrial-country markets or because they sub- consumer prices would more than offset lower sidize their exports by selling at higher prices in producer prices on the 40 percent of sugar that their domestic markets. The world market has is domestically produced. In the United States, shrunk to a trade residual, with an estimated 80 producer losses would be some $200 million percent of world production being sold in high- greater than consumer gains. Western Europe priced, protected markets (figure 3.9). would show a net gain of $1.5 billion, with The benefits of sugar policy reform are sub- consumer gains of $4.3 billion exceeding pro- stantial—particularly with multilateral reform. ducer losses of $3.3 billion. Exporting coun- Presently, developed countries are protecting tries that presently enjoy preferential access to their sugar producers at great cost to them- the European Union and the United States now selves and to developing countries with export collect some $800 million by selling into pro- potential. A recent study of the global sugar tected markets at high prices. However, the and sweetener markets estimated that remov- value of this preferential access is less than it ing all trade protection and support would appears, because many of these producers have bring annual global welfare gains of $4.7 bil- high production costs and would not produce lion. In countries with the highest protection— at all at world-market prices. The rise of world Europe, Indonesia, Japan, and the United sugar prices following full liberalization would States—net imports would increase by 15 mil- partially offset the loss of preferences and lion tons per year. World sugar prices would allow some preferred producers to compete. rise about 40 percent, while prices in heavily The net loss to preferred producers from full protected countries would decline: in Japan by liberalization is estimated to total about $450 65 percent, in Western Europe by 40 percent, million per year (Borrell and Pearce 1999, and in the United States by 25 percent. Brazil- Sheales and others 1999). ian producers would gain the most from liber- A similar situation occurred in EU wheat alization—about $2.6 billion per year—but markets as high domestic prices encouraged this gain would be partially offset by higher production and reduced net imports from consumer prices. Japan’s net gain from lower about 5 million tons in the 1970s to net exports 128 A G R I C U L T U R A L P O L I C I E S A N D T R A D E of 20 million tons in the early 1990s, before by 13 percent. Uzbekistan would increase its policy reforms reduced net exports. Subsidized exports by 5.8 percent and Australia by 2.7 wheat exports from the European Union con- percent, while exports from the United States tinue to depress world prices. Wheat is one of would decline by 3.5 percent. Cotton produc- the most protected products in the European tion in the United States would decline by 6.7 Union; total production support averaged al- percent; in the European Union, by 70.5 per- most $10 billion annually during 1999–2001, cent. In effect, cotton production in the Euro- corresponding to a protection rate of almost pean Union would fall back to levels that ex- 50 percent. isted prior to the Common Agricultural Policy. World trade in cotton shows severe policy Groundnuts (peanuts) are one of the distortions, but, unlike sugar, the distortions world’s main oilseed crops. Widely cultivated come through producer support rather than in developed and developing countries, they from border measures such as tariffs and quo- provide livelihood and cash income to many tas (Baffes 2003). The United States provides poor farmers in the developing world, espe- the greatest support to its producers—$3 bil- cially in Sub-Saharan Africa and Asia. In Sene- lion annually. The European Union provides gal, for example, an estimated one million about $0.6 billion each year to its producers. people (one-tenth of the population) are in- Producer prices in the United States were 91 volved in groundnut production and process- percent higher than the world-market price ing. Groundnuts account for about 2 percent in 2001–02. In Greece they were 144 percent of GDP and 9 percent of exports. China is the higher; in Spain, 184 percent higher. High- world’s largest exporter of groundnuts, fol- producer support encouraged U.S. cotton pro- lowed by the United States and Argentina. duction to grow about 25 percent faster than Sub-Saharan Africa (where the major produc- world production after 1970, and EU produc- ers are The Gambia, Malawi, Nigeria, Sene- tion accelerated once Greece and Spain joined gal, South Africa, and Sudan) has lost ground the (then) European Community in 1981 and in world edible groundnut markets, account- 1986. While the United States and European ing for only 5 percent of the world market in Union were maintaining high support, sev- 2001, compared to 17 percent in 1976. In the eral cotton-producing developing countries oil segment of the market, Senegal is the world’s (especially those in Sub-Saharan Africa) un- largest exporter. Governments in Sub-Saharan dertook substantial policy reform to increase Africa taxed production until the early 1990s. the efficiency of their cotton sectors. Price These taxes, borne by domestic groundnut and export prospects of developing-country users and taxpayers, had an important domes- exporters—especially in Sub-Saharan Africa— tic cost (Diop and others 2003). would be greatly improved if support in devel- Historically, world groundnut markets have oped countries were reduced or eliminated. been distorted by heavy government interven- Removal of protection and support would tion designed to stimulate production through cause a drop in production in the United States subsidies and price supports or to protect pro- and European Union and thus boost prices. ducers by controlling imports. China and India Simulations show that with full liberalization have price-control schemes and impose very in the cotton sector—removal of trade barriers high tariffs on imports. Since the mid-1990s, and production support, along with liberaliza- all major exporters have gradually liberalized tion in all other commodity sectors—cotton their groundnut sectors, in part to fulfill their prices would increase over the next 10 years by commitments under WTO agreements. Results an average of 13 percent over the price that are mixed, however, and trade in groundnuts would have prevailed in the absence of re- remains heavily distorted. Both China and forms. World cotton trade would increase by 6 India have removed some import restrictions percent. Africa’s cotton exports would increase and allowed wider private-sector participation 129 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 in importing groundnuts. But tariffs on countries modeled in our analysis (The Gam- groundnut products remain very high in both bia, Malawi, Nigeria, Senegal, and South countries; the removal of trade distortions by Africa) would experience aggregate net welfare China and India is essential to successful re- gains of $72 million, with Senegal and Nigeria form of groundnut markets. gaining most. The increase in world prices The U.S. groundnut policy, highly distorted after trade liberalization would lead to a total by large subsidies and prohibitive tariffs be- gain for African groundnut producers of some tween 1930 and 2001, was recently reformed, $124 million in profits. These figures are siz- but with high and redundant tariffs still in able for small African economies. The rest of place. The 2002 Farm Bill eliminated some the world would experience a net welfare loss unsustainable features of previous legislation because consumers would face higher prices (high support prices and production quotas) for groundnut oil. but introduced new distortions that have the Rice is the most important food grain in the potential to depress world market prices and world. Production and consumption are con- subsidize producers (for example, through centrated in China, India, and Indonesia. Con- countercyclical payments and a price floor sumers in low-income, food-deficit countries mechanism that becomes effective when world get 28 percent of their calorie intake from rice. prices are low). Prohibitive tariffs of almost The rice market is a mature market, with sta- 150 percent remain. tic demand in the North and demand in devel- Full trade liberalization would raise market oping economies growing with demographics prices by about 19 percent for groundnuts, 18 rather than income. Prospects for growth in percent for meal, and 17 percent for oil. Be- trade therefore rely on policy reforms. cause the current U.S. peanut program is Tariff and related border protection is very mostly a domestic affair, liberalization of the high, averaging about 40 percent globally and U.S. market would not have a far-reaching ef- rising to 200 percent in some markets. Total fect on world prices or on exports of the poor- OECD-zone support is more than $26 billion, est developing countries. As a bloc, the OECD and in Japan support is a staggering 700 per- countries would experience welfare losses after cent of production cost (at world prices). Tar- trade liberalization—moderate gains in the iff escalation is prevalent (from paddy to United States offset by losses in Canada, the milled rice) in many countries, including the European Union, and Mexico, which would European Union, where the tariff on milled lose from trade liberalization because, with rice is prohibitive, except for small preferential few policy distortions in these markets, they import quotas granted to a few countries. For would be penalized by higher world prices example, the tariff on milled rice imports into after liberalization. the European Union is 80 percent, compared Although the net world welfare gains of to 46 percent for brown rice. In Mexico, paddy liberalizing groundnut markets are moderate, rice enters with a 10 percent tariff, whereas they are still significant for small agrarian brown and milled rice enter with a 20 percent economies such as Malawi and other West tariff. This pattern of protection depresses African countries. In China and India, gains world prices for high-quality, milled long-grain to consumers would be partially offset by rice and discriminates against the milling sec- losses to producers under full trade liberaliza- tors of exporting nations such as Thailand, the tion. Specifically, buyers in India and southern United States, and Vietnam (Wailes 2003). China, where groundnuts and groundnut oil Global reforms—elimination of all border are heavily used in food, would reap signifi- barriers and support—would lead to average cant gains from liberalization. price increases of about 33 percent, rising to Liberalization of the value-added markets— 90 percent for medium- and short-grain rice. oil and meal—would result in even larger wel- Producers in Cambodia, China, and Vietnam fare gains in African countries. The African would be the main beneficiaries, along with 130 A G R I C U L T U R A L P O L I C I E S A N D T R A D E consumers in most of high-income Asia. Since agriculture. Its reforms have led to higher pro- most production is by small farmers in these ductivity and growth rates, no changes in rural countries, the gains would be very pro-poor as population, and a much more dynamic and en- well. Following trade liberalization, net rice vironmentally sustainable agricultural sector importers could be negatively affected by the (box 3.6). Particularly noteworthy is the fact resulting world price increase wherever the that New Zealand farmers are able to compete consumer prices rise following reform; that is, effectively on world markets, expanding their wherever the current ad valorem tariffs are share of world trade in dairy products from 6.7 lower than the potential world price increase. percent in 1985 to 9.5 percent in 2001. Estimates show that in Indonesia, Nigeria, and the Philippines, three large rice importers, Harbinson splits the difference consumer prices would fall after the reforms. Despite the large potential gains from liberal- The tale of these five commodities has an ization, many of the proposals for the Doha important moral for those who would pro- Round are modest. Proposals range from the mote development. Cutting back on subsidies Japanese suggestion to impose an “average and other protection that primarily benefit rel- cut,” which can be predicted to have little ef- atively wealthy farmers in rich-country mar- fect, to the more ambitious proposal of the kets (and in some cases middle-income coun- Cairns Group. try markets) can open up opportunities for The Harbinson proposal, named for Stuart poor farmers in Africa, Asia, and Latin Amer- Harbinson, the chairman of the WTO negoti- ica. The effects on incomes in poor countries ating group on agriculture, takes the middle would be strong and immediate. In many cases ground (DRIFE 2003).13 For industrial coun- the gains would be a substantial order of mag- tries, it proposes average tariff cuts of 60 per- nitude greater than development assistance to cent on bound tariffs above 90 percent, a 50 these same countries. percent cut on bound tariffs between 15 and 90 percent, and a 40 percent cut on bound tariffs below 15 percent.14 For the developing Proposals for reforms in the countries and for products not considered Doha Round strategic, it proposes average tariff cuts from bound rates of 40 percent for tariffs above 120 The potential gains for developing percent, a cut of 35 percent for tariffs between countries are large 60 percent and 120 percent, a cut of 30 per- One way to evaluate reform proposals is to cent for tariffs between 20 percent and 60 per- compare their likely results with the potential cent, and an average reduction of 25 percent gains from full removal of all barriers, which in tariffs below 20 percent. These cuts would would yield global welfare gains of $400–900 be implemented by industrial countries in billion, more than half of which would go to equal installments over five years and over ten developing countries. If all trade barriers were years for developing ones (WTO 2003). The dismantled, agriculture and food would ac- Harbinson cuts look significant—some groups count for 70 percent of these gains. A major have called them radical—but their impact, de- share—60 percent—would derive from reforms pending on how they are interpreted, would in developing countries. The largest gains are to not be as significant as first appears. be had from tariff reforms in agriculture under- For the industrial world, the results would taken in a context of a global reform program. depend on whether countries achieve the aver- Can agriculture adjust to new prices? The ex- age cuts by reducing lower tariffs by greater perience of New Zealand, which implemented percentages (which would have relatively little the most far-reaching reforms of any industrial effect) or cut all tariffs at the average rate. The country, suggest that the answer is yes. New “average cuts” called for under the Uruguay Zealand has almost no tariffs or subsidies in Round were interpreted loosely, with many 131 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Box 3.6 Fewer subsidies, stronger agricultural sector T here is a strong belief among policymakers in OECD countries that trade reform in agriculture would destroy their rural communities and the agri- 1986–87, output of the agricultural sector has grown by more than 40 percent in constant terms. The share of farming in GDP rose from 14.2 percent cultural sector. Yet, as the experience of one OECD in 1986–87 to 16.6 percent in 1999–2000, and country shows, protection and subsidies are not a nec- growth in the farming sector has outpaced economic essary condition for the continued growth of the farm growth of New Zealand as a whole. The reform also sector. Indeed, the removal of protection can be ac- prompted greater competition and lower input costs companied by faster agricultural growth and increases among suppliers, and brought environmental benefits in productivity, achieved without a significant decline through reduced waste. Although land values fell in the farming population or its standard of living. during late 1980s and early 1990s, they recovered Today, New Zealand has the lowest level of during the later part of the decade. The share of farm support among OECD countries—its producer rural population has remained constant since the support, estimated to be around 1 percent of the abolition of subsidies. value of agricultural production, is primarily dedi- Some of the most impressive effects of subsidy cated to research funding. This was not always the removal have been the changes in agricultural pro- case. Producer support reached 33 percent of output ductivity. Since 1986, the annual average rate of in 1983, when almost 40 percent of the income of productivity growth in agriculture has reached 5.9 an average sheep or cattle farmer came from govern- percent, compared with 1 percent prior to subsidy ment subsidies. Yet, these policies were clearly unsus- abolition. The fact that total lamb production has in- tainable, as the loss of preferential access to the creased while the number of sheep has declined by 29 British market and an escalating inflation spiral led percent attests to the increased efficiency of the sec- the government to abandon most payments to agri- tor. However, some studies, such as Morrison Paul, cultural producers. Johnston, and Frengley (2000), have questioned the Government deregulation was quick and sub- positive effects of the reforms on productivity. The stantial. Nearly all subsidies were removed in 1984. latter, using an unbalanced panel of 32 farms be- The sectors involved included wheat, egg, milk, pota- tween 1969 and 1991, found that agricultural reform toes, honey, raspberries, hops, tobacco, apples, poul- caused changes in the composition of output—a shift try, pork, and other meats. Altogether, almost 30 dif- out of wool and lamb and into beef and deer—but ferent production subsidies and export incentives did not affect technical efficiency. On the other hand, were abolished (Bell and Elliott 1993). The govern- work using aggregate data, such as Kalaitzandonakes ment made only limited efforts to soften the impact and Bredahl (1994), has confirmed improvements in on farmers; those who decided to exit the agricul- technical efficiency following the reforms. tural sector received a one-time “exit grant” of ap- Overall, the removal of support did not have a proximately two-thirds of annual income. grave effect on New Zealand’s farmers. Instead, the At the time, estimates pointed to 8,000 farms policy of liberalization created a more vibrant, diversi- (10 percent of total) going out of business, prompt- fied, and sustainable rural economy in New Zealand. ing widespread opposition to the government’s plan. However, only 800 farms exited the market, and those that remained became more dynamic. Since Source: World Bank staff. countries reducing already-low tariffs by high duced applied rates to much lower levels.15 If percentages to avoid cutting higher tariffs sig- cuts were applied to the bound rates, such nificantly (see chapter 2, box 2.2). countries would get credit for past unilateral For the developing countries, the key issue reforms, but the reductions would not lead to is reductions from the bound, not the applied, significant tariff reductions. rates. Most developing countries have bound The Harbinson proposals would imply sub- their tariffs at relatively high rates, but re- stantial tariff cuts in the United States and Eu- 132 A G R I C U L T U R A L P O L I C I E S A N D T R A D E Table 3.14 The Harbinson proposals could greatly reduce applied tariffs in the European Union and the United States Tariffs in the European Union and United States before and after average reduction from applied tariffs (percent) United States European Union Before Harbinson After Harbinson Before Harbinson After Harbinson Average Peaks Average Peaks Average Peaks Average Peaks Raw 5.5 350.0 2.7 140.0 13.2 131.8 6.9 52.7 Intermediate 7.1 159.3 3.8 63.8 16.6 284.8 8.3 113.9 Final 11.7 180.8 6.2 72.3 26.8 506.3 13.1 202.5 Overall 8.8 350.0 4.6 140.0 19.7 506.3 9.9 202.5 Note: The analysis excludes cigarettes and alcoholic drinks. Source: WTO Integrated Database. Table 3.15 The Harbinson proposals would not significantly reduce protection in the developing world—if reductions were taken from bound rates Tariffs in selected areas before and after average reductions from bound rates (percent) Costa Rica India Jordan Korea Bound rates Average Peak Average Peak Average Peak Average Peak Before Harbinson 49.0 245.0 115.3 300.0 21.5 180.0 50.8 917.0 After Harbinson 33.8 147.0 72.3 180.0 14.9 108.0 33.2 550.2 Current applied rates 13.1 154.0 36.7 115.0 18.5 120.0 42.7 917.0 Note: The analysis excludes cigarettes and alcoholic drinks. Source: WTO Integrated Database. ropean Union at the end of the program under the end of 10 years, the Harbinson reform an optimistic scenario in which all tariffs were would leave bound tariffs significantly above cut by the average rate from the applied rates applied rates. For Jordan and Korea, bound (table 3.14).16 rates after 10 years would be marginally below Under this optimistic scenario, the average the current applied rates. Because these results effective tariffs in the European Union and the would hold for most developing countries, ex- United States would be halved by the end of the isting levels of protection in the developing reform process. EU tariffs would come down to world would not be significantly reduced under about 10 percent from 20 percent, while U.S. the Harbinson proposals. tariffs would fall below 5 percent from 9 per- cent. Even so, the average agricultural tariffs in Cushioning adjustment: The impact both areas would remain significantly higher of reforms on net food importers than manufacturing tariffs—which stand at 4.2 Serious reforms in global trade policies would and 4.6 percent respectively. Tariff peaks would lead to price increases for many products remain above 140 percent in the United States now protected. These price changes could lead and above 200 percent in the European Union. to balance-of-payments problems for low- For the developing countries, the optimistic income developing countries that are net agri- scenario reduced the bound rates by the aver- cultural importers. Currently, the developing age cut. Four country examples are given in countries as a group—low- and middle- table 3.15 above. Cuts from bound rates do income alike—enjoy a trade surplus in agri- not significantly lower protection in most de- culture. But many countries are net importers, veloping countries. In India and Costa Rica, at and they could be negatively affected. Of 58 133 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Box 3.7 The potential impact of real preferences G iven the high level of agricultural protection in many industrial countries, the value of prefer- ences should be very high and should lead to high tected sectors such as grains and meat products. Join- ing NAFTA or the European Union implies more than simple preferential access—for example, membership rates of export expansion in the countries that re- in a trade bloc offers a more secure and predictable ceive them. After Spain and Portugal joined the environment for investment than is usually provided European Union, and after Mexico joined NAFTA, by unilateral preferences—but the experiences of exports rose dramatically, especially in highly pro- Mexico, Portugal, and Spain illustrate the potential tected milk products (see figures below). response of many developing countries if they were Milk and milk products are the most protected given free access with few other restrictions. of all commodities, and, at $42 billion, they have the highest level of OECD support. However, this highly protected subsector responds similarly to other pro- Source: COMTRADE. Exports of milk products shot up after Mexico, Spain, and Portugal joined regional trade blocs Exports of milk products from Mexico, 1961–2001 Exports of milk products from Spain, 1961–2001 (millions of dollars) (millions of dollars) 50 400 350 40 300 1994 Mexico joins NAFTA 30 250 200 1986 Spain joins EU 20 150 ᮡ 100 10 50 ᮡ 0 0 1961 1965 1969 1973 1977 1981 1985 1989 1993 1997 2001 1961 1965 1969 1973 1977 1981 1985 1989 1993 1997 2001 Exports of milk products from Portugal, 1961–2001 (millions of dollars) 120 110 100 90 80 70 60 1986 Portugal joins EU 50 40 30 20 10 ᮡ 0 1961 1965 1969 1973 1977 1981 1985 1989 1993 1997 2001 countries classified as low income in 2000–01, (China). Another 35 percent went to the oil 29 were net importers; of 89 classified as exporting countries—Algeria, Saudi Arabia, middle-income, 51 were net importers. and the United Arab Emirates. Excluding Among the middle-income countries, the these and small island states, Egypt and Oman total net imports of the net importers were al- account for 57 percent of remaining imports. most $56 billion; 46 percent of the imports Thus the impact of agricultural price increases went to high-income, industrialized develop- on the middle-income countries would be ing countries such as Hong Kong (China), limited, particularly as a proportion of their Republic of Korea, Singapore, and Taiwan trade. 134 A G R I C U L T U R A L P O L I C I E S A N D T R A D E Among low-income countries, oil-producing senting mainly tropical products not produced Angola, Nigeria, and Yemen account for almost in the United States, enters the country duty- 32 percent of the total deficit. Twelve countries free—here preferences have no effect. Of prod- in conflict account for another 21 percent. Only ucts in the GSP, most agricultural products 14 low-income countries are real net food im- with nonzero tariffs are not eligible for prefer- porters; their total net imports were only $2.8 ences—only 34 percent of imports covered by billion in 2000–01. In this group, three coun- the GSP were eligible for preferences; only 26 tries account for 80 percent of the net imports: percent received them. Bangladesh, Pakistan, and the Democratic Re- Preferences are more generous in other, public of Korea. The rest of the low-income mainly regional, programs. U.S. preferences countries have a deficit of just $565 million, a for Mexico and the LDCs are much more ex- small percentage of their trade. These countries tensive than for the rest of the world, and the would gain from price increases, because their eligibility ratio is almost 100 percent. How- exports are also predominantly agricultural, ever, this measure reveals little about the actual as well as from other aspects of a multilateral coverage of these schemes because it records trade negotiation. Nonetheless, the international only products actually exported and not those community should be prepared to provide assis- that would have been exported if granted pref- tance to countries to help them adjust to and erences or lower tariffs. For example, the total take advantage of new trade opportunities. exports of agricultural products with nonzero rates from the 64 GSP countries come to no Can tariff preferences substitute more than the exports of Mexico, which re- for reform? ceives almost full preferences (table 3.16). Some have argued that the poor are not harmed Tighter rules of origin also complicate pref- by the protection practices of rich countries be- erences. For example, seafood imports under cause the Quad countries are generous in grant- Europe’s Everything But Arms preference ing trade preferences. To be sure, the levels of scheme for least developed countries have protection in industrial countries are moder- stricter rules of origin than do its other prefer- ated by tariff and quota preferences. However, ence programs, the GSP and Cotonou agree- as we saw earlier in this chapter, most of the ments. Similarly, the NAFTA agreement, the poor live not in the least developed countries, world’s most extensive preferential trade which get deep preferences, but in Asia, which regime, is associated with very detailed and gets fewer preferences, if any. Thus deep prefer- product-specific rules of origin (box 3.8). ences do not reach the majority of the world’s Although preferences may help some very poor living on less than $1 day. Aside from the poor countries, they are no substitute for LDCs, many of the countries that enjoy prefer- multilateral reform that will benefit all the ences are not among the world’s poorest. For world’s poor. example, a significant portion of the EU’s low- tariff sugar quota benefits Mauritius, the rich- Summary: A pro-poor agenda est country in Sub-Saharan Africa. Half of the for policy change countries that benefit from U.S. sugar quotas Realizing the development promise of the Doha are net sugar importers. Rules governing pref- Agenda will require the international commu- erences are typically complex and cumbersome, nity to tackle some of the most difficult prob- preventing many producers from taking advan- lems of agricultural trade. Agriculture remains tage of them (see chapter 6). one of the most distorted areas of international The United States is the only country that trade, and those distortions impede develop- collects data on the effect and degree of use of ment. A pro-poor program of trade reform preferences. Agricultural exports from all de- would contain several important elements: veloping countries total about $25 billion; of A reduction in the use of specific duties that total, approximately $15 billion, repre- and greater transparency is necessary to bring 135 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Table 3.16 U.S. trade preferences—a plethora of programs U.S. trade preferences for agricultural products, 2002 (millions of dollars) Share of (a) Share of (b) Share of (b) for which duty for which no eligible for Eligible but Preference Country group (number is greater than preference is preference not requesting received of countries in group) Total value (a) zero (b) available (c) (d=b–c) preference (e) (f=d–e) ATPA (Andean) (4) 2,242.6 870.2 106.7 763.4 256.4 507.0 U.S. LDCs (40) 369.0 65.6 0.0 65.6 12.2 53.3 Non-LDC AGOA (15) 600.5 168.9 0.4 168.5 20.0 148.5 Non-LDC CBI (19) 3,005.3 1,391.3 0.7 1,390.6 10.8 1379.8 Jordan 1.2 1.0 0.1 0.9 0.1 0.8 Mexico 6,319.6 3,866.9 0.0 3,866.9 13.8 3,853.1 Other GSP countries (64) 9,769.6 3,662.0 2,408.5 1,253.6 300.6 952.9 Non-GSP developing 2,906.5 939.9 855.8 84.1 0.7 83.5 Total developing 25,214.3 10,965.7 3,372.1 7,593.5 614.6 6,979.0 Source: U.S. International Trade Commission. Box 3.8 Rules of origin in preferential schemes are complicated—and often contradictory R ules of origin are a key element in determining the extent to which countries are able to use the preferences available to them. EU rules of origin are particularly with regard to the nature and extent of “cumulation” and the “tolerance rule.” In this regard the rules of origin for the Everything But product-specific and sometimes complex. For some Arms scheme differ from those of the Cotonou products a change of tariff heading is required. Agreement—and also from those of other free-trade Others must meet a value-added requirement. Still agreements. The Cotonou Agreement, for example, others are subject to a specific manufacturing-process provides for full cumulation—inputs from other requirement. In some cases these requirements are Cotonou countries can be freely used. The GSP combined. For certain industrial products, alternative allows more limited diagonal cumulation, which methods of conferring origin are specified—for ex- may occur only within four regional groupings: ample, change of tariff heading or satisfaction of a ASEAN, CACM, the Andean Community, and value-added requirement. Although clearly more SAARC. The EU agreement with South Africa con- flexible, such an approach is not available for any tains a general tolerance rule of 15 percent, whereas agricultural products. For many products the EU those with Mexico and Chile allow only 10 percent. rules require a change of chapter, which is even more The rules of origin for the U.S. GSP scheme de- restrictive than a change of heading. In certain cases fine a 35 percent value-added criterion that is com- the EU rules provide for a negative application of the mon across all included products. In later bilateral change of tariff classification by proscribing the use trade agreements, such as the NAFTA and the re- of certain imported inputs. For example, the rule of cently signed free-trade agreement with Singapore, origin for bread, pastry, cakes, biscuits, and so on the United States has stipulated extensive and often requires a change of tariff heading except from any very complicated product-by-product rules of origin heading in chapter 11 (products of the milling indus- which run to several hundred pages. In any event, try). Hence, bakery products cannot use imported the common rule applied in the GSP is that sensitive flour and still qualify for the preferential rates. products are excluded from preferences. Although the European Union has sought to harmonize the processing requirements for each product, some of the general rules vary substantially, Source: World Bank staff. 136 A G R I C U L T U R A L P O L I C I E S A N D T R A D E agricultural protection regimes closer to the with some arrangement for reducing tariff tariff structures used for manufacturing. All escalation on agricultural products. specific, mixed, composite, and seasonal tar- The combination of tariff walls and domes- iffs should be replaced with transparent ad tic subsidies that annually channel some $248 valorem duties. Not only will this make the billion to producers in the industrial countries protection clear, but also it will eliminate dis- must be dismantled, as must the high levels crimination against lower-priced exports from of protection in developing countries. Export developing countries. Since tariff peaks are subsidies must be further reduced and ideally very high—and will stay high under the exist- eliminated. Discipline should also extend to ing reform proposal—the peaks must be capped, food aid (see box 3.9). Finally, border barriers Box 3.9 Food aid principles tween legitimate food aid and commercial interests is F ood aid recipients constitute a special group of low- income, food-importing countries with urgent needs arising from natural disasters, disease, and civil con- difficult to make. Thus, although the actual food aid budgets of the five largest donors in 1998 were $2.9 flict. In June 2003, FAO identified 37 countries requir- billion, Trueblood and Shapouri (2002) estimate the ing food assistance, most of them in Sub-Saharan annual cost of an insurance scheme to provide food Africa, but others in Asia, the Middle East, Europe and security for 67 needy countries would have cost less Central Asia, and Central America and the Carib- than $450 million per year from 1988 to 1999.19 bean.17 Overall, food aid accounts for a relatively small Any WTO agreement should tighten the URAA proportion of world trade, around 2 to 4 percent of provisions to facilitate genuine food aid while pre- traded cereal volumes during the period 1995– 2000.18 venting the abuse of aid to circumvent export subsidy Though needed and effective immediately after restrictions. Proposals include limiting food aid to disasters, food aid raises development and trade con- grants only or to in-kind provision only in response cerns when extended for longer periods or driven to appeals from the United Nations or other appro- by supply. From a commercial standpoint, food aid priate international bodies. Donations in cash or may disguise export subsidies, or it may be used for channeled through international agencies would be developing commercial export markets or promot- most desirable.20 Several principles, some beyond ing strategic objectives. Furthermore, it may alleviate the purview of the WTO, should govern the provision pressure on governments to reform policies and pro- of food aid: mote self-sufficiency. When given in kind, food aid may be detrimental • Food aid should be in the form of full grants and to local producers by lowering prices and by altering provided only for needs of well-defined vulnerable traditional dietary preferences. When distributed out- groups or in response to an emergency as deter- side of normal indigenous commercial channels, as is mined by the United Nations. usually the case, in-kind food aid also undermines the • Cash aid should be provided unless in-kind food aid development of those channels and disrupts move- is a more appropriate response to the crisis (for ex- ment of food to the deficit areas from surplus regions ample, because marketing channels are not func- in the country and neighboring countries. These events tioning, in-kind aid can be better targeted). can then increase the likelihood and severity of future • Food aid should never be used as surplus disposal famine situations. by industrial countries. The trade aspects of food aid are regulated by • An impact assessment on marketing and local in- many agreements and conventions. The Uruguay centives should be undertaken when food aid is Round Agreement on Agriculture (URAA, Section provided, and designs should be altered or mitiga- 10.4) requires that food aid not be tied to commercial tion should be undertaken if significant negative im- exports of agricultural products, that it accord with pacts are observed. the FAO Principles of Surplus Disposal and Consulta- tive Obligations, and that it be given under genuinely concessional terms. Nevertheless, the distinction be- Source: World Bank staff. 137 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 against processed foods, which constitute the countries take advantage of new trading op- expanding part of agricultural and food trade, portunities that arise with trade liberalization. must be brought explicitly into the negotia- tions. Policies governing such products should be aligned with those governing other manu- Notes factured products. Reform of these policies 1. Global poverty rates have been estimated on a consistent basis at $1 a day. Unfortunately, the poverty will yield immense global benefits, especially data are not separated for rural and urban populations. in developing countries. The only source of data where the poverty rates can be Decoupling subsidies can be positive. Re- separated between rural and urban households is based ducing subsidies without lowering border bar- on the national poverty rates that vary across countries, riers will have only marginal effects. Similarly, and the country coverage of these surveys is limited. decoupling subsidies from direct production Data used here cover the surveys for 52 country house- hold surveys conducted between 1990 and 2001. The will have no effect if border barriers are not sample has a higher share of rural population than the slashed. However, if border protection is re- overall average and both ratios are given in the tables duced and subsidies decoupled from produc- for reference. tion requirements, the effects would be posi- 2. A comprehensive analysis of (a) protection indi- tive. To succeed, the decoupling programs cators (tariff protection, nontariff barriers, and trade- must have characteristics that most past ef- distorting domestic policies such as market price forts have lacked (see box 3.5). supports and export subsidies), and (b) performance indicators (export structure and output) requires con- A global effort should be made on particu- sistent information that is available only for the OECD lar commodities with large development con- countries and then only for some product groups. sequences. Certain individual commodities can Even for the OECD, the focus of data is more the have important effects on both developing and protection of selected commodities than the overall industrial countries. Sugar, cotton, wheat, and trade regime. Thus, the measures covered by OECD groundnuts all illustrate ways in which policy data systems and the tariff data from the WTO are not fully consistent. Definitions of the agricultural sector regimes—particularly in the OECD coun- also vary. The OECD database focuses on key raw tries—can adversely affect developing coun- commodities that have high protection; others exclude tries when allowed to operate over long peri- fisheries, which have become the biggest food trade ods of time. item. Many agricultural items are covered under food A program of development assistance to processing and thus are classified under manufacturing manage the adjustment to reform—particu- rather than agriculture. Because processed foods con- stitute a growing share of consumption and trade, their larly in food-importing countries—is a prior- absence from the data seriously understates trade in ity. The effects of tariff and subsidy reform are agricultural products. Finally, trade regimes in agricul- unlikely to affect most countries adversely, but ture include complicated duty structures, extensive use the risk that a handful of countries may ex- of quotas and other restrictions, and complicated and perience a net terms-of-trade loss cannot be changing subsidy schemes, all of which make it im- treated lightly. Adjustment is not likely to be possible to devise simple measures of protection and distortions. costly. Careful analysis shows that most net Information for the developing countries is more food importers are either high-income indus- limited and is only partially consistent. In the analysis trialized countries or major oil exporters. presented in this chapter, partial data will be patched Many of the remaining net food importers together to give a picture of agricultural trade regimes have high tariff walls, so that reducing the tar- and export performance in industrial and developing iffs could offset all or most of the increase in countries. For the purposes of this study, the agricultural sec- the global price. Nonetheless, such countries tor is defined broadly to include fisheries and processed would lose the revenues associated with the food products in all subgroups. For example, the high tariffs and so would experience some dis- seafood and seafood products subgroup includes raw, location. Development assistance can also help frozen, and processed seafood. This classification al- 138 A G R I C U L T U R A L P O L I C I E S A N D T R A D E lows us to include all stages of processing and to con- ment on Agriculture. The proposal also spells out struct data series that are economically consistent. See propositions on special and differential treatment and annex I for the details of the coverage and definition of the role of nontrade-concerns. subgroups. 14. These are average cuts, so the actual cuts in 3. Of 20 categories of farms tracked by the U.S. De- each line could be lower. partment of Agriculture, 12 lose money from farming 15. This is also true of many industrial countries alone. Most of the money-losing categories consist of but the difference between the bound, and applied smaller farms. USDA, Agricultural Income and Finance rates is much smaller. Outlook, September 26, 2002. 16. The European Union and United States were se- 4. OECD (2002) The Incidence and Income Trans- lected because there are tariff equivalents for the spe- fer Efficiency of Farm Support Measures. cific duties. The data for the European Union is for 5. From the trade data, it is very difficult to separate 1999, the last year for which the tariff equivalents were out food processing from raw agricultural trade. The available. definition used here treats food processing within agri- 17. http://www.fao.org/docrep/005/y9643e/y9643 culture and manufacturing excludes food processing. e04.htm 6. Annual GDP growth in industrial countries 18. http://www.foodgrainsbank.ca/downloads/fjfa_ slowed from 3.0 percent in the 1980s to 2.3 percent in foodaid.pdf the 1990s. In the developing countries, during the same 19. Trueblood, Michael, and Shahla Shapouri. period, annual GDP growth accelerated from 3.1 per- 2002. “Safety Nets: An Issue in Global Agricultural cent to 3.7 percent. Unless there was a significant Trade Liberalization.” Agricultural Outlook (Eco- change in income elasticities between the 1980s and nomic Research Service/U.S. Department of Agricul- 1990s, the changes in GDP growth rates are not large ture). March. http://www.ers.usda.gov/publications/ enough to cause the shift in import growth rates. But agoutlook/Mar2002/ao289f.pdf faster liberalization in developing countries can explain some of the shift. 20. WTO, Committee on Agriculture Special Ses- 7. Annex 2 Table 4 shows the ad valorem and non- sion. 2002. “Negotiations on Agriculture: Overview.” ad valorem rates separately, as well as the proportion TN/AG/6. Pages 58–61. December 18, 2002. http:// www. of the tariff lines to which the average applies. wto.org/english/tratop_e/agric_e/negoti_ modoverview_e. 8. In the European Union and United States, very pdf high tariffs are all specific. The variance and peaks for Canada and Japan probably do not reflect the real peaks because specific duties are excluded. References 9. For example, in the European Union the duties Aksoy, M. Ataman, Z. Ersel, and B. Sivri. 2003. “De- on wine are 13 Euros per hectoliter, which corresponds mand Growth versus Market Share Gains: De- to about 12 cents per bottle. For wines that come from composing Export Growth in the 1990s.” developing countries such as Bulgaria and Moldova, Mimeo. World Bank, Washington, D.C. CIF prices per bottle are less than $1, which gives a tar- Baffes, John. 2003. “Cotton Market Setting Policies, Is- iff rate of about 12 percent, a high rate. For a $10 dol- sues, and Facts.” Mimeo. World Bank, Washing- lar bottle from California, the tariff rate would be just ton, D.C. March. 1.2 percent, a very low one. Baffes, John, and Harry de Gorter. 2003. “Decoupling 10. Individual country details are given in annex 2. Support to Agriculture: An Economic Analysis and 11. A recent OECD publication argues that the ad- Recent Experience.” Annual World Bank Confer- ministration of ad valorem rates could cause difficul- ties for the customs administration; the developing ence on Development Economics. Paris, May countries have been administering such rates with much 15–16. lower administrative capacity (OECD 2002a). Beghin, John, and M. Ataman Aksoy. 2003. “Agricul- 12. Additional distortion is produced by circum- tural Trade and the Doha Round. Lessons from vention, possibly through the subsidy elements in ex- Commodities Studies.” Annual World Bank Con- port credits, export restrictions, and revenue-pooling ference on Development Economics. Paris, May arrangements in major products. 15–16. 13. The Harbinson proposal presents the current Bell, B., and R. Elliott. 1993. “Aspects of New Zea- status of agricultural negotiations on establishing nu- land’s Experience in Agricultural Reform Since meric targets, formulas, and other ‘modalities’ for 1984,” Ministry of Agriculture and Forestry, countries’ commitments to increase market access, de- Auckland, New Zealand. crease export subsidies, and reduce distorting domestic Borrell, Brent, and David Pearce. 1999. “Sugar: the support as mandated by the Uruguay Round Agree- Taste Test of Trade Liberalization.” Center for In- 139 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 ternational Economics, Canberra and Sydney, Other Standards.” Mimeo. World Bank, Wash- Australia, September. ington, D.C. Cato, James C., and C. A. L. Dos Santos. 2000. “Costs Jansen, H. Tarp, Sherman Robinson, and Finn Tarp. to Upgrade the Bangladesh Frozen Shrimp Pro- 2002. “General Equilibrium Measures of Agricul- cessing Sector to Adequate Technical and Sanitary tural Policy Bias in Fifteen Developing Coun- Standards.” In L. J. Unnevehr, ed., The Econom- tries.” TMD Discussion Paper 105. International ics of HACCP. Saint Paul, Minn.: Eagan Press. Food Policy Research Institute, Washington, D.C. DRIFE (Danish Research Institute of Food Economics). Kalaitzandonakes, N. G., and M. E. Bredahl. 1994. 2003. “Note on the Harbinson Draft on Modali- “Protectionism, Efficiency, and Productivity ties in the WTO Agriculture Negotiations.” Agri- Growth.” In M. Hartmann, P. M. Schmitz, and cultural Policy Research Division, Frederiksberg, H. von Witzke, eds., Agricultural Trade and Denmark. Economic Integration in Europe and in North De Gorter, H. 2003. “Agricultural Support Programs America. Kiel, Germany: Wissenschaftsverlag and Farm Household Incomes.” Mimeo. World Vauk Kiel KG. Bank, Washington, D.C. Kudat A., M. I. Ajwad, and B. Sivri. 2003. “A Synthe- Diop, Ndiame, John Beghin, and Mirvat Sewadeh. sis of Trade and Poverty Country Studies in Diag- 2003. “Groundnut Policies, Global Trade Dy- nostic Trade Integration Studies.” Mimeo. World namics, and the Impact of Trade Liberalization.” Bank, Washington, D.C. Mimeo. World Bank, Washington, D.C. March. Lindland, J. 1997. “The Impact of the Uruguay Round Federated Farmers of New Zealand. 2002. “Life after on Tariff Escalation in Agricultural Products.” Subsidies: The New Zealand Farming Experience Food Policy 22: 487–500. 15 Years Later.” Wellington, New Zealand. Minot, Nicholas, and Lisa Daniels. 2002. “Impact of Gallezot, J. 2003. “La progressivité tarifaire de l’UE vis Global Cotton Markets on Rural Poverty in à vis des produits agricoles et agroalimentaires.” Benin.” International Food Policy Research Insti- Mimeo. Institut National de Recherche Agrono- tute, Washington, D.C. November. mique, Unité Mixte de Recherche Économie Minot, Nicholas, M. Kherallah, and P. Berry. 2000. Publique, Paris-Grignon. January. “Fertilizer Market Reform and the Determinants Golub, S., and J. Michael Finger. 1979. “The Process- of Fertilizer Use in Benin and Malawi.” MSSD ing of Primary Commodities: Effects of Devel- Discussion Paper 40. International Food Policy oped-country Tariff Escalation and Developing Research Institute, Washington, D.C. October. Country Export Taxes.” Journal of Political Mitchell, Donald. 2003. “Sugar Policies: Opportunity Economy 83 (3): 559–77. for Change.” Mimeo. World Bank, Washington, Henson, Spencer, Ann-Marie Brouder, and Winnie Mi- D.C. March. tullah. 2000. “Food Safety Requirements and Morrison Paul, Catherine J., Warren Johnston, and Food Exports from Developing Countries: The Gerald Frengley. 2000. “Efficiency in New Case of Fish Exports from Kenya to the European Zealand Sheep and Beef Farming: The Impacts of Union.” American Journal of Agricultural Eco- Regulatory Reform.” Review of Economics and nomics, No. 5. Statistics 82 (2): 325–37. Hoekman, Bernard, Francis Ng, and Marcelo Olar- Nogues, Julio J. 2002. “Comment to ‘Trade, Growth, reaga. 2002. “Reducing Agricultural Tariffs ver- and Poverty—A Selective Survey,’ by Andrew sus Domestic Support: What’s More Important Berg and Anne Krueger, and ‘Doha and the World for Developing Countries?” Mimeo. World Bank, Poverty Target,’ by L. Alan Winters.” Commen- Washington, D.C. September. tary presented at the Annual World Bank Confer- Ingco, M. 1997. “Has Agricultural Trade Liberaliza- ence on Development Economics, May. tion Improved Welfare in the Least-Developed Nogues, Julio J. 2003. “Agricultural Exports in a Pro- Countries? Yes.” Policy Research Working Paper tectionist World: Assessing Trade Strategies for 1748. World Bank, Washington, D.C. MERCOSUR.” Paper prepared for the Trade and Integrated Framework. 2003. “Malawi Diagnostic In- Integration Division of the Inter-American Devel- tegration Study, Integrated Framework Diagnos- opment Bank. tic.” Trade Integration Studies, Washington, OECD (Organisation for Economic Co-operation and D.C., May. Development). 1999. “Distributional Effect of Jaffee, Steven. 2003. “From Challenge to Opportunity: Agricultural Support in Selected OECD Coun- The Transformation of Kenyan Vegetable Trade tries.” Directorate for Food, Agriculture, and in the Context of Emerging Food Safety and Fisheries, Committee for Agriculture, Paris. 140 A G R I C U L T U R A L P O L I C I E S A N D T R A D E _____. 2001. “The Uruguay Round Agreement on ume 4 of A Synthesis of the Economics in De- Agriculture: an Evaluation of Its Implementation veloping Countries. Baltimore: Johns Hopkins in OECD Countries.” Paris. University Press. _____. 2002a. “A Positive Agenda.” Directorate for Sheales, T., S. Gordon, A. Hafi, and C. Toyne. 1999. Food, Agriculture, and Fisheries, Committee for “Sugar International Policies Affecting Market Agriculture. Paris. Expansion.” ABARE Research Report 99.14. _____. 2002b. “Methodology for the Measurement of Australian Bureau of Agricultural and Resource Support and Use in Policy Evaluation.” Paris. Economics, Canberra, Australia. November. _____. 2002c. “Agricultural Policies in OECD Coun- USDA (U.S. Department of Agriculture). 2002. “Agri- tries: Monitoring and Evaluation.” Paris. cultural Income and Finance Outlook.” Washing- _____. 2002d. “Farm Household Income Issues in ton, D.C. September 26. OECD Countries: A Synthesis Report.” Direc- Von Braun J., and E. Kennedy. 1994. Agricultural torate for Food, Agriculture, and Fisheries, Com- Commercialization, Economic Development, and mittee for Agriculture. Paris. Nutrition. Baltimore: Johns Hopkins University _____. 2002e. “The Incidence and Income Transfer Ef- Press. ficiency of Farm Support Measures.” Directorate Wailes, Eric. J. 2003. “Global Trade and Protection for Food, Agriculture, and Fisheries, Committee Regimes in Rice Trade.” Mimeo. University of for Agriculture. Paris. Arkansas, Fayetteville. April. Quiroz, J., and L. Opazo. 2000. “The Krueger-Schiff- Watkins, Kevin. 2003. “Northern Agricultural Policies Valdés Study 10 Years Later: A Latin American and World Poverty: Will the Doha ‘Development Perspective.” Economic Development and Cul- Round’ Make a Difference?” Annual World Bank tural Change 49 (1): 181–96. Conference on Development Economics. Paris, Rae, A., and T. Josling. 2003. “Processed Food Trade May 15–16. and Developing Countries: Protection and Trade World Bank. 2000. Global Economic Prospects and Liberalization.” Food Policy 28: 147–66. the Developing Countries 2001. Washington, Ravallion, Martin. 2000. “On the Urbanization of D.C.: World Bank. Poverty.” Mimeo. World Bank, Washington, D.C. _____. 2001. Global Economic Prospects and the De- Rosengrant, M. W., and S. Meijer. 2002. “Agricultural veloping Countries 2002. Washington, D.C. Trade Liberalization to 2020: Impact on Trade, WTO (World Trade Organization). 2003. “Negotia- Price, and Economics Benefits.” International tions on Agriculture First Draft of Modalities for Food Policy Research Institute, Washington, D.C. the Further Commitments.” WTO Document No: September. TN/AG/W/1/Rev.1. Geneva. Schiff, Maurice, and Alberto Valdés. 1992. The Politi- cal Economy of Agricultural Pricing Policy. Vol- 141 4 Labor Mobility and the WTO: Liberalizing Temporary Movement Globalization is driving the movement more readily and rapidly. Temporary move- of people across borders ment, in particular by highly skilled workers, With globalization—the dramatic expansion of has seen the largest growth in the past decade. cross-border trade and investment—has come an upsurge in international labor mobility. Both rich and poor countries can benefit Falling costs of transportation and communica- Both developed and developing countries have tion have reduced the distances between peo- much to gain from an increased flow of work- ples, and the drive for better lives has motivated ers. Rich countries benefit because they gain workers to move to areas where jobs are more workers whose skills are in short supply. Also, plentiful and pay is better. Foreign-born persons as demographics drive up the average age in now account for 10 percent of the total popula- rich countries, migration allows an influx of tion in the United States, 5 percent in Europe, younger workers who contribute to pension and 1 percent in Japan. In Canada and Aus- systems that would otherwise be actuarially tralia, foreign-born persons represent 17 and 24 unviable. Poor countries gain from higher percent of the total population, respectively.1 wages as well as from the remittances that Even so, today’s movement of people is still accrue from migration. In 2001, worker re- well below levels experienced in the late nine- mittances alone provided some $70 billion to teenth century, and migration rates, now ham- developing countries, nearly 40 percent more pered by restrictive policies, are well below than all development assistance and signifi- cross-border flow of goods and investment. By cantly more than net debt flows to developing 2000, according to the United Nations, 175 countries. Returning workers also often bring million persons were living outside their coun- new skills back to the sending country. To be try of birth—about 3 percent of the world’s sure, there are costs to both receiving and send- population. By contrast, global exports of ing countries: labor markets and social services goods reached almost a third of GDP, and fi- may be strained in the rich countries, and de- nancial flows were well above 10 percent veloping countries may lose skilled workers (OECD 2001c; World Bank 2003; United Na- who have been educated with public resources. tions 2000). Nonetheless, if a temporary visa system were While long term and settlement migration introduced in rich countries permitting move- are still predominant in most developed coun- ment of labor up to 3 percent of the total labor tries, migrant flows are now more diverse and force, world incomes would rise by nearly complex, with migrants moving back and forth $160 billion (Walmsley and Winters 2002). 143 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 The GATS could facilitate temporary to maintain immigration and labor market movements policy flexibility, countries have made GATS Temporary movement of certain types of work- commitments far below the degree of TMNP ers—service suppliers—is included under the access already afforded under domestic laws World Trade Organization (WTO) General and regulations. An important corollary of this Agreement on Trade in Services. This is de- tension is that the extent of TMNP liberali- signed to facilitate the movement of people in a zation for some sectors and categories of way analogous to the movement of goods and workers where labor demand routinely exceeds capital. This type of temporary movement— supply (for example, in tourism, information called Mode 4 in the GATS—is treated as other technology, and medical-related services) may services in the global negotiations. They allow be significantly greater than in other categories countries to negotiate fixed limits accorded to of labor, particularly unskilled labor. all foreign workers on a most-favored-nation A second tension stems from the fact that (MFN) basis. Some developing countries see the strong regional character of migration pat- temporary movement under GATS Mode 4 as terns creates domestic political support for their key interest in services trade and are ex- programs that favor neighboring countries, pecting real progress in the context of the Doha while Mode 4 commitments necessarily are Development Agenda negotiations. open to all countries on an MFN basis. Pref- erential migration schemes are commonly ne- However, progress has been minimal gotiated at the bilateral and regional levels, because of policy restrictions and MFN-based liberalization would under- To date, however, even judging by the relatively mine these. Because the many bilateral labor limited liberalization of trade in services during agreements are usually untied to trade policy the Uruguay Round, little has been done to or other agreements, they afford governments loosen conditions governing the temporary a greater degree of flexibility to adjust pro- movement of natural persons supplying services grams to evolving migration trends and labor- (Mode 4). Mode 4 today accounts for less than market needs. 2 percent of the total value of services trade. While the potential gains from increasing Present commitments refer almost exclusively temporary labor mobility, including for ser- to higher level personnel. More than 40 percent vice suppliers under GATS Mode 4, could be of Mode 4 commitments are for intracorporate sizeable, the analysis presented in this chapter transferees whose mobility is intimately related cautions that expectations of far-reaching for- to foreign direct investment; another 50 percent ward movement need to be tempered because of commitments cover executives, managers of the political sensitivity of such trade in re- and specialists, and business visitors. All this ceiving countries. That sensitivity has become means that the Mode 4 liberalization achieved more pronounced in the context of decelerat- to date has been of limited significance for de- ing worldwide economic growth and height- veloping countries whose comparative advan- ened security concerns. tage lies in the export of medium- and low- skilled, labor-intensive services. Expanding Mode 4 requires changes Two fundamental tensions hamper progress to realize its modest potential on Mode 4 labor mobility. The first is that gov- Tensions notwithstanding, present levels of ernments are reluctant to undertake perma- Mode 4 access fall far short of even its rela- nent commitments when employment demand tively modest potential. One possible response varies with cyclical conditions, and when sev- is for developing countries to actively expand eral OECD countries are facing difficulties in their requests and offers in the Doha Round. integrating existing immigrant communities Also, WTO members could adopt rules that into their labor market and societies. Wanting would provide greater clarity and predictabil- 144 L A B O R M O B I L I T Y A N D T H E W T O : L I B E R A L I Z I N G T E M P O R A R Y M O V E M E N T ity. And to help regularize entry and exit while together have accounted until recently for up to ensuring improved security, countries could half of Africa’s migratory flow. Almost every- adopt a GATS visa system that would facili- where, most migrants tend to stay within their tate national visas for up to one year, subject regions, reflecting the importance of culture to appropriate checks and strict rules of and language and the lower costs associated administration. with geographical proximity. Around the world, migration is on the rise The bigger picture: Global Five forces have governed world migration migration and remittance trends since the mid-nineteenth century: A lthough on an upward trend over the last two decades, migration is still far below its historic peak. The greatest migratory flows • Wage and opportunity gaps between rich and poor countries took place between the middle of the nine- • Regional conflicts and political instabil- teenth century and the onset of World War I, ity in developing countries when an estimated 10 percent of the world’s • The relative share of young adults in labor force relocated permanently across bor- the populations of sending and receiving ders (World Bank 2001). Mass migration was countries a major factor in equalizing incomes across • Numbers of migrant stock residing in re- countries throughout this period, by some ceiving countries estimates exerting a greater influence than ei- • Reductions in the cost and inconvenience ther trade or capital movements (Lindert and of travel. Williamson 2001). Since World War II, globalization has led These forces are still driving South-North, to more unrestricted movement of both goods and South-South, migration. Successful devel- and capital, while international policies toward opment and poverty eradication in the develop- migration have become more restrictive. As a ing world almost certainly will release part of result, the overall scale of labor migration re- the poverty constraint on potential emigrants mains relatively smaller than that of capital or while simultaneously reducing the motivation trade flows. Only 3 percent of the world’s pop- of many to move. In regions where development ulation—some 175 million people—live outside has been slower and poverty more obstinate, their country of citizenship, and the number rising populations, dwindling opportunities, of permanent legal immigrants to the United and lower travel costs will combine to impel States is less today than it was in 1914, both in emigration. The shrinking share of young adults absolute terms—850,000 vs. 1.2 million—and in the developed countries, particularly in Japan as a percentage of the total population—0.35 and Western Europe, and the rising share of percent vs. 1.5 percent. By contrast, global ex- young people in South Asia, Africa, and other ports of goods represent almost one-third of parts of the world are complementary drivers of world GDP (World Bank 2003; OECD 2001c). labor movement. Growing numbers of young While South-North migratory flows receive people in the developing world have acquired the most attention, much cross-border labor the education and training needed to assume mobility—representing roughly half of the total skilled positions in developed economies. And number of migrants—takes place between de- as the numbers of the foreign-born grow in de- veloping countries. While poorly measured and veloped countries, their presence makes it easier less well understood than flows into the North, and more attractive for newcomers to join them some patterns are evident: South Asians typi- (Hutton and Williamson 2002). cally travel to the Middle East and East Asia, Wage differentials remain high. The average while South Africa, Nigeria, and Côte d’Ivoire hourly wage in manufacturing is about $30 in 145 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Germany, while in some parts of China and skilled labor (box 4.1). In some service occu- India it is only 30 cents. Between the United pations, particularly those most directly re- States or France and newly emerging countries lated to population aging (medical care and such as Thailand or Malaysia the gap is ten- associated personal care services), where there fold. Meanwhile, the supply of labor is swelling is no substitute for human labor, the demand in developing countries—particularly in South for—and benefits of—movement of lower- Asia and Africa, where poverty is concentrated. skilled labor are likely to continue to increase Each year 83 million people are added to the (Winters 2003). world’s population, 82 million of them in the The foreign and foreign-born make up a developing world (World Bank 2001). growing share of the population of most The continuing demographic transition in major industrial countries, rising over the last industrial countries adds to these pressures. decade from 4.6 to 5.4 percent in the Euro- As their populations age and average levels of pean Economic Area, and from 7.9 to 10.4 training and education rise, developed coun- percent in the United States (table 4.1) Be- tries face a declining ratio of workers to re- cause the population of developing countries tirees and an increasing scarcity of lower- is about five times greater than that of devel- Box 4.1 Population aging and migration T he combined demographic effects of the baby boom that marked the immediate post-war period, the fall in fertility rates that began in guard balance and equity in the systems of social protection—decisions related to the length of work- ing life, levels of contributions and benefits, and pro- OECD countries in the late 1960s, and longer life ductivity advances. expectancy have led to a striking acceleration in One solution receiving increased consideration population aging in virtually all advanced industrial in several countries is to increase levels of permanent societies. immigration to modify population structures and Population aging is much more marked in mitigate the social and economic costs of aging. Im- Europe and Japan than in North America, but all migration has advantages. It can quickly increase the three regions will be affected. According to demo- economically active population because new immi- graphic projections by the United Nations, the grants tend to be younger and more mobile. Also, populations of the European Union and Japan are fertility rates among immigrant women are often expected to fall by 10 percent and 14 percent, re- relatively high, which can help boost population spectively, between 2000 and 2050, representing a growth. This has only a limited impact in the short decline of some 55 million in all. In both Japan and run, however. the European Union, the dependency ratio (defined Immigration alone cannot provide the answer to as the ratio of pensioners to workers) is expected to population aging, as demonstrated by simulations decline from five to one today to three to one in produced by the United Nations (2000). The simula- 2015. For the United States, projections still point to tions show that maintaining steady dependency ra- an increased total population over the same period, tios until 2050 would require an enormous increase but the dependency ratio also rises. in migration flows—for the United States and the Eu- Recent research has considered the economic ropean Union, migration balances would have to be and fiscal impact of these demographic trends in the at least 10 times the annual averages of the 1990s. OECD area (OECD 2000, 2001c, 2002; Visco Such scenarios seem implausible by historical stan- 2000). Without offsetting measures, the growing de- dards, and in light of the likely political reactions. pendency could place enormous strains on social se- curity, Medicare, and pensions systems. Far-reaching decisions are required over the medium and long term to meet shifting labor demands and to safe- Source: OECD (2001f). 146 L A B O R M O B I L I T Y A N D T H E W T O : L I B E R A L I Z I N G T E M P O R A R Y M O V E M E N T Table 4.1 Migration is rising in many OECD countries Migration flows and stocks of foreign and foreign-born population in OECD countries, annual averages, 1990–2000 (thousands, except where otherwise noted) 1990–94 1995–99 2000 Immigration Australia Permanent 99 87 92 Temporary 104 154 224 Canada Permanent 236 204 227 Temporary workers a 64 69 86 European economic area b 1,614 1,352 1,426 Japan 244 251 346 United States Permanent 1,209 747 850 Temporary c 1,357 1,893 2,741 Net migration per thousand inhabitants Australia 4.3 5.1 5.4 Canadad .. 5.4 5.1 European economic areae 3.1 1.7 2.5 Japan –0.03 –0.04 0.3 United States 3.3 3.3 3.1 Asylum seekers Australia 9 9 12 Canada 30 26 36 Central and Eastern Europe 3 13 26 European economic area 516 326 427 United States 136 105 57 Acquisitions of nationality Australia 107 102 80 Canada 130 160 205 European economic areaf 460 690 720 Japan 12 16 18 United States 315 680 900 1990 (percent) 2000 (percent) 2000 (thousands) Stock of foreign population European economic areag 4.6 5.4 20,381 Japan 0.9 1.3 1,686 Stock of foreign-born population Australia 22.8 23.6 4,517 Canadah 16.1 17.4 4,971 United States 7.9 10.4 28,400 .. negligible a. Inflows of foreign workers entering Canada to work temporarily (excluding seasonal workers) provided by initial entry. b. Includes Austria, Greece, Italy, and Spain. No 2000 data for Denmark available; 1999 data substituted. c. Excluding visitors, transit migrants, foreign government officials, and students. d. Fiscal years (July to June). e. Data relate to 1999–2000 average instead of 2000. f. Excluding Greece and Ireland. g. Excluding Greece. No 2000 data available for France; 1999 data substituted. h. Data are for 1991 instead of 1990 and for 1996 instead of 2000. Source: OECD (2002f). oped countries, migrants comprise a larger historical and institutional backgrounds in var- share of the total population in rich countries ious countries. Some countries, such as Aus- (6 percent) than in poor countries (1 percent). tralia, Switzerland, and the United Kingdom, The uneven composition of immigration explicitly give priority to foreign workers, so flows reflects differing policy objectives and that this group accounts for around half of all 147 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 immigration. Other OECD countries, because costs and thus greater access to information they tend to restrict work-related migration, on migration opportunities via global media, implicitly give priority to nonselective migra- the Internet, and diaspora networks in receiv- tion arising from family reunification (which ing countries are breaking down barriers to mi- accounts for approximately 80 percent of flows gration (Nielson 2002). into the United States and France, for example) or requests for asylum (approximately half in Remittances by migrants are an important the Nordic countries) (OECD 2002f). source of income for many developing Because legal immigration is restricted, ille- countries gal migration has risen noticeably in recent Remittances from foreign workers, both per- years, as have trafficking in human beings and manent and temporary, are the second-largest expenditures to combat both phenomena. Ille- source of external funding for developing gal migration into the European Union soared countries, after foreign direct investment tenfold in the 1990s, reaching half a million (FDI). In 2001, workers’ remittances to devel- people annually by the end of the decade. In oping countries stood at $72.3 billion, consid- the United States, an estimated net inflow of erably higher than total official development 300,000 undocumented workers occurs each assistance and private non-FDI flows, and year, although this could well underestimate 42 percent of total FDI flows to developing the actual scale of illegal migration. countries that year (table 4.2). For most of the Newer factors are compounding the more 1990s, remittance receipts exceeded official familiar drivers of migration. The developing development assistance (World Bank 2003). world’s rising share of educated workers—those As with actual movements of people, the who have completed secondary education— data on payments are susceptible to measure- has jumped from one-third to nearly one-half ment problems—not all flows, even from legal over the past three decades. Increasingly, the migration, are captured in the balance of pay- growing pool of skilled developing-country ments accounts, and in situations where sub- labor is meeting industrial-country shortages, stantial illegal migration occurs, the bulk of as the marketplace for skills widens to encom- the international resource flows also may be pass the entire globe. Meanwhile, continued missed. Such difficulties notwithstanding, ini- declines in transportation and communication tial estimates of these flows can be derived Table 4.2 Workers’ remittances are the second-largest source of external funding for developing countries Remittance receipts and payments by developing countries in 2001 (billions of dollars) Lower middle Upper middle All developing Low income income income Total remittance receipts 72.3 19.2 35.9 17.3 As percentage of GDP 1.3 1.9 1.4 0.8 As percentage of imports 3.9 6.2 5.1 2.7 As percentage of domestic investment 5.7 9.6 5.0 4.9 As percentage of FDI inflows 42.4 213.5 43.7 21.7 As percentage of total private capital inflows 42.9 666.1 44.9 20.2 As percentage of official development assistance 260.1 120.6 361.7 867.9 Other current transfers a 27.2 6.1 14.0 7.1 Remittances and other current transfers 99.5 25.3 49.9 24.4 Total remittance payments 22.0 1.2 1.7 19.1 Excluding Saudi Arabia 6.9 1.2 1.7 4.0 a. Other current transfers include gifts, donations to charities, pensions received by currently retired expatriate workers, and so on. They also may include personal transfers by migrant workers to families back home. Source: World Bank (2003). 148 L A B O R M O B I L I T Y A N D T H E W T O : L I B E R A L I Z I N G T E M P O R A R Y M O V E M E N T Figure 4.1 Workers’ remittances are an important source of income for many developing countries Top developing-country recipients of workers’ remittances, 2001 (billions of dollars and percent of GDP) India Lesotho Mexico Jordan Philippines Albania Morocco Nicaragua Egypt, Arab Rep. Yemen, Rep. Turkey Moldova Lebanon Lebanon Bangladesh El Salvador Jordan Cape Verde Dominican Republic Jamaica El Salvador Yugoslavia, FR (Serb./Mont.) Colombia Morocco Yemen, Rep. Pakistan Dominican Republic Brazil Vanuatu Ecuador Philippines Yugoslavia, FR (Serb./Mont.) Honduras Thailand Uganda China Ecuador Sri Lanka Sri Lanka 12 10 8 6 4 2 0 0 5 10 15 20 25 30 Source: World Bank (2003). from balance of payments statistics by com- the Caribbean were the largest recipient of bining workers’ compensation (transfers re- remittances in nominal terms in 2001, but rel- lating to work abroad of less than one year) ative to the size of GDP, South Asia was the and workers’ remittances (transfers made by largest recipient, with remittances of nearly 2.5 workers whose stay abroad exceeds one year) percent of GDP. Remittance flows to countries (World Bank 2003). in Sub-Saharan Africa also were significant, ac- In nominal terms, the top recipients of re- counting for 1.3 percent of GDP (table 4.3). mittances included several large developing Workers’ remittances are spread more economies—India, Mexico, and the Philip- evenly among developing countries than are pines—although as a share of GDP, remit- capital flows—the top 10 recipient countries in tances were larger in other low-income coun- 2001 received 60 percent of total remittances tries in 2001 (figure 4.1). Broken down along to developing countries as compared with a regional lines, countries in Latin America and top 10 share of 68 percent of GDP, 72 percent Table 4.3 Remittances are a significant source of income in all regions of the developing world Workers’ remittances received by developing countries, by region, 1999–2002 (in billions of dollars and as percentage of GDP) 1999 2000 2001 2002 (billions of dollars) $ billions % GDP $ billions % GDP $ billions % GDP $ billions % GDP Total 67 1.2 66 1.1 72 1.3 80 1.3 East Asia and Pacific 11 0.7 10 0.7 10 0.6 11 0.6 Europe and Central Asia 8 0.9 9 0.9 9 0.9 10 1.0 Latin America and Caribbean 17 1.0 19 1.0 23 1.2 25 1.5 Middle East and North Africa 12 2.2 11 1.9 14 2.3 14 2.2 South Asia 15 2.6 13 2.3 14 2.3 16 2.5 Sub-Saharan Africa 4 1.3 3 0.8 3 1.0 4 1.3 Sources: IMF, Balance of Payments Yearbook; World Bank, World Development Indicators (2001). 149 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 of exports, and 74 percent of FDI. Remittances temporary mobility, the picture is not a simple also are more stable than private capital flows, one. Developed countries are major exporters, which tend to move in a pro-cyclical manner. as well as importers, of labor. Similarly, some developing countries are significant importers. By some value indicators—for example, com- Temporary movement of workers pensation of employees—developed countries M any people move only temporarily— students, tourists, business visitors ex- ploring or conducting trade and investment account for most of the flow in both direc- tions. On the other hand, developing countries are the major receivers of remittances (see fig- activities, and people working abroad under a ure 4.1). Temporary movement is by no means range of schemes. People working or conduct- unidirectional. Relative to the overall size of ing business are thus a subset of temporary labor markets, the number of temporary for- movement, and GATS Mode 4, temporary eign workers remains small for most countries, movement of natural persons as service sup- except for the Arab states of the Persian Gulf. pliers, is a further subset of this group. The areas of highest growth are short-term Most of the developed economies experi- movements (from three to six months) (OECD enced significant growth in certain types of 2001b). Movements of both skilled and un- temporary migration during the 1990s (table skilled workers appear to be concentrated in 4.4). In the United States, for example, the av- the service sectors of major receiving countries, erage number of temporary immigrants per notably in construction, commerce, catering, year doubled between 1990–94 and 2000, at education, health care, services to households, which point the total was more than three times and other services. In developing countries, larger than permanent immigrants (see table foreign workers tend to be concentrated in pri- 4.1). mary activities (agriculture, fishing, and min- The absence of global figures on temporary ing) as well as in manufacturing, although the foreign workers and the limitations of existing share in services (particularly tourism-related) migration data2 make analysis difficult and de- is rising in several countries (UNCTAD 2001). finitive conclusions impossible. However, ac- Although the volume of global trade repre- cording to the OECD (2001), some trends sented by temporary foreign workers remains have begun to emerge. small compared to overall trade in goods and Although available statistics are insufficient services, it is very important for some indus- to identify conclusively the primary traders in tries and for some countries. Indeed, exports Table 4.4 Temporary movement is rising in rich countries Entries of temporary foreign workers in selected OECD countries, 1992–2000 (thousands) 1992 1993 1994 1995 1996 1997 1998 1999 2000 Australia 40.5 14.9 14.2 14.3 55.7 81.7 92.9 99.7 115.7 Canada 70.4 65.4 67.5 69.5 71.5 75.4 79.5 85.4 93.7 France 18.1 — — — 13.6 12.9 11.8 13.4 15.4 Germany 332.6 69.1 53.9 61.7 271.0 267.7 244.0 274.1 331.6 Italy — — — — — — — 18.7 24.52 Japan — — — — 124.1 143.5 151.7 156.0 183.9 Korea, Rep. of 8.3 12.4 33.6 47.0 81.4 105.0 75.4 111.0 122.5 Sweden — — — — — — — 15.0 19.4 Switzerland 127.8 — — — 63.4 47.4 40.3 46.1 50.3 United Kingdom 57.6 25.9 26.3 32.9 78.7 89.7 98.8 107.9 134.1 United States 143.0 112.5 130.7 147.5 191.2 — 342.7 422.5 505.1 — Not available. Note: Definitions vary among countries and the figures are not strictly comparable. Source: OECD (2002f). 150 L A B O R M O B I L I T Y A N D T H E W T O : L I B E R A L I Z I N G T E M P O R A R Y M O V E M E N T Box 4.2 Temporary labor movement and the East Asian crisis of 1997–98 I ncreased labor migration, particularly the tempo- rary movement of unskilled labor, was an impor- tant dimension of structural change and globaliza- to undertake so-called 3-D jobs (difficult, dirty, and dangerous) shunned by nationals in export-oriented industries has helped economic recovery. The large tion within much of Northeast and Southeast Asia in real gains to workers who migrate to more developed the 1980s and 1990s. Conservative estimates suggest countries were documented in a recent study: un- that the number of migrants doubled or even tripled skilled Indonesian workers can earn $2 or more per in most net labor importing countries from the early day in neighboring Malaysia compared to 28 cents 1980s to the onset of the financial crisis in 1997 per day at home (World Bank 2001). (Athukorala and Manning 1999). By the onset of The migration of skilled and professional work- the crisis, some 2 million overseas migrant workers ers also was an important part of the international- were employed in Northeast Asian economies out- ization of East Asian labor markets during the side China, and an even larger number—some 3 to 1990s. FDI and associated trade flows resulted in sig- 4 million—worked in Southeast Asia. nificant skilled migration from developed to develop- The Asian financial crisis changed the context ing countries—as well as considerable out-migration within which those labor movements occurred, pos- of skilled and professional migrants from countries ing a major threat to economic growth. International in the region. Workers from the Philippines domi- capital flows reversed, many firms went bankrupt, nated the latter flows, reflecting both slower rates of and unemployment rates rose steeply in the crisis- FDI and economic growth domestically, and a mod- affected countries (World Bank 2000). Demand for ern sector incapable of absorbing enough graduates labor plummeted in the modern sectors of labor- from the well-developed education sector. receiving countries. Yet a recent examination of Several countries have sought to develop a more labor mobility patterns before and after the financial coherent migration policy in light of the social effects crisis suggests that the economic turmoil did not of the cutback in labor demand and unemployment alter the fundamental conditions underpinning high that accompanied the crisis. This may be good news levels of intraregional migration, especially of un- for many migrant workers who need protection skilled contract workers (Manning 2002)—a widen- against exploitation. In-migration of unskilled over- ing wage gap and imbalances between supply and seas workers, meanwhile, can be expected to con- demand for labor across East Asia. tinue to support production in both tradable and Countries in the region—especially the net labor nontradable industries in the more affluent East importers: Hong Kong (China), Malaysia, Republic Asian economies. of Korea, and Thailand—have continued to rely on inflows of migrant workers. Many such migrants are unskilled. Indeed, the willingness of migrant workers Source: World Bank staff. of labor services in many developing countries and vertically through relocation or so-called account for a substantial share of output and greenfield investments—accounts for much trade. In several developing countries (mainly new temporary movement, helped along by lower income), such flows dwarf the value of advances in transportation, communications, total services exports (OECD 2001b). More- logistics, and organization that have altered all over, migration flows are resilient—for exam- phases of the business process. Today’s compa- ple, they were much more stable than capital nies not only can but must respond quickly to flows in the wake of the recent Asian financial emerging opportunities by forming specialized crisis (box 4.2). task teams, regardless of where the personnel Burgeoning cross-border investment—hori- are based. Business functions previously han- zontally through mergers or joint ventures dled by expatriate staff, resident representa- 151 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Table 4.5 Foreign-born workers meet skill ogy, but also in education and health-related shortages in rich countries services, which cannot be met quickly enough Percentage of foreign workers in selected sectors, 2001 by domestic training programs and institutions Health (OECD 2001b) (table 4.5). Recent growth in Information professions Education technology (except nurses) the movement of highly skilled workers has been supported by specific programs designed Austria 3.0 28.3 6.3 to address key national shortages (box 4.3). Belgium 3.3 6.0 4.0 France 2.8 4.2 1.7 Germany 2.8 5.7 4.4 Italy 0.3 0.4 1.0 Bilateral and regional approaches Netherlands Norway 1.6 4.3 2.8 4.3 1.4 5.7 to labor mobility Switzerland United Kingdom United States 13.0 5.9 19.3 6.4 16.5 7.8 R egional trade agreements, which include provisions on labor mobility, range from a few that provide no-cost movement of all types (foreign-born) 8.3 18.3 13.2 of workers and service providers—although Source: OECD (2002f). with caveats that allow exclusions from certain public services and restrictions based on public tives, or correspondent firms now may be ac- health and safety—to others, the majority, that complished by expert personnel on temporary target only intracorporate transferees and busi- assignments. Shorter product lifecycles, higher ness visitors.3 Some—such as the EU, EEA, and customer expectations, and stiffer competition CARICOM—allow a relatively high degree of force companies to be ready to send expert freedom of movement with few special pro- personnel abroad on short notice—or to bring cedures. Others, notably the North American them to the firm’s home country for specialized Free Trade Association (NAFTA), allow for tasks in engineering, production, and market- some regulated mobility and involve relatively ing. Companies entering new markets often detailed special procedures implemented among wish to bring experienced staff from other lo- a few parties. Still others (such as APEC) are cations to assist with the establishment phase aimed at facilitating existing mobility, adding (OECD 2001b). Indeed, among those tem- some special procedures, but retaining maxi- porary foreign workers who would fall under mum flexibility to continue existing national GATS Mode 4, the mobility of intracompany practices (OECD 2002b) (box 4.4). transferees—key personnel involved in the The differing approaches to labor mobility establishment or operation of enterprises in regional trade agreements reflect a range of abroad—has shown the fastest growth. factors, including the degree of geographical Investment liberalization can also create a proximity of the parties and the extent of simi- demand for exports of skilled labor from the larities in their levels of development, as well as host country. Foreign firms in India, for ex- other cultural and historical ties. Agreements ample, have become aware of the skills avail- among countries enjoying geographic proxim- able in India and are now meeting demand for ity and similar levels of development generally various services in their home countries with adopt a more liberal approach to labor mo- supply from India (OECD 2001b). The grow- bility (EU, EFTA, EEA, Trans-Tasman Travel ing importance of South-South FDI is an Arrangement) as compared to agreements important vector of more highly skilled tem- among geographically distant members of dif- porary labor movement between developing fering levels of development (APEC, U.S.- countries (World Bank 2003). Jordan). But this is not always the case (MER- Temporary labor is used to meet skills COSUR, SAARC) (OECD 2002b). shortages in developed countries, particularly The range of special provisions found in var- in information and communication technol- ious regional trade agreements, but not in the 152 L A B O R M O B I L I T Y A N D T H E W T O : L I B E R A L I Z I N G T E M P O R A R Y M O V E M E N T Box 4.3 Recent initiatives to facilitate temporary movement of highly skilled workers A ustralia has established a number of business- sponsored temporary entry programs, supported by business service centers for employers seeking Japan announced a plan in November 2000 to recruit 30,000 skilled IT engineers and researchers from overseas by 2005. In the United Kingdom high- skilled foreign workers. Canada initiated a pilot pro- volume nonimmigrant visa employers with a proven gram related to software development workers under track record receive streamlined and fast-tracked visa which Human Resources Development Canada pre- approval. Simplified fast-track procedures are now identified a general need within the labor market for applied for issuing work permits for certain occupa- such workers. This enabled suitably qualified appli- tions, and the list of occupations susceptible to labor cants with a job offer from a Canadian employer shortages has been extended. The maximum length and any necessary visa (depending upon country of of a work permit also has been extended from four origin) to be automatically validated (that is, not to five years. The United States raised the annual subjected to labor market needs tests). Under a pilot quota of H-1B visas for professional and skilled project, spouses of “highly skilled foreign workers” workers by nearly 70 percent in 2000, providing who are admitted to Canada for at least six months temporary admission for 195,000 people over the also are permitted employment authorizations with- next three fiscal years. The 7 percent ceiling on the out being subjected to labor market testing. proportion of visas going to nationals of any given France published a decree in 1998 permitting country also was dropped. Faced with adverse cy- companies to hire foreign workers skilled in com- clical developments in its labor market, the United puter science if the company can demonstrate its in- States recently set the annual quota of H-1B visas at ability to fill the post with a local candidate. Ger- 65,000, the level set prior to the dot com boom of many offered 20,000 employment permits (“green the late 1990s. cards”) for up to five years to computer and infor- mation technology specialists recruited outside the European Union. By August 2000, 13,000 green cards had been issued. Source: OECD (2001b, 2002f). general GATS provisions related to the tempo- or to trade in goods or investment rary movement of service suppliers, includes: (Group of Three) • Specific reference to key personnel in re- • Access to the labor market (EU, EFTA, lation to investment (EU-Mexico, FTAA) EEA, Trans-Tasman Travel Arrangement) • Extension of WTO treatment to non- • Full national treatment and market ac- WTO members (AFTA); and nondiscrim- cess for service suppliers (ANZCERTA) inatory conditions for workers, extend- • Commitments on visas (NAFTA; U.S.- ing beyond service suppliers (Euro-Med) Jordan, which extends the commitment (OECD 2002b). beyond service providers) • Special market access or facilitated ac- To assess the degree of liberalization offered cess for service providers and others in a regional trade agreement, provisions re- (APEC, Canada-Chile, CARICOM, Eu- lated to labor mobility should be considered in rope Agreements, NAFTA) conjunction with provisions in the same agree- • Separate chapters dealing with all tem- ments related to supply of services. Generally, porary movement, including movement the right to labor mobility does not automati- related to investment (Japan-Singapore) cally entail the right to practice a certain pro- 153 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Box 4.4 A trade facilitation approach to labor mobility: NAFTA and APEC F or professionals, “Trade NAFTA” (TN) visas are available to citizens of Canada or Mexico for entry into the United States, provided that the pro- ing it on a best-effort basis. Each government pre- clears applicants (a process that can be customized by each country, such as requiring formal sponsor- fession is on the NAFTA Chapter 16 list, the candi- ship by a business organization). Once home authori- date meets the specific criteria for that profession ties have provided clearance, details of the candidate (typically a university degree in a relevant field of are sent to all participating economies, which must study), the prospective position requires someone in offer a response within two weeks. Economies can that capacity, and the candidate is going to work for refuse clearance for an individual without providing a United States employer.4 TN status lasts for one reason, but this will only restrict travel to that partic- year and is renewable. The requirements for entry to ular economy rather than vetoing the entire applica- the United States differ, however, for Mexican and tion. Following responses from other economies, the Canadian nationals. Canadians are not required to card is issued by the home economy authorities (who have a visa or prior approval, but can receive TN also have the sole right to cancel the card). Fees can status at the port of entry. The candidate must have be charged for issuance of the card and can vary a letter from a U.S. employer offering a job, or re- among participating economies. questing an intracompany transfer, with a job de- The card is valid for three years from the date of scription. For Mexicans, the employer must file a issue and provides multiple short-term business en- labor condition application (I-29 Petition for Non- tries, stays of two or three months on each arrival, Immigrant Workers), and the candidate must apply and access to special immigration processing coun- for a visa at the U.S. Embassy in Mexico. ters on arrival and departure. The cards are the size Anecdotal evidence suggests that industry expe- of a credit card, are manual or machine-readable, rience with TN visas has been positive, with evidence and must contain the signature and photograph of of difficulties confined to some confusion among the cardholder as well as the list of countries for border officials as to how the TN operates and the which entry has been approved. Cardholders are still need for a regularly updated NAFTA professions list. required to present their passports. Separate applica- The APEC Business Travel Card offers accred- tions for visas and work permits are not required. ited business travelers visa-free travel and expedited Additionally, all economies retain the right to refuse airport processing for travelers visiting participating entry to cardholders at the border. economies. After an initial pilot, the scheme was Some 3,400 cards had been issued by mid-2002, made permanent in March 1999. Current partici- with initial assessments indicating that the scheme is pants include Australia, Chile, China, Chinese working effectively, with strong support from the Taipei, Hong Kong (China), Indonesia, Republic of business community and only a small percentage of Korea, Malaysia, New Zealand, and the Philippines. applicants refused. There is no limit on the number Brunei Darussalam, Peru, and Thailand have signed of cards that can be issued. the Operating Framework but are yet to issue cards. The scheme allows considerable flexibility for individual economies. Its Operating Framework is not binding, with members committed to implement- Source: OECD (2001b). fession. National regulations regarding licens- In most agreements, the signatories retain ing and recognition of qualifications are still broad discretion in matters of residency and applied and candidates must meet all criteria visas. Some agreements specify that the agree- and conditions (OECD 2002c).5 ment cannot be invoked to challenge national Few agreements provide immigration rights decisions to refuse entry (Euro-Med), or pro- or supersede national immigration practices. vide remedies only where a pattern of restric- 154 L A B O R M O B I L I T Y A N D T H E W T O : L I B E R A L I Z I N G T E M P O R A R Y M O V E M E N T tive practice can be proved (Canada-Chile, migration in most people. Demographic, edu- NAFTA) (OECD 2002c). cational, and labor market conditions in both Overall, progress in facilitating movement the source and destination countries affect mi- of less-skilled temporary foreign workers has gration decisions, as well as laws and policies not been extensive at the regional level. Indeed, in both countries, information and informa- regional trade agreements tend to replicate the tion flows, chain migration effects (among two key biases found in GATS favoring highly family members or those from the same origin skilled (mostly professional) workers and the area), transport and transaction costs, capital close links between investment and the special- constraints (which may influence potential ized skills such investments require. migrants’ ability or willingness to incur these Additionally, bilateral foreign worker pro- costs), and “exogenous” factors such as civil grams, which have been designed to fill both unrest and climate. There may well be sub- skilled and unskilled labor shortages, have stantial “disutility” costs associated with relo- existed in a number of countries for some cation from one’s social-cultural-linguistic time. Such agreements often cover seasonal context into an alien one. These costs can in workers in agriculture and tourism; project fact be among the most important factors workers in construction; and various other in cross-border migration. The fact that the employment-specific workers. world’s poorest countries supply a very small share of internationally mobile workers lends credence to the oft-observed notion that bind- Understanding the impact of ing poverty constrains out-migration. temporary foreign workers Moreover, migration decisions often are de- picted as essentially on-off and unidirectional What are the determinants of migration? in nature when in practice people migrate for Economic models of migration tend to focus a host of economic and non-economic rea- on the economic incentives facing migrants. sons. They initially may intend to stay tem- In the absence of legal restrictions on immi- porarily and then return or move on to a third gration, cross-border labor mobility is often country, or they may intend to settle. Global- assumed to depend on the size of the gap in ization increases the number and complexity labor and income that existed between in- of these flows. For this reason, some analysts dustrialized and developing countries (wages, prefer to talk about migration and migrants working conditions, social security arrange- rather than immigration and immigrants ments), and on the extent of information on (Home Office 2001). that gap available to potential migrants. Mi- Economic analysis of the temporary move- gration would increase when the gap widened ment of foreign workers straddles the two or when more information on the gap became worlds of trade and migration. This is in light available. Accordingly, it is not rare to see mi- of significant differences in the nature and skill gration described in terms of a pent-up flood: profile of worker categories concerned and by if the tap is opened a little bit, more immi- the sharply differing lengths of time such work- grants will come in; if it is closed, fewer will ers can spend in foreign labor markets. Thus a come. Yet there are strong reasons, rooted in business visitor going abroad to assess future observed trends in international migration, to opportunities or to conclude contract negotia- believe that such a characterization is not fully tions may stay only a few days. Such a transac- accurate. tion is largely akin to cross-border trade in ser- Labor mobility tends to be more complex vices (so-called Mode 1 trade under GATS) and than either trade or capital mobility. Even very differs little from goods trade to the extent that large differences in economic returns (mea- it does not involve lasting factor movement. sured by wages) are not sufficient to induce Not so for revolving teams of contract-based 155 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 workers in construction services or for intra- gressive relaxation of barriers to labor mobil- company transferees, who may be deployed ity (table 4.6). Winters (2003a) suggested that abroad for several years (and yet may consti- if developed countries were to raise to 3 per- tute a transaction for the purposes of the cent of their labor force their quotas on the in- GATS). In the latter cases, such “trade” in ser- ward movements of temporary workers from vices has more in common with the economics developing countries, they would realize an of migration, as migrating workers reduce the overall gain of $150 billion each year. His supply of labor in the sending country while work concludes that the gains from such lib- adding to it in receiving countries. Further- eralization would be shared equally by devel- more, the temporary movement of labor is oped and developing countries. Most impor- often tied to longer-term flows of capital (in the tant, from a development perspective, these form of foreign direct investment). findings suggest that the largest benefits (to The essence of international trade lies in both sending and receiving countries) would securing the gains from cross-country differ- come from the movement of lower-skilled ences in costs, prices, endowments, or tastes. workers, as those workers are spread more The larger such differences, the greater the po- evenly over the economy, benefiting more sec- tential gains from removing obstacles to such tors, than highly skilled ones (OECD 2002a). trade. The disparity between the abundance of Winters and Walmsley (2002) recognize labor in developing countries and its scarcity that adjusting wage levels in response to com- in the developed world suggests that signifi- petition from low-skilled developing country cant pro-development returns—potentially workers could entail high social costs. Such greater than those stemming from the full lib- findings underscore the long-term importance eralization of trade in goods—could be had if of enhancing the educational levels and human medium- and less-skilled workers in develop- capital of lower-skilled individuals to ensure ing countries were allowed to provide their that fewer developed country nationals are services temporarily in developed countries. competing directly with unskilled workers from poorer countries. To manage the complex What are the gains from political economy that TMNP liberalization temporary movement? could entail and to minimize adverse (or exces- Although labor remains far less mobile than sively concentrated) distributional effects for goods and capital, the increasing diversity of vulnerable workers in receiving countries, the migrants’ nationalities and the migration authors suggest that the liberalization process channels used, as well as the growing share of should be incremental, with the most sensitive temporary and skilled workers in total migra- sectors exempted. Affected workers may be tion flows, does indicate a growing promi- helped through adjustment schemes similar to nence for migration in the broader context of the U.S. Trade Adjustment Assistance Act, the economic globalization.6 Links between labor assistance provisions of NAFTA, Canada’s mobility and the liberalization of trade and in- General Adjustment Assistance Programme, vestment have gained in visibility, as modern and Australia’s Special Adjustment Assistance trade agreements have proliferated and broad- (Borjas 2000; Borjas, Freeman, and Katz 1997; ened at both the regional and multilateral OECD 2002a). levels. Such developments have sparked an in- Because liberalization of merchandise trade terest within the research community in mea- has reduced price differentials between devel- suring the potential effects of liberalizing oped and developing countries to a ratio of labor flows. two to one—whereas service prices and wage Several recent studies have drawn attention differentials continue to differ by a factor of to the potential benefits arising from a pro- ten or more—the gains from liberalizing cross- 156 L A B O R M O B I L I T Y A N D T H E W T O : L I B E R A L I Z I N G T E M P O R A R Y M O V E M E N T Table 4.6 The distribution of costs and benefits associated with Mode 4 trade Sending countries Receiving countries Benefits Costs Benefits Costs In a situation of saturated In a situation of labor Temporary admission of Temporary admission of labor markets, departure of shortages, departure of foreign workers is a response foreign workers could delay workers exerts a downward workers exerts only an to labor needs (shortages in the structural adjustment of pressure on unemployment, upward pressure on (high) some sectors or geographic the economy. and an upward pressure on wages. areas). In a situation of saturated (low) wages. The effects of temporarily Entry of foreign service labor markets, arrival of Once abroad, workers losing human capital and providers results in increased workers exerts an upward proceed to income transfers public investment (education competition (wider choice of pressure on unemployment (compensation of employees and training expenses) better services at lower price) and a downward pressure and remittances), which are depend on the scarcity of at lower cost (activity stays on wages. a major source of capital the workers’ skills (see cost in the country). Departure of workers inflows and investment for of removing a doctor v. Temporary admission of generates a replacement cost many developing countries. a low-skilled worker). workers is a partial and a loss of human capital Upon return of the workers, Replacement of scarce substitute for permanent and investment (training). global human capital of the resources could generate immigration (less sensitive country is increased. high costs. and lesser use of public infrastructure and services). In a situation of saturated labor markets, return of workers exerts an upward pressure on unemployment, and a downward pressure on wages. Maximizing benefits Maximizing benefits Provide adequate infrastructure and career opportunities to Create a possibility for workers to change visa status maximize the use of competence acquired abroad once the (become permanent resident) to avoid departure of the most worker is back. useful ones. Create incentives for return. Facilitate temporary admission of workers and create incentives to attract workers in specific sectors or geographic Negotiate commitments from other members in sectors areas (where shortages exist). where the national labor market is saturated. Adjust the national economy to new competitive conditions. Adopt structural adjustments in sectors of labor shortages to prevent outflows of workers. Overall benefits For business: a source of increased flexibility, profitability, and competitiveness; an instrument for facilitating trade and penetrating new markets For individuals: acquisition of vocational skills and know-how, including the learning of a foreign language; improved quality of life while abroad (compensation for expatriation); increased employment opportunities upon return Source: OECD (2001, 2002a). border labor movements and other trade in five years, after which they would return to services could be much greater than from fur- their home countries. The author suggests that ther liberalizing trade in goods. Rodrik (2002) if admissions were capped at 3 percent of the proposes a scheme under which skilled and developed countries’ labor force, the scheme unskilled workers from developing countries could generate direct income gains of as much could apply for temporary visas entitling them as $200 billion annually. Returnees’ invest- to work in developed countries for three to ments and the transfer of their experience 157 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 would produce further gains for sending coun- of trade growth reduces the demand for skilled tries (OECD 2002d). labor in skill-scarce developing countries while The source of the gains identified in the increasing the demand for skilled labor in skill- studies discussed above is a narrowing of abundant developed countries, the result could wage differentials between rich and poor be a widening of the gap in labor income of countries—a politically sensitive subject. In skilled workers between the North and the the regulated labor markets of many WTO South while narrowing the gap in labor income member countries, domestic legislation limits of unskilled workers. Ghose (2002) concludes or opposes downward pressure on wages. that “in a world without restrictions on labor Moreover, many recipient (developed) coun- mobility, increased trade worsens the brain tries require equality of treatment for tempo- drain from developing countries but has uncer- rary workers (equal wages and social pro- tain effects on overall migration. Trade and tection parity requirements) with nationals at flows of skilled labor are complements while comparable levels of skill and experience trade and flows of unskilled labor are substi- (OECD 2002d). For these and other reasons, tutes.” Under such circumstances, measures to most models assume no more than a halving increase the mobility of skilled workers can re- of differentials. To the extent that national inforce the initial comparative advantage of the provisions sustain domestic wages, the overall trading countries, so that skill-abundant coun- gains of Mode 4 liberalization may be lower, tries become ever more skill abundant, while but in such cases adjustment costs will be con- labor-abundant countries become ever more comitantly lower as well. (low-skill) labor abundant. Trade expansion Other factors that may affect the overall combined with the unrestricted mobility of gains of greater mobility of temporary labor skilled workers could conceivably put the accu- are the cost of creating or scaling up tempo- mulation of human capital beyond the reach of rary visa schemes7 and the possibility that many developing countries.8 temporary workers might join the ranks of the Much of the existing literature on brain unemployed in the developed countries. Over- drain focuses on the short-term costs in send- shadowing such technical factors, however, ing countries. Indeed, such costs can be sub- are the doubts and fears arising from the re- stantial—higher education is subsidized heav- cent rise in legal and illegal migration in the ily in developing countries and skilled migrants OECD area. The bursting of the dot-com bub- carry away scarce human capital built through ble of the late 1990s and the security implica- public investments. But some of the purported tions of the ongoing war against terrorism disadvantages associated with the migration of have compounded those fears, making signif- skilled workers from developing to developed icant liberalization of temporary movement countries can be partially mitigated if the less probable than was believed possible a few movement is temporary. Temporary migrants years ago. may generate sizeable remittance flows which, along with the accumulation and subsequent And the costs? repatriation of embodied knowledge and Temporary movement may help address sev- global experience when workers return, could eral concerns often associated with the politi- substantially increase the benefits associated cal debate over more permanent migration in with greater temporary movement. Key to both developing and developed countries (see achieving this outcome are changes in the de- table 4.6). One such concern revolves around sign of both trade and immigration policies in the possible impact on developing countries receiving countries and stepped-up efforts on of a “brain drain,” and its longer-term conse- the part of sending countries to increase return quences for capital accumulation and growth migration of skilled workers while pursuing in the developing world. Indeed, if the pattern enhanced temporary movement of lower- 158 L A B O R M O B I L I T Y A N D T H E W T O : L I B E R A L I Z I N G T E M P O R A R Y M O V E M E N T skilled workers in developed country markets. significant impact on a developing country’s Doing so would help to ensure that the growth economy and impede its development (WTO of trade does not exacerbate developing coun- 1998, Devan and Tewari 2001). Moreover, be- try skill shortages. cause skilled migrants are often from rela- tively affluent households that do not require For sending countries, TMNP has risks regular income support, they may be less and rewards 9 likely than unskilled migrants to send back re- The impact of the temporary movement of mittances (Ghose 2002). As long as the move- workers on the sending country can be con- ment of such workers remains temporary, sidered at three levels. First, there are eco- however, it is certainly preferable to the more nomic effects of removing the worker from lasting brain drain caused by permanent emi- the labor market (departure). Next, during the gration (WTO 1998, OECD 2002d). stay abroad, the worker will maintain con- Temporary movement of workers can help tacts with the home country, remitting funds to ease the strain on domestic labor markets— to family or making direct investments. Fi- work abroad can be an escape route from un- nally, there is the economic impact of the mi- employment, and can reduce a country’s over- grant’s return to the home country. Each of all unemployment rate (Werner 1996). This these effects is considered in turn. is most often the case for unskilled workers, although in some countries the number of Departure: Temporary migration and domes- people trained for certain occupations exceeds tic labor markets. A genuine risk associated the absorptive capacity or needs of the local with sending workers abroad is that scarce re- market—as is the case for Indian engineers, sources, such as human capital, will be lost— for example.11 Whether out-migration sus- often at a substantial public cost in education tains wages in the sending country (Ghosh and training investments.10 Workers who go 1998) or exacerbates an existing skills short- abroad are generally young, highly motivated, age (Werner 1996), its effect will be less if the well educated, and not easy to replace, espe- migration is temporary than if it were perma- cially in developing countries, where wages nent—thereby reducing not only the potential are lower, career paths are limited, and work- risk but also the potential reward of tempo- ing conditions less satisfactory than abroad rary mobility as a policy tool (OECD 2002d). (PSI/EI 1999, OECD 2002d). An important Receiving countries can play a part in pre- corollary of the “brain drain” is indeed the venting particularly harmful shortages of risk that developing countries will indirectly skilled personnel in developing countries. For subsidize industrial country R&D by export- example, the British government has published ing the human capital embedded in locally a code of conduct for trusts under the Na- trained workers. Indeed, the cost of providing tional Health Service (NHS) that prohibits university education to professionals who the recruitment of nurses from countries where then move to wealthier countries for increased they are in short supply, such as in South Africa opportunities may represent a net resource or the West Indies. The code is not legally loss for developing nations. Tax policy can be binding (OECD 2002d). used to recover some of the loss—for example, Through appropriate policies and incen- by requiring students who choose to leave the tives, sending countries can encourage skilled country to repay their education expenses be- migrants to return (box 4.5). To date, incen- fore departing. tive policies have a mixed record, often failing Further, because highly skilled workers to address the real reasons behind workers’ earn more, consume more, pay more taxes, decisions to settle abroad permanently—ac- and are more productive than the unskilled, culturation, better career opportunities, and their departure, even if temporary, can have a access to and integration within personal and 159 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Box 4.5 Initiatives to encourage return migration U p to one-third of R&D professionals from the developing world reside in the OECD area. Foreign studies constitute a major channel for migra- returns of migrants and capital. Developing countries such as India, for example, have the capacity to invest in R&D and human infrastructure, and thus tion, especially in science and technology—79 per- are more able to draw migrants back. China recently cent of 1990–91 PhD graduates in science and tech- launched a project to develop 100 universities into nology from India and 88 percent of those from world-class institutions that not only provide higher China were still working in the United States in education, but also academic employment and 1995, compared to only 11 percent of Koreans and research opportunities. An alternative for less- 15 percent of Japanese. The migration of skills can developed countries is to create a good communica- be slowed through the return of expatriates to their tion network among expatriates by linking them country of origin. Returnees contribute to economic to counterparts in their county of origin. Scientific development through their valuable management diaspora and other expatriate knowledge networks experience, entrepreneurial skills, and access to can help sending countries reap benefits and know- global networks. The Ministry of Science and Tech- how from emigrants overseas. Forty-one expatriate nology in China, for example, estimates that most knowledge networks have been identified around the of China’s Internet-based ventures were started by world. The FORS Foundation, for example, seeks to returning overseas students. involve Romanian scientists in Romania and abroad Taiwan (China), the Republic of Korea, and in contributing to economic reform in Romania. Singapore have been successful in fostering return Grassroots initiatives in South Africa and Latin migration by opening up their economies and em- America have been developed to connect researchers ploying policies to foster domestic investments in in- abroad to networks in their home countries. novation and R&D. Korea has focused on upgrading The worldwide network of Indian professionals its research institutions, such as the Korea Institute has been investing in skill development at home to for Science and Technology, as a way of attracting raise endowments and bolster the finances of some returnees (UNDP 2001). The government of Taiwan of India’s institutions of higher education. The Indian (China) likewise played an important role in drawing government also has contributed to the emergence of back American-trained scientists and engineers, who private networks among Indian professionals abroad have subsequently helped to develop the country’s through legislative and tax rules that encourage re- information technology sector. A National Youth mittances and investment. The Return of Qualified Commission has been established there as a clear- African Nationals Program, conducted by the Inter- inghouse for potential employers and returning national Organization for Migration, has attempted scholars seeking employment, and an airfare subsidy to encourage the return of qualified nationals and is granted to the graduating student, spouse, and up helped them to reintegrate. Even Switzerland has to two children, if they decide to return to Taiwan promoted networking and contact among Swiss sci- (China). The Commission also has established chan- entists in the United States through Swiss-List.com, nels of communication with overseas scholars to an online network. France has a similar network. simplify recruitment when the need arises (Cultural Division of Taipei 1998). Opportunities for research, innovation, and entrepreneurship at home are needed to stimulate Source: World Bank staff. professional networks. Chinese Taipei ad- worked for a time in California’s Silicon Val- dressed those reasons by investing heavily in ley. Today the park accounts for nearly 10 per- research and education. More than half of the cent of Chinese Taipei’s gross national product enterprises in Chinese Taipei’s Hsinchu Sci- (Devan and Tewari 2001). The Hsinchu ex- ence Park were created by engineers who had ample illustrates the positive economic impact 160 L A B O R M O B I L I T Y A N D T H E W T O : L I B E R A L I Z I N G T E M P O R A R Y M O V E M E N T that workers can have upon their return from gration may thus do more to encourage remit- a temporary stay abroad (OECD 2002d). tances than permanent migration. Many visa regimes require proof of a fixed-term contract Staying in touch: Temporary migration as a (pre-established and limited duration); it source of increased financial inflows. Because could be argued that this type of regulation banning temporary movement by workers is may therefore have benefits for sending and neither feasible nor desirable, sending govern- receiving countries alike (OECD 2002d). ments have an interest in optimizing the bene- fits of such movement, for example, by offer- Returning: Returning migrants as a source of ing incentives for the repatriation of foreign increased human capital. Where labor mobil- earnings (OECD 2002d). ity is only temporary, the net benefits of de- The economic impact of remittances and as- parture may be partially offset by the effects sociated labor receipts depends significantly on on return. Yet the balance need not be zero, as the use to which the funds are put in the send- the country’s stock of human capital will have ing country. They may be consumed, invested, grown between the time of departure and the or saved. Moreover, if they are consumed, the time of return (OECD 2002d). impact will differ again depending on the na- The additional skills (languages, experi- ture (consumer or capital goods) and origin ence, know-how) acquired by temporary (local or imported) of the goods consumed workers can be put to work upon their return, (Werner 1996). An ILO study on Indonesia thereby contributing to economic growth and shows that income from temporary work development. Indeed, as endogenous growth abroad was applied by workers and their fam- theories suggest, increases in human capital ilies to pay off debts, sustain consumption, can yield significant positive externalities and raise savings, and finance investments, in that durably affect the long-term growth prospects order (ILO 1996). In India, expatriate engi- of developing countries. In particular, the ac- neers working either permanently or temporar- cumulation of human capital can be instru- ily in Silicon Valley have accounted for most of mental in helping developing countries to the investments made in the cities of Bangalore move into more skill-intensive production. Yet and Hyderabad, which have become new poles just as with firms, taking advantage of such of growth for the country, establishing India as capital requires a suitable enabling environ- an export powerhouse in software design and ment. In particular, the home country must be IT-outsourcing industries (Devan and Tewari able to provide the infrastructure and career 2001, OECD 2002d). The current Mexican opportunities necessary to meet the aspirations administration has reshaped the government’s workers may have developed during their stay attitude toward migrants to the U.S., including abroad (OECD 2002d). greater advocacy on their behalf with the U.S. Temporary movement may be an impor- government. The Ecuadorian government has tant vector for enhancing two-way trade and a program designed to increase the earnings of investment flows between sending and receiv- its citizens working abroad, whose remittances ing countries. The Indian experience confirms are the country’s largest source of foreign ex- such linkages, for while India may have ex- change after petroleum. ported a number of its skilled workers in re- An important feature of TMNP is the ob- cent years, the flow has not been one-way; the served tendency for workers to retain closer country saw its IT exports increase from $150 links with their home country if the length of million in 1990 to $4 billion in 2000. The re- their stay abroad is relatively short and pre- cent surge in FDI directed toward India’s high- determined. Remittances seem to be greatest tech centers is similarly related to the presence when a worker expects to return at a fixed abroad of a large group of scientists, engi- date (Galor and Stark 1991). Temporary mi- neers, and entrepreneurs (Nielson 2002). 161 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 For receiving countries, temporary enacted to contain such flows may inhibit the movement is politically sensitive but ability of undocumented workers to maintain usually beneficial closer two-way links with sending countries While the temporary movement of workers is (in part because of a reluctance to incur the not likely to unduly disrupt the sending coun- high costs and attendant risk of illegal reentry) try’s labor market, the potential effects of such through formal temporary migration channels. mobility on segments of the receiving country’s As a result, undocumented workers become labor market may at times be more significant. particularly vulnerable to various forms of There, the concern arises that mobile foreign work-related abuse and often become caught workers may be in direct competition with na- in a poverty trap (Papademetriou 2001). tionals of the host country working perma- nently in the same occupations. Even if the mi- Impact on the receiving country labor market. grant’s stay is temporary, the growing number Migrants, especially workers involved in tem- of foreign workers and the continuous influx porary movement, tend to concentrate in sec- of workers over different time horizons under tors and regions characterized by labor short- contract-based flows could increase competi- ages at both the high and low end of the skills tion in the labor market (OECD 2002d). spectrum. It may thus be less likely for them to From this angle it is easy to see why immi- compete directly with native workers than is gration can be controversial in receiving coun- commonly assumed.12 In the majority of re- tries. There is evidence that unskilled migra- ceiving countries, temporary foreign workers tion reduces the relative wages of unskilled are found mainly in the following five sectors: workers in industrial countries (Borjas, Free- (a) health (especially doctors and nurses; doc- man, and Katz 1997). An inflow of unskilled tors are more likely to practice in remote/rural workers from the South will benefit highly areas); (b) education, particularly higher edu- skilled workers in the North. Their jobs are cation (that is, academic and research staff); not threatened by the latter, and the presence (c) information technology; (d) catering; and of immigrants will lower prices for many (e) agricultural labor. goods and services consumed by the skilled It is important, however, to understand workers. But the same inflow will reduce real why migrants (including temporary migrants) wages of unskilled northern workers (World are concentrated in these sectors. In health Bank 2001), and over time contribute to a de- and education, wages in most countries are set terioration in income distribution. Against this by policy or collective bargaining, and rela- latter trend, however, demographic and educa- tively clear procedures for recognizing foreign tional trends in affluent countries will combine credentials are in place. Migration in these in the coming decades to raise the relative sectors benefits the public sector—and hence wages of unskilled labor in the absence of mi- the general public—as workers become tax- gration (see box 4.1). As these demographic ef- payers and consumers of public services. In IT fects will likely be large, scope may therefore and other private sector professions prone exist for increased flows of unskilled labor in to labor shortages, wages are more likely an environment of relative wage stability. market-determined. But supply is constrained Despite the acknowledged benefits of tem- by lags in training home-country specialists. In porary migration, the norm in receiving coun- the absence of migration, firms would bid up tries is to continue to impede the movement wages and after a lag, supply would respond. of low-skilled or unskilled workers through But with flexible work permit systems, firms various restrictions. Such restrictions can con- can import migrants, especially on temporary tribute to the recent sharp rise observed in un- contracts. In low-paying sectors such as ca- documented low-skilled workers throughout tering and domestic services, unskilled local the OECD area. The stricter border controls workers are typically unwilling or unable to 162 L A B O R M O B I L I T Y A N D T H E W T O : L I B E R A L I Z I N G T E M P O R A R Y M O V E M E N T fill the available jobs. The effect of temporary significant income transfers to the most highly foreign workers in these sectors again is to qualified workers at the expense of the rest of benefit firms, but it is not likely that workers the workforce and of the country’s consumers in receiving countries will be significantly dis- (Hodge 1999, OECD 2002d). However, the advantaged: if migrants do not fill these jobs, impact of temporary workers on the work- they simply tend to go unfilled or are not cre- force in receiving countries is the subject of ated in the first place (Home Office 2001). significant debate, in particular with regard to In all three cases, the receiving country wages and working conditions (box 4.6). tends to benefit overall from filling labor mar- Temporary foreign workers may also bring ket gaps through migration/temporary move- more direct benefits. Intracompany transferees ment. The result of such mobility is likely to consume the bulk of their income in the host include reduced inflationary pressures and an country (housing, food, clothing) and make increase in the overall efficiency of firms. use of that country’s services (banks, trans- Expansion of such temporary schemes has portation, communications). Their income thus become a preferred means of responding therefore generates wealth to the host country, to labor shortages in receiving countries, which would not be the case if the services whether these are seasonal, cyclical, regional, were provided remotely across borders (under sectoral, or skills-related (Werner 1996). From online outsourcing, for example), or if con- an economic viewpoint, the ability to bring in sumers required such services abroad. The foreign labor is essential, since human capital presence of the temporary foreign worker’s limitations can depress investment and create employer in the host country is itself a source Box 4.6 Wages and conditions T he basic question is whether temporary workers should receive the same wages and conditions as nationals employed in the same industry. In many their nursing shortages—the need for better wages and working conditions (OECD 2002f). Proponents of wage parity argue that because a countries (particularly OECD countries), this is a temporary foreign worker in an OECD country faces legal obligation—and 50 WTO members have in- living costs in that country, there is no justification cluded this stipulation in their Mode 4 commitments. for paying lower wages. Moreover, temporary for- But some developing-country advocates contend eign workers, particularly women in domestic ser- that such requirements undermine the comparative vices and other lower-skilled activities, can be highly advantage upon which Mode 4 trade should be vulnerable to exploitation if not fully subjected to based—the relatively inexpensive labor of sending local labor laws. countries. Equal treatment, however, does not always re- Such arguments are met with fierce resistance sult in equitable outcomes. In many countries, tem- from unions in developed countries, which fear that porary workers are required to contribute to social cheaper temporary foreign workers could undermine security programs in the receiving country from the hard-won gains of workers in developed coun- which they receive no, or minimal, benefits. One tries. To prevent foreign “strike breakers,” 22 WTO alternative in this regard could be for social secur- members also have reserved the right to suspend ity charges from temporary migrant workers to be Mode 4 commitments in the event of labor-manage- paid into separate funds and reimbursed upon ment disputes. Even where foreign workers may not workers’ return to their home country. actually be paid less, their presence may act as an impediment to reform. For example, some claim that the temporary employment of foreign nurses has allowed governments to ignore the root causes of Source: Nielson (2002) and OECD (2001b). 163 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Box 4.7 E-commerce and temporary movement S ome WTO members, in particular developing countries, have expressed concern that the growth of trade in services via information and communica- countries with a large pool of skilled labor, a range of lower skilled “back-office” services are now traded over the Internet—among them basic data tions technologies will become a substitute for trade entry and customer call centers. via temporary movement of service suppliers (Mode Technological developments may be changing 4). In some cases, services are now delivered over the the nature of temporary movement, rather than re- Internet that previously required physical presence. moving the need for it, with increasing numbers of This tends to be more common in knowledge-inten- employees managing their international responsibili- sive fields of activity and can be attractive to compa- ties through a combination of regular communica- nies wishing to pay developing-country wages rather tion link-ups and frequent, shorter business trips to than local (developed country) wages, as is generally the local operations, referred to as “virtual assign- required for workers temporarily relocated. Still, the ments.” Replacing longer-term assignments with Internet is not always a good substitute. Security and more frequent shorter ones, in addition to virtual confidentiality requirements may limit its use, and in working, helps companies manage the costs of inter- some countries the infrastructure is not yet capable national responsibilities. ICT also may play a role in of fulfilling contracts remotely. encouraging employees to accept longer-term assign- Trade in services over the Internet can offer ser- ments by reducing the sense of isolation from friends, vice suppliers from developing countries the opportu- family, and cultural context. nity to participate in global trade, notwithstanding their lack of commercial presence in foreign markets. While some of this trade may be in knowledge- intensive areas, primarily benefiting those developing Source: OECD (2001b). of wealth (jobs preserved, intermediate con- an alternative to sending qualified personnel sumption of goods and services, business taxes, abroad (box 4.7). and so on) (OECD 2002d). In fact, by focusing on the physical mobil- Temporary movement as a first step toward ity of temporary movement of foreign workers permanent migration. Temporary movement critics may miss a key point: A country cannot can be a first step to permanent residence, restrain international competition in hopes of either legally (by changing visa categories) preserving the market share of domestic pro- or illegally (overstaying). Overstaying is a risk ducers simply by blocking the entry of tempo- with all forms of temporary entry, including rary foreign workers. For example, India is for tourists. Administered schemes for tempo- widely recognized for expertise in computer rary movement arguably could help discour- services, thus explaining the heavy flow of In- age employers from using undocumented dian computer specialists to more advanced workers by making available legal temporary countries. It is likely that if those countries foreign workers for seasonal activities. Where were to refuse temporary visas to these work- temporary workers are permitted to apply for ers or introduce stricter limitations on their permanent-resident status, their temporary issuance, the work could still be outsourced stay may serve as a useful preselection of can- to India over electronic networks via Modes 1 didates for future migration (OECD 2002a, and 2 (Chadah 2000, OECD 2002d). Indeed, Nielson 2002). some developing countries point to outsourc- Available data show little evidence of large- ing of work, including over the Internet, as scale transfer of workers from temporary to 164 L A B O R M O B I L I T Y A N D T H E W T O : L I B E R A L I Z I N G T E M P O R A R Y M O V E M E N T Box 4.8 Boosting intra-EU labor mobility T he Treaty of Rome recognizes the principle of free movement for nationals of EU countries wishing to reside or work within the area formed is partly attributable to linguistic and cultural barri- ers, but it is also a result of structural rigidities in the labor markets of individual member states. by the signatory states. More recently, measures In 2002, the European Commission launched have been taken to facilitate intra-European mobility. an action plan for mobility and skills with a view to These include a directive on free movement of non- facilitating geographic mobility in the period up to workers, students, and retired persons, and a series 2005 by removing remaining administrative and legal of directives on mutual recognition of skills and ac- barriers, increasing the portability of supplementary cess to certain public service jobs previously reserved pension rights of migrant workers, and improving for nationals. existing regimes of skills recognition in the regulated Nevertheless, intra-European mobility remains professions. very low, involving less than 0.2 percent of the total population of the Union, a level seven times lower than movements among the nine major census areas in the United States. The low mobility within Europe Source: OECD (2002f). permanent status. The U.K. work permit sys- little doubt that tightening immigration con- tem allows employees to apply for permanent trols because of heightened concerns over na- settlement after four years of continuous em- tional security is likely to have a chilling effect ployment, but in practice, a relatively small on TMNP liberalization. Meaningful expan- proportion seem to settle permanently (in sion of temporary worker programs requires 1998, 3,160 work permit holders settled in that security clearance be quick and reliable. the United Kingdom against approximately The challenge politically is to separate the se- 70,000–80,000 work permits approved each curity arguments from labor market or service year). Indeed, even where all, or most, barriers export considerations, and to strike an accept- have been removed, floods of foreign workers able balance between economic efficiency and generally have failed to materialize, as intra- national security (Mattoo 2003). EU labor flows show in the context of a single It estimated that up to 250,000 information- labor market (OECD 2001b) (box 4.8). technology professionals from India work in the United States, some on temporary visas and Security concerns. Any attempt to facilitate others on work permits. But with almost all individual mobility must confront today’s in- visa applications taking longer to process, In- creased concerns about national security. dian technology companies are taking steps While all countries ultimately share such con- to adapt to the increasingly limiting conditions. cerns, they tend to pose a greater challenge for For example, Infosys, a leading provider of policy officials in major receiving countries. software services, is ensuring that the bulk of Brought into sharper focus since September its outsourcing activities are undertaken on site 11, 2001, security considerations are changing in India rather than in the United States or the balance between the facilitation and en- Europe. The spectacular recent growth of IT forcement aspects of immigration controls, outsourcing in developing countries, while with measures to facilitate the entry of foreign minimizing the need for labor movement, is workers increasingly scrutinized to ensure nonetheless proving controversial, with fears of they do not become conduits for entry by ille- an exodus of white-collar jobs in service indus- gal or undesirable individuals. There can be tries. A case in point is the recent set of legisla- 165 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 tive measures passed by the states of Connecti- set in process following the WTO meetings in cut, Maryland, New Jersey, and Washington, Doha in November 2002. restricting outsourcing of state government ser- These discussions go by the label of “Mode vices. Similar concerns, and calls for similar 4” negotiations, in reference to the classifi- policy responses, have been voiced in Europe. cation of the modes of service delivery in the Besides restricting the movement of work- GATS agreement. Mode 1, or “cross-border ers, delays in travel can harm the competitive- supply,” is analogous to trade in goods; Mode ness of firms. There is evidence that the com- 2 is “consumption abroad” (for example, petitiveness of subsidiaries of U.S. companies tourism or study abroad); Mode 3 is “commer- established in China has been adversely af- cial presence” (as in the supply of a service fected as tightened security has hampered the through a subsidiary or branch in another coun- ability of U.S. companies to obtain visas for try); and Mode 4 is “temporary movement of Chinese nationals to conclude deals, under- individual service suppliers.”13 WTO members take training, and even attend strategic semi- can elect to commit to providing market access nars and meetings in the United States. Parent and/or national treatment for each mode of companies in the United States are complain- supply for any number of around 160 possible ing about lost contracts and the move of Chi- services sectors and sub-sectors. nese clients to European companies that can Mode 4 is defined as the supply of a service offer faster and more predictable issuance of by a service supplier of one WTO member, visas. through presence of natural persons of a mem- While recognizing the importance of border ber in the territory of another member on a security in an environment of heightened risk, temporary basis. While there is some debate care must be taken that the granting of visas about what exactly this means, Mode 4 ser- and work permits does not become a disguised vice suppliers generally: barrier to trade. India’s minister for trade and commerce recently termed the denial of visas • Gain entry for a specific purpose (for ex- and restrictions on the movement of natural ample, to fulfill a service contract as self- persons as an indirect method by developed employed or as an employee of a service nations of denying market access to developing supplier); nations. Care also must be taken to reconcile • Are confined to one sector (as opposed the need for increased security at entry points to workers who enter under general mi- with that of allowing commerce to flow as gration or asylum programs who can freely as possible. This includes recourse to new move among sectors); technologies, notably biometrics, a system of • Are temporary (that is, they are neither fingerprint and retinal recognition, and more migrating on a permanent basis nor seek- traditional methods such as permanent resi- ing entry to the labor market). “Tem- dent cards. porary” is not defined under the GATS, but permanent migration is explicitly ex- cluded, and thus this issue is left to the Mode 4 and the WTO discretion of each country. In practice, A s noted above, some types of temporary foreign workers—service suppliers—are covered under the WTO General Agreement the time frames set out in WTO members’ commitments on Mode 4 range from several weeks to up to three to five years, on Trade in Services (GATS). Greater freedom varying among countries, sectors, and for the temporary movement of individual ser- professions. Thus, for example, Japan al- vice suppliers is being negotiated under the lows foreign business travelers to stay for GATS, as part of the multilateral negotiations a maximum of 90 days, but certain cate- 166 L A B O R M O B I L I T Y A N D T H E W T O : L I B E R A L I Z I N G T E M P O R A R Y M O V E M E N T gories of intracorporate transferees can Measurement of Mode 4 trade suffers stay as long as five years. from poor data, tepid commitments, and • Are service suppliers. Being a services a range of barriers agreement, GATS Mode 4 only covers ser- There are two ways to measure Mode 4 trade: vice suppliers—there are no parallel WTO by value or by number of service suppliers (see rules covering movement of people related box 4.9). Services trade statistics face a num- to agriculture or manufacturing.14 ber of conceptual and practical problems • Are service suppliers at all skill levels, al- and, despite progress, reliable figures are some though in practice WTO members’ com- way off. Nonetheless, available estimates—and mitments are limited to the higher skilled they are very rough—suggest that, in terms of (see below) (Nielson 2002). the monetary value of trade, Mode 4 is the Box 4.9 Measuring Mode 4 is still imprecise Value of trade: Balance-of-payments statistics • Migration statistics consider “temporary” to be 12 Balance-of-payments statistics capture some labor- months or less; under the GATS it is undefined but related flows of relevance to the estimation of trade in practice can be up to 6 years. under Mode 4: • Migration categories generally do not distinguish “Compensation of employees” (wages, salaries, between service and non-service activities. and other compensation received by individuals • It is not always possible to judge whether the ac- working abroad for less than one year). This mea- tivities covered by some visa categories are com- sures both overestimates (includes workers other mercial and would qualify as the supply of a ser- than service providers) and underestimates (excludes vice under the GATS (for example, occupational business visitors and individuals staying more than a trainees, professional exchange programs). year abroad) trade under Mode 4. • Some visa categories include persons both consum- “Workers’ remittances” (transfers from workers ing and supplying services (for example, exchange who stay abroad for a year or longer). This measure visitors encompass exchange students and visiting overestimates (covers all expatriates, regardless of lecturers). the sector in which they work) and underestimates (only a residual income after expenditure and savings Neither of these sources—value of trade and in the host country, and many such remittances are numbers of people—capture the dynamic effects of not effected through official channels) trade under Mode 4 and its essential role in facilitating trade Mode 4. under other modes (for example, Mode 3, commer- Statistics on trade in services are available only cial presence; Mode 1, cross-border supply). for some services sectors and traditionally have not Some national figures for entries under specific been broken down by modes. Figures for Mode 4 visa programs may closely correspond with Mode 4 are likely to be significantly underestimated. (for example, temporary medical practitioner visas), but because of the above problems, no aggregate The number of people: Migration and labor statistics figures are available for all entrants falling under Statistics on the number of people moving under Mode 4 at the national level. Additionally, given the Mode 4 are scarce and highly imprecise. Statistics are absence of detailed temporary entry visa regimes in available for temporary foreign workers for several many countries, aggregate global estimates of the countries, but they are not an exact match to GATS number of people moving to supply services under Mode 4. Main problems include: Mode 4 are not possible. • Business visitors may be excluded or hidden under tourist visas (a significant part of Mode 4 trade). Source: OECD (2001b) and Nielson and Cattaneo (2003). 167 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Table 4.7 TMNP is the smallest of the four modes of international service supply Service exports by mode of supply, 2001 (billions of dollars and percentage of total) 1997 2001 Percentage Percentage Mode of international service supply Value of total Estimate of total Proxy 1 Cross-border supply 890 41.0 1,000 28.2 BOP: commercial services minus travel 2 Consumption abroad 430 19.8 500 14.1 BOP: travel exports 3 Commercial presence 820 37.8 2,000 56.3 FATS statistics turnover 4 Movement of natural persons 30 1.4 50 1.4 BOP: compensation of employees Total 2,170 100.0 3,550 100.0 BOP is balance of payments. FATS is Foreign Affiliate Trade in Services. Source: IMF, Balance of Payments Yearbook. smallest of the four modes of services supply A look at members’ current GATS sched- (table 4.7). ules shows that levels of commitments vary Negotiations on Mode 4 first took place strongly across modes of supply. Within a during the Uruguay Round of trade talks held given sector, trade conditions for Mode 4 tend from 1986 to 1993, but they were not partic- to be significantly more restrictive than condi- ularly successful—in fact, they served primar- tions for other modes. No developed country ily to facilitate exploratory business visits and has scheduled a “none” entry (signifying un- the movement of high-level personnel within fettered access) for its Mode 4 commitments, multinational corporations. While the Uruguay and only 1 percent of market-access commit- Round negotiations were formally concluded ments undertaken by developing countries are in December 1993, negotiations in several fully liberal. This compares with one out of areas—basic telecommunications, financial two entries for Mode 2 (consumption abroad) services, maritime transport services, and the being full commitments.15 movement of natural persons—were extended Many schedules have established links beyond the end of the Round because of wide- across modes of supply. Members’ schedules are spread dissatisfaction with the level of liberal- mostly biased in favor of intracorporate trans- ization achieved in those areas. Further negoti- ferees, hence making the economic value of ations on Mode 4, concluded on June 30, such commitments dependent on access condi- 1995, produced no major breakthrough. Only tions for Mode 3 (table 4.8). Such commitments Australia; Canada; the European Communi- are of limited interest to WTO members which, ties; and its member states, India, Norway, and given their level of economic development, are Switzerland improved on the commitments not significant foreign investors. Schedules are they made in the Uruguay Round, and these also more open for highly skilled labor, where improvements were annexed to the Third Pro- developing countries tend to be net importers, tocol to the GATS. The improvements mainly since their comparative advantage lies with rel- concern access opportunities for additional atively unskilled labor-intensive services. categories of services suppliers, usually inde- As of April 2002, an overview of members’ pendent foreign professionals in a number of horizontal commitments shows that the major- business sectors, or the extension of such pro- ity of the entries scheduled—nearly 280 out fessionals’ permitted duration of stay. of a total of 400—concern executives, man- 168 L A B O R M O B I L I T Y A N D T H E W T O : L I B E R A L I Z I N G T E M P O R A R Y M O V E M E N T Table 4.8 Most Mode 4 commitments the commitments scheduled by developed and by WTO members are in management developing countries. Both groups seem to have categories been equally hesitant in undertaking very lib- Entries by WTO members that have made Mode 4 eral commitments for Mode 4 (box 4.10).16 commitments in the horizontal section of their GATS schedules as of April 2002, by type of natural person The periods for which entry may be permit- ted have not always been indicated. This is sur- Number of Percentage of entries entries prising because it might be expected that, in the absence of a definition of “temporary” in the Intracorporate transferees, GATS, members would provide more precision of which 168 42 Executives 56 in their schedules. Where time limits have been Managers 55 specified, the relevant periods are shorter for Specialists 56 business visitors than for executives, managers, Others 1 Executives 24 28 and specialists. The focus of existing commit- Managers 42 ments on employed persons is reflected also in Specialists 44 members’ frequent use of employment links as Business visitors, of which 93 23 Commercial presence 41 an entry criterion: “Pre-employment,” usually Sales negotiations 52 of one year, is one of the most recurrent re- Contract suppliers 12 3 strictions. Numerical quotas and economic Other 17 4 Total 400 100 needs tests rank next in terms of frequency of limitations. While most of the quotas relate to Source: Mattoo and Carzaniga (2003). the total staff of a company, some members also have reserved the right to operate quotas agers, and specialists. Of these, some 170 en- based on parameters, such as senior staff or tries explicitly relate to intracorporate transfer- wages. Significant administrative discretion re- ees. Only 17 percent of all horizontal entries sults from the frequent scheduling of economic may cover low-skilled persons as well (“busi- needs tests without indication of the criteria on ness sellers” and “other”). It is also revealing which they are operated; with such entries, the that few significant differences exist between relevant government agency grants access to Box 4.10 Key impediments to Mode 4 trade F ive policy impediments discourage Mode 4 trade. tests also may cause significant delays in hiring procedures. • Issuance and renewal of visas and work permits • Quantitative restrictions on the movement of nat- may be cumbersome, expensive, stringent, and ural persons with a view to protecting local labor lack transparency. markets. • Social security contributions (lack of tax credits in • Economic needs tests and labor certification the home country), double taxation burdens requirements, whereby prospective employers placed on foreign workers, non-portability of must certify that no domestic workers were avail- pension and other social contributions. able prior to hiring a foreign worker. Particularly • Lack of recognition of qualifications, educational troublesome is the lack of transparency and the degrees, training, and experience, especially in high degree of administrative discretion applied regulated professions. to such tests, which reduces the predictability of trading conditions. The administration of such Source: Mattoo (2003). 169 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 foreign natural persons provided that unspeci- Providing clearer information on economic fied economic conditions are met. needs tests (where entry of foreigners is sub- ject to an assessment of needs in the domestic What’s on the table in the current market), such as criteria used, responsible au- negotiations? thorities, likely timeframe for determinations, Proposals related to Mode 4 in the current ser- and record of recent decisions (Nielson 2002). vices negotiations by both developed and de- veloping countries address many of the issues Greater transparency. Existing access is not identified above.17 Six proposals relate specif- always used because service suppliers lack ically to Mode 4; others raise Mode 4 in the information on the necessary requirements context of sectoral proposals. Some propose and procedures. WTO members could provide ways to expand existing market access, either one-stop information on all relevant proce- through the development of sectoral commit- dures and requirements via a dedicated website ments or by expanding access available to one covering all WTO members, through notifica- group (such as intracorporate transferees) or tions to the WTO, or by creating a one-stop the categories of personnel that benefit from contact point at the national level. Other sug- favorable Mode 4 access. Other proposals seek gestions include prior consultation on regula- to improve the level of access by removing ob- tory changes, timely responses to applications, stacles to existing commitments, such as lack and the right of appeal (Nielson 2002). of information or cumbersome and inappro- priate administrative procedures. Some make Adoption of a GATS visa. This would facili- links to the development of broader regulatory tate entry of Mode 4 workers, including avoid- disciplines under GATS Article VI.4, or raise ance of the detailed visa procedures currently specific barriers such as economic needs tests required in many countries (often not sepa- or recognition of qualifications. rated from permanent migration). India has The negotiating proposals on Mode 4 put forward the idea of a GATS visa, which tabled by WTO members pursue two core ob- would be issued rapidly, be time-limited, cover jectives. One class of proposals, favored by de- both independent service suppliers and intra- veloping countries, focuses on widening mar- corporate transferees, feature rights of appeal, ket access. Another, preferred by developed and be backed up by a bond, with sanctions countries, aims at increasing the effectiveness for abuse. The main idea behind the proposal of existing market-access commitments (Niel- is to distinguish between temporary and per- son 2002). Together, such proposals provide a manent flows of migrants in the administra- useful roadmap of what an improved and more tion of entry procedures.18 The key elements equitable outcome on Mode 4 trade could of a GATS visa scheme are presented in box comprise within the framework of the Doha 4.11 (Nielson 2002). Development Agenda. Key issues under dis- cussion include: Enhanced market access commitments. There are several additional areas where expanded Greater clarity and predictability in WTO market access for specific groups would sub- members’ commitments. Common definitions stantially increase the scope for developing for main personnel categories are included in countries’ Mode 4 entry: many WTO members’ commitments. Many members refer to “executives, managers, spe- • Commitments for particular service sec- cialists,” but there is no common understand- tors in high demand (such as ICT, pro- ing of who is covered by these categories; use of fessional services) rather than the current a worker category nomenclature developed by blanket treatment for Mode 4 entry across the ILO could be useful in this regard. all sectors; 170 L A B O R M O B I L I T Y A N D T H E W T O : L I B E R A L I Z I N G T E M P O R A R Y M O V E M E N T Box 4.11 Elements of a possible GATS visa/permit regime Coverage: Either all categories of service providers demonstrated experience of performing services at covered by sectoral and horizontal commitments senior level; proof of qualifications for some senior under Modes 3 and 4 (visas), or only intracorporate levels of personnel; contracts above a certain value transferees (including at trainee level) and key per- not subject to economic needs tests. sonnel providing services pursuant to a contract Role of companies: A company-specific GATS between two businesses (permits). visa for personnel working for well-known and rep- Duration of stay: Less than 12 months; no sin- utable companies. Following certification by immi- gle visit to exceed 365 days; 3 years for intracorpo- gration authorities, companies could self-administer rate transferees. Stays of less than 3 months (but transfers. possibly multiple entries over the course of a year) Appeal rights: Appeal against rejection, with a would not require a visa. decision within one month. Procedure: A separate body dealing with GATS Renewal: Simple procedures with fees reflecting visas as contact point within the overall immigration administrative costs. framework; a one-source availability of all relevant Prevention of abuse: Declaration of intention rules and regulations; information on the status of not to establish a permanent residence; inability to applications to be available upon request; authorities change to another visa category during life of the required to provide notification of delays; expedited GATS visa; payment of bonds by sponsoring com- security checks; consultation mechanism for any pany to local embassy or consulate; imposition of changes to the rules. special safeguard of one year’s duration against any Time for issuance: 2 to 4 weeks from filing of WTO member whose companies have a pattern of application to issuance of visa, but with procedures visa abuse. for issuance in one day or at port of entry under special circumstances. Conditions: For intracorporate transferees, proof of employment with current employer for a Sources: OECD (2001), drawing on Chanda (1999), Zutshi defined period (6 months) and performance bonds; (2000), and European Services Forum (2001). • Better access for some groups, in par- fessional service providers working on a con- ticular intracorporate transferees, via tract basis. “blanket” applications by companies or by charging companies for streamlined Of the six proposals tabled specifically on processing (including via a GATS visa); Mode 4 by WTO members to date, four are • More access for other types of skilled, by developed countries—Canada, the Euro- but not necessarily highly skilled, per- pean Union, Japan, and the United States— sonnel such as “technical support person- whereas only two are from developing coun- nel,” “nonprofessional essential per- tries—Colombia and India. The fact that so sonnel,” and trainees (future executives) few developing country members of the WTO (Nielson 2002); have articulated negotiating proposals in an • Progressively reducing the range of ad- area of obvious export interest is somewhat missible worker categories subject to surprising. This lack of interest may connote a labor market/economic needs tests, with preference for the guaranteed access afforded no economic needs tests applied to in- to sending countries by bilateral guest worker tracompany transferees or to certain pro- programs (an outcome that appears to mirror 171 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 the tendency for some developing countries to porting countries the financial and knowledge pursue preferential bilateral trade agreements benefits would be greatest if service suppliers rather than multilateral agreements). It also return home after a certain period abroad, and may reflect the difficulties many developing for importing countries the temporary move- countries have faced in identifying their export ment would create fewer domestic problems interests in services trade, an area of high de- than immigration. mand in trade-related capacity building. The However, there are a number of significant dearth of negotiating proposals need not, how- issues and concerns to be addressed. Experience ever, imply that individual developing coun- with bilateral or regional temporary worker tries are not formulating specific requests for schemes might highlight some of the practical greater access for their workers to developed- means of tackling policy challenges and con- country markets in the context of ongoing cerns associated with temporary movement— bilateral request-offer negotiations under the issues such as the operation of bonding re- GATS. quirements, avoidance of double taxation of temporary workers, repatriating social security Discovering mutual interests is essential and pension contributions to the sending coun- not only for the success of Mode 4 try, ensuring that the temporary nature of entry negotiations but also for the GATS not be abused, and on-site inspections of work as a whole sites employing TMNP workers.19 The success of the GATS negotiations may de- None of these issues are insurmountable— pend on progress on Mode 4 trade. As Mattoo but they require a new level of policy dialogue (2003) notes, liberalizing advances in the mul- and coordination among trade, labor, and mi- tilateral trading system have always derived gration authorities, both at the national and from the reciprocal exchange of market-access international level, to find workable solutions concessions. It is important that developing (Nielson 2002). countries understand the potential of, and press for, enhanced access in an area of natural com- parative advantage. Such an understanding, if Notes not opposed by the OECD countries, should 1. Two definitions of migrants are used: Europe and Japan usually refer to country of citizenship in defining enable developing countries to engage more ef- “foreign,” whereas in the United States, Australia, and fectively in the GATS negotiations. Canada country of birth is the relevant definition. Furthermore, there is reason to believe that 2. Available statistics are incomplete and not read- reduced barriers to the temporary movement ily comparable between countries. While most migra- of service providers will produce substantial tion systems distinguish between temporary and foreign global benefits. Significant gains already are migration, the definition varies among countries. To a being realized, for example, in the software in- certain extent, statistics on highly skilled workers tend to be better, because data on such workers are collected dustry—some 60 percent of India’s burgeon- in connection with their temporary visas. Work permits ing exports are provided through the move- and visas are valuable sources of data (OECD 2001b). ment of software engineers to the site of the The situation is even more difficult for statistics on consumer. And with greater liberalization of those temporary foreign workers falling under GATS barriers to the movement of people, many Mode 4—for example, Mode 4 entrants usually cannot more developing countries could “export” at be separated from broader groups, and even when mi- least the significant labor component of ser- gration data provide occupations—such as “man- agers”—they are not disaggregated by sector. Further, vices such as construction, professional ser- many business visitors may enter under tourist visas vices, environmental services, and transport. and not appear in employment-related figures, particu- A benefit of the temporary nature of such larly where no short-term business visitor visa exists. In- movement is the potential for both the host dustry surveys can be a useful, but limited, source of country and the home country to gain. For ex- data for Mode 4 (Nielson and Cattaneo 2003). 172 L A B O R M O B I L I T Y A N D T H E W T O : L I B E R A L I Z I N G T E M P O R A R Y M O V E M E N T 3. This section draws heavily upon the chapter on 10. In health services, the World Health Organiza- labor mobility prepared by Julia Nielson of the OECD tion has suggested offsetting earnings generated by mi- Secretariat in the study “Regional Trade Agreements grant service workers against (1) reduced domestic ac- and the Multilateral Trading System” prepared for the cess to these services, (2) loss in the quality of services, Trade Committee of the OECD (OECD 2002c). and (3) loss of public investment (Scholtz 1999). 4. H-1B visas are also available for professionals 11. Circumstances may even induce a deliberate entering the United States. Some main differences be- policy of encouraging migration as a way of combating tween H-1B and TN visas include: H-1B visas include unemployment (Abella and Abrerar-Mangahas 1997). requirements to show that temporary hires will not ad- The effectiveness of such a strategy may be limited by versely affect U.S. workers; TNs are granted for one the reluctance of workers to accept a job abroad as a year, but renewals are unlimited, whereas H-1B visas substitute for one at home (OECD 2002d). have a three-year duration with one renewal (up to six 12. Borjas (2000) suggests that immigration may years). Similar conditions to TNs are applied to traders contribute to improving domestic-factor use by com- and investors and intracompany transferees under pensating for the reluctance of native workers to move E1/E2 and L1 visas, respectively (OECD 2001b, citing from areas of relative labor surplus to areas of short- Globerman 2000). age. Such findings hold especially in health-related pro- 5. Provisions facilitating mutual recognition are in- fessions, with obvious social benefits for populations in cluded in some agreements (for example, EFTA), and more geographically remote areas. others have complementary arrangements. For exam- 13. Mode 4 is defined in Article I:2(d) as entailing ple, the ANZCERTA Services Protocol, the Trans- “the supply of a service . . . by a service supplier of one Tasman Travel Arrangement, and the Trans-Tasman Member, through presence of natural persons of a Mutual Recognition Arrangement together provide Member in the territory of any other Member.” The that persons registered to practice an occupation in one Annex on Movement of Natural Persons Supplying Ser- country can practice an equivalent profession in an- vices under the Agreement (hereinafter the Annex) other (OECD 2002e). specifies that two categories of measures are covered: 6. This section of the chapter relies heavily on those affecting natural persons who are “service suppli- “Service providers on the move: economic impact of ers of a Member”; that is, self-employed suppliers who Mode 4” prepared by Olivier Cattaneo and Julia Niel- obtain their remuneration directly from customers; and son of the OECD Secretariat for the Trade Committee those affecting natural persons of a Member who are of the OECD (OECD 2002d). “employed by a service supplier of a Member in respect 7. With the exception of the “settlement countries” of the supply of a service.” These natural persons can (Australia, Canada, New Zealand, United States), or be employed either in their home country and be pre- others with significant migration (Germany, United sent in the host market to supply a service, or employed Kingdom), many WTO members do not currently have by a service supplier in the host country. specialized regimes in place to deal with temporary en- The Annex clarifies that the GATS does not apply trants as service providers (OECD 2002d). to measures affecting individuals seeking access to the 8. Not everyone agrees that permitting workers to employment market of a member, or to measures re- move abroad temporarily, or indeed to emigrate perma- garding citizenship, residence, or permanent employ- nently, reduces the sending country’s welfare. Stark and ment. There is no specified timeframe in the GATS of Wang (2001) suggest that emigration can have the op- what constitutes “temporary” movement; this is de- posite effect—that is, improve the welfare of those left fined negatively, through the explicit exclusion of per- behind. They argue that migration opportunities create manent presence. A cursory look at members’ sched- a strong incentive to acquire greater skills through edu- ules shows that the maximum length of stay permitted cation. Only a portion of graduates will emigrate, while under Mode 4 varies with the underlying purpose. many will remain behind, better educated than they Thus, while business visitors generally are allowed to would have been if immigration opportunities had not stay up to 90 days, the presence of intracorporate been provided (Winters 2003b). Such effects thus can transferees, another frequently scheduled category, generate spillover benefits in sending countries, effects tends to be limited to periods of between two and five that are likely to be felt intergenerationally (Comman- years. The Annex does provide for the possibility that der, Kangasniemi, and Winters 2002). commitments, and therefore access conditions, may be 9. This section of the chapter draws heavily on scheduled by “categories of natural persons,” thereby “Service providers on the move: economic impact of introducing an additional element of flexibility. Mode 4” prepared by Olivier Cattaneo and Julia Niel- The Annex also clarifies that, regardless of their son of the OECD Secretariat for the Trade Committee obligations under the Agreement, members are free to of the OECD (OECD 2002d). regulate the entry and stay of individuals in their terri- 173 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 tory, including through measures necessary to protect country but does not have a commercial presence in the integrity of their borders and to ensure the orderly that market. movement of natural persons across those borders, 17. This section of the chapter relies heavily on provided that the measures concerned “are not applied Nielson (2002) and OECD (2001b). in such a manner as to nullify or impair the benefits ac- 18. Although the Indian proposal for the adoption cruing to any Member under the terms of a specific of a GATS visa has helped to broaden the scope of commitment.” The operation of visa requirements only Mode 4 discussions among trade and immigration of- for natural persons of certain members, but not for ficials, the odds of seeing such a scheme adopted in the others, is not per se regarded as nullifying or impairing DDA seem remote. Indeed, the sobering experience such benefits. emerging from attempts to implement the APEC Busi- 14. This is a strange distinction—are temporary ness Travel Card, epitomized by the reluctance of three foreign workers engaged in picking apples temporary key APEC Members (Canada, Japan, and the United agricultural workers or suppliers of fruit-picking ser- States) to implement the scheme, suggests a long road vices? Is an employee of General Electric’s consumer ahead in liberalizing TMNP at the multilateral level credit arm engaged in service or manufacturing activi- (OECD 2001b). It should be noted that from the point ties? (Nielson 2002) of view of migration authorities, TMNP represents a 15. Calculated on a sample of 37 sectors deemed small proportion of those crossing borders every day. representative for various services areas. (See docu- The additional resources required to create special ment S/C/W/99, March 2, 1999). The shallow level of treatment for such persons—which a GATS visa would commitments for Mode 4 is to a certain extent also re- entail—are hard to justify in the face of other priorities, flected in the pattern of horizontal limitations, which notably in border security, arising for larger groups of apply across all sectors: there are five times as many migrants. Such resources also could be well beyond the limitations scheduled for Mode 4 than for Mode 2. administrative capacities of many developing country In turn, this reflects many members’ basic method WTO members (Nielson 2002). See OECD (2001b) for to scheduling Mode 4 entries. Contrasted with other more discussion of the potential impact of a GATS visa modes, the “negative list” approach to scheduling lim- scheme. itations has been turned upside down: schedules start 19. Winters and others (2002, pp. 43–50) provide a with a general “unbound” which is then qualified by useful summary of such programs and the means to en- liberalization commitments, mostly limited to specified force them in France, Germany, the United Kingdom, types of persons (for example, managers), movements and the United States. (intracorporate), and stays (up to four years). Commitments are often exclusively governed by what is inscribed in the horizontal part of the schedule, References so that identical access conditions apply to all sched- Abella, Manolo I., and M. A. Abrera-Mangahas. 1997. uled sectors. Commitments usually are based on func- Sending Workers Abroad: A Manual for Low- tional or hierarchical criteria, related either to the type and Middle-Income Countries. Geneva: Interna- of person involved (executive, manager, specialist) or tional Labor Office. to the purpose of their movement (for example, to es- Athukorala, P., and C. Manning. 1999. Structural tablish business contacts, negotiate sales, set up a com- Change and International Migration in East Asia. mercial presence). Besides, no generally agreed defini- Melbourne: Oxford University Press. tions or precise descriptions exist of the types of Borjas, George J. 2000. Labour Economics, 2nd ed. natural persons to which access is granted, which can Boston: McGraw Hill. detract from the predictability of entry conditions. Borjas, George, Richard B. Freeman, and Lawrence F. 16. Access conditions scheduled by countries ac- Katz. 1997. “How Do Immigration and Trade Af- ceding to the WTO after 1995 also are substantially fect Labour Market Outcomes?” In Brookings identical to the ones scheduled by Uruguay Round par- Papers on Economic Activity, pp. 1–67. ticipants. This contrasts with the situation in the three Chadah, Rajesh. 2000. GATS Negotiations and Devel- other modes of supply, for which recently acceded oping Countries: A Case Study of India. Mimeo. members have generally undertaken deeper commit- Washington, D.C.: The World Bank. ments. The only detectable difference with regard to Chanda, R. 1999. “Movement of Natural Persons and Mode 4 is a relatively higher number of commitments Trade in Services: Liberalising Temporary Move- scheduled by recent WTO members for “contract sup- ment of Labour under the GATS.” New Delhi: In- pliers”—that is, employees of a foreign enterprise who dian Council for Research on International Eco- have completed a contract to supply a service in a nomic Relations. Available at www.icrier.res.in. 174 L A B O R M O B I L I T Y A N D T H E W T O : L I B E R A L I Z I N G T E M P O R A R Y M O V E M E N T Chanda, Rupa. 2002. “Movement of Natural Persons and Washington, D.C.: Oxford University Press and the GATS: Major Trade Policy Impedi- and World Bank. ments.” In Bernard Hoekman, Aaditya Mattoo, Mattoo, Aaditya, and Antonia Carzaniga, eds. 2003. and Philip English, eds., Development, Trade, and Moving People to Deliver Services. New York the WTO. Washington, D.C.: World Bank. and Washington, D.C.: Oxford University Press Commander, S., M. Kangasniemi, and L. Alan Winters. and World Bank. 2002. “The Brain Drain: Curse or Boon? A Sur- Nielson, Julia. 2002. “Service Providers on the Move.” vey of the Literature.” Paper prepared for Inter- Evian Group Compendium, Lausanne: Evian national Seminar on International Trade, May, Group. Stockholm. OECD (Organisation for Economic Co-operation and Devan, J., and P. S. Tewari. 2001. “Brains Abroad.” Development). 1998. Migration, Free Trade, and The McKinsey Quarterly, No. 4. Regional Integration in North America. Paris. European Services Forum. 2001. “Draft Model Sched- ———. 2000. Globalisation, Migration, and Develop- ule Covering the Temporary Entry of Natural Per- ment. Paris. sons under the General Agreement on Trade in ———. 2001a. Migration Policies and EU Enlarge- Services.” July 20. Brussels: European Science ment: The Case of Central and Eastern Europe. Foundation. Paris. Galor, O., and O. Stark. 1991. “The Probability of ———. 2001b. “Service Providers on the Move: A Return Migration, Migrants’ Work Effect, and Closer Look at Labor Mobility and the GATS.” Migrants’ Performance.” Journal of Development TD/TC/WP/Final. Paris. Economics, No. 35. ———. 2001c. Trends in International Migration. Ghose, Ajit K. 2002. “Trade and International Labor 2001 SOPEMI Report. Paris. Mobility.” Employment Paper 2002/33. Interna- ———. 2002a. International Mobility of the Highly tional Labor Office, Geneva. Skilled. Paris. Ghosh, Bimal. 1998. Gains from Global Linkages: ———. 2002b. “Labor Mobility in Regional Trade Trade in Services and Movements of Persons. Agreements.” TD/TC/WP/Final. Paris. New York: St. Martins’ Press. ———. 2002c. Migration and the Labour Market in Globerman, Steve. 2000. Trade Liberalisation and the Asia: Recent Trends and Policies. Paris. Migration of Skilled Professionals and Managers: ———. 2002d. “Service Providers on the Move: The The North American Experience. London: Black- Economic Impact of Mode 4.” TD/TC/WP/Final. well Publishers. Paris. Hodge, James. 1999. “Examining the Cost of Services ———. 2002e. “Service Providers on the Move: Mutual Protection in a Developing Country: the Case of Recognition Agreements,” TD/TC/WP/48. Paris. South Africa.” Paper presented at the World Ser- ———. 2002f. Trends in International Migration. vices Congress, November, Atlanta. 2002 SOPEMI Report. Paris. Home Office. 2001. “Migration: an Economic and So- Papademetriou, Demetrios G. 2000. “Labor Mobility cial Analysis.” RDS Occasional Paper 67. Lon- and Human Resources Development Policies.” don: Home Office. Globalisation, Migration, and Development. Paris: Hutton, T., and J. G. Williamson. 2002. “What Funda- OECD. mentals Drive World Migration?” NBER Working PSI (Public Services International). 1999. “The WTO Paper 9159. National Bureau of Economic Re- and the GATS: What Is at Stake for Public search, Cambridge, Mass. September. Health?” Common Concerns for Workers in Edu- International Labor Office (ILO). 1996. Emigration cation and the Public Sector Series. Brussels. June. Pressures and Structural Change: Case Study of Rodrik, Dani. “Feasible Globalisations.” NBER Work- Indonesia. Geneva. ing Paper W9129. National Bureau of Economic Lindert, Peter H., and Jeffrey G. Williamson. 2001. Research, Cambridge, Mass. August. “Does Globalization Make the World More Un- Scholtz, Peter. 1999. “International Trade Agreement equal?” National Bureau of Economic Research and Public Health: WHO’s Role.” Paper prepared (NBER) Paper 8228. Cambridge, Mass. for conference on Increasing Access to Essential Manning, Chris. 2002. “Structural Change, Economic Drugs in a Globalized Economy, Geneva, World Crisis, and International Labor Migration in East Health Organization, April 11–12. Asia.” The World Economy 25(3): 359–85. Stark, Oded, and Yong Wang. 2001. Inducing Human Mattoo, Aaditya. 2003. “Introduction and Overview.” Capital Formation: Migration as a Substitute for In Aaditya Mattoo and Antonia Carzaniga, eds., Subsidies. Reihe Okonomie Economics Series Moving People to Deliver Services. New York 100. Vienna: Institute for Advanced Studies. 175 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 UNCTAD (United Nations Commission on Trade and ment of Natural Persons.” Mimeo. Brighton: Uni- Development). 2001. “Movement of Natural Per- versity of Sussex. sons under the GATS: Perspectives for the New Winters, Alan, T. L. Walmsley, Z. K. Wang, and R. Negotiations.” Mimeo. Geneva. Grynberg. 2002. “Liberalising Labor Mobility United Nations. 2000. “Replacement Migration: Is it a under the GATS.” Economics Discussion Paper Solution to Declining and Ageing Populations?” 87, University of Sussex, Brighton, U.K. October. Population Division, Department of Economic World Bank. 1995. World Development Report 1995: and Social Affairs, New York: United Nations. Workers in an Integrating World. Washington, Visco, Ignazio. 2000. “Immigration, the Labour Mar- D.C. ket, and Development.” Paper presented at con- World Bank. 2000. East Asia: Recovery and Beyond. ference on Migration Scenarios for the 21st Cen- Washington, D.C.: The World Bank. tury, Rome, July. ———. 2001. Globalization, Growth, and Poverty: Werner, Heinz. 1996. Temporary Migration for Em- Building an Inclusive World Economy. Washing- ployment and Training Purposes, Social Cohesion, ton, D.C. and Quality of Life. Brussels: Council of Europe. ———. 2001. Global Economic Prospects and the De- Winters, Alan. 2001. “Assessing the Efficiency Gain veloping Countries 2002: Making Trade Work for from Further Liberalisation: A Comment.” In the World’s Poor. Washington, D.C. Roger Porter, Pierre Sauve, Arvind Subramaniam, ———. 2002. Global Economic Prospects and the De- and Americo Beviglia-Zampetti, eds., Efficiency, veloping Countries 2003: Investing to Unlock Equity and Legitimacy: The Multilateral Trading Global Opportunities. Washington, D.C. System at the Millennium. Washington, D.C.: ———. 2003. Global Development Finance 2003: Brookings Institution Press. Striving for Stability in Development Finance. Winters, Alan. 2003a. “The Economic Implications of Washington, D.C. Liberalising Mode 4 Trade.” In Aaditya Mattoo WTO (World Trade Organization). 1998. “Presence of and Antonia Carzaniga, eds., Moving People to Natural Persons (Mode 4).” Background note by Deliver Services. New York and Washington, the Secretariat, S/C/W/75, Council for Trade in D.C.: Oxford University Press and World Bank. Services. Geneva. December 8. ———. 2003b. “GATS Mode 4: The Temporary Move- Zutshi, B.K. 2000. “Liberalisation of Temporary Mo- ment of Natural Persons.” Mimeo. Brighton: Uni- bility of Natural Persons.” Intervention before the ersity of Sussex. European Services Forum International Confer- Winters, Alan O., and Terrie L. Walmsley. 2002. “Re- ence on GATS 2000 negotiations, Brussels, No- laxing the Restrictions on the Temporary Move- vember 27. 176 5 Reducing Trading Costs in a New Era of Security Security measures can drive up vulnerable to cost increases related to security transport costs threats. Limited budget resources, dependence In the wake of September 11 and worldwide on foreign trade and investment, and outdated worries about terrorism, governments every- infrastructure and technology present serious where have enacted security measures that challenges for these countries. could, if not managed properly, drive up trade Fortunately, new security protocols being costs and shut out exports from developing deployed at ports, customs offices, and border countries. This action has focused attention on posts around the world have the potential to the search for greater efficiency in interna- streamline trade transactions as well as pro- tional transportation, the need for cooperation mote safety and security. However, a global in adopting collective measures to promote framework must be established to ensure that transport security, and the imperative of im- the needs of developing countries are addressed proving customs regimes, port facilities, and as security regimes take shape. The G-8 and logistics management. developing-country partners should take the The cost of moving goods between destina- lead in drafting such a framework. tions and across international borders is often as important as formal trade barriers in deter- Regulations hamper competition in mining the cost of landed goods—and ulti- international transport systems mately of market share. The costs of transport and raise costs among many points are as significant as tar- Anticompetitive regulations and private com- iffs. Other delays are equally costly. One study mercial practices inflate trade costs by restrict- estimates that every day spent in customs adds ing international air and maritime transport nearly 1 percent to the cost of goods. In devel- services to developing countries. The share of oping countries, transit costs are routinely two trade shipped by air has grown to 30 percent to four times higher than in rich countries. for U.S. imports in 1998, but international air transport is one of the service sectors that is But they hold out the promise of most heavily shielded from international com- facilitating and securing trade petition. By denying entry to efficient outside A study of the trade effects of September 11 carriers, bilateral air service agreements in- estimated that world welfare declined by $75 crease export costs for developing countries. billion per year for each 1 percent increase in Though international airline alliances increase costs to trade from programs to tighten border network efficiency, they can be harmful if they security. Developing countries are particularly impede effective competition. City-pair routes 179 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 on which more than two passenger airlines or developing countries. In particular, appropri- dedicated freight airlines operate can cut costs ate legal and regulatory frameworks are by an average of more than 10 percent. needed to ensure competition. Developing Maritime transport is often subject to prac- countries need to address such domestic re- tices such as cargo reservation schemes and form to take advantage of the opportunities limitations on port services that protect inef- offered by a liberalized trading system. ficient service providers. Such competition- restricting practices among shipping lines and —and new multilateral efforts could prove port operators can increase freight rates up to beneficial 25 percent on some routes. Rising concentra- Multilateral efforts to reduce transport fric- tion in the market for port terminal services tions could include revamping competition- has increased the risk that private firms may restricting regulations in air and maritime capture the benefits of government reforms. transport. Such an effort might include revisit- Abusive practices by private operators are of ing antiquated exemptions of transport from special concern in developing countries, where OECD antitrust legislation. Involving develop- traffic volumes are lower and competitive ing countries more centrally in global security forces inherently more limited. planning, together with a program of appropri- ate technical assistance, would help developing Investments in improving ports, customs, countries mitigate security-driven cost increases and trade-related institutions can have a that would otherwise reduce their participation substantial payoff in the global market. A commitment to multi- Building capacity in trade-related services can lateral efforts on trade facilitation would also provide the great gains in this new environ- have a high payoff—the World Customs Orga- ment. If the countries now below the world nization (WCO), the multilateral development average in trade-facilitation capacity could be banks, bilateral donors, and private groups are raised halfway to the average, trade among 75 all important players. The leadership of the countries would increase by $377 billion an- G-8 should join multilateral and other develop- nually, according to new analyses outlined in ment institutions in a plan to facilitate and ex- this chapter. Facilitating trade to improve pand trade, strengthen security, and promote export-led growth therefore depends on policy domestic development. reform, technical assistance, and moderniza- Broad trade facilitation goals do not fit tion of infrastructure. All trading partners can neatly into the disciplines of the World Trade benefit when barriers are removed and capac- Organization (WTO). In contrast to stroke- ity is strengthened—with many of the benefits of-the-pen tariff reductions, improving ports, of reform and modernization flowing directly customs, and logistics involves a continuing to developing countries. process of institutional changes that move countries toward best practice. The lion’s Domestic policy reform is now even share of the agenda requires national action, more important— supported by multilateral development agen- Domestic policy reform is needed to ensure cies to promote—and in some cases finance— that the benefits of modernized customs, port institutional changes. However, if the Doha facilities, and related investments in informa- Round propels the WTO into a supporting role tion technology are realized. Streamlining reg- in the broader trade-facilitation agenda, nego- ulations to remove technical barriers and lib- tiations on simplified and harmonized trade eralizing transport and telecommunications procedures could advance best practice in ad- can promote domestic competition and signif- ministering fees and formalities in trade and in icantly lower transport costs while expanding reducing the costs and uncertainty of transit the availability and choice of services in many trade, especially for land-locked countries. 180 R E D U C I N G T R A D I N G C O S T S I N A N E W E R A O F S E C U R I T Y Most importantly, obligations undertaken port costs determine potential access to for- by developing countries should be carefully tai- eign markets, which in turn explains up to 70 lored to long-term implementation capacity. percent of the variance in countries’ GDP per Any new agreement should include innovative capita. Among the problems that add to the procedures for settling disputes before they costs of trade are: move toward WTO-sanctioned action. • Frequent reloading of goods • Port congestion affecting turnaround Why transport, trade facilitation, time for feeder vessels and logistics matter • Complicated customs-clearance proce- T he costs of transporting developing-coun- try exports to foreign markets are a much greater hindrance to trade than are tariffs. A dures • Complex and nontransparent adminis- trative requirements, often pertaining to comparison of countries’ “transport cost inci- documentation dence” (the share of international shipping • Limited use of automation leading to costs in the value of trade) and their tariff in- high costs for processing information cidence (the trade-weighted ad valorem duty • Uncertainty about the enforceability of actually paid) shows that for 168 out of 216 legal trade documents such as bills of U.S. trading partners, transport cost barriers lading or letters of credit. outweigh tariff barriers. For the majority of Sub-Saharan African countries, the tariff inci- Policies to remove nontariff barriers and ac- dence was relatively insignificant, at less than celerate the flow of goods and services across 2 percent, while their transport cost incidence borders—in short, to facilitate trade—are thus exceeded 10 percent (World Bank 2001). A at the forefront of today’s trade-policy debate.1 doubling of shipping costs is associated with Cross-country evidence suggests that high slowdowns in annual growth equivalent to transport costs tax growth in countries with more than one-half of a percentage point. underdeveloped transport links (World Bank Trade-related transaction costs—freight 2001). Inefficient internal transport systems charges as well as other logistical expenses— can widen income inequalities within countries are a crucial determinant of a country’s ability by separating the hinterland regions from the to participate in the global economy. Trans- global marketplace. Box 5.1 The evolving definition of trade facilitation OECD: “Simplification and standardization of pro- APEC: “Trade facilitation generally refers to the sim- cedures and associated information flows required to plification, harmonization, use of new technologies, move goods internationally from seller to buyer and and other measures to address procedural and ad- to pass payments in the other direction.” ministrative impediments to trade.” UN/ECE: A “comprehensive and integrated approach APEC: “The use of technologies and techniques which to reducing the complexity and cost of the trade will help members to build up expertise, reduce costs transactions process, and ensuring that all these ac- and lead to better movement of goods and services.” tivities can take place in an efficient, transparent, and predictable manner, based on internationally accepted Source: Wilson and others (2002), citing various institutional norms, standards, and best practices.” sources. 181 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 The new international security will go into force on July 1, 2004, for vessels in dimension in trade international trade, contains detailed security- related requirements for shipping companies, T he terror and tragedy of September 11, 2001, have emphasized the need for re- forms in border and transport infrastructure. port authorities, and governments, together with guidelines on meeting the requirements. Terrorist attacks can seriously disrupt the pas- The new rules cover security plans, security of- sage of people, goods, and modes of transport ficers, and certain security equipment. across borders. Measures designed to stop ter- In the United States, the Maritime Trans- rorism can add certainty and stability to the portation Security Act of 2002 (MTSA), signed global economy, raise investor confidence, and by President Bush in November 2002, is in- facilitate trade. Secure trade is now as impor- tended to improve safeguards at the country’s tant as free trade—and the two need not be 361 sea and river ports and to improve intelli- mutually exclusive.2 gence on cargo and personnel entering U.S. Since the September 11 attacks, billions ports. Many of the requirements imposed by of dollars have been spent to enhance port the IMO protocol also are mandated by the security, install airport security equipment, MTSA. Port-security efforts have been ex- strengthen customs authorities, and bolster tended with the introduction of the Anti- border security. While much attention has Terrorism and Port Security Act of 2003. been devoted to new security protocols in the In April 2002, the trade community and the United States, security plans in other parts of U.S. Customs Service (USCS) launched the the world also have been revised and strength- Customs-Trade Partnership Against Terrorism ened.3 The G-8 has committed itself to in- (C-TPAT) to improve security along the entire creasing security for all transport modes and transport chain. The initiative encompasses to promoting policy coherence and coordina- manufacturers, warehouse operators, and tion among international organizations such shipping lines. Participation in the voluntary as the International Civil Aviation Organiza- scheme is open to all importers, airfreight con- tion (ICAO), International Maritime Organi- solidators, carriers, and non-vessel-owning zation (IMO), and WCO. common carriers that agree to comply with the The bombing of the VLCC Limburg off supply-chain security profile. Under the pro- the coast of Yemen in 2002 was a stark re- gram, importers or carriers provide USCS with minder of weaknesses in global maritime sys- documentation relating to security measures at tems, which handle 95 percent of world trade. each step along the route of goods—from the The event alarmed the shipping world and factory to the warehouse, the port, and the prompted sweeping new security proposals, ocean carrier.4 several of which are outlined below. The United States has imposed new controls to increase the screening of freight containers The security of maritime transport has arriving at and leaving ports with goods bound been strengthened, but the costs and for the United States. Almost 90 percent of benefits of the new security programs all freight is transported in containers, 244 mil- have yet to be assessed lion of which move annually among the A series of measures aimed at strengthening world’s seaports. The Container Security Ini- maritime security and suppressing acts of ter- tiative (CSI), introduced in January 2002 by rorism was adopted by the IMO at its diplo- the USCS, is designed to prevent terrorists matic conference in December 2002. These in- from concealing personnel or weapons of mass cluded changes to the 1974 Safety of Life at destruction in U.S.-bound cargo. Participating Sea Convention (SOLAS), which covers 98 countries agree to help the USCS identify and percent of the world’s fleets. The International screen high-risk containers at the earliest stage. Ship and Port Facility Security Code, which Beginning with the world’s 20 busiest ports, 182 R E D U C I N G T R A D I N G C O S T S I N A N E W E R A O F S E C U R I T Y CSI initiative will be extended until 100 per- Under the USCS’s 24-Hour Advance Cargo cent of containerized cargo is covered.5 Manifest Rule, which took effect on February The bilateral agreements that underpin the 2, 2003, carriers must provide cargo manifests CSI may discriminate against ports not covered electronically via the Automated Manifest Sys- by CSI. The European Commission, concerned tem (AMS) 24 hours before loading a container that the United States had approached only bound for a U.S. port. USCS will use the infor- some large European ports, argued that the CSI mation to identify containers that pose a po- could divert trade to Rotterdam, for example, tential risk and determine whether containers and create competitive distortions among ports can be cleared for loading. Ships unable to meet in the European Union—violating EU fair trade the requirements risk receiving “no load” or- rules. Although the top nine northwest Euro- ders and thus being detained at the port of ori- pean ports handle 80–90 percent of Europe’s gin. Failure by a shipper to comply with the no- containerized cargo bound for the United tification requirement carries a fine and the States, the other 11 that also export to the possibility of seizure and forfeiture of the cargo. United States would be affected. In a recent de- Even freight not bound for the United States— velopment, the European Union has given the a shipment from Hong Kong to Canada via European Commission the power to negotiate the United States, for example—must meet the a maritime security agreement with the United requirements. Canada’s Customs and Revenue States to replace the bilateral deals with eight Agency adopted a similar manifest rule for ma- EU countries. In return, the Commission has rine cargo imports in April 2003. decided to drop legal action against EU mem- The U.S. Food and Drug Administration bers that signed deals with Washington.6 (FDA) has proposed registration of an esti- The CSI measure is especially important for mated 400,000 domestic and foreign food fa- countries that send a substantial share of their cilities to prevent a threat to the U.S. food exports to the United States—for example, 20 supply as mandated by the Bioterrorism Act of percent of Malaysian exports are to the United 2002. Starting December 12, 2003, importers States. By not joining the CSI, Malaysian must file advance notice of food shipments goods could lose competitiveness in the global with the FDA. Estimates by the FDA suggest market—a risk not many nations are willing that the U.S. food industry could lose as much to take. Countries that do not implement the as $6.5 million in perishable imports if the rule required procedures would have a competitive for importers is adopted.7 Many agricultural disadvantage because their shipments would commodities such as bananas and broccoli are undergo more complex examinations and thus still growing on the stalk, vine, or tree the day be cleared more slowly. before loading, and in some cases as few as six The WCO passed a resolution on Security hours before.8 Such cargo may spoil if ship- and Facilitation of the International Trade Sup- ments are held up because of documentation ply Chain in June 2002 to enable ports in all requirements. In the highly competitive market 161 member nations to develop programs simi- for agricultural commodities, this risk could lar to the CSI and consider adopting stricter prompt importers in other countries to move security measures. These measures are intended away from U.S. suppliers. The Bioterrorism to enhance security and improve facilitation Act may also be harmful to Indonesia’s small through a comprehensive reform of customs. and medium enterprises, which are big ex- With nations seeking reciprocal inspection porters of food and agricultural products. rights, Japanese officers have been positioned at The USCS intends to extend the advance the ports of Los Angeles and Long Beach to electronic cargo reporting requirement to im- screen high-risk cargo containers bound for ports and exports transported by air and on Japan. Canadian customs inspectors also have land. Final rules are expected by October 1, been posted at Newark, New Jersey, and Seattle. 2003. Since September 11, airlines have spent 183 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 $43 billion on security measures—among them protect cargo through programs of container more thorough baggage checks, greater in- security, container risk assessment, and ad- flight inspection, and new regulations for se- vance electronic information on container con- cure cockpit doors.9 A new passenger data tent. The group also will endeavor to intro- collection system, the Advance Passenger In- duce more effective baggage screening in formation System (APIS), was recently intro- airports in the region, improve coordination duced by the United States; already it has among immigration officials, establish new raised ethical questions about a passenger’s cyber-security standards, develop an advanced right to privacy.10 With regard to air cargo, passenger information system, and devise sys- new security proposals are expected this year tems for tracking and monitoring potential from the U.S. Transport Security Administra- threats. APEC’s new counterterrorism task tion. Road and rail transport operators also force will coordinate these activities.13 have also been the subject of new measures to forestall attacks.11 Developing countries may have a hard Canada has tightened security at airports, time meeting new security requirements ports, and border crossings to prevent ship- Balancing new security priorities with eco- ments to the United States from being delayed. nomic and trade objectives is complicated. Se- The Canadian government will spend $112.7 curity proposals can affect global supply chains million over the next five years to improve se- by requiring costly changes in business prac- curity at maritime borders. Canada’s Customs tices, process redesigns, and new equipment. Self-Assessment and Partners in Protection Critics fear that developing nations could be programs, like C-TPAT in the United States, squeezed out of the global trading system be- are based on the hypothesis that if companies cause of their limited capacity to implement the adopt secure practices, inspectors will be free new international initiatives. High transport to focus on shipments from companies whose costs, poor infrastructure, and the high costs practices are uncertain. With respect to air of border clearance already pose a large obsta- transport, the Canadian Air Transport Secu- cle to their development. Customs services in rity Association has improved luggage screen- many less-developed countries lack qualified ing by installing explosive-detection equip- personnel to operate advanced security equip- ment at many large airports in Canada.12 In ment and the ability to execute the necessary May 2003, the EU adopted a brief that sug- reforms in their domestic administration. In re- gested high security standards on maritime sponse to new security demands, for example, transport to be applied across the member shippers are adding extra cycle time to their states, new requirements on passenger ships on supply chain rather than risk delays or fines.14 domestic voyages, and heightened security of The USCS 24-hour rule has affected ports the entire maritime transport security chain. that accept cargo as few as six hours before Accounting for nearly 60 percent of world departure, dealers in perishable commodities GDP and half of all trade, the 21 countries of that are harvested and loaded within 24 hours, the Asia Pacific Economic Cooperation group and shipments of emergency replacement parts (APEC) have had to adopt new technologies to and medical supplies. Holding additional in- strengthen security without impeding trade. At ventories to hedge against delays and disrup- a recent meeting in Bangkok, APEC adopted tions requires more storage space and more Secure Trade in the APEC Region (STAR)—a operating capital. set of measures to protect cargo, ships making The 24-hour rule also has introduced extra international voyages, international aviation, costs for Indian exporters. Almost 35 percent and people in transit. Ports in the APEC region of outbound trade from India is headed to the now must upgrade security to meet STAR United States, including 600,000 containers. standards. At a meeting in February 2003 in Exporters must now pay additional costs to Thailand, APEC announced its commitment to local agencies that help them with documenta- 184 R E D U C I N G T R A D I N G C O S T S I N A N E W E R A O F S E C U R I T Y tion. While large manufacturers can provide detailed commodity descriptions, the small- Figure 5.1 Customs clearance takes scale and cottage industry units, which are big longer in the developing world than in the exporters, may be unable to provide the cor- OECD, lowering the competitiveness of rect description that is required at a level con- developing-country trade sistent with the Harmonized Tariff Schedule Average days required for customs clearance by sea, (HTS) codes of USCS. by region Descriptions such as “freight of all kinds” are no longer acceptable. If officials are uncer- Developed (7) tain about their contents, containers may miss their scheduled carrier sailing dates. Electronic filing and paperless clearance are additional East Asia challenges. Most transactions handled by ocean and Pacific (9) carriers are still conducted by fax or phone. Many shippers in India still use manual type- Latin America and writers—obviously hindering their ability to Caribbean (4) provide data in electronic form. Many govern- ments may be unaware of the potentially nega- Africa (5) tive trade-related repercussions from inaction on security. Despite limited finances and capacity to facilitate trade, some developing countries South Asia (1) such as Sri Lanka have adopted cargo security measures that are on par with ports in many 0 2 4 6 8 10 12 developed countries. The same applies to Note: The number in parenthesis indicates the number of Bangladeshi facilities that boast advanced de- countries selected from each region to calculate the average. tection devices. These measures, however, are Developed includes France, Germany, Greece, Netherlands, Spain, Sweden, United States; East Asia and Pacific focused on imports into the country, emphasiz- includes China, Hong Kong (China), Indonesia, Malaysia, ing the need to enhance inspection of exports. Philippines, Singapore, Taiwan (China), Thailand, Vietnam; Latin America and Caribbean includes Argentina, Brazil, Airlines and airports throughout Asia are Chile, Mexico; Africa includes Mozambique, South Africa, working toward the goal of screening all Egypt, Guinea Bissau, Angola; South Asia includes India. Source: International Exhibition Logistics Associates checked baggage. The Agency for Air Trans- (http://www.iela.org). port Security in Africa (ASECNA) is investing $27 million to modernize member states’ air- port security infrastructure.15 positioning technology, and other electronic Security-driven improvements can measures—could accelerate global trade while benefit trade improving security (Reddy 2002). Sharing New programs to combat terrorism and cor- information among terminal operators, ship- ruption clearly will involve investment in new pers, and customs brokers can help expedite technology and infrastructure—possibly rais- the movement of freight through terminals ing the costs of trade in the short to medium without any new physical investment. By re- term. At the same time, the prospect of reduc- ducing delays in container clearance through ing future threats through technology-inten- customs, the need for shippers to pay “tea sive customs inspections should be viewed as money”17 to officials would be diminished— an investment in greater trade efficiency.16 Au- contributing to port efficiency (figure 5.1). In tomated technology—such as bar codes, wire- addition, simplification of customs procedures less communications, radio frequency ID tags, can increase the chances of detection of fraud tamper-proof seals for containers with global and criminal activities. 185 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Security-inspired modernization can bring about overdue improvements to ocean ship- Figure 5.2 Higher trade costs reduce ping. The USCS Automated Commercial Envi- global welfare ronment (ACE) project, which replaces paper Overall welfare losses, by region, from a one-percentage- documents with electronic methods of identify- point ad valorem increase in trade costs ing high-risk containers, is expected to save Eastern Europe U.S. importers $22.2 billion and the U.S. gov- ernment $4.4 billion in administrative costs Latin America over 20 years. Hong Kong recently launched North Africa electronic filing for cargo manifests for all and Middle East modes, which will enhance the efficiency and North America accuracy in submitting these documents. Pak- istan has introduced electronic filing of a single North Asia shipping document at Port Qasim as part of an effort by its customs service to streamline Oceania clearance and reduce transaction costs. Ac- cording to recent research, automated customs South Asia can lower the direct costs of customs clearance Sub-Saharan by the equivalent of 0.2 percent of the value of Africa traded goods. By accounting for the indirect Western Europe benefits of reduced delays, costs are reduced by 1 percent of merchandise value. (Hertel, 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Walmsley, and Ikatura 2001). Note: The measure of welfare loss—“equivalent variation” Implementation of these measures, which divided by GDP—was devised by Walkenhorst and Dihel (2002). “High-risk” regions will suffer greater welfare involve important changes throughout the sup- losses than lower risk regions from an identical increase in ply chain, may prove a difficult task for many frictional costs of trade. Similarly, some sectors are more sensitive than others to increases in frictional costs. developing countries. But if the costs of com- Source: Walkenhorst and Dihel (2002). plying with new security-inspired measures can be recovered later through greater efficiencies in the supply chain, the end result will be a global trading system that works better for fects of September 11 on international trade everyone—securing trade and smoothening indicates that even countries not directly in- trade flows simultaneously. volved in a terrorist event may expect their income to decline by $75 billion per year as Can the impact of security measures a result of a 1 percent ad valorem increase in be quantified? frictional costs to trade.19 While Western Eu- The recent introduction of the new security rope and North America suffer the greatest protocols and their even more recent imple- loss in absolute terms, other regions, such as mentation make it difficult to quantify their South Asia, North Africa, and the Middle East, impact on trade. Leonard (2001) estimated the are the main losers when income losses are re- new security-related costs at 1–3 percent of the lated to the size of the economies (figure 5.2).20 value of traded goods, while analysis by the A one-percentage-point increase in trade costs OECD (2002a, 2002b) suggests a more modest would cost South Asia $6 billion, more than impact.18 one-half of one percentage point when ex- Security-driven frictional costs of transport, pressed as a percentage of GDP. Regions with handling, insurance, and customs can affect high trade-to-GDP ratios and sectors with elas- trade even in the medium and long run. tic import demand incur the greatest trade and Walkenhorst and Dihel’s 2002 study of the ef- income losses in relative terms. 186 R E D U C I N G T R A D I N G C O S T S I N A N E W E R A O F S E C U R I T Y The threat of terrorism is not the only eral international, private, and professional source of frictional costs. War and epidemic organizations, has focused on facilitating disease, too, call for extraordinary measures trade—with security one of its themes. that often disrupt trade. The 2003 war in Iraq Development institutions, in partnership imposed significant costs on manufacturers, with national governments, have a role to play shippers, wholesalers, and retailers stemming in risk assessment, training, development of from supply-chain disruptions, blockages of human capital, and improving customs ad- vital sea routes, and delays in shipments. This ministration and infrastructure in their client was especially true for manufactures linking countries. They also can help track interna- factories in Asia with markets in North Amer- tional initiatives and assess implications for ica and Europe. New protocols at seaports developing countries. and airports have been implemented this year Because containers travel by sea, road, and to prevent the spread of severe acute respira- rail, their regulation is especially problematic. tory syndrome, or SARS, a virus that has The container may be subject to IMO regula- swept large parts of Asia. The outbreak has tions when on ship, but on land national gov- sharply reduced passenger travel to, from, and ernments may impose a different set of legis- within Asia, especially Hong Kong and China. lations. The interdependence and linkages Tourism has fallen sharply.21 among different transport modes call for a co- ordinated security approach among sectors A coordinated action plan on trade and and modes. A ship may be owned by a com- security is clearly needed pany in one country, crewed by national of a Even though the costs of compliance could be second country, and carry the cargo of a third large and disproportionate for smaller coun- to a port of a fourth. Regional and bilateral tries, all participants in the global trading sys- partner-ships among countries and stakehold- tem have an incentive to invest in counterter- ers can strengthen information exchange, co- rorism efforts. As noted above, the initial costs operation in training, and sharing of best prac- of new security procedures will pay off in the tices, resulting in mutual enhancement of long run through efficiency gains, better man- security efforts.22 The United Nations Interna- agement of information, and greater use of tional Drug Control Program (UNDCP) con- electronic commerce. It is easy to square en- tainer security project and the United Nations hanced security with improved trade facilita- Economic Commission for Europe (UNECE) tion at a theoretical level, however; it may well supply chain model are particularly promising be more difficult in practice. in this regard. The importance of partnerships at various The APEC STAR initiative has stressed levels in the global security campaign is clear. greater cooperation between governments and The IMO, ICAO, and other organizations private business to protect the global econ- should step up their technical cooperation ac- omy.23 Sustained dialogue between governments tivities to help developing countries improve and industry is needed to implement measures their capacity to bolster security and trade. to protect supply chains against security threats. Assistance should be coordinated so as to The private sector needs to be directly involved ensure absorption and nonduplication of with governments in crafting the most efficient capacity-building initiatives provided by the ways of complying with the requirements and developed world. The WCO has conducted a to ensure the integrity of trade from the point of survey of members’ capacity-building needs to manufacture to the port of delivery. In October ensure that security measures do not impede 2001, for example, a joint venture plan between development. Initiated by the World Bank in Boeing and Israel’s El Al airline was aimed at 1999, the Global Facilitation Partnership for integrating airlines’ security concerns into the Transportation and Trade, which includes sev- early stages of the aircraft production process, 187 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 with a view to providing a higher level of secu- proved viable for almost all transport modes rity at a lower cost (OECD 2002a). and generally have brought greater efficiency A risk-assessment template should be de- and lower prices for consumers. However, pub- veloped to ensure that high-risk areas are tar- lic and private barriers remain pervasive in air geted for special security programs. The mea- and maritime transport—restricting competi- sures adopted should be those that distort tion and increasing costs. In general, they trade the least and provide the greatest bene- should be replaced with systems that rely on fits, especially for exports from developing private provision of services. nations. Since it is impossible to screen and inspect all containers, procedures to identify International air transport services are high-risk containers—by detecting irregulari- heavily protected ties in shipping patterns—could be deployed. Efficient air transportation is an important de- Emphasis on detecting corrupt practices such terminant of an economy’s export competitive- as bribery will be needed to prevent controls ness. This is especially true for high-value, non- from being evaded. bulky manufactures, perishable horticultural A formula for cost-sharing that is optimal and agricultural products, and time-sensitive for all also must be developed. The Hong intermediate inputs traded within international Kong Shippers Council (HKSC) and the production networks. Efficient air cargo ser- ASEAN Federation of Forwarders Associa- vices play a critical role in attracting invest- tions (AFFA) have urged USCS to subsidize ment—including foreign direct investment the cost of its new requirements and U.S. im- (FDI)—in these sectors, which can be an im- porters to share with Asian exporters the bur- portant source of employment and economic den of providing information. growth. Caution must be exercised to ensure that The share of world trade shipped by air has security barriers do not become trade barriers. grown continuously over the past decades— One possible solution may be a new intergov- for example, from 7 percent of U.S. imports in ernmental program with the mandate to plan 1965 to 30 percent in 1998. In terms of ton- coordinated and comprehensive trade-related miles shipped worldwide, air cargo has grown security programs. Such a program could en- by almost 10 percent annually from 1970 to sure the win-win outcome that is achievable in 1996, while ocean shipping grew only 2.6 per- security and trade, while recognizing the spe- cent per year over the same period (World cial needs of developing countries. The G-8, in Bank 2001). More than 20 percent of African cooperation with developing countries, is one exports enter the United States by air, and, for logical forum for development of a coordi- a quarter of all product groups, the share of nated “Action Plan for Security and Trade.” air-shipped exports exceeds 50 percent.24 For many developing countries, the cost of air transportation often far exceeds the costs The anticompetitive effects observed on developed-country routes. Amjadi of international transport and Yeats (1995) found, for example, that regulations air transport costs made up between 10 and C osts rise and fall with public policies and private practices. For a long time, many transport services came under the aegis of pub- 50 percent of the value of African exports to the United States, a much higher propor- tion than for U.S. imports from non-African lic monopolies, and state-owned enterprises ex- countries. erted a powerful force in the transport sectors High air freight rates on developing-country of many countries. Such public monopolies are routes are primarily due to two factors. First, becoming increasingly difficult to jusify. Private the cost of serving developing countries may be entry and competitive market structures have higher. Developing countries are farther from 188 R E D U C I N G T R A D I N G C O S T S I N A N E W E R A O F S E C U R I T Y the world’s economic centers, increasing the Liberalizing air services can help cost of operating aircraft. And overall trade reduce costs— volumes tend to be smaller on routes serving What are the implications of these findings for less-developed countries, preventing operators public policy? First, there remain significant from reaping economics of scale and scope. policy-induced barriers to competition in air Thin traffic densities also may adversely affect cargo services. The complex system of air ser- the quality of air transportation, as services vice agreements (ASAs) still governs the mar- may be offered less frequently. Second, differ- ket for international air cargo services. ASAs ent degrees of competition in the provision of are typically negotiated bilaterally, although air services may affect the markup that air recent years have seen the emergence of re- cargo operators may be able to charge on a gional arrangements. Among other things, particular route. The extent of competition they designate the airlines allowed to operate among cargo carriers again depends on traffic on city-pair routes and the number and fre- volumes—as economies of scale and scope quency of flights they can operate. limit the number of providers that can be sus- Over time, ASAs have become increasingly tained on a particular route. Competition also liberal. For example, the so-called Bermuda- may be influenced by government policies—in type agreements do not regulate capacity on particular, restrictive market-access agreements each route but allow the designated airlines to for the provision of air transport. negotiate the number and frequency of flights. In an econometric investigation conducted “Open skies” agreements are even less restric- for this report, we attempted to quantify the tive, allowing all airlines to fly on all routes determinants of air transport costs using a between two countries without any ex ante sample of 139 randomly selected city-pair controls on capacity. routes in the Western Hemisphere. Preliminary More liberal ASAs can be a way of promot- results suggest that distance is a key determi- ing competition and thus lowering air cargo nant of international air cargo freight rates— freight rates. Moreover, greater freedom in de- most likely due to the cost of fuel and the signing air transport networks could allow air capital cost of operating aircraft. Across the service operators to reap greater economies of sample, a one-percentage-point increase in scale and scope, offering additional cost sav- city-pair distance leads to a 0.72 percent in- ings. Indeed, it is thought that liberalization in crease in prices—a higher distance elasticity protected air service markets may lead to con- than that typically found for maritime trans- solidation among airlines, as operators seek to port. Countries located far from economic cen- generate larger scale and network economies. ters are therefore at a disadvantage. Consolidation may not necessarily be associ- Moreover, the investigation confirms that ated with lessened competition, as fewer opera- there are sizeable economies of scale in the tors may compete on a larger number of routes. provision of air transport. On average, a 10 But it does suggest that liberalization needs to percent increase in city-pair traffic volumes be accompanied by competition policies that leads to a drop of slightly more than 1 percent ensure a review of mergers, acquisitions, and in the observed freight rate. In view of the other forms of private cooperation on eco- wide variance in freight traffic volumes, the nomic efficiency grounds (World Bank 2001). scale effect can be quite large—and in most cases it works against poorer nations. Finally, —but important challenges remain competition among airlines is found to exert Notwithstanding the benefits of air-service lib- downward pressure on freight rates. City-pair eralization, thin traffic densities and the asso- routes on which more than two passenger air- ciated lack of economies of scale are likely to lines or dedicated freight airlines operate remain a key obstacle to substantially lower- enjoy, on average, 10.7 percent lower prices. ing air cargo freight rates in the developing 189 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 world. Moreover, liberalization may lead air- by other criteria). Cargo sharing with trad- lines to foster the adoption of hub-and-spoke ing partners can be done unilaterally, or on networks, which may lower prices on well- the basis of bilateral and multilateral agree- connected hub routes but could actually raise ments. Although more and more countries freight rates on thin spoke routes. have phased out such requirements, countries Overcoming these challenges may call for ranging from Benin to India still have in place broader policy reforms. Countries have long reservation policies that at least nominally re- recognized the need for universal service poli- strict the scope of trade. cies in a variety of service sectors to ensure Cooperative agreements among maritime that remote and poor regions are offered ser- carriers on technical or commercial matters vices at affordable prices. These policies in- are another type of practice that restrains clude special service obligations imposed on competition. For example, liner conference operators, universal service funds, and various agreements set uniform freight tariff rates and forms of subsidies. The universal service con- conditions of service, often employing exclu- cept could be extended by international action sive contracts and other loyalty-inducing in- to remote and poor countries within conti- struments to prevent the entry of outside ship- nents. The necessary action could come in the ping lines. Private cooperation can improve form of tax breaks offered by developed coun- network coordination, generate economies of tries on air cargo service provided to certain scope, and provide a wider range of services to developing country locations or through the consumers of shipping lines. But a recent establishment of an international fund for the study of the impact of price-fixing and coop- provision of universal air services. erative working agreements on liner freight rates for U.S. imports, found that liberaliza- Regulations in maritime transport also tion of certain port services would lead to an restrict competition average price reduction of 8 percent and cost Various trade barriers have been imposed on savings of up to $850 million (Fink and others international maritime transport that protect 2002a). Private practices continue to have a inefficient service providers and hamper effec- strong impact on liner freight rates; breaking tive competition. Public policy restrictions in- up carrier agreements could cause prices to de- clude cargo reservation schemes that require cline further by 20 percent, with additional part of the cargo carried in trade with other cost savings of $2 billion (table 5.1). states to be transported only by ships carrying Seaport services have recently witnessed a a national flag (or other ships deemed national trend toward increased private-sector partici- Table 5.1 Elimination of anticompetitive private practices can cut costs drastically Cumulative Breakup of effect of the cooperative Breakup of breakup of Liberalization working price-fixing private carrier Cumulative of port services agreements agreements agreements total effect Average percentage price reduction 8.3 5.3 15.7 20.0 26.4 Projected total savings for all U.S. imports (in millions of dollars) 850.4 544.1 1618.4 2063.0 2712.5 Note: The average percentage price reductions are computed from the sample of 59 countries included in the study, while the pro- jected total savings apply to all U.S trading partners. Given the functional form of the underlying regression equation, the individual effects do not sum to the cumulative effects. See Fink and others (2001) for additional explanatory notes. Source: World Bank (2001). 190 R E D U C I N G T R A D I N G C O S T S I N A N E W E R A O F S E C U R I T Y pation and greater competition within and transaction gain from fast, easy, and low-cost among ports. Different ownership and opera- trading conditions.25 tion structures have emerged with respect to port management, provision of infrastructure, High logistic costs affect competitiveness and the supply of services. With the emergence Although the incidence of logistics costs varies, of large global port operators, there is now a with some developing countries more efficient greater risk of abuse of market power—abuse than others in providing trade services, a study that could reduce the benefits gained from by the United Nations Commission on Trade port liberalization. Creating regulatory capac- and Development (UNCTAD) estimates that ity and strengthening institutions is necessary average customs transactions in developing for the success of port reforms. countries involve 20 to 30 parties; 40 docu- ments; and 200 data elements, 30 of which had to be repeated at least 30 times. Subramanian Trade facilitation and Arnold (2001) broke down the cost of T o be effective, trade-facilitation measures must include or be accompanied by im- provements in domestic regulatory procedures international shipment into five categories: ocean freight, inland transport cost, and three indicators of logistics costs—custom inspec- and institutional structures. Above all, they tion, cargo handling and transfer, and process- must include development assistance to raise ing of trade documentation. Their analysis trade-related capacity—individual, institu- shows that logistics accounted for no less than tional, and social—in developing nations. One a third of the cost of door-to-door shipment of example of the need for capacity building and containerized carpets from Nepal to Germany technical assistance in the trade-facilitation and teabags from India to the United King- agenda relates to the technology required for dom. Using a similar methodology, a World expediting cargo clearance. International com- Bank (1997) study of the performance of merce depends increasingly on information Brazilian ports reported that per-container technology—most basically for the electronic costs for administrative procedures and cus- transmission of trade information. Modern toms clearance could be reduced by more than customs methods of profiling consignments or 20 percent (from $1,727 to $1,320) if interna- traders based on risk-assessment techniques tional best practices were followed. can help expedite cargo clearance. Compiling For small economies, higher logistics costs a unique set of computerized information for translate directly into higher import and export each shipment enables data to be processed prices. To remain competitive in industries before cargo arrives, thus expediting clearance where profit margins are thin, exporters must and speeding delivery. either pay lower wages to workers, accept Measures to speed the flow of goods and lower returns on capital, or enhance productiv- services across international borders have ity. The pressure on factory prices and produc- played a critical role in the expansion of global tivity is even higher for countries exporting trade over the past decades. Cross-border trade products that have a high import content, such in raw materials, components, and intermedi- as domestic export-oriented firms producing ate goods by multinationals with integrated garments or electronic final goods using im- production and distribution facilities now con- ported materials—as is the case with many de- stitutes more than one-third of world trade. veloping countries. In these cases, where small This expansion would not have been possible differences in transaction costs can determine without precisely timed maritime container whether the export venture is commercially services, express door-to-door delivery of air viable or not, logistical efficiency can mean freight, new information and communications greater retention of the value-added benefits of technologies, and other services. All parties to a trade-led growth. Particularly in labor-intensive 191 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Box 5.2 The logistics needs of a German car part manufacturer in Tunisia A s the subsidiary of a large German car part sup- plier, Leoni Tunisie S.A. produces cable and elec- tronic components for Daimler Chrysler and other their return journey. Eight trucks carrying approxi- mately 350 tons of finished parts leave Tunisia each week. The company considers the chain to be effi- European car manufacturers. The just-in-time supply cient and reliable. chains in the car industry put high demands in the Even so, the just-in-time demands of the indus- logistics system. Leoni has outsourced all logistics try are posing a threat to Tunisia as a production needs to an international forwarder that has a legal base—more and more clients require six-day cycles. subsidiary in Tunisia. Because internal production processes have been A full production and logistics cycle lasts about streamlined, further time savings depend on logistics nine days. Raw materials and intermediate products efficiency. Leoni Tunisie recently lost an internal are sourced from across Europe, Asia, and the company competition for a new factory with a po- United States. They are consolidated at Leoni’s head- tential for 1,700 jobs to Leoni’s Romanian sub- quarters in Germany and shipped to about a dozen sidiary. The reasons cited for the loss were not wage factories in various countries. Several trucks leave competitiveness or the investment environment—the Germany for Tunisia each week. The trailers are company regards Tunisia as very competitive—but cleared and sealed by German customs on the firm’s Eastern Europe’s logistics advantage. The land jour- premises, where they are picked up by the logistics ney between Romania and Germany takes one day provider. The forwarder drives the trailers to Genoa less in each direction. According to the CEO of or Marseilles (2–2.5 days), places them without a Leoni Tunisie, Tunisia will need to economize on lo- driver on RoRo ferries (20–24 hours by sea), claims gistics costs (including better air cargo connections them at Rades’s port, and delivers them to a factory or high-speed ferries to Europe) if it is to retain its in Sousse (2–3 hours by land). Once assembled, the competitive advantage in time-sensitive industries. finished components are cleared by a Tunisian cus- toms officer on the premises before they are sent on Source: Mueller-Jentsch (2002). industries in developing countries, high trans- enable firms to outsource stages of produc- port costs may preclude wage growth, thus af- tion to geographically dispersed locations. Re- fecting the standard of living of workers. search by Hummels (2001) showed that deliv- A study that examined the effect of higher ery times had a pronounced effect on imports shipping and port charges in the garment in- of intermediate products, suggesting that rapid dustry of Bangladesh estimated that exports delivery of goods is crucial for the maintenance could rise by 30 percent, raising hard-currency of multinational vertical product chains. Hum- earnings by 125 percent, if port inefficiencies mels found that each day saved in shipping were reduced. One recent estimate, based on time due to a faster transport mode and faster comparisons between air and ocean freight customs clearance was worth almost one per- rates for U.S imports, puts the per-day cost for cent, ad valorem, for manufactured goods. Un- shipping delays at 0.8 percent of the value of certain order-to-delivery times impose implicit trade for manufactured goods—with only a production costs as well. If logistics services small fraction attributable to the capital costs are unreliable and infrequent, firms are likely of the goods while on the ship (box 5.2). to maintain higher inventory holdings at every Minimizing transit time is particularly im- stage of the production chain, requiring addi- portant in modern commerce, given the trend tional working capital. Forgone earnings can toward just-in-time production systems that be significant for firms in countries with high 192 R E D U C I N G T R A D I N G C O S T S I N A N E W E R A O F S E C U R I T Y real interest rates. Gausch and Kogan (2001) fields. Freund and Weinhold (2000) found that found that inventory holdings in the manufac- a 10-percentage-point increase in the relative turing sector in many developing countries are number of web hosts in one country would two to five times higher than in the United have increased trade flows by 1 percent in 1998 States. Their estimates further show that de- and 1999. Fink, Mattoo, and Neagu (2002b) veloping countries could reduce the unit cost found that a 10 percent decrease in the bilat- of production by as much as 20 percent by re- eral calling price was associated with an 8 per- ducing inventory holdings by half. cent increase in bilateral trade. Moenius (2000) estimated the effect of bi- Better measures of trade facilitation are laterally shared and country-specific standards yielding some positive results on goods trade, finding that shared standards With increased attention to the benefits of re- generally promoted trade. Hertel, Walmsley, ducing nontransport barriers to trade, efforts and Itakura (2001) quantified the impact on have been made to assess the importance of trade of greater harmonization of e-business trade facilitation. Because empirical measures of standards and of automating customs proce- trade facilitation are lacking, however, progress dures between Japan and Singapore, conclud- has been limited. ing that such reforms would increase trade Several recent studies quantify the benefits flows between these countries as well as with of improved trade facilitation, modeled as a the rest of the world. In agricultural trade, reduction in the costs of international trade or Wilson and Otsuki (2003) find that the major as an improvement in the productivity of the exporters of nuts and cereals would gain international transportation sector. UNCTAD $38.8 billion if divergent national food-safety (2001) considers trade facilitation in the standards relating to aflatoxins were replaced broader context of creating an environment by the Codex international standard. conducive to developing e-commerce usage and applications. The results show that a re- Those results show clearly that when trade duction of one percentage point in the cost of is facilitated, trade volumes rise maritime and air transport services could in- In their study of APEC manufacturing trade, crease Asian GDP by $3.3 billion.26 If trade Wilson, Mann, and Otsuki (2003a) incorpo- facilitation is expanded to include improve- rate indicators linked to multiple categories of ments in wholesale and retail trade services, trade facilitation into a single model, thus al- an additional $3.6 billion could be gained by lowing a synthetic analysis that prioritizes a one-percentage-point improvement in the areas for reform. The authors estimate the re- productivity of that sector. APEC (1999) lationship among four indicators and trade found that the shock-derived reduction in flows using a gravity model.28 Wilson, Mann, trade costs ranged from 1 percent of import and Otsuki (2003b) expand the scope of the prices for industrial countries and the Repub- analysis to include 75 countries worldwide, lic of Korea, Taiwan, and Singapore, to 2 per- including 52 developing countries. The indica- cent for developing countries.27 The study tors in the analysis are: estimated that APEC merchandise exports would increase by 3.3 percent from trade- • Port efficiency (through measurements facilitation efforts. of port infrastructure) Empirical studies of the impact of enhanced • Customs environment (including nontar- e-commerce and telecommunication access, iff fees) improved customs procedures, and harmo- • Regulatory environment (including trans- nized or improved standards also demonstrate parency of government policy and con- the benefits of trade facilitation in specific trol on corruption) 193 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 • Use of e-commerce by businesses, a would come from improved customs regimes, proxy for the service-sector infrastruc- and about $107 billion (2.8 percent) from ture necessary to implement e-business. more efficient ports. But the largest gain— $154 billion, or 4.0 percent—would come The study estimates the potential increase in from enhancing infrastructure in the services trade following improvements in each trade- sector (figure 5.3). Reforms and improve- facilitation area. The authors examine a sce- ments affecting exports would have a greater nario in which trade-facilitation capacity in effect on trade growth than would changes af- below-average countries is raised halfway to the fecting imports, suggesting that the export- average for the entire set of countries. This ap- promotion effect of trade facilitation should proach recognizes that in some countries trade not be underestimated in designing capacity- is already being facilitated at levels approaching building efforts. the global best, leaving little room for improve- The impact of individual trade-facilitation ment, whereas in others a standardized im- measures differs significantly from region to provement of, say, 10 percent in trade facilita- region, but improvements in service-sector in- tion would be quite difficult, requiring frastructure would provide the largest gain in additional capacity-building assistance. Figure sum of imports and exports in all regions— 5.3 illustrates the projected gain, by trade-facil- particularly in South Asia (figure 5.4). The po- itation indicator, for the 75 countries examined. tential gains from improvements in port effi- Preliminary findings suggest that better ciency also are great—again, particularly in trade facilitation would increase trade among South Asia. the 75 countries by approximately $377 bil- The gains to be expected from domestic re- lion dollars—an increase of about 9.7 percent. form alone (figure 5.5) show similar patterns About $33 billion (0.8 percent) of the gain to those projected for global reforms, imply- ing that priority areas for domestic reform are consistent with reforms to raise capacity glob- ally. In assigning priorities for capacity build- Figure 5.3 Facilitating trade in less- ing, enhancing port efficiency and improving efficient countries would bring service-sector infrastructure appear to be most significant gains Trade gains from raising handling capacity in 75 important for domestic and global action. below-average countries halfway to the global average (percent change and dollar gain) Reforming domestic policies 4.5 160 is indispensable 4 140 The benefits from investment in modern cus- 3.5 toms, port facilities, and new technology, how- 120 ever, can only be realized if an appropriate reg- 3 Export 100 ulatory framework is in place. In this regard, promotion 2.5 ᮡ domestic reforms have an important role to 80 2 play, by making better use of existing resources 60 and improving the efficiency of services. Liber- 1.5 Import promotion alization of transport and telecommunication 40 1 networks can help encourage domestic compe- 0.5 ᮡ 20 tition and produce substantial cost reductions. Liberalization also offers the added advan- 0 0 Ports Customs Regulations Services tage of widening the availability and choice Source: Calculations based on table 4 in Wilson, Mann, of services in many developing countries. As and Otsuki (2003b). discussed earlier, the prevalence of anticom- petitive practices by transport service providers 194 R E D U C I N G T R A D I N G C O S T S I N A N E W E R A O F S E C U R I T Y Figure 5.4 The impact of individual trade-facilitation measures differs significantly from region to region Trade gains (exports plus imports) under the “halfway to average” scenario; gains from domestic and partners’ reforms (percent change) 20 18 16 14 Customs 12 10 Regulations 8 Ports Services 6 ᮡ ᮡ ᮡ 4 ᮡ 2 0 OECD East Asia Europe and Latin America Middle East South Asia Sub-Saharan Central Asia and the and Africa Caribbean North Africa Source: Wilson, Mann, and Otsuki (2003b). calls for the development of efficiency-oriented hensive fashion.29 Exploratory work on the competition policies. Replacement of ineffi- trade-facilitation agenda centered on ways to cient public monopolies in international trans- simplify trade procedures, to harmonize them port systems with private operators can thus to conform to a rule-based multilateral frame- increase competition. work, and to integrate the work of other in- Cross-country comparisons provided in ternational organizations involved in trade- Wilson, Mann, and Otsuki (2003b) reveal sig- facilitation into the WTO framework. But nificant difference in countries’ potential for progress on trade facilitation at WTO’s next gains from trade facilitation. Guatemala, for two ministerial conferences was limited.30 example, has great potential to reap trade gains by reforming its service-sector infrastructure. The Doha Declaration raised the issue In contrast, the potential importance of regu- of negotiations on trade facilitation latory reform predominates in Indonesia. In The Doha ministerial raised the possibility of Nigeria, reform of the customs system would launching multilateral negotiations on trade have the most productive outcome. In all coun- facilitation at the Cancun ministerial meeting tries, domestic reform would have greater im- in September 2003.31 The “Doha agenda” on pact on total trade than would reforms by trad- trade facilitation referenced simplifying trade ing partners. procedures, enhancing technical assistance and capacity building, and recognizing limitations in capacity associated with a country’s level of Trade facilitation and the development. The Doha ministerial declara- WTO agenda tion was explicit in recognizing the need to in- T he Singapore Ministerial Declaration in 1996 empowered the WTO for the first time to look at trade facilitation in a compre- crease trade-related capacity in developing countries and address the issue of implementa- tion costs associated with capacity building, 195 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Figure 5.5 Domestic reforms alone would produce many of the same gains as global reform Trade gains (exports plus imports) under the “halfway to average” scenario; gains from domestic reforms only (percent change) 16 14 12 10 Regulations 8 Customs 6 Ports Services ᮡ ᮡ 4 ᮡ ᮡ 2 0 OECD East Asia Europe and Latin America Middle East South Asia Sub-Saharan Central Asia and the and Africa Caribbean North Africa Source: Wilson, Mann, and Otsuki (2003b). particularly in developing and least-developed vention have not been ratified by all WCO countries. member countries. To the proponents, a bind- Some proposals under discussion envisage ing framework would help establish effective either building new or strengthening existing compliance with the recommendations of the rules and principles of transparency, simplifi- revised Kyoto convention. cation, efficiency, proportionality, nondiscrim- The challenges to negotiations on trade fa- ination, and due process within the country for cilitation center on two concerns. First, it is independent litigations involving trade dis- not clear that binding multilateral rules on putes, as well as recourse to the WTO Dispute customs and border-crossing procedures are Settlement Undertaking, if need be, to ensure actually needed; implementing institutional nondiscriminatory treatment of traders in mem- changes requires country ownership and vol- ber countries. untary actions. Second, it is not clear that any Not everyone agrees that trade facilitation new rules could be enforced through con- measures should be on the agenda at Cancun. ventional dispute-settlement proceedings and Proposals advanced by the proponents of a penalties, since violations of those rules often rules-based approach to trade facilitation focus stem from the limited capacity of governments on the role and importance of the WTO. Are to meet their obligations. Rules alone are not national governments free, they ask, to follow likely to produce the desired reforms or mod- the recommendations of the WCO’s revised ernizations. Those depend on capacity building, Kyoto convention on customs procedures? Are and capacity building depends on resources— they likely to do so? Or should these recom- financial and other. mendations be a reference point for WTO Simplifying administrative and procedural rules?32 The WTO has many more members requirements in customs and border-crossing than WCO, the proponents point out, and the procedures—unlike negotiations on tariff cuts, revised recommendations of the Kyoto con- for example—depends directly on improve- 196 R E D U C I N G T R A D I N G C O S T S I N A N E W E R A O F S E C U R I T Y ments in physical infrastructure, institutional building must have a broad scope. The proce- reform, and other complex development ob- dural and administrative burdens on traders jectives that typically involve large commit- are often aggravated by overlapping and du- ments of resources for training and capital plicative informational requirements from sev- equipment. In an environment of constrained eral ministries, departments, or agencies. New resources, expenditures on such improvements ways of minimizing such requirements include must be made in accordance with the priori- the single-window concept: official controls are ties of the national development strategies. administered by a single agency. This and other innovations often require coordination of sev- The trade-facilitation agenda requires eral government agencies, as well as changes a broad multilateral effort in domestic regulatory procedures and institu- Trade facilitation necessarily involves both in- tional structures. frastructure investments and revisions in legal Multilateral agreement on standards of frameworks, technology training, and other transparency and acceptable fees as well as measures. A little capacity building could go a new ways to publish and disseminate applica- long way. Expediting customs clearance proce- ble trade laws, rules, fees, and schedules—per- dures by providing transparent and clear guide- haps through a new information clearinghouse lines, for example, reduces the discretionary or inquiry center for WTO members—could power of customs officials, thereby reducing help countries facilitate trade. Discussions also the scope for corruption by establishing a right could include development of harmonized and of independent judicial appeal (box 5.3). Assis- quantifiable measures of “timely release of tance to modernize customs procedures by up- goods.” Consideration of wider use of the “sup- grading information technology and applying plier’s declaration of conformity” with techni- risk-based criteria in reviews of documentation cal regulations for low-risk goods—along with and cargo, allowing for self-assessment and a parallel program to expand information tech- audit-based (as opposed to transactions-based) nology systems and databases—could decrease release of goods can also have high payoffs and simplify documentation requirements. (box 5.4). Strengthening the provisions of GATT Article Because behind-the-border policies affect V (Freedom of Transit) could be particularly cross-border trade, trade-related capacity beneficial to land-locked countries. Box 5.3 Tackling corruption in customs: Peru E liminating customs corruption in Peru required changes in laws, regulations, and human- resources policy. retained personnel were increased by nearly 10 times the previous salaries. To increase the competence of customs officers, After firing corrupt employees, the Peruvian cus- hiring for professional positions was limited to uni- toms service established a uniform code of conduct versity graduates. The service established a training and contracted with a university to develop tests of academy, offering up to one year of training to new employee competence. Continued employment was and incumbent employees. In addition to these ef- contingent on passing the tests. Although substantial forts, customs embarked on a program to bring new pressure was exerted on the customs service to rehire skills and knowledge to the organization through discharged employees, it was able to resist the pres- external recruitment of mid-career professionals— economists, auditors, statisticians, and information sure with support from government allies. As a result technology experts. of these efforts, employee corruption decreased, while competence levels improved. Salaries of Source: Wilson et. al (2002). 197 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Box 5.4 Customs reform in Lebanon L ebanon’s ongoing customs reform project is part of a fiscal reform project under the sponsorship of the United Nations Development Programme and program monitoring the days required for clearance. The computerization process was accompanied by staff training and a restructuring of work procedures. the World Bank. An inefficient customs service has Clearance operations were set up on the shop floors long been a major logistical bottleneck in Lebanon. of some of the main importers and exporters, with Procedures were nontransparent and time-consum- inspectors using risk-assessment criteria to conduct ing. Most of the 7–12 days needed for container de- selective inspections. To inform users of their rights livery were due to customs delays. Such deficiencies and responsibilities and to further streamline the in- not only imposed unnecessary economic costs but spection process, the customs service published a also bred corruption. summary of its border regulations. Under the reform project, customs clearance was Preliminary results of the ongoing reform indi- reduced from 13 to 5 basic steps—entry and accep- cate that although the percentage of consignments tance of declaration, inspection of goods for verifica- cleared without inspection had quadrupled between tion of declared information, assessment of informa- 1997 and 1999 (from 10 to 40 percent) and the aver- tion, automatic calculation of taxes, and payment of age days needed for clearance had declined from six taxes. Clearance procedures were then aligned with to four as a result of more selective testing, the aver- international procedures, with UN and EU standards age rate of tariff collection remained constant. translated into Arabic for the first time. A one-page administrative document replaced 26 complex and outdated forms. These reforms enabled the clearance process to be computerized, with a new software Source: Mueller-Jentsch (2002). Any negotiation on trade-facilitation mea- costs of goods. High costs also result from reg- sures, however, must carefully consider how the ulations and practices that impede effective WTO’s dispute-settlement provisions could be competition in international transport systems. tailored to ensure that the capacity constraints of developing countries are taken into account Open skies and universal service can when enforcing commitments undertaken by lower international transport costs for governments. Important capacity-building work developing countries done by several institutions outside the WTO Improving competition in air transport will re- framework provides a model for a future mul- quire revisions in air service agreements to re- tilateral agenda to raise trade-facilitation ca- duce barriers to entry into markets. Moving pacity in developing countries. It is difficult to progressively toward “open skies” agreements see how this level of institutional change and would allow all airlines to fly on all routes be- assistance coordination could be “enforced” tween two countries without any ex ante con- through dispute settlement mechanisms. trols on capacity. This would lower air cargo freight rates and could allow air service opera- tors to obtain greater economies of scale and Lowering transport costs, scope, offering additional cost savings. It may increasing security, and well be that liberalization in protected air ser- facilitating trade vice markets may lead to consolidation among T he costs of moving goods across interna- tional borders are a crucial determinant of a country’s export competitiveness. Every day airlines, as operators seek to generate larger scale and network economies. (Consolidation may not necessarily reduce competition, how- spent in customs adds almost 1 percent to the ever, as fewer operators may compete on a 198 R E D U C I N G T R A D I N G C O S T S I N A N E W E R A O F S E C U R I T Y larger number of routes.) Even with open entry, dependence and linkages among different thin traffic densities and the associated lack of transport modes call for a coordinated ap- economies of scale are likely to remain key ob- proach to security among sectors and modes. stacles to lowering air freight rates in the de- Regional and bilateral partnerships among veloping world. If liberalization leads airlines countries can strengthen channels for informa- to adopt hub-and-spoke networks, prices could tion exchange and cooperation in training and fall on well-connected hub routes, while rising sharing of best practices, resulting in mutual on some spoke routes. To reduce this risk by enhancement of security efforts. Other regions cross-subsidizing transport to remote and poor could follow APEC’s lead by looking for ways areas within continents, the concept of univer- to design collaborative programs with the pri- sal service should be embraced internationally. vate sector to implement security measures. Rich countries could offer tax breaks on air Third, a risk-assessment template would cargo service provided to certain developing ensure that high-risk areas are targeted for country locations. Alternatively, an interna- special security programs. The measures tional fund for the provision of universal air adopted should be those that distort trade the services could be established. least and provide the greatest benefits, espe- For maritime transport, one avenue to im- cially for exports from developing nations. provement would be to subject the industry to Fourth, a formula for cost-sharing must be MFN treatment in routes as part of the larger developed. The Hong Kong Shippers Coun- GATS discussion on services. Doing so would cil (HKSC) and the ASEAN Federation of undermine the competition-restricting liner Forwarders Associations (AFFA) have urged codes that prevent new entries in designated the USCS to subsidize the cost of its new shipping routes. Another avenue would be to requirements and U.S. importers to share review exemptions in U.S. and EU antitrust with Asian exporters the burden of providing law for maritime transport. information. Security can be increased without Trade facilitation depends on capacity jeopardizing trade flows from building and development assistance developing countries Capacity building and development assistance Even though the costs of compliance with new are necessary if countries are to make the most security measures could be large and dispro- of trade-facilitation measures—whether those portionate for smaller countries, all partici- measures stem from security imperatives or pants in the global trading system have an in- multilateral trade talks. Attempts to build trade centive to invest in counterterrorism. Such capacity may require several elements—from investments are likely to pay off in the long building basic transport infrastructure to mak- run through efficiency gains, better manage- ing legislative changes and training regulators. ment of information, and greater use of elec- Some developing countries may require only tronic commerce. To ensure that they do, sev- technical assistance to expedite cargo clearance eral steps must be taken. through electronic trade documentation. Oth- First, technical assistance must be increased. ers will need much more help. No single pack- The IMO, ICAO, and other organizations age will meet the needs of all countries. should step up their technical cooperation ef- Whether or not trade facilitation becomes forts to provide more training in risk assess- part of multilateral trade negotiations, mea- ment, customs administration, and infrastruc- sures that lower transport costs, remove barri- ture planning in their client countries. ers to goods and services moving across bor- Second, nations must coordinate trade-re- ders, and build capacity in trade facilitation lated actions not only with other countries, but must be pursued. Success will depend first on also with their own private sectors. The inter- governments and the private sector in devel- 199 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 oping countries, but also on the G-8, UN far as security and its costs are concerned, criticized agencies, the WCO, the World Bank, and other U.S. government aid of $10 billion to its airlines to international development institutions. Multi- conform to increased security measures. European car- riers believe that the aid has helped U.S. carriers slash lateral efforts to support domestic policy re- rates on very competitive North Atlantic routes. form and institutional improvements in devel- 7. Another proposal under consideration is the fil- oping countries are particularly important if ing of a bill of lading by U.S. Agricultural exporters 24 investments in trade facilitation are to yield hours before loading the containerized freight. their full potential—a potential that is great 8. The Agricultural Ocean Transport Coalition has indeed. urged Customs to require no more than 12 hours ad- vance notice for agricultural products and 6 hours for perishable products. 9. U.S. VISIT, a new entry-exit system to be in- Notes stalled in U.S. airports and seaports by January 1, 2004, 1. These sections draw on Wilson and others will be based on visas that include biometric features (2002), among other sources. such as fingerprints and photographs to identify foreign 2. A study of the effect of security on private in- visitors. The EU has also earmarked Euro 140 million vestment and growth by Poirson (1998) spanning 53 to fund biometric identification technology for visas. developing countries from 1984–95 indicates that en- hanced security fosters private investment and growth 10. A U.S.-EU dilemma arose over reservation in developing economies. Private investment in the records demanded by the United States that violated short run increased by 0.5 to 1 percentage point of EU’s data privacy rules. An interim agreement was GDP, in relatively insecure countries that adopted se- reached, after the United States assured the European curity measures to the levels in “best practice” regions. airlines of “appropriate handling” of the records, Moreover, economic growth received a boost by 0.5 to which include not only names but also the passenger’s 1.25 percentage points per year in the long term. itinerary, contact phone number, and other details, 3. The newly created Department of Homeland Se- such as credit card numbers. curity includes Customs, Immigration and Naturaliza- 11. The United States has initiated “smart border” tion Services (INS), Border Patrol, and the federal Agri- programs with Canada and Mexico, that use modern cultural Inspection Service. The Department provided technology to enhance security and expedite movement $170 million in port security grants in June 2003. across borders. Under discussion is a plan that would include an addi- 12. Canada levied a C$24 (US$15) Air Traveller’s tional $1 billion for the Transportation Security Ad- Security Charge on all round-trip tickets in April 2002, ministration, $200 million to $700 million more for to finance the increased airport security measures. The the Coast Guard, and an increase in federal grants to tax—the highest security tax in the world—contributed local police and fire departments for counterterrorism to a 10.2 percent decline in passenger traffic across training. Canada since the beginning of 2002, and resulted in a 4. Some overseas suppliers are covered under the steep fall of 50 percent on some short routes. C-TPAT because they are subsidiaries of U.S. compa- 13. Recognizing the lack of resources to buy new nies enrolled in the initiative. technology, the United States intends to provide fi- 5. The Swedish port of Goteborg has become the nancing to developing countries with transportation twelfth to join the Container Security Initiative (as of security projects. Two security experts from the United May 2003). Those already participating include: Rot- States have arrived in Indonesia to assist in upgrading terdam, LeHavre, Bremerhaven, Hamburg, and Ant- cargo security and assess the implementation of secu- werp in Europe; Singapore, Hong Kong, and Yoko- rity measures at the country’s seaports and airports. hama in Asia; and Vancouver, Montreal, and Halifax The United States announced a joint initiative with in Canada. These ports are at different stages of imple- Thailand to transform Laem Chabang port into a safe mentation of the CSI framework. CSI is now moving transportation port. into its second phase, which will include Turkey, Dubai, 14. Given that a ship carries thousands of contain- and about 20 other nations in Asia, Latin America, ers at any time, inspection of the cargo could cause de- Europe, and Africa. lays. While the scanning process is quite fast, the prob- 6. On a related note, Europe’s largest air cargo car- lem lies with the turnaround time of the containers riers, which are calling for a level playing field among targeted for scanning. It would take time to transport the United States, Europe, and the rest of the world as the container to and from the scanning area, and con- 200 R E D U C I N G T R A D I N G C O S T S I N A N E W E R A O F S E C U R I T Y tainers that are late for loading would tie up hauling However, assumptions are made regarding such in- equipment and reduce stowage efficiency. creases as varying across regions and sectors according 15. In other developments: to exposure to terrorism risks following the Septem- • The Japanese Ministry of Land, Infrastructure, ber 11 attacks. For example, high-risk regions (North and Transport (MLIT) is set to introduce anti- America, Middle East, North Africa) are assumed to ex- terrorist legislation that will prevent foreign perience increases in frictional costs that are two and a ships from entering Japanese ports unless they half times as high as cost increases in low risk regions. have a security crew on board and can provide The figure shows only the uniform increase in frictional identification. costs to trade. • Hong Kong’s customs authorities have created a 21. Since a large part of the airfreight is transported terrorist response system, acquiring mobile x-ray in the bellies of passenger planes, a cutback in passen- machines and a radiation detector to scan cargo ger flights has an impact on cargo. and beefing up its intelligence capabilities with 22. Australia and New Zealand are strengthening more staff and equipment. their Pacific regions border control relationship by co- • The ICAO has adopted resolutions designed to operating and exchanging information regarding assure the safety of passengers, ground crew per- smuggling, air and sea cargo security approaches, sonnel, and the public. Its Regulated Agent Re- SARS, and general border protection issues. gime requires parties in the flight chain to imple- 23. In its “Cargo Security White Paper,” the Na- ment measures to strengthen air-cargo security. tional Customs Brokers and Forwarders Association of • The Australian government’s Aviation Transport America (NCBFAA) has outlined ways for the trading Security Bill aims to provide screening of all bag- industry to assess risks, build information links to help gage checked on international flights. A $100 government officials, and use technology to improve million federal plan to protect the nation’s mar- cargo security. It recommends building a “chain of cus- itime gateways also has been enacted. tody dataset” to verify people connected to a shipment • The New Zealand government will be allocating and assess cargo security throughout the supply chain. $5.9 million next year and $1.9 million in future 24. See Amjadi and Yeats (1995). years to the Ministry of Foreign Affairs and 25. This part draws extensively from the WTO Trade, for security. (1999). 16. A recent online survey by BDP International in- 26. See UNCTAD 2001, table 8, page 33. dicated by a three to two margin that exporters believed 27. APEC (1999). the implementation of the 24-hour rule would enhance 28. See Global Competitiveness Report 2001–2002, security. About 23 percent of those surveyed said that World Competitiveness Yearbook 2001–2002, and the impact was extreme, 30 percent reported moderate Kaufmann, Kraay, and Zoido-Lobaton (2002), for the to significant costs of compliance, half did not know list of countries in the dataset. how to recover costs, and 42 percent plan to absorb ex- 29. The ICC, a nongovernmental organization that penses. With respect to implementation of the advance has long advocated trade facilitation, promoted the manifest filing rule, USCS has issued less than 400 “No- subject on the WTO agenda at the Singapore minister- Load” directives for violations of cargo description re- ial meeting. quirements in its first three months of enforcement. 30. The Ministerial conference in Geneva (1998) 17. Tea money refers to the use of illegal or unfair concentrated on the perceived threat to the global means, such as bribery to gain an advantage in busi- economy due to the ensuing Asian financial crisis. Al- ness. Ports and airports all over the world are places though there were several proposals in favor of and where tea money comes in handy to expedite deliveries against launching trade negotiations in the period prior and shipments. to the Singapore ministerial meeting in 1999, trade fa- 18. Estimates by Leonard were made soon after the cilitation was overshadowed by other events at the events and could reflect the major disruptions faced Seattle ministerial (Woo 2002). during the period. 31. The Doha declaration states: “Recognizing the 19. This figure is comparable to the estimates of case for further expediting the movement, release, and $30–58 billion losses for the insurance industry by the clearance of goods, including goods in transit, and the OECD (2002b). need for enhanced technical assistance and capacity 20. The authors employ four alternative scenarios building in this area, we agree that negotiations will to quantify the trade and welfare impacts, in which all take place after the fifth session of the ministerial on frictional costs are increased by 1 percent ad valorem. the basis of a decision to be taken, by explicit consen- 201 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 sus, at the session on the modalities of the negotiations. Mueller-Jentsch, Daniel. 2002. Transport Policies for In the period until the fifth session, the Council for the Euro-Mediterranean Free-Trade Area. An Trade in Goods shall review and, as appropriate, clar- Agenda for Multi-modal Transport Reform in the ify and improve relevant aspects of Articles V, VIII, and Southern Mediterranean. World Bank Technical X of the GATT 1994 and identify the trade-facilitation Paper 527. Washington, D.C. needs and priorities of members, in particular develop- Moenius, Johannes. 2000. “Three Essays on Trade ing and least-developed economies. We commit our- Barriers and Trade Volumes.” Ph.D. dissertation. selves to ensuring adequate technical assistance and University of California, San Diego. support capacity building in this area.” WTO (2001). OECD (Organisation for Economic Development and 32. See WTO (2002) and Messerlin and Zarrouk Co-operation). 2002a. “The Impact of the Ter- (2000). rorist Attacks of 11 September 2001 on Interna- tional Trading and Transport Activities.” Unclas- sified Document TD/TC/WP (2002)9/Final. Paris. References ———. 2002b. “Economic Consequences of Terror- Amjadi and Yeats. 1995. “Have Transportation Costs ism.” OECD Economic Outlook 71: 117–40. Contributed to the Relative Decline of Sub-Saharan Paris. African Exports? Some Preliminary Evidence.” Poirson. 1998. “Economic Security, Private Investment World Bank Working Paper 1559. and Growth in Developing Countries.” IMF APEC (Asia Pacific Economic Cooperation). 1999. “As- Working Paper WP/98/04. sessing APEC Trade Liberalization and Facilita- Reddy, R. 2002. “Friction over Security Gaps,” Intelli- tion.” Update, Economic Committee. September. gent Enterprise. October 8 (Available at http:// Fink, Carsten, Aaditya Mattoo, and Ileana Cristina www.intelligententerprise.com/021008/516 Neagu. 2002a. “Trade in International Maritime infosc1-1.shtml) Services: How Much Does Policy Matter?” World Subramanian U., and J. Arnold. 2001. Forging Subre- Bank Economic Review 16(1): 81–108. gional Links in Transportation and Logistics in ———. 2002b. “Assessing the Impact of Communica- South Asia. World Bank, Washington, D.C. tion Cost on International Trade.” World Bank UNCTAD (United Nations Commission on Trade and Working Paper 2929. Development). 2001. E-Commerce and Develop- Freund and Weinhold. 2000. “On the Effect of the In- ment Report. ternet on International Trade.” Board of Gover- Walkenhorst, Peter, and Nora Dihel. 2002. “Trade Im- nors of the Federal Reserve System International pacts of the Terrorist Attacks of 11 September Discussion Paper 693. 2001: A Quantitative Assessment.” Paper prepared Gausch and Kogan. 2001. “Inventories in Developing for the Workshop on The Economic Consequences Countries: Levels and Determinants, a Red Flag of Global Terrorism. Berlin, June 12–13. on Competitiveness and Growth.” World Bank Wilson, John S., and Tsunehiro Otsuki. 2003. “Food Working Paper 2552. Safety and Trade: Winners and Losers in a Non- Hertel, T., T. Walmsley, and K. Ikatura. 2001. “Dy- harmonized World.” Journal of Economic Inte- namic Effects of the ‘New Age’ Free Trade Agree- gration 18 (2): 266–87. ment between Japan and Singapore.” Journal of Wilson, John S., Catherine L. Mann, and Tsunehiro Economic Integration 24: 1019–49. Otsuki. 2003a. “Trade Facilitation and Economic Hummels, D. 2001. “Time as a Trade Barrier.” Unpub- Development: Measuring the Impact.” Working lished. Department of Economics, Purdue Univer- Paper 2933. World Bank, Washington, D.C. sity, West Lafayette, Ind. ———. 2003b. “Trade Facilitation and Capacity Kaufmann, Kraay, and Zoido-Lobaton. 2002. “Gover- Building: Global Perspective.” Unpublished. World nance Matters II: Updated Indicators for 2000–01.” Bank, Washington, D.C. World Bank Working Paper 2772. Wilson, John S., Catherine L. Mann, Yuen Pau Woo, Leonard, J. 2001. “Impact of the September 11, 2001, Nizar Assanie, and Inbom Choi. 2002. Trade Fa- Terrorist Attacks on North American Trade cilitation: A Development Perspective in the Asia- Flows.” E – Alert, Manufacturers Alliance, Ar- Pacific Region. Singapore: APEC Secretariat. lington, Virginia. Woo, Yuen Pau. 2002. “Trade Facilitation in the Messerlin, Patrick A., and Jamel Zarrouk. 2000. WTO: Singapore to Doha and Beyond.” In Will “Trade Facilitation: Technical Regulations and Martin and Mari Pangestu, eds., Options for the Customs Procedures.” The World Economy 23 Next Trade Round: View from East Asia. Cam- (4): 577–593. bridge, U.K.: Cambridge University Press. 202 R E D U C I N G T R A D I N G C O S T S I N A N E W E R A O F S E C U R I T Y World Bank. 1997. “Multilateral Freight Transport: ———. 2001. “Doha WTO Ministerial Declaration.” Selected Regulatory Issues.” Report 16361-BR. WT/MIN(01)/DEC/1. Geneva. November. ———. 2001. Global Economic Prospects 2002: Mak- ———. 2002. “Compendium of WCO Capacity Build- ing Trade Work for the World’s Poor. Washing- ing Tools.” G/C/W/445. Geneva. ton, D.C. WTO (World Trade Organization). 1999. “Develop- ment Aspects of Trade Facilitation.” WT/COMTD/ W/57. Geneva. 203 6 Development and the Doha Agenda Historically, many WTO rules evolved to re- T he challenges confronting developing countries seeking to expand their inter- flect the perceived interests of developed coun- national trade are primarily domestic. tries in an era when the participation of devel- Countries that have expanded their share of oping countries was limited. Many rules reflect global markets have generally shared certain the status quo practices that have already been conditions: a progressively more open domes- adopted in industrial countries. The wider lati- tic trade regime; a supportive investment cli- tude accorded agricultural subsidization re- mate; and complementary policies relating to flects the use of such support policies in many developed countries. The same is true for the education, health, and infrastructure. Most of permissive approach historically taken toward this agenda is national and requires domestic the use of import quotas on textile products— policies to deal with prevailing constraints to in principle prohibited by General Agreement increasing trade. The World Trade Organiza- on Tariffs and Trade (GATT) rules. New disci- tion (WTO) negotiating agenda is necessarily plines adopted in the WTO often mirror regu- limited to a narrow subset of issues that over- latory practices of rich countries. For example, laps only partially with priority development the recent inclusion of rules on the protection concerns for most countries (Finger 2001). of intellectual property rights has led to the per- In this sense, the WTO is not a comprehen- ception that the WTO contract demands regu- sive development institution. It is a negotiat- latory changes in developing countries without ing forum in which governments make trade any corresponding changes in regulatory poli- policy commitments that improve access to cies in industrial countries.1 each others’ markets and establish rules gov- As developing countries have become more erning trade. Developing countries can gain actively involved in the WTO, the challenge is from both functions: first, because trade open- to design rules that promote development. ness, growth, and poverty reduction are mutu- Meeting that challenge means evaluating the ally reinforcing; and, second, because a rules- implications of various ways to achieve this based world trading system protects small objective. Rarely is this a straightforward players that have little ability to influence the process, especially when it comes to the policies of large countries. Rules can reduce “behind-the-border” regulatory policies that uncertainty by placing mutually agreed limits are increasingly the subject of multilateral dis- on the policies that governments may adopt— cussions. Negotiating pro-development rules thus potentially helping to increase domestic in such a context requires the active engage- investment and lower risks. ment of developing countries. 205 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 The developing countries’ traditional ap- Nondiscriminatory trade liberalization proach has been to seek differential treatment. for poor countries—and poor people— “Special and differential treatment” (SDT) pro- is critical visions in the WTO span three core areas: pref- Recent initiatives by developed countries to ex- erential access to developed-country markets, tend duty- and quota-free market access for the typically without reciprocal commitments from least developed countries (LDCs) could, if fully developing countries; exemptions or deferrals implemented, make preferences more effective. from some WTO rules; and technical assistance But because offering deep, unilateral prefer- to help implement WTO mandates. What con- ences to larger countries is not politically feasi- stitutes a developing country is not defined in ble, preferences can do little for the majority of the WTO—a country’s status is a matter of self- the poor in non-LDCs. Providing opportunity declaration. All in all, the current system has for all of the world’s poor, therefore, requires not worked especially well, and many countries multilateral, nondiscriminatory liberalization are seeking a new approach. of trade, so that all developing countries can develop their comparative advantage. Most Trade preferences have been disappointing of the gains from trade liberalization result in delivering market access: the dilemma from a country’s own reforms. As reciprocity Developed countries grant preferences volun- in the exchange of liberalization commitments tarily rather than as part of a binding multilat- is the engine of the WTO process, both low- eral negotiation. Those preferences often come and middle-income countries should harness laden with restrictions, product exclusions, and reciprocity to gain market access. administrative rules. Preference programs often Elements of a development-supportive trade cover only a share of exports from developing regime would include a binding commitment countries, and among those eligible countries by developed countries to abolish export sub- and products, only a fraction of preferences are sidies, decouple agricultural support, and sig- actually utilized. Products and countries with nificantly reduce—or eliminate—tariffs on export potential often do not receive prefer- products of export interest to developing coun- ences, whereas eligible countries and product tries. Negotiations should target tariff peaks, categories often lack export capacity. specific tariffs, and tariff quotas, while aiming Another problem is that preferences, even for a significant overall reduction in the aver- when effective, are likely to divert trade away age level of applied tariffs. The pursuit of these from other excluded developing countries be- objectives would be more supportive of devel- cause the exports of developing countries tend opment than one that continues to emphasize to overlap more with each other than with nonreciprocal preferential access to markets. those of developed countries. Finally, preferences do little to help the Negotiating WTO rules that support majority of the world’s poor. Most of those development is a major challenge living on less than $1 per day live in countries Trade-policy disciplines that can be imple- like China, India, Pakistan, and Association mented through a stroke of pen, such as tariff of Southeast Asian Nations (ASEAN) member reductions, are fundamentally different from countries, which receive limited preferences regulatory disciplines and administrative rules in products in which they have a comparative that require institutional changes. In contrast advantage. Meanwhile, many middle-income to tariff reforms, administrative rules may countries justify relatively high barriers to trade require substantial resources to establish or on SDT grounds, to the detriment of poorer de- strengthen implementing institutions. Domes- veloping countries whose access is impeded. tic rules and regulations must be customized to 206 D E V E L O P M E N T A N D T H E D O H A A G E N D A local circumstances. Thus, rules relating to reg- ing countries are granted access to developed- ulatory practices are unlikely to be a develop- country markets at tariffs lower than the most- ment priority for every country, nor are the favored-nation (MFN) rates through policies benefits to global partners likely to be propor- such as the Generalized System of Preferences tional in all countries. The experience after the (GSP). Second, they may be temporarily ex- Uruguay Round with implementation of agree- empted from certain disciplines or granted ments by developing countries has demon- greater discretion to apply restrictive trade strated that limiting recognition of differential policies. Third, they may request technical as- capacities and levels of development to uni- sistance from high-income countries to imple- form transition periods is inadequate, as are ment trade rules and related reforms. nonbinding offers of technical assistance. Al- The intellectual foundation of SDT was laid lowing for greater differentiation among devel- in the 1960s by Raoul Prebisch and Hans oping countries in determining the reach of Singer, who argued that developing-country WTO rules is important. exports were concentrated mainly in com- modities with volatile and declining terms of Aid for trade must be complemented by trade. They called for import-substitution poli- action in developing countries cies, supported by protection of infant indus- Development assistance must play an impor- tries at home, and preferential access to export tant role in helping to expand and improve the markets. Although the rationale for these poli- trade capacity needed for countries to benefit cies remains controversial (see, for example, from better access to markets. Low-income Bhagwati 1988), in 1968 the Generalized Sys- countries confront many challenges in identi- tem of Preferences (GSP) was launched under fying and addressing trade-related policy and United Nations Conference on Trade and De- public investment priorities. Those priorities velopment (UNCTAD) auspices. This called should be made explicit in the form of a na- on developed countries to provide preferen- tional development strategy. That strategy, not tial access to developing-country exports on a a WTO agenda, should drive technical and voluntary basis.2 Because GSP programs vio- financial assistance. Diagnostic trade-integra- late the GATT’s MFN rule, GATT contracting tion studies completed for several LDCs reveal parties waived the MFN requirement in 1971 that action is required in areas lying far be- for 10 years, thereby placing GSP within the yond the scope of WTO agreements. Trade fa- GATT framework. In 1979, at the conclusion cilitation and logistics are especially impor- of the Tokyo Round, permanent legal cover tant. Additional development assistance could for the GSP was obtained through the so- help low-income countries address such prior- called Enabling Clause,3 which called for pref- ities. Such assistance should also help coun- erential market access for developing countries tries adjust and adapt to a gradual reduction and limited reciprocity in GATT negotiating in trade preferences and address the effects of rounds to levels “consistent with development possible increases in world food prices. needs.” It also confirmed that developing countries should have greater freedom to use restrictive trade policies. An important feature Special and differential treatment of the Enabling Clause was that SDT was to be and the WTO phased out when countries reached a certain T he idea that developing countries should receive SDT has a long history in the GATT/WTO system. It has three related di- level of development. That level was never de- fined, however, leaving eligibility for trade preferences to the discretion of preference- mensions. First, for certain products, develop- granting countries. 207 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 The existence of the GSP and limited reci- language in various WTO agreements. The procity in GATT negotiations affected the proposals included calls for improved preferen- patterns of MFN trade liberalization in both tial access to industrialized countries, further the Kennedy (1964–67) and Tokyo Rounds exemptions from specific WTO rules, and (1973–79) (see chapters 2 and 3). The end re- binding commitments on developed countries sult was larger tariff reductions in goods pri- to provide technical assistance to help imple- marily of export interest to industrialized ment multilateral rules. Despite intensive talks economies.4 Average levels of trade protection during 2002, no agreement on these proposals in developing countries were reduced rela- emerged. One reason was that many of the tively little. Lack of engagement by developing proposals sought to convert nonbinding, “best countries also facilitated the emergence of re- endeavors” language into obligations binding strictive quota regimes for textiles under the on developed countries. Another was disagree- Multi-fiber Arrangement (MFA) and the effec- ment over what types of provisions would pro- tive removal of GATT disciplines on agricul- mote development. The latter issue is funda- ture-related trade policies (Hudec 1987). mental, of course, but it was never the focus of Under the pre-WTO trade regime, new rules explicit analysis and discussion in the relevant extending the original GATT treaty were ap- WTO committee (Keck and Low 2003). plied on a voluntary basis. Extensions were called “codes,” whose disciplines bound only the contracting parties that signed them (Hoek- Market access for development man and Kostecki 2001). This approach to rule extension was removed with the creation of the WTO. In contrast to the GATT, all WTO I nternational trade helps raise and sustain growth—a fundamental requirement for re- ducing poverty—by giving firms and house- agreements and disciplines, with the exception holds access to world markets for goods, of rules on government procurement and trade services, and knowledge; lowering prices and in civil aircraft, apply to all members regardless increasing the quality and variety of consump- of level of development—although in many tion goods; and fostering specialization of eco- cases transition periods apply to developing nomic activity in areas where countries have a countries. A consequence of this so-called Sin- comparative advantage. Through the diffusion gle Undertaking and the expansion in the cov- and absorption of technology, trade fosters the erage of multilateral rules to new areas, such investment and positive externalities that are as intellectual property and trade in services, associated with learning. Policies that shelter was that developing-country governments were economic agents from the world market im- confronted with a significant implementation pede these benefits and dynamic gains. While agenda as well as new policy constraints.5 adjustment costs and measures to safeguard In the runup to WTO’s 1999 Seattle minis- the interests of poor households should not be terial meeting, SDT and implementation con- neglected in the design of policies, openness to cerns figured prominently. The 2001 Doha trade is associated with higher incomes (Irwin Ministerial Declaration emphasized the impor- 2002). Moving toward an open trade policy tance of SDT, stating that “provisions for spe- and identifying the needed complementary cial and differential treatment are an integral domestic policies should consequently figure part of the WTO agreements.” Paragraph 44 centrally in the design of national poverty- called for a review of SDT provisions with a reduction strategies. Many developing coun- view to “strengthening them and making them tries have pursued unilateral liberalization of more precise, effective, and operational.” On their trade regimes in the last two decades. the basis of this mandate, developing countries They have also concentrated on obtaining pref- made over 85 suggestions to strengthen SDT erential access to rich-country markets. 208 D E V E L O P M E N T A N D T H E D O H A A G E N D A Preferences result in limited market country’s GSP eligibility for a given product access and are uncertain may be removed if annual exports of that prod- Trade preferences granted by developed coun- uct reach $100 million7 or if there is significant tries are voluntary. They are not WTO obli- damage to domestic industry. In the European gations. Donor countries determine eligibility Union, products classified as “sensitive” only criteria, product coverage, the size of prefer- benefit from a 3.5-percentage-point reduction ence margins, and the duration of the prefer- of the MFN tariff rate, except for clothing, for ence. Developed-country governments rarely which the reduction is 20 percent.8 Most chem- have granted deep preferences in sectors where icals, almost all agricultural and food products, developing countries had the largest export po- and all textiles, apparel, and leather goods are tential. Indeed, preferences tend to be the most classified as “sensitive.”9 The European Union limited for products protected by tariff peaks also excludes from GSP eligibility certain prod- (Hoekman, Ng, and Olarreaga 2002). ucts from large countries—regardless of their Developing countries often obtain only lim- per capita income. Examples include Brazil, ited preferences in sectors where they have a China, India, and Indonesia. Finally, the Euro- comparative advantage (table 6.1).6 In some pean Union has a safeguard clause allowing cases, developing countries face higher average preferences to be suspended if imports “cause tariffs because of the composition of their ex- or threaten to cause serious difficulties to a ports. Some subcategories include tariff peaks Community producer.” that further restrict access for developing coun- In numerous instances, products or coun- tries. The primary reason for this pattern of tries have been removed from GSP eligibility, ei- protection is that in some sectors there is strong ther as the result of specific criteria having been domestic opposition to liberalization in devel- satisfied (see above) or because of lobbying by oped countries. However, it is also partly a con- domestic interest groups in importing coun- sequence of the limited engagement by devel- tries. The resulting uncertainty can only have a oping countries in reciprocal negotiations. negative impact on incentives to invest in ex- Benefits are often limited by design. Market port sectors. Binding multilateral liberalization share or value thresholds limit the extent to commitments under the WTO are more secure. which recipients can export on preferential The uncertainty of unilateral preferences also terms. In the United States, for example, a arises from conditions that may be attached Table 6.1 Developing countries rarely receive significant preferences in sectors in which they would have a comparative advantage Import revenues, market shares, and tariff rates for key products without GSP preferences in the European Union and United States in 2001 (percent, except where otherwise noted) Average tariff Total imports GSP recipients’ LDC Average rate faced by (billions of dollars) market share market share tariff rate GSP recipients EU U.S. EU U.S. EU U.S. EU U.S. EU U.S. Dairy products 1.4 1.2 15 11 1 1 9.9 13.4 15.9 19.7 Textiles and yarn 15.3 9.6 42 21 3 1 5.4 7.8 4.6 7.2 Apparel and clothing 48.7 60.8 54 47 8 7 10.2 15.3 8.8 15.9 Leather products 6.1 7.6 74 24 1 1 2.3 10.4 1.9 11.5 Footwear 6.5 16.1 67 18 1 1 7.5 10.6 7.4 10.0 Ceramics and glassware 6.2 8.7 27 13 1 1 5.1 6.3 3.8 8.2 Note: GSP countries only; LDCs may obtain deeper preferential treatment. China is included under EU GSP but excluded by the United States. Source: World Integrated Trade Solution. 209 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Table 6.2 Utilization rates for preference-eligible products with high MFN tariffs are low Preference use by GSP recipients in the U.S. market, 2001 (percent, except where otherwise noted) Imports from Share under Average applied Total imports GSP recipients Share imported all preference tariffs on Category (billions of dollars) (billions of dollars) under GSP programs all imports Cut flowers 0.72 0.43 3 95 6 Prepared fish 0.72 0.47 7 13 9 Cane or beet sugar 0.56 0.41 29 77 6 Fruit, nuts 0.78 0.34 20 33 15 Unmanufactured tobacco 0.75 0.58 3 7 68 Acrylic alcohols 1.56 0.73 55 94 5 Ethers, ether-alcohols 1.78 0.34 84 84 5 Carboxylic acids 1.64 0.73 4 4 5 Trunks, suitcases 4.59 1.17 0 6 10 Articles of leather 2.68 0.52 18 21 8 Plywood and panels 1.10 0.46 18 22 5 Footwear 13.87 2.39 0 1 11 Apparel, knitted 25.00 11.50 0 25 14 Apparel, not knitted 35.10 15.90 1 15 14 Hats, headgear 0.96 0.33 0 2 7 Articles of jewelry 5.40 2.10 54 70 6 Note: Table reports data on imports (at the 4-digit HS classification) of products on which the United States applied tariffs that exceeded 4 percent in 2001, and where GSP recipient countries had significant exports to the United States. Source: U.S. International Trade Commission. to eligibility. Such conditions often relate to produced with yarn spun in either the prefer- worker protection, human rights, intellectual ence-granting or the beneficiary country. Al- property, and the environment.10 though rules of origin are necessary for pref- Recent initiatives by the European Union, erences to work and are beneficial in ensuring the United States, and several other industrial- that value is added and employment created in ized countries to provide either full or much in- the recipient country, it is important to ensure creased duty- and quota-free access to their that rules of origin are not intentionally or in- markets for exports from LDCs clearly im- advertently protectionist. proves the situation. However, excessively re- Rules of origin and associated paperwork strictive rules of origin remain an important and administrative requirements are likely to impediment to full use of these deeper prefer- be a major reason that many eligible products ences, which, moreover, do not extend to many do not enter developed-country markets under poor countries with substantial trade capacity. preference provisions—instead exporters pay the applicable MFN tariff. Except for certain The use of preferences is limited alcohols, sugar, flowers, and jewelry, less than All preferential programs, whether unilateral one-third of eligible exports from beneficiary or reciprocal (under free-trade agreements), countries entered the United States under any impose significant administrative costs related preference program in 2001 (table 6.2). An in- to enforcement of rules of origin.11 These rules dicator of the restrictiveness of these rules is are imposed to prevent transshipment—that that only 65 percent of eligible apparel exports is, reexport of products produced in non- from the Caribbean and Central America enter eligible countries. Rule-of-origin requirements the United States under all preference pro- and related inspection procedures can be quite grams, despite a preference margin of more costly. They also may be explicitly protection- than 14 percent. ist in intent. An example is the so-called triple Limited use of preference programs is also transformation rule in textiles, which requires observed in other countries. Sapir (1997) imported clothing to be made from textiles showed that in 1994, only one-half of Euro- 210 D E V E L O P M E N T A N D T H E D O H A A G E N D A Table 6.3 Actual use of preference programs is declining Quad country imports from GSP beneficiaries (billions of dollars) and ratio of use of available preferences (percent), 1994–2001 Rate of use Eligible for Receiving of preferences Year Total imports Dutiable imports preference preference (percent) 1994 448 283 162 83 51.1 1995 539 331 195 108 55.1 1996 585 351 178 100 56.0 1997 575 346 200 100 50.1 1998 543 311 183 74 40.6 1999 548 290 166 68 40.7 2000 623 308 171 72 42.0 2001 588 296 184 71 38.9 Source: Inama (2003). pean imports that could potentially benefit preferential access to the EU market for sugar from GSP entered under this preferential and has been granted a relatively large quota regime, reflecting the combined effect of rules (Mitchell 2003). Such benefits are obtained at of origin and tariff quotas. In 2001, imports high cost to EU taxpayers and consumers, and by the Quad (Canada, the European Union, to other excluded developing countries. Japan, and the United States) from GSP bene- There also is evidence that GSP programs ficiaries totaled $588 billion, of which $296 are associated with success stories in countries billion were subject to duties and $184 billion with the capacity to benefit from access oppor- were covered under various preferential pro- tunities. Ozden and Reinhardt (2003b) com- grams (table 6.3). Only $71 billion of the eli- pare the export performances of U.S. GSP gible exports actually received preferential beneficiaries with those of countries removed treatment (approximately 39 percent of eligi- from eligibility (those said to have “graduated”). ble exports). The share for LDCs, however, is Their results suggest that countries removed higher at approximately 60 percent (Inama 2003), reflecting less restrictive treatment. Figure 6.1 The benefits of U.S. trade Who benefits from preferences? preferences are distributed unequally Top 10 beneficiaries of U.S. generalized system of A relatively small number of mostly middle- preferences, 2001 (percentage of total GSP benefits) income countries are the main beneficiaries of preference programs. These countries have the Chile 4% capacity to exploit the opportunities offered India All other 11% by meeting the administrative requirements. In 26% 2001, 10 of the 130 eligible countries ac- Indonesia counted for 77 percent of U.S. non-oil imports 13% under GSP provisions (figure 6.1). The same countries accounted for only 49 percent of all Russia 4% Philippines imports from GSP-eligible countries. How- 7% Venezuela ever, some small countries have benefited sig- 6% nificantly from preferential access to markets Turkey South Africa 4% 5% where high tariffs, subsidies, or other policies Thailand are used to drive the domestic price of the 20% product to levels well above the world market Source: USITC Dataweb. price. An example is Mauritius, which has 211 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 among these programs in product coverage, Figure 6.2 Countries “graduating” from eligibility criteria, and administrative rules (es- U.S. generalized system of preferences pecially rules of origin) have important impli- have better export performance than cations, not only for the countries who benefit those still in program from them, but also for those excluded. Export performance of countries dropped from U.S. GSP program in 1976–2000 vs. those remaining in program The United States, for example, has imple- (percent) mented the African Growth and Opportunity 50 Act, the Caribbean Basin Initiative, and the 45 Andean Trade Promotion Act, as well as sev- 40 eral reciprocal free-trade agreements (with Israel, Jordan, and Mexico). Major EU pro- 35 Dropped from GSP grams include the Cotonou convention cover- 30 ᮡ In GSP ing the African, Caribbean, and Pacific (ACP) 25 ᮡ countries, and the Everything-But-Arms initia- 20 tive, which covers LDCs. The European Union 15 also has concluded a large number of preferen- 10 tial trade agreements with neighboring coun- tries in Europe, North Africa, and the Middle 5 East (Schiff and Winters 2003). 0 These unilateral and reciprocal programs Exports/GDP Industrial Growth rate exports/GDP of exports differ in several important respects from the Source: Ozden and Reinhardt (2003b). GSP. First, they include sectors excluded by standard GSP programs—for example, apparel and food products. Thus, by 2009 Everything But Arms will cover all exports of beneficiary from the GSP outperform those remaining eli- countries (the 49 LDCs) without exception— gible for GSP treatment (figure 6.2). Countries all duties and quotas will have been removed. that are not on GSP tend to have higher ratios Similarly, the Caribbean, Andean, and African of exports to GDP, as well as higher export programs of the United States include apparel, growth rates. This could be interpreted as evi- in contrast to its GSP program. Second, the ad- dence that for some countries—the successful ministrative requirements of these deeper pref- ones—GSP played a role in generating the ini- erential schemes tend to be more relaxed re- tial export expansion. While great care is re- garding rules of origin and competitive needs quired in attributing causality—clearly many tests (USTR 2002). other factors will be important in determining Notwithstanding these improvements, the export performance—one reason for the better overall impact of these programs has not yet performance of countries that were removed been very significant, with the exception of from GSP is probably their own trade policies. apparel exports to the United States from cer- Because import protection is equivalent to tain African countries (more on this below). taxation of exports, liberalization is a precon- The share of LDCs in total imports of the dition for substantially expanding exports. United States and the European Union has not increased significantly in recent years (figure Preferences have a hierarchy 6.3). In the case of Everything But Arms, this The foregoing discussion has focused primar- may reflect, in part, that the products that ily on GSP. In practice, unilateral preferences matter most to a number of LDCs—bananas, granted by the European Union and the rice, sugar—will be liberalized only in 2006 United States are implemented under many or 2009. Most of the products exported by different programs (box 6.1). The differences LDCs already were eligible for duty-free entry 212 D E V E L O P M E N T A N D T H E D O H A A G E N D A Box 6.1 EU and U.S. preference programs United States Duty-free treatment is granted on all products except Generalized System of Preferences (GSP): The U.S. textiles and apparel, certain footwear, petroleum and GSP program has existed since 1976. Criteria for related products, certain leather products, canned eligibility include not aiding international terrorists tuna, rum and sugar, syrup, and molasses. The main and complying with international environmental, differences with GSP are that the Andean scheme labor, and intellectual property laws. Unlike the Eu- covers more products, has more liberal qualifying ropean GSP (see below), the U.S. program grants rules, and is not subject to competitive need limits. complete duty- and quota-free access to eligible ATPA rules of origin permit inputs from CBERA products from eligible countries. China and several beneficiaries. ATPA was renewed in 2002 as the An- “graduated” countries are not eligible—among them dean Trade Promotion and Drug Eradication Act Hong Kong (China), Republic of Korea, Malaysia, (ATPDEA) and expanded to include tuna, leather Singapore, Taiwan (China), and Malaysia. and footwear products, petroleum products, and ap- Textiles and apparel, footwear, and many parel—subject, however, to restrictive rules of origin. agricultural products are not eligible for the GSP. For example, if apparel is assembled from U.S. fab- Certain products from certain countries can be ex- rics, no quotas or duties apply, but if local inputs are cluded if the total exports pass the “competitive used, duty-free imports are subject to a cap of 2 per- needs limit”—$100 million per tariff line or $13 mil- cent of total U.S. imports (increasing to 5 percent in lion if the exporting country has more than a 50 per- equal annual installments). cent share of U.S. imports. Total imports to the African Trade Preferences: The African Growth United States under GSP provision totaled $14.5 bil- and Opportunity Act (AGOA), passed in 2000, of- lion in 2001, or 1.5 percent of total U.S. non-oil im- fers beneficiary Sub-Saharan African countries duty- ports and 13 percent of all non-oil exports to the free and quota-free market access for essentially all United States from GSP recipients. In most eligible products. AGOA excludes textiles but extends to sectors where the MFN tariff rate is above 5 percent, duty- and quota-free treatment for apparel made in the share of exports entering under the program Africa from U.S. yarn and fabric. If regional fabric from eligible countries is only 30–40 percent, in part and yarn are used, there is a cap of 1.5 percent of as a result of rules-of-origin requirements. U.S. imports, increasing to 3.5 percent over eight Caribbean Trade Preferences: The Caribbean years. African LDCs are exempt from all rules of Basin Economic Recovery Act (CBERA), commonly origin for a limited period of time, helping to signifi- known as the Caribbean Basin Initiative (CBI), was cantly expand apparel exports from countries such as enacted in 1984 and modified in 1990. Twenty-four Lesotho. countries are eligible. Duty-free treatment is granted on all products other than textiles and apparel, cer- European Union tain footwear, handbags, luggage, petroleum and re- Generalized System of Preferences: Preferences under lated products, certain leather products, and canned GSP are available to all developing countries, in- tuna. In 1998, only 18 percent of exports from bene- cluding China. Overall, 36 percent of tariff lines are ficiary countries were in eligible product categories. eligible for reduced tariffs, and 32 percent are eligible The 2000 Caribbean Basin Trade Partnership Act for duty-free access. Twelve percent of tariff lines (CBTPA), provides NAFTA-equivalent treatment for (mostly agricultural) are excluded and subject to certain items (mainly apparel) excluded from duty- full MFN duty. Excluded products include meat, free treatment under the CBI program. dairy products, cereals, sugar, wine, and products for Andean Trade Preferences: The Andean Trade which the European Union sets minimum import Preferences Act (ATPA) extends preferences to Bo- prices. Approximately 36 percent of all products are livia, Colombia, Ecuador, and Peru. Enacted in 1991 classified as “sensitive”—often those with the highest as part of U.S. efforts to reduce narcotic production MFN tariffs (Panagariya 2002). Sensitive products and trafficking, it was modeled after the CBI and has similar eligibility requirements and product coverage. (Continues on next page) 213 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Box 6.1 (continued) are subject to a flat 3.5-percentage point-reduction ucts are duty and quota free. Preferences are less in the MFN tariff, implying that the higher the duty, comprehensive for agricultural products. In 2000 the smaller the proportionate impact of the prefer- duties were still applied to 856 tariff lines (837 of ence. Specific duties are reduced by 30 percent; if a which were agricultural products). Of these, 116 product is subject to both ad valorem and specific lines were excluded from the Cotonou Agreement, duties, the specific duty is not reduced. although specific protocols govern access for sugar Additional tariff reductions are available under and bananas on a country-specific basis. An addi- special incentive schemes for the protection of labor tional 301 tariff lines were eligible for reduced rights (an additional 5-percentage-point reduction), duties, subject to specific quantitative limits (tariff the environment (an additional 5 percentage points), quotas) set for the ACP countries as a group. The and for countries that combat drug production and remaining 439 products were eligible for reduced trafficking (duty-free access for certain products). duties without limits on exported quantities. Currently only one country, Moldova, has requested Everything But Arms: Introduced in March and satisfied the requirement relating to labor rights 2001, this program grants duty-free access to im- (but not the environment). A group of Latin Ameri- ports of all products from the LDCs, with the excep- can countries and Pakistan benefit from arrange- tion of arms and munitions, without any quantitative ments relating to drugs. restrictions. Liberalization was immediate except for Countries can be excluded from the GSP (based three major products: fresh bananas, rice, and sugar. on their development level) or from particular prod- Tariffs on these three items will be reduced gradually uct categories. Sectoral exclusions are determined by to zero (in 2006 for bananas; in 2009 for rice and specific criteria based on shares of EU imports from sugar), while tariff quotas for rice and sugar will be GSP-beneficiary countries and certain indicators increased annually. Access to the EU market is gov- of development and specialization.a For example, erned by the rules of its GSP scheme. A key feature Argentina is excluded from preferences for live ani- of the program is that in contrast to the GSP, prefer- mals and for edible products of animal origin, while ences for the LDCs are granted for an unlimited Thailand is excluded from preferences for fishery period and are not subject to periodic review. products. ACP countries (Cotonou Agreement): ACP countries are granted preferences that often exceed a. Some ad hoc exclusions are applied to China, the CIS coun- those available under the GSP. Most industrial prod- tries, and South Africa in the fisheries and iron and steel sectors. under GSP or Cotonou provisions. As a result, proximately $1.1 billion, compared to $750 Everything But Arms had no immediate im- million from Andean countries and $9.5 bil- pact (Brenton 2003). In the case of the United lion from Caribbean countries. These coun- States, export market shares of countries eligi- tries accounted for some 20 percent of the ble under the three primary deep preference $58 billion U.S. apparel import market. This programs have not increased (figure 6.4).12 growth is mainly a result of exemptions from The primary exception is apparel, which quotas and tariffs imposed on other exporters. shows remarkable export growth, especially in In the case of AGOA, rules of origin are re- the case of AGOA. Total exports of apparel moved temporarily for some countries for a since 1996 increased by more than 200 percent limited period, providing an extra advantage. for AGOA countries, and approximately 60 A crucial issue is how these regions will fare percent for Caribbean and Andean countries. when remaining quotas (mostly faced by coun- As a result, in 2002, apparel exports to the tries in South and East Asia) are phased out United States from AGOA countries were ap- at the end of 2004, as required by the WTO 214 D E V E L O P M E N T A N D T H E D O H A A G E N D A Figure 6.3 Preferences have not increased the share of the least developed countries in imports into the European Union and the United States Share of LDCs in total imports of European Union and United States, 1966–2002 (percent) 1.0 0.9 0.8 United States ᮡ 0.7 0.6 0.5 ᮡ 0.4 European Union 0.3 0.2 0.1 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Source: WITS. Agreement on Textiles and Clothing. This is therefore suggest that a reciprocal liberaliza- also the time when liberal rules of origin under tion strategy supported by complementary do- AGOA are set to expire. Competitive pressures mestic policies may have a much larger impact are likely to increase substantially, giving rise to than unilateral preferences. a need for adjustment and for investment pro- grams to improve productivity and diversifica- tion of the export base. Extending the liberal rules of origin under AGOA would help reduce Figure 6.4 Market shares of countries the impact of the abolition of the remaining eligible for three U.S. “deep preference” import quotas on textiles and clothing. programs have not increased The available evidence suggests that pref- Shares of AGOA, Andean, and CBI countries in U.S. imports (excluding oil and previous metals), 1996–2002 erences by industrialized countries have the (percent) greatest effect on developing-country exports 2.0 if they are granted on a reciprocal basis as part CBI 1.8 ᮡ of a deep regional free-trade agreement. Span- 1.6 ish exports to the European Union and Mexi- can exports to the United States rose dramati- 1.4 cally following accession to the European 1.2 Union and NAFTA, respectively (figure 6.6). 1.0 AGOA This supply response is not just the result of ᮡ 0.8 removing import barriers by northern part- 0.6 ners, but also of the “regime change” that oc- 0.4 curred in these countries and the consequent ᮡ 0.2 Andean change in risk premiums, uncertainty, and in- vestment incentives. A large part of the regime 0.0 1996 1997 1998 1999 2000 2001 2002 change involved changes in investment, regu- latory, and administrative policies, not just Source: US ITC. preferential trade liberalization. These data 215 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 and sell at low margins in competitive mar- Figure 6.5 Preferred countries’ apparel kets, even small preferences in certain cate- exports to the United States have risen gories could make a difference for some coun- Growth of apparel exports to United States, 1996–2002 (percent) tries, fueling those conflicts of interest. As an example, 36 countries in Sub-Saharan 250 Africa are eligible to export apparel products into the United States without any tariffs or 200 AGOA quantitative restrictions under AGOA. These ᮡ countries risk losing preference margins if 150 MFN protection is reduced in the United States. Many of these countries temporarily face no 100 rules of origin requirements. Twenty-four coun- CBI tries in the Caribbean and Central America ᮡ enjoy similar privileges under the CBPTA. 50 Through bilateral NAFTA preferences and uni- lateral Caribbean and African preferences, ben- ᮡ Andean 0 eficiary countries managed to increase their 1996 1997 1998 1999 2000 2001 2002 share of U.S. apparel imports to around 32 per- Source: US ITC. cent (figure 6.8). Such countries may be con- cerned about the erosion of their preferences if MFN protection is reduced. Preferences have unintended consequences Sugar is a product for which preference ero- Numerous side effects of unilateral preferences sion will have important consequences for sev- also must be considered when assessing the eral countries. Quota allocations in protected case for them. Preferences can lead to more markets, such as the European Union, are cur- protectionist trade policies in recipient coun- rently very concentrated in a few countries that tries. Ozden and Reinhardt (2003b) show that tend to have high costs relative to other pro- U.S. GSP recipients implement more protec- ducers. For example, Mauritius has 38 percent tionist trade policies than countries removed of EU quotas (Mitchell 2003). Given the ex- from the GSP program (figure 6.7). Although tension by 2009 of duty- and quota-free access their finding does not prove causality, prefer- to the EU market for all LDCs, Mauritius will ences can decrease the incentives for domestic confront much greater competition in the EU exporters to mobilize in favor of more liberal market. trade policies. Because domestic trade policies Most of the academic research on prefer- affect developing countries’ growth prospects ence programs has concluded not only that more than barriers in their export markets, they generally yield modest export increases the perverse-incentive effect of unilateral pref- (at best), but also that a significant portion erences may be quite damaging. Similarly, of these gains is because of trade diversion preferential market access may lower the in- from nonbeneficiaries. Multilateral liberaliza- centives for developing countries to participate tion would reduce some of the detrimental ef- actively in multilateral negotiations, in part be- fects of preferential access to highly distorted cause they believe that they will not receive markets. For example, moving to free trade in any further concessions in the multilateral sugar markets not only would result in esti- process or because of concerns about erosion mated global welfare gains (the sum of pro- of preferences. The latter may create conflicts ducer, consumer surpluses, and tax revenues) of interest between preferred and nonpreferred of $4.7 billion, but also would yield a 38 per- developing countries. To the extent that coun- cent increase in world sugar prices and boost tries specialize in similar product categories sugar trade by about 20 percent. Brazil alone 216 D E V E L O P M E N T A N D T H E D O H A A G E N D A Figure 6.6 Agricultural exports from Mexico and Spain rose dramatically after the two countries joined regional trade blocs Mexican exports of certain agricultural products, 1961–2001 Mexican exports of certain agricultural products, 1961–2001 (billions of dollars) (millions of dollars) 8 300 7 250 6 200 5 4 Agricultural 150 products, total 3 Meat ᮡ 100 Milk 2 ᮡ ᮡ 50 1 0 0 61 65 69 73 77 81 85 89 93 97 01 61 65 69 73 77 81 85 89 93 97 01 19 19 19 19 19 19 19 19 19 19 20 19 19 19 19 19 19 19 19 19 19 20 Spanish exports of certain agricultural products, 1961–2001 Spanish exports of certain agricultural products, 1961–2001 (billions of dollars) (millions of dollars) 1,400 15 Agricultural products, total 1,200 13 Meat ᮡ 1,000 ᮡ 10 800 8 600 Milk 5 400 ᮡ 3 200 0 0 65 69 73 77 81 85 89 93 97 01 61 65 69 73 77 81 85 89 93 97 01 61 19 19 19 19 19 19 19 19 19 20 19 19 19 19 19 19 19 19 19 19 20 19 would experience a real income gain of some efficient way to provide development assis- $1.6 billion. Although countries such as Mau- tance (Beghin and Aksoy 2003). ritius could lose significantly, coordinated How much preferences are likely to be global liberalization across all products would eroded as a result of further multilateral trade offset some of the lost rents. World sugar price liberalization will depend both on the benefits increases alone would offset about one-half of countries currently obtain from preference pro- the lost quota rents for countries that currently grams and the speed with which preferences are have preferential access. Moreover, the loss in eroded. The foregoing discussion suggests that preference rents would be much less than is the overall benefits of unilateral market access commonly expected, because many of the ben- preferences are limited by exclusions for sensi- eficiaries are high-cost producers. Indeed, the tive products, rules of origin, and limited sup- cost to the European Union and United States ply capacity. The fact that a substantial share of of providing each $1 of preferential access has total exports from eligible countries under been estimated to be more than $5—a very in- AGOA and Everything But Arms preferences 217 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 of free-trade agreements—and, as noted ear- Figure 6.7 The trade policies of countries lier, of the impact of rules of origin. As MFN in the U.S. generalized system of tariffs are by definition not associated with preferences are more protectionist than rules of origin, such liberalization may well re- those of countries not in the program sult in export gains for countries in products Import policies of countries in U.S. GSP versus those of that in principle benefit from preferences. countries dropped from GSP (percent) Thus, it may be more beneficial for developing 60 countries to obtain more secure MFN reduc- tions on their key exports than to seek to pre- 50 serve preference margins on products with rel- atively high MFN tariffs (Laird, Safadi, and 40 Turini 2003). The available evidence and analysis sug- 30 In GSP gests that preference erosion is unlikely to be 20 Dropped ᮡ a major issue for many countries, given that from GSP automatic compensation will result from ᮡ 10 broad-based multilateral liberalization of mar- ket access. However, specific developing coun- 0 tries and sectors in these countries may be Duties/imports Average Imports/GDP hurt, and resources for adjustment need to be nominal tariff mobilized and allocated. Governments should Source: Ozden and Reinhardt (2002). prepare by determining where adjustment needs are likely to be most significant, so that technical and financial assistance can be pro- vided. Actions of the type discussed in chap- do not enter duty free (Inama 2003, Brenton ters 2 and 5 to facilitate trade, complemented 2003) is one illustration. by the adoption of more liberal rules of origin, One way of obtaining a sense of the magni- will help to attenuate the impact of preference tude of possible preference erosion is first to erosion. assume that LDCs obtain full preferential ac- The low share of exports entering under cess to Quad markets and then to assess the preferences, and the recent research suggesting impact of a reduction in MFN tariffs. A recent that rules of origin play a role in that low exercise along these lines suggests that the ag- share, suggest that the rules used to determine gregate impact of a 40 percent reduction in av- origin should be simplified. The recent experi- erage MFN tariffs would lower LDC exports ence under AGOA, under which several bene- by about $440 million, or 1.6 percent of total ficiary countries significantly expanded ap- exports (IMF 2002). This estimate does not parel exports to the United States after origin take into account terms-of-trade changes from restrictions were relaxed, illustrates this point. the MFN tariff reductions, which on average The WTO includes an Agreement on Rules of can be expected to be positive for exporting Origin that aims to foster the harmonization countries, reducing the losses from preference of the rules used by members. The agreement erosion. Moreover, in practice, it is likely that calls for a work program to be undertaken by part of the rents from preferences accrues to a Technical Committee, in conjunction with the importing countries, and especially to in- the World Customs Organization, to develop termediaries (Tangermann 2002). Account also a classification system regarding changes in should be taken of the benefits to countries tariff subheadings based on the Harmonized with preferences of the erosion in preferential System (Hoekman and Kostecki 2001). The access of other countries—especially members harmonization program provides a potential 218 D E V E L O P M E N T A N D T H E D O H A A G E N D A Figure 6.8 Countries enjoying preferences have increased their exports of apparel to the United States U.S. apparel imports, 1989–2002, by source (millions of dollars) Africa and Caribbean Africa and Caribbean 70 without preferences under preferences 60 NAFTA countries under preferences ᮡ ᮡ 50 ᮡ 40 ᮡ 30 NAFTA countries 20 without preferences 10 Rest of the world 0 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Source: U.S. International Trade Commission. solution to rules-of-origin problems. While textile and apparel products, tariff peaks, and the harmonized rules are intended to be ap- tariff escalation would not only be benefi- plied in cases of nonpreferential commercial cial to developing countries (and developed- policy—tariffs, import licensing, antidump- country consumers), but also would facilitate ing—they could be applied to preferential further trade reforms in developing countries. trade as well. A recent proposal by Canada to Developed and developing countries alike use a common value-added criterion and to could affirm their commitment to poverty alle- allow for comprehensive cumulation (extend- viation by accepting an ambitious program of ing to major players such as China) is an al- liberalization that would include the abolition ternative approach. Yet another option is to of export subsidies, substantial decoupling of emulate the “visa” procedure for textile prod- agricultural support, and significant reduction ucts used under AGOA—this could help to of MFN tariffs on labor-intensive products of substantially reduce uncertainty for traders. export interest to developing countries. The Indeed, minimizing uncertainty is a critical fea- program must include significant trade liberal- ture of making preferential regimes effective.13 ization by developing countries, a major source of the total potential gains. MFN liberalization The ultimate goal is MFN-based, should extend to middle-income countries, reciprocal liberalization which are among the most dynamic markets in MFN-based market access will have the great- the world and where trade barriers are often est beneficial impact on development.14 One substantially higher than in developed coun- reason is that it implies that elements of “re- tries, and would usefully extend to LDCs as verse SDT”—special opt-outs and exemptions well. that benefit interest groups in industrialized In defining negotiating modalities to pursue countries at the expense of developing coun- desired MFN liberalization, WTO members tries—will be removed. Eliminating agricul- should set a concrete timetable and agree on tural subsidy programs, high protection for specific benchmarks for product coverage and 219 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 maximum tariffs. The challenge is to identify ing these questions, it is helpful to distinguish reciprocal commitments that make economic between (a) agreements and disciplines that sense and support development. The overuse pertain to the core business of the WTO— of nonreciprocity in past market-access negoti- traditional trade policies such as tariffs, quo- ations has excluded developing countries from tas, and export subsidies—and (b) rules whose the major source of gains from trade liberal- implementation requires significant resources ization—namely the reform of their own poli- or the existence of well-functioning comple- cies. Nonreciprocity is also a reason why tariff mentary institutions. With respect to the first peaks today are largely on goods produced in category, developing countries would benefit developing countries. A willingness to pursue from abiding by the same trade-policy disci- liberalization at home is critical to increase plines that apply to developed countries. The developing countries’ participation in global overwhelming conclusion in the economic lit- trade, particularly South-South trade, which is erature is that traditional trade-policy instru- subject to significant barriers (see chapter 3).15 ments should not be used in pursuit of devel- Services are of great importance to develop- opment objectives.17 ment—it is difficult for firms to be competitive A country’s trade policy is a key link in the in the absence of efficient services sectors. Sub- transmission of price signals from the world stantial opportunities exist to expand develop- market to the national economy. Undistorted ing-country services exports and to liberalize price signals from world markets, in com- further access to developing-country markets. bination with an exchange rate that reflects While the latter would bring the greatest gains, macroeconomic conditions, encourages effi- temporary access to service markets (particu- cient resource allocation consistent with com- larly labor markets) in developed countries parative advantage. An open trade regime gives also would generate large gains for developing consumers and firms access to a greater vari- countries (see chapter 4). In addition, binding ety of goods and services, including capital the current set of liberal policies applied to and intermediate goods, and contributes to cross-border trade (GATS Modes 1 and 2) productivity growth through access to global would assist governments in pursuing domes- technology and by forcing domestic firms to tic reforms. Many developing countries have become more efficient. begun to exploit opportunities offered by the Although numerous arguments have been Internet and telecommunication networks to developed that potentially provide a rationale provide services through cross-border trade. for intervention to protect infant industries— Currently such trade is generally free of re- most of which revolve around some type of strictions, but that freedom is not locked in market failure or externality—trade policy is through the GATS (Mattoo 2003). rarely if ever an appropriate instrument. More- over, well-known political economy problems are associated with protection of infant indus- Toward a new regime tries. The prospect of protection can give rise to for WTO rules unproductive rent-seeking behavior with asso- S everal WTO agreements offer developing countries some latitude to pursue restric- tive trade policies and provide transition peri- ciated scope for (legal) lobbying and (illegal) corruption (Bhagwati 1988). Moral hazard problems can easily arise because the reward for ods and technical assistance to help in imple- an industry doing well is the removal of protec- menting agreements (box 6.2).16 tion, which can generate perverse incentives for Should trade rules apply to all developing firms to underperform so as to retain protection. countries? If so, should account be taken of Economic first principles suggest that a sub- differences in national capacities to implement sidy-type policy generally will be less distorting and benefit from multilateral rules? In answer- (more efficient) than trade policy in offsetting 220 D E V E L O P M E N T A N D T H E D O H A A G E N D A Box 6.2 Major WTO provisions allowing developing countries greater freedom to use restrictive trade policies Infant industry protection. GATT Articles XVIII:a grounds, and those that distort the incentive to trade and XVIII:c allow for removal of tariff concessions in a major way. Nonspecific subsidies (defined as or use of quotas if necessary to establish an industry those for which access is general or eligibility auto- in a developing country. Compensation must be of- matic on the basis of clear, objective criteria) are fered to countries that would be negatively affected. permitted and cannot be countervailed. Subsidies Balance-of-payments protection. Article XVIII:b contingent on export performance or on the use of allows a nation to impose trade measures to safe- domestic rather than imported goods are generally guard its balance of payments. In contrast to Articles prohibited. Permitted measures that create “serious XVIII:a and c, surveillance and approval procedures prejudice”—defined to exist if the total ad valorem are less burdensome, and compensation need not be subsidization of a product exceeds 5 percent, if subsi- offered to affected countries. Not surprisingly, no dies are used to cover operating losses of a firm or country has invoked the infant industry provisions of industry, or if debt relief is granted for government- the GATT since 1967, but numerous countries have held liabilities—can be countervailed or disputed. made use of Article XVIII:b. In the Uruguay Round, Subsidies that can be shown to have had a negative Article XVIII:b was revised and surveillance proce- effect on a partner’s exports likewise may be counter- dures were tightened. WTO members must now vailed or disputed. LDCs and countries with GNP publicly announce time schedules for the removal per capita below $1,000 are exempted from the pro- of restrictive import measures taken for balance of hibition on export subsidies. Developing countries payments purposes and must, in principle, use price- that have become competitive in a product—defined based measures (such as tariffs). as having a global market share of 3.25 percent— Subsidies: The WTO Agreement on Subsidies must phase out any export subsidies over a two-year and Countervailing Measures (ASCM) attempts to period. distinguish between subsidies (defined as financial contributions by government) that can be justified on the grounds of market failure, or on noneconomic Source: Hoekman and Kostecki (2001). an externality. From an economic viewpoint, abide by the same rules on the use of traditional the drafters of the GATT were therefore justi- trade polices that pertain to developed coun- fied in placing relatively stringent constraints on tries, however, will benefit consumers and en- the use of trade policy, and in particular, on the hance welfare in developing countries. use of quantitative restrictions and local content The WTO does not identify or constrain requirements. Moreover, in cases in which im- what governments can do in the realm of non- port competition proves too fierce, the WTO trade policy to maximize the benefits of trade. allows for safeguards—emergency protection As emphasized by Stern (2002), any credible through safeguards. However, WTO safeguard poverty reduction strategy must rest on two provisions impose conditions that enhance cer- pillars: a good investment climate to propel tainty and help ensure that interventions are growth, and empowerment of poor people made only for good cause. through participation in decisions that shape The foregoing does not imply that develop- their lives. Although a sound trade policy is a ing countries should be forced to sign away major element of any successful development their ability to use trade policies—all countries strategy, other factors are equally important. have the right under the WTO to impose tariffs Institutions to manage the distributional impli- or export taxes if they so desire. Committing to cations of trade reforms and to ensure that 221 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 consumers, enterprises, and farmers have ac- holders in developing countries who might in- cess to competitive markets for goods and ser- form their governments of their preferences vices are critical to harnessing greater trade for and clarify the implications of various propos- development. als (Finger 2001; Hoekman 2002). As a result, Nor does the WTO constrain the ability of the “trade-related” aspects of international is- governments to address market failures through sues are often identified by constituencies in subsidies or taxes. In many cases the interven- major trading powers and may not have much tion required to address an externality will be relevance for development. horizontal (general), not sector- or industry- Because countries differ widely in their do- specific, and thus be nonactionable. For exam- mestic priorities and in their capacities to im- ple, food subsidies, household energy subsidies, plement change, it is as important as it is dif- and health and education programs are not vul- ficult to identify how and where development nerable to WTO action. Similarly, factor-use will be promoted by proposals to expand the subsidies—for example, for the wages of work- reach of the WTO into (new) regulatory areas. ers taken directly off the unemployment rolls— With respect to behind-the-border policies are permissible unless they are configured in affecting trade, it is difficult to design generic such as way as to make them de facto subsidies rules that apply to all. Even if negotiators get to specific sectors. Overall, therefore, the sorts the economics right, there is a danger that of subsidies of most use in fighting poverty and good policies may be resisted. Dealing with offsetting market failures are not constrained these types of issues in the context of negotia- by WTO disciplines (McCulloch, Winters, and tion may also induce a predominant focus on Cirera 2001). costs, as reforms will be seen as concessions to foreign interests, as opposed to being in the na- Several policy options are open for future tional interest (Finger 2002). Given their com- rules and regulations plexities and the limited negotiating resources Some developing countries and many civil so- available in most low-income countries, such ciety groups have charged that some WTO issues may be best left off the negotiating table agreements do not support development. In (Winters 2002). However, if negotiations are such cases, the appropriate solution may be to launched on these topics, it will be important reopen (renegotiate) agreements where mem- to recognize that for many countries they may bers perceive the rules to be unbalanced or not have a high priority. detrimental to their interests. This has been the approach taken by some proponents of the so- Differential interests and capacities must called Development Box in the Agreement on be recognized Agriculture (box 6.3). It is also the approach As noted previously, the challenge is to get the that has been suggested by the chair of the rules right from a development perspective. WTO General Council to address several pro- Even if this is done, in the sense that con- posals made by developing countries on SDT. stituencies in developing countries are enthusi- In addition to the Agreement on Agriculture, astic about what has been negotiated and gov- an agreement often viewed as unbalanced is ernments regard implementation as being in the Agreement on Trade-Related Aspects of In- the national interest, other issues may consti- tellectual Property Rights (TRIPS).18 tute a higher priority for investment of scarce One lesson that emerged from the Uruguay administrative and financial resources. Round is that it is important but difficult to These observations suggest the need for dif- assess what makes sense from a development ferentiation among developing countries in perspective. In large part this is a reflection of determining the reach of WTO rules that are the absence of strong trade interests and stake- resource intensive; that is, those that require 222 D E V E L O P M E N T A N D T H E D O H A A G E N D A Box 6.3 A “development box” for the Agreement on Agriculture? institutions that promote marketing, quality con- T he primary focus of the Uruguay Round Agree- ment on Agriculture was to bring this sector back into the trading system—that is, to reimpose trol, or otherwise strengthen the competitiveness of poor farmers. multilateral disciplines on trade-distorting domestic • A “special” safeguard provision, available only support policies. A major feature of the agreement to developing countries, to provide rapid, time- was to distinguish between permitted subsidies (the limited protection against import surges that hurt “green box”) and subsidies subject to reduced com- poor producers, especially dumped imports of mitments and disciplines. Because the objective of subsidized goods. negotiators in the Uruguay Round was to reduce dis- • Acceptance that some products—especially key torting types of support, it does not cover the types staple commodities critical to food security—would of market imperfections likely to be found in devel- not be subject to liberalization commitments. oping countries, nor does it recognize their need to pursue “second-best” policies where they do not have the institutional capacity to pursue the most While often defended as examples of needed prefer- efficient policies to combat poverty. As a result, ences, proposals for a development box effectively several countries have sought to introduce a “de- involve changing the terms of the Agreement on velopment box” into the agreement, which would Agriculture. Many of the provisions have been in- identify a set of measures to enhance food security, cluded in the “Harbinson” draft, which suggests ap- stimulate agricultural production, and reduce rural proaches for future liberalization commitments in poverty in developing countries. Examples of such the Doha negotiations on agriculture (WTO 2003). proposals: The draft also contains a large number of provisions permitting developing countries far greater leeway in • Direct and indirect investment and input subsidies protecting agriculture through border measures such or other supports to households below the na- as tariffs and tariff quotas than would be the case tional poverty level to encourage agricultural for developed countries. Such policies may be justi- and rural development. Such supports could be fied because low-income developing countries do product-specific as well as general, as long as not have the fiscal capacity to support agriculture they are effectively targeted to the rural poor. through less trade-distorting direct-income supports, • Programs supporting product diversification in but they may lead to the same inefficiencies that small, low-income developing countries currently have undermined competitiveness of many industries dependent on limited commodities for their ex- nurtured behind high protective barriers. While get- ports, including programs involving government ting the economics right is important and requires assistance for risk management. careful analysis, the efforts to alter the terms of the • Domestic foodstuffs at subsidized prices in tar- Agreement on Agriculture is arguably the most ap- geted programs aimed at meeting food require- propriate method of addressing rules that are not ments of the poor, whether urban or rural, as perceived to support development. part of an overall effort to enhance food security. • Transportation subsidies for agricultural products and farm inputs to poor remote areas. • Programs involving government assistance for es- Sources: Ruffer, Jones, and Akroyd (2002); Hoekman, Michalo- tablishment of agricultural cooperatives or other poulos, and Winters (2003). 223 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 significant complementary legal, administra- rules. Country classification is inevitably a sen- tive, and institutional investments or capacity sitive issue, as is the question of determining or that will result in large transfers from de- the “coherence” of WTO rules with national veloping countries. The basic rationale for dif- development priorities and identifying the im- ferentiation is that certain agreements may not plications of different types of WTO rules. be development priorities for some countries What constitutes “resource intensive,” for ex- or may require that other preconditions be ample? And what agreements would give rise satisfied before implementation can be benefi- to large implementation costs? These questions cial. Such preconditions can be proxied by the will require analysis. Determining implementa- attainment of a minimum level of per capita tion criteria would require input from relevant income, institutional capacity, and economic development institutions, national and interna- scale (size). Some WTO disciplines may not be tional, to strengthen policy coherence at both appropriate for very small countries, for ex- levels. One advantage of an agreement-specific ample, in that the regulatory institutions re- “implementation audit mechanism” is that it quired may be unduly costly.19 could avoid explicit country classifications, giv- Several options could be considered to take ing rise instead to a monitoring process to sup- country differences into account in WTO agree- port developing countries in managing trade ments.20 Options include: reforms and implementing WTO agreements by recognizing competing demands on scarce • Adopting a rule of thumb that makes a resources. group of countries eligible to opt out of Involving development agencies may reduce provisions that entail substantial imple- the risk of inducing countries to adopt and pur- mentation costs until specific criteria or sue a program of trade and regulatory reform benchmarks have been met. This rule that may not, in fact, be suited to the country. would require renegotiating the current During the Uruguay Round many countries set of country groups recognized in the were concerned about avoiding possible “cross- WTO (the LDCs, all developing coun- conditionality” between WTO and interna- tries, and developed countries). It would tional financial institutions; this led to a minis- also require establishing criteria to iden- terial declaration on “coherence” to call for tify which agreements are affected. “avoiding the imposition on governments of • Establishing an agreement-specific ap- cross-conditionality or additional conditions” proach under which application of rules resulting from cooperation between the WTO to any given country is determined by and the international financial institutions.21 agreement-specific criteria, possibly linked Indeed, care would be required to avoid “cross- to availability of technical assistance and conditionality.” an action plan for implementation. Several alternative options may therefore be • Adopting a country-specific aproach that feasible in recognizing differences in the ability places trade-reform priorities in the con- to benefit from implementation of resource- text of national development plans as intensive rules. Deciding on the best approach defined in Poverty Reduction Strategy Pa- will require considerable thought and discus- pers, and relying on multilateral monitor- sion. What matters most at this point is that ing to establish a cooperative framework WTO members acknowledge the issue. It might under which countries may be assisted in be expeditious to make a decision in Cancun to gradually adopting WTO norms. consider alternative approaches along the lines sketched out above. Given the steady expan- A common feature of these possibilities is sion of the WTO into regulatory areas, doing that they require a narrower definition of eligi- so would help make development relevance bility for temporary exemptions from WTO more than a slogan. 224 D E V E L O P M E N T A N D T H E D O H A A G E N D A WTO commitments must be enforced external shocks, suggest that more attention through monitoring and dispute settlement and resources should be devoted to costing out While getting rules right is important, so is en- the implementation requirements of proposed forcement and accountability. Two aspects of rules and to calculating their costs and bene- enforcement of commitments are particularly fits. The multiagency Integrated Framework important. The first is regular information on for Trade-Related Technical Assistance could implementation of agreements and comple- be used to support this objective in LDCs. mentary actions being pursued in other forums Developing-country think tanks and policy and organizations. The second concerns the research networks—for example, the Global ability of developing countries to use WTO Development Network—also have an impor- dispute-settlement mechanisms in instances in tant role to play in assisting governments with which partners have not respected the terms of the required assessments at the national level, an agreement. Developing countries may be at supported by bilateral and multilateral devel- a disadvantage in using such mechanisms be- opment institutions. cause of resource constraints and lack of retal- iatory power. Proposals to address such biases Dispute settlement are discussed below. Whatever agreements eventually emerge from the Doha Round, enforcement will be impor- tant. But how well can low-income countries Strengthened assessment and monitoring defend their rights in the WTO? During 1995– Strengthening mechanisms for regular monitor- 2002, 305 bilateral disputes were brought to ing of implementation of agreements and per- the WTO, entailing over 1,800 “grievances”— formance of both developed and developing specific allegations of violation of a WTO pro- countries would help improve transparency vision (Horn and Mavroidis 2003). Developing and accountability. This should extend to the countries brought 123 of the 305 disputes, provision of information on national trade- about one-third. Most of these complainants related priorities by developing countries, the were middle-income economies. Low-income funding and investment requirements these pri- countries (defined by Horn and Mavroidis as orities involve, and the extent to which in- those with a per capita income below $800) ternational and bilateral donors have provided were complainants in only 18 cases and respon- assistance. A first step in compiling informa- dents in only 21. LDCs did not participate at tion on assistance provided was taken in 2002 all—they never acted as complainant or respon- by the WTO and the Development Assistance dent. Thus, well over half of the WTO mem- Committee of the Organisation for Economic bership does not participate in WTO dispute Co-operation and Development (OECD), settlement.22 building on a database of bilateral and multi- That many developing countries have not lateral development projects. The WTO’s Trade participated in WTO dispute settlement re- Policy Reviews provide a potential mechanism flects the manifold challenges of such partici- for bolstering monitoring, although the publi- pation. A first challenge to defending rights cation schedule of reports would need to be through the WTO is obtaining knowledge that increased for timely information to be made a WTO provision may have been violated by available to WTO members. The type of mech- a partner government. A second is convincing anisms to put in place would depend in part on the government to bring the case forward— the approach taken to determine the reach of enterprises alone have no legal standing before resource-intensive rules. the WTO. A third is expectation of a positive The weaker social safety nets and insurance payoff from bringing a case—a function of the mechanisms in the developing world, as well remedy available and the likelihood that the as higher rates of poverty and vulnerability to trading partner will actually implement it. 225 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Small countries cannot credibly threaten re- done, it is important to halt the emerging trend taliation—the ultimate threat that can be made toward escalating retaliation and the use of against a member that does not comply with a trade sanctions. WTO panel recommendation—because raising import barriers will have little impact on the Aid for trade: addressing target market, while being costly in welfare national priorities terms.23 Thus, pressure to comply with panel Trade capacity is critical in ensuring that rulings is largely moral. In practice, the system low-income countries are able to benefit from has worked rather well, in that recourse to re- trade opportunities. Numerous studies—most taliation has rarely been required to enforce recently the Diagnostic Trade Integration Stud- multilateral dispute-settlement decisions. This ies undertaken in LDCs under the auspices is a reflection of the repeated nature of WTO of the Integrated Framework initiative—have interactions and the resulting value that gov- identified institutional weaknesses and an ad- ernments attach to maintaining a good reputa- verse investment climate as a major source of tion. Nonetheless, asymmetry in enforcement comparative disadvantage.27 ability can affect incentives to use the system. The reports reveal that many countries re- The classic recommendation by economists to main largely without the appropriate institu- address the problem is to change the rules so tional frameworks and systems to manage that nonimplementation of panel recommen- their trade policy and that trade costs severely dations would be punished by withdrawal of limit the competitiveness of developing-coun- market-access commitments by all WTO mem- try firms in export markets. Transport costs bers. But suggestions to this effect have always are often the single most important component been resisted (Hudec 1987, 2002). of cost for exporters in several of the countries. Retaliation involves raising barriers to trade, The transport sectors (air, ports, trucking) are which is generally detrimental to all parties. The often plagued by a lack of domestic competi- power of retaliation may also be captured by tion (see chapter 5). protectionist interests in an importing country. These anticompetitive conditions impinge A superior approach would be to strengthen directly on the welfare of the poor. In most compensation provisions. Developing countries LDCs, the majority of poor households derive have proposed, for example, that WTO panels their income from agriculture and agro- should be authorized to recommend payment of processing. Agricultural trade integration can financial compensation in cases where a devel- therefore increase both productivity and rural oping country loses its trade in a product as a incomes by providing better access to modern result of actions by a developed country that are inputs and technologies and by encouraging inconsistent with WTO norms.24 Such sugges- exports.28 In addition to the fragmentation of tions have a long history (Hudec 2002).25 Mex- markets and the remoteness of many farming ico recently suggested allowing countries that communities, binding constraints include trans- have won a dispute but where implementation portation bottlenecks that, combined with nu- has not occurred to auction off the resulting re- merous informal fees and internal checkpoints, taliation rights.26 lead to high transaction costs (in Ethiopia, While compensation or fines would be less Guinea, and Nepal), lack of basic post-harvest distorting than trade sanctions, they may not marketing infrastructure (in Ethiopia, Guinea, be very effective in inducing compliance, as the Malawi, Mauritania, and Yemen), lack of costs would disperse among all taxpayers. water control infrastructure (in Ethiopia and Other options should therefore be considered, Senegal), and a pervasive degradation of nat- including stronger surveillance mechanisms ural resources (Tsikata 2003). To improve com- and greater opportunities for interested parties petitiveness and extend the benefits of trade to to bring cases in national forums. Whatever is the poorer segments of the population, better 226 D E V E L O P M E N T A N D T H E D O H A A G E N D A domestic market integration should be a key In fact, liberalized market access on an MFN policy objective in many LDCs. basis has the greatest potential payoff, in Development assistance can play an impor- terms of development and poverty reduction, tant role in bolstering trade capacity, thereby al- of any issue on the Doha agenda. lowing countries to benefit from liberalized ac- Because duty- and quota-free access to cess to international markets. Additional funds major markets can help to offset the many dis- are needed to address both policy and public- advantages that confront firms in poor coun- investment priorities, to help low-income coun- tries, developed countries should continue to tries adapt to a reduction in trade preferences grant trade preferences to LDCs and similarly following further nondiscriminatory trade lib- disadvantaged countries, emulating Europe’s eralization, and to assist poor net-importing Everything But Arms initiative. To maximize countries to deal with the potential detrimental the benefits of such deep preferences, con- effects of a significant increase in world food certed action should be taken to minimize the prices, should these materialize. The world trade-restricting effect of rules of origin and community made general commitments to this related administrative requirements. effect at the International Conference on Fi- But preferences are best viewed as a transi- nancing for Development in Monterrey in tional instrument. What matters most in any March 2002—the need now is to translate the effort to expand exports on a sustained basis “Monterrey consensus” into identification and is a good investment climate, including open- financing of trade-related investment priorities. ness to trade. Indeed, most of the potential Although more “aid for trade” would be gains from trade reform could be realized by beneficial, it is important to avoid a situation unilateral liberalization. in which a desire by donor countries to see de- Traditional forms of preferential treatment veloping countries implement certain WTO cannot be the primary instruments to enhance agreements diverts assistance away from recip- the development-relevance of the WTO. In- ients’ own development priorities. This is one stead, countries must commit to engage in the of the risks of suggestions to make technical reciprocal exchange of market-access conces- assistance a requirement of new WTO rules sions. Continuing to exempt developing coun- and to link implementation of WTO agree- tries from trade-policy disciplines is not likely ments to the provision of such assistance. Ide- to achieve development goals. ally, identification and delivery of trade-related With regard to behind-the-border regulation, technical assistance should be embedded in the the same principle should apply as with trade- national policy-setting processes used by gov- policy disciplines: ensuring that the rules sup- ernments and the donor community—for ex- port development.29 Getting the rules right re- ample, the Poverty Reduction Strategy Papers. quires each country to think hard about what This would ensure that trade priorities are is in its national interest. Even where the po- considered for funding along with other devel- tential benefit of an agreement may be clear, opment priorities. poor countries may not have the resources re- quired to implement it immediately, perhaps because other issues command higher priority Putting development into the or because complementary policies and insti- Doha agenda tutions must be put in place before implemen- P utting development into the Doha agenda requires actions by developing and devel- oped countries alike. Interests and priorities tation will be beneficial. In this regard, a new concept of special and differential treatment, one that would establish clear criteria and differ from country to country, but improved mechanisms to link implementation to de- access to markets in agriculture, manufac- velopment priorities and capacities, could be tures, and services would be most beneficial. most fruitful. 227 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Promoting the trade and development but mostly because of the ancillary investments that are prospects of low-income countries requires ac- required to allow the rule to be applied. 6. Note that China and Mexico are included in the tion on many fronts. The key needs—to estab- European Union’s GSP recipients’ list while they are lish priorities for reforming domestic policy excluded from the U.S. list. The United States has a free and enhancing trade-related capacity—grow trade area agreement (FTA) with Mexico and does not more acute as MFN trade barriers are reduced, grant GSP status to China. On the other hand, South and the value of preferences is eroded. A pre- Africa, Turkey, and Eastern European countries have condition for effective use of additional de- FTAs with the European Union and, hence, are ex- velopment assistance—whether in the form of cluded from its GSP while they receive GSP status from grants or development loans—is that the nec- the United States. essary priorities are determined appropriately. 7. The limit is $13 million if the exporting country has more than 50 percent market share. In many low-income countries much more 8. For most textile and clothing products, the 20 should be done to integrate trade priorities into percent reduction is less than 3.5 percentage points. national development strategies and invest- 9. According to EU Council regulation 2501/2001, ment allocation decisions. there are some nonsensitive agricultural and food It should be possible to move rapidly to- products. According to Annex IV, these are artichokes, ward providing greater access for all countries castor oil, frogs’ legs, grapefruit, green tea, inactive to others’ markets in goods and services. As yeasts, licorice extract, malt beer, papayas, pepper, and far as rule-making is concerned, agreement on sweet potatoes. 10. U.S. intellectual property advocacy groups have an approach that recognizes the significant used GSP eligibility as an instrument to induce a num- differences among countries will give develop- ber of countries (including El Salvador, Honduras, ment a much more solid place in the WTO. Panama, Paraguay, Poland, and Turkey) to take actions Most important is to address what is by far in such areas. the most urgent challenge of expanding trade 11. Rules of origin are intended to prevent trade de- for development—identifying and dealing with flection and to determine where a good originates for trade constraints within countries, whether duty purposes when two or more countries are in- those constraints derive from policies, poor in- volved in the production of a product. The general rule is that the origin of a product is the one in which the frastructure, or limitations in capacity. last substantial transformation took place, that is, the country in which significant manufacturing or process- ing occurred most recently. Significant or substantial is Notes defined as the level of transformation sufficient to give 1. See World Bank (2002), Finger and Schuler the product its essential character. Various criteria can (2000), and Hoekman and Kostecki (2001) for a dis- be used to determine if a substantial transformation cussion and references to the literature. occurred. These include a change in tariff heading (as 2. Hudec (1987) and Finger (1991) review the when, as a result of whatever processing was per- background in some depth, noting that SDT was heav- formed the good is classified under another category of ily influenced by foreign policy considerations, espe- the Harmonized System), the use of specific processing cially the Cold War. operations, tests based on the value of additional ma- 3. The full name is Decision on Differential and terials embodied in the transformed product, or the More Favorable Treatment, Reciprocity and Fuller Par- ticipation of Developing Countries. See Hudec (1987) amount of value added in the last country where the for further discussion of the history. good was transformed. 4. This was first documented by Finger (1974, 12. These categories make up around 65–75 per- 1976). cent of the exports of the AGOA and Andean Program 5. Because WTO rules are often based on those pre- beneficiaries and tend to dominate general patterns vailing in OECD countries, implementation costs are with their volatile prices. asymmetrically distributed. Post-Uruguay Round re- 13. A major benefit of Everything But Arms is that search—for example, Finger and Schuler (2000) and the preferences are not time-limited. Finger (2001)—revealed that the costs associated with 14. This was the approach used in the GATT before complying with certain WTO disciplines can be signif- the creation of the GSP. See Hudec (1987) and Finger icant. This is in part because of the rules themselves, (1991). 228 D E V E L O P M E N T A N D T H E D O H A A G E N D A 15. Given that many developing countries either with the number of incompatible measures a country’s have not bound tariffs at all or have high tariff bind- exporters encounter and the volume of trade, and takes ings, this will automatically imply that credit will be into account that there is likely to be a threshold below given for past reductions in applied tariffs and—pro- which it is not worth bringing a case. vided formulas are used—that autonomous liberaliza- 23. See Bown (2002) for a recent analysis and ref- tion (reduction in applied tariff rates) will not prejudice erences to the relevant literature. future WTO negotiations. See Francois and Martin 24. WT/GC/W/162 (2003) for an in-depth analysis of alternative formula- 25. For a discussion of a proposal by Brazil and based approaches. Uruguay to reform the dispute settlement system to in- 16. This section draws in part on Hoekman, clude financial compensation, see Dam (1970, 368–73). Michalopoulos, and Winters (2003). 26. In their analysis of this proposal, Bagwell, 17. See, for example, Hoekman, Michalopoulos, Staiger, and Mavroidis (2003) conclude that the prob- and Winters (2003), Noland and Pack (2003), Irwin ability of the winning country being compensated is (2001, 2003), and Hausmann and Rodrik (2002). highest if such rights extend to the losing party. 18. Research on the effects of TRIPS suggests that 27. The Diagnostic Trade Integration Study (DTIS) net transfers from low- to high-income countries could is an analytical tool developed to reexamine the policy be substantial (World Bank 2001). and institutional constraints to trade for LDCs, and to 19. For example, despite remarkable reductions in identify technical assistance needs for the purpose of customs clearance times that have been achieved by enhancing LDCs’ integration into the global economy. some LDCs (sometimes from weeks to days or hours, As of July 2003, DTIS reports had been completed as in Senegal), the customs regimes in many participat- for Burundi, Cambodia, Ethiopia, Guinea, Lesotho, ing countries are characterized by long clearance times, Madagascar, Malawi, Mali, Mauritania, Nepal, Sene- a plethora of informal fees, and inadequate perfor- gal, and Yemen. See www.worldbank.org/trade. mance monitoring indicators (see chapter 5). Many 28. Increasing productivity in agriculture is critical countries are struggling to implement the Agreement for the transformation of these economies. Reducing on Customs Valuation and to work with and reform the price of food products increases the real income of other institutions whose actions impinge on customs the whole population and allows higher household efficiency such as security and enforcement. spending in nonagricultural products, thus favoring 20. These options are discussed further in Stevens diversification. (2002), Prowse (2002), Wang and Winters (1999), and 29. The GATT/WTO may have it right when it Hoekman, Michalopoulos, and Winters (2003). comes to rules on the use of trade policy; for example, 21. Declaration on the Contribution of the World the ban on the use of quotas and the focus on binding Trade Organization to Achieving Greater Coherence in tariffs. This is much less clear when it comes to other Global Economic Policymaking, December 15, 1993. agreements, such as TRIPS. The solution in such cases 22. Busch and Reinhardt (2002) found that devel- is to reopen existing agreements, something that can oping countries accounted for around 30 percent of readily be done. Many of the agreements are already complaints under both GATT and WTO, but that the subject to ongoing negotiations. share of cases against developing countries had risen from 8 percent to 37 percent during the period covered by their study. This suggests that the shift to the References WTO—with the associated expansion of disciplines on Bagwell, K., R. Staiger, and Petros C. Mavroidis. 2003. developing countries—has given rise to a significant in- “The Case for Auctioning Countermeasures in the crease in the probability of engaging in dispute settle- WTO.” Mimeo, Columbia University, New York. ment. However, Holmes, Rollo, and Young (2002) Beghin, John, and M. Ataman Aksoy. 2003. “Agricul- concluded that the simple hypothesis that disputes will tural Trade and the Doha Round: Preliminary be proportionate to trade shares is not borne out by the Lessons from Commodity Studies.” Mimeo. data. They also found that income per head or mea- World Bank, Washington, D.C. sures of openness did not help to explain the incidence Bhagwati, Jagdish. 1988. Protectionism. Cambridge, of disputes, suggesting that there is no significant bias Mass.: MIT Press. between small and large countries, or between richer Bown, Chad. 2002. “The Economics of Trade Dis- and poorer countries, in terms of participation in the putes, the GATT’s Article XXIII and the WTO’s system as complainants or respondents. Horn, Mavroi- Dispute Settlement Understanding.” Economics dis, and Nordström (1999) similarly conclude that the and Politics 14: 283–323. evidence for “bias” is not particularly strong once one Brenton, Paul. 2003. “Integrating the Least Developed controls for the fact that disputes should be correlated Countries into the World Trading System: The 229 G L O B A L E C O N O M I C P R O S P E C T S 2 0 0 4 Current Impact of EU Preferences under Every- Hoekman, Bernard, Constantine Michalopoulos, and thing But Arms.” Policy Research Working Paper L. Alan Winters. 2003. “Development and More 3018. World Bank, Washington, D.C. Favorable and Differential Treatment of Develop- Brenton, Paul, and M. Manchin. 2002. “Making EU ing Countries.” Mimeo, Development Research Trade Agreements Work: The Role of Rules of Group, World Bank, Washington, D.C. Origin.” CEPS Working Document 183. Brussels. Hoekman, Bernard, Francis Ng, and Marcelo Olar- Busch, M., and E. Reinhardt. 2003. “Developing Coun- reaga. 2002. “Tariff Peaks in the Quad and Least tries and GATT/WTO Dispute Settlement.” Mi- Developed Country Exports.” World Bank Eco- meo. Emory University, Atlanta, Georgia. nomic Review 16: 1–22. Dam, Kenneth. 1970. The GATT: Law and the Inter- Holmes, P., J. Rollo, and A. Young. 2002. “Emerging national Economic Organization. Chicago: Uni- Trends in WTO Dispute Settlement: Back to the versity of Chicago Press. GATT?” Mimeo. University of Sussex, U.K. Evenett, Simon. Editor. 2003. The Singapore Issues Horn, Henrik, and Petros C. Mavroidis. 2003. “Which and the World Trading System. Berne: State Sec- WTO Provisions are Invoked by and Against De- retariat for Economic Affairs. veloping Countries?” Mimeo. Institute for Inter- Finger, J. Michael. 1974. “GATT Tariff Concessions national Economics, Stockholm. and the Exports of Developing Countries: United Horn, Henrik, Petros C. Mavroidis, and Hakan Nord- States Concessions at the Dillon Round.” Eco- ström. 1999. “Is the Use of the WTO Dispute Set- nomic Journal 84 (335). tlement System Biased?” CEPR Discussion Paper, ———. 1976. “Effects of the Kennedy Round Tariff Center for Economic Policy Research, London. Concessions on the Exports of Developing Coun- Hudec, Robert. 1987. Developing Countries in the tries.” Economic Journal 86 (341). GATT Legal System. London: Trade Policy Re- ———. 1991. “Development Economics and the Gen- search Centre. eral Agreement on Tariffs and Trade.” In J. de ———. 2002. “The Adequacy of WTO Dispute Settle- Melo and André Sapir, eds., Trade Theory and ment Remedies: A Developing Country Perspec- Economic Reform: North, South, and East. Ox- tive.” In Bernard Hoekman, Aaditya Mattoo, and ford, U.K.: Basil Blackwell. Philip English, eds., Development, Trade, and the ———. 2001. “Implementing the Uruguay Round WTO: A Handbook. Washington, D.C.: World Agreements: Problems for Developing Coun- Bank. tries.” The World Economy 24 (9): 1097–1108. IMF (International Monetary Fund). 2002. “Financing Finger, J. M. 2002. “The Doha Agenda and Develop- of Losses Due to Preference Erosion.” Washing- ment: A View From the Uruguay Round,” ton, D.C. Mimeo. Manila: Asian Development Bank. Inama, Stefano. 2003. “Trade Preferences and the Finger, J. Michael, and Philip Schuler. 2000. “Imple- WTO Negotiations on Market Access.” Mimeo. mentation of Uruguay Round Commitments: The United Nations Commission on Trade and Devel- Development Challenge.” The World Economy 23: opment (UNCTAD), Geneva. 511–25. Irwin, Douglas. 2002. Free Trade Under Fire. Prince- Francois, J., and Will Martin. 2003. “Formula Ap- ton: Princeton University Press. proaches for Market Access Negotiations.” The ———. 2003. “Interpreting the Tariff-Growth Corre- World Economy 26: 1–28. lation of the Late Nineteenth Century.” American Hausmann, R., and Dani Rodrik. 2002. “Economic Economic Review 91 (May 2002): 165–69. Development as Self-Discovery.” CEPR Discus- Keck, Alexander, and Patrick Low. 2003. “Special and sion Paper 3356, Center for Economic Policy Re- Differential Treatment.” Mimeo. World Trade In- search, London. stitute, Berne. Hoekman, Bernard. 2002. “Strengthening the Global Laird, Sam, Raed Safadi, and Alessandro Turini. 2003. Trade Architecture for Development.” World Trade “The WTO and Development.” Mimeo. United Review 1: 23–46. Nations Commission on Trade and Development Hoekman, Bernard, and Michel Kostecki. 2001. The (UNCTAD), Geneva. Political Economy of the World Trading System. Mattoo, Aaditya. 2003. “Shaping Future Rules for Oxford, U.K.: Oxford University Press. Trade in Services: Lessons from the GATS.” In Hoekman, Bernard, Aaditya Mattoo, and Philip English, Takatoshi Ito and Anne O. Krueger, eds., Trade eds. 2002. Development, Trade and the WTO: A in Services in the Asia Pacific Region. NBER. Handbook. Washington D.C.: World Bank. Chicago: University of Chicago Press. 230 D E V E L O P M E N T A N D T H E D O H A A G E N D A Mitchell, Donald. 2003. “Sugar Policies: Opportunity Stevens, Christopher. 2002. “The Future of SDT for for Change.” Mimeo. World Bank, Washington, Developing Countries in the WTO.” Mimeo. In- D.C. stitute of Development Studies, University of Sus- McCulloch, N., L. Alan Winters, and Xavier Cirera. sex, U.K. May. 2001. Trade Liberalization and Poverty: A Tangermann, S. 2002. “The Future of Preferential Handbook. London: Center for Economic Policy Trade Arrangements for Developing Countries Research. and the Current Round of WTO Negotiations on Noland, Marcus, and Howard Pack. 2003. Industrial Agriculture.” Food and Agriculture Organiza- Policy in an Era of Globalization: Lessons from tion, Rome. Asia. Washington, D.C.: Institute of International Tarr, David. 2002. “Arguments For and Against Uni- Economics. form Tariffs.” In Bernard Hoekman, Aaditya Ozden, Caglar, and E. Reinhardt. 2003a. “The Perver- Mattoo, and Philip English, eds., Development, sity of Preferences.” Policy Research Working Pa- Trade, and the WTO: A Handbook. Washington, per 2955, World Bank, Washington, D.C. D.C.: World Bank. ———. 2003b. “First Do No Harm: GSP and the Ef- Tsikata, Yvonne. 2003. “Integrating into the Global fect of Trade Preferences on Developing Country Economy: Challenges, Experiences and Policy Exports.” Mimeo. World Bank, Washington, D.C. Lessons from the Integrated Framework.” Mimeo. Panagariya, A. 2002. “EU Preferential Trade Policies World Bank. and Developing Countries.” The World Economy USTR (United States Trade Representative). 2002. 25 (10): 1415–32. “Report on U.S. Trade and Investment Policy To- Prowse, Susan. 2002. “The Role of International and ward Sub-Saharan Africa and Implementation of National Agencies in Trade-Related Capacity the African Growth and Opportunity Act.” Re- Building.” The World Economy (9): 1235–61. port Submitted by the President to the Congress, Rodrik, Dani, ed. 2003. In Search of Prosperity: Ana- Washington, D.C. lytic Narratives on Economic Growth. Princeton, Wang, Z. K., and L. Alan Winters. 2000. “Putting N.J.: Princeton University Press. ‘Humpty’ Together Again: Including Developing Ruffer, Tim, S. Jones, and S. Akroyd. 2002. “Develop- Countries in a Consensus for the WTO.” CEPR ment Box Proposals and Their Potential Effect Policy Paper 4. London. March. on Developing Countries.” Mimeo. Report for Winters, L. Alan. 2002. “Doha and the World Poverty U.K. Department for International Development, Targets.” Paper presented at the Annual Bank London. Conference on Development Economics, World Sapir, André. 1997. “The Political Economy of EC Bank, Washington, D.C. April. Regionalism.” European Economic Review 42: World Bank. 2001. Global Economic Prospects and the 717–32. Developing Countries 2002: Making Trade Work Schiff, M., and L. Alan Winters. 2003. Regional Inte- for the Poor. Washington, D.C.: World Bank. gration and Development. Oxford, U.K.: Oxford World Bank. 2002. Global Economic Prospects and the University Press and Washington, D.C.: World Developing Countries 2003: Investing to Unlock Bank. Opportunities. Washington, D.C.: World Bank. Stern, Nicholas. 2002. A Strategy for Development. Washington D.C.: World Bank. 231 Appendix 1 Regional Economic Prospects East Asia and Pacific As has been the case for some years, growth in 2002 was strongest in the transition Recent developments economies, China and Vietnam, where sus- tained strength in exports, consumption, and T HE YEAR 2002 was one of solid recovery in much of East Asia. GDP growth in investment pushed 2002 GDP growth to 8 developing East Asia rose from 5.5 per- percent and 7 percent, respectively. But growth cent in 2001 to 6.7 percent in 2002. However, also strengthened in several of the countries the renewed economic slowdown in the devel- that had been hardest hit by 2001’s fall in oped world, relatively high oil prices in the world-trade growth and high-tech demand, first part of 2003, and the SARS outbreak in exceeding 6 percent in Korea, 5 percent in the region, dampened the pace of the regional Thailand, and 4 percent in both Malaysia and recovery in the first half of 2003. However, the Philippines. Prompt policy action helped East Asian growth is expected to rebound pro- mute the impact of the Bali terrorist attack on gressively, as economies get beyond the short- Indonesia’s economy, which nevertheless man- run impact of SARS, oil prices wane, and aged only 3–4 percent growth for a second global growth revives. year. Overall, the robust performance sup- A smart rebound in exports was one impor- ported continued reductions in poverty. tant driver in the regional recovery in 2002— The strong growth momentum of 2002 regional exports rose nearly 14 percent in dol- became more diffused in the first quarter of lar terms, after having been flat in 2001 as a 2003, continuing in some countries, waning result of the global slowdown and the deep re- in others, and falling sharply in yet others. cession in world high-tech demand. Intrare- The strongest performances were in China and gional exports were especially strong, in par- Thailand, where year-on-year first-quarter ticular to China, which is emerging as a major growth reached 9.9 percent and 6.7 percent, re- hub for regional production and trade net- spectively, supported by both domestic demand works. Robust household consumption pro- and export growth. First-quarter growth was vided further support for the recovery, and, in 4.5 percent in the Philippines—at the higher several cases, the year saw the start of a end of expectations. However, some signs of a stronger trend in fixed investment as well. But downshift were already emerging in Malaysia, the regional picture was not uniformly upbeat, where first-quarter growth dipped to 4 percent with more modest growth rates, or even con- from over 5 percent in the latter part of 2002, tinued outright recession in some of the smaller while in Indonesia growth continued at a mod- or lower-income economies. est 3.4 percent pace (figure A1.1). 233 G L O B A L E C O N O M I C P R O S P E C T S forward by about 1 percent in dollar terms for Figure A1.1 Some slowing of growth 2002 as a whole, despite a strong rebound on into the first quarter of 2003 a quarter-to-quarter basis during much of Real GDP, percent change, year/year 2002. However, global semiconductor sales 7 peaked in the three months prior to November Thailand 6 ᮡ 2002, and trended lower in December and early 2003. These erratic developments in 5 Philippines ᮡ global and sectoral demand have had some im- 4 pact in slowing East Asian export growth in ᮡ 3 the first part of 2003. Exports for East Asia as Indonesia a whole started the year strongly, up 20 per- 2 cent over year ago (oya) in the first quarter in ᮡ 1 Malaysia dollar terms, but by April–May dollar export 0 growth rates had slowed to below 10 percent in the majority of cases, or even below 5 per- –1 Q1 Q3 Q1 Q3 Q1 cent in some. China and Thailand were the 2001 2001 2002 2002 2003 only countries where dollar exports continued Source: National agencies. to grow at rates of over 30 percent and 20 per- cent, respectively. The World Health Organization’s March Second-quarter data for China, at 6.7 per- alert about SARS sparked extraordinary pub- cent growth, shows the domestic effects of lic concern throughout East Asia and around SARS quite clearly. An increased deceleration the world. The number of cases worldwide was observed in the high-income or newly in- rose from very few in mid-March to 8,360 by dustrializing economies (NIEs) of the region. the end of May (most of which were in China GDP in the first quarter fell by 1–2 percent and Hong Kong, with smaller numbers in Tai- in Korea and Hong Kong (at a quarter-on- wan, China, and Singapore), and then leveled quarter seasonally adjusted annual rate), off at around 8,465 by mid-June, indicating while it rose at only 0.7 percent in Singapore that the outbreak had been brought under (on the same basis). In Korea, policy efforts to control, at least for the time being. Most of the restrict excessive growth in consumer borrow- economic impacts of the outbreak were the re- ing and the impact of security concerns with sult of public perceptions and fears about the regard to North Korea, contributed to a fall in disease, and from precautions taken against it, consumer spending and machinery and equip- rather than from the disease itself. The princi- ment investment. Exports also slowed from pal way SARS appears to spread is through the end of 2002, though they increased droplet transmission, and so worst affected strongly year-on-year in dollar terms. The were service industries, which depend on face- weaker trends in GDP growth and trade for to-face interaction between service providers the NIEs continued into the second quarter. and customers, especially tourism, and related Among the factors contributing to slower sectors such as restaurants and hotels, retail East Asian growth in 2003 was the downturn sales, business travel, and transportation. in growth in most parts of the developed world April tourist arrivals in Hong Kong and Sin- in the last quarter of 2002 and early 2003. The gapore were down by 65–70 percent from year- still hesitant and uncertain pace of recovery in earlier levels, for example, while passenger traf- the global high-tech industry, to which East fic in all Asia-Pacific air carriers fell by 45 Asia is a key supplier, has been another factor percent in that month. With fears about SARS affecting exports. World semiconductor sales, being at their height in April and May, before which had slumped 31 percent in 2001, inched easing substantially in June, the economic im- 234 R E G I O N A L E C O N O M I C P R O S P E C T S pacts will likely be concentrated in the second agricultural primary commodities such as rice, quarter, especially in economies with large rubber, palm oil, coconut products, and lum- tourism sectors, such as Hong Kong (China), ber from late 2001 on. International capital Singapore, Thailand, and Malaysia. As noted, markets had been afflicted by high levels of there was a significant impact in China, which volatility and risk aversion through much of had the largest number of SARS cases, though 2002, but were showing intriguing signs of growth there is still expected to reach 7–8 per- a turnaround late in 2002 and early 2003. cent for the year as a whole, because of the very Spreads on emerging market and high-yield strong momentum of the economy going into corporate debt fell sharply, while flows to 2003, and the relatively small role of tourism. emerging market funds began to increase. Of course many East Asian countries already had Short-run outlook achieved low spreads because of the dramatic Developing East Asian growth is expected to increase in their foreign reserves and the large dip to around 6.1 percent in 2003 from 6.7 fall in their net foreign indebtedness in recent percent in 2002, before picking up pace in years. However, the improvement in general 2004 and later years. That is, the impacts of emerging market sentiment will be good news SARS and the global slowdown are expected for countries such as the Philippines and In- to be modest. Several factors should support donesia, which tend to face higher spreads be- regional activity in the near term. For one cause of their more precarious fiscal positions. thing, China is an increasingly important mar- At the level of domestic policy, low infla- ket for other East Asian countries, and contin- tion, a shift to greater exchange rate flexibility, ued growth there should provide some sup- and a marked improvement in external balance port for other East Asian economies’ exports, sheets have all enhanced the ability of central despite the disruption caused by SARS. Chi- banks to implement supportive monetary poli- nese imports were still rising at a 30–40 per- cies in many East Asian countries. Real interest cent year-on-year dollar rate in April-May. rates have fallen to historically low levels in re- Exports in several Southeast Asian coun- cent years and this should provide an environ- tries have been boosted by higher prices for ment conducive to growth. Most regional stock Table A1.1 East Asia and Pacific forecast summary Growth rates/ratios (percent) 1991–2000 2001 2002 2003 2004 2005 2006–15 Real GDP growth 7.7 5.5 6.7 6.1 6.7 6.6 6.2 Consumption per capita 5.5 4.1 6.1 5.3 6.5 6.5 5.8 GDP per capita 6.4 4.5 5.8 5.2 5.8 5.7 5.4 population 1.2 0.9 0.9 0.9 0.9 0.8 0.8 Gross Domestic Investment/GDPa 28.8 30.5 33.0 33.7 34.4 35.2 30.4 Inflationb 6.8 2.6 3.3 3.7 4.6 2.8 General gvt. budget balance/GDP –0.9 –3.3 –3.4 –3.4 –3.2 –2.9 Export Market Growthc 8.3 –2.2 3.9 6.9 8.2 8.1 Export volumed 11.5 2.7 15.7 14.6 13.7 11.3 Terms of trade/GDPe –0.1 0.1 –0.1 0.6 –0.7 –1.1 Current account/GDP 0.4 2.7 3.2 1.9 2.1 1.2 Memorandum items East Asia excluding China 4.6 2.3 4.4 3.9 5.0 5.4 4.9 a. Fixed investment, measured in real terms. b. Local currency GDP deflator, median. c. Weighted average growth of import demand in export markets. d. Goods and non-factor services. e. Change in terms of trade, measured as a proportion of GDP (percentage). Source: World Bank baseline forecast July 2003. 235 G L O B A L E C O N O M I C P R O S P E C T S markets rebounded quite sharply between late down of overcapacity, and continued macro- March (precisely when the SARS crisis was economic stability. And structural and institu- starting) and June, indicating that financial tional reform efforts to improve the investment markets were looking beyond the near-term climate will also be important. These include disruptions while currencies in most flexible reducing barriers to foreign direct investment exchange rate economies have also appreciated in services; strengthening infrastructure and against the dollar. In several countries govern- the provision of other public services impor- ments also felt able to undertake small pro- tant for business; and improving the regu- grams of fiscal stimulus to bolster economic ac- latory, legal, and judicial frameworks. Also tivity during the temporary SARS shock, in important will be continued financial and particular in Malaysia, Hong Kong, Singapore, corporate sector restructuring and reform, in- and Korea. cluding reforms to improve financial supervi- sion and regulation, and strengthen corporate Long-term outlook governance. The years since the 1997 financial crisis While some governments have undertaken have been ones of extraordinary volatility and counter-cyclical fiscal policies in 2003, given uncertainty in the world economy. East Asian the significant public debt built up after the economies have actually come through this pe- crisis, policy attention in the medium term is riod reasonably well. Simple average growth in becoming more focused on the need for fiscal the five crisis countries was 4.6 percent in consolidation, and also, to some extent, on 1999–2002, while including China and Viet- better addressing the risks of implicit or con- nam it reached 5.1 percent. Contributing to tingent liabilities. Looking forward, there is this reasonably positive experience has been a scope to focus on better public administration broad array of efforts at policy reform—albeit and financial accountability, improve public often gradual and incomplete—accompanied service delivery, and address questions of by a gradual strengthening of domestic de- governance more broadly—that is, to provide mand. Robust household consumer spending growing volumes of critical public goods helped underpin growth during 2001’s export while maintaining sound fiscal positions. De- slowdown, and it also continued to bolster the veloping East Asia is expected to achieve per regional recovery through much of 2002. Ef- capita growth of close to 5.5 percent in the forts to recapitalize and restructure the finan- medium to longer term, given continued cial sector have been successful enough for steady efforts to improve structural policies banks to foster the emergence of new con- and the quality of institutions. sumer credit markets, a positive development from a long-run standpoint, although bank management and regulators will need to ensure South Asia it does not become a source of vulnerability. Investment spending has been strong in Recent developments G China and Vietnam, but in general it has re- DP GROWTH in the South Asia region mained erratic and, on the whole, still rela- declined to 4.2 percent in 2002 from tively weak in the crisis countries. But even 4.9 percent in 2001, a downward revi- here, there were some signs of an emerging sion from our previous estimate of an ac- pickup in 2002, for example in housing con- celeration of growth as published in Global struction. A number of factors should support Development Finance—2003. The slowdown a stronger investment revival in due course, in- largely reflects adverse weather conditions cluding continued domestic economic growth, and a decline in agricultural output in India, improvement in corporate profitability and re- Nepal, and near stagnation in Bangladesh. duction in corporate indebtedness, the running Additionally, Nepal experienced a plunge in 236 R E G I O N A L E C O N O M I C P R O S P E C T S of remittances is a reflection of incentives in- Figure A1.2 Industrial production in troduced by the government to channel remit- selected South Asian countries tances through official sources. In Pakistan 3-month moving average, percentage change, year/year and Sri Lanka, increases in remittances are 20 Pakistan largely attributed to the improvements in their 15 ᮡ domestic security situations and to progress India in macroeconomic stabilization. High interest 10 ᮡ rate differentials in India may have con- 5 tributed to a rise in banking transfers there. While net FDI inflows to Pakistan rose 0 markedly in 2002, attracted by increased macroeconomic stabilization and progress in –5 ᮡ reforms, net inflows to India declined. In Bangladesh –10 Bangladesh, FDI inflows fell by over 60 per- cent, mainly reflecting the determination of an –15 absence of export markets (gas exploration) Ju 998 98 Ju 99 99 Ju 00 00 Ju 01 01 Ju 02 02 03 and the lack of progress in reforms of the in- 19 19 19 20 20 20 20 20 20 20 1 n. ly n. ly n. ly n. ly n. ly n. Ja Ja Ja Ja Ja Ja frastructure sector, such as in ports, power, Source: National agencies. and telecommunications. India’s and Pakistan’s nominal exchange rates appreciated relative to the U.S. dollar tourism receipts and a sharp decline in manu- during 2002, while the exchange rates of facturing output, as the domestic insurgency Bangladesh and Sri Lanka remained relatively intensified. Pakistan and Sri Lanka both en- flat. Nepal’s exchange rate is pegged to the In- joyed a rise in growth rates in 2002 over dian rupee. Throughout much of South Asia, 2001, because of strong government con- inflation remained broadly stable, as feed- sumption in Pakistan and a recovery in the through effects of higher oil prices were offset services sector—along with improved political by easing price pressures tied to generally stability tied to progress on peace talks and a weaker domestic demand conditions. In Sri yearlong cease-fire—in Sri Lanka. Lanka, inflationary pressures were reduced Industrial production in the main econo- from 2001, but remained close to 10 percent mies continued to register gains of 5–10 per- on average in 2002. cent entering 2003 (figure A1.2). Driven largely There was some progress in the consolida- by a recovery in India’s exports, export vol- tion of persistent fiscal deficits, which are ume growth accelerated for the region on av- prevalent throughout the region (in large part erage—despite sluggish external demand— because of weak revenue collection). By con- thus reducing the region’s net trade deficit. taining outlays and raising revenues, Bangla- Current account balances in the two largest desh reduced its budget deficit to 4.7 percent economies, India and Pakistan, posted sur- of GDP in 2002 from just below 6 percent of pluses, and the region’s aggregate external bal- GDP in 2001. It improved its financing profile ance strengthened. But exports from Nepal by reducing its reliance on more expensive do- declined significantly, because of weak exter- mestic financial markets and increasing its re- nal demand and heightened competition. liance on external financing sources. Sri Lanka A number of regional economies experi- also achieved a two-point reduction in its bud- enced a significant increase in remittances get deficit, from 10.8 percent of GDP in 2001 during 2002 over 2001, notably Bangladesh, to 8.8 percent in 2002, and was able to shift India, Nepal, Pakistan, and Sri Lanka. In the composition of its financing to longer-term Bangladesh, the reported increase in inflows instruments. Pakistan brought its underlying 237 G L O B A L E C O N O M I C P R O S P E C T S budget deficit down slightly to 5.1 percent of with a forecasted acceleration of private con- GDP. India’s general government fiscal deficit sumption growth, is projected to support was little changed from 11 percent of GDP (in- growth in 2003. With higher domestic de- cluding both central and state deficits), despite mand, import volume growth is forecast to ac- increased revenue collection. celerate on aggregate for the region in 2003, leading to a narrowing of the current account Near-term outlook surplus. Growth is forecast to accelerate throughout the region in 2003, up to an average of 5.4 Medium-term prospects percent, assuming a return to trend agricul- South Asian growth is expected to main- tural production, a recovery in external de- tain an average of close to 5.4 percent over mand, and continued improvements in politi- the medium term, assuming normal weather cal stability and regional security. Domestic conditions—leading to recovery and accelera- demand, especially private consumption and tion of agricultural output—and a continued fixed investment, is expected to accelerate, recovery in external demand. Declining oil spurred by recovery in agricultural incomes. prices, both in nominal and real terms, should Growth in government spending is expected reduce pressures on current account balances to accelerate less strongly. A projection of for the region, which is overwhelmingly a net higher growth in India underpins this forecast, energy importer. India should benefit from a as it represents nearly 80 percent of the re- recovery in domestic demand (particularly in gion’s aggregate GDP. Aside from a recovery the manufacturing sector) and firming export in agricultural output, growth in India is likely volume growth. Bangladesh is expected to to be supported by continued strong expan- witness strengthening domestic demand and sion in the services sector especially in infor- recovering exports, and on the supply side, a mation technologies now burgeoning in the strong performance of SMEs and export-ori- Bangalore area. A continued recovery in Pak- ented manufacturing units (provided proposed istan’s gross fixed investment rates, coupled structural reforms are carried out). Both Pak- Table A1.2 South Asia forecast summary Growth rates/ratios (percent) 1991–2000 2001 2002 2003 2004 2005 2006–15 Real GDP growth 5.2 4.9 4.2 5.4 5.4 5.4 5.4 Consumption per capita 2.0 3.5 1.5 2.9 3.3 3.5 3.3 GDP per capita 3.3 3.1 2.5 3.7 3.7 3.8 4.1 population 1.9 1.7 1.7 1.6 1.6 1.5 1.3 Gross Domestic Investment/GDPa 21.3 22.0 22.8 22.7 22.5 22.3 25.0 Inflationb 7.8 3.0 1.8 5.1 5.8 4.7 General gvt. budget balance/GDP –10.5 –8.3 –9.8 –9.7 –9.3 –8.5 Export Market Growthc 7.7 0.2 2.9 6.0 7.5 7.3 Export volumed 11.5 9.1 3.5 5.9 7.4 8.0 Terms of trade/GDPe –0.1 0.0 0.5 0.7 0.4 0.1 Current account/GDP –1.5 –0.5 0.5 0.4 0.2 –0.1 Memorandum items GDP growth: South Asia excluding India 4.4 3.2 3.6 4.8 5.3 5.6 5.2 a. Fixed investment, measured in real terms. b. Local currency GDP deflator, median. c. Weighted average growth of import demand in export markets. d. Goods and non-factor services. e. Change in terms of trade, measured as a proportion of GDP (percentage). Source: World Bank baseline forecast July 2003. 238 R E G I O N A L E C O N O M I C P R O S P E C T S istan and Sri Lanka are projected to benefit underlying assumptions, not least of which is a from continued macroeconomic stability and larger contribution to growth by the private an associated acceleration of growth. The sector. This in turn reflects the expectation of peace process is expected to yield significant progress with fiscal consolidation and contin- economic gains in Sri Lanka. Similarly, Nepal ued structural reforms, including reforms in is likely to experience strengthening growth, trade, banking, privatization, and infrastruc- assuming continued improvement in the secu- ture. These factors, combined with the im- rity situation there, with a recovery in domes- provement in human capital indicators in re- tic demand, exports, and in tourism receipts. cent years—such as rising literacy rates and Further, recent steps toward improving rela- school enrollments and declining infant mor- tions between India and Pakistan may lead to tality rates—will lead to an increase in produc- greater stability in the sub-region, paving the tivity. Despite a projection of declining infant way for increased business confidence and sta- mortality rates, overall the South Asian popu- bility. Throughout the region, growth should lation growth rate is projected to decelerate as be underpinned by continued firm expansion birth rates are expected to decline at faster in services and industrial production. rates. Lower population growth in the coming Recovery in external demand and a gradual decade, along with the forecast growth rates, return to lower oil prices is likely to be more implies that per capita GDP growth will be than offset by generally firming import de- close to 4.0 percent per year. mand, which is expected to lead to a moderate decrease in the region’s aggregate current ac- Risks count surplus to an average balance near zero There are a number of risks to the forecast. as a share of GDP over the medium term Persistent fiscal deficits continue to be a risk (2004–05). At the individual country level, in a number of the region’s economies, India India and Pakistan’s projected current account in particular, as they can undermine fiscal sus- surpluses are expected to roughly balance the tainability, contribute to a growing debt-to- projected deficits in Bangladesh, Nepal, and GDP ratio, and lead to higher interest rates, Sri Lanka. thereby crowding out private investment and The fiscal positions of the South Asian diverging public outlays from investment to economies are forecast to improve moderately, interest payments and limiting the scope for assuming some progress in raising budget rev- both fiscal and monetary policies. Fiscal con- enues and in the management of government solidation is required, which would not only expenditures. Inflation is projected to increase mitigate such vulnerabilities, but also provide somewhat, albeit still at generally moderate for broader scope of action in macroeconomic levels, because of the assumed pick-up in policies to pursue sustained higher growth. growth and assumptions of a more accom- The forthcoming phase-out of the Multi-Fibre modative monetary stance in a number of Arrangement (MFA) in 2005 will imply greater countries, though falling oil prices are ex- competition for the region’s textile exporters. pected to partially offset these factors. While India and Pakistan appear to be gearing up for the impending increase in competition, Long-term outlook the impact of the MFA phase-out on other Long-term growth in South Asia is forecast South Asian regional economies is more un- to average about 5.4 percent, in line with the certain, particularly in Bangladesh, Nepal, GEP 2003 growth forecast. This forecast is and Sri Lanka (where garment exports repre- somewhat higher than the 5.2 percent average sent about 75 percent, 25 percent, and just real growth posted during the 1990s. The over 50 percent of total merchandise exports, higher projected growth over the coming respectively). Given the importance of the decade through 2015 reflects a number of agricultural sector to the region, the threat of 239 G L O B A L E C O N O M I C P R O S P E C T S severe weather conditions and associated poor macropolicies, and this has rapidly thwarted harvests remain a significant risk to growth negative expectations for the future. outcomes. Political risks and uncertainties A second domestic factor explaining the re- also remain a concern, because of both inter- gion’s enhanced resilience to crisis is found in nal and external factors. Heightened domestic the substitution of inward-oriented develop- and regional instability could undermine ment policies (through the maintenance of growth prospects and slow the pace of eco- high trade barriers and other perverse price in- nomic reforms. Remittances could be affected centives) by more liberal trade and market- by increased instability in the Middle East. friendly policies. This shift has helped the And significantly higher-than-forecast energy economies to diversify and has broadened the prices would pose an additional burden on regional export base while diminishing its current account positions. dependency on a narrow set of commodity prices. As shown in figure A1.3a, during the 1970s, exports of agricultural products, oil Latin America and the Caribbean and other natural resources for the region as a whole, accounted for about one-quarter and Recent developments one-fifth, respectively, of total exports; two T he Latin American region has begun to decades later, both sectors accounted for recover from last year’s recession. The about one-tenth of total exports, and manu- upturn this year reflects the tentative re- facturing has become an important source of covery in Argentina and Uruguay as well as foreign exchange. calming of pre-election jitters in Brazil at the Additional important elements contribut- end of 2002. As a result, regional growth in ing to the success of this outward-oriented 2003 is projected to reach 1.8 percent, com- development strategy have been the NAFTA pared to last year’s contraction of 0.8 percent. agreement—that directly benefited Mexico, Although the region is on a favorable recovery but generated spillovers to other countries— path, its growth rate remains well below poten- and increased intraregional integration. In- tial and below that of other regions. In addition deed, the revamping of Mercosur and other to Argentina and Uruguay, Chile, Mexico, Co- sub-regional integration arrangements are lombia, and Brazil, countries with generally being discussed in Latin America, together stronger policy frameworks, registered a some- with the plans for a hemispheric free-trade what improved growth performance. area. Chile and Mexico suffered least from the The fact that the 2002 recession was rela- 2001–2 economic downturn, thanks to their tively short reflects both domestic structural effective integration in the global economy policy, renewed confidence, and external fac- and good macroeconomic policies. Chile re- tors. Domestic macropolicies have improved cently signed an FTA with the United States significantly: the region managed to reduce in- that is anticipated to be ratified by the two flation to single-digit figures (for 2003–Q1, governments by end-2003, and become opera- the regional inflation rate was 8.3 percent), tional in 2004: the FTA grants free access to proving that the commitment of central banks 87 percent of Chilean exports to U.S. markets. to low inflation is, although not universal, A final factor is the achievement of health- quite widespread across the region. Similarly, ier current account positions. The Latin Amer- balanced fiscal policies have been applied and ican region has witnessed a drop of more than the expected 2003 regional public deficit 3.5 points in the ratio of the current account should be about 2 percent of GDP. For ex- balance to GDP (from –4.4 percent in 1998 to ample, the new president of Brazil has reiter- –0.8 percent in 2002) resulting from a reduc- ated the country’s commitment to balanced tion of the region’s borrowing needs and the 240 R E G I O N A L E C O N O M I C P R O S P E C T S Figure A1.3a Sectoral export shares (selected countries) Agriculture Shares in percent of total exports 70 60 50 1970s 40 ᮡ 1980s 30 26 18 1990s 20 ᮡ 10 ᮡ 10 0 Latin America Argentina Brazil Mexico Chile Colombia Peru Venezuela and the Caribbean Oil and other natural resources Shares in percent of total exports 70 60 50 40 30 1980s 21 22 1970s 20 ᮡ 1990s 11 ᮡ ᮡ 10 0 Latin America Argentina Brazil Mexico Chile Colombia Peru Venezuela and the Caribbean Source: World Bank, DECPG estimates from GTAP 5 databases. reluctance of global investors to enter emerg- ous, net oil importers suffer less from oil price ing markets. The adoption of more flexible ex- surges, and usually non-oil commodity prices change rate regimes is probably an important surge with a falling dollar. Moreover, the weak- structural change. Anticipated increases in GDP ening of the dollar against the euro should help growth are not expected to directly translate manufacturing exporters with currencies fol- into large unsustainable current account lowing the dollar by raising the competitive- deficits. ness of their goods in the euro zone and en- Despite the sluggish global economy, some abling their share of exports to the EU to rise. specific improvements in the external environ- The recent dramatic reduction of spreads on ment are also contributing to recovery in the yields of foreign sovereign debt for the region region. A weaker U.S. dollar is the first of these is another favorable global financial develop- improvements: servicing the region’s dollar- ment. The reduction in spreads—highlighted in denominated external debt becomes less oner- figure A1.3b (in Brazil spreads narrowed from 241 G L O B A L E C O N O M I C P R O S P E C T S Figure A1.3b Latin American yield spreads EMBI Global, Jan. 2000 = 100 160 Brazil Dominican Republic ᮡ 140 ᮡ Latin America and the Caribbean 120 Chile ᮡ Mexico ᮡ Venezuela 100 ᮡ ᮡ Argentina ᮡ 80 60 40 02 02 02 02 02 02 02 02 2 2 2 2 03 03 03 03 03 03 00 00 00 00 20 20 20 20 20 20 20 20 20 20 20 20 20 20 .2 .2 .2 .2 n. b. ril ay ne ly g. n. b. ril ay ne . pt ct ov ec . ar ar Ju Ap Ap Ja Fe Au Ja Fe M O M Ju Se Ju N D M M Source: World Bank, DECPG estimates from J.P. Morgan-Chase data. a peak of 2,067 bp in October 2002 to 754 bp (Brazil and Colombia) and, with some excep- in April 2003, Mexican spreads have reached tions, all countries should experience a recov- an all-time low, those of Peru and Colombia ery of domestic demand and record positive are at their lowest levels since 1998 and 2000, rates of growth. The regional average growth respectively)—has not yet been accompanied rate should climb to 3.7 percent in 2004 and by large capital inflows. For Latin America, to 3.8 percent in 2005. Even if current account average monthly flows during the first half of deficits increase in most of the expanding 2003 registered $5.3 billion, modestly above economies of the region, low world interest the $4.3 billion average of the same period in rates, lower perceived risks for Latin Ameri- 2002. In fact the bond price rally may be tied can assets (bonds, equities, or direct invest- to an excess of global investors’ demand for ments), the continuation of fiscal prudence in higher yield instruments over the supply of the region, and flexible exchange rates should new issues by borrowers. However, it should maintain a cap on these incipient deficits. be mentioned that Mexico has been quite ac- Even though the 2003 –0.5 percent ratio of tive in international capital markets and that current account deficit to GDP for the region even Brazil has already been able to place is- is unlikely to be maintained, external deficit sues twice this year. ratios are not foreseen anywhere near the 1995–99 average of more than 3 percent, but Near-term outlook rather a modest –0.7 percent for 2004 and –1 Improving world trade growth coupled with percent for 2005. increased OECD economic growth led by the Clearly some uncertainty remains, given United States should further boost the export- that these projections assume that no (domes- led recovery of Latin America in 2004–05. tic or external) adverse development reverses Also, by 2004, the region’s largest economies the easing of financial pressure on the region’s will have surpassed the worst of their crises most vulnerable countries. High levels of debt (Argentina and Uruguay) or potential for crises still burden fiscal authorities and tight mone- 242 R E G I O N A L E C O N O M I C P R O S P E C T S tary policies required to keep inflation under as a source of imports, as a destination for ex- control may limit expansion. This is especially ports, and as a source of external finance. As true for Brazil; however, thanks to its en- shown in figure A1.3c, the United States has hanced credibility, the central bank in this become the region’s major export market des- country may be able to ease its stance in the tination, increasing from 36 percent of total near future. exports in the 1970s to almost 50 percent in Furthermore, some country risks persist. the 1990s. Several challenges await Argentina: a) restruc- Given the lackluster growth forecasts for turing the banking system to reactivate credit the Euro zone, a stronger link to the United (which is still contracting); b) replacing in- States may be considered positive; however, efficient taxes, including on export and sav- geographical diversification may reduce risks. ings, with new efficient taxes; c) building a In particular, European markets have lost im- solid regulatory and institutional framework portance for Latin American exporters and to protect property rights and deal with other this may be corrected by pushing for more bi- governance issues; d) renegotiating public util- lateral agreements and a more comprehen- ities tariffs that have been held well below the sive lift of European restrictions to market rate of inflation; and e) eventual restructuring access. of the defaulted debt. The Republica Bolivari- The rising share of intraregional trade is a ana de Venezuela started to export oil again, positive sign and may help Latin American but its political crisis is not over and macro- countries negotiate trade and other integra- economic instability, price controls, and other tion agreements as a block, vis-à-vis OECD distortions need to be addressed. Recently, the countries. In fact, trade integration should be government endorsed an Economic Stability a priority among the long-term development agreement that pledges to make every effort to policies—Latin America, compared to East improve its external accounts by reestablish- Asia, shows much lower trade-to-GDP ra- tios—and deeper trade integration agreements ing the level of oil income and implementing bring additional benefits in terms of increased the necessary adjustments to the currency con- foreign investment and also boost credibility trol regime; however, previous similar at- to sound macro policies. tempts have failed. On the internal front, some of the major The Caribbean countries, and this also ap- impediments of the past, such as large fiscal plies to some small Central American coun- imbalances and perverse price incentives, have tries, are still facing the difficult transition been removed (though not consistently in all from being tropical agriculture export-oriented countries). However, smaller fiscal deficits economies to becoming more diversified. Their have been obtained mainly through expendi- preferential trade agreements are expiring or ture restraints, with potential long-term issues are severely eroded, and their successful at- on infrastructure development and poverty tempts at diversifying toward tourism and fi- eradication, and most countries still have rela- nancial services received a severe set-back in tively (with respect to more developed coun- 2002 from which they have not yet recovered. tries) low tax revenues to GDP ratios. Low tax revenues result from a strong reduction of in- Long-term prospects ternational trade duties (in itself a positive de- In the longer term, Latin American countries velopment), inefficient collection, evasion, and could achieve higher growth rates if they over- other governance issues. Low revenues have come several critical structural constraints been compensated by recurring to unstable (table A1.3). On the external front, a number non-tax revenues (oil royalties or other nat- of countries rely heavily on the United States ural resources form of taxation, privatiza- 243 G L O B A L E C O N O M I C P R O S P E C T S Figure A1.3c Export shares by destination market (selected countries) United States Shares in percent of total exports 80 70 60 48 1980s 50 44 40 36 1970s ᮡ 1990s 30 ᮡ ᮡ 20 10 0 Latin America Argentina Brazil Mexico Chile Colombia Venezuela and the Caribbean European Union Shares in percent of total exports 1970s 50 1980s ᮡ 45 40 1990s ᮡ 35 29 30 25 ᮡ 25 20 15 15 10 5 0 Latin America Argentina Brazil Mexico Chile Colombia Venezuela and the Caribbean Intra-Regional Shares in percent of total exports 50 45 40 35 1970s 1980s 30 ᮡ 1990s 25 20 ᮡ 20 16 ᮡ 15 12 10 5 0 Latin America Argentina Brazil Mexico Chile Colombia Venezuela and the Caribbean Source: World Bank, DECPG estimates from GTAP 5 databases. tion), generating a situation of inadequate and skewed income distribution, and develop ma- volatile government income. ture financial markets necessary to generate In the long term, beyond macroeconomic enough resources to become less dependent on stability and commitment to sound fiscal and foreign finance, allowing important invest- monetary policies, LAC countries will have to ments in physical infrastructure and human tackle governance issues, attempt to correct a capital to be financed domestically. 244 R E G I O N A L E C O N O M I C P R O S P E C T S Table A1.3 Latin America and the Caribbean forecast summary Growth rates/ratios (percent) 1991–2000 2001 2002 2003 2004 2005 2006–15 Real GDP growth 3.4 0.3 –0.8 1.8 3.7 3.8 3.8 Consumption per capita 2.4 –0.9 –3.5 –0.1 1.8 1.9 2.3 GDP per capita 1.7 –1.2 –2.3 0.4 2.3 2.5 2.5 population 1.7 1.6 1.5 1.4 1.4 1.3 1.2 Gross Domestic Investmenta 19.8 19.1 18.0 17.6 18.4 18.4 22.6 Inflationb 12.0 5.5 4.7 4.1 4.0 4.0 General gvt. budget balance/GDP –3.0 –1.8 –2.9 –2.0 –1.0 –0.6 Export Market Growthc 9.4 –1.2 0.5 5.0 8.6 7.2 Export volumed 8.7 1.0 2.2 9.2 11.2 10.2 Terms of trade/GDPe 1.7 –0.2 0.1 –0.4 0.1 –0.7 Current account/GDP –2.7 –2.7 –0.8 –0.5 –0.7 –1.0 Memorandum items GDP growth: LAC excluding Argentina 3.2 1.2 1.0 1.5 3.6 3.9 Central America 4.4 1.5 1.9 2.4 3.1 3.8 Caribbean 4.0 3.1 3.0 0.9 2.4 4.1 a. Fixed investment, measured in real terms. b. Local currency GDP deflator, median. c. Weighted average growth of import demand in export markets. d. Goods and non-factor services. e. Change in terms of trade, measured as a proportion of GDP (percentage). Source: World Bank baseline forecast July 2003. Eastern Europe and Central Asia Growth in the Central and Eastern Euro- pean Countries (CEECs), excluding Turkey, Recent developments was unchanged in 2002 relative to 2001 at 2.9 G estimated to have expanded DP IS NOW percent. Including Turkey, GDP growth aver- by 4.6 percent in 2002 for ECA, an aged 4.5 percent for the group, swinging up upward revision from the 4.1 percent sharply from a 0.8 percent contraction posted projected in the last forecast presented in in 2001. Regional growth was underpinned by Global Development Finance 2003 (World expanding domestic demand, often spurred Bank 2003). Growth prospects proved to be by fiscal policy (Hungary, Czech Republic, more resilient than previously anticipated, pri- Poland, Slovenia, Slovakia) and/or easing marily because of the strength of domestic de- monetary policies (Czech Republic, Latvia, mand, which was more than able to offset Lithuania, Romania). Furthermore, despite lackluster growth in the region’s main export tepid external demand, export growth re- markets. The firming of ECA regional eco- mained firm in many of the CEECs. nomic growth in 2002—over the 2.2 percent In the Commonwealth of Independent growth posted in 2001—was driven by the States (CIS), GDP growth continued to slow to huge 15 percentage point swing in Turkey’s an average of 4.7 percent in 2002, down from performance (from a contraction of 7.4 per- 5.8 percent in 2001, and following the spike in cent in 2001, following the financial crisis, to CIS average growth of 8.4 percent posted in an upswing of 7.8 percent in 2002). The ECA 2000. The slowdown largely reflects a deceler- average growth excluding Turkey registered ation of growth in Russia, as the effects of the a slowing to 3.9 percent in 2002, contrasted devaluation of the 1998 crisis and the rents with 4.5 percent in 2001. This latter trend from high energy prices eroded. Diminished largely reflected the aggregate slowdown in import demand from Russia—representing an the CIS region. important export market for the remaining CIS 245 G L O B A L E C O N O M I C P R O S P E C T S countries—contributed to the easing of growth in the rest of the region. In addition, growth Figure A1.4 Industrial production in decelerated in Turkmenistan, because of a poor selected ECA countries cotton harvest and slower growth of natural 3-month moving average, percentage change, year/year gas exports (the result of pipeline constraints), 30 Turkey [right scale] 60 and suffered a decline in Kyrgyz Republic, tied ᮡ 45 to an accident at its largest gold mine and 20 a temporary decline in exports. The South 30 Caucasus countries (Armenia, Azerbaijan, and 10 Georgia) and Belarus, however, experienced an Poland ᮡ 15 acceleration of growth. 0 0 Near-term outlook –10 –15 ECA aggregate growth is forecast to slow ᮡ moderately to 4.3 percent in 2003. This pro- Russia –20 –30 jected growth rate for 2003 is higher than an- 01 01 1 2 2 02 2 03 03 00 0 0 00 20 20 20 20 20 20 20 ticipated in GDF 2003, and reflects stronger .2 .2 ay g. b. ay g. b. ay ov ov Au Fe Au Fe M M M N N than expected growth in a number of coun- tries, particularly in Russia, which has exhib- Source: National agencies. ited firming domestic demand during the first quarter of 2003 (figure A1.4). The decelera- tion of ECA aggregate growth between 2002 particular, growth is projected to accelerate and 2003 is primarily because of a projected moderately in Albania, the Czech Republic, moderation of growth in Turkey following the Poland (which represents 13 percent of the re- sharp upswing in 2002. gion’s GDP), and Slovenia. Growth in the re- Growth in Turkey is projected to decelerate maining economies is forecast to either remain largely because of base effects following the flat or decelerate moderately. strong recovery in 2002. Other factors affecting In the CIS, growth is projected to Turkey’s near-term outlook include the contin- strengthen in 2003, as domestic demand has ued required fiscal consolidation, the expected begun to accelerate in Russia, underpinning slowdown in inventory building, weaker tour- growth there, which in turn should support ism revenues because of the Iraq conflict, and growth in other CIS countries dependent on limited foreign investment. The current account Russia’s import demand. First-quarter data for deficit has been rising rapidly, driven by rising 2003 show strong growth in energy exports imports, the recovery in domestic demand, and and industrial activity in Russia, spurring higher oil prices. Export growth has remained stronger investment, especially in the energy relatively robust despite the recent real appreci- sector. This, coupled with an increase in pri- ation of the lira and weak external demand. As vate consumption—boosted by strong growth of end-April, the Turkish government appeared in real incomes and falling unemployment—is on track for the Fund’s program, but the pace leading to higher output. Further, the recent of reforms will need to accelerate to sustain appreciation of the euro against the dollar has growth. led to increased import prices for Russian im- In the CEECs, excluding Turkey, growth in ports from Europe, relative to US dollar- 2003 is projected to accelerate moderately (by denominated oil export revenues, which in turn 0.5 percentage points) because of continued is stimulating increased demand for cheaper penetration in new export markets and an ex- domestic products. While oil prices have de- pected boost to consumer confidence because clined in recent months, they are still high rel- of progress in the EU accession process.1 In ative to the average over the last few years and 246 R E G I O N A L E C O N O M I C P R O S P E C T S are projected to average $26.5/bbl in 2003 2005—from $26.5/bbl in 2003 to $22/bbl and (given the spike during the first quarter), up $20/bbl in 2004 and 2005, respectively—and a from $24.9/bbl in 2002 (and well above the corresponding decline in the growth impetus $21/bbl reference price used for the Russian through fiscal linkages. budget). The southern-tier energy exporters of Kazakhstan and Azerbaijan are expected to Long-term prospects continue to post high growth rates, driven by Higher investment rates and ongoing restruc- the ongoing oil sector investment boom, and turing of the capital base are expected to con- supported by strong FDI inflows. Manufactur- tribute to stronger growth in the CEE countries ing and services related to investment in the oil during the second decade of transition than sector are expected to continue to expand as posted during the previous decade. Further, well. GDP growth in most other CIS econo- continued improvements in the policy environ- mies is anticipated to either remain flat or de- ment, including greater macroeconomic stabil- celerate in 2003, with the exceptions of Geor- ity, are expected to underpin the projected gia, where a rise in investment is projected, higher growth rates. The EU accession process linked to the construction of oil and gas tran- and coming membership will continue to act as sit pipelines, and the Kyrgyz Republic, which an anchor for structural reforms and will help is anticipated to benefit from a revival in its attract significant inflows of FDI. While struc- gold production. tural reforms are being pursued in many CIS countries, in general, implementation is not as Medium-term prospects advanced or as widespread as in the CEE sub- ECA regional growth is expected to first region’s economies, and in some cases there is accelerate to 4.5 percent in 2004, and then to significant resistance to structural reforms. decelerate to 4.1 percent in 2005, reflecting di- This implies lower long-run growth in com- vergent trends at the sub-regional levels: accel- parison. The recent boom in hydrocarbon erating growth in the CEECs and slowing ac- rents has provided an impetus to growth, facil- tivity in the CIS. itating the introduction of a number of reforms Growth in the CEECs (including Turkey) is to oil-exporting countries, and contributing to projected to accelerate from 3.5 percent in an increase in investment outlays (particularly 2003 to 4.3 and 4.7 percent in 2004 and 2005, in the energy sector). However, given the respectively, in part because of a gradual volatility of energy market prices, these buildup in external demand. The first round of economies will not be able to sustain recently new EU members, in particular, is expected to achieved higher growth rates until diversifica- continue to receive significant inflows of FDI tion from energy becomes more broadly based. (in addition to EU transfers)—which will re- Given the degree of energy dependence in main an important source of external finance many of the CIS economies, particularly Rus- and support for long-term growth. Sustained sia, the projected softening of oil prices—to an growth in Turkey assumes the country will re- average nominal price of about $19 per barrel main committed to the Fund’s program, and for the 2005–10 period, in the underlying fore- that structural reforms will contribute to a cor- cast—implies a ratcheting down of the sub- rection in internal and external balances. These region’s growth from recent high rates. factors, along with the assumption of declining interest rates, are expected to help spur domes- Risks tic demand in Turkey. There are three main risks to the forecast: Growth is expected to slow in the CIS from 5.3 percent in 2003 to 4.6 and 3.4 percent in • Global trade and growth prospects: 2004 and 2005, respectively, assuming signifi- More sluggish than anticipated world cant decline in oil price in both 2004 and growth prospects, and/or a delayed re- 247 G L O B A L E C O N O M I C P R O S P E C T S Table A1.4 Europe and Central Asia forecast summary Growth rates/ratios (percent) 1991–2000 2001 2002 2003 2004 2005 2006–15 Real GDP growth –1.6 2.2 4.6 4.3 4.5 4.1 3.4 Consumption per capita 0.0 3.2 5.4 5.1 4.8 4.5 3.1 GDP per capita –1.8 2.0 4.5 4.3 4.4 4.0 3.3 Population 0.2 0.1 0.1 0.1 0.1 0.1 0.1 Gross Domestic Investment/GDPa 24.6 21.7 20.8 21.1 21.5 22.0 28.8 Inflationb 76.0 6.1 4.0 5.8 6.5 2.7 General gvt. budget balance/GDP –4.0 –8.4 –9.3 –8.4 –7.1 –6.2 Export Market Growthc 6.6 4.6 2.4 8.2 7.2 7.4 Export volumed 10.0 5.5 6.7 8.1 8.5 8.4 Terms of trade/GDPe 0.8 0.9 –1.6 1.7 0.0 –0.3 Current account/GDP –2.5 –1.4 1.1 0.6 0.1 –0.5 Memorandum items GDP growth: transition countriesf –2.5 4.5 3.9 4.5 4.4 3.8 Central and Eastern Europef 0.6 2.9 2.9 3.4 4.2 4.4 CIS –4.4 5.8 4.7 5.3 4.6 3.4 a. Fixed investment, measured in real terms. b. Local currency GDP deflator, median. c. Weighted average growth of import demand in export markets. d. Goods and non-factor services. e. Change in terms of trade, measured as a proportion of GDP (percentage). f. Excluding Turkey. Source: World Bank baseline forecast July 2003. covery, especially in the EU, could under- dermine confidence of foreign investors mine or reduce the export led component and result in difficulties in maintaining ac- of growth, especially for the CEECs; cess to financing; • Domestic policies and investor confi- • Energy prices: A sharper decline in oil dence: Delays in fiscal consolidation in price (induced, for example, by low world countries with large budget deficits growth outcomes), could translate into a (Turkey, Hungary, Poland, Czech Repub- marked deceleration in growth for both lic), which would risk diminished use CIS energy exporters and countries de- of automatic stabilizers; skewed fiscal and pendent on Russia’s consumer markets. monetary policies (as witnessed in Poland, For the ECA region’s energy exporters, ef- for example); crowding out of private in- fective management of large government vestment; and a slowdown of structural oil revenues and continued structural re- reforms. In Turkey, failure to achieve sub- forms are required to pave the way for stantial decline in real interest rates would sustained long-term growth, economic di- result in significantly lower growth out- versification, and employment creation. turns. For the EU accession candidate Medium- and long-term prospects de- countries with large fiscal deficits, overall pend largely on rapid diversification of fiscal consolidation as well as public ex- the production base and exports. penditure restructuring will be necessary to join the European exchange rate mech- anism and to absorb EU transfers, which Sub-Saharan Africa require national co-financing; the growth outlook will also partly depend on a re- Recent developments A covery in external demand, which could SUBDUED EXTERNAL environment, to- cushion the adjustment process. Coun- gether with poor weather and home- tries with large twin fiscal and external grown problems of governance and deficits (Turkey, Croatia, etc.) could un- civil strife, held real growth in Sub-Saharan 248 R E G I O N A L E C O N O M I C P R O S P E C T S Africa (SSA) to 2.8 percent in 2002, down per up 13 percent, cotton up 24 percent, and from 3.2 percent in 2001. Faced with Europe’s cocoa up a sharp 107 percent. Year over year, faltering economy and rising geopolitical un- agricultural export prices gained an average of certainty, real export growth slumped to just 20.5 percent and though metals and minerals 0.7 percent, the worst outcome in a decade, declined a further 2.3 percent, the export- while net exports contributed –0.8 percent to weighted average price of non-energy com- GDP growth. Meanwhile, domestic absorp- modities for the region was up 13.5 percent, tion was flat at 3.6 percent. Notably, invest- while non-oil exporters’ terms of trade strength- ment spending was relatively resilient, espe- ened by 4.2 percent. Moreover, higher fre- cially in South Africa and a number of oil quency data on metals indicate signs of recent exporters. Early indications from the first half strength as well. No sustained upward trend is of 2003 are that performance will be similar in anticipated given highly competitive new sup- the current year and growth in SSA is expected pliers coming on stream, but at least key export to remain at 2.8–3 percent. markets appear to be stabilizing around present In domestic economies, adverse weather levels (figure A1.5a). compounded by civil strife has seriously dis- Tourism has been affected not only by rupted food production for more than half the weak income growth in the OECD, especially region’s population, and as many as 40 mil- Europe, but also by security concerns arising lion persons are facing acute hunger.2 Drought from the September 2001 terrorist attacks and has been particularly severe in the horn of the run-up to the Iraq war. If the rest of the Africa, comprising Ethiopia and Eritrea and world needed a reminder that travel to the re- parts of the Sudan, leaving over 15 million ur- gion is risky, it came in the form of a terrorist gently in need of food aid. But disruptions attack in Kenya in November 2002 that left have occurred in numerous other countries as fifteen dead, including three Israeli tourists. well because of weather or civil strife, includ- Barring further negative shocks, the low point ing Angola, Burundi, Democratic Republic of for tourism was likely reached at the time of Congo, the Gambia, Malawi, Mauritania, the Iraq war when travel to SSA was down by Senegal, and Zimbabwe. Apart from the hu- nearly one-quarter compared to the year be- manitarian crises, with agriculture represent- fore. Nevertheless, that virtually guarantees a ing nearly one-fifth of GDP—over one-quarter excluding South Africa—the macroeconomic impact is to reduce household incomes and Figure A1.5a SSA commodity price expenditure. On average, consumption grew outlook favors non-oil exporters by just 2.4 percent in 2002 (0.1 percent per Prices in US$, indices 2002 = 100 capita), down from 3.1 percent in 2001. 115.0 For the region as a whole, the terms of trade Metals and minerals strengthened in 2002, thanks both to gains in ᮡ ᮡ export prices and a decline in the cost of man- 105.0 Agriculture ufactured imports. Oil prices were up 2.4 per- cent over 2001 which, with net energy exports accounting for some 8 percent of GDP, con- 95.0 tributed 0.2 percent to incomes. Non-energy ᮡ commodities are also enjoying a significant re- Energy 85.0 bound, albeit from very low levels after the dizzying declines of the late 1990s. This re- bound is primarily the result of supply con- 75.0 straints rather than growth in demand. Com- 2002 2003 2004 2005 pared to low points reached since 2000, the Source: DECPG Commodities Group. price of gold in 2002 was up 14 percent, cop- 249 G L O B A L E C O N O M I C P R O S P E C T S mediocre year at best and the World Travel been relatively few beneficiaries—Nigeria, and Tourism Council (WTTC) forecasts a South Africa, Gabon, Lesotho, and Kenya ac- slight decline of 0.2 percent in GDP and em- count for nearly 93 percent of U.S. imports ployment in the tourism sector in 2003. At the under AGOA preferences—and three-quarters same time, however, South Africa was the of the total consists of oil. But non-oil exports world’s fastest growing tourist destination in grew rapidly in 2002. Despite a slump in the 2002, with 20 percent growth in arrivals over U.S. economy that led to an overall decline of the year before, and the momentum continued nearly 16 percent in AGOA exports, nontradi- into 2003, though the pace is likely to slow tional exports were sharply higher—textiles with the stronger rand. Thus, superimposed and apparel more than doubled from 2001, on the overall pattern of growth has been a while transportation equipment and agricul- southward shift in the industry’s center of tural products were up 80 percent and 38 per- gravity. Whether this will be reversed remains cent respectively.3 to be seen. In South Africa, the sharp fall of the rand be- In spite of the overall disappointing results, ginning in 2000 provided a strong stimulus to there were some positive developments. Aver- growth, but much of that was reversed through age per capita income rose for a fourth succes- 2002 and into 2003 as the currency bounced sive year in 2002, which is the longest sustained back because of tight money and the unwinding increase in over two decades. Moreover, the of the Reserve Bank’s net open forward position slowdown was largely attributable to a small which had unnerved investors. Growth in 2002 group of poor performers. In Nigeria, a lower was a still-robust 3.0 percent, though the con- OPEC quota and budget gridlock offset much tribution to growth from trade declined to neg- of the potential gain from higher oil prices, ative territory, while domestic demand soared Ethiopia and Eritrea suffered through another by 4.1 percent. Though the domestic economy year of increasingly savage drought, and deep- slowed in the first quarter of 2003, growth re- ening political crises paralyzed Côte d’Ivoire mained relatively strong at 3.5 percent (saar), and Zimbabwe. For this group of countries— particularly investment, which was up an im- representing around one-third of the region’s pressive 8 percent. Given the momentum evi- population and GDP—growth fell by two- dent in the first quarter, domestic demand thirds, from 1.7 percent to 0.6 percent in the should remain relatively strong, though with year. This same group also contributed to most the rand remaining firm, net exports will be a of the retrenchment in exports. By contrast, drag on growth and GDP is expected to grow elsewhere in the region, GDP growth slowed only around 2.8 percent. Especially worthy of only marginally. As usual, politics played an note is an auspicious turnaround in the labor overarching role and countries in conflict or ex- market in 2002, with formal sector payrolls in- periencing civil disruption were at the bottom creasing after seven straight years of decline, of the league. Even here there is a glimmer of even though the increase of 70,000 jobs is small hope, though, in the growing institutional compared to the official estimate of 4.8 million strength of initiatives such as the African Union, unemployed. NEPAD, and the East African Community. In Nigeria, slower growth in 2002 reflected Also encouraging is further evidence of OPEC production constraints and the impact Africa’s potential competitiveness, given the of budget gridlock that limited spillovers from right incentives and opportunities. An impres- oil production to the rest of the economy in sive recent example is the growth of nontradi- spite of strong prices. Hydrocarbons constitute tional exports under the U.S. African Growth almost all Nigerian exports, but gas is taking and Opportunity Act (AGOA), which extends an increasing share, and in 2002 a one-third in- preferential access to imports from a growing crease in liquid natural gas (LNG) production list of eligible countries. So far, there have partially offset a 6 percent fall in oil. Produc- 250 R E G I O N A L E C O N O M I C P R O S P E C T S tion in the first quarter of 2003 was up around medium term as the global recovery consoli- 5 percent from 2002, somewhat below the av- dates. Energy sectors are expected to contribute erage increase for OPEC as a whole.4 Nigeria significantly to regional growth, especially with would like an increase in its quota from the a moderate rebound in Nigeria from a weak current level of 2.1mb/d that is well below ca- performance in 2002. But, more generally, pacity. But a significant rise in the near term is West Africa is emerging as an energy hot spot, unlikely, as slow growth in the industrial econ- with mainly offshore exploration and develop- omies will constrain the global sector. Never- ment activities underway from Angola through theless, higher production in 2003 in addition the Gulf of Guinea, and extending to nontradi- to further growth in LNG and continuing price tional energy producers from São Tomé and strength, will underpin a stronger performance Principe, to Mauritania. Major energy-related this year. The most encouraging recent news infrastructure projects including the West from Nigeria concerned the presidential elec- African Gas Pipeline and Nigeria’s plan to end tion which has strengthened the fragile political gas flaring by 2004 will also have a direct im- process. But, equally important, meaningful pact on investment and incomes. Though link- progress on economic reform has been frus- ages to non-energy sectors are generally weak, tratingly slow. On balance, growth this year increased fiscal revenues will permit relatively and in the medium term is likely to be only expansionary fiscal policies. Overall, oil ex- moderate. porters’ growth is expected to reach 3.6 percent in 2003 and accelerate further to 3.9 percent by Near-term outlook 2005. Despite some retrenchment in non-oil Though the forecast calls for world economic commodity prices in the second quarter, non- growth to accelerate in the second half of oil exporters will benefit from better terms of 2003, the pace of recovery is expected to re- trade in the current year and relative stability main sluggish, especially in Europe where the after that. Policy improvements are expected to economy stalled in the first quarter and the have a cumulative impact on competitiveness risk of outright recession has risen. Even with and will help attract investment. Realistically, a pickup in momentum going into the second though, this is a slow process and many of the half, the EU is expected to grow by only 0.9 benefits will be realized only in the longer term percent in 2003 and 1.9 percent in 2004. SSA (figure A1.5b). should benefit from a more rapid expansion of European imports thanks to the appreciation Long-term prospects of the euro, but the 4.1 percent increase in EU In the longer run, SSA will continue to face import demand anticipated over the next two formidable obstacles to growth from low sav- years is barely half the pace of the late 1990s. ings and investment rates, limited quantity Moreover, travel and tourism will continue to and quality of infrastructure and human capi- reflect security concerns, fanned recently by tal, and especially HIV/AIDS. As a result of the UK government’s decision in May (since re- the growing severity of the HIV/AIDS epi- tracted) to ban commercial flights to Kenya, demic, population growth for the region has though, as noted, there will be differences been revised downward by 0.3 percent to 1.9 within the region. However, for SSA overall percent per annum. Total GDP growth has the external environment will provide only been lowered by the same amount. The expec- modest support for growth in the current year, tation is that per capita GDP growth will re- with exports rising 2.9 percent. A further ac- main at 1.6 percent. This expectation may be celeration to 4.9 percent is expected in 2004 as optimistic given the long-run performance of the recovery continues to build. the region, but even so it is barely half of what The forecast anticipates stronger growth for is needed to achieve the MDGs. There is little both energy and non-energy exporters in the doubt that SSA will continue to lag behind 251 G L O B A L E C O N O M I C P R O S P E C T S ing governance and policymaking in general. Figure A1.5b Non-oil exporters For the region as a whole, real per capita in- increasingly drive SSA growth come in absolute terms peaked in 1974, de- GDP growth, percent clining since then at an average annual rate of 5.0 0.7 percent. Yet differences across the region Oil exporters Non-oil exporters X RSA are striking. The worst performances have in- 4.0 ᮡ evitably been associated with failed states and civil conflict, including Sierra Leone where 3.0 growth has averaged –3.0 percent over the pe- ᮡ riod, Democratic Republic of Congo at –5.6 2.0 percent, and Liberia at –5.8 percent. But, at the other end of the spectrum, long-term growth 1.0 averaged 4.4 percent in Mauritius and 5.6 per- cent in Botswana. In between, there are success 0.0 stories of spectacular turnarounds. Mozam- 2002 2003 2004 2005 bique, after a protracted civil war and average Note: RSA = Republic of South Africa. Source: World Bank staff simulations. growth of –1.8 percent during the 1980s, aver- aged 5.6 percent from 1992–2001. Already Mozambique, and other countries such as other developing regions over the forecast pe- Uganda and Tanzania, have made substantial riod, with outcomes well behind best practice progress and have results to show for it, while in other regions. many others are at an earlier stage of the The expectation of a reversal of SSA’s process. By early 2003 almost all countries in lengthy downward spiral hinges on strong as- the region were participating in the PRSP sumptions about ending conflicts and improv- process, with 15 having completed PRSPs and Table A1.5 Sub-Saharan Africa forecast summary Growth rates/ratios (percent) 1991–2000 2001 2002 2003 2004 2005 2006–15 Real GDP growth 2.3 3.2 2.8 2.8 3.5 3.8 3.5 Consumption per capita –0.3 0.6 0.1 0.5 0.8 1.1 1.2 GDP per capita –0.3 0.8 0.5 0.5 1.3 1.6 1.6 Population 2.6 2.4 2.3 2.3 2.2 2.2 1.9 Gross Domestic Investment/GDPa 17.4 19.3 19.8 20.2 20.0 20.1 20.8 Inflationb 10.5 5.8 4.2 3.7 4.2 4.7 General gvt. budget balance/GDP –4.5 –1.6 –1.0 –1.6 –1.6 –1.7 Export Market Growthc 7.3 0.4 2.0 5.0 7.3 7.1 Export volumed 4.5 3.8 0.7 2.9 4.9 5.4 Terms of trade/GDPe 0.1 –1.5 0.5 –0.4 –0.8 –0.5 Current account/GDP –2.0 –2.2 –2.2 –2.7 –2.5 –2.6 Memorandum items GDP growth: SSA excluding South Africa 2.9 3.6 2.7 2.8 3.8 4.3 Oil exporters 2.6 4.3 3.0 3.6 3.8 3.9 CFA countries 2.6 3.1 1.9 2.2 3.1 3.5 a. Fixed investment, measured in real terms. b. Local currency GDP deflator, median. c. Weighted average growth of import demand in export markets. d. Goods and non-factor services. e. Change in terms of trade, measured as a proportion of GDP (percentage). Source: World Bank baseline forecast July 2003. SSA is Sub–Saharan Africa. CFA is Communaute Financiere Africaine. 252 R E G I O N A L E C O N O M I C P R O S P E C T S another 25 having I-PRSPs. Meanwhile, re- for oil exporters at $20 billion, 4.8 percent of gional initiatives are enhancing the credibility GDP. Algeria continued expenditures under of governments and strengthening intra- the PSRE (Programme de soutien a la relance regional cooperation. At the same time, how- économique), boosting the construction and ever, deep-seated conflicts in West and Central services sectors and offsetting weakness in the Africa remain to be resolved and in many areas agricultural sector, which was affected by a political processes remain fragile. severe drought, particularly devastating for In principle, higher standards of gover- cereal crops. Private sector growth in Saudi nance and improved policies will encourage Arabia and several Gulf countries is expected higher savings and investment, and raise pro- to remain weak as a result of the disruption ductivity and growth. Yet, at the same time it caused by the war in Iraq, but companies from will remain a struggle to overcome low levels Saudi Arabia and Kuwait will benefit from of human and physical capital, poor infra- subcontracting work associated with the re- structure, HIV/AIDS, and negative percep- construction of Iraqi infrastructure. In Iran, tions of international investors. Moreover, the the lifting of import restrictions in the past region remains highly dependent on primary year has allowed the non-oil industrial and commodity exports, hence exposed to high ex- manufacturing sector to raise production, and ternal volatility. While these factors indicate domestic demand has been a strong driver for downside risks to the projections, achieving growth. the moderate improvement in performance Developments were more adverse for diver- envisaged by the forecast seems a plausible sified exporters, particularly those in the Mid- baseline expectation. dle East. The prospects for war in Iraq led to continued stagnation of tourism, which had not fully recovered from the events of Septem- Middle East and North Africa ber 11. These factors and the waning of exter- nal demand, particularly in Europe in late Recent developments 2002, drought conditions in several countries, T HE OVERARCHING event in the Middle and a weak investment climate in several East and North Africa in 2002–03 was North Africa countries all contributed to a fall the buildup toward the war in Iraq, in growth from 3.8 percent in 2001 to 2.8 per- provoking continued high oil prices, further cent in 2002. Growth in Egypt appeared set shocks to the tourism sector, and declining for recovery in late 2002 with a recovery in confidence in the private sector. Uncertainty tourist arrivals and increasing business confi- before the war had an even larger impact than dence, but tourist arrivals fell in the first half the war itself. However, the economic shocks of 2003. The Egyptian exchange rate was resulting from the conflict highlighted several floated in January 2003, but the central bank weaknesses in regional economies—the slack- is maintaining higher interest rates to prevent ening of the investment climate, weakness in further depreciation of the pound, curbing pri- the private sector, and the relatively poor vate investment. Production in Jordan plum- prospects for new foreign direct investment. meted over late 2002 and into early 2003 In the rise of uncertainty surrounding the (figure A1.6). Contraction in the agricultural situation in the Gulf in 2002 and early 2003, sector was caused by drought conditions in oil prices surged and oil exporters lifted oil Tunisia, the fourth consecutive year of production. Combined with fiscal expansion drought. Exports, particularly from the manu- programs, it led to an acceleration of output facturing sector, were adversely affected by the growth to 3.2 percent in oil-exporting coun- waning of demand from the European Union tries in 2002. High oil prices helped to keep in late 2002. A tighter fiscal stance, in the face current account balances in surplus in 2002 of a widening current account deficit, also con- 253 G L O B A L E C O N O M I C P R O S P E C T S porters recover from the adverse impacts of Figure A1.6 Some diversified economies slowing external demand and drought condi- hard-hit by events leading to Iraq war tions. The diversified exporters should begin Percent change to recover in 2004 as external demand im- 30.0 proves and as growth in oil exporters is 20.0 Jordan IP (ch% y/y) buoyed by further fiscal expansion, somewhat offsetting the impacts of lower oil prices and 10.0 ᮡ production. 0.0 Most of the acceleration in 2003 will occur in the oil-exporting countries, which are ex- –10.0 pected to grow more quickly in 2003 as a re- ᮡ –20.0 Egypt: Remittances sult of fiscal pump-priming and increased oil (ch% y/y) production quotas. GDP growth in oil ex- ᮡ –30.0 Egypt: Tourism receipts porters should reach 3.9 percent in 2003–04. (ch% y/y) –40.0 Oil production levels for the oil exporters in 2003 will be higher than 2002, ensuring that 01 01 02 02 02 02 03 20 20 20 20 20 20 20 the oil sector will make large, positive contri- 3 4 1 2 3 4 1 Q Q Q Q Q Q Q Source: Egypt, Ministry of Foreign trade; Haver Analytics. butions to GDP volume growth. High prof- itability in oil, and in services sectors, such as telecommunications and banking, has led to strained growth. However, Tunisia continued surges in stock markets in Iran and Saudi to pursue reforms in the financial sector and Arabia, particularly as a result of thin mar- continued to attract FDI in 2002 thanks to the kets. Firms in the Gulf Cooperation Council award of a second GSM license when global (GCC) countries are expected to benefit from flows to developing countries were shrinking. contracting agreements as the reconstruction Conversely, Morocco had positive agricultural of Iraq gets underway in the second half of growth in 2002, but the industrial moderniza- 2003, helping to boost the weak private sec- tion program under the Euro-Med Association tor. Evidence is also emerging that citizens Agreement and the privatization programs of oil-exporting countries, particularly in the have tended to stagnate. The exception to this Gulf, are traveling less and boosting domestic generally gloomy picture is Jordan, with buoy- consumption. ant GDP and export growth despite a slow- Saudi Arabia and Iran have begun to push down in export market growth and a stagnat- through needed reforms in several sectors such ing tourism sector. as mining, capital markets, and insurance, as well as privatizing state concerns to encourage Short-term prospects foreign investment in the oil sector. But Alge- Prospects in 2003–04 will be shaped by the ria has abandoned proposed hydrocarbons re- path of oil prices and public expenditures in forms. Moving into 2004, the impetus to oil-exporting countries; the recovery of exter- growth from the oil sector should wane in nal demand; the speed of recovery in the these countries as oil prices and production tourism, trade, and transportation sectors in fall, but continued fiscal stimulus and a re- the Middle East post-war; weather conditions; bound in private sector activity should main- and the policy response to slackening of the tain GDP growth. Growth in Algeria will be investment climate in the diversified exporters. supported in 2004 by the online debut of a Growth in the region is expected to rise slightly major gas project, but fiscal stimulus under in 2003 to 3.3 percent as oil exporters increase the PSRE is being absorbed more slowly than production and exports and diversified ex- expected. 254 R E G I O N A L E C O N O M I C P R O S P E C T S Growth in the diversified exporters is fore- gion, particularly given the very high rates of cast to remain sluggish in 2003 at 2.4 percent, growth expected in the labor force, as the age caused by stagnation in tourism in the first half structure of the population shifts to reduce of the year because of the Iraq war. The agri- the dependency ratio, constituting a “demo- cultural sector in Tunisia will expand in the graphic gift.” The region has achieved macro- wake of good rains, and agricultural expansion economic stability for the most part, with low will continue in Morocco, but slow import de- inflation and stable external debt positions, mand in Europe will keep the manufacturing but growth in the region is not reaching its po- sectors depressed. The countries most affected tential, with many countries still stuck with by the Iraq war were border nations Syria and per capita GDP growth of around 1.5 percent. Jordan. Both countries have extensive trade Prospects for growth and sustainable employ- links with Iraq, and Jordan sourced its oil from ment can be improved through reforms in Iraq at concessional rates. Prior to the war, Jor- trade regimes and a strengthened investment dan’s exports to Iraq were growing strongly. climate. Traditional trade liberalization, such The disruption to oil supplies from Iraq and to as lowering tariff and non-tariff barriers, is haulage routes through Iraq has meant that vital considering that the Middle East and output growth in the first half of 2003 in these North African (MENA) region is one of the countries was adversely affected. The Jordan- more highly protected developing regions. ian government has decreased fuel subsidies While MENA’s export products (excluding oil) and broadened the tax base to ease the impact are not highly integrated into global trade on the fiscal accounts of higher oil prices, markets, the region has strong potential for affecting the current balance. Jordanian ex- non-energy exports. The Mediterranean coun- ports to other destinations, particularly to the tries are well on the way to trade reforms United States through the “Qualified Industrial through their commitments over the coming Zones,” are unaffected by the disruption decade to the Euro-Med agreements, bilateral caused by the conflict, but uncertainty in the trade agreements, and WTO membership. But private sector has decreased confidence. trade reforms must also include ‘behind the In response to this range of negative border’ reforms (trade facilitation, services lib- growth factors, central banks in several coun- eralization, and improved competitiveness) de- tries (Jordan, Morocco, Tunisia) have reduced signed to increase productive efficiency. interest rates to stimulate domestic demand. Reform of the investment climate is needed Growth conditions should improve later in to ensure sustainable employment growth, be- 2003 and into 2004 as confidence returns to cause trade reforms increase growth only when the tourism market, the European economy they stimulate new investment. Much of the rebounds, and the effects of lower interest impetus to growth in recent years in many rates filter through to domestic investment. countries has come from the public sector, and Diversified economies are expected to grow much of the FDI flowing into the region is tar- 3.7 percent in 2004. This is below the average geted to extractive industries or parastatals that of the diversified exporters during the 1990s, have been privatized. To capture greater bene- where GDP growth achieved annual growth fits from trade and financial market integra- of 4.2 percent. The lackluster investment cli- tion, policy-makers in the region should focus mate in these countries requires further atten- on structural and microeconomic impediments tion in the reform process. to efficient resource allocation and to improve competition. Policy reforms are required to en- Long-term prospects sure a sound regulatory environment in prod- Reducing unemployment through higher GDP uct and factor markets. Improving governance, growth remains the key challenge for the re- improving the quality of public institutions, 255 G L O B A L E C O N O M I C P R O S P E C T S Table A1.6 Middle East and North Africa forecast summary Growth rates/ratios (percent) 1991–2000 2001 2002 2003 2004 2005 2006–15 Real GDP growth 3.4 3.2 3.1 3.3 3.9 3.5 4.3 Consumption per capita 0.2 4.8 1.8 1.0 1.3 1.3 2.3 GDP per capita 1.2 1.3 1.1 1.3 1.9 1.5 2.5 Population 2.2 1.9 1.9 1.9 1.9 1.9 1.7 Gross Domestic Investment/GDPa 21.4 21.3 21.6 21.8 22.0 22.1 26.2 Inflationb 7.5 2.6 3.8 4.0 4.0 4.0 General gvt. budget balance/GDP –1.2 –0.8 –3.0 –2.3 –3.2 –2.2 Export Market Growthc 7.5 –1.1 2.2 8.3 8.1 7.9 Export volumed 5.7 3.6 1.2 4.2 5.5 5.2 Terms of trade/GDPe 0.3 –1.8 0.1 –2.2 –1.7 –1.1 Current account/GDP –1.9 4.2 3.5 0.3 –1.5 –2.6 Memorandum items GDP growth: Oil exporters 3.0 2.9 3.2 3.9 3.9 3.3 Diversified exporters 4.2 3.8 2.8 2.4 3.7 3.8 a. Fixed investment, measured in real terms. b. Local currency GDP deflator, median. c. Weighted average growth of import demand in export markets. d. Goods and non–factor services. e. Change in terms of trade, measured as a proportion of GDP (percentage). Source: World Bank baseline forecast July 2003. and enforcing public accountability are neces- Notes sary if a vibrant public sector is to evolve. Com- 1. In April 2003, eight transition countries signed bating bureaucratic delays and inefficiency, im- the accession treaty to join the EU in May 2004, along proving the quality of infrastructure services, with Malta and Cyprus. and fighting corruption are also essential ele- 2. The UN’s FAO’s Global Information and Early Warning System identifies 25 countries representing 56 ments of a better investment climate. percent of total Sub-Saharan population where agricul- With only one-third of the labor force active tural production has been seriously disrupted. (ftp://ftp. today, women represent a huge untapped re- fao.org/docrep/fao/005/y9304e/y9304e00.pdf) Accord- source in the region. Experience from around ing to the UN WFP, 40 million persons in SSA are fac- the world suggests that women, particularly the ing acute hunger because of drought, disease, and con- young and well-educated, can reap gains from flict (http://www.wfp.org/index.asp?section=3). trade and investment climate reforms. These 3. http://www.agoa.gov/resources/TRDPROFL03. pdf. AGOA rules of origin are to be tightened in 2004 gains are already evident in the garment and and it remains to be seen how long lasting the benefits textile industry in Egypt, Jordan, Morocco, and will be. Tunisia. Provided that economic and social bar- 4. According to data from the Energy Information riers to women are dismantled, women can Agency of the US Department of Energy. See http:// more widely participate in economic life, www.eia.doe.gov/emeu/ipsr/t11a.xls. thereby boosting economic growth and produc- tivity in the region. 256 Appendix 2 Global Commodity Price Prospects C ommodity prices increased signifi- coming years, in particular the Caspian, Rus- cantly from the lows reached shortly sia, West Africa, and several deepwater loca- after the terrorist attacks on Septem- tions. Much of the moderate growth in world ber 11, 2001 (figure A2.1). Crude oil prices oil demand is expected to be captured by non- rose 78 percent from the December 2001 lows OPEC producers, thus rising supply competi- to the highs in February 2003, just prior to the tion, both inside and outside OPEC, is ex- start of the war in Iraq, but have since de- pected to lead to lower prices. clined. Agricultural prices were up 29 percent The rise in agricultural prices since October from the lows to recent monthly highs, while 2001 was caused mostly by reduced supplies metals and minerals prices rose 15 percent. from earlier low prices and severe El Niño- The decline of the dollar since early 2002 (10 related droughts in 2002 (in Australia, Ca- percent on a real-trade weighted basis) con- nada, the Middle East, and the U.S.), which tributed to the rise in commodity prices. Pe- reduced grain and oilseed production. Cocoa troleum and most agricultural prices are now supplies were disrupted by conflict in Côte expected to decline on rising supplies, while d’Ivoire, while production was reduced for metals and minerals prices are expected to natural rubber, robusta coffee, cotton, and continue their recovery because of higher de- vegetable oils because of earlier low prices. mand in the foreseen economic recovery. Most of the sharp agricultural price increases The increase in crude oil prices resulted in 2002 and 2003 are expected to be reversed from strong OPEC production discipline, ex- as surplus production capacity once again re- tremely low inventories, cold winter weather, sults from higher prices. More rapid economic and supply disruptions in Venezuela, Iraq, and growth would strengthen demand somewhat Nigeria. Higher output from other OPEC and moderate the price declines. However, in- members leading up to the war in Iraq pre- come elasticities for most agricultural commod- vented prices from spiking sharply higher, and ities are low, and with weak demand growth use of strategic stocks was not required. Crude agricultural prices are expected to decrease. oil stocks remain low and the return of Iraqi Fertilizer prices generally increased in 2003 exports has been delayed, thus prices are likely along with the recovery in agricultural com- to remain relatively firm for the balance of modity prices. Higher prices for natural gas— 2003. The return of Iraqi exports and rising a key input in nitrogen fertilizer production— capacity in both OPEC and non-OPEC coun- caused nitrogen fertilizer prices to rise sharply. tries is expected to lead to lower prices in In addition, production capacity utilization in 2004 and beyond. Large increases in produc- the fertilizer industry increased to five-year tion are expected in a number of regions in the highs and further contributed to the price in- 257 G L O B A L E C O N O M I C P R O S P E C T S A2.2). Such declines in commodity prices rela- Figure A2.1 Commodity price trends tive to manufactures prices pose real challenges Index, January 2000 = 100 for developing countries that depend on pri- 140 mary commodities for a substantial share of Crude oil ᮡ their export revenues. For example, 57 percent of merchandise exports from Sub-Saharan 120 Africa in 2000 came from primary commodi- ties and fuels. The situation is not expected to 100 improve, with real non-oil commodity prices ᮡ expected to increase only modestly through 2015 and crude oil prices expected to decline ᮡ 80 Agriculture by 23 percent from 2002 levels. Multilateral Metals and minerals trade negotiations could lead to higher agricul- tural prices if reforms reduce production subsi- 60 Jan. July Jan. July Jan. July Jan. July dies and tariffs in major consuming and pro- 2000 2000 2001 2001 2002 2002 2003 2003 ducing countries; however, little progress on Source: World Bank. reforms has thus far been achieved. (Specific commodity prices and price indices forecasts for 2003, 2004, 2005, 2010, and 2015 in cur- creases. The recent downturn in agricultural rent and constant dollars are given in appendix commodity prices is expected to be reflected in tables A2.14–16. The forecasts do not reflect lower fertilizer prices in 2004 and 2005. the effects of a multilateral trade agreement be- The modest recovery in metals and miner- cause of the uncertainty of such an agreement.) als prices resulted from production cuts be- ginning in 2001, and weakening of the U.S. Beverages dollar. Demand growth has been weak, and The World Bank’s index of beverage prices stocks of most metals remain high. The one (composed of coffee, cocoa, and tea prices) is exception is nickel, where strong demand for stainless steel, low inventories, and tight sup- plies, caused prices to almost double since the Figure A2.2 Real commodity prices lows in 2001. A recovery in metals demand is Index, 1990 = 100 expected to send most metals markets into 250 deficit and allow prices to increase over the Forecast Crude oil next several years. If global economic growth 200 ᮡ accelerates more quickly than projected, met- als and minerals prices would increase more 150 rapidly in the near term. Over the longer term, Agriculture real prices are expected to decline as produc- ᮡ tion costs continue to fall because of new tech- 100 ᮡ nologies and improved managerial practices. There is also little constraint on primary re- Metals and 50 minerals source availability. Real commodity prices declined significantly 0 from 1980 to 2002, with the World Bank’s 70 75 80 85 90 95 00 05 10 15 index of agricultural prices down 47 percent, 19 19 19 19 19 19 20 20 20 20 crude oil prices down 43 percent, and metals Source: World Bank. and minerals prices down 35 percent (figure 258 G L O B A L C O M M O D I T Y P R I C E P R O S P E C T S Surpluses over the past four seasons have Figure A2.3 Beverage prices kept the coffee market depressed, and this sit- US cents per kilogram uation has often been referred to as the “coffee 600 crisis” by the popular press. Attempts to deal Arabica coffee with the surpluses have either been largely un- 500 successful or abandoned. The Association of ᮡ Coffee Producing Countries (ACPC), which 400 urged coffee-producing countries to join its ex- Tea port retention scheme, ceased operating last 300 ᮡ year. The International Coffee Organization (ICO), in an effort to reduce coffee availability 200 and thus push prices higher, called for the re- moval of low-quality coffee beans. This plan ᮡ 100 Cocoa too has met resistance because there is no well- ᮡ Robusta coffee defined compensation mechanism in place. 0 Jan. Jan. Jan. Jan. Jan. Jan. Jan. In addition, improved roasting methods have 1997 1998 1999 2000 2001 2002 2003 made it easier to remove the harsh taste of nat- Sources: International Coffee Organization, International ural arabicas and robustas, enabling roasters Cocoa Organization, International Tea Committee. to produce the same coffee quality with lower- quality green beans, thus putting into question expected to increase by about 6 percent in ICO’s proposal. 2003, largely reflecting coffee price increases Global coffee production during the (arabica up 7 percent and robusta up 33 per- 2003–04 season is expected to be about 107 cent) in response to reduced output from million bags, down from last season’s 123 mil- Brazil (figure A2.3). Cocoa prices, which have lion bags (table A2.1). Almost all of the re- been (and are likely to be) extremely volatile duction is because of reduced Brazilian output because of the political unrest in Côte d’Ivoire, (from 52 million bags in 2002 to 34 million are expected to remain unchanged. Fears that bags in 2003), which is partly because of less tea prices might suffer a major setback result- favorable weather conditions and partly be- ing from the military conflict in Iraq did not cause of the strength of the Brazilian currency. materialize and the three-auction average for Still, Brazil will account for one-third of 2003 is expected to remain at its 2002 level. global coffee output while Colombia and Viet- nam are expected to reach 12 and 11 million Coffee. Despite the increase in coffee prices bags, respectively, and be the second and third expected in 2003, (robusta up US$0.22 to largest coffee suppliers of arabica and robusta, US$0.88/kg and arabica up US$0.10 to respectively. US$1.46/kg) prices will remain near historical Coffee prices are projected to increase in lows—at about one-third of their 1960 real 2004, with arabica up 9 percent and robusta levels. Low coffee prices reflect both the surge up 5 percent. Over the longer term, real coffee in supplies and weak demand. During the past prices are expected to increase relative to the five seasons, global coffee production has av- 2002 depressed levels but remain well below eraged 114 million 60 kg bags, compared to the historical highs of the 1970s and more re- 99 million bags during the five prior seasons cent highs of the mid-1990s. By 2015, real when coffee prices peaked. Per capita con- arabica and robusta prices are projected to in- sumption in the major importing countries has crease about 50 and 70 percent, respectively, been stagnant at 4.6 kgs of green coffee equiv- over their 2002 levels. Prices would still be alent during the past ten years. only about half of their 1990s peaks. 259 G L O B A L E C O N O M I C P R O S P E C T S Table A2.1 Coffee production in selected countries (million bags) 1998 1999 2000 2001 2002 2003 Brazil 35.6 30.8 34.1 35.1 51.6 33.6 Colombia 10.9 9.5 10.5 12.0 10.9 11.8 Vietnam 7.5 11.0 15.3 12.3 10.3 10.8 Indonesia 7.0 6.7 6.5 6.2 6.0 6.1 México 5.0 6.2 4.8 4.2 4.4 4.7 Guatemala 4.3 4.4 4.6 3.5 3.8 3.8 Ethiopia 3.9 3.8 3.7 3.8 3.0 3.3 Uganda 3.6 3.1 3.2 3.5 3.1 3.2 World 108.4 113.4 116.6 110.1 122.8 107.1 Note: Years refer to crop years beginning in April. Source: U.S. Department of Agriculture. Cocoa. Cocoa prices have staged a remark- the increase is expected to come from Ghana, able recovery, going from a 30-year low of the world’s second-largest cocoa supplier (from US$0.86/kg in February 2000 to a 16-year 341 to 450 thousand tons). Côte d’Ivoire’s high of US$2.28/kg in February 2003. Prices share is expected to remain largely unchanged have been extremely volatile, especially during at 1.26 million tons. Cocoa prices for 2003 are the last two years, with month-to-month price expected to remain at their 2002 levels, but a changes often exceeding 10 percentage points. small decline is expected in 2004 as production While the recovery in prices is a result of the continues to increase, an assessment which is return to normal supply levels, the volatility is based on the assumption that the strong prices a reflection of the political instability in Côte enjoyed during the last two seasons will pro- d’Ivoire, the world’s dominant supplier. vide further incentives to cocoa growers to Global cocoa production is expected to maintain their trees and increase production. reach 3 million tons during the marketing sea- The degree of volatility in cocoa prices is likely son ending in September 2003, up from last to remain high until the political unrest in Côte season’s 2.85 million tons (table A2.2). All of d’Ivoire is settled. Table A2.2 Beverages global balances Annual growth rates (percent) 1970 1980 1990 1999 2000 2001 1970–80 1980–90 1990–00 Coffee (Thousand bags) Production 64,161 86,174 100,181 116,581 110,104 122,759 2.1 1.4 1.2 Consumption 71,536 79,100 96,300 106,343 108,186 110,750 1.0 2.0 0.2 Exports 54,186 60,996 76,163 90,394 86,823 88,974 0.8 2.4 1.7 Cocoa (Thousand tons) Production 1,554 1,695 2,506 2,812 2,850 2,996 0.5 4.6 1.2 Grindings 1,418 1,556 2,335 3,014 2,858 2,976 0.2 4.5 2.6 Stocks 497 675 1,791 1,111 1,137 1,127 2.4 13.9 –4.7 Tea (Thousand tons) Production 1,286 1,848 2,516 2,895 3,021 3,000 4.1 2.9 1.5 Exports 752 859 1,132 1,330 1,391 1,419 24 2.4 1.6 Notes: Time reference for coffee (production and exports) and cocoa are based on crop years (October to September for cocoa and April to March for coffee). For coffee consumption and tea time is calendar year. Sources: US Department of Agriculture, International Coffee Organization, International Cocoa Organization, International Tea Committee, and World Bank. 260 G L O B A L C O M M O D I T Y P R I C E P R O S P E C T S Tea. The three-auction average tea price is additional 2.1 percent in 2005 as grain and expected to remain largely unchanged in 2003 oilseed prices decline from recent highs. Grains vs. 2002 at about US$1.50/kg and therefore prices have increased almost 15 percent from not to recover from the 20 percent decline be- the lows in 2001 and fats and oils prices have tween 2000 and 2002. The weakness in tea increased 26 percent. Over the longer term, prices is expected to persist because of over- real food prices are projected to decline 2.7 supply and a trend of slow growth of con- percent from 2003 to 2015. sumption. Production in 2002 was about the same as in 2001, but production is expected to Fats and oils. Prices of fats and oils are ex- increase in 2003. The rapid increase in pro- pected to increase almost 7 percent in 2003, duction in Vietnam has contributed to already which gives a cumulative increase of 20 per- ample supplies and threatens to depress prices. cent since 2001. However, prices have recov- Vietnam has doubled production since 1990. ered less than half of the decline experienced Tea prices have been volatile because of the from 1997 to 2001. The price increase is ex- uncertainties associated with the war in Iraq pected to be greatest in groundnut oil (up 60 and concerns that imports would be disrupted. percent). Price increases are expected to be less In addition, excessive rains in Sri Lanka, the in the two major oilseed crops, soybean and largest exporter, disrupted supplies. By 2015 palm, with soybean oil up 16 percent and palm real tea prices are expected to be slightly lower oil up 9 percent. than in 2002. Global production of the 17 major fats and oils is expected to increase by 1.4 percent in Food the season starting October 2003, following The World Bank’s food price index is expected last season’s increase of 2.5 percent. Demand to rise 4.3 percent in 2003 and be up 11.2 per- in 2003–04, to be fueled by increased imports cent from the low in 2000. However, the index by China and India, is projected to outpace is still well below highs reached in 1997 (figure production by at least 1 percentage point. A2.4). Following recent increases, the index is Global soybean production has increased expected to decline 2.4 percent in 2004 and an by more than 5 percent per year since 1990, with the most rapid increase in Brazil and Argentina (table A2.3). Argentina and Brazil Figure A2.4 Food prices have been increasing production at nearly 10 Index, 1990 = 100 percent per year since 1990. 130 Global palm oil production has doubled every eight years during the past three decades 120 with the largest increases coming from In- donesia and Malaysia (table A2.4). 110 Table A2.3 Soybean production 100 (millions of tons) United 90 Year Argentina Brazil States World 80 1990 11.5 15.8 52.4 104.1 1995 12.4 24.2 59.2 124.9 2000 27.8 39.0 75.1 175.1 70 2001 30.0 43.5 78.7 184.3 Jan. Jan. Jan. Jan. Jan. Jan. Jan. 2002 35.0 51.0 74.3 194.0 1997 1998 1999 2000 2001 2002 2003 Note: Argentina, Brazil, and the U.S. account for about 83 Source: World Bank. percent of global production. Source: USDA. 261 G L O B A L E C O N O M I C P R O S P E C T S Table A2.4 Palm oil production Rice prices are projected to rise about 4 (million tons) percent in 2003 and an additional 3.0 percent Year Indonesia Malaysia Nigeria World by 2005. Rice prices are well below historical norms relative to other food grains, and this 1980 0.69 2.58 0.43 4.59 should increase import demand for rice rela- 1985 1.24 4.13 0.39 7.04 1990 2.41 6.10 0.58 11.03 tive to wheat. Lower Indian exports this year 1995 4.22 7.81 0.66 15.22 because of drought will also contribute to the 2000 7.05 10.8 0.74 21.87 price increases. Global rice stocks are low and 2001 8.03 11.8 0.77 23.92 2002 9.02 11.9 0.78 25.03 prices could increase significantly if a poor 2003 9.60 12.7 0.79 26.59 crop reduces stocks further. Over the longer Source: Oil World. term, real rice prices are projected to rise 4.6 percent by 2015 vs. 2003, while most other grains prices are projected to decline. Grains. Global grain stocks, relative to use, Wheat prices are expected to decline in are expected to recover slightly from last 2003–05 as production recovers from severe year’s lows (excluding China where data is drought. Prices increased from US$112/ton in very uncertain). However, stocks remain low 1999 to US$148/ton in 2002, but are expected and there is still a risk that prices could rise to decline to US$133/ton by 2005. Production sharply if yields in the coming crop year are in the major exporters (U.S., EU, Canada, significantly below trends. If yields are near Australia, and Argentina) is expected to in- trend, then prices should decline and stocks crease 20 percent in the 2003–04 crop year should continue to rebuild. and stocks are expected to increase 17 per- Maize prices are projected to rise 6.7 per- cent. However, global wheat stocks remain cent in 2003 and then decline 5.7 percent in low (table A2.5) and there is a substantial risk 2004 as production increases and stocks re- that prices could rise if the drought persists. build (table A2.5). Production in the U.S., the major producer with 40 percent of world pro- Sugar. Sugar prices averaged 15.2 cents/ duction, is projected to increase 12 percent in kilogram in 2002 (figure A2.5). They are ex- 2003–04 compared to the previous year. Real prices are projected to decline about 4 percent Figure A2.5 Sugar prices from 2003 to 2015 as yields continue to grow U.S. cents per kilogram faster than consumption, as was the case dur- 30 ing the 1990s. 25 Table A2.5 Global grain stocks-to-use 20 percentages (excluding China) Maize Rice Wheat Total grains 15 1997–98 10.1 9.3 16.5 13.0 1998–99 11.5 10.2 18.0 14.0 10 1999–00 11.4 11.8 17.1 13.6 2000–01 11.6 13.6 18.6 14.4 5 2001–02 10.4 13.1 20.4 14.9 2002–03 6.3 10.0 15.8 12.0 2003–04 8.9 10.2 16.9 12.5 0 90s Low 6.1 8.6 13.9 9.8 Jan. Jan. Jan. Jan. Jan. Jan. Jan. 1997 1998 1999 2000 2001 2002 2003 Note: Data for 2003–04 is the USDA’s May 2003 estimate for wheat and maize and World Bank estimate for rice. Source: International Sugar Organization. Source: USDA. 262 G L O B A L C O M M O D I T Y P R I C E P R O S P E C T S Table A2.6 Foods global balances (million tons) Annual growth rates (percent) 1970 1980 1990 2000 2001 2002 1970–80 1980–90 1990–00 Grains Production 1,079 1,430 1,769 1,839 1,872 1,807 2.88 1.55 1.04 Consumption 1,114 1,451 1,717 1,862 1,902 1,906 2.58 1.78 1.02 Exports 119 212 206 233 237 237 6.35 0.13 0.94 Stocks 193 309 490 536 506 407 7.24 3.83 –0.56 Soybeans Production 42.1 62.2 104.1 175.1 184.3 194.0 6.84 1.87 5.08 Consumption 46.0 68.1 104.3 172.2 182.3 194.2 6.53 2.04 4.99 Exports 12.3 20.8 25.4 55.5 55.1 63.2 5.24 0.80 2.88 Stocks 3.4 10.3 20.6 30.6 32.0 31.0 13.83 –0.66 0.20 Sugar (raw equivalent) Production 70.9 84.7 109.4 130.4 134.7 143.3 2.80 1.59 3.26 Consumption 65.4 91.1 106.8 130.3 134.9 136.6 3.30 1.40 3.00 Exports 21.9 27.6 34.1 37.7 40.7 46.6 3.26 0.83 3.12 Stocks 19.6 19.5 19.3 37.3 34.0 32.2 3.96 –0.77 4.52 Fats and oils Production 39.8 58.1 80.8 117.2 120.1 121.8 3.68 3.54 3.70 Consumption 39.8 56.8 80.9 116.8 120.9 123.8 3.55 3.69 3.64 Exports 8.8 17.8 26.9 38.3 41.0 42.2 7.05 4.19 3.39 Stocks 5.2 9.3 12.2 14.8 13.6 12.1 7.09 2.44 0.69 Notes: Time references for grains, soybeans and sugar are based on marketing years, shown under the year in which production began, and vary by country; for fats and oils, crop years begin in September. Fats and oils includes the 17 major fats and oils. Sources: USDA and Oil World. pected to increase slightly in 2003 and 2004 Global consumption grew by 3.0 percent per as supplies are curtailed and stocks reduced. annum during the 1990s (table A2.6). High crude oil prices have contributed to the price increase by diverting sugar cane produc- Raw Materials tion to ethanol production in Brazil for use The index of agricultural raw materials prices as vehicle fuel. Prices are projected to average (composed of tropical hardwoods, cotton, and about US$0.16/kg in 2003 and 2004, and rise natural rubber) declined sharply during the slightly in 2005. The longer-term price pros- Asia crisis and then stabilized before declining pects are not encouraging for producers unless again as supplies of commodities continued global policy reforms are agreed in the current to increase (figure A2.6). Prices reached a low round of multilateral trade negotiations. With- in 2001 and have since recovered because of out reforms, nominal prices are expected to re- higher cotton and natural rubber prices. Nom- main low except when supplies are reduced by inal prices are projected to increase an addi- drought in a major producing country. tional 6 percent by 2005 from 2003 levels, Brazil, the world’s lowest cost and largest while real prices are projected to rise 10 per- sugar exporter, with about one-third of world cent from 2003 to 2015. sugar exports, has increased production and exports dramatically since 1990 and is ex- Cotton. Cotton prices are expected to in- pected to continue expanding. This has put crease 28 percent in 2003, following declines downward pressure on prices, as Brazilian in the two previous years that took prices to exports have increased from 1.3 million tons 30-year lows. The price recovery is due mostly in 1990–91 to 14.2 million tons in 2002–03. to an 11 percent reduction in supplies in the 263 G L O B A L E C O N O M I C P R O S P E C T S The A Index cotton price is expected to aver- Figure A2.6 Raw materials age US$1.30/kg during 2003 and remain at Index, 1990 = 100 approximately the same level during the next 130 two seasons, as the market appears to have reached a balance. By 2015, real prices are 120 projected to increase 30 percent relative to 110 2002 levels. 100 Natural Rubber. Rubber prices are ex- pected to increase 23 percent in 2003, after 90 falling to historical lows in 2001 following the 80 Asian financial crisis. The recent strength in rubber prices reflects increased demand as well 70 as supply controls by Thailand and Indonesia, 60 the dominant natural rubber suppliers with a Jan. Jan. Jan. Jan. Jan. Jan. Jan. combined 60 percent of global output. Con- 1997 1998 1999 2000 2001 2002 2003 sumption in 2002 increased 3.6 percent over Source: World Bank. 2001 and preliminary figures for 2003 indicate that it will stay strong. China, the world’s dominant natural rubber consumer, has been 2002–03 marketing season (table A2.7). Most the major source of increased demand (table of the reduction came from China and the A2.8). In the 12-month period ending May U.S., the world’s two dominant cotton suppli- 2003, Chinese rubber demand increased 7 per- ers, which account for over 40 percent of cent. Strong demand was also present by other global output. main buyers, notably the U.S., Japan, and Ger- The 2003 increase in cotton prices is ex- many. The demand for natural rubber has also pected to lead to a strong supply response, ac- been aided by lower demand for synthetic rub- cording to the International Cotton Advisory ber, whose prices increased considerably be- Committee. They estimate the 2003–04 global cause of high crude oil prices (crude oil is a cotton production will be 9 percent higher major cost component of synthetic rubber). than this season’s crop. Most of the increase is Natural rubber prices are expected to re- expected to come from China (almost 1 mil- main above US$0.90/kg for the next two to lion tons). Global consumption is expected to three years. Over the longer term, real prices stay slightly higher than production, causing are projected to increase slightly over the 2002 stocks to fall for a second consecutive season. levels. Table A2.7 Cotton production in selected countries (million tons) 1998 1999 2000 2001 2002 2003 China 4.50 3.83 4.42 5.32 4.92 5.80 United States 3.03 3.69 3.74 4.42 3.75 3.71 India 2.71 2.65 2.38 2.69 2.35 2.68 Pakistan 1.48 1.91 1.82 1.80 1.70 1.80 Uzbekistan 1.00 1.13 0.98 1.06 1.03 0.99 Franc Zone 0.90 0.93 0.70 1.03 0.93 0.95 World 18.55 19.09 19.46 21.51 19.20 20.96 Notes: Years refer to crop years that begin in August. Source: International Cotton Advisory Committee. 264 G L O B A L C O M M O D I T Y P R I C E P R O S P E C T S Table A2.8 Natural rubber consumption ing Japan, and has become a significant ply- (thousand tons) wood producer and exporter. The partial ban 1999 2000 2001 2002 on log exports from Asian and African ex- porters, intended to increase domestic pro- China 997 1,123 1,224 1,332 cessing, has raised the prices of logs, and United States 1,116 1,142 1,010 1,046 Japan 733 753 729 774 somewhat restricted supplies, while depressing India 617 638 631 675 prices of sawnwood and plywood relative to Korea, Rep. of 325 331 327 321 logs. However, the bans have not been totally Germany 226 250 245 254 France 253 262 262 241 effective and illegal exports continue. Tropical World 6,771 7,129 6,973 7,223 timber prices are expected to continue to re- Sources: LMC International, International Rubber Study cover, up 3 percent in 2004 and up 7 percent Group. in 2005, with demand in China, Japan, and Europe important factors determining the rate of price increase. Real tropical timber prices Tropical Timber. Tropical timber prices re- are projected to increase 28 percent from covered in 2002 and 2003 from sharp declines 2003 to 2015, but stay below the highs of the in 2001, with nominal prices up 9 percent in 1990s as new technology allows better utiliza- 2002 and expected to be up an additional 5 tion of timber. percent in 2003. The initial price increases were supported by the decline of the dollar vs. the Yen and Euro, but the price recovery ap- Fertilizers pears to have stalled in 2003 as demand has weakened in Asia and Europe. China has be- come the largest tropical log importer, displac- F ertilizer prices generally increased in 2003 as demand increased because of the rise in agricultural commodity prices. Among the Table A2.9 Raw materials global balances Annual growth rates (percent) 1970 1980 1990 2000 2001 2002 1970–80 1980–90 1990–00 Cotton (thousand tons) Production 11,740 13,832 18,970 19,461 21,510 19,200 1.2 3.1 0.8 Consumption 12,173 14,215 18,576 19,886 20,194 21,000 1.1 3.1 0.2 Exports 3,875 4,414 5,081 5,857 6,496 6,500 0.9 2.8 0.5 Stocks 4,605 4,895 6,645 9,637 10,585 8,780 1.7 2.8 1.4 Natural rubber (thousand tons) Production 3,140 3,820 5,080 6,730 7,190 7,110 1.8 3.2 3.1 Consumption 3,090 3,770 5,190 7,340 7,080 7,390 1.6 3.2 3.3 Net Exports 2,820 3,280 3,950 4,930 5,140 5,040 1.3 2.1 1.8 Stocks 1,440 1,1480 1,500 1,930 2,040 1,760 0.6 0.2 3.7 Tropical timber (thousand cubic meters) Logs, production 210 262 300 279.5 283.3 n.a. 1.5 1.7 0.5 Logs, imports 36.1 42.2 25.1 18.6 17.9 n.a. 0.2 5.1 5.4 Sawnwood, production 98.5 115.8 131.8 109.1 106.2 n.a. 1.2 1.7 2.0 Sawnwood, imports 7.1 13.2 16.1 23.1 22.5 n.a. 5.0 2.6 3.3 Plywood, production 33.4 39.4 48.2 58.1 55.5 n.a. 1.2 2.0 0.5 Plywood, imports 4.9 6.0 14.9 19.0 19.2 n.a. 0.7 9.1 3.6 n.a. = Not available. Notes: Time reference for cotton is based on crop year beginning in August; for natural rubber and tropical timber, time refers to calendar year. Sources: International Cotton Advisory Committee, International Study Rubber Group, FAO, and World Bank. 265 G L O B A L E C O N O M I C P R O S P E C T S Potash prices, as represented by muriate of Figure A2.7 Fertilizer prices potash (MOP), remained constant because US$s per ton prices are set by annual contracts, and have 210 not kept up with changed market fundamen- tals. Fertilizer demand is expected to fall in TSP 2004 and 2005 in response to the recent 170 ᮡ downturn in agricultural prices and this should cause most fertilizer prices to weaken. Urea prices rose about 38 percent in 2003 130 due partly to higher prices for natural gas. ᮡ Demand increased by an estimated 4 percent MOP resulting from higher planted crop area and 90 higher application rates. Nitrogen production capacity utilization increased to about 85 per- ᮡ Urea cent in 2002 from about 81 percent in 2001, 50 Jan. Jan. Jan. Jan. Jan. Jan. Jan. and is at the highest level in several years. In 1997 1998 1999 2000 2001 2002 2003 response to higher prices and demand, global Source: Fertilizer Week. production and exports both increased about 4 percent in 2002 after declining in the previ- ous year. Prices are expected to decline about 4 percent per year in 2004 and 2005 as de- three major types of fertilizer, nitrogen prices mand weakens and natural gas prices begin to (as represented by urea) increased most rapidly decline resulting from lower crude oil prices. because of higher prices of natural gas used in By 2015, real urea prices are expected to fall production in addition to demand increases. 9.5 percent from 2003 levels as the industry Phosphate fertilizer prices, as represented by expands production capacity more rapidly than triple super phosphate (TSP), increased after demand. falling for several years as demand increased MOP prices remained unchanged in 2003, and production capacity utilization increased. but new contract prices are likely to increase in Table A2.10 Fertilizers global balances (million tons Annual growth rates (%) 1970 1980 1990 1999 2000 Est. 2001 1970–80 1980–90 1990–00 Nitrogen Production 33.30 62.78 82.28 87.75 84.62 82.3 6.53 3.12 0.28 Consumption 31.76 60.78 77.18 84.95 81.62 n.a. 6.86 2.60 0.56 Exports 6.77 13.15 19.59 23.94 24.70 24.6 7.23 5.10 2.34 Phosphate Production 22.04 34.51 39.18 32.51 31.70 30.7 3.72 1.70 –2.10 Consumption 21.12 31.70 35.90 33.46 32.65 n.a. 3.85 1.39 –0.90 Exports 2.92 7.51 10.50 12.70 12.11 n.a. 8.37 5.01 1.44 Potash Production 17.59 27.46 26.82 25.01 25.54 25.9 3.97 –0.03 –0.49 Consumption 16.43 24.24 24.68 22.12 22.16 n.a. 3.93 0.05 –1.07 Exports 9.45 16.72 19.82 22.65 23.41 23.2 4.89 0.73 1.68 n.a. = Not available. Notes: All data are in marketing years. Source: FAO. The data for 2001 are estimated by World Bank staff from industry sources. 266 G L O B A L C O M M O D I T Y P R I C E P R O S P E C T S 2004 in response to improved demand and the highest capacity utilization rates in five years. Figure A2.8 Index: Metals prices and Production rose about 3 percent in 2002, with exchange rates Index, January 1990 = 100 most of the increase coming from Canada, which accounts for 40 percent of world ex- 130 ports and one-third of production. Prices are 120 $/Euro expected to increase by about 3 percent in ᮡ 2004 and remain at the higher level in 2005. 110 Increased domestic production in China, along 100 with large surplus global production capacity, is expected to keep price increases small. By 90 2015, real prices are projected to fall 3.5 per- cent compared to 2003. 80 TSP prices increased 7 percent in 2003 ᮡ 70 after falling 23 percent from 1998 to 2001 Metals & ᮡ minerals $/Aus$ and increasing 6 percent in 2002. Production 60 increased by about 7 percent in 2002, with Jan. Jan. Jan. Jan. Jan. Jan. Jan. Jan. 1990 1992 1994 1996 1998 2000 2002 2004 production in the U.S.—the world’s largest producer with a 30 percent share—increasing Source: World Bank, Datastream. by 13 percent, according to industry sources. Exports declined because of a sharp drop in Chinese imports, which were replaced by in- of the U.S. dollar, rising when the Euro or creased domestic production. TSP prices are Australian dollar appreciate and falling in the expected to decrease marginally in 2004 and opposite case (figure A2.8). 2005 as demand weakens; however, surplus Most metals markets are expected to re- production capacity is smaller than for other main in surplus or a balanced position in major fertilizers and is expected to remain 2003, and slip into deficit in 2004 as demand tight over the next several years. Thus real recovers. During the upturn of the next eco- prices are projected to remain about constant nomic cycle metals prices could rise signifi- between 2003 and 2015. cantly, as is typical during a recovery. How- ever, higher prices will induce development of new capacity and the restart of idle facilities, Metals and Minerals and prices will eventually recede. Real prices M etals and minerals prices have rallied a number of times since the lows of Oc- tober 2001, often on investor expectations are expected to decline in the longer term (fig- ure A2.9), as production costs continue to fall from new technologies and improved manage- that a global economic recovery would lead to rial practices, and there is little constraint on higher demand for metals. However, prospects primary resource availability. The one excep- for a strong economic recovery have kept tion is nickel, where new supply prospects over being pushed back and the price rallies have the next few years are quite limited, which been short-lived. Yet the index for metals and could lead to much higher prices. minerals is up 13 percent since October 2001 on improving fundamentals—notably pro- Aluminum ducer cutbacks, some modest reduction of in- Aluminum prices have been relatively steady ventories, and weakening of the U.S. dollar. the past year (figure A2.10), despite extremely As major producers and consumers do not high inventories and a market in surplus. have their currencies linked to the dollar, the Three main factors have limited an expected metal prices in dollars fluctuate with the value widening surplus and supported prices. First, 267 G L O B A L E C O N O M I C P R O S P E C T S minum production growth—where much of Figure A2.9 Index: Metals and minerals the recent increase has occurred—as it is a Index, 1990 = 100 large importer of alumina. Finally, tightness in 160 scrap supplies has generated higher demand for primary aluminum. 140 1990$ ᮡ If these conditions continue into 2004, the large surplus that had been forecast may not 120 occur. This may limit the price declines that some had forecast. However, world aluminum 100 production in May was the highest on record, 80 with Chinese production up 29 percent for the first five months of this year. There is also 60 the possibility that shut-in capacity could be Nominal restarted. The aluminum market is expected to move ᮡ 40 into deficit in 2005, but there are a number of 20 uncertainties in the near term, e.g., the extent 60 965 970 975 980 985 990 995 000 005 110 115 of demand growth, reactivation of idle capac- 19 1 1 1 1 1 1 1 2 2 2 2 ity, and the size of Chinese net exports. Real Source: World Bank. prices for primary aluminum are expected to slightly decline in the long term following a modest recovery during the next economic several production cuts have occurred in cycle. New low-cost capacity in a number of North America and elsewhere because of elec- countries, e.g. Canada and the Middle East, is tric power-related difficulties. Second, tight- expected to meet the relatively strong growth ness in alumina supplies has resulted in high in demand, although new investment will con- alumina prices, which may slow Chinese alu- tinue to require low-cost power supplies. There is not expected to be any constraint on alumina supply over the forecast period, and Figure A2.10 Aluminum price and LME stocks several new alumina capacity expansions are $/ton tons underway, e.g., Australia and Brazil. 1,800 1,400,000 Stocks Copper [right scale] 1,700 1,200,000 Copper prices have risen more than 20 percent ᮡ Price from the lows of October 2001, largely be- 1,600 [left scale] ᮡ 1,000,000 cause of a number of production cutbacks and 1,500 curtailments that began in 2001. This has 800,000 helped reduce the large surplus that emerged 1,400 in 2001, and LME copper stocks have fallen 600,000 about 30 percent from the peak in 2001—yet 1,300 they remain relatively high (figure A2.11). In 400,000 the first quarter of 2003, the global copper 1,200 market moved into deficit according to the In- 1,100 200,000 ternational Copper Study Group, because of Jan. Jan. Jan. Jan. Jan. Jan. Jan. lower world production and relatively strong 1997 1998 1999 2000 2001 2002 2003 demand, particularly in China where con- Sources: Platts Metals Week, London Metal Exchange. sumption rose more than 20 percent from a year earlier. 268 G L O B A L C O M M O D I T Y P R I C E P R O S P E C T S stocks, strong demand for stainless steel, and Figure A2.11 Copper price and tight supplies. A strike at Inco’s operations in LME stocks Sudbury, Canada, on June 1, 2003, briefly sent $/ton tons prices above US$9,500/ton, but prices receded 2,800 Stocks 1,000,000 after Russia’s Norilsk agreed to release 24,000 [right scale] 2,600 900,000 tons from inventory. This followed an an- ᮡ Price [left scale] 800,000 nouncement by the company in April to release 2,400 ᮡ 16,000 tons. 700,000 2,200 Demand for nickel rose 6 percent in 2002 600,000 because of strong growth of stainless steel pro- 2,000 500,000 duction, led by China, which increased stain- 1,800 less steel output by around 20 percent. Growth 400,000 for both stainless steel and nickel is expected to 1,600 weaken slightly this year, mainly because of the 300,000 1,400 slowdown in Europe, before strengthening in 200,000 2004. The nickel market is expected to slip 1,200 100,000 into deficit this year and remain so in 2004 and Jan. Jan. Jan. Jan. Jan. Jan. Jan. 1997 1998 1999 2000 2001 2002 2003 2005, mainly because of a dearth of major new Sources: Platts Metals Week, London Metal Exchange. projects to come on stream over this period. Nickel producers have had a number of setbacks with pressure acid leach (PAL) tech- nology at new laterite deposits (a high pro- Demand outside of China and neighboring portion of potential new developments have Asian countries remains relatively weak, and this type of ore-body). Technical problems and the market could remain in surplus in 2003. substantial cost overruns have significantly Much will depend on the extent of the eco- nomic recovery and continuation of produc- tion cuts in Latin America and the U.S. The Figure A2.12 Nickel price and LME market is expected to move into deficit in stocks 2004 as demand recovers, which will put up- $/ton tons ward pressure on prices. However, the restart 11,000 Stocks 70,000 of idled capacity in Chile and the U.S. could [right scale] ᮡ prevent prices from moving sharply higher. 10,000 60,000 In the medium term, the market is expected to return to balance as new capacity is ex- 9,000 50,000 pected to meet the projected growth in global Price 8,000 40,000 consumption of around 3.5 percent per year, [left scale] which will be mainly driven by strong growth 7,000 ᮡ 30,000 in China and other Asian countries. Over the longer term, increases in new low-cost capac- 6,000 20,000 ity are expected to result in a continued de- cline of real prices. A major uncertainty over 5,000 10,000 the forecast period will be the volume of Chi- 4,000 0 nese imports. Jan. Jan. Jan. Jan. Jan. Jan. Jan. 1997 1998 1999 2000 2001 2002 2003 Nickel Nickel prices have risen about 75 percent from Sources: Platts Metals Week, London Metal Exchange. October 2001 (figure A2.12), because of low 269 G L O B A L E C O N O M I C P R O S P E C T S Table A2.11 Metals and minerals global balances (thousand tons) Annual growth rates (%) 1970 1980 1990 2000 2001 2002 1970–80 1980–90 1990–00 Aluminum Production 10,257 16,027 19,362 24,485 24,477 26,099 3.2 1.9 2.4 Consumption 9,996 14,771 19,244 24,903 23,561 24,944 3.2 1.8 2.6 LME Ending Stocks n.a. 68 311 322 821 1241 n.a. –0.3 0.3 Copper Production 7,583 9,242 10,809 14,820 15,889 15,336 1.9 1.1 3.2 Consumption 7,294 9,400 10,780 15,176 14,876 14,963 2.5 1 3.5 LME Ending Stocks 72 123 179 357 799 856 7.4 –5.6 7.1 Nickel Production n.a. 717 842 1,107 1,145 1,177 n.a. 1.6 2.8 Consumption n.a. 742 858 1,172 1,178 1,206 n.a. 1.5 3.2 LME Ending Stocks 2 5 4 10 19 22 n.a. –0.5 9.6 n.a. = Not available. Sources: World Bureau of Metal Statistics, London Metal Exchange, and World Bank. limited the expected ramp-up of production at hedging). In addition, increased investment new projects in Australia. In addition, Inco demand resulting from declining equity mar- has temporarily suspended some construction kets and the U.S. dollar helped support prices. work at its US$1.4 billion Goro project in More recently, much of the movement in gold New Caledonia, after costs escalated by 30–45 prices seems to have been largely currency re- percent. The company’s current review of the lated (figure A2.13). project may delay start-up of production into Producer dehedging totaled about 4.5 mil- 2006. These difficulties at laterite projects will lion ounces in the first quarter of 2003 (6 per- likely impact development of forthcoming PAL operations. Cost estimates for future develop- ments are being raised, which will likely result Figure A2.13 Gold price and $/euro Index, January 1997 = 100 in higher long-term nickel prices. 375 1.20 With no new major greenfield projects on the immediate horizon, nickel prices could 1.15 $/Eur [right scale] jump significantly over the next couple of 350 ᮡ years before new supplies bring the market 1.10 back into balance. Over the longer term, large Gold [left scale] 1.05 325 new projects are planned for development, ᮡ and a new generation of technology and oper- 1.00 ational practices may help to reduce costs. In 300 0.95 addition to the risks of higher costs, a major uncertainty for the nickel market is the pace of 0.90 275 demand growth in China. 0.85 Gold 250 0.80 In 2002, gold prices climbed above their four- Jan. Jan. Jan. Jan. Jan. Jan. Jan. 1997 1998 1999 2000 2001 2002 2003 year trading range of roughly US$250–$300/ toz, largely because of the buyback of hedged Sources: Platts Metals Week, Datastream. positions by gold producers (referred to as de- 270 G L O B A L C O M M O D I T Y P R I C E P R O S P E C T S Table A2.12 Gold global balance (tons) 2002 1Q03 2001 2002 (% y/y.) 1Q02 2Q02 3Q03 4Q04 1Q03 (% y/y.) Jewelry 3,037 2,688 –11.5 655 638 657 738 586 –10.5 Other Fabrication 476 485 1.9 125 117 117 126 151 20.8 Bar Hoarding 248 252 1.6 80 53 61 58 35 –56.3 Net Producer Hedging 151 423 180.1 31 104 149 139 145 367.7 Implied Net Investment – 130 n.a. 40 48 24 17 64 60.0 Total Demand 3,912 3,978 1.7 931 960 1,008 1,078 970 4.2 Mine Production 2,623 2,587 –1.4 570 637 727 653 572 0.4 Official Sector Sales 529 556 5.1 163 118 83 191 151 –7.4 Old Gold Scrap 708 835 17.9 198 205 198 234 248 25.3 Implied Net Disinvestment 52 – n.a. – – – – – Total Supply 3,912 3,978 1.7 931 960 1008 1078 970 4.2 n.a. = Not available. Sources: Gold Field Minerals Service and World Bank. cent of producer hedges), about the same level Finally, official central bank sales continue of reduction that occurred in each of the third to take place. An important determinant of and fourth quarters of 2002, according to medium-term prices will be the decision by Gold Field Mineral Services. Many companies central banks whether to further stem official have indicated a desire to further reduce their gold sales when the Washington Agreement hedges, and shareholder sentiment generally expires in 2004 (the European Central Bank appears to be against hedging. This was evi- and 14 European central banks agreed in Sep- denced in the first quarter of 2003 when, de- tember 1999 to sell only 400 tons of gold per spite high prices, little new hedging took place. year, and not more than 2,000 tons in total, However, hedging of gold is unattractive at for the subsequent five years). current low interest rates. It is expected that producer dehedging will slow in the second half of this year and in Petroleum 2004, and remove much of the support under gold prices. And at some point, higher interest rates may trigger another bout of hedging. S ince late 1999, the average oil price (for Brent, Dubai, and WTI) has generally been above US$25/bbl, with the exception of the Higher gold prices have had a negative im- slump following the September 11, 2001, at- pact on consumer demand. In the first quarter tacks (figure A2.14). Excluding the slump, oil of 2003, jewelry demand fell by more than 10 prices averaged about US$27.1/bbl, compared percent (table A2.12), with declines in both to US$17.6/bbl over the 1986–99 period. The developing and developed regions. In the higher prices are mainly because of strong largest consuming country, India, demand fell production discipline by OPEC, but have also 13 percent, following a 20 percent drop in been supported by periods of low stocks, sup- 2002. High prices will continue to weaken the ply disruptions, and cold weather. price-sensitive jewelry demand market, and Following the collapse of prices in 1998, stimulate investment in new production, and OPEC began adjusting production quotas as from scrap. Over the medium term prices are required to maintain prices within a band of expected to fall below US$300/toz as supplies US$22–$28/bbl for its basket of crudes. By and from all sources exceed demand. Even below large the organization has been successful, US$300/toz, mine production is expected to though its market share has slowly eroded. For continue to increase moderately as new low- OPEC-10 (excluding Iraq), its crude oil pro- cost operations come on stream. duction as a share of total world oil supply fell 271 G L O B A L E C O N O M I C P R O S P E C T S output, and then oil inventories fell precipi- Figure A2.14 Oil price and OECD stocks tously after Venezuela’s oil exports ceased in $bbl mil bbl December because of strikes, and as cold 35 WB price 2,800 weather raised peak-winter demand. At end- 2,750 winter 2003, oil inventories were near historic ᮡ 30 2,700 lows. 2,650 With the loss of Venezuela’s production and impending loss of Iraq’s exports, other 25 2,600 OPEC producers raised production signifi- 2,550 cantly, particularly from the Gulf. Saudi Ara- 20 2,500 bia’s production rose from 7.7 mb/d in the Stocks ᮡ 2,450 fourth quarter of 2002 to more than 9.0 mb/d 15 2,400 by March 2003, and the rest of OPEC (ex- 2,350 cluding Venezuela and Iraq) added more than 1 mb/d over this period, with the largest in- 10 2,300 Jan. Jan. Jan. Jan. Jan. Jan. Jan. creases from Kuwait, UAE, and Algeria. At 1997 1998 1999 2000 2001 2002 2003 the same time, Venezuela’s production began Sources: International Energy Agency, World Bank. to recover, although it appears that some 0.4 mb/d of capacity was permanently lost as a re- sult of the strikes. from 35 percent in 1996–97 to 30 percent in The disruption to oil supplies from the war 2002. in Iraq was limited to Iraqi exports of about The escalation of prices in 2002 resulted 2 mb/d. Higher output from other OPEC from large OPEC production cuts (figure members was sufficient to prevent a sharp A2.15), augmented by expectations of supply spike in prices, and emergency stocks in con- disruption as the U.S.-led coalition prepared suming countries were not withdrawn. Oil for war in Iraq. The physical market tightened prices peaked in early March just before the in the second half of 2002 from lower OPEC conflict commenced at US$34.2/bbl. Figure A2.15 OPEC-10 production and quotas mb/d 30 29 Plus Iraq production ᮡ 28 27 26 OPEC-10 OPEC-10 production 25 quotas ᮡ 24 ᮡ 23 22 21 Jan. 1997 Jan. 1998 Jan. 1999 Jan. 2000 Jan. 2001 Jan. 2002 Jan. 2003 Sources: International Energy Agency, OPECNA. 272 G L O B A L C O M M O D I T Y P R I C E P R O S P E C T S Iraq’s exports did not restart soon after the mand, about 1.5 mb/d, will be captured by war ended because of widespread looting and strong gains in non-OPEC supply of more problems with pumping facilities and pipe- than 1 mb/d per year. Large increases are ex- lines. Because of broader problems with elec- pected from Russia, the Caspian Sea, West tricity, water, and other facilities that service Africa, and the Western Hemisphere, includ- the oil sector, it is unlikely that Iraq’s pre-war ing the U.S. because of significant develop- production of around 2.5 mb/d will be reached ments in the deepwater Gulf of Mexico. BP re- this year. ports that between 2002 and 2007, 5 mb/d of The delay in resumption of Iraqi exports new supply are likely to come on stream from and the low level of oil inventories eases the these regions alone. task for OPEC this year of maintaining prices This will leave little room for growth in within its band. However, the difficulty man- OPEC production. With the build-up of new aging oil prices is expected to deepen in 2004, capacity in many OPEC countries, including as Iraq oil exports exceed pre-war levels. Iraq, oil prices are expected to decline. By OPEC will have to absorb Iraq back into its 2006–07, oil prices are expected to fall to quota system at some point, and quotas for all US$18/bbl (figure A2.16) as significant vol- members may need to be adjusted. A number umes of new production begin from the Cas- of OPEC members are raising capacity and pian, and as production and export capacity will likely request higher quotas, e.g., Algeria increase more broadly from the FSU, West and Nigeria. The expansion of OPEC capacity Africa, and other regions. will occur when non-OPEC producers are ex- A risk to the forecast is that OPEC will pected to capture virtually all of the growth in maintain strong production discipline over the world oil demand. Consequently, oil prices are next few years to keep prices at or above expected to fall to the lower end of OPEC’s US$25/bbl. If successful, it would further im- price band in 2004. pact oil demand growth and stimulate even Downward pressures on oil prices are ex- greater supplies from competing sources. It is pected to continue in subsequent years, as felt that OPEC would only prolong a decline much of the moderate growth in world oil de- in oil prices that is expected by mid-decade. Figure A2.16 World Bank oil price $/bbl 50 45 40 1990$ 35 ᮡ 30 25 20 15 ᮡ 10 Nominal 5 0 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2019 2015 Source: World Bank. 273 G L O B A L E C O N O M I C P R O S P E C T S Table A2.13 Petroleum global balance (million barrels per day) Million barrels per day Annual growth rates (%) 1970 1980 1990 2000 2002 2003 1970–80 1980–90 1990–00 Consumption OECD 34.0 41.5 41.5 47.8 47.6 48.2 2.0 0.0 1.3 FSU 5.0 8.9 8.4 3.6 3.7 3.8 5.9 –0.6 –7.2 Other non-OECD 6.8 12.3 16.1 24.8 25.6 25.9 6.1 2.7 4.1 Total 45.7 62.6 66.0 76.2 76.9 77.9 3.2 0.5 1.3 Production OPEC 23.5 27.2 24.5 30.7 28.6 29.8 1.5 –1.0 1.9 FSU 7.1 12.1 11.5 7.9 9.4 10.2 5.4 –0.5 –2.6 Other non-OPEC 17.4 24.6 30.9 38.0 38.6 39.1 3.5 2.3 1.9 Total 48.0 63.9 66.9 76.6 76.5 79.1 2.9 0.5 1.3 Stock Change, Misc. 2.3 1.3 0.9 0.4 –0.3 1.2 Sources: British Petroleum, International Energy Agency, and World Bank. In the longer term, demand growth will ture. In addition, new areas continue to be de- only be moderate, as it has been the past two veloped (e.g., deep water offshore), and devel- decades (table A2.13), but new technologies, opment costs are expected to continue to fall environmental pressures, and government poli- from new technologies (shifting supply curves cies could further reduce this growth. Prices outward). In addition, OPEC countries are in- below US$20/bbl are sufficiently high to gen- creasing capacity, and will add to the supply erate ample development of conventional and competition in coming years. Consequently, non-conventional oil supplies, and there are no real oil prices are expected to continue their apparent resource constraints far into the fu- long-term decline. 274 G L O B A L C O M M O D I T Y P R I C E P R O S P E C T S Table A2.14 Commodity prices and price projections in current dollars Actual Projections Commodity Unit 1970 1980 1990 2000 2002 2003 2004 2005 2010 2015 Energy Coal, Australia $/mt n.a. n.a. 39.67 26.25 27.06 26.00 26.50 27.00 29.50 32.00 Crude oil, average $/bbl 1.21 36.87 22.88 28.23 24.93 26.50 22.00 20.00 19.50 22.00 Natural gas, Europe $/mmbtu n.a. 3.40 2.55 3.86 3.05 3.75 3.00 2.65 2.75 3.00 Natural gas, US $/mmbtu 0.17 1.55 1.70 4.31 3.35 5.25 3.75 3.50 3.25 3.50 Non-Energy Commodities Agriculture Beverages Cocoa c/kg 67.5 260.4 126.7 90.6 177.8 177.0 172.0 167.0 160.0 150.0 Coffee, other milds c/kg 114.7 346.6 197.2 192.0 135.7 145.5 158.7 165.4 210.1 230.4 Coffee, robusta c/kg 91.4 324.3 118.2 91.3 66.2 88.2 92.6 92.6 104.7 125.0 Tea, auctions (3) average c/kg 83.5 165.9 205.8 187.6 150.6 150.0 155.0 160.0 170.0 170.0 Food Fats and oils Coconut oil $/mt 397.2 673.8 336.5 450.3 421.0 442.0 460.0 470.0 500.0 530.0 Copra $/mt 224.8 452.7 230.7 304.8 266.3 305.0 380.0 420.0 450.0 475.0 Groundnut oil $/mt 378.6 858.8 963.7 713.7 687.1 1100.0 1000.0 890.0 795.0 796.0 Palm oil $/mt 260.1 583.7 289.8 310.3 390.3 425.0 415.0 415.0 420.0 445.0 Soybean meal $/mt 102.6 262.4 200.2 189.2 175.2 193.0 183.0 175.0 185.0 195.0 Soybean oil $/mt 286.3 597.6 447.3 338.1 454.3 527.0 485.0 450.0 460.0 480.0 Soybeans $/mt 116.9 296.2 246.8 211.8 212.7 241.0 225.0 210.0 225.0 235.0 Grains Maize $/mt 58.4 125.3 109.3 88.5 99.3 106.0 100.0 95.0 105.0 112.0 Rice, Thailand, 5% $/mt 126.3 410.7 270.9 202.4 191.9 199.0 202.0 205.0 220.0 230.0 Sorghum $/mt 51.8 128.9 103.9 88.0 101.7 106.0 100.0 95.0 105.0 112.0 Wheat, US, HRW $/mt 54.9 172.7 135.5 114.1 148.1 143.0 135.0 130.0 145.0 155.0 Other food Bananas, US $/mt 166.1 377.3 540.9 424.0 528.6 410.0 425.0 440.0 530.0 555.0 Beef, US c/kg 130.4 276.0 256.3 193.2 212.7 211.6 218.2 220.5 222.0 220.0 Oranges $/mt 168.0 400.2 531.1 363.2 555.0 645.0 550.0 500.0 510.0 530.0 Shrimp, Mexico c/kg n.a. 1,152 1,069 1,513 1,052 1,200 1,275 1,350 1,550 1,650 Sugar, world c/kg 8.2 63.16 27.67 18.04 15.18 16.00 15.40 15.00 19.00 21.00 Agricultural raw materials Timber Logs, Cameroon $/cum 43.0 251.7 343.5 275.4 n.a. 275.0 280.0 285.0 320.0 350.0 Logs, Malaysia $/cum 43.1 195.5 177.2 190.0 163.4 185.0 188.0 205.0 245.0 265.0 Sawnwood, Malaysia $/cum 175.0 396.0 533.0 594.7 526.5 550.0 570.0 610.0 700.0 780.0 Other raw materials Cotton c/kg 67.6 206.2 181.9 130.2 101.9 130.1 129.0 132.3 141.1 143.3 Rubber, RSS1, Malaysia c/kg 40.7 142.5 86.5 69.1 77.1 95.0 90.0 94.8 88.2 90.4 Tobacco $/mt 1,076 2,276 3,392 2,976 2,740 2,700 2,750 2,800 2,950 3,000 Fertilizers DAP $/mt 54.0 222.2 171.4 154.2 157.5 177.0 175.0 170.0 170.0 175.0 Phosphate rock $/mt 11.00 46.71 40.50 43.75 40.38 38.00 38.00 38.00 40.00 42.00 Potassium chloride $/mt 32.0 115.7 98.1 122.5 113.3 112.5 115.0 116.0 118.0 120.0 TSP $/mt 43.0 180.3 131.8 137.7 133.1 142.0 140.0 140.0 146.0 154.0 Urea, E. Europe, bagged $/mt n.a. n.a. 119.3 101.1 94.4 130.0 128.0 126.7 125.0 130.0 Metals and minerals Aluminum $/mt 556 1,456 1,639 1,549 1,350 1,390 1,425 1,500 1,600 1,700 Copper $/mt 1,416 2,182 2,661 1,813 1,559 1,650 1,800 1,900 2,000 2,050 Gold $/toz 35.9 607.9 383.5 279.0 310.0 330.0 300.0 280.0 300.0 300.0 Iron ore, Carajas c/dmtu 9.84 28.09 32.50 28.79 29.31 31.95 32.00 31.00 32.00 32.50 Lead c/kg 30.3 90.6 81.1 45.4 45.3 47.0 51.0 55.0 60.0 62.5 Nickel $/mt 2,846 6,519 8,864 8,638 6,772 8,200 8,500 8,000 6,700 6,800 Silver c/toz 177.0 2,064 482.0 499.9 462.5 460.0 480.0 500.0 525.0 550.0 Tin c/kg 367.3 1,677 608.5 543.6 406.1 470.0 500.0 525.0 540.0 550.0 Zinc c/kg 29.6 76.1 151.4 112.8 77.9 80.0 92.0 100.0 105.0 110.0 n.a. = Not available. Note: Projections as of June 24, 2003 Source: World Bank, Development Prospects Group. 275 G L O B A L E C O N O M I C P R O S P E C T S Table A2.15 Commodity prices and price projections in constant 1990 dollars Actual Projections Commodity Unit 1970 1980 1990 2000 2002 2003 2004 2005 2010 2015 Energy Coal, Australia $/mt n.a. n.a. 39.67 26.97 28.06 26.87 27.49 27.59 28.84 29.93 Crude oil, average $/bbl 4.31 46.80 22.88 29.01 25.84 27.39 22.82 20.44 19.06 20.58 Natural gas, Europe $/mmbtu n.a. 4.32 2.55 3.96 3.16 3.88 3.11 2.71 2.69 2.81 Natural gas, US $/mmbtu 0.61 1.97 1.70 4.43 3.48 5.43 3.89 3.58 3.18 3.27 Non-Energy Commodities Agriculture Beverages Cocoa c/kg 240.6 330.5 126.7 93.1 184.3 183.0 178.4 170.6 156.4 140.3 Coffee, other milds c/kg 408.8 440.0 197.2 197.3 140.7 150.4 164.7 169.0 205.4 215.5 Coffee, robusta c/kg 325.7 411.7 118.2 93.8 68.6 91.2 96.1 94.6 102.4 116.9 Tea, auctions (3) average c/kg 297.7 210.6 205.8 192.8 156.1 155.0 160.8 163.5 166.2 159.0 Food Fats and oils Coconut oil $/mt 1416.0 855.3 336.5 462.7 436.5 456.9 477.2 480.2 488.8 495.7 Copra $/mt 801.6 574.7 230.7 313.1 276.1 315.3 394.2 429.1 439.9 444.3 Groundnut oil $/mt 1349.5 1090.1 963.7 733.3 712.4 1137.0 1037.5 909.4 777.1 744.6 Palm oil $/mt 927.1 740.9 289.8 318.8 404.6 439.3 430.5 424.0 410.6 416.2 Soybean meal $/mt 365.7 333.1 200.2 194.4 181.6 199.5 189.9 178.8 180.8 182.4 Soybean oil $/mt 1020.8 758.6 447.3 347.4 471.0 544.7 503.2 459.8 449.7 449.0 Soybeans $/mt 416.8 376.0 246.8 217.7 220.5 249.1 233.4 214.6 219.9 219.8 Grains Maize $/mt 208.2 159.0 109.3 91.0 102.9 109.6 103.8 97.1 102.6 104.8 Rice, Thailand, 5% $/mt 450.3 521.4 270.9 208.0 198.9 205.7 209.6 209.5 215.1 215.1 Sorghum $/mt 184.7 163.6 103.9 90.4 105.5 109.6 103.8 97.1 102.6 104.8 Wheat, US, HRW $/mt 195.7 219.3 135.5 117.2 153.5 147.8 140.1 132.8 141.7 145.0 Other food Bananas, US $/mt 592.1 478.9 540.9 435.7 548.0 423.8 440.9 449.6 518.1 519.1 Beef, US c/kg 465.0 350.3 256.3 198.5 220.6 218.7 226.4 225.3 217.0 205.8 Oranges $/mt 599.1 508.0 531.1 373.2 575.5 666.7 570.6 510.9 498.5 495.7 Shrimp, Mexico c/kg n.a. 1,462 1,069 1,554 1,090 1,240 1,323 1,379 1,515 1,543 Sugar, world c/kg 29.32 80.17 27.67 18.5 15.7 16.5 16.0 15.3 18.6 19.6 Agricultural raw materials Timber Logs, Cameroon $/cum 153.3 319.5 343.5 283.0 n.a. 284.2 290.5 291.2 312.8 327.4 Logs, Malaysia $/cum 153.8 248.2 177.2 195.2 169.4 191.2 195.0 209.5 239.5 247.9 Sawnwood, Malaysia $/cum 623.9 502.7 533.0 611.1 545.9 568.5 591.4 623.3 684.3 729.6 Other raw materials Cotton c/kg 241.1 261.7 181.9 133.8 105.7 134.4 133.8 135.2 137.9 134.0 Rubber, RSS1, Malaysia c/kg 145.2 180.8 86.5 71.0 79.9 98.2 93.4 96.9 86.2 84.6 Tobacco $/mt 3,836 2,889 3,392 3,058 2,841 2,791 2,853 2,861 2,884 2,806 Fertilizers DAP $/mt 192.5 282.1 171.4 158.5 163.3 183.0 181.6 173.7 166.2 163.7 Phosphate rock $/mt 39.2 59.3 40.5 45.0 41.9 39.3 39.4 38.8 39.1 39.3 Potassium chloride $/mt 114.1 146.9 98.1 125.9 117.5 116.3 119.3 118.5 115.4 112.2 TSP $/mt 153.3 228.8 131.8 141.5 138.0 146.8 145.2 143.1 142.7 144.1 Urea, E. Europe, bulk $/mt n.a. n.a. 119.3 103.9 97.8 134.4 132.8 129.5 122.2 121.6 Metals and minerals Aluminum $/mt 1,982 1,848 1,639 1,592 1,400 1,437 1,478 1,533 1,564 1,590 Copper $/mt 5,047 2,770 2,661 1,863 1,617 1,705 1,867 1,941 1,955 1,918 Gold $/toz 128.1 771.6 383.5 286.7 321.4 341.1 311.2 286.1 293.3 280.6 Iron ore c/dmtu 35.1 35.7 32.5 29.6 30.4 33.0 33.2 31.7 31.3 30.4 Lead c/kg 108.0 115.0 81.1 46.6 46.9 48.6 52.9 56.2 58.7 58.5 Nickel $/mt 10,147 8,275 8,864 8,876 7,021 8,475 8,818 8,174 6,549 6,360 Silver c/toz 631.0 2619.4 482.0 513.7 479.5 475.5 498.0 510.9 513.2 514.5 Tin c/kg 1309.6 2129.3 608.5 558.5 421.0 485.8 518.7 536.4 527.9 514.5 Zinc c/kg 105.5 96.6 151.4 115.9 80.7 82.7 95.5 102.2 102.6 102.9 n.a. = Not available. Note: Projections as of June 24, 2003 Source: World Bank, Development Prospects Group. 276 G L O B A L C O M M O D I T Y P R I C E P R O S P E C T S Table A2.16 Weighted indices of commodity prices and inflation Actual Projectionsa Index 1970 1980 1990 2000 2002 2003 2004 2005 2010 2015 Current dollars Petroleum 5.3 161.2 100.0 123.4 109.0 115.8 96.2 87.4 85.2 96.2 Non-energy commoditiesb 43.8 125.5 100.0 86.9 83.0 88.8 89.7 91.1 97.7 102.6 Agriculture 45.8 138.1 100.0 87.7 86.5 92.7 92.6 93.7 102.0 107.6 Beverages 56.9 181.4 100.0 88.4 84.6 89.1 92.4 93.6 106.1 111.6 Food 46.7 139.3 100.0 84.5 90.1 94.0 91.7 89.8 96.5 101.1 Fats and oils 64.4 148.7 100.0 96.2 101.2 112.4 107.7 104.2 108.4 114.1 Grains 46.7 134.3 100.0 79.5 88.1 89.8 87.0 85.0 93.2 98.8 Other food 32.2 134.3 100.0 77.7 82.2 81.2 81.2 80.8 88.7 91.8 Raw materials 36.4 104.6 100.0 91.4 83.2 93.6 93.8 98.8 106.1 113.0 Timber 31.8 79.0 100.0 111.0 98.1 103.3 106.8 114.6 132.2 146.7 Other Raw Materials 39.6 122.0 100.0 78.0 73.1 86.9 84.9 88.0 88.3 90.0 Fertilizers 30.4 128.9 100.0 105.8 100.5 100.5 102.6 101.7 105.4 110.9 Metals and minerals 40.4 94.2 100.0 83.0 72.8 78.1 81.6 83.6 86.4 89.5 Constant 1990 dollarsc Petroleum 18.9 204.5 100.0 127.0 117.2 119.7 99.8 89.3 83.3 89.9 Non-energy commodities 156.3 159.2 100.0 89.4 89.3 91.7 93.1 93.1 95.5 96.0 Agriculture 163.3 175.2 100.0 90.3 93.0 95.8 96.0 95.8 99.7 100.7 Beverages 202.8 230.2 100.0 90.9 91.0 92.1 95.9 95.7 103.7 104.4 Food 166.5 176.7 100.0 87.0 96.9 97.2 95.1 91.8 94.4 94.6 Fats and oils 229.5 188.6 100.0 99.0 108.8 116.2 111.7 106.4 105.9 106.7 Grains 166.6 170.4 100.0 81.8 94.7 92.8 90.2 86.9 91.1 92.4 Other food 114.9 170.5 100.0 80.0 88.4 84.0 84.2 82.5 86.7 85.9 Raw materials 129.8 132.7 100.0 94.0 89.5 96.7 97.3 100.9 103.7 105.7 Timber 113.3 100.3 100.0 114.2 105.5 106.8 110.8 117.1 129.2 137.3 Other Raw Materials 141.1 154.8 100.0 80.2 78.6 89.8 88.1 89.9 86.3 84.2 Fertilizers 108.3 163.6 100.0 108.9 108.1 103.9 106.5 103.9 103.0 103.7 Metals and minerals 143.9 119.5 100.0 85.4 78.3 80.7 84.6 85.5 84.5 83.7 Inflation indices, 1990=100d MUV indexe 28.05 78.81 100.00 97.17 92.99 96.75 96.39 97.87 102.30 106.91 % change per annum 10.88 2.41 -0.29 –2.18 4.05 -0.37 1.53 0.89 0.88 US GDP deflator 33.59 65.93 100.00 123.56 127.91 129.96 132.69 135.87 152.83 172.07 % change per annum 6.98 4.25 2.14 1.75 1.60 2.10 2.40 2.38 2.40 a. Commodity price projections as of June 24, 2003. b. The World Bank primary commodity price indices are computed based on 1987–89 export values in US dollars for low- and middle-income economies, rebased to 1990. Weights for the sub-group indices expressed as ratios to the non-energy index are as follows in percent: agriculture 69.1, fertilizers 2.7, metals and minerals 28.2; beverages 16.9, food 29.4, raw materials 22.8; fats and oils 10.1, grains 6.9, other food 12.4; timber 9.3 and other raw mterials 13.6. c. Computed from unrounded data and deflated by the MUV index d. Inflation indices for 2002–2015 are projections as of June 10, 2003. MUV for 2001 is an estimate. Growth rates for years 1980, 1990, 2000, 2002, 2005, 2010 and 2015 refer to compound annual rate of change between adjacent end-point years; all others are annual growth rates from the previous year. e. Unit value index in US dollar terms of manufactures exported from the G-5 countries (France, Germany, Japan, UK, and US) weighted proportionally to the countries’ exports to the developing countries Source: World Bank, Development Prospects Group. Historical US GDP deflator: US Department of Commerce. 277 Appendix 3 Global Economic Indicators G L O B A L E C O N O M I C P R O S P E C T S Table A3.1 Growth of Real GDP, 1971–2015 GDP in 1995 prices and exchange rates, average annual growth (percent) GDP in 2002 Growth percent (current billions Estimate Forecast of dollars) 1971–80 1981–90 1991–00 2002 2003 2004–2005 2006–2015 World 32,016 3.7 3.0 2.6 1.9 2.0 2.9 3.2 High income economies 25,937 3.5 3.1 2.5 1.6 1.5 2.5 2.7 Industrial countries 25,190 3.4 3.1 2.4 1.6 1.5 2.4 2.6 G-7 countries 21,350 3.4 3.1 2.3 1.4 1.4 2.3 ... United States 10,446 3.3 3.2 3.2 2.4 ... ... ... Japan 3,997 4.5 4.1 1.4 0.2 ... ... ... G-4 Europe 6,172 2.9 2.4 1.9 0.8 0.7 1.9 ... Germanya 1,990 2.7 2.2 1.7 0.2 ... ... ... Euro Area 6,633 3.2 2.3 2.0 0.9 0.7 1.9 2.2 Non-G-7 industrial 3,840 3.2 2.9 3.0 2.4 1.8 3.0 ... Other high income 747 7.6 4.9 5.4 2.4 2.1 4.3 4.5 Asian NIEs 531 9.5 7.4 6.0 3.0 2.2 4.6 ... Low and middle income economies 6,079 5.0 2.6 3.3 3.3 4.0 4.9 4.6 excluding CE.Eur / CIS 5,147 5.6 3.1 4.7 3.0 4.0 5.0 ... Asia 2,403 5.3 6.9 7.0 6.1 5.9 6.3 ... East Asia and Pacific 1,757 6.7 7.3 7.7 6.7 6.1 6.6 6.2 China 1,237 6.2 9.3 10.1 8.0 ... ... ... Indonesia 173 7.9 6.4 4.2 3.8 ... ... ... South Asia 646 3.1 5.9 5.2 4.2 5.4 5.4 5.4 India 515 3.0 6.1 5.4 4.4 ... ... ... Latin America and Caribbean 1,658 5.9 1.1 3.4 –0.8 1.8 3.8 3.8 Brazil 452 8.5 1.5 2.7 1.5 ... ... ... Mexico 637 6.7 1.8 3.5 1.0 ... ... ... Argentina 102 3.0 –1.5 4.5 –10.9 ... ... ... Europe and Central Asia 1,114 3.9 1.5 –1.6 4.6 4.3 4.3 3.4 Russian Federationb 347 3.9 1.5 –4.0 4.3 ... ... ... Turkey 183 4.1 5.2 3.6 7.8 ... ... ... Poland 188 5.1 –1.7 3.7 1.3 ... ... ... Middle East and North Africa 587 7.0 2.5 3.4 3.1 3.3 3.7 4.3 Saudi Arabia 193 10.3 0.3 2.3 0.8 ... ... ... Iran 108 2.6 2.7 4.2 6.3 ... ... ... Egypt 90 6.6 5.5 4.3 3.0 ... ... ... Sub-Saharan Africa 316 3.5 1.7 2.3 2.8 2.8 3.7 3.5 Republic of South Africa 104 3.5 1.3 1.7 3.0 ... ... ... Nigeria 44 4.7 1.1 2.6 1.9 ... ... ... Notes: growth rates over intervals are computed using compound average methods; a. data prior to 1991 covers West Germany b. data prior to 1992 covers former Soviet Union Figure A3.1 Real GDP growth Percent 8.0 1991–2000 2006–2015 6.0 ᮡ ᮡ 4.0 2.0 0.0 –2.0 High-income East Asia South Latin America Europe and Middle East Sub- countries and Pacific Asia and the Central Asia and North Saharan Caribbean Africa Africa Source: World Bank data and staff estimates. 280 G L O B A L E C O N O M I C I N D I C A T O R S Table A3.2 Growth of real per-capita GDP, 1971–2015 GDP in 1995 prices and exchange rates, average annual growth (percent) GDP per capita Growth percent 2002 (current Estimate Forecast dollars) 1971–80 1981–90 1991–00 2002 2003 2004–2005 2006–2015 World 5,297 1.8 1.3 1.2 0.7 0.8 1.8 2.2 High income economiesa 27,185 2.6 2.5 1.8 1.2 1.1 2.1 2.5 Industrial countries 27,710 2.6 2.5 1.8 1.2 1.1 2.1 2.4 G-7 countries 30,256 2.7 2.5 1.7 1.0 1.1 2.0 … United States 36,224 2.2 2.2 2.2 1.7 … … … Japan 31,437 3.3 3.5 1.2 0.1 … … … G-4 Europe 23,856 2.6 2.1 1.5 0.8 0.7 1.9 … Germanyb 24,123 2.6 2.0 1.4 0.3 … … … Euro Area 21,721 2.7 2.1 1.7 0.8 0.7 1.9 2.3 Non-G-7 industrial 18,879 2.2 2.2 2.4 2.0 1.5 2.8 … Other high income 16,577 5.0 3.1 3.8 1.1 0.9 3.1 4.2 Asian NIEs 15,891 7.2 5.9 4.7 2.0 1.3 3.7 … Low & middle income economies 1,194 2.9 0.7 1.7 1.9 2.7 3.6 3.4 excluding CE.Eur / CIS 1,066 3.2 1.0 3.0 1.5 2.5 3.6 … Asia 742 3.0 4.9 5.4 4.8 4.6 5.1 … East Asia and Pacific 956 4.6 5.6 6.4 5.8 5.2 5.7 5.4 China 966 4.3 7.7 9.0 7.2 … … … Indonesia 817 5.4 4.4 2.5 2.4 … … … South Asia 461 0.7 3.6 3.3 2.5 3.7 3.8 4.1 India 491 0.7 3.9 3.6 2.8 … … … Latin America and Caribbean 3,149 3.3 –0.9 1.7 –2.3 0.4 2.4 2.5 Brazil 2,593 5.9 –0.4 1.2 0.3 … … … Mexico 6,314 3.6 –0.3 1.7 –0.8 … … … Argentina 2,694 1.3 –2.9 3.2 –12.0 … … … Europe and Central Asia 3,374 2.9 0.6 –1.8 4.5 4.3 4.2 3.3 Russian Federationc 2,405 3.2 0.9 –3.9 4.6 … … … Turkey 2,626 1.7 2.8 2.0 6.3 … … … Poland 4,859 4.2 –2.4 3.5 1.2 … … … Middle East and North Africa 1,917 4.0 –0.6 1.2 1.1 1.3 1.7 2.5 Saudi Arabia 8,727 5.1 –4.8 –1.1 –2.2 … … … Iran 1,641 –0.6 –0.7 2.5 4.6 … … … Egypt 1,354 4.4 2.9 2.4 1.4 … … … Sub-Saharan Africa 460 0.7 –1.1 –0.2 0.5 0.6 1.5 1.6 Republic of South Africa 2,392 1.2 –1.2 –0.2 1.7 … … … Nigeria 328 1.7 –1.9 –0.2 –0.8 … … … Notes: growth rates over intervals are computed using compound annual methods; a. regional aggregates computed as sum(GDPi)/sum(POPi), where “i” indicates country in the region, and are unweighted by pop- ulation or other measures; b. data prior to 1991 covers West Germany; c. data prior to 1992 covers former Soviet Union Figure A3.2 Real per capita GDP growth Percent 8.0 6.0 1991–2000 2006–2015 ᮡ 4.0 ᮡ 2.0 0.0 –2.0 High-income East Asia South Latin America Europe and Middle East Sub- countries and Pacific Asia and the Central Asia and North Saharan Caribbean Africa Africa Source: World Bank data and staff estimates. 281 G L O B A L E C O N O M I C P R O S P E C T S Table A3.3 Inflation: GDP Deflators, 1971–2005 Deflators in local currency units; 1995=100; percentage changea Growth percent Estimate Forecast 1971–80 1981–90 1991–00 2001 2002 2003 2004–2005 World 9.1 5.9 3.7 2.2 1.5 1.6 1.8 High income economies 8.9 5.3 2.0 1.5 1.0 1.0 1.2 Industrial countries 8.7 4.6 2.0 1.5 1.0 1.0 1.2 G-7 countries 8.3 4.2 1.7 1.2 0.8 0.8 1.0 United States 7.0 4.3 2.1 2.4 1.1 1.5 1.5 Japan 7.8 2.0 0.1 –1.6 –1.6 –2.1 –0.5 G-4 Europe 9.9 5.7 2.6 1.9 2.2 1.8 1.5 Germanyb 5.3 2.6 2.6 1.4 1.6 1.2 0.9 Euro Area 9.6 6.1 2.8 2.4 2.3 1.9 1.5 Non-G-7 industrial 11.1 7.1 3.3 3.1 2.3 2.4 2.1 Other high income 19.6 33.7 3.7 0.1 –0.6 0.5 1.8 Asian NIEs 9.5 4.7 2.4 –0.4 –1.4 –0.7 1.3 Low and middle income economies 9.8 8.8 11.8 5.2 4.0 4.0 4.1 excluding CE.Eur / CIS 11.6 10.7 9.4 4.8 4.1 4.0 4.2 Asia 10.2 7.7 7.6 2.9 2.6 4.4 4.9 East Asia and Pacific 9.8 6.6 6.8 2.6 3.3 3.7 3.7 China 0.9 5.4 6.3 0.0 0.4 ... ... Indonesia 20.6 8.8 14.9 12.6 7.2 ... ... South Asia 11.9 9.0 7.8 3.0 1.8 5.1 5.2 India 8.9 8.5 8.0 3.0 2.6 ... ... Latin America and Caribbean 15.6 20.4 12.0 5.5 4.7 4.1 4.0 Brazil 39.7 330.8 206.1 8.8 8.5 ... ... Mexico 18.1 63.7 18.1 5.4 4.9 ... ... Argentina 117.7 439.5 10.2 –1.1 30.8 ... ... Europe and Central Asia 0.2 1.0 76.0 6.1 4.0 5.8 4.6 Russian Federationc 0.0 31.6 104.5 18.0 10.7 ... ... Turkey 32.8 46.4 71.7 57.1 48.2 ... ... Poland 4.4 72.1 24.6 3.9 –3.5 ... ... Middle East and North Africa 11.7 8.4 7.5 2.6 3.8 4.0 4.0 Saudi Arabia 23.8 –3.1 3.8 –2.4 2.4 ... ... Iran 19.3 15.6 26.6 8.8 5.2 ... ... Egypt 11.0 13.1 8.7 3.8 4.0 ... ... Sub-Saharan Africa 10.7 10.0 10.5 5.8 4.2 3.7 4.5 Republic of South Africa 13.3 15.1 9.9 7.6 8.5 ... ... Nigeria 13.4 16.6 28.6 6.0 15.0 ... ... Notes: growth rates over intervals are computed using compound annual squares method; a. High-income group inflation rates are GDP-weighted averages of local currency inflation; LMIC groups are medians; world is GDP-weighted average of the two groups. b. data prior to 1991 covers West Germany c. data prior to 1992 covers former Soviet Union. Figure A3.3 GDP inflation Percent 15.0 76% 10.0 1991–2000 ᮡ 2004–2005 5.0 ᮡ 0.0 High-income East Asia South Latin America Europe and Middle East Sub- countries and Pacific Asia and the Central Asia and North Saharan Caribbean Africa Africa Source: World Bank data and staff estimates. 282 G L O B A L E C O N O M I C I N D I C A T O R S Table A3.4 Current Account Balances, 1971–2005 Expressed as shares of GDP (percent) 2002 Current Shares percent Account (billions of Estimate Forecast U.S. dollars) 1971–80 1981–90 1991–00 2001 2002 2003 2004–2005 Worlda –106 –0.1 –0.6 –0.2 –0.5 –0.3 –0.5 –0.6 High income economies –177 –0.1 –0.3 0.0 –0.6 –0.7 –0.7 –0.7 Industrial countries –238 –0.2 –0.5 –0.1 –0.9 –0.9 –0.9 –0.9 G-7 countries –304 –0.1 –0.4 –0.3 –1.4 –1.4 –1.4 –1.4 United States –481 0.0 –1.9 –1.8 –3.9 –4.6 ... ... Japan 112 0.3 2.3 2.4 2.1 2.8 ... ... G-4 Europe 54 0.1 0.3 –0.2 0.1 0.9 1.0 1.1 Germanyb 46 0.5 2.4 –0.9 0.0 2.3 ... ... Euro Area 54 0.0 0.3 0.5 0.5 0.8 1.1 1.3 Non-G-7 industrial 65 –1.1 –0.8 1.2 1.9 1.7 1.8 2.0 Other high income 62 7.6 6.9 3.2 7.2 8.3 6.8 6.4 Asian NIEs 62 1.8 1.2 4.1 9.0 11.6 10.6 10.5 Low & middle income economies 70 0.2 –1.7 –1.6 0.1 1.2 0.5 –0.1 excluding CE.Eur / FSU 58 0.6 –1.8 –1.6 0.2 1.1 0.3 –0.2 Asia 59 –0.3 –1.7 –0.2 1.8 2.5 1.5 1.2 East Asia and Pacific 56 –0.1 –1.5 0.4 2.7 3.2 1.9 1.6 China 29 0.1 0.2 1.6 1.5 2.4 ... ... Indonesia 7 –2.3 –3.3 –0.4 4.7 4.2 ... ... South Asia 3 –0.4 –2.0 –1.5 –0.5 0.5 0.4 0.0 India 3 0.2 –1.7 –1.1 –0.7 0.7 ... ... Latin America and Caribbean –14 –1.4 –1.7 –2.7 –2.7 –0.8 –0.5 –0.9 Brazil –8 –4.4 –1.6 –2.2 –4.6 –1.7 ... ... Mexico –14 –3.9 –0.8 –3.7 –2.9 –2.2 ... ... Argentina 9 –0.4 –2.2 –3.1 –1.7 8.8 ... ... Europe and Central Asia 13 –0.8 –0.5 –2.5 –1.4 1.1 0.6 –0.2 Russian Federationc 33 2.1 3.5 4.7 11.3 9.5 ... ... Turkey 0 –2.1 –1.3 –1.1 2.3 0.0 ... ... Poland –7 –0.9 –1.4 –3.7 –3.0 –3.5 ... ... Middle East and North Africa 20 6.8 –1.5 –1.9 4.2 3.5 0.3 –2.0 Saudi Arabia 12 21.1 –7.3 –6.5 7.6 6.2 ... ... Iran 4 11.8 –0.4 1.9 3.5 3.1 ... ... Egypt –1 –0.6 –2.6 1.5 –0.4 –1.3 ... ... Sub-Saharan Africa –8 –2.1 –2.7 –2.0 –2.2 –2.2 –2.7 –2.6 Republic of South Africa 0 –1.3 0.4 –0.2 –0.3 0.3 ... ... Nigeria 0 1.5 –0.6 –0.2 –0.7 0.0 ... ... a. Current account as defined in BOP version 5.0, world represents statistical discrepancy; shares over intervals are period averages; b. data prior to 1991 covers West Germany; c. data prior to 1992 covers former Soviet Union. Figure A3.4 Current account balances-to-GDP ratio Percent of GDP 2004–2005 2.0 1991–2000 ᮡ ᮡ 1.0 0.0 –1.0 –2.0 –3.0 –4.0 High-income East Asia South Latin America Europe and Middle East Sub- countries and Pacific Asia and the Central Asia and North Saharan Caribbean Africa Africa Source: World Bank data and staff estimates. 283 G L O B A L E C O N O M I C P R O S P E C T S Table A3.5 Exports of goods, 2002 Merchandise exports (FOB), millions of dollars; average annual volume growth 1993–2002 (percent); effective market growth (EMG) 1993–2002 (percent) Exports Growth EMG2 Exports Growth EMG2 Exports Growth EMG2 World 6,285,029 6.3 6.6 Europe and High income Central Asia (continued) countries 4,666,260 6.1 6.9 All developing Belarus 8,035 –2.9 7.8 countries 1,618,770 7.1 7.2 Bulgaria 3,512 5.1 6.3 Industrial Czech Republic 26,977 4.5 6.3 countries 4,431,191 6.2 6.8 Asia 663,348 11.4 7.0 Estonia 5,992 11.0 7.6 Georgia 748 24.5 6.5 G-7 countries 2,857,740 5.4 7.0 East Asia 595,377 12.0 7.0 Hungary 27,305 3.9 6.1 Canada 263,777 7.3 8.9 China 325,574 17.3 6.8 Kazakstan 10,590 17.2 8.0 France 339,050 6.1 5.9 Fiji 511 3.6 7.3 Latvia 2,718 7.4 8.5 Germany 592,010 5.9 6.6 Indonesia 56,922 5.1 7.0 Lithuania 4,365 2.4 7.9 Italy 284,552 5.5 6.2 Malaysia 93,364 8.2 7.1 Poland 36,115 13.5 6.0 Japan 395,662 3.6 7.8 Philippines 34,455 10.4 7.4 Romania 12,820 1.2 5.8 United Kingdom 279,140 5.4 6.4 Thailand 68,902 8.0 7.1 Russian United States 703,549 5.3 7.0 Vietnam 16,159 18.4 6.9 Federation 79,059 –0.4 7.9 Slovak Republic 18,023 7.1 8.7 Other industrial 1,253,846 7.4 6.2 South Asia 67,972 7.1 6.9 Turkey 35,904 8.4 6.0 Australia 64,988 6.2 6.5 Bangladesh 6,374 11.6 7.0 Ukraine 19,260 –0.3 7.5 Austria 69,422 7.9 5.9 India 46,129 7.6 6.8 Uzbekistan 7,005 4.3 7.2 Belgium1 200,696 6.7 5.9 Nepal 733 8.2 6.4 Denmark 53,531 2.1 5.8 Pakistan 9,699 3.0 6.8 Middle East and Finland 7,507 7.7 7.0 Sri Lanka 5,036 7.4 7.2 N. Africa 161,597 2.3 6.6 Greece 3,158 10.8 5.8 Algeria 17,648 2.7 6.4 Iceland 2,150 2.8 6.3 Latin America 351,966 5.6 7.7 Egypt, Ireland 88,970 9.7 6.2 Argentina 25,348 2.8 6.7 Arab Rep. 7,323 6.1 6.3 Korea, Rep. 162,412 14.4 7.5 Bolivia 1,241 7.1 5.9 Iran, Islamic Rep. 24,203 –0.3 6.4 Netherlands 198,624 6.3 5.8 Brazil 60,362 –2.9 5.9 Jordan 2,466 7.0 5.8 New Zealand 14,860 3.7 7.1 Chile 18,340 5.1 6.7 Morocco 7,427 3.6 6.1 Norway 61,761 2.8 6.3 Colombia 12,803 4.5 7.3 Oman 11,508 5.0 7.9 Portugal 26,731 9.2 5.6 Costa Rica 5,007 11.0 7.8 Saudi Arabia 73,691 1.3 6.8 Spain 121,572 7.7 6.1 Dominican Syrian Arab Rep. 6,173 5.1 5.4 Sweden 85,304 3.7 6.1 Republic 5,142 24.5 7.9 Tunisia 7,676 5.0 5.5 Switzerland 92,161 3.7 6.1 Ecuador 5,279 3.9 7.7 Yemen, Rep. 3,483 9.4 8.8 El Salvador 2,526 17.2 7.8 Other high income 554,674 7.0 8.3 Guatemala 2,850 7.4 8.7 Sub-Saharan Bahrain 5,529 3.4 5.6 Jamaica 1,466 2.4 7.7 Africa 91,060 3.1 6.8 Hong Kong, Mexico 160,834 13.5 9.0 Angola 7,582 4.4 8.7 China 200,285 6.3 10.2 Panama 5,778 1.2 7.3 Cameroon 1,917 –1.2 6.2 Israel 29,435 7.6 7.3 Paraguay 2,176 –0.4 6.3 Côte d’Ivoire 4,275 2.8 5.6 Kuwait 15,477 6.2 6.7 Peru 7,551 –0.4 7.3 Ethiopia 326 10.3 5.1 Singapore 125,172 8.8 7.5 Trinidad and Gabon 2,403 –2.1 8.3 Taiwan, China 130,666 6.9 7.1 Tobago 4,214 7.1 8.2 Ghana 1,402 6.9 6.4 United Arab Uruguay 1,850 8.4 6.7 Kenya 2,028 5.6 5.7 Emirates 38,386 3.1 5.6 Venezuela 26,735 –0.3 8.8 Madagascar 962 10.9 5.9 Nigeria 16,333 0.7 7.4 Europe and South Africa 31,085 2.4 6.2 Central Asia 306,543 5.6 7.2 Sudan 1,332 22.1 6.3 Armenia 435 2.8 5.2 Zambia 1,057 6.1 5.0 Azerbaijan 2,333 7.1 4.3 Zimbabwe 1,519 –0.4 6.2 1. Includes Luxembourg. 2. Effective market growth (EMG) is a weighted average of import volume growth in the country’s export markets. Source: see technical notes. 284 G L O B A L E C O N O M I C I N D I C A T O R S Figure A3.5a Merchandise exports as share of GDP, 2002 Percent 40 30 World 20 10 0 Industrial Sub-Saharan East Asia South Asia Latin America Europe and Middle East countries Africa and Pacific and the Central Asia and Caribbean North Africa Source: World Bank data. Figure A3.5b Annual growth rate of export volumes, 1993–2002 Percent 15.0 10.0 World 5.0 0.0 Industrial Sub-Saharan East Asia South Asia Latin America Europe and Middle East countries Africa and Pacific and the Central Asia and Caribbean North Africa Source: World Bank data. 285 G L O B A L E C O N O M I C P R O S P E C T S Table A3.6 Imports of goods, 2002 Merchandise imports (CIF), millions of dollars; average annual volume growth 1993–2002 (percent); merchandise imports share of GDP (percent) Imports Growth Share Imports Growth Share Imports Growth Shares World 6,232,809 6.6 19.4 Europe and Sub-Saharan Central Asia (continued) Africa (continued) All developing Belarus 7,909 7.8 59.4 Zimbabwe 798 –7.7 6.4 countries 1,441,537 7.5 23.5 Bulgaria 6,106 6.3 43.6 Czech Republic 40,176 6.3 65.1 High income Asia 579,564 9.6 23.7 Estonia 4,768 7.6 83.6 countries 4,791,272 6.4 18.5 East Asia Georgia 885 6.5 27.9 China 501,706 10.2 28.3 Hungary 34,619 6.1 62.5 Industrial Fiji 613 –0.1 36.2 Kazakstan 7,106 8.0 27.8 countries 4,620,406 6.6 17.6 Indonesia 31,242 1.2 17.5 Latvia 3,623 8.5 45.3 Malaysia 75,043 7.2 78.9 Lithuania 6,188 7.9 47.1 G-7 countries 3,099,228 6.6 14.5 Philippines 32,547 7.1 41.8 Poland 53,993 6.0 29.6 Canada 226,994 6.7 30.9 Thailand 64,317 4.8 50.8 Romania 15,527 5.8 35.1 France 316,953 5.8 22.1 Vietnam 16,599 19.5 44.8 Russian Germany 486,062 4.4 24.4 Federation 63,956 7.9 19.3 Italy 244,491 4.6 20.7 South Asia 77,858 5.6 11.5 Slovak Republic 11,980 8.7 49.6 Japan 302,246 4.8 7.6 Bangladesh 8,514 9.1 17.0 Turkey 29,933 6.0 17.5 United Kingdom 330,862 6.7 21.1 India 52,114 7.0 9.6 Ukraine 14,079 7.5 32.1 United States 1,191,620 9.4 11.4 Nepal 1,521 5.4 27.9 Uzbekistan 3,523 7.2 32.5 Pakistan 9,937 –0.8 16.6 Other industrial 1,166,814 5.9 30.4 Sri Lanka 5,772 5.9 33.3 Middle East and Australia 70,178 8.1 17.6 N. Africa 114,003 1.4 17.2 Austria 67,958 5.3 33.2 Latin America 329,697 7.8 19.7 Algeria 11,378 4.2 19.8 Belgium1 184,178 5.6 74.8 Argentina 8,988 –4.7 8.8 Egypt, Arab Rep. 14,364 4.3 15.0 Denmark 46,896 2.9 27.2 Bolivia 1,492 2.2 17.8 Iran, Islamic Rep. 20,260 –2.1 16.3 Finland 5,277 3.8 4.0 Brazil 47,237 8.4 10.4 Jordan 4,349 3.5 46.2 Greece 10,207 4.8 7.8 Chile 15,827 4.9 23.8 Morocco 10,630 3.1 29.4 Iceland 2,206 3.0 25.9 Colombia 12,304 7.2 14.3 Oman 5,614 4.1 24.4 Ireland 52,381 8.7 42.8 Costa Rica 6,533 11.3 38.7 Saudi Arabia 30,418 0.0 15.8 Korea 151,768 9.2 31.8 Dominican Syrian Arab Rep. 4,576 4.0 5.0 Netherlands 177,228 6.0 42.4 Republic 8,428 14.6 37.4 Tunisia 9,864 4.5 47.4 New Zealand 13,132 3.8 23.6 Ecuador 6,240 11.0 32.8 Yemen, Rep. 2,550 2.4 25.9 Norway 35,619 2.2 18.6 El Salvador 4,702 11.8 32.7 Portugal 39,530 3.3 32.8 Guatemala 4,895 7.7 24.4 Spain 155,190 7.9 23.7 Jamaica 3,063 7.2 38.1 Sub-Saharan Sweden 65,809 5.7 27.5 Mexico 168,731 11.0 26.6 Africa 79,907 3.5 25.3 Switzerland 89,257 2.4 33.3 Panama 6,435 1.5 61.3 Angola 3,383 4.6 29.4 Paraguay 2,460 0.6 34.1 Cameroon 1,866 5.7 20.8 Other high Peru 7,271 0.6 12.5 Côte d’Ivoire 2,339 1.1 22.5 income 525,229 6.4 70.4 Trinidad and Ethiopia 1,494 5.6 23.0 Bahrain 3,972 –0.8 46.7 Tobago 3,541 5.5 38.1 Gabon 1,113 2.5 25.3 Hong Kong, Uruguay 2,067 13.3 12.1 Ghana 2,513 6.1 45.1 China 205,337 6.9 127.1 Venezuela 15,204 –1.8 13.0 Kenya 3,202 6.8 27.4 Israel 33,317 5.8 29.6 Madagascar 1,076 8.3 22.1 Kuwait 7,341 0.0 21.6 Europe and Nigeria 11,570 4.7 24.7 Singapore 110,256 6.4 126.8 Central Asia 318,748 7.2 30.5 South Africa 26,712 3.5 24.8 Taiwan, China 112,957 5.9 40.0 Armenia 807 5.2 33.0 Sudan 1,306 6.5 9.9 United Arab Azerbaijan 1,787 4.3 26.6 Zambia 1,021 2.8 27.3 Emirates 41,882 9.2 104.1 1. Includes Luxembourg. Source: see technical notes. 286 G L O B A L E C O N O M I C I N D I C A T O R S Figure A3.6a Merchandise imports as share of GDP, 2002 Percent 40 30 20 World 10 0 Industrial Sub-Saharan East Asia South Asia Latin America Europe and Middle East countries Africa and Pacific and the Central Asia and Caribbean North Africa Source: World Bank data. Figure A3.6b Annual growth rate of import volumes, 1993–2002 Percent 15.0 10.0 World 5.0 0.0 Industrial Sub-Saharan East Asia South Asia Latin America Europe and Middle East countries Africa and Pacific and the Central Asia and Caribbean North Africa Source: World Bank data. 287 G L O B A L E C O N O M I C P R O S P E C T S Table A3.7 Direction of merchandise trade, 2002 (percentage of world trade) High-income importers Low- and middle-income importers Latin Middle America All Sub- East Europe East and low- Other All Other All Saha- Asia and and the and United indus- indus- high- high- ran and South Central North Carib- middle- Source of exports States EU-15 Japan trial trial income income Africa Pacific Asia Asia Africa bean income World High-income econ. 12.3 27.4 3.0 6.5 50.7 5.5 56.2 0.8 6.5 0.7 3.7 1.5 4.0 17.1 73.4 Industrial 10.7 26.2 2.2 6.2 46.6 4.3 50.9 0.7 4.1 0.5 3.5 1.4 3.8 14.1 65.0 United States 0.0 2.3 0.9 3.3 6.9 1.0 7.9 0.1 0.8 0.1 0.2 0.2 2.5 3.9 11.8 EU-15 3.7 20.6 0.7 2.2 27.4 1.3 28.7 0.5 0.9 0.2 3.1 0.9 0.8 6.5 35.2 Japan 2.1 1.0 0.0 0.3 3.9 1.2 5.2 0.0 1.4 0.1 0.1 0.1 0.2 1.9 7.1 Other industrial 4.3 2.0 0.4 0.3 7.1 0.3 7.4 0.0 0.3 0.1 0.1 0.1 0.1 0.8 8.2 Other high-income 1.6 1.2 0.7 0.3 4.1 1.2 5.3 0.1 2.3 0.2 0.1 0.1 0.2 3.1 8.4 Low and middle- income economies 6.7 6.5 2.2 1.0 17.1 3.1 20.3 0.5 1.7 0.4 1.9 0.5 1.3 6.4 26.6 Sub-Saharan Africa 0.3 0.5 0.1 0.0 0.9 0.1 1.0 0.2 0.1 0.0 0.0 0.0 0.1 0.4 1.4 East Asia and Pacific 2.2 1.4 1.5 0.4 6.0 2.4 8.3 0.1 0.9 0.2 0.2 0.2 0.2 1.9 10.2 South Asia 0.3 0.3 0.0 0.0 0.7 0.2 0.8 0.0 0.1 0.0 0.0 0.0 0.0 0.3 1.1 Europe and Central Asia 0.3 2.8 0.1 0.2 3.4 0.2 3.6 0.0 0.2 0.0 1.5 0.1 0.0 1.9 5.5 Middle East and North Africa 0.3 0.8 0.3 0.1 1.7 0.3 2.0 0.1 0.3 0.1 0.1 0.1 0.0 0.6 2.6 Latin America and Caribbean 3.3 0.7 0.1 0.3 4.4 0.1 4.5 0.0 0.2 0.0 0.1 0.1 0.9 1.3 5.8 World 19.0 33.8 5.1 7.5 67.8 8.6 76.5 1.3 8.2 1.1 5.6 2.0 5.3 23.5 100.0 a. Expressed as a share (percent) of total world exports. World merchandise exports in 2002 amounted to some $6,300 billion. b. Other high-income group includes the Asian newly industrializing economies, several oil exporters in the Gulf region, and Israel. Source: IMF, Direction of Trade Statistics. 288 G L O B A L E C O N O M I C I N D I C A T O R S Table A3.8 Growth of current dollar merchandise trade, by direction 1993–2002 (average annual percentage growth) High-income importers Low- and middle-income importers Latin Middle America All Sub- East Europe East and low- Other All Other All Saha- Asia and and the and United indus- indus- high- high- ran and South Central North Carib- middle- Source of exports States EU-15 Japan trial trial income income Africa Pacific Asia Asia Africa bean income World High-income econ. 5.6 1.2 1.8 4.0 2.7 3.9 2.8 1.5 8.3 3.7 11.1 0.1 5.7 6.5 3.5 Industrial 6.2 1.2 1.4 4.1 2.6 3.5 2.7 1.6 8.2 2.5 11.0 0.1 5.7 6.2 3.3 United States 0.0 1.9 0.7 5.7 3.5 3.7 3.5 0.8 8.2 5.2 2.9 –2.1 7.1 6.1 4.3 EU-15 7.4 1.1 2.8 2.6 1.9 4.7 2.0 1.5 6.6 1.3 12.1 0.6 4.7 6.4 2.7 Japan 2.2 –1.8 0.0 0.1 1.1 1.6 1.2 –2.9 7.2 –0.5 3.3 –3.9 –0.3 4.1 1.9 Other industrial 7.7 2.4 0.7 4.3 5.3 2.9 5.2 4.5 6.7 4.7 7.6 3.2 3.3 5.4 5.2 Other high-income 2.4 2.8 3.2 2.6 3.2 5.8 3.7 1.2 8.3 7.1 14.7 0.6 5.2 7.5 5.0 Low and middle- income economies 11.3 7.3 7.4 12.2 9.2 7.6 8.9 12.2 15.8 9.8 12.2 5.7 9.2 11.4 9.5 Sub-Saharan Africa 5.6 5.5 18.1 4.2 6.4 18.5 6.9 11.4 25.7 7.6 15.1 7.8 13.7 13.4 8.4 East Asia and Pacific 14.6 11.1 9.5 14.1 12.2 7.4 10.6 16.6 17.4 14.9 13.2 8.8 19.7 15.6 11.3 South Asia 10.8 4.8 0.0 7.5 7.0 10.1 7.5 11.1 13.9 8.1 5.9 3.2 20.1 9.1 7.9 Europe and Central Asia 17.8 9.7 2.8 12.7 10.2 15.9 10.4 10.0 9.9 8.6 13.4 4.6 9.8 11.8 0.0 Middle East and North Africa 1.8 2.5 2.9 2.1 3.0 5.3 3.3 15.3 16.3 4.8 0.0 3.9 –0.2 7.3 4.2 Latin America and Caribbean 11.3 2.5 2.1 16.7 9.2 3.5 9.0 4.6 13.9 11.2 12.8 5.4 8.1 8.6 8.9 World 7.3 2.1 3.8 4.8 3.9 5.1 4.1 4.3 9.4 5.7 11.5 1.2 6.4 7.6 4.8 Note: Growth rates are compound averages. Source: IMF, Direction of Trade Statistics. 289 G L O B A L E C O N O M I C P R O S P E C T S Table A3.9 Structure of long-term debt, 2001 Share of long-term debt (%): concessional debt; nonconcessional debt at variable interest rates; nonconcessional debt at fixed interest rates Non-concessional Non-concessional Concessional Variable Fixed Concessional Variable Fixed All developing countries 19.0 48.1 32.9 Europe and Central Asia (continued) Czech Republic 2.0 72.8 25.2 Asia 32.0 44.5 23.5 Estonia 1.3 97.6 1.1 East Asia 24.2 52.0 23.8 Georgia 74.8 11.9 13.2 China 20.7 45.3 34.0 Hungary 0.5 63.6 35.9 Indonesia 27.4 63.8 8.8 Kazakhstan 2.9 86.1 11.0 Korea, Rep. .. .. .. Kyrgyz Republic 65.2 27.7 7.1 Malaysia 8.3 56.9 34.8 Latvia 1.6 82.1 16.3 Myanmar 80.1 10.9 9.0 Lithuania 3.0 51.7 45.3 Papua New Guinea 34.1 55.9 9.9 Moldova 19.9 54.7 25.4 Philippines 25.1 42.5 32.4 Poland 11.2 76.7 12.2 Thailand 16.2 65.6 18.1 Romania 2.6 72.4 25.0 Vietnam 73.2 11.7 15.1 Russian Federation 0.4 34.7 64.9 Slovak Republic 3.3 48.8 47.8 South Asia 52.0 24.9 23.0 Tajikistan 85.6 12.5 1.9 Bangladesh 96.4 0.0 3.6 Turkmenistan 19.4 67.4 13.3 India 38.9 29.4 31.7 Turkey 5.0 50.3 44.7 Nepal 99.8 0.0 0.2 Ukraine 24.4 50.8 24.8 Pakistan 68.8 24.8 6.3 Uzbekistan 23.5 68.1 8.4 Sri Lanka 81.7 10.2 8.1 Middle East and Latin America and North Africa 36.1 32.8 31.1 the Caribbean 4.6 59.1 36.3 Algeria 13.0 49.7 37.3 Argentina 1.3 38.5 60.2 Egypt, Arab Rep. 77.2 6.1 16.7 Bolivia 54.4 32.9 12.7 Jordan 48.4 34.1 17.5 Brazil 1.4 72.4 26.2 Morocco 31.6 44.7 23.7 Chile 0.9 94.1 5.0 Oman 18.5 64.7 16.8 Colombia 2.7 56.9 40.4 Syrian Arab Rep. 93.0 0.0 7.0 Costa Rica 15.1 23.1 61.8 Tunisia 25.4 31.6 43.0 Dominican Republic 35.2 33.0 31.9 Yemen, Rep. 94.1 2.0 4.0 Ecuador 15.8 32.5 51.7 El Salvador 31.8 35.3 32.9 Sub-Saharan Africa 48.3 15.1 36.6 Guatemala 40.3 31.7 27.9 Angola 21.1 8.5 70.5 Jamaica 19.4 20.7 59.9 Botswana 69.0 7.1 23.9 Mexico 0.8 62.1 37.2 Côte d’Ivoire 38.9 46.8 14.3 Nicaragua 59.0 20.0 21.0 Cameroon 58.6 12.2 29.2 Panama 4.2 42.7 53.1 Ethiopia 91.4 0.1 8.5 Paraguay 32.7 47.4 19.8 Gabon 42.3 8.7 48.9 Peru 15.0 61.5 23.5 Ghana 78.6 5.4 16.0 Trinidad and Tobago 0.7 38.3 60.9 Kenya 78.4 5.1 16.6 Uruguay 3.0 47.0 50.0 Madagascar 72.8 5.0 22.2 Venezuela, RB 0.1 51.9 48.0 Nigeria 4.6 5.6 89.9 Senegal 87.0 5.5 7.5 Europe and Central Asia 5.7 54.3 40.0 South Africa 0.0 55.5 44.5 Armenia 74.1 12.7 13.2 Sudan 50.1 17.9 32.0 Azerbaijan 55.4 41.0 3.6 Zambia 78.5 7.8 13.7 Belarus 12.6 59.7 27.7 Zimbabwe 45.7 20.3 34.0 Bulgaria 4.2 86.0 9.8 Note: Nonconcessional debt data are available only for countries which report to the World Bank’s Debtor Reporting System. For aggregate figures, missing values are assumed to have the same average value as the available data. Republic of Korea is not included in the aggregate figures. 290 G L O B A L E C O N O M I C I N D I C A T O R S Figure A3.9a Structure of long-term debt, by group, 2001 Percent Concessional 100 80 ᮡ 60 Variable rate 40 ᮡ 20 Fixed rate ᮡ 0 Severely Moderately Severely Moderately Other indebted indebted indebted indebted countries low-income low-income middle-income middle-income countries countries countries countries Source: World Bank data. Figure A3.9b Structure of long-term debt, by region, 2001 Percent Concessional 100 ᮡ 80 Variable 60 rate ᮡ 40 Fixed 20 ᮡ rate 0 Sub-Saharan East Asia South Asia Latin America Europe and Middle East and Africa and Pacific and the Central Asia North Africa Caribbean Source: World Bank data. Figure A3.9c Top ten ratios of nonconcessional debt to GDP, 2001 Percent Congo, Dem. Rep. Liberia Congo, Rep. Belize Nigeria Angola Panama Kazakhstan Bulgaria Côte d’Ivoire 0.0 20.0 40.0 60.0 80.0 100.0 Source: World Bank data. 291 G L O B A L E C O N O M I C P R O S P E C T S Table A3.10 Long-term net resource flows to developing countries, 2001 (millions of U.S. dollars) Private Official Total Percent Debt flows millions $ GDP Total (net) FDI Portfolio Total ODA Other All developing countries 207,063 3.3 169,003 –8,648 171,693 5,958 38,060 37,875 184 Asia 54,524 2.4 40,616 –16,898 52,979 4,535 13,909 12,359 1,550 East Asia 45,028 2.7 36,817 –15,025 48,913 2,930 8,211 6,649 1,561 China 45,635 3.9 43,238 –4,017 44,241 3,015 2,396 874 1,522 Indonesia –6,446 –4.6 –7,312 –4,198 –3,278 164 866 955 –89 Korea, Rep. 9,046 2.1 9,279 –4,084 3,198 10,165 –233 –56 –177 Malaysia 3,004 3.4 855 951 554 –650 2,149 1,413 736 Myanmar 210 .. 145 –63 208 0 65 65 0 Papua New Guinea 150 5.2 2 –61 63 0 148 78 70 Philippines 2,014 2.8 2,077 –98 1,792 383 –62 164 –226 Thailand –2,977 –2.6 –3,052 –6,891 3,820 18 76 534 –458 Vietnam 1,946 5.9 710 –590 1,300 0 1,236 1,222 14 South Asia 9,496 1.5 3,798 –1,873 4,066 1,606 5,698 5,710 –12 Bangladesh 1,305 2.8 304 230 78 –4 1,002 1,007 –5 India 4,383 0.9 3,534 –1,608 3,403 1,739 849 813 36 Nepal 255 4.6 19 0 19 0 236 236 0 Pakistan 1,639 2.8 –308 –561 383 –130 1,947 2,066 –120 Sri Lanka 525 3.4 243 71 172 0 282 271 11 Latin America 79,898 4.2 72,067 500 69,309 2,258 7,831 3,495 4,336 Argentina –2,770 –1.0 –3,897 –7,030 3,214 –81 1,127 –225 1,353 Bolivia 1,217 15.3 637 –26 662 0 580 549 32 Brazil 26,159 5.1 23,337 –1,781 22,636 2,482 2,823 398 2,424 Chile 5,634 8.5 5,727 1,470 4,476 –219 –93 8 –101 Colombia 4,768 5.8 3,597 1,312 2,328 –43 1,171 81 1,090 Costa Rica 451 2.8 630 176 454 0 –179 –31 –148 Dominican Republic 1,719 8.1 1,729 530 1,198 0 –10 –31 21 Ecuador 1,549 7.4 1,444 113 1,330 1 105 54 51 El Salvador 982 7.1 674 406 268 0 308 132 175 Guatemala 577 2.8 403 –52 456 0 174 102 71 Jamaica 1,344 17.3 1,385 771 614 0 –41 –4 –38 Mexico 27,429 4.4 28,079 3,198 24,731 150 –650 –64 –586 Nicaragua 820 .. 13 –119 132 0 807 822 –15 Panama 1,806 15.0 1,799 1,287 513 0 7 –17 24 Paraguay 6 0.1 –14 –93 79 0 19 –13 33 Peru 2,317 4.3 1,400 294 1,064 42 917 184 733 Trinidad and Tobago 819 9.0 830 –5 835 0 –11 3 –14 Uruguay 880 4.7 796 478 318 0 84 –37 121 Venezuela 1,734 1.4 2,644 –730 3,448 –74 –910 0 –910 Europe and Central Asia 40,319 4.0 36,162 5,774 30,130 258 4,157 7,347 –3,191 Armenia 174 8.2 74 4 70 0 101 110 –9 Azerbaijan 348 6.1 216 –11 227 0 132 140 –7 Belarus 87 0.7 83 –13 96 0 5 20 –16 Bulgaria 1,028 7.6 1,043 360 692 –9 –15 170 –185 Czech Republic 5,481 9.6 5,194 –346 4,924 616 287 150 137 Estonia 658 11.9 625 54 539 32 34 43 –9 Georgia 307 9.6 173 13 160 0 134 147 –13 Hungary 3,935 7.6 3,952 1,378 2,440 134 –17 31 –48 Kazakhstan 5,009 22.6 4,947 2,128 2,763 55 62 57 6 Kyrgyz Republic 52 3.4 –73 –78 5 0 125 145 –19 Latvia 977 12.7 880 697 177 6 97 80 17 Lithuania 625 5.3 521 91 446 –16 104 91 14 Moldova 125 8.4 70 –27 94 4 55 64 –9 Poland 6,205 3.4 9,611 4,205 5,713 –307 –3,406 462 –3,868 Romania 3,002 7.6 2,633 1,468 1,157 8 369 236 133 Russian Federation 526 0.2 1,488 –1,524 2,469 543 –962 414 –1,376 Slovak Republic 447 2.2 303 –1,173 1,475 0 144 63 81 Tajikistan 179 17.0 39 17 22 0 141 134 6 Turkmenistan .. .. .. .. .. .. .. .. .. Turkey 2,329 1.6 906 –2,281 3,266 –79 1,423 –302 1,725 Ukraine 815 2.1 426 368 792 –734 389 126 263 Uzbekistan 346 3.0 46 –25 71 0 300 217 83 Middle East & N. Africa 9,543 1.4 7,462 2,134 5,460 –132 2,082 2,553 –471 Algeria –595 –1.1 243 –953 1,196 0 –838 –143 –695 Egypt, Arab Rep. 2,073 2.1 2,067 1,519 510 39 5 130 –125 Iran, Islamic Rep. 1,101 1.0 1,049 1,016 33 0 52 –28 80 292 G L O B A L E C O N O M I C I N D I C A T O R S Table A3.10 Long-term net resource flows to developing countries, 2001 (continued) (millions of U.S. dollars) Private Official Total Percent Debt flows millions $ GDP Total (net) FDI Portfolio Total ODA Other Middle East & N. Africa (continued) Jordan 548 6.2 –114 –70 100 –145 662 472 190 Morocco 2,346 6.9 2,633 –17 2,658 –8 –286 46 –332 Oman –557 –2.8 –867 –905 42 –3 309 9 301 Syrian Arab Rep. 167 2.1 204 –1 205 0 –37 –25 –12 Tunisia 1,568 7.8 1,108 666 457 –15 460 247 213 Yemen, Rep. –118 –1.3 –210 –5 –205 0 92 113 –22 Sub-Saharan Africa 22,778 7.2 12,697 –157 13,815 –961 10,081 12,121 –2,040 Angola 1,793 18.9 1,924 –222 2,146 0 –131 –109 –23 Botswana 41 0.8 55 –2 57 0 –14 8 –22 Côte d’Ivoire 173 1.6 137 –110 246 1 36 155 –120 Cameroon 254 3.0 –16 –91 75 0 270 277 –7 Ethiopia 857 13.7 10 –10 20 0 847 860 –13 Gabon –12 –0.3 170 –30 200 0 –182 –45 –138 Ghana 818 15.4 244 154 89 0 574 567 7 Kenya 188 1.6 –37 –43 5 0 225 371 –146 Madagascar 257 5.6 9 –2 11 0 248 247 1 Nigeria –454 –1.1 920 –184 1,104 0 –1,374 92 –1,466 Senegal 439 9.5 167 41 126 0 273 294 –21 South Africa 6,842 6.0 6,627 427 7,162 –962 215 211 4 Sudan 716 5.7 574 0 574 0 142 143 –1 Zambia 518 14.2 126 54 72 0 392 401 –9 Zimbabwe 56 0.6 –28 –33 5 0 84 71 13 Note: Republic of Korea is not included in the aggregate figures. Figure A3.10a Distribution of long-term net resource flows, 2001 Percent 100 80 Official Private 60 ᮡ 40 ᮡ 20 0 Sub-Saharan East Asia South Asia Latin America Europe and Middle East Africa and Pacific and the Central Asia and Caribbean North Africa Source: World Bank data. Figure A3.10b Change in share of private long-term flows, 1990–2001 Percent 80 60 40 20 0 Sub-Saharan East Asia South Asia Latin America Europe and Middle East Africa and Pacific and the Central Asia and Caribbean North Africa Source: World Bank data. 293 Technical Notes The principal sources for the data in this ap- lions of current U.S. dollars, while growth pendix are the World Bank’s central databases rates are based on constant price data, which and several International Monetary Fund data- are derived from current values deflated by rel- bases, combined with data sourced from evant price indices or unit value measures. Ef- the OECD and from Oxford Economics Inc. fective market growth (EMG) in table A3.5 is (OEF), covering the industrial and other high- the export-weighted growth of each country’s income economies. The cut-off date for data trading partner imports. updates was July 16, 2003. Data revisions and Tables A3.7 and A3.8. The IMF’s Direction new releases since that time have not been in- of Trade database serves as the underlying corporated in the tables. Regional aggregates source for the bilateral trade share and growth are based on the classification of economies by information highlighted in these tables. income group and by region, following the Growth rates are compound annual averages, Bank’s standard definitions (see country classi- and are computed from current U.S. dollar fication tables that follow). measures of trade. Debt and finance data (tables A3.9 and Table A3.9. Long-term debt covers public A3.10) cover the 138 countries that report to and publicly guaranteed debt but excludes the Bank’s Debtor Reporting System (DRS), IMF credits. Concessional debt is debt with an supplemented by data for non-DRS countries, original grant element of 25 percent or more. for which commercial market information has Nonconcessional variable interest-rate debt been utilized. Small countries have generally includes all public and publicly guaranteed been omitted from the tables, but are included long-term debt with an original grant element in the regional totals. Current price data are of less than 25 percent, whose terms depend reported in U.S. dollars. on movements in a key market interest rate. This item conveys information about the bor- Notes on tables rower’s exposure to changes in international Tables A3.1 through A3.4. Historic data interest rates. sourced from the databases noted above, Table A3.10. Long-term net resource flows while projections are consistent with those are the sum of net resource flows on long-term highlighted in chapter 1 and appendix 1. debt (excluding IMF) plus non-debt-creating Tables A3.5 and A3.6. Merchandise trade flows. Foreign direct investment refers to the data is sourced from combined IMF, World net inflows of investment from abroad. Portfo- Bank, OECD, and OEF sources. Merchandise lio equity flows are the sum of country funds, exports and imports exclude trade in services. depository receipts, and direct purchases of Imports are reported on a cost-insurance-and- shares by foreign investors. freight basis. Trade values are expressed in mil- 294 Classification of Economies G L O B A L E C O N O M I C P R O S P E C T S Table 1 Classification of economies by income and region, July 2003 Europe and Middle East Sub-Saharan Africa Asia Central Asia and North Africa East and Eastern Income Southern West East Asia South Europe and Rest of Middle North group Subgroup Africa Africa and Pacific Asia Central Asia Europe East Africa Americas Low- Angola Benin Cambodia Afghanistan Azerbaijan Yemen, Haiti income Burundi Burkina Faso Indonesia Bangladesh Georgia Rep. of Nicaragua Comoros Cameroon Korea, Dem. Bhutan Kyrgyz Congo, Dem. Central Rep. India Republic Rep. African Lao PDR Nepal Moldova Eritrea Republic Mongolia Pakistan Tajikistan Ethiopia Chad Myanmar Uzbekistan Kenya Congo, Rep. Papua New Lesotho Côte d’Ivoire Guinea Madagascar Equatorial Solomon Malawi Guinea Islands Mozambique Gambia, The Timor-Leste Rwanda Ghana Vietnam Somalia Guinea Sudan Guinea-Bissau Tanzania Liberia Uganda Mali Zambia Mauritania Zimbabwe Niger Nigeria São Tomé and Principe Senegal Sierra Leone Togo Middle- Lower Namibia Cape Verde China Maldives Albania Turkey Iran, Islamic Algeria Bolivia income South Africa Fiji Sri Lanka Armenia Rep. Djibouti Brazil Swaziland Kiribati Belarus Iraq Egypt, Arab Colombia Marshall Bosnia and Jordan Rep. Cuba Islands Herzegovina Syrian Arab Morocco Dominican Micronesia, Bulgaria Republic Tunisia Republic Fed. Sts. Kazakhstan West Bank Ecuador Philippines Macedonia, and Gaza El Salvador Samoa FYRa Guatemala Thailand Romania Guyana Tonga Russian Honduras Vanuatu Federation Jamaica Serbia and Paraguay Montenegro Peru Turkmenistan St. Vincent Ukraine and the Grenadines Suriname Upper Botswana Gabon American Croatia Lebanon Libya Argentina Mauritius Samoa Czech Oman Belize Mayotte Malaysia Republic Saudi Arabia Chile Seychelles N. Mariana Estonia Costa Rica Islands Hungary Dominica Palau Latvia Grenada Lithuania Mexico Poland Panama Slovak St. Kitts and Republic Nevis St. Lucia Trinidad and Tobago Uruguay Venezuela, RB Subtotal 152 25 23 24 8 26 1 9 6 30 296 C L A S S I F I C A T I O N O F E C O N O M I E S Table 1 Classification of economies by income and region, July 2003 (continued) Europe and Middle East Sub-Saharan Africa Asia Central Asia and North Africa East and Eastern Income Southern West East Asia South Europe and Rest of Middle North group Subgroup Africa Africa and Pacific Asia Central Asia Europe East Africa Americas High- OECD Australia Austria Canada income Japan Belgium United Korea, Rep. Denmark States New Zealand Finland Franceb Germany Greece Iceland Ireland Italy Luxembourg Netherlands Norway Portugal Spain Sweden Switzerland United Kingdom Non-OECD Brunei Slovenia Andorra Bahrain Malta Antigua and French Channel Israel Barbuda Polynesia Islands Kuwait Aruba Guam Cyprus Qatar Bahamas, Hong Kong, Faeroe United Arab The Chinac Islands Emirates Barbados Macao, Greenland Bermuda Chinad Isle of Man Cayman New Liechtenstein Islands Caledonia Monaco Netherlands Singapore San Marino Antilles Taiwan, Puerto Rico China Virgin Islands (U.S.) Total 209 25 23 36 8 27 28 14 7 41 a. Former Yugoslav Republic of Macedonia. b. The French overseas departments French Guiana, Guadeloupe, Martinique, and Réunion are included in France. c. On 1 July 1997 China resumed its exercise of sovereignty over Hong Kong. d. On 20 December 1999 China resumed its exercise of sovereignty over Macao. Source: World Bank data. Definitions of groups economies have reached a preferred or final stage of develop- For operational and analytical purposes, the World Bank’s ment. Classification by income does not necessarily reflect main criterion for classifying economies is gross national in- development status. come (GNI) per capita. Every economy is classified as low in- This table classifies all World Bank member economies, come, middle income (subdivided into lower middle and upper and all other economies with populations of more than middle), or high income. Other analytical groups, based on 30,000. Economies are divided among income groups accord- geographic regions and levels of external debt, are also used. ing to 2002 GNI per capita, calculated using the World Bank Low-income and middle-income economies are sometimes Atlas method. The groups are: low income, $735 or less; referred to as developing economies. The use of the term is lower middle income, $736–2,935; upper middle income, convenient; it is not intended to imply that all economies in $2,936–9,075; and high income, $9,076 or more. the group are experiencing similar development or that other 297 G L O B A L E C O N O M I C P R O S P E C T S Table 2 Classification of economies by income and indebtedness, July 2003 Income Sub- Not classified group group Severely indebted Moderately indebted Less indebted by indebtedness Low- Afghanistan Lao PDR Bhutan Azerbaijan Timor-Leste income Angola Liberia Cambodia Bangladesh Benin Madagascar Cameroon Equatorial Burkina Faso Malawi Ghana Guinea Burundi Mauritania Haiti Eritrea Central African Moldova Kenya Georgia Republic Myanmar Mali India Chad Nicaragua Mongolia Korea, Dem. Comoros Niger Papua New Rep. Congo, Dem. Nigeria Guinea Lesotho Rep. Pakistan Senegal Mozambique Congo, Rep. Rwanda Tanzania Nepal Côte d’Ivoire São Tomé and Togo Solomon Islands Ethiopia Principe Uganda Vietnam Gambia, The Sierra Leone Uzbekistan Yemen, Rep. Guinea Somalia Zimbabwe Guinea-Bissau Sudan Indonesia Tajikistan Kyrgyz Zambia Republic Middle- Lower Brazil Bolivia Albania Kiribati Marshall Islands income Cuba Bulgaria Algeria Macedonia, Micronesia, Fed. Sts. Ecuador Colombia Armenia FYRa West Bank and Gaza Guyana Honduras Belarus Maldives Iraq Jamaica Bosnia and Morocco Jordan Kazakhstan Herzegovina Namibia Peru Philippines Cape Verde Paraguay Serbia and Russian China Romania Montenegro Federation Djibouti South Africa Syrian Arab Samoa Dominican Sri Lanka Republic St. Vincent Republic Suriname and the Egypt, Arab Swaziland Grenadines Rep. Tonga Thailand El Salvador Ukraine Tunisia Fiji Vanuatu Turkey Guatemala Turkmenistan Iran, Islamic Rep. Upper Argentina Chile Botswana St. Lucia American Samoa Belize Croatia Costa Rica Trinidad and Mayotte Gabon Dominica Czech Republic Tobago N. Mariana Islands Lebanon Estonia Libya Venezuela, RB Palau Panama Grenada Lithuania Uruguay Hungary Mauritius Latvia Mexico Malaysia Oman Slovak Poland Republic Saudi Arabia St. Kitts and Seychelles Nevis 298 C L A S S I F I C A T I O N O F E C O N O M I E S Table 2 Classification of economies by income and indebtedness, July 2003 (continued) Income Sub- Not classified group group Severely indebted Moderately indebted Less indebted by indebtedness High- OECD Australia Korea, Rep. income Austria Luxembourg Belgium Netherlands Canada New Zealand Denmark Norway Finland Portugal Franceb Spain Germany Sweden Greece Switzerland Iceland United Kingdom Ireland United States Italy Japan Non- Andorra Isle of Man OECD Antigua and Israel Barbuda Kuwait Aruba Liechtenstein Bahamas, The Macao, Chinac Bahrain Malta Barbados Monaco Bermuda Netherlands Brunei Antilles Cayman Islands New Caledonia Channel Islands Puerto Rico Cyprus Qatar Faeroe Islands San Marino French Singapore Polynesia Slovenia Greenland Taiwan, China Guam United Arab Hong Kong, Emirates Chinad Virgin Islands (U.S.) Total 209 50 39 55 65 a. Former Yugoslav Republic of Macedonia. b. The French overseas departments French Guiana, Guadeloupe, Martinique, and Réunion are included in France. c. On 20 December 1999 China resumed its exercise of sovereignty over Macao. d. On 1 July 1997 China resumed its exercise of sovereignty over Hong Kong. Source: World Bank data. Definitions of groups either of the two key ratios exceeds 60 percent of, but does This table classifies all World Bank member economies, and not reach, the critical levels. For economies that do not report all other economies with populations of more than 30,000. detailed debt statistics to the World Bank Debtor Reporting Economies are divided among income groups according to System (DRS), present-value calculation is not possible. 2002 GNI per capita, calculated using the World Bank Atlas Instead, the following methodology is used to classify the method. The groups are: low income, $735 or less; lower non-DRS economies. Severely indebted means three of four middle income, $736–2,935; upper middle income, key ratios (averaged over 1999–2001) are above critical $2,936–9,075; and high income, $9,076 or more. levels: debt to GNI (50 percent); debt to exports (275 Standard World Bank definitions of severe and moderate percent); debt service to exports (30 percent); and interest to indebtedness are used to classify economies in this table. exports (20 percent). Moderately indebted means three of the Severely indebted means either: present value of debt service four key ratios exceed 60 percent of, but do not reach, the to GNI exceeds 80 percent or present value of debt service to critical levels. All other classified low- and middle-income exports exceeds 220 percent. Moderately indebted means economies are listed as less indebted. 299 he Doha Development Agenda of the Fourth Ministerial T The World Trade Organization (WTO) round of trade negotiations Conference of the WTO opened many contentious and initiated in November 2001 in important questions. Global Economic Prospects 2004: Doha, Qatar, was intended to be a Realizing the Development Promise of the Doha Agenda analyzes the “development round.” Those good most critical multilateral trade issues and suggests policy options intentions are now being tested. that would raise living standards in developing countries and Trade ministers from all over the reduce global poverty. world are discussing ways to reduce trade barriers—barriers that greatly The fourteenth annual edition of Global Economic Prospects harm development and poverty reduction. Global Economic Prospects • explores the short-, medium-, and long-term outlook for the global economy, including driving forces, commodity prices, 2004 explores the tough issues and capital flows, and their implications for major regions under discussion—protection of agriculture, trade in labor-intensive • reviews recent trends in exports from developing countries, trade barriers that work to the disadvantage of poor people, manufactures, labor services, and and policies to reduce protection and other inequities in the special treatment for developing world trading system countries, among others—to present • examines trade in agriculture—the most important and politically options that would indeed reduce contentious sector for global poverty reduction—including key poverty and advance development. lessons from development experience, possible changes to the The global talks will not be easy and current system of subsidies and protection, and the potential may take time, but allowing poor for liberalization in both rich and poor countries people greater access to world • investigates the temporary movement of labor—so-called markets will offer them new Mode 4 of the General Agreement on Trade in Services— opportunities to improve their living evaluating its advantages and disadvantages to both the home standards. If agreements are to be and the host countries genuinely pro-development, these • discusses trade facilitation in light of post-9/11 concerns for discussions must be informed by security to suggest new policies that would promote greater clear analysis of measures that are and more-secure trade likely to benefit poor people the • reviews the special treatment of developing countries in most.That is the purpose of this the world trading system and the role of trade preferences, Global Economic Prospects. exemptions from WTO rules, and technical assistance to implement WTO trade regulations. —Nicholas Stern Senior Vice President and Global Economic Prospects 2004 provides essential information for Chief Economist those concerned with developments shaping today’s global economy. THE WORLD BANK ISBN 0-8213-5582-1