GLor ____ n C Financing the Poorest Countries 2002 2IX2 Ilt 4 7r ~a ~c4 v - -  Global Development Finance Financing the Poorest Countries ANALYSI S AND SOlOlrr A THE WORLD BANK @ 2002 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 All rights reserved. 1 2 3 4 04 03 02 The findings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank cannot guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply on the part of the World Bank any judgment of the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is copyrighted. No part of this work may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or inclusion in any information storage and retrieval system, without the prior written permission of the World Bank. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. Cover design by W. Drew Fasick, Serif Design Group Cover photo: Curt Carnemark, World Bank Photo Library ISBN 0-8213-5085-4 ISSN 1020-5454 Table of Contents The Report Team vii Preface viii Acronyms and Abbreviations ix Overview 1 Chapter 1 Challenges for Developing Countries during the Coming Global Recovery 5 Recession and recovery in the industrial world 7 Bust and boom in world trade 13 Regional developments 19 Risks to the forecast 26 Notes 28 References 29 Chapter 2 Private Capital Flows to Emerging Markets 31 The global slowdown reduced capital market flows to developing countries 31 Net resource flows 32 Capital market flows 32 Trends in FDI 37 Emerging market financial crises in 2001 43 The prospects for capital market flows and FDI 47 Annex 2.1: Forecasts of private flows to developing countries 49 Annex 2.2: Measuring resource flows to developing countries 51 Notes 51 References 52 Chapter 3 The Poor Countries' International Financial Transactions 55 Poor countries have benefited from the growth of global capital flows 55 Financial integration in the poor countries 55 FDI to the poor countries 59 Improved investment climate is associated with rapid growth of FDI 61 Effective competition policies are critical 63 The participation of foreign banks in poor countries' financial systems 64 Capital outflows 69 Annex 3.1: Econometric analysis of foreign bank participation 78 Notes 81 References 83 111 GLOBAL DEVELOPMENT FINANCE Chapter 4 Strengthening Official Financial Support for Developing Countries 89 Mixed results from aid have led to a fall in aid 89 The policy framework 89 Trends in aid 90 The macroeconomic impact of aid 96 Conditionality and adjustment lending 101 Aid and debt relief 104 Strengthening the effectiveness of official guarantees 107 Annex 4.1 110 Notes 111 References 113 Appendix 1 Debt Burden Indicators and Country Classifications 119 Appendix 2 Commercial Debt Restructuring 133 Appendix 3 Official Debt Restructuring 151 Appendix 4 Regional Economic Developments and Prospects 165 East Asia and Pacific 166 Europe and Central Asia 170 Latin America and the Caribbean 174 Middle East and North Africa 178 South Asia 183 Sub-Saharan Africa 186 Appendix 5 Global Commodity Price Prospects 191 Summary tables Tables 1.1 Global conditions affecting growth in developing countries and world GDP growth 7 1.2 Initiating factors: turning points to downturn and recovery in OECD recessions 10 1.3 Developing-country forecast summary, 1991-2004 20 2.1 Net long-term resource flows to developing countries, 1991-2001 32 2.2 Capital market commitments to developing countries, 1991-2001 33 2.3 Debt ratios during recessions, East Asia and Latin America 33 2.4 International equity placement and performance of stock markets 34 2.5 Capital market commitments and spreads for developing countries 38 2.6 Projected capital market flows to developing countries 47 2A.1 How representative is the forecasting model? 49 2A.2 Comparison of forecasts with actual capital market flows to developing countries 50 2A.3 Statistics for the forecast of FDI 51 3.1 Net external financial flows to developing countries, 1999 56 3.2 Net long-term capital flows to poor countries, 1986-99 56 3.3 Annual change in policy performance and FDI as ratio to GDP, 1991-99 61 3.4 FDI as ratio to GDP and policy performance index in poor countries 61 3.5 Mining sector performance in three countries, before and after reforms 65 3.6 Cumulated outflows during 1980-99 70 3.7 Volatility of capital flows, 1990-99 70 3.8 Cumulated outflows as a share of GDP, 1999 71 iv TABLE OF CONTENTS 3A.1 Foreign bank presence and domestic bank performance 78 3A.2 Panel-VAR results for all developing countries 80 3A.3 Summary of impulse response functions, all developing countries 80 3A.4 Results of panel-VAR regression for poor countries 80 3A.5 Summary of impulse response functions, poor countries 80 4.1 Net official aid to developing countries, by type and source, 1990-2001 93 4.2 Trends in aid allocation 95 4.3 Forgiveness of ODA claims, 1970-2000 105 4.4 Impact of HIPC Initiative in 24 decision-point cases 106 4.5 Export credit commitments to HIPCs, 1990-2000 110 Figures 1.1 World and industrial and developing country GDP growth, 1997-2004 6 1.2 Manufacturing production in the G-3 countries 2000-02 8 1.3 U.S. manufacturing output, high-tech and non-high-tech industries 9 1.4 Consumer confidence in the United States, the Euro Area, and Japan 9 1.5 OECD GDP growth and fiscal balance, 1970-2000 12 1.6 GDP growth in the industrial countries, 2001-04 12 1.7 World export growth, 1999-2001 13 1.8 World industrial production and import volumes 14 1.9 Shipping cost index (Baltic Dry) 15 1.10 Real non-oil commodity prices since 1980 18 1.11 Per capita agricultural production 18 1.12 Oil prices and OECD oil stocks 19 1.13 GDP growth in developing regions 21 1.14 Forecasting the 2001 U.S. slowdown 27 1.15 Two recessions in the United States, 1990-91 and 2001 27 2.1 Performance of developing-country stock markets by sector 34 2.2 Bank lending standards and bank credit to developing countries, 1990-2001 35 2.3 Corporate default rate and risk premiums, 1990-2001 37 2.4 FDI and M&A in developing countries, 1991-2001 38 2.5 FDI as ratio to GDP, 1991-2001 40 2.6 Regional trends of FDI flows, 1991-2001 42 2.7 North-South and South-South FDI, 1991-1999 42 3.1 Five-year rolling correlation between savings and investment, 1974-1999 57 3.2 FDI-to-GDP ratios, 1991-2000 59 3.3 Foreign direct investment in mining exploration and government policies 64 3.4a Foreign bank presence in poor countries 65 3.4b Foreign bank presence in Africa 65 3.5 Effect of greater foreign bank presence on intermediation costs and domestic bank profitability 66 3.6 Effect of greater foreign bank presence on international bank lending to poor countries 67 3.7 Effect of greater foreign bank presence on nonperforming loans 68 3.8 Capital outflows from developing countries, 1985-99 70 3.9 Cumulated outflows and minerals exports 73 3.10 Capital account restrictions 74 V GLOBAL DEVELOPMENT FINANCE 4.1 ODA from donor countries in relation to their GNP, 1990-2000 94 4.2 Compliance with conditionality and economic performance 101 4.3 NPV of external debt of the 24 countries that reached their HIPC decision point 105 Boxes 1.1 The Doha Development Agenda 17 2.1 Evidence of changes in the appetite for risk and capital market flows 36 2.2 The concentration of FDI flows 39 2.3 Round-tripping of capital flows between China and Hong Kong 41 2.4 Financial market contagion from the Argentine crisis 44 2.5 Moral hazard and rescue packages 46 3.1 Improving market access through future-flow securitization 58 3.2 The investment climate and domestic investment 60 3.3 Capital outflows from the middle-income countries 72 3.4 Narrowly focused capital controls in emerging markets 75 4.1 The PRSPs 91 4.2 The Financing for Development (FfD) process 92 4.3 The relationship between private and multilateral flows in poor countries 98 4.4 Official guarantees and the Mozal project 109 Vi The Report Team HIS REPORT WAS PREPARED BY THE ECO- Mick Riordan, and Virendra Singh, and benefited nomic Policy and Prospects Group, and from the guidance of the Bank's regional chief drew on resources throughout the Devel- economists. Appendix 5 was prepared by John opment Economics Vice-Presidency, the Economic Baffes, Betty Dow, Don Mitchell, and Shane Policy Sector Board, the World Bank operational Streifel. The financial flow and debt estimates regions, the International Finance Corporation, were developed in a collaborative effort by Punam and the Multilateral Investment Guarantee Asso- Chuhan, Nevin Fahmy, Shelley Fu, Ibrahim Lev- ciation. The principal author was William Shaw, ent, and Gloria Moreno of the Financial Data with direction by Uri Dadush. Chapter 1 was led Team along with Himmat Kalsi, Eung Ju Kim, and by Hans Timmer, with contributions by John Malvina Pollock of the Economic Policy and Pros- Baffes, Betty Dow, Caroline Farah, Fernando pects Group. The report was prepared under the Martel Garcia, Bernard Hoekman, Robert Key- general direction of Nicholas Stern. fitz, Annette I. De Kleine, Robert Lynn, Donald Many others from inside and outside the Bank Mitchell, Mick Riordan, Virendra Singh, Shane provided input, comments, guidance, and support Streifel, Dominique van der Mensbrugghe, and at various stages of the report's publication. Get- Bert Wolfe. Chapters 2-4 were largely prepared by ard Caprio, Paula Donovan, Guy Pfeffermann, the international finance team of the Economic and Sanjivi Rajasingham were discussants at the Policy and Prospects Group, including Gholam Bankwide review. Sebastian Edwards, Shahrokh Azarbayejani, Shweta Bagai, Maria Pia lannarello, Fardoust, Jan Willem Gunning, Jim Hanson, and Himmat Kalsi, Eung Ju Kim, Aparna Mathur, Stephen O'Connell provided extensive reviews of Sanket Mohapatra, Shoko Negishi, Bilin Neyapti, individual chapters. Comments were provided by Malvina Pollock, Dilip Ratha, and Jeff Ziarko. Jehan Arulpragasam, Amarendra Bhattacharya, Additional contributions and background papers Jaime Biderman, Gerard Caprio, Haydee Celaya, were provided by Dilek Aykut, Punam Chuhan, James Emery, Alan Gelb, Ian Goldin, Charleen and Barry Eichengreen (chapter 2); Sara Calvo, Gust, Daniel Kaufman, Jeni Kiugman, Stefan Koe- Stijn Claessens, Susan Collins, Sebastian Edwards, berle, Jacob Kolster, Richard Newfarmer, John Simon Evenett, Nagesh Kumar, Jeffrey Lewis, Page, Enrique Rueda-Sabater, Sudhir Sherry, Philip Deepak Mishra, Koh Naito, Claudine Ndayiken- Suttle, Axel van Trotsenburg, and Ultich Zachau. gurutse, Andrew Powell, Jaya Prakash Pradhan, Comments were also received from the Interna- Felix Remy, Tony Thompson, Esen Ulgenerk, tional Monetary Fund. Mark Feige edited the re- Aristomene Varoudakis, and Peter van der Veen port to highlight the main messages. Awatif (chapter 3); and Paul Collier, David Dollar, Robert Abuzeid and Katherine Rollins provided assis- Keyfitz, and Dan Morrow (chapter 4). Appendix tance to the team. Robert King managed dissemi- 1 was prepared by Ibrahim Levent, appendix 2 by nation and production activities by the Economic Eung Ju Kim, and appendix 3 by Malvina Pollock. Policy and Prospects Group. Book design, editing, Appendix 4 was prepared by Caroline Farah, production, and dissemination were coordinated Robert Keyfitz, Annette w. De Kleine, Robert Lynn, by the World Bank Publications team. vii Preface GLOBAL DEVELOPMENT FINANCE WAS debt stock and its components; the computation of formerly published as World Debt Tables. flows, aggregate net resource flows, and aggregate The new name reflects the report's ex- net transfers; and the relation between net resource panded scope and greater coverage of private fi- flows and the balance of payments. Exact defini- nancial flows. tions of these and other terms used in Global De- Global Development Finance consists of two velopment Finance are found in the Sources and volumes: Analysis and Summary Tables and Coun- Definitions section. try Tables. Analysis and Summary Tables contains The economic aggregates presented in the ta- analysis and commentary on recent developments bles are prepared for the convenience of users; in international finance for developing countries. their inclusion is not an endorsement of their value Summary statistical tables are included for selected for economic analysis. Although debt indicators regional and analytical groups comprising 148 can give useful information about developments in countries. debt-servicing capacity, conclusions drawn from Country Tables contains statistical tables on them will not be valid unless accompanied by care- the external debt of the 136 countries that report ful economic evaluation. The macroeconomic in- public and publicly guaranteed debt under the formation provided is from standard sources, but Debtor Reporting System. Also included are tables many of them are subject to considerable margins of selected debt and resource flow statistics for in- of error, and the usual care must be taken in inter- dividual reporting countries, as well as summary preting the indicators. This is particularly true for tables for regional and income groups. the most recent year or two, when figures are pre- For the convenience of readers, charts on pages liminary or subject to revision. x to xii summarize graphically the relation between viii Acronyms and Abbreviations CIS Commonwealth of Independent States Mercosur Southern Cone Common Market (Argentina, CPPR Country Portfolio Performance Review Brazil, Paraguay, Uruguay; Bolivia and DAC Development Assistance Committee Chile are associate members) (of the OECD) MILIC moderately indebted low-income country DCB debt conversion bond mimic moderately indebted middle-income country DDSR debt and debt service reduction MUV manufacturing unit value DRS Debtor Reporting System (of the World Bank) MYRA multiyear rescheduling agreement El eligible interest bond NAFTA North American Free Trade Agreement EMBI Emerging Market Bond Index NBC National Bank of Commerce (Tanzania) EPZ export processing zone NGO nongovernmental organization EU European Union NIE newly industrialized economy FDI foreign direct investment NPV net present value FfD Financing for Development OA official aid FLIRB front-loaded interest reduction bond ODA official development assistance FRN floating-rate note OECD Organisation for Economic Co-operation G-7 Group of Seven (Canada, France, and Development Germany, Italy, Japan, United Kingdom, OPEC Organization of Petroleum Exporting United States) Countries GATS General Agreement on Trade in Services PRSC Poverty Reduction Support Credit GDP gross domestic product PRSP Poverty Reduction Strategy Paper GNI gross national income REER real effective exchange rate HIPC heavily indebted poor countries SDR special drawing right (of the International HIV human immunodeficiency virus Monetary Fund) IBRD International Bank for Reconstruction and SILIC severely indebted low-income country Development (of the World Bank Group) SIMIC severely indebted middle-income country ICT information and communications technology SMEs small and medium enterprises IDA International Development Association U.N. United Nations (of the World Bank Group) UNCTAD United Nations Conference on Trade IFC International Finance Corporation and Development IMF International Monetary Fund URR unremunerated reserve requirement LIBOR London interbank offered rate VAR vector autoregression LILIC less indebted low-income country WTO World Trade Organization LIMIC less indebted middle-income country XGS exports of goods and services M&A mergers and acquisitions Dollars are current U.S. dollars, unless otherwise specified. ix Debt stock and its components Total external debt (EDT) Short-term debt Long-term debt Use of IMF (LDOD) credits by debtor Private Public and publicly nonguaranteed debt guaranteed debt by creditor Official creditors Private creditors Multilateral Bilateral Commercial Bonds Other banks x Aggregate net resource flows and net transfers (long-term) to developing countries Loan disbursements minus Principal repayments equals Foreign direct in- Debt service Net resource flows -plus- vestment (FDI), -equals- Aggregate net (LTDS) on debt portfolio equity resource flows flows, and official grants minus minus Loan interest Interest payments an I proft Eand FDI profits equals equals -Net transfers on Aggregate net debt transfers Note: Includes only loans with an original maturity of more than one year (long-term loans). Excludes IMF transactions. xi Aggregate net resource flows (long-term) and the balance of payments Credits Debits * Exports of goods and services * Imports of goods and services * Income received * Income paid Current account * Current transfers * Current transfers Including workers' remittances and private grants * Official unrequited transfers (by foreign * Official unrequited transfers (by national governments) government) * Official unrequited transfers (by foreign * Official unrequited transfers (by national governments) government) * Foreign direct investment (by nonresidents) * Foreign direct investment (by residents) (disinvestment shown as negative) (disinvestment shown as negative) *Capital and Capial ad -IS *Portfolio investment (abroad by finacia accuntresidents) (amortizations shown as negative) *b NetOther long-term capital outflow (by sresidents) (amortizations shown as negative) *Short-term capital inflow *Short-term capital outflow Reserve account Net changes in reserves F1 Aggregate net resource flows FR Net resource flows on debt (long-term) x11   Overview: International Finance and the Poorest Developing Countries HE INTEGRATION OF DEVELOPING COUN- percent in real terms over the same period. The tries into the global economy increased poor countries now receive about the same level of sharply in the 1990s with improvements in FDI as middle-income countries, relative to the size their economic policies; the massive expansion of of their economies. In addition, the global expan- global trade and finance driven by technological sion of international banks coupled with the liber- innovations in communications, transport, and alization of domestic financial systems in the poor data management; and the lowering of barriers to countries increased the average share of foreign trade and financial transactions. Many of the poor- bank assets to more than 40 percent of total assets, est developing countries' participated strongly in more than double the share of 1995 and compara- this process despite their limited access to capital ble to that of many middle-income countries that markets. This report analyzes the interaction be- have recently benefited from increased foreign tween the global expansion of finance and im- bank participation (see pages 64-66). provements in domestic policies in the poor coun- tries over the 1990s, and the implications for -good policies and governance, along with growth and poverty reduction. Three main mes- strong institutions, are critical to using private sages are developed: (a) a strong investment cli- flows productively mate is critical to attracting foreign capital and A rise in private flows can have a substantial im- using it productively; (b) poor countries' increas- pact on investment in the poor countries and, if ing integration in the global economy means that productively used, on growth. However, the policy they face similar policy challenges as middle-income framework must be right. Improvements in the in- countries, including how to deal with capital mo- vestment climate (a term that refers to the numer- bility; and (c) achieving the Millennium Develop- ous ways in which government affects the produc- ment Goals will require a substantial rise in aid tivity of investment, including policies, governance, flows, an increased allocation of aid to countries and the strength of institutions) have boosted the with good policies, and improvements in policies impact of international financial transactions on by both developing countries and donors. productivity in the poor countries. Domestic firms in countries with strong investment climates are A greater integration of poor countries more able to absorb the foreign technology and and private capital- skills that come with FDI (see pages 62-63). Better The surge in foreign direct investment (FDI) flows policies have enabled some poor countries to at- and the decline in aid have transformed external fi- tract more diversified FDI flows-the share of nance to the poor countries. FDI flows to the poor countries that export natural resources in the poor countries rose from 0.4 percent of the gross domes- countries' FDI dropped from half in 1991 to 20 tic product (GDP) in the late 1980s to 2.8 percent percent toward the end of the decade. Countries in the late 1990s in response to the globalization of that established the competitive conditions re- production and improvements in domestic policies quired to attract foreign banks experienced an im- (see pages 59-61). Aid to these countries fell by 20 provement in the efficiency of their domestic banks of teir conmies Inaddiion thegloal epan GLOBAL DEVELOPMENT FINANCE and thus a decline in the cost of financial interme- Good policies and strong governance diation (see pages 66-69). are also key to improving aid effectiveness Earlier empirical studies consistently found a I,weak relationship between aid and investment, Poor countries face similar challenges with even less of an impact of aid on growth. How- from g lobalization as middle-income ever, more recent research shows that aid makes countries an effective contribution to growth and poverty re- T he events of the past year underlined the risks duction in countries with good economic policies, of capital mobility for the middle-income sound institutions, and strong governance, but has emerging markets. The current global economic little effect in countries with poor policies. A dou- slowdown, exacerbated by the bursting of the high- hlng of aid flows would help ensure that develop- tech bubble at the end of 2000 and the terrorist at- ing countries achieve the Millennium Development tacks in September 2001, is exceptionally deep and Goals, provided that this aid is allocated to coun- broad (see pages 7-11). Capital market flows once tries with good policies and large numbers of poor again proved to be procyclical: the growth slow- people (pages 99-100). down in industrial countries reduced both emerg- Aid continued to decline in 2001, and the pros- ing markets' export revenues and their access to ex- pects for a substantial rise in the medium term are ternal finance (see pages 32-36). By contrast, the limited (pages 90-94). Most countries with good level of FDI in 2001 was virtually unchanged from policies can continue to absorb additional aid re- the previous year despite adverse global conditions, sources without seriously impairing the effective- including a drop in global FDI flows (see pages ness of that aid (see pages 96-99). Aid does not, in 37-40). The crisis in Argentina illustrates how general, increase the volatility of government re- open capital accounts can compound the effects of sources, and appropriate policies can ensure that unsustainable macroeconomic policies and high aid does not contribute to inflationary pressures or public sector debt, thus seriously complicating sta- cause excessive exchange-rate appreciation. It is bilization efforts (see pages 43-47). true that even in many countries with good poli- The poor countries are also vulnerable to capi- cies, lack of administrative capacity lowers the tal mobility. While most still impose restrictions on marginal productivity of aid as aid levels rise. capital account transactions, controls have had However, recent research indicates that aid levels only limited success in controlling capital outflows to most countries with strong economic programs in the context of a weak investment climate, where are well below the threshold where aid becomes domestic investment opportunities are limited and ineffective. fears of confiscation or reduction in the value of assets provide considerable incentive to put money abroad (see pages 69-78). Poor countries with bet- ter than average policies (as measured by the Better aid policies by donors also World Bank) had more success in retaining domes- contribute to poverty reduction tic capital: a rough estimate of the stock of their here is evidence that donors have made pro- capital outflows relative to GDP was about one- T gress in improving their own policies, through sixth the size in poor countries with worse than av- increasing resources to debt relief for good per- erage policies. Capital outflows have been more formers, easing complex administrative require- volatile in the poor countries than in the middle-in- ments that can strain limited government capacity, come countries, while volatility can be more costly and reducing the share of tied aid (see pages (in terms of welfare) in poor countries because 101-104). Modifications of adjustment assistance more people live close to subsistence and have little have helped to preserve the use of conditionality in private insurance or public safety nets. Thus poli- channeling aid resources to good performers and cymakers in poor countries need to recognize the supporting the credibility of government policies, potential impact of capital mobility on both stabi- while ensuring adequate government flexibility lization policies and long-term development, and domestic stakeholder commitment to the pro- 2 OVERVIEW: INTERNATIONAL FINANCE AND THE POOREST DEVELOPING COUNTRIES gram. Here also, recipient government policies are Cambodia, Cameroon, Cape Verde, Central African Repub- key: strong leadership and effective administration lic, Chad, Comoros, the Democratic Republic of Congo, the by the government can help promote aid coordina- Republic of Congo, C6te d'lvoire, Djibouti, Eritrea, Ethi- opia, The Gambia, Georgia, Ghana, Guinea, Guinea-Bissau, tion and make it easier for donors to adopt more Guyana, Haiti, Honduras, Kenya, Kiribati, the Kyrgyz Re- flexible policies. public, the Lao People's Democratic Republic, Lesotho, Liberia, Madagascar, Malawi, Maldives, Mali, Mauritania, Moldova, Mongolia, Mozambique, Myanmar, Nepal, Nicaragua, Niger, Nigeria, Pakistan, Rwanda, Samoa, Sio Tom6 and Principe, Senegal, Sierra Leone, Solomnon Islands, Note Somalia, Sri Lanka, Sudan, Tajikistan, Tanzania, Togo, 1. The poor countries are defined to represent developing Tonga, Uganda, Vanuatu, Vietnam, Republic of Yemen, countries with relatively low per capita income and almost Zambia, and Zimbabwe. These countries' average per capita no access to international capital markets. The group in- income is under $500 per year compared with $2,900 for cludes all IDA-only countries plus a few blend countries that other developing countries. And most of them are small; only have had few IBRD loans over the past few years. The coun- Pakistan, Bangladesh, Nigeria, Vietnam, Ethiopia, and the tries included are Afghanistan, Albania, Angola, Armenia, Democratic Republic of Congo have more than 50 million Bangladesh, Benin, Bhutan, Bolivia, Burkina Faso, Burundi, people. 3  Challenges for Developing Countries during the Coming Global Recovery HE CURRENT GLOBAL ECONOMIC SLOW- their exposure in emerging markets in reaction to down is exceptionally deep and broad. increased uncertainty, reduced value of portfolios in Global growth in 2001, at 1.2 percent, was industrial countries, and increased default provi- 2.7 percentage points lower than in 2000 (figure sions; high-tech sectors, with many firms decimated 1.1). In the last 40 years the deceleration in gross after the high-tech bubble burst; and tourism indus- domestic product (GDP) was sharper only in tries, suffering from the aftermath of the terrorist 1974, during the first oil crisis. The current slow- attacks. As in every severe downturn, poor people down is also broad in that the deceleration is pay a high price. Without buffers or safety nets to equally rapid for industrial countries and develop- rely upon, their ability to satisfy basic needs is im- ing countries. The slowdown in economic activity mediately at stake when incomes decline. coincides with an unprecedented 14 percentage The current sharp deceleration in economic point deceleration of world trade, from record activity largely follows a typical investment and growth of 13 percent in 2000 to a 1 percent de- inventory cycle, even if it was triggered by other cline in 2001 (table 1.1). However, contrary to factors, such as the bursting of the high-tech bub- many earlier downturns, inflationary pressures re- ble or the terrorist attacks. Likewise, the standard mained very subdued and this allowed monetary investment cycle is expected to play a major role authorities to loosen their policies substantially. in recovery. The steep decline in investment and The bursting of the high-tech bubble at the end stock building in recent quarters carries seeds for a of 2000 and the terrorist attacks in September forceful cyclical recovery. As capital stocks and in- 2001 made the deceleration of the global economy ventories are adjusted downward to reflect lower so exceptionally sharp. The unpredictable charac- growth expectations, the decline in investment and ter of these events made it difficult to anticipate the stock-building tends to become less steep and ac- depth of the downturn. Nevertheless, after the ter- tivity starts to rebound. The rebound will be fur- rorist attacks the expectations-a deeper recession ther fueled by aggressive monetary and fiscal stim- and a delay of the recovery by one or two quar- ulus, especially in the United States. The current ters-appear to be materializing.' Several of the synchronism of the cycles in different parts of the strong market reactions to the terrorist attacks world will likely be reflected in a strong global re- have been reversed and signs of a recovery in the covery, even if recovery in individual countries is United States and the high-tech sectors have started not exceptionally vigorous. to mount. The economic consequences of the terrorist Even during this unusually synchronized down- attacks probably delayed this rebound by about turn, the intensity and character of the economic two quarters, implying strong growth in the sec- malaise differ across countries, sectors, and income ond half of 2002. Weak growth in the second groups. Especially hard hit are countries dependent half of 2001 and the first half of 2002 is expected on commodity exports, with many commodity to keep global growth in 2002 at 1.3 percent, prices at historical lows; highly indebted emerging slightly above growth rates for 2001. This outlook economies, because private investors have reduced implies a downward adjustment since the publica- 5 GLOBAL DEVELOPMENT FINANCE Figure 1.1 World and industrial and developing country GDP growth, 1997-2004 Percentage change Forecast 6 Developing countries World 3 - Industrial 0 1 9 countries, 0 1997 1998 1999 2000 2001 2002 2003 2004 Source: World Bank Economic Policy and Prospects Group calculations. tion of Global Economic Prospects 2002 (World achievable in the absence of additional adverse Bank 2001), mainly reflecting more pessimistic shocks to the global economy. Such a recovery views on Japan and Latin America. World trade would be supported by modest inflation-median could very well decline in 2002 for a second year inflation in the developing world is around 5.5 in a row. However, an anticipated acceleration in percent, only half the average rate during the the second half of 2002 will likely result in a strong 1990s-relatively low interest rates after the re- recovery in annual growth for 2003. Although cent easing of U.S. monetary policy, rapidly grow- global GDP growth in 2003 of 3.6 percent would ing import demand in the industrial countries, and fall short of the strong 3.9 percent performance a slight rebound in real commodity prices. Ex- of 2000, advances in world trade are expected to porters of high-tech products are likely to benefit breach 8 percent. more than average from this recovery. The main Not all economies will benefit immediately from risks to this favorable outlook are to be found in the robust global rebound. Argentina's financial financial markets. The fragile Japanese banking strains have resulted in defaults and devaluation, sector may trigger more adverse developments than heralding a protracted period of painful adjustment; is currently assumed, and the full complement of but there is also hope that a new base can be cre- ramifications stemming from financial crises in Ar- ated for resumption of long-term growth. As finan- gentina and Turkey remains uncertain. cial weakness in Japan has worsened during the Many developing countries, even those that global downturn, a recovery of the external environ- currently do not have large financial imbalances, ment can probably not avert, but only alleviate, face difficult challenges. The global downturn and structural adjustments. Commodity exporters, in- country-based policy responses to slowing growth cluding oil producers, have experienced large terms- have reversed the trend of declining fiscal deficits of-trade losses that will limit their short-term ability in many countries, and deterioration of deficits to rebound. The speed of recovery toward normal tend to persist well after economic growth has re- trends in tourism is uncertain, leaving the prospects turned to normal levels. Some oil exporters-such cloudy for many of the developing countries that are as Nigeria, the Rep6blica Bolivariana de Vene- heavily dependent on this revenue source. zuela, and Indonesia-are particularly vulnerable, On average, however, developing countries' as oil prices are expected to continue their down- growth is expected to be robust in 2003 and 2004, ward trend. Furthermore, the global downturn im- reaching 5 percent per year. A strong recovery seems plies a deterioration of the current account for 6 CHALLENGES FOR DEVELOPING COUNTRIES DURING THE COMING GLOBAL RECOVERY Table 1.1 Global conditions affecting growth in developing countries and world GDP growth (percentage change from previous year, except interest rates and oil prices) Current Forecasts E002e 2002 2003 2004 2q01 00 0 Global conditions World trade (volume) 1.8 8.3 7.3 Inflation (consumer prices) G-7 OECD countries.b 0.9 1.6 1.8 United States 1.5 2.4 2.6 Commodity prices (nommal J... - I Commodity prices, except I iJ..ii...1 - Oil price (dollars, weigh -I . .. dollars a barrel *0 S I' Oil price, percent changt - Manufactures export un-' . lu i. r -. Interest rates LIBOR, 6 months (dollar . per,,.r 1 - 4 1 4 . EURIBOR, 6 months (ut... r. ' 4 -4 i 4 i 4 World GDP (growth) I I High-income countries - 1 1 OECD countries 4 3 .. United States I 1 1 . Japan Euro Area 4 1 2 I T Non-OECD countries - 4 4 1 i Developing countries East Asia and Pacific Europe and Central A.. Latin America and the . .rI.I. I, 0i.i . Middle East and Nort , i . 22. South Asia 5 Sub-Saharan Africa . 7 7 Memorandum itens East Asian crisis-affec-. ... 2 . , Transition countries o- i1 . 4 4 i 4 34 '11 Developing countries, Excluding the transitu... ... 23 I 3 Excluding China and I.- J. L .1 I1 2 4 . a. The G-7 countries are Canada, France, Germany, Italy, Japan, the United Kingdom, and the United States. b. Unit value index of manufactures exports for G-5 countries (G-7 minus Canada and Italy) to developing countries, expressed in dollars. c. London interbank offered for dollars. d. Interbank offered rate for euros. e. Indonesia, the Republic of Korea, Malaysia, the Philippines, and Thailand. Source: World Bank Economic Policy and Prospects Group, February 2002 forecast; Global Economic Prospects (GEP) 2002 projections of October 2001. many developing countries. Together with limited into-or came close to-recession in the course of availability of international private capital, this 2001. Aggregate annual growth in the industrial could generate new financial strains, which could world decelerated from 3.4 percent in 2000 to 0.9 impede further recovery. percent in 2001. With almost all recessions having started in the second half of 2001, it is unlikely that aggregate annual growth in 2002 will exceed 2001 growth, even with a solid rebound in the second Recession and recovery in the half of the year. Indeed, measured growth is likely industrial world to decline further, to only 0.8 percent. The advance T he United States, Japan, Germany, and several in output in 2003, in contrast, is expected to return smaller industrial countries in Europe entered to 3.1 percent, assuming that no major crisis evolves 7 GLOBAL DEVELOPMENT FINANCE from the fragilities in the Japanese banking system The second phase began at the end of 2000 or other sources of tension in the forecast. Growth when the recession in durable goods had begun to in 2004 is assumed to fall back to near its long-term bottom out, but the high-tech bubble burst yet fur- trend of 2.5 percent. ther, forcing stock markets into sharp decline In the fall of 2000 the downturn still had while high-tech production started to fall at dra- characteristics of a soft landing, with cyclical cor- matic rates (figure 1.3). Japanese output, highly rections that did not suggest one of the most se- dependent on high-tech exports, declined precipi- vere decelerations in economic activity in decades. tously. The fall in exports and the accompanying However, in two steps-the first initiated by the drop in equity prices exacerbated the bad-loan burst of the high-tech bubble at the end of 2000, problems in the Japanese banking sector, which and the second by terrorist attacks in September could nor escape the spiral of defaults and thin 2001-the global economy decelerated further. margins in a deflationary environment. In Europe, signals were mixed in the beginning of this phase. A three-phase slowdown- Since European growth in 2000 hardly exceeded At the root of the simultaneous economic down- its long-term capacity trend, the internal cyclical turn in all major industrial countries was a severe forces were much weaker than in the United States. slowdown in manufacturing sectors (figure 1.2). However, the slowdown in world trade affected That slowdown went through three phases. The the manufacturing sectors, while the European first phase began in the middle of 2000 with the telecommunications industry shared the fate of the slowdown in the United States, which was partly global high-tech sectors as future profitability was a reaction to the tightening of monetary policy suddenly reassessed. The European Central Bank by the Federal Reserve Board, a move designed hesitated to ease monetary policy in the face of in- to slow an economy that had been growing well flationary pressures originating from temporary above capacity. Production of traditional durables increases in food prices due to livestock diseases, declined, and production in high-tech sectors high oil prices, and a weak euro. The slowdown, started to slow. The latter was partly a reaction to first apparent in Germany, gradually spread to sev- the high-tech investment bubble that had been eral other European countries. swelling since 1998, especially in the United States, The terrorist attacks in September 2001 and then burst. Japan and the European economies marked the start of the third phase. At that time the clearly lagged in the downturn, recessions in manufacturing production had more Figure 1.2 Manufacturing production in the G-3 countries 2000-02 Percentade change, three-montenthree-month, seasonally adjusted annual rate t y T Germany 12A % 6/ 0 - . % -6- droUnited States equit p -18p I I I I I I secto I I Jan. March May July Sepi. Nov. Jan March May July Sept Nov. Jan. 2000 2000 2000 2000 2000 2000 2001 2001 2001 2001 2001 2001 2002 Source: National statistics and World Bank Economic Policy and Prospects Group calculations. 8 CHALLENGES FOR DEVELOPING COUNTRIES DURING THE COMING GLOBAL RECOVERY Figure 1.3 U.S. manufacturing output, high-tech and non-high-tech industries Percentage change, three-month moving average, seasonally adjusted annual rate 100 - 75 7.5 - - 60 50 - - 45 25 - High-tech (night axis) g - 30 0.0 - - 15 -2.5 - -50 - Excluding high-tech 0 (left axis) -75 _ - -15 -100 -30 Jan March May July Sept. Nov. Jan March May July Sept Nov Jan 2000 2000 2000 2000 2000 2000 2001 2001 2001 2001 2001 2001 2002 Source: Federal Reserve, through Datastream. or less bottomed out, albeit for Japan and the within weeks, and commodities prices fell 7 percent United States at still large declining rates. The pe- within one month. Industrial production dipped riod immediately after the terrorist attacks was once again, although it seemed that the high-tech characterized by an extraordinary, but temporary, cycle was less affected (figure 1.3). While these first loss of consumer confidence and deterioration of market reactions were reversed within one quarter, business sentiment (figure 1.4). Equity prices plum- economic recovery will probably be delayed by meted 15 percent immediately after the attacks, about two quarters as a result of supply disruptions spreads on junk bonds jumped 200 basis points and shaken confidence. Figure 1.4 Consumer confidence in the United States, the Euro Area, and Japan EC: diffusion Index, U.rS. and Japan Indexes, January 2001=100 115 1 Euro Area 2 0 01 ---_ (night axis) 0 105 - -2 Japan (left axis) 95 -4 -6 -10 65 - Jan. 2001 March 2001 May 2001 July 2001 Sept. 2001 Nov. 2001 Jan 2002 Source: U.S.. Conference Board; Japan: ERISA; Euro Area: European Comm ssion. 9 GLOBAL DEVELOPMENT FINANCE The prolongation and deepening of the down- rent account surpluses of developing countries start turn in the aftermath of the terrorist attacks made to decline, a deterioration of government balances this recession comparable in intensity to the reces- could increase tensions in global capital markets. sions of the early 1980s and 1990s, at least for in- dustrial countries. Although the downturn in indi- -largely driven by investment cycles vidual countries has not necessarily been as deep as The deep recessions and subsequent recoveries in during those two severe recessions, its simultane- the United States during the last three decades ous character made the current slowdown espe- were primarily the reflection of inventory and in- cially sharp for the industrial world as a whole. vestment cycles.' Table 1.2 summarizes the main Experience during the last decades suggests that sources of change in GDP growth at the beginning the turning point to positive growth will probably and end of recessions. In the majority of U.S. re- be triggered by the investment cycle, and that reces- cessions since the 1970s, changes in investment or sions of this magnitude tend to result in a dete- inventories were the main source of changes in rioration of fiscal balances that typically lasts for GDP growth, both at the start and close of each three or more years. The sharp fall in private recession. With the structural decline in invento- spending implies an improvement of the current ries through the use of new technologies and just- account in the short run, despite increased fiscal in-time supply systems, the inventory cycle, still deficits. The mirror image of the industrial coun- dominant in the 1970s and 1980s, has become less tries' reduced current account deficit is the ten- important. The investment cycle was the main dency of current account surpluses to narrow and contributing factor in the current recession, and deficits to widen in the developing world. The re- investment will likely be the force that brings GDP mainder of this section will discuss triggers of turn- growth out of negative territory. As capital stocks ing points in economic activity and the behavior of adjust downward, the decline in investment rates government balances in the industrial world. In- will soften, reversing the downward spiral. creased trade linkages have made developing coun- Table 1.2 highlights the fact that net exports tries more dependent on these turning points in the have been a relatively more important factor de- industrial countries' business cycles, and as the cur- termining the dynamics of recessions in Europe Table 1.2 Initiating factors: turning points to downturn and recovery in OECD recessions (changes in contribution to growth, at seasonally adjusted annualized rates) wRNeorto o ntriaain- NoesDxre s t change in GDP g dotruti te beg meaurd or dontrn" aerae f oe r wo uater piortnth tunnepind (2) frcessions.y"n turin paority tof U.S. avrae- Svennteperes tr m Da instrent ofcanges in stocgrswth,nettexportseoftgoodsndndlosrsofceac MourceesW through the use and Pr technologiscandlju in-im suppl se toi dominan in90 the 1970s an .180s has becom Ss cotiuigfatrih urrent reesin and invetmen wills likely2 be 0 the forc tha brigsGD adjus downward,sen th deln m vsmn -ae Table 1.20 higliht the fact tha ne0epot haeueerrrlaielnmr imotn fato d- iNduestDriaountries'ontbuins cyclpedues n pnnreepesd as the curbmn nge dnamicsgrof adcreessions in Erope Tabled 1.2 Initiadotin : ator: foeotoqatrpirth turning points,odwtr and (2 o recovery " inrin OECDt recessionsag (chw ang e rs olbtong toProwthipa sresonallyiauted aonualtzed, rate etepniue,Igosfie netet =hnei stoks Ending quarter GDP growth Pnuipa sourcess Suite Str ates oo i olc n rspcsGopcacltos Mi1900,17 :-. - 46 Q ,17 . :36 I . CHALLENGES FOR DEVELOPING COUNTRIES DURING THE COMING GLOBAL RECOVERY than in the United States. The inventory cycle has risen significantly, depressing investment in high- never been as important in Europe as in the United risk assets, especially in the United States. In Japan, States. This could reflect the less pronounced do- financial markets are burdened by the accumulated mestic business cycles in Europe, which has more debt of failed businesses, which has reached Y50 automatic stabilizers in place, as well as greater trillion ($420 billion) since 1999, of which Y16 regional diversity in monetary and fiscal policies. trillion accrued during 2001. This has exacerbated Note that the recent downturn in Europe was trig- the "bad loan" problems of the commercial bank- gered mainly by swings in international trade, ing system, adding new nonperforming assets al- rather than by changes in domestic consumption, most as quickly as "old" nonperforming loans are investment, or inventories. It is thus likely that the written off. Under these circumstances, additional international trade cycle will also be an important Bank of Japan liquidity is unlikely to greatly in- ingredient of the recovery, in which case Europe crease the willingness of Japanese commercial will lag behind the United States in the rebound. banks to lend, and signs of a credit crunch for the Japan is the odd one out in this picture. Re- small-business sector may be emerging. cessions were avoided during the 1980s due to Fiscal policy also offers promise for boosting strong, continuous growth in investment and pro- growth, especially in the United States. The U.S. ductivity. However, investment growth has been Congress approved more than $40 billion in emer- declining since the early 1990s, when structural gency and industry-support funds in the immedi- growth rates fell, financial bubbles burst, and ate aftermath of September 11. Moreover, tax re- problems in the banking sector began to mount. ductions enacted earlier in 2001 will continue to This trend was so strong that it overwhelmed the be implemented over the next few years. In the tendency for investment to experience sharp cycli- Euro Area, automatic stabilizers will tend to in- cal changes. As a result, investment failed to play crease public deficits, but the constraints inherent the standard role of initiating a turning point in in the Stability and Growth Pact of the European economic activity. This is one reason why Japan Union could limit government support for slowing staggered from one recession into another during economies.3 In Japan debate continues regarding the 1990s, and why it is not easy to identify a the degree and nature of supplemental budget pro- source that could reverse the current downturn. grams, against the background of Prime Minister Junichiro Koizumi's stated limits to bond-m-arket Policy is supportive, but will operate funding of such efforts. On balance, fiscal stimulus with some delay- is likely to be a significant additional driving force Policies will play an important role in the recovery for recovery in the major industrial economies, of the industrial countries. Monetary policy has particularly for the United States. now turned highly expansionary in the United However useful and needed the fiscal stimulus States, and with some delay, has eased in the Euro may be in the short term, increased deficits could Area. In Japan the economy remains in a state of become a burden in the medium run. Historically, deflation (consumer prices have declined for the deficits that originated in severe downturns tend to past two years), and interest rates can hardly fall last well beyond the recovery in economic activity any further. Given the lack of headroom for alter- (figure 1.5). After the brief and steep recession fol- native action, the Bank of Japan initiated a pro- lowing the first oil crisis in the mid-1970s, the aver- gram of liquidity injections-potentially weakening age fiscal deficit (as a share of GDP) in the OECD the yen as a way to combat deflation and stimulate turned from positive to negative, never again to re- exports. turn to positive territory. After the second oil crisis, The effects of monetary easing are likely to be it took a decade for the deficits to come back close felt with some lag, and should provide a needed to precrisis levels, and after the Gulf War this took fillip to demand for consumer durables and hous- five years. The stubbornness of deficits is partly ing across the Organisation for Economic Co- due to the vicious circle of higher debt and increas- operation and Development (OECD) countries. ing debt service, and partly due to the temptation But there is concern that the eventual impact of to see recessions as unique, temporary phenomena lower interest rates on business investment may be and a subsequent recovery as a permanent im- limited. In particular, investor risk aversion has provement. While the deterioration of government finacia maretsare urdned y te acumuate GLOBAL DEVELOPMENT FINANCE Figure 1.5 OECD GDP growth and fiscal balance, 1970-2000 GDP percentage change; fiscal balance percentage of GDP 6 GDP growth 4 2 0 I Fiscal balance 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 Source: OECD. deficits is often abrupt, the restoration tends to be The U.S current account deficit which al smoothed out over time. Of course, many regional ready diminished to $420 billion in 2001 from differences and different policy decisions deter- $445 billion in 2000 as a result of recession and mined the trend in the average deficit. Neverthe- faling oil prices is expected to deteriorate only less, the historical pattern of persistent deficits is modestly over the next two years The adjustment clear, and the main challenge in the current reces- in 2002 and coming years is expected to be accom sion is to keep the necessary stimulus confined to panied by a gradual weakening of the dollar and a the short run. In the medium run, improvement in widening of current account deficits in some Euro the industrial countries' fiscal deficits will facilitate a resumption of capital flows toward developing countries. Figure 1.6 GDP growth in the industrial -auguring a strong recovery in 2003 countries 2001-04 Taking into account the likely impact of the inven- Pecentagechange tory and investment cycles, and the policy re- sponses, we anticipate that the United States will come out of the recession in the beginning of 2002 and European countries will follow one or two quarters later, but Japan will hardly reach positive growth during the year-resulting in annual 2002 growth rates of 1.3, 1.2, and -1.5 percent respec- tively for these countries (figure 1.6). As industrial production, investment, and global trade pick up rapidly over the course of the year, 2003 is ex- pected to provide a much rosier picture, with GDP U 2001 U2002 U2003 92004 growth climbing to 3.7, 3.3, and 1.7 percent in the 2 three industrial centers. If banking problems in Uniled Stales Euro Area Japan remain unsolved, a relapse into low or nega- Soue Word Bank Economic Policy and Prospec tive growth after a temporary export-led recovery Group calcula in that country cannot be excluded. 12 CHALLENGES FOR DEVELOPING COUNTRIES DURING THE COMING GLOBAL RECOVERY pean countries. The Japanese current account sur- of a year ago, and this deterioration intensified into plus declined substantially in 2001 because the the fourth quarter. latest recession in Japan was driven mainly by a The regions most affected by the fall-off in decline in exports instead of a deceleration in in- trade were East Asia-from depressed world de- vestment. Because Japanese investment is also not mand for high-tech goods and associated slippage likely to recover strongly in the near future, the in intraregional trade-and Latin America, due current account surplus is expected to widen again to the extensive trade relations between Mexico when world trade, and Japanese exports, rebound. and the United States. Central European economies The current account deficit for the industrial coun- continued to witness robust (although slowing) tries as a whole is expected to decline from $280 trade growth, while Sub-Saharan African countries billion in 2000 to $240 billion by 2004, most of were more affected by falling commodity prices the improvement being realized in the near term. than by declines in volume. Merchandise imports The mirror image of this development is a reduced are now expected to rebound strongly in the sec- current account surplus in the developing coun- ond half of 2002, together with a recovery of tries, partly reflecting declining oil prices and partly world industrial production (figure 1.8). By 2003 reflecting reduced export opportunities. growth rates could approach double-digit levels again, of which 3 percentage points will be positive carryover from 2002 .5 North American exports are expected to return to 9 percent growth in 2003, Bust and boom in world trade European exports to 7.5 percent, while Japanese W orld trade, already undergoing the sharpesttrade flows are expected to achieve growth of 6.5 deceleration on record, suffered additional percent. The high-tech exporters are likely to expe- setbacks following the terrorist attacks of Septem- rience the most rapid recovery, with particularly ber 11. These events delayed the expected recovery fast export growth expected for East Asia (near 10 in output, which will in turn delay the rebound in percent), boosted by China's accession to the WTO. merchandise trade for one or two quarters. More- over, security concerns disrupted trade flows, as did Trade logistics disrupted ... air transport increased shipping and insurance costs, although continues to suffer- medium-term effects arising from these develop- The disruption of the global transportation sys- ments are more uncertain. The attacks also reduced tem resulting from the terrorist attacks appears developing countries' revenues from international tourism. However, longer-run prospects for global trade have improved after a first important step to- ward a new round of trade negotiations was made at the World Trade Organization (WTO) minister- Percentage change, year-over-year ial conference in Doha, Qatar, in November 2001. 18 The record deceleration of merchandise trade growth in 2001 was due to a collapse in high-tech markets and recessions in the manufacturing sec- 6 tors of the industrial countries. Import demand de- clined sharply in the United States and Japan dur- 0 ing the first half of 2001, while European import demand fell in the second half. High-tech-intensive F 1999 U 2000 0 firsi half, 2001 ihird quarter, 2001 merchandise exports from the East Asian newly in- -12 dustrialized economies (NIEs-Hong Kong (China); Singapore; and Taiwan (China) declined much World Industrial more rapidly than merchandise exports from the counies counies rest of the world (figure 1.7).4 Trade flows also Note: Exports are fora sample of countries representing 79 percent of world exports. slowed in the developing world, although not as Source: Datastream and World Bank Economic Policy and Prospects Group sharply as in the NIEs. By the third quarter of 2001, calculations. developing-country export volumes were near levels 13 GLOBAL DEVELOPMENT FINANCE Figure 1.8 World industrial production and import volumes Percentage change at seasonally adjusted annualized rates 10 - Forecast 20 5 - 10 0 0 Industrial production -5 (left axis) Import volume -10 (night axis) -10 1 -20 Q2, 1993 02, 1994 02,1995 02,1996 02,1997 02,1998 02, 1999 02, 2000 01, 2001 Q2, 2002 Source: Datastream and World Bank Economic Policy and Prospects Group calculations. to have had only temporary adverse impacts on ica, and between North America and Europe. There trade growth, but uncertainties continue to loom. is anecdotal evidence suggesting that costs have Air cargo has suffered more than other transport risen substantially more on less-traveled routes, par- modes. After September 11, U.S. airspace was com- ticularly those close to the conflict zone around the pletely shut down for several days to domestic and Middle East and South Asia. For example, insur- international passenger and cargo traffic, and ca- ance rates on traffic through the Suez Canal in- pacity utilization and revenues in air transport re- creased dramatically after September 11. mained significantly below preattack levels for sev- Security concerns following the terrorist at- eral months. Other parts of the world, especially tacks had a more pronounced impact on the cost of South Asia and the Middle East, also suffered inter- air transport. In September, the air cargo index for ruptions in transportation, albeit less severe than transportation across major routes increased by an those in the United States. There is evidence to sug- average of 17 percent, with cargo costs from the gest, however, that the physical constraints on United States increasing by 22 percent. By October, trade from the security response to the attacks are the global index had declined by only 2 percent, abating. with costs still nearly 15 percent higher than before The attacks had the immediate effect of in- September 11. It is likely that a significant portion creasing insurance and security costs. Maritime of the rise in air cargo rates may be longer lasting. shipping costs rose for 10 to 15 days in the after- Developing countries' exports will be more math of September 11, rising on average 7 percent affected by rising transportation costs than will ex- according to the most widely available shipping cost ports from industrial countries, because developing indexes. One of these indexes, the Baltic Dry Index, countries tend to specialize in exports of primary shows a price spike shortly after September 11 (fig- goods and labor-intensive manufactures, which ure 1.9). However, costs declined quickly thereafter. have higher trade margins (international transport The Baltic Dry Index resumed its sharp downward costs) than the high-tech exports from industrial trend in a matter of days, continuing to track the countries.' One estimate of the effects of a sus- decline in world trade volumes over the last year. rained increase in the cost of trade on world trade Furthermore, the available data on seaborne ship- flows suggests that, if the terrorist attacks caused ping costs generally cover the major trade routes- a 10 percent increase in the port-to-port costs of for example, those between Asia and North Amer- merchandise trade, world trade could decline by 14 CHALLENGES FOR DEVELOPING COUNTRIES DURING THE COMING GLOBAL RECOVERY Figure 1.9 Shipping cost Index (Baltic Dry) Dollars 1,600 1,500 1,400 1,300 1,200 1,100 1,000 900 800 I I I I I I Jan. 2001 March 2001 May 2001 July 2001 Sept. 2001 Nov. 2001 Jan 2002 Note: The index is computed for the third day of each month. Source: The Balic Exchange through Datastream. about 1 percent, approximately $60 billion (rela- $ billion support package quickly enacted in the tve to a projection where the terrorist attacks have aftermath of September t1. Aside from declines in no lasting impact on costs).7 Developing countries' volume, price effects may also be important as re- trade would fall by 1.6 percent, and industrial sorts and hotels drop their prices in order to entice countries' exports would fall by 0.8 percent. visitors. In the first eight months of 2001 world tourism -and world tourism arrivals and revenues was on track for an increase of 2.5 to 3 percent for approach record lows the year as a whole, but after September 11 expec- The terrorist attacks also reduced developing coun- tations were adjusted to only 1 percent growth, im- tries' foreign exchange revenues from international plying a decline of more than 20 percent (annual- tourism, which amount to 7 percent of total ex- ized) in fourth quarter momentum.' Assuming a 20 ports of goods and services, about equivalent to percent drop in tourism revenues during a period of revenues from high-tech exports or exports of agri- six months, the loss in export revenues for develop- cultural and food products. The World Tourism ing countries could amount to $14 billion. The im- Organization reports that travel reservations world- pact on employment could be particularly severe, wide in November 2001 stood 12 to 15 percent because tourism services tend to be highly labor in- below the levels of a year earlier.' Anecdotal evi- tensive. Short-term impacts probably far exceed the dence suggests that the fall in tourism revenues longer-term consequences, since past trends idi- may well have reached double-digit rates, as both care that demand for travel and tourism services re- tourist arrivals and expenditures collapsed. Directly covers relatively rapidly from setbacks. Even so, after September 11, 40 percent of booked vacation countries near the conflict zone in South Asia and trips with Caribbean countries as the destination the Middle East may suffer a more sustained re- were canceled. Airlines have substantially trimmed duction of revenues. The impact of any decline in their schedules to other destinations as well. Sev- tourism revenues will vary enormously among de- eral mid-size carriers in Europe have failed in the veloping countries. For example, tourism can con- last few months, and some carriers in the United stitute as much as 70 percent of goods and services States are threatened with bankruptcy despite the exports in some small island economies, and also 15 GLOBAL DEVELOPMENT FINANCE has become a key export sector in many Sub- coffee, oiseeds, sugar, and raw materials such as Saharan African countries. Revenues from tourism rubber. Continued rapid technological progress for the 14 Sub-Saharan African countries with the contributed to supply increases in a number of highest dependency on tourism revenues average commodities," and improved policies in some de- 22 percent of total export revenues.'0 In absolute veloping countries contributed to large increases in terms, Turkey is the largest recipient of tourism rev- exports.12 Coffee prices were especially hard hit enues, and the sharp fall in these receipts since Sep- (down 30 percent in 2001 compared with 2000) tember 11 has complicated efforts to overcome the due to a 20 percent increase in global production financial crisis. over the past three years with little increase in con- sumption. Cotton prices declined 20 percent in Improved prospects for a development round 2001 due to large production increases in China of multilateral trade negotiations and the United Stares, and rice prices fell 15 per- The Doha Development Agenda-which emerged cent due to the large exports from Thailand and from the WTO Ministerial Conference held in Vietnam. Copper prices fell by 12 percent in 2001, Doha, Qatar, in November 2001-demonstrates and prices would have declined even further if the increased prominence of development con- major producers had not cut production by about cerns in WTO deliberations, in turn reflecting in- 5 percent in an effort to prevent additional price creased participation by developing countries in declines. the international trading system. Doha launched The price declines have been especially hard negotiations on market access for manufactures, for exporters in Africa, where non-oil commodi- dispute settlement, WTO rules, environmental ties often account for 70 percent or more of ex- policies, and intellectual property protection. port revenues. Ethiopia, for example, derives These negotiations will complement ongoing talks nearly two-thirds of total export revenues from on market access in agriculture and services, which coffee, and Mali derives about 40 percent of total are mandated by the Uruguay Round agreements. exports from cotton. Moreover, the prices of com- Negotiations will be launched on four so-called modities that account for a large share of Sub- Singapore issues--competition, investment, trade Saharan exports (such as cocoa, coffee, and cop- facilitation, and transparency in government pro- per) have fallen by more than the prices of com- curement-at the next WTO ministerial meeting modities exported by other developing countries in 2003, if consensus can be reached on the modal- (figure 1.10). Since 1980, the index of real non-oil ities of such negotiations at that time. Completing commodity export prices of Sub-Saharan African negotiations by January 1, 2005, as envisaged in countries has declined by 10 percent relative to the the Doha Ministerial Declaration, represents a index of all developing countries. On top of that, major challenge (box 1.1), but success in doing so the African index tends to be more volatile over would imply large welfare gains for both develop- the price cycle, implying a sharper fall during a ing and industrial countries. downturn. African producers have been unable to make up for the decline in prices through higher Secular declines and cyclical swings in volumes, since African agricultural production has commodities prices been flat over the past two decades, while agricul- Non-oil commodities. The global economic slow- tural production increased rapidly in developing down, a strong dollar, and large supplies of most countries as a whole (figure 1.11). Sub-Saharan commodities reduced the average dollar price of Africa's non-oil commodity export revenues developing countries' non-oil primary commodity dropped at least $3 billion between 1997 and exports by 9 percent in 2001. Demand for metals 2001-equal to 3.6 percent of non-oil export rev- was most affected by the economic slowdown, enues in 1997 and 25 percent of total official de- while agricultural commodities continued to face velopment aid to these countries in 1999. large supply increases despite falling prices. Non- We expect a recovery of only 15 percent in oil commodity prices are now one-third below non-oil commodity prices from current cyclical their cyclical high of 1997. Currency depreciation lows over the interval through 2004. This will leave in major commodity exporters in East Asia and non-oil commodity prices 22 percent below their Latin America resulted in sharp price declines for 1997 level. The short-term recovery will be driven 16 CHALLENGES FOR DEVELOPING COUNTRIES DURING THE COMING GLOBAL RECOVERY Box 1.1 The Doha Development Agenda T he Doha agenda has great potential to be beneficial the it e use to,r6 ohniegotiarion andbinding, from a development perspective. A great deal of re ipute settleient to address behind-the-border olicies i search has documented that there is still a large market- de WTO is a qiestioh that developing countries need to access agenda and that dealing with this agenda will answer for theriiselves. The DQha mrunisterial meeting re significantly increase real incomes and reduce poverty in vealed that many-countries had an ariswer to thtrquestion, developing countries (World Bank 2001). Research also ut thait iad others did not. suggests that care is required to determine the develop- The Doha declaration contains numerouscbrnmit ment relevance and payoffs of extending the WTO into, ments by high-income WTO memnbers to providetechnucal regulatory areas (Hertel, Hoekman, and Martin 2002). assistance. However, there is no -ntionof thI:maghitude The key areas of concern for developing countries in the of assistance tha willbe offered, nor is there discussion new trade round will be market access, regulatory issues, (4 any mecharism to determine what the,needs are and and the magnitude and effectiveness of the technical assis- liov.they should'be addressed (that iS, what the delivery tance that was promised in Doha. n1 -echanism rhight be). Embedding technical -assisfate in a Improving market access remains a key goal of multi lbroader devel6pment framework is critical in etsuring that lateral trade negotiations. Industrial countries will need to the assistance focuses on the ptiority needs of each country mobilize the political will to reduce remaining pockets of and is consistent with its development strategy. The sepa- protection in key sectors such as agriculture, labor-based rate section iin he,Doha declaration on technical coopera- services, and labor-intensive manufactures. Developing i -on and ckpacit uilding provides scope to a i bfs countries also need to be willing to liberalize access to - Ldirection: 1Miustersf"instructtlie-Secretariat, in co6rdia their markets for goods and services. The relatively high ctionwith other relevant agetcies, to support doiestic barriers to trade in goods and services that continue to efforts for mainstreaming trade into national plans for prevail in many developing countries implies that they- economic development and strategies for povei-ty.reduc have a lot to bring to the table in a mercantilist sense:. ion" (paragraph 38). A concerted effort will be needed to Identifying a set of "concessions" that are of interest to aeslfre Aid istargeted at, national priorities, aid-o-ensure politically powerful groups in OECD countries and that Brhatassistancei§ providedir neffective manne'byK are beneficial to developing countries is the major chal- agecies witlh comparative advantage in an area. lenge confronting policymakers in the coming years. The Ensuring that the new round of trade -iegofiations research and development communities need to help iden- aieves a pro-development negotiating outcomneis a tify what such issues might be and assist in mobilizing t nmajor challeige Resistance to liberaizaton ofagiiiIture affected constituencies.'3 Ild'textiesis very strong.. Conversely, many -inowjcome As far as multilateral rule-making on regulatory issue countries are,unwilling to extend the reach-of the WTO to is concerned, better understanding of the issues in develop. _over issues suchas competition and investment policies. ing countries is required, not just by government officials A major questioi confronting WTO inembers is whether a but also by the private sector and civil society. Despite five deal should ;5e constructed thatinvolves linking old market years of studying trade and investment-competition link' access isu6RSd disciplines oi new issues such As§invest ages in WTO working groups set up for that purpose, ent ind mpetitior. The feasibility of any such linkage many low-income countries were fearful in Doha of will depend greatly on what is done in the coiing years to launching negotiations in these areas. There is clearly a address deveoping -country concerns regarding im-plemen- need to provide greater assistance to build capacity and tation of Uruguay Round agreements and the magnitude undertake analysis in developing countries to determine and effectiveness of the technical assistance thar'was the merits and implications of multilateral disciplines. ,, promised in Doha. by a rebound in global economic activity, reduced a recovery that is further delayed will have only a supplies and stocks in response to current low limited negative impact on prices. The potential for prices, and some weakening of the dollar. There is unexpected supply increases may be a greater risk. uncertainty associated with the factors that underlie During the 1990s rapid technological progress, the recovery of commodities prices, but the impacts combined with improved policies, led to the emer- of the uncertainties on prices differ markedly. While gence of major producers in a relatively short pe- the timing of the rebound of demand is uncertain, riod of time, resulting in sharp declines in prices (as 17 GLOBAL DEVELOPMENT FINANCE spiked briefly to $31 a barrel immediately follow- Figure 1.10 Real non-oil commodity prices since 1980 ing September 11, but when it became apparent Inde, 190 10; dflatd byMUVthat there were no immediate threats to oil sup- Index, 1980 = 100; deflated by MUV plies, prices quickly fell, ending the year at $18.5. 100 World oil demand grew little in 2001, and actually fell by 1 percent year-on-year in the second half of 80 the year as a result of the after-effects of the attacks 80 (such as reduced jet travel, for example), the deep- Developing countries ening economic slowdown, and mild weather. With non-OPEC (Organization of Petroleum Exporting 60 Countries) production growing moderately overall 60 (increases occurred mainly in the Commonwealth of Independent States, or CIS), oil inventories have Sub-Saharanrisen back to a more comfortable range compared 40 ........... ..Iwith the low levels of 2000 (figure 1.12). 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 Note- MUV is the unit value of manufactures exports from the G-S countries to developing countries, expressed in U.S dollars. Source: World Bank Economic Policy and Prospects Group calculations. within its target range of $22 to $28 a barrel. But, ________________________________________________ with the changed political environment after Sep- tember 11 and as the economic slowdown wors- ened, OPEC chose not to activate its "automatic in the case of coffee). While such supply increases mechanism" that reduces output when the price are difficult to predict, they remain an important of oil falls below $22 for 10 consecutive days. In- risk to the forecast. Conversely, abnormal weather stead, OPEC countries relied on reducing their pro- conditions are more likely to lead to higher prices, duction above quota (estimated at 0.54 million since bad harvests tend to result in much larger falls barrels a day in November) to help support prices. in production than would be the case when good With oil prices well below $20 a barrel in weather conditions boost production. November, OPEC agreed to reduce quotas by 6.5 Oil prices. The global economic slowdown percent or 1.5 million barrels per day (mb/d) be- contributed to a reduction of oil prices from $28.2 ginning January 1, 2002-but only if non-OPEC a barrel in 2000 to $24.4 in 2001."4 Oil prices producers firmly committed to reducing produc- tion by 0.5 mb/d. OPEC threatened a price war if a deal could not be reached. Non-OPEC producers Figure 1.11 Per capita agricultural production responded in part, with major producers Norway Figureand the Russian Federation each agreeing to cut Index, 1980 = 100 production by 0.15 mb/d. While non-OPEC cuts 150 fell short of the 0.5 mb/d demanded, they were Developing countries large enough for OPEC to follow through on its 140 proposed cuts, which will last "as long as neces- 130 sary" according to OPEC's secretary general. We expect oil prices to average $20 a barrel 120 in 2002, somewhat above current levels but well below the 2001 average. It will be difficult to lift 110 Sub-Saharan Africa prices to 2000 levels, mainly because of the under- 100 lying weakness in demand and because non-OPEC capacity has been increased during the recent pe- 90 ii , ,,, ,,,riod of high prices. But with an economic recovery 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 in the second half of 2002, oil demand is expected Source: United Nations Food and Agricultural Organization Statistical Office to increase marginally, following sharp declines in (FAOSTAT). the prior year. Non-OPEC supplies are expected to rise by 1 mb/d, excluding any temporary, volun- 18 CHALLENGES FOR DEVELOPING COUNTRIES DURING THE COMING GLOBAL RECOVERY Figure 1.12 Oil prices and OECD oil stocks Price in dollars per barrel Stock in million barrels 35 2,850 - - 2,800 30 2,750 2,700 25 2,650 2,600 20 2,550 15 --0 stocks 2,500 2,450 10 2,400 Jan. 1996 Jan. 1997 Jan 1998 Jan. 1999 Jan 2000 Jan. 2001 Jan 2002 Source: International Energy Agency; World Bank staff estimates. tary reductions. Consequently, OPEC will be re- Regional developments quired to produce less oil in 2002. If oil producers evere recession in the rich countries, unprece- maintain low levels of output throughout the year, Scdented deceleration in world trade, weak cor- oil inventories could begin to tighten; that would modity prices, and heightened risk perceptions and help firm prices later in 2002 and into 2003, to av- increased selectiveness in financial markets af- erage $21 for the latter year. In 2004 non-OPEC fected all developing regions during 2001." GDP supplies are expected to capture much of the ex- growth for the aggregate of developing and transi- pected growth in demand, and oil prices are ex- tion countries fell from a record 5.4 percent in pected to weaken, to $19 a barrel, as OPEC mem- 2000 to 2.8 percent in the year, and per capita bers continue to lose market share. The increase in growth declined to 1.4 percent, both rates well non-OPEC supply is expected to exceed the rise in below the averages of the 1990s (table 1.3). The demand when global economic growth solidifies. intensity of the international effects differed across The risks to the price forecast are mainly on countries and regions, tied to-among other fac- the downside, since the agreement between OPEC tors-market orientation and product speciali- and non-OPEC producers is likely to be fragile zation in patterns of trade; initial conditions in under expected weak demand conditions. How- domestic financial markets, and different policy ever, while the potential for supply disruptions is measures adopted in response to the slowdown. thought to be small, disruptions could have a large Country-specific conditions are likely to shape the impact if they do occur. The major uncertainties in- recovery onto differing paths of growth by region clude the prospects for exports from Iraq, which following the expected rebound in industrial- will depend on that country's reactions to changes country activity and trade. in the sanctions regime, and any military conflict in The movement from boom to bust in the ex- the Middle East due to the war on terrorism. The ternal environment is reflected distinctly in the fall impact of the latter could be extremely significant. of export market growth from 13 percent to 1.1 For example, the loss of 5 mb/d of Iranian produc- percent, and the concomitant decline in developing- tion in 1980 caused a 150 percent rise in prices country export performance from 15 percent to 4 within several months, and the similar-size loss of percent-although this movement still implies a Iraq and Kuwait production in 1990 caused a tem- pick-up in market share for the group. Terms of porary doubling of prices within three months. trade, expressed as a proportion to GDP, dropped 19 GLOBAL DEVELOPMENT FINANCE Table 1.3 Developing-country forecast summary, 1991-2004 (percent per year) Estimate Forecast Growth rates/ratios g biii T M) 2001 2002 2003 2004 Real GDP growth 2.8 3.2 5.0 4.9 Consumption per capita 0.5 1.8 3.0 3.1 GDP per capita 1 1.4 1.8 3.7 3.5 Population 1io 1.4 1.4 1.3 1.3 Gross domestic investment/GDP' 235 2m 24.2 24.6 24.8 25.0 Inflationb y5.3 4.4 4.2 4.1 Central government budget balance/GDP 3. 4O -.2 -3.2 -3.5 -3.5 -3.1 Export market growth, o 1.1 2.5 7.7 7.4 Export volumed 714.3.7 6.1 9.6 9.4 Terms of trade/GDP ,-0.2 -G(pr4 -0.1 -1.1 -0.2 -0.3 Current account/GDP n ,20020.4 -0.2 -0.4 -0.7 Memorandum Ttems GDP growth: developing excluding the transition countries cc2.6 3.2 5.2 5.0 Excluding China and India e1.8 2.2 4.4 4.2 Excluding transinon, Chmna, India 3t u 2p 1 48 1.3 2.0 4.4 4.3 a. Fxed investment, measured m real terms. b. Local currency GDP deflator, median. c. Weighted average growth of import demand in eport markets. d. Goods and nonfactor services. e. Change n terms of trade, measured as a proportion to GDP (percent). Source: World Bank baselmne forecast, February 2002. by 0.1 percent. These developments pushed export epr revnue lor tha ion as revenues into negative territory (a decline of 1.3 te rice il fellishapydorese ctes, percent), and contributed to se cowies ith a though, finncigg- dificltie ar not as ping gregate Current account surplus to 0.4 percent of ccumtedubntia current aonturpuses GDP in the year. At the same time, however, under- an rsve overehe ast s evee s More lying inflation trends have continued on a path of neabe are countries tate nal onpnon-oil deceleration, central government budgt balances co moities export reies have narrowed from the averages of the 1990s, and blino . ecn frgoa D) a ags a general improvement in the investment climate in fsa siu. Osa ters; these ountries ua he many countries, including new emphases on gover- lstn-cre bor o ts. ost nance and institutional reforms, have helped main- potrthuguthedvlinwrdhaesn tain the flow of FDI into selected developing and terepr eeusfl oeta 10blina transition economies at high levels. These factors thprcofilelsapy.Frhseoure, have opened the door-for those countries with a prsiga financing pre for coures favorable climate-to pursue countercyclical policy such as Ey and inaethatiha amassed options to help mitigate the full brunt of the exter- mlae f aialace nal shocks of 2001. For example, large levels of re- serves, low inflation, and manageable government nrbeaecutista eedlreyo o-i debt enabled many countries in East Asia to reduce cmoiiseprs ro ors,ohrsrie interest rates and to implement fiscal stimuli. Other reipsantasfs;hsecuressalyav countries, with weaker initial conditions (includ- lesta-rdwohybrwigttu.M t ing, for example, Indonesia), several countries in prsigaetefncngrolmfrcutis Latin America, and Turkey, were forced to persist suhaTrkyndAgtiahthdamse in isal onoldaton o evn o tghenfusam, er timge thnatciaccs tobainterntoalcpta.a grw2oelmtd.Dciei0xor eeps(2 CHALLENGES FOR DEVELOPING COUNTRIES DURING THE COMING GLOBAL RECOVERY Figure 1.13 GDP growth in developing regions Percent 8 7 6 _ _2001 U 2002 0 2003 2004 5- 4- 2- East Asia South Asia Latin America Central and Middle East Sub-Saharan and Pacific and the Caribbean Eastern Europe/ and North Africa Africa Commonwealth of Independent States Source: World Bank Economic Policy and Prospects Group data and projections. For developing and transition countries as a Recovery in the developing world is likely to group, recovery is anticipated to build momentum begin, and to be strongest, in East Asia, where over the course of 2002. Growth is expected to countries have benefited from domestic stimuli reach 3.2 percent in 2002, and rise to 5 percent and where strong dynamics in the high-tech sec- during 2003-04 (table 1.3). A rebound in export tors could once again work in their favor (figure market growth to rates near 8 percent by 2003 1.13). In contrast, little recovery for the aggregate would suggest a return of export performance to- of Latin American counres is anticipated, given ward double-digit gains. Terms of trade for the ag- their much less favorable starting points, since fi- gregate of developing countries is likely to worsen nancial strains remain elevated and commodity in the short to medium term, since it is tied in part prices are expected to rebound only modestly. Sub- to the large weight of oil exporters in the group, as dued commodity prices will also continue to re- well as to anticipated increases in the dollar cost strain economic growth in Sub-Saharan Africa. of manufactures imports from the industrial coun- The war on terrorism could hamper growth tn tries."6 Nonetheless, strong export volume growth South Asia and the Middle East and North Africa should underpin domestic investment, with posi- in the short run as trade and tourism flows remain tive multiplier effects, and falling inflation should disrupted, while at the same time financial flows boost real incomes and consumption. A gradual to frontline states should ease current account ten- return of private capital to emerging markets will sions. In the medium run, necessary fiscal austerity accentuate these developments, so that by 2003 in South Asia is expected to dampen growth rates growth will be returning toward 5 percent. More- in the region somewhat. Recovery in Central and over, recent developments, including the Doha Eastern Europe will hinge critically upon develop- Round, China's accession to the WTO (and the ments in the European Union (EU), suggesting Russian Federation's expressed interest in the or- somewhat delayed recovery relative to East Asia, ganization), offer promise of a broader scope for while the Russian Federation and other countries fuller participation in global trade, which will ben- of the CIS are likely to see recent stronger rates efit the new members and their trading partners of growth-linked in large measure to the prtce of alike. o il-fade gradually over the next years. 21 GLOBAL DEVELOPMENT FINANCE East Asia and Pacific cial difficulties in Indonesia-and to a lesser degree, Growth in East Asian developing countries slowed in the Philippines-were being addressed through to 4.6 percent in 2001 from the 7.4 percent regis- agreements with the International Monetary Fund tered during the 2000 boom. The growth slow- and multilateral development banks. down in the region, excluding China, was more East Asia may be the first developing region dramatic-from 7 percent in 2000 to 2.3 percent. to emerge from the current global downturn, and Chinese growth remained above 7 percent, boosted growth there is expected to pick up to 5.2 percent in by large-scale fiscal stimulus. 2002-reflecting the positive impact of looser mon- The collapse of global demand for high-tech etary and fiscal positions and improvement in exter- products, compounded by progressively weaker nal conditions. But the strength of recovery will economic conditions in the United States and Ja- hinge upon the revival of world trade and rise in pan, hit exports, industrial production, and invest- global demand for technology-based products. ment in most countries quite hard and raised un- There are some early signs of encouragement in employment rates. Regional export volume growth the information and communications technology slowed sharply to 3 percent-in contrast to the ro- (ICT) sector, as world semiconductor sales appear bust 22 percent advance of 2000-with the largest to have reached a trough. Industrial production is growth decline occurring in the five countries most now rising across key ICT-producing economies of affected by the 1997-98 Asian crisis. Manufactur- the region-notably Korea, but also Malaysia, the ing output in the larger countries, excluding China, Philippines, Thailand, and the NIEs. As demand dropped by some 7.5 percent, fixed investment is unlikely to gain substantial momentum until slowed by 4 percentage points, and liquidation of the second half of 2002, however, a more robust unwanted inventories played a substantial role in export-led recovery in East Asia is not likely until the downturn, subtracting more than 1 percentage 2003, with GDP growth expected to reach about 7 point from the regions' output in the year. The percent, before moderating toward potential growth high-income, high-tech-dependent entrep6t centers of 6.5 percent in 2004. Challenges will remain dur- of the NIEs were battered into recession despite ing recovery, especially the potential widening of strong monetary and fiscal stimuli; this led to a fiscal balances and the need to re-address fragile sharp compression of East Asia's intricate network banking systems in several countries. China's recent of intraregion trade. The events of September 11 accession to the WTO offers the broader region only exacerbated the difficult external env'ironment both substantial opportunities, in an opening of the facing the region, especially for tourism revenues, large Chinese market to the region, and potential as tourist arrivals in the five leading Association competitive pressures in third markets, because of Southeast Asian nations countries are thought to these open wider to Chinese products. have fallen by 10 to 15 percent in October (year- over-year). Latin Ametica and the Catibbean Low and declining inflation rates allowed most Regional GDP grew 0.6 percent in 2001 in Latin countries to use fiscal and monetary stimuli to miti- America and the Caribbean, a substantial slow- gate the downturn. For example, the Republic of down from the 3.8 percent advance registered in Korea lowered interest rates by 140 basis points, 2000. The weak growth performance reflects ad- stepped up fiscal outlays-with the central govern- verse external conditions alongside a progressive ment balance deteriorating from a surplus of 1.3 worsening of the political and economic situation in percent of GDP in 2000 to a small deficit in 2001- Argentina. Output in Latin America, excluding Ar- and tapped international capital markets for gross gentina, increased by 1.3 percent in the year. Fol- flows of some $21 billion in the year. These mea- lowing September 11, economic conditions wors- sures provided cushion for domestic demand while ened for the region as Argentina's crisis deepened, increasing reserve levels. Similar policy measures by commodity prices fell, secondary market spreads several other economies in the region (with the rose, and capital flows fell from already subdued exception of Indonesia) yielded a widening of the levels in July and August. The Caribbean region average fiscal deficit to 3 percent of GDP from 2.5 witnessed a steep decline in tourist bookings, while percent in 2000., while the current account surplus weakening labor markets in North America led to a position diminished by 1.5 percent of GDP. Finan- slackening of remittances to Central American and 22 CHALLENGES FOR DEVELOPING COUNTRIES DURING THE COMING GLOBAL RECOVERY Caribbean countries. Few countries (among them, industrial world eases. By that time private capital Chile and the Rep6blica Bolivariana de Venezuela) flows will have increased again, and earlier recov- were able to pursue countercyclical fiscal policy or ery in industrial countries should boost the price of monetary expansion to mitigate the growth slow- the region's primary commodities and the volume down, due to generally high public debt and rela- of manufactured exports. tively large external financing requirements. These developments translated into a rise in regional un- Europe and Central Asia employment, with falling inflation rates in most Europe and Central Asia grew by 2.2 percent in countries, but little change in real interest rates or 2001, contrasted with 6.4 percent growth in 2000. fiscal balances. The sharp deceleration was due to a 7.5 percent International developments were a major con- contraction in Turkish output, the fall in Russian straint on external revenues in 2001. The regional growth to 4.8 percent following robust 8.3 percent trade balance moved from a deficit of $35 billion performance in 2000, and a 0.9 percentage point in 1998 to a surplus of almost $10 billion in 2000 deceleration in Central and Eastern European our- on the back of rising surpluses for major oil ex- put. Growth for the region, excluding Turkey, porters. During 2001, however, aggregate dollar amounted to 4.4 percent, down from 6.2 percent exports declined 1.5 percent and imports fell 1 per- in 2000. Most transition economies witnessed de- cent, narrowing the trade surplus by about $3.6 dining inflation and interest rates, reflecting lower billion. Oil exporters saw their surpluses diminish import prices and falling international interest while Argentina and Brazil raised their surpluses rates. However, adoption of accommodative fiscal significantly. In combination with these trends, a and monetary policies in the face of slowing growth softening of receipts from tourism and remittances led to a slight deterioration of fiscal deficits in sev- contributed to a widening of the region's current eral Central European countries. account deficit by $5 billion. With declines in fi- Developments during the year served to nar- nancing from international capital markets, the row current account surpluses for those countries current account deficit was balanced by a draw- recently attaining positive balances (for example, down of reserves and increased support from the Kazakhstan, the Russian Federation, and Ukraine) international financial institutions. and widened deficits for countries whose external The outlook for 2002 has dimmed, with GDP balances have remained persistently negative (such now expected to rise by 0.5 percent-assuming as Bulgaria, Croatia, Romania, and the Slovak that the repercussions of the Argentine default and Republic). This reflects delayed spending of oil devaluation have been discounted by financial mar- revenues (as in the Russian Federation and Ka- kets, and that regional contagion remains limited. zakhstan), and a deterioration in the external en- The forecast revision is also due to a much weaker vironment, particularly weaker external demand fourth quarter 2001 outturn for most countries- from the EU area. There are exceptions. In Turkey, implying delay to the recovery, the growth-eroding the current account deficit shifted into a $3 billion effects of crisis for Argentina itself, and a decidedly surplus in 2001, as net external finance plummeted weaker outlook for private-capital market and and the February 2001 crisis resulted in drastic business-related foreign direct investment (FDI) in- measures to reduce domestic demand, and to switch flows. Fiscal deficits were deteriorating sharply at expenditure, including a 56 percent depreciation of the end of 2001 for a number of countries due to the lire. In Poland compressed domestic demand slowing growth and continued declines in the (linked to previously tight monetary policy, easing prices of commodity exports, and government debt as of late 2001) has contained imports, translating levels have risen. Hence fiscal consolidation may be into a narrowing of the current deficit, from $10 required in 2002 to avoid excessive debt burdens, billion to $7 billion in 2001. and this may constrain governments' ability to sup- Growth in the region is expected to pick up port growth through increased spending. Growth modestly in 2002, to 3.2 percent from 2.2, but is expected to recover to 3.8 percent in 2003-yet largely based on the assumed strength of recovery with considerable downside risks, should Argen- in Turkey. In contrast, among the transition econ- tina's output decline become more protracted- omies, growth in the CIS is anticipated to decline maintaining growth at that rate during 2004, as the to 3.8 percent in 2002, driven principally by a sharp 23 GLOBAL DEVELOPMENT FINANCE decline in Russian oil revenues. Growth may ease and slower import growth are expected to keep moderately in Central and Eastern Europe from India's current account deficit well below 2 percent 2.9 percent to 2.8, while recovery in the Euro Area of GDP. FDI inflows ballooned to $4.5 billion in develops only gradually and fiscal consolidation the year, twice the level of any previous fiscal year. may be necessary for potential accession countries Given a comfortable foreign reserve position, India to the EU. The region as a whole should see an ac- is unlikely to face tight constraints in external fi- celeration of growth to between 4 and 4.5 percent nance. But increasing direct government spending in 2003-04, as the eventual pickup in Europe in- and subsidies, in India as well as in Bangladesh and creases demand for the region's exports, although Pakistan, will tend to widen central government fis- continued sluggish oil markets will partially con- cal deficits-to 5.3 percent, 6.3 percent, and 5.3 strain growth in the CIS. percent respectively-and these deficits are likely to remain impediments to a more robust acceleration South Asia of growth in the medium term. Although South Asia is relatively less integrated Output in the region is expected to gain mo- into the global economy than most developing re- mentum over 2002-03, partly on the strength of gions, trends in the external environment served to global trade recovery, although political and mili- restrain the pace of growth during 2001. Growth tary tensions in the region create large uncertain- rose from a 4 percent advance in 2000 to 4.3 per- ties. Removal of sanctions by the United States on cent in 2001, as a decline in manufacturing out- India and Pakistan and a potential pick-up in rex- put offset general improvement in agricultural per- tile and clothing exports linked to eventual opening formance (agriculture accounts for 50 percent or of rich-country markets are additional favorable more of output for all countries of the region). Ex- factors that could support the medium-term out- port market growth declined abruptly and sharply, look. And hoped-for progress in addressing struc- leading to a fall in regional export growth to 1.1 rural reforms across countries of the region should percent from the strong 12.3 percent outturn of support gains in productivity. Regional output is 2000. Indian exports, for example, dropped by 2 expected to register growth of 4.9 percent in 2002, percent over the period from April to September before rising somewhat faster over 2003-04 at a compared with the levels from a year earlier. pace above 5 percent. Manufacturing output in that country showed no growth in the first half of the calendar year. Pak- Middle East and North Africa istan will clearly pay a toll in economic activity for Middle East and North Africa region GDP slowed the duration of the military activities in Afghani- to 3.1 percent in 2001, following above-average stan, but it will also receive adequate financial sup- growth performance of 4.2 percent during 2000. port from the international community to reduce Cutbacks in oil production by OPEC members of debt-servicing requirements, possibly establishing the region to support oil prices within a target a foundation for renewed growth. band, coupled with volatility-and recent sharp Given the size and relative self-sufficiency of declines-in the oil price, depressed growth among the Indian economy, tepid domestic demand is the the major hydrocarbon producers. For example, main culprit behind the current sluggishness of following a rise of some 4.5 percent in 2000, GDP growth, although external factors have played a in Saudi Arabia advanced by slightly less than 2 greater role than was typical in the past. Investment percent in 2001. Ar the same time, progressive is slowing, in part due to the slackening of export weakening of conditions in continental Europe growth, and capital goods output dropped 8 per- (the dominant export market for countries of the cent during the first half of fiscal 2001. However, Maghreb and several countries of the Mashreq) positive developments on the inflation front, with dampened export performance substantially-Mo- the consumer price index moving below 3 percent, roccan export growth dropped into negative tern- provided some headroom for easing of monetary tory during the first half of the year. These trends policy in response to increasingly weak conditions. were exacerbated by declines in revenues from The recent fall in oil prices, continued growth of tourism and remittances due to heightened security software exports (albeit at reduced 30-percent rates), concerns after September 11. Against this back- 24 CHALLENGES FOR DEVELOPING COUNTRIES DURING THE COMING GLOBAL RECOVERY ground output growth for the oil exporters of the protracted relative to other developing regions, ris- region dropped from 3.6 percent in 2000 to 2.5 ing by 3.3 percent over 2003 and 2004. percent in 2001; and with the exception of Mo- rocco, which was recovering from severe drought Sub-Sabaran Africa conditions, growth among the diversified exporters Growth in Sub-Saharan Africa eased to 2.6 percent of the region slowed to 3.2 percent from 4.7 per- in 2001 from 3.1 percent in 2000, as the global cent in 2000. slowdown exacted a toll on commodity prices and An important consequence of these develop- growth in the region's export markets. The slowing ments has been a substantial waning of external of Sub-Saharan Africa's aggregate growth was surpluses across the region. This is most evident moderate because oil exporters enjoyed relatively among the oil-exporting countries, where current high oil prices for much of the year, and favorable account balances that ballooned to some $59 bil- weather conditions boosted agricultural produc- lion (13 percent of GDP) with the jump in oil prices tion in several countries (for example, cocoa pro- in 2000, dropped quickly to less than $40 billion duction in West Africa increased sharply). But on the back of slumping prices and curtailment of terms-of-trade losses as a proportion to GDP were exports. Although public spending levels were ad- 1 percent, the worst performance outside of the justed in many countries, fiscal deficits increased. Middle East and North Africa region, and export In the case of Saudi Arabia, despite public sector market growth fell from 11 percent in 2000 to wage restraint, the 2002 budget foresees a deficit of 1 percent. These fundamental conditions were re- some $6 to $7 billion, contrasted with a surplus of flected in African high-frequency data covering similar magnitude in 2000. Similar adverse fiscal production, trade, and financial markets, which trends are affecting countries such as the Arab Re- indicate that, as elsewhere, economic conditions public of Egypt, Morocco, and Tunisia, and may deteriorated sharply over the course of the year. broaden across the diversified exporters as external Growth of regional export volumes dropped by revenue shortfalls become more acute in the near 5.4 percentage points, to 2.1 percent, and revenues term. by 24 percentage points, to -4.3 percent from 2000 Some countercyclical policy actions have been outturns. Moreover, weak tourism demand in the possible. Improved inflation performance in Egypt critical year-end period-and in the wake of Sep- has allowed a full percentage point reduction in tember 11-further affected a number of countries the central bank discount rate; and exchange rates dependent on tourism, especially Kenya and Tanza- have been falling relative to the dollar as well as nia. In South Africa GDP registered growth of 1.2 the euro over the second half of 2001 in Egypt, percent (seasonally adjusted annual rate) in the Morocco, Tunisia, and the Republic of Yemen. third quarter, down from a recent peak of 3.4 per- These measures may help to mitigate the effects of cent in the fourth quarter of 2000. A deterioration the global slowdown to a modest degree; but given in the country's trade balance coupled with a de- the importance of the EU as an export market and dine in equity capital flows precipitated a sharp a principal source of remittance and tourism in- fall in the value of the rand, which lost nearly a come, recovery there will be necessary for a return third of its value over the fourth quarter. of more buoyant external conditions in the Middle Looking to 2002, the projected decline in oil East and North Africa. prices will adversely affect fiscal and external bal- Given difficult conditions in the external en- ances of hydrocarbon exporters, but at the same vironment, near-term prospects appear muted: time it will provide a degree of relief to the large growth recovery in the EU is likely to lag behind number of oil-importing countries of Sub-Saharan that of North America and East Asia; underlying Africa. Oil contributes 70-80 percent of export rev- demand for hydrocarbons will require some time enues for Angola, the Republic of Congo, Gabon, to reach 1999-2000 levels, and uncertainty asso- and Sudan, and more than 90 percent for Nigeria ciated with the war on terrorism will likely remain and Equatorial Guinea. It is also the source of a a dampening factor for regional dynamism. GDP majority of government revenues, pointing to a dif- growth is anticipated to fall to 2.7 percent in 2002, ficult period of fiscal consolidation. At the same while recovery over the following years may be time lower oil prices, if sustained, reduce the attrac- 25 GLOBAL DEVELOPMENT FINANCE tiveness of FDI flows into new production facilities of 1991) resulted in a 0.5 percent annual decline of in southern and western Africa. Elsewhere, rev- GDP in 1991 over 1990. From Spring until late enues from tourism are also likely to remain de- Fall of 1990 international organizations forecast an pressed pending a resumption of faster growth in increase of around 2 percent,18 implying an average the industrial countries (even without concerns forecast error of 2.5 percentage points. In 1991 the over security in the wake of the September 11 at- forecast errors were reduced to on average 0.3 per- tacks), and the recovery in non-oil commodity centage points. The recent U.S. recession-reflected prices is expected to be relatively muted. This bal- in the 1.1 percent GDP growth in 2001, compared ance of factors suggests that regional output should to the more than 4 percent growth in 2000-pro- only maintain growth of 2.6 percent in the year. vided an almost identical picture. The average fore- Both export revenues and the terms of trade cast error in 2000 (for growth in 2001) was 2 per- may decline slightly in 2002, requiring a further centage points, and it dropped to 0.3 percentage 3 percentage point reduction in import growth. points in 2001 (figure 1.14). However, for the 19 Sub-Saharan Africa countries This experience implies that uncertainty may that have fulfilled the conditions for debt relief be relatively small for the 2002 growth rate fore- under the Heavily Indebted Poor Countries Initia- casts, but substantially larger concerning the tive, a reduction in debt service payments by $656 strength of the recovery in 2003. Figure 1.15 shows million compared to the average of recent years that the current cycle, including the baseline fore- will provide some offset to reduced export rev- cast, is expected to have a recovery pattern similar enues. Conditions in export markets (particularly to the 1990-91 cycle. Although the recent recession in Europe) are expected to improve progressively seems more shallow, the deceleration in growth was through the year, setting the stage for 3.6 percent actually quite similar, as could be the acceleration. GDP growth over 2003-04, when oil prices may With the larger share of high-tech production in stabilize and non-oil commodity prices rise by a the current cycle, and possible further stimulus cumulative 15 percent. packages, the recovery could even turn out to be sharper. However, there are also significant down- side risks to this prediction. The prospects for high- tech industries depend, to a large extent, on the Risks to the forecast sentiment in financial markets, which is notori- U ncertainties involved in macroeconomic fore- ously difficult to predict. Continued nervousness casts are sizeable, and substantial forecast er- about future profitability could make the recovery rors are virtually impossible to avoid. Errors in more fragile than is currently forecast. GDP growth forecasts made one year ahead tend The prospects after the coming recovery are to average around 1.5 percentage points.17 Once even more uncertain, particularly given that earlier leading indicators or partial data are available, recessions were often followed by a second dip. the accuracy of forecasts improves dramatically. For example, in the beginning of 1993 U.S. GDP Current-year forecasts of GDP growth typically growth again fell below zero following the Euro- have errors substantially below 1 percentage point. pean recession. Since the current regional cycles It is extremely difficult to predict cyclical develop- are much more synchronized than a decade ago, ments well in advance, partly because the timing such a strong double dip is not foreseen in the of turning points is highly uncertain. baseline. However, the cumulated financial imbal- The prediction of recessions or severe down- ances in the U.S. economy could set off another re- turns is particularly difficult, since they are often versal in market sentiment, leading to a sharper triggered by the burst of a speculative bubble or slowdown after the current recovery than is antici- other unforeseeable events. Even if some tensions pated in the baseline forecast. In other words, a were observable in advance, the timing of their un- major risk is that the cyclical pattern could be winding is close to random. The U.S. recession in more pronounced than is assumed in the baseline, the early 1990s provided an example of how fore- with a stronger recovery, but a substantial reversal casters can fail to anticipate recessions. The con- in the medium run. traction of the U.S. economy (that started in the Although the recovery may be stronger than third quarter of 1990 and ended in the first quarter currently anticipated, possible downside risks de- 26 CHALLENGES FOR DEVELOPING COUNTRIES DURING THE COMING GLOBAL RECOVERY Figure 1.14 Forecasting the 2001 U.S. slowdown Month forecast was made, percent 4 M WB Mg IMF 3 ION OECD -Consensus 2 Actual outcome 0 Jan. April July Oct Jan. Apr. July Oct. Jan 2000 2000 2000 2000 2001 2001 2001 2001 2002 Source: World Bank; International Monetary Fund; OECD Secretariat. Figure 1.15 Two recessions in the United States, 1990-91 and 2001 Annualized quarterly growth, trough of recession = 0 10 8 Ol0, 1988-04, 1994 A- 04, 1998-04, 2001 6 1, 2002-Q3, 2004 (f) 0 -2 1 1-14 t I l l l l l l l l l i l i i l l -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 Source: World Bank Economic Policy and Prospects Group. serve more attention, since they often pose more right Japanese banking crisis in the short run, the serious challenges than do upside risks. Because downside risks are significant. the baseline forecast does not anticipate new major Japan is mired in deep recession and deflation, adverse shocks to the global economy, assumes with corporate profits declining sharply and bank- only limited contagion from the breakdown of the ruptcies mounting, and is beset by heightening Argentine economy, foresees an uninterrupted re- concerns about credit availability and the sound- covery of Turkey's economy, and excludes an out- ness of the banking system. Commercial banks 27 GLOBAL DEVELOPMENT FINANCE have become hesitant to lend, while the banks' a degree of budgetary flexibility during severe recessions. capital base is being eroded by falling equity While the projected downturn in European economic activ- prices--commercial bank stocks dropped 45 per- ity could not be described as a severe recession, the Septem- ber I1I attacks would certainly qualify as unusual events cent during 2001. Credit availability for smaller outside the control of member states. And some flexibility in companies is tight, a flight to quality into Japanese fiscal positions may be witnessed in the short run. government bonds has ensued, and Japanese sov- 4. On a momentum basis (quarter-over-quarter), these ereign debt has been downgraded by Moody's and economies experienced the deceleration earlier, with a de- othercline of 9 percent (seasonally adjusted annualized rate) in othe crditratig aences.the last quarter of 2000 and the first quarter of 2001, be- However serious these problems are, the fore reaching 25 percent decline at the trough in the second probability of a full-blown crisis seems to be rela- quarter. tively low in the near term because the economy 5. Almost 40 percent of each year's annual growth rate will benefit from the recovery of export demand, is determined by the quarterly growth pattern in the previ- ous year. The contribution of the previous year's quarterly possibly fueled by a weakening of the yen. Such a growth to the current year's annual growth is called "carry- depreciation could help to fight deflation through over. an increase in exports and a rise in import prices. 6. The average trade margin for total exports from in- This could have a negative impact on emerging dustrial countries toward developing countries is 3.8 per- countries in Asia who compete with Japanese ex- cent, but is 5.5 percent for developing-country exports to- ward industrial countries. porters, depend on Japanese imports, or are recipi- 7. The impacts of higher international trade margins ents of Japanese FDI. However the adverse im- were evaluated using the World Bank's global computable pacts are likely to be limited in the case of a general equilibrium model of world trade (van der Mens- modest depreciation, since the yen has appreciated brugghe 2001). in recent years, most countries in the region have 8. World Tourism Organization, Tourism Industry Takes Action to End Crisis, November 12, 2001. www. adopted flexible exchange rate systems, and a world-tourism.org. gradual real depreciation of the yen seems war- 9. World Tourism Organization third quarter 2001 ranted from a structural perspective. news release. Other information confirms the sharp drop in Whatever happens in the current rebound of tourism: two months after September 11 worldwide travel the global economy, the challenges are formidable reservations were 12 to 15 percent below levels of the previ- ous year. in the medium term. The escalation of Japan's fis- 10. Not all of the countries highly dependent on travel cal deficit has limited the scope for large injections services are tourist destinations. A few countries affected by of public funds for re-capitalization or closure of conflict (for example Sierra Leone and Rwanda) are depen- institutions. The major risk of a severe credit crunch ent on revenues from travel services, probably due to the presence of staff from international organizations and non- is growing rather than shrinking. A sharp fall in governmental organizations, as well as the presence of peace- Japanese domestic demand would be a major set- keepers. The data from IMFs Balance of Payments database back for developing economies in East Asia, with, lack sufficient detail to separate out the different purchasers for example, 15 percent of Chinese exports and of travel services for these countries. 11 percent of Korean exports going to Japan. 11. An interesting example of the impact of technology 11 percen commodity production is the new technique for cutting two-by-fours from logs. In the past a curved log could not be used to produce a straight board without huge wastage. Notes However, lasers and computers are now used to scan a log 1. See Global Economic Prospects 2002 (World Bank and cut with the curve of the log. The two-by-fours are then 2001). pressed and dried to produce a straight board from a crooked 2. The so-called accelerator mechanism makes inven- log. tory and investment cycles much more pronounced than 12. Vietnam reformed coffee marketing, which re- cyclical developments in other components of aggregate de- suIted in a large increase in the producer's share of interna- mand. Firms generally attempt to keep the stock of invento- tional prices and led to a significant increase in exports. ries and capital goods at a desired ratio to GDP. This implies 13. To some extent this has already begun. A notewor- that the flows of inventory accumulation and investment are thy difference between Doha and previous ministerials was linked to changes in GDP. Thus as stocks reach desired lev- the active involvement of representatives of development els, the change in inventory accumulation and investment ministries on national delegations. National development from the previous period can be quite large, generating communities and stakeholders represent a potentially pow- sharp changes and turning points in GDP growth. erful constituency in many European countries. 3. The Stability and Growth Pact, setting out the rules 14. The price used to represent oil market conditions is for budgetary behavior in stage three of the European the average of West Texas Intermediate and Brent and Dubai Union's (EU's) Economic and Monetary Union, provides for crudes, and is roughly equivalent to the Brent price. 28 CHALLENGES FOR DEVELOPING COUNTRIES DURING THE COMING GLOBAL RECOVERY 15. See appendix 4 for a fuller treatment of recent Hertel, Thomas W., Bernard M. Hoekman, and Will Mar- macroeconomic and financial developments and prospects tin. 2000. "Developing Countries and a New Round of for the developing regions. WTO Negotiations." World Bank, Washington, D.C. 16. The manufacturing unit value (MUV) of exports in Processed. dollar terms from the G-5 countries to developing countries IMF (International Monetary Fund). 1990a. World Eco- is anticipated to rise by 3.6 and 3.7 percent respectively in nomic Outlook. Washington, D.C. 2003-04, reflecting market expectations for a likely weak- . 1990b. World Economic Outlook. Washington, ening of the dollar against the euro over the next years, D.C. counterbalanced by a trend of strength relative to the yen. Loungan, Prakash. 2000. "How Accurate Are Private Sec- For non-oil developing-country exporters, such develop- tor Forecasts? Cross-Country Evidence from Consen- ment is likely to offset part of the firming of non-oil com- sus Forecasts of Output Growth." IMF Working Paper modity prices, while mitigating gains from lower fuel im- WP/00/77. IMF, Washington, D.C. port prices. For hydrocarbons exporters, the up-trend in OECD (Organisation for Economic Co-operation and De- MUV will serve to pressure terms of trade yet further. velopment). 1990a. OECD Economic Outlook 47 17. See, for example, Batchelor 2001 and Loungani (June). 2000. . 1990b. OECD Economic Outlook 48 (December). 18. IMF 1990a and 1990b. The World Bank did not van der Mensbrugghe, Dominique. 2001. "Linkage Techni- produce annual forecasts at that time. cal Reference Document." World Bank, Washington, D.C. World Bank 2001. Global Economic Prospects 2002. Wash- ington, D.C. References World Tourism Organization. 2001: "Tourism Industry The word processed describes informally reproduced works Takes Action to End Crisis." November 12. www. that may not be commonly available through libraries. world-tourism.org News bulletin. Third quarter 2001 news release. Batchelor, R. A. 2001. "How Useful Are Forecasts of Inter- governmental Agencies? The IMF and OECD versus the Consensus." Applied Economics 33 (2): 225-35. 29  2 Private Capital Flows to Emerging Markets The global slowdown reduced capital plined domestic monetary polices, and lower market flows to developing countries short-term debt. Finally, there is more work to be T he global economic slowdown in 2001 trans- done on private sector involvement in crisis pre- lated into reduced private capital flows to de- vention and resolution. Recent experience has un- veloping countries. The reevaluation of prospective derlined the importance of a cleat definition of the returns in technology investments severely reduced limits on official resources and of the role and re- demand for developing countries' technology sponsibilities of the official sector, debtor coun- stocks. Further, the global slowdown and collapse tries, and their private creditors. This challenge of equities prices increased the riskiness of the debt points to the need to consider more ambitious pro- of highly leveraged corporations, reduced investors' posals for facilitating orderly workouts of prob- appetite for risk, and increased economic uncer- lematic private sector debts, and the recent pro- tainty. All of these had the effect of tightening bank posal by the International Monetary Fund (IMF) lending criteria and reducing access by speculative- to provide for a standstill of debt payments to grade borrowers, which sharply depressed bank allow time for an orderly restructuring will, no lending to developing countries. By contrast, bond doubt, be debated in the year ahead. issues by developing countries remained stable, be- cause the share of developing-country investment- No significant recovery in capital flows grade borrowers is greater among bond issuers than until 2003 bank borrowers. The level of foreign direct invest- Capital market flows are forecast to decline further ment (FDI) in 2001 was virtually unchanged from in 2002. Investors are likely to remain cautious the previous year, with changes in flows largely dri- about emerging markets, because low growth and ven by changes in the domestic economic environ- recession in industrial countries limits demand for ment, by large privatization transactions, or by a developing countries' exports, financing constraints few major private sector acquisitions. on banks and other investors remain tight, and the appetite for risk remains low. The recovery antici- Financial crises highlighted the problems of pared to begin in the second half of 2002, coupled rescue packages with low interest rates, should spark a rise in capi- The crisis in Argentina highlighted the challenges tal market flows in 2003-04. Nevertheless, the in- facing the international community in assisting crease in flows will remain modest, since commod- countries in crisis. Fixed exchange rate regimes are ity exports will continue to experience low export vulnerable to asymmetric shocks. There are severe revenues, investors will remain concerned after the costs associated with hanging on to a pegged, over- string of emerging market crises since the mid- valued exchange rate. The success of multilateral 1990s, and low rates of capacity utilization will re- rescue packages depends critically on strong ad- duce the need for capital in some of the more cred- justment by recipient countries. Contagion can be itworthy developing countries. FDI flows should contained through prudent external financial man- remain high, and perhaps rise somewhat, over the agement, including flexible exchange rates, disci- next few years, while growth in developing coun- 31 GLOBAL DEVELOPMENT FINANCE tries accelerates and they continue to enjoy the ben- lower interest rates and increased access by devel- efits from sustained improvements in policies over oping countries helped to cushion the impact of the past 10 years. FDI flows are likely to remain the the global recession. FD1, which is less sensitive to largest source of external finance for developing cyclical changes in output than capital market countries. flows, was little changed from the previous year, and remained only $16 billion below the peak level of 1999. Net resource flows The global slowdown has depressed capital Capital market flows flows to developing countries eveloping countries' access to capital markets Developing countries' net long-term flows (gross D deteriorated substantially in 2001. Total cap- inflows of capital less amortization) fell to an esti- ital market commitments (bank loans, bond issues, mated $196 billion in 2001, $65 billion below the and portfolio equity) declined to an estimated previous year's level and $145 billion less than the $171 billion, about one-quarter less than the level peak in 1997 (see table 2.1, and see annex 2.2 for in 2000 (see table 2.2). External factors played the a definition of the measurement of capital flows predominant role in reducing external finance. used). Expressed as a share of gross domestic prod- The slowdown in industrial countries led to a de- uct (GDP), net long-term flows have fallen from dine in developing countries' export revenues, the 5.3 percent in 1997 to 3.1 percent in 2001. Deteri- impact of which was only in part mitigated by orating prospects for developing countries, the col- the drop in international interest rates. Because lapse in the price of technology stocks, the crises in most developing-country borrowers are specula- Argentina and Turkey, and increased concern over tive grade, they were hurt by a widespread retreat risk have reduced demand for developing-country from speculative-grade investments. Slower growth debt. Speculative-grade borrowers saw a sharp fall and the collapse of technology stock prices in- in access, with much higher spreads and sharply creased uncertainty and sharply reduced the wealth reduced flows. By contrast, investment-grade bor- of investors in high-risk assets, and thus reduced rowers enjoyed improved terms from the decline in their appetite for risk. Private flows failed to com- interest rates.' The decline in access to capital mar- pensate for adverse cyclical conditions; the fall in kets exacerbated the impact of the global growth developing countries' market access exacerbated slowdown on developing countries. This experi- the impact on growth of reduced demand for their ence contrasts sharply with the early 1990s, when exports. Tabled2.1 Net long-term resource flows to developing countries, 1991-2001 (billions of dollars) a prEsiate. inpta 2000et (see tabl 2.)xenaZatr paeh preomian roeenreucndetrnlniane clin inceciW orpi Ban the di mostt deelpigooutr borrwer arespeula theire appetit foBis.Pivtnlosfaldtocm pest3o2des ylca odtos h ali PRIVATE CAPITAL FLOWS TO EMERGING MARKETS Table 2.2 Capital market commitments to developing countries, 1991-2001 (billions of dollars) Note: The data in this table are gross commitments, and thus differ significantly from the data in table 2.1 which are gross disbursements minus amortization. The data on equity placements refer only to initial offerings of equity transactions marketed across borders, and do not include net purchases of securities by foreigners in domestic stock markets (which are included in the line "equity flows" in table 2.1). a. Estimate. Slowdown in world trade partially offset by debt ratios improved in the early 1990s reces- lower interest rates sion-see table 2.3). The growth slowdown in industrial countries re- duced developing countries' export revenues, but The impact of the technology crash the direct impact on borrowing capacity, at least The reevaluation of prospective returns in technol- for investment-grade borrowers, was softened by ogy sectors also had a role in reducing flows to de- the fall in interest rates. The drop in world trade veloping countries. By the middle of 2000, markets growth coupled with the continued fall in com- perceived that the investment boom in telecommu- modity prices (see chapter 1) reduced developing nications had created massive overcapacity, and countries' export revenues by almost 1 percent in that many of the newly formed Internet companies dollar terms in 2001.2 The export revenues of the would be unlikely to generate the profits required East Asian and Latin American regions, which ac- to justify the investments made. This reevaluation counted for almost three-fourths of developing of the likely profits from technology investments countries' private-source debt in 2000, fell by 2 led to a general drop in technology stocks, while percent in 2001 (compared with a rise of 20 per- the slowdown depressed equities prices in general. cent in the previous year). This decline would have The technology-heavy Nasdaq index fell 21 per- increased the aggregate debt to exports ratio of the cent in 2001, and an index of global information two regions by 3 percentage points (from 123 to technology and telecommunications stocks (the 126 percent), if there had been no net borrowing in Morgan Stanley Global Industry Indices) fell 28 2001. However, slower growth in industrial coun- percent. By contrast, the more broad-based Dow tries also resulted in a significant fall in short-term Jones industrial index fell 7 percent. interest rates, because the demand for funds de- just as the boom in global stock markets in clined and central banks in the United States and 1995-99 encouraged greater equity placements Europe cut policy rates. The fall in interest rates re- from developing countries, it appears that the sulted in improved terms on new lending for many sharp fall in stock markets is now associated with developing countries. For example, in 2001 the in- a decline in placements. Developing-country aver- terest rate on new bond issues by investment-grade age stock market prices, after falling by 33 percent sovereign borrowers among developing countries in 2000, dropped another 5 percent in 2001. The fell by 130 basis points, compared with the previ- ous year. At unchanged debt levels, the two regions Table 2.3 Debt ratios during recessions, would have seen a decline in the ratio of interest East Asia and Latin America payments to exports from 7.6 percent in 2000 to (percent) 7 percent in 2001.3 Thus, the direct impact of the growth slowdown on borrowing capacity was rela- tively modest, particularly in comparison with the sharp deterioration in debt ratios during the reces- sion of the mid-1970s and early 1980s (although Source: Wortld Bank. 33 GLOBAL DEVELOPMENT FINANCE Figure 2.1 Performance of developing-country stock markets by sector Percent change from previous year -10 6Mi A i a -20ME EEK_ -30 - g 2000 2001 -40 ,- All sectors Financial Technology and Utilities Energy and Other telecommunications industry services Note: Data for 2001 are until July. Source: Bloomberg; Morgan Stanley Capital International; and World Bank. technology sector, which accounts for about one- Table 2.4 International equity placement and third of Morgan Stanley's emerging stock market performance of stock markets index, suffered the largest price declines (figure 2.1). Capital market flows were pro-cyclical in re- sponse to booms and busts in equities prices. Inter- (billions ofdollars) national equity placements by developing countries China fell by 72 percent in 2001, to only $10 billion. All Other countries developing-country regions experienced a decline (percent change over p-r, . .: in equity placements, but China alone accounted All developing countries for some three-fourths of the total fall (table 2.4). 44 China China had received over 60 percent of develop- Nasdaq i 3 1 ing countries' equity placements in 2000, largely in Soarce: 2.4Inernaptoal DATA;ityacement & and sI technology sectors. A retreat from speculative-grade investments- The growth slowdown and collapse of technology highly leveraged corporations, a fall in investors' prices also reduced capital market flows by reduc- appetite for risk, and increased uncertainty about ing the demand for speculative assets in general. economic prospects: Spreads on global high-yield debt in 2001 were 203 basis points higher than the average in 2000, 1. Speculative-grade corporations tend to be and shot up by about 400 basis points in the more highly leveraged, and thus more likely to aftermath of the September 11th terrorist attacks default during recessions (they have less access (figure 2.3 ).4 Since about two-thirds of develop- to loans to support operations, but need to al- ing-country sovereign borrowers (and a much locate a growing share of declining revenues to larger share of private borrowers) are speculative meet fixed debt service payments). The global grade, this implied a general decline in flows to de- default rate of corporations with speculative- veloping countries. The retreat from speculative- grade credit ratings reached 9.8 percent in grade assets reflected an increase in the riskiness of 2001, the highest level since 1992 (Moody's 34 PRIVATE CAPITAL FLOWS TO EMERGING MARKETS Figure 2.2 Bank lending standards and bank credit to developing countries, 1990-2001 Billions of dollars Percent 200 60 Bank lending to developing countries 150 40 100 20 50 0 Tightening of bank lending standards 0 1 l l 1 l -20 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Note: Data for 2001 are until the second quarter. Tightening of lending materials refers to the share of banks in the United States that reported a tightening of their standards and terms on commercial and industrial loans over the past three months, as reported in the quarterly survey of the U.S. Federal Reserve Board. Source: Capital DATA Loanware; U.S. Federal Reserve Board. Investor Service). Therefore, when growth 2001 may have increased the range of out- slows banks tend to tighten their credit stan- comes that investors feel they should consider. dards to restrict loans to speculative-grade Increased uncertainty can cause risk-averse in- borrowers, both in reaction to the deteriora- vestors to reduce the share of high-risk assets tion in the banks' portfolios while default rates in their portfolios. increase and in anticipation of the impact of recession on highly leveraged corporations. For all of these reasons, the past year has seen a The percentage of U.S. banks tightening their widespread retreat from speculative-grade borrow- lending conditions exceeded that of the reces- ers. Because their share in total developing-country sion of the early 1990s (figure 2.2), and the borrowers is three times that of industrial-country volume of global cross-border bank lending borrowers, the decline in loan commitments to de- commitments fell by 13 percent in 2001. While veloping countries was relatively large. Bank lend- bank credit contracted in all categories of ing to developing countries dropped to $93 billion credit risk, the most severe pull back was from in 2001, or less than 75 percent of the 2000 fig- the high-risk borrowers.5 ure-the second-lowest annual level since 1994. 2. Reduced demand for speculative-grade assets The decline in commitments was biased against also may have reflected investors' reduced new entrants to the market: the share of hank appetite for risk after their wealth declined credit attributed to refinancing rose from 26 per- (see box 2.1), exacerbated by the events of cent in 2000 to 34 percent in 2001. The cost of September 11. Investors in high-risk assets refinancing for investment-grade borrowers rose have experienced a sharp fall in wealth: since minimally. By contrast, the cost of refinancing for its peak in early 2000, the market capitaliza- borrowers rated below-investment-grade rose tion of the Nasdaq stock index has fallen by sharply and loan maturities fell. Unlike the case for over $3 trillion. bonds (see next paragraph), the decline in bank 3. Reduced demand for speculative assets may lending affected most developing countries. Even also reflect increased uncertainty about eco- excluding Argentina and Turkey, which are suffer- nomic prospects. The collapse of technology ing severe domestic crises, and Brazil, which had stocks and the industrial countries' plunge from been greatly affected by developments in Argentina 3.4 percent growth in 2000 to 1 percent in during most of 2001, the decline in bank lending 35 GLOBAL DEVELOPMENT FINANCE Box 2.1 Evide f hangs in the a ite and capital pkt lws C hanges in investors' ppetesi fr isk are often assoei- and measure e rdi rif , ilYdin preads bYt drn ated with changes ii dceoping-Zountry acess o ris - re an hi -sk asses diffetence between the private capital flows. Theipp(ite for risk under ondi- iskier sma1e- stek and .h S P , rei n - tions of uncertainty in par ids on the level o wealth chang velt i a d shag a h ice optos sels- (Gusl o1 99 hh%% . thi1 in il r cia moel.mn -Wir -h - -f--is48rte4c i- - i agers' heh.i% i-r). 1. cu%e eiLh 1llar of inCome b Ceomes- J7.!_kUbs1 ;e iS I liakh tr lli4--1 9' Citl.ii more important ri %,ealth.deIl nq is.Iv-rse nvistors the appetite fIr r 7ik during periods when i is l IilI t.by: are leyo willine to undertaki- acr r:sks .u lower kIfs of such delines ',curred. for -xample the RusIan Jeala:r wea th. Clark 1 1991 tinds rhit tne reason for capia rie n of August 19. n addit.unon the J.P. gorgan indeX flows trom rich ti poor o unmls,'s rhat the highei wethh recisters subsantial r e n rik aversion durn Julc s,iuia I of rich coururies in, cqor%,ma1k theniore illog to when the A gentlne e-risis derurarid and thn rnwedi- undert ike risk% n%estnent. T e C. ;nerse1 o1 th et1' 7 t 7 was import.nt during the RuM iEh de. iluation of 98.Fn when hu losses siffered h inveors in Rhu4ian Fcuri- 0 of ties reduccd the appete lor rUikumar and Persand lili... . :i * r ' 2001. hisriturc oI Intcrn.at.-n.il F1irnc 1- rd bpMralh n flows to d eloping nri r, clip i p n 'enerld. l p Howe% er, apart fro1 CriNei tIiire ]earl% related to change, in in.estor.' %calth. it i' Jitti,7ilt t 1 eilinine 1ath1 r an p1 M sn i whether hanges in the appeEt trlirik ha..e hid an im- 111. 3nir3 r (L iingi1sN ii e portant impaict -:n nirke acczss ThFIppttt Io ri- Nl ~i-ird FT i~cs~i h nledua.l~ foJui extremek diftticll to nilwurt.'i theSpm \r 1 h - rr r Chase it-turititr. I P. Nlor.gin, in tdi C JuI-e-T pHowetha, theen in eni fas d ae[The-oie nirn- Boston, dr pro% ide statictical aVr*hoa t and har.m audeasug 1 q . the in atW Aha k u vestors ippetite tor risk These In ndioZ gi: 71113ue in len l ilS Ir F Ceti M a lo deS T - measures. tt niarkeaang in srets b aq 1somb.iin f 0thi tween recenik is-tued and htt tr.dmnnt u m.ts.rM hpp.urLi- ieshr- was about 25 percent. Bank lending is less tolerant The stability in bond volume in 2001 was sup- of changes in risk than are bond markets, reflect- ported by increased borrowing by higher quality ing banks' high leverage and the greater concentra- borrowers (rated either investment grade or just tion of their loan portfolio compared to investors below), including China, Hungary, Malaysia, Mex- in bonds. ico, and Poland, as well as smaller borrowers, such as Colombia, Latvia, Panama, and Uruguay. -might benefit developing-country bonds Perhaps surprisingly, the reduced demand for high- Reduced capital flows partially reflect a fall risk assets may have helped support developing in demand countries' bond issues, which remained stable in Declines in the demand for capital played a mod- 2001, at $68 billion. Developing-country bond is- est role in determining the volume of capital mar- suers have higher credit ratings, on average, than ket commitments in 2001. Most developing coun- developing-country bank borrowers. Thus bond is- tries' access to foreign capital is constrained by the sues were less affected by increased uncertainty and willingness of foreign investors and lenders to sup- reduced appetite for risk. Moreover, the decline in ply funds. However, a few countries could borrow interest rates and a slight reduction in investment- more even at the current interest rate, but do not grade spreads implied a significant reduction in in- because their demand for capital is low. For exam- terest rates for investment-grade borrowers, thus ple, during 1998-99 the demand for funds from encouraging more of them to come to the market, the East Asian crisis countries collapsed with the 36 PRIVATE CAPITAL FLOWS TO EMERGING MARKETS Figure 2.3 Corporate default rate and risk premiums, 1990-2001 Risk premium (basis point) 1,555 1,485 Default rate (percent) 1,200 - Developing-country Indusial-country 14 risk premium high-risk premium -12 1,000 -10 800 1 8 600 6 400 4 200 -- 2 Industrial-country high-risk corporate default rate 0 l l l l1I 0 Dec Dec. Dec Dec Dec Dec Dec. Dec Dec. Dec Dec Dec 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Source: Moody's Investor Service; Merrill Lynch. 30 percent fall in investment, and they ran large $16 billion per month during the first half of 2001 current account surpluses. Capital market com- (compared with $19 billion per month in 2000), mitments to the crisis countries fell to about $30 and then dropped to only $9 billion per month fol- billion per year during this period, compared with lowing the September 11 terrorist attacks (table $74 billion in 1997. It appears that demand also 2.5). Spreads on developing countries shot up to remained low in the five crisis countries in 2001, 924 basis points in the aftermath of the attacks, since investment fell slightly and the government compared with 716 basis points in the first half of deficit improved by almost 1 percent of GDP. Cap- 2001, although the rise in spreads (excluding Ar- ital market commitments fell to $34 billion. Thus gentina and Turkey, the two major countries most low demand from the crisis countries most likely affected by domestic economic crises) was modest. reduced the level of capital market commitments Commitments recovered during the last quarter, but compared with what would have happened with a remained well below the 2000 level. The average robust recovery. Nevertheless, there was no repeat spread excluding Argentina and Turkey fell to 400 of the experience of the 1998-99 period, when the basis points (100 basis points below the average of drop in capital market commitments in the crisis the previous year) while interest rates fell and opti- countries had a noticeable impact on the total mism about an early recovery increased. for developing countries. A few of the richer oil- exporting developing countries also reduced their capital market commitments in 2001, presumably choosing to increase saving in response to contin- Trends in FD1 ued high oil prices. et FDI to developing countries is estimated Nat $168 billion in 2001, almost unchanged Capital market commitments declined until from the previous year, and just 8 percent below late in the year the peak reached in 1999. The stability of FDI The overall decline in capital market commitments flows was achieved in the face of a significant fall accelerated in 2001 while the global slowdown in global FDI flows. Changes in FDI flows to de- deepened. Capital market commitments fell to about veloping countries in 2001 were driven more by 37 GLOBAL DEVELOPMENT FINANCE Table 2.5 Capital market commitments and spreads for developing countries 20 2001 January-June July-August September-October November-December (monthly average, billions of dollars) Capital market commitments 15.8 12.7 9.3 16.6 Bonds 6.9 4.1 2.5 6.8 Banks 7.7 7.9 6.7 9.1 Equity 1.2 0.6 0.2 0.7 (basis points) Developing-country spreads 716 844 924 865 without Argentina and Turkey 5 440 416 447 404 Note: Developing-country spreads refer to J. P. Morgan Chase's Emerging Market Bond Index Global, which uses country weights based on market capitalization of outstanding debt. Source: Dealogic; J. P. Morgan Chase; World Bank staff calculations. domestic economic developments (for example de- decline in profits and tightening of bank credit cisions over privatization transactions and policy standards. For example, FDl outflows from the improvements) in a few of the large FDI recipients United States dropped from $19 billion in 1980 to than by changes in the global economy. only $1 billion during the 1982 recession year, and then recovered to $13 billion in 1984. Global FDI in downturn- Preliminary estimates from the United Nations Con- -but developing countries were less affected ference on Trade and Development (UNCTAD) in- The past years have seen considerable stability in dicate that global FDI flows fell massively in 2001, FDl flows to developing countries, although their to $760 billion from about $1.3 trillion in the pre- share of global FDl flows was cut in half in the vious year. Global mergers and acquisitions (M&A) wake of the Asian crisis. Essentially, the trends ob- activity show a 45 percent drop in 2001. Slow served since FDl flows plateaued in the late 1990s growth or recession is often associated with a de- have remained constant. Developing countries' cline in FDI outflows (paralleling the decline in share of global FD1 flows turned up with the drop domestic investment) since multinational corpora- in global flows, but remained well below the 36 tions face stringent financing constraints with the percent level reached in 1997 (see figure 2.4). FD1 Figure 2.4 FDd and M&A in developing countries, 1991-2001 FDI M&A Billions of dollars Percent Billions of dollars Percent 200 40 90 - 20 80 18 150 30 70 --16 60 14 12 50 100 20 10 30 - 50 60 2 102 0 0 0 L0 1991 1992 1993 1994 1995 19961997 1998 1999 2000 2001 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 - Developing countries - Share of total (right axis)f Source: World Bank, Global Development Finance: Country Tables and sources cited therein, various years; UNCTAD, World Investment Report 2001 World Bank staff estimates for 2001. 38 P R I VA TE CAPI TAL FLO W S TO E MER G IN G M A R K E T S flows continue to decline relative to developing rather than the impact of the economic slowdown countries' GDP, down to 2.3 percent in 2001 from or other global factors. Eight out of the top ten re- 3 percent in 1998. FDI flows remain highly con- cipients saw changes (either increases or decreases) centrated: as has been true for the past few years, in FDI flows of 20 percent or more from the previ- the top 10 recipients of FDI received over 70 per- ous year. These changes were driven largely by in- cent of total FDI to developing countries (box 2.2). ternal factors, often privatization, private sector The stability of FDI flows in 2001 largely re- M&A transactions, or general domestic economic flects offsetting changes in a few large countries conditions. In Mexico the sale of Banamex-Accival Box 2.2 The concentration ofED1flws M .t V-DI tIon iI hai e ren.iined conricentraed in Narrec t desiredbv toreig m est~rs. tek% de' eloping countries through-ut tc 1990s, Th uirgin F D I hn kth the itrdcon ut . h. n the share of the top Il has never lallen below 6-1 i¶ 7 m©tI prhap th rrci spectacular eami-ple per,enr.~ Arket sue appears in be a maior e\plananiil ot t,iio nir FDIl oks to;NIMexico'kkere Conricentra': ot the top 110 dr . pmgc:untri Norrh American are ilsi mngthe top 1O countries in terms ·Fr cremenr:,FD! ioha iricreased to eontinies t CDP, but marker size is not the oni iactor The :3nerage ik irb ssron . 9coii 'n.p.-ograms-that i iraliblize rh. fL& ratio ot FDI t GDP G n the top 1< recipients is ulmos-i . vr pexan FD he Repulbi,f Korea full percentage point bigher than in de\ cluping countries - rose:m > äl reh ¯ t i n crir as a group i figure 2.5. \hile Brazil, China, and ASu o bg r e.irules alon account for ibout halt of de' eloping countries' torn nvestmentmee\crid Bank F0i0i la,1 thit make up nl a hrtle. more than oine-thrd ot J lgrt d d sns on pri ing countrie GDP. \While FDI flo,is to India-the out , ~zah i 0 F rr largest de eloping countr\-hac ncreased oer the - mdfnretmn"e 1 'M%s, the Lountrk reminms 1-th on the lihr of de elopm atznn cu \ ór Ir countrv FDI recipients. i cenrK: tnd , -ith m ilir i th t .- F Di is ilso concentrated im relation to other indica ED e oun t e s dmr- c tor,, ot conmic eis itn Of the 16 larger F iu tir I percelit.]llmtmeint af ýc FDI ~recipielC c 9s iM ent l5:1 are als) the de eloping countries with tiie largest ex- ore unurpi in bie"aus tor- ports. UNCTA D 21 l1 de' eloped a more comprehen1 n itcoin estorsre blÙIo re'pond to the indc\ that measures FD[ inflrn s relant e to econonnc s iz am ctorscmarket tize annrestimient cmat. Nore- represented hs an unn eighte i ,rage 11 tchreje o rr [h LD. er, E .! DirinwC t.nd rtocr?. 4-rJij mJ...fl A uiintrn \ harr im t orld FDI intlnt s to its shreimn \\·crld Bank 20e1 rhapr .;Bosu orth and 6e|Im 1 9 k orld GDP. empl. ment, and ecports. By ti iure, lihe con.errarwnof capita RmarkeH1 fownsã I t FDI is m ldl concenrrared; nl 30 our ut 1d CdLping h1gher than F, tbe rop l rec'ents accounej r counrries had shares of FDI that equaled or'é ceded thxr cent of total .low s. A.cess to eapita marketlowk cdends i k(rage shares of-t %% orld GDP, enploy ment d Npo- 011 tresence of relbti .h wel-devdoped:hnancial när Onl, half the t,p 11. FDi recipieints received mor FP! kes Fsnun.and Fernandez-Aris 2'105.ldut Mle than expected, based on fhen shares uClot g.ji econo,ic the poorest devdopig .ount.Ies reiini iinim ic ti tt The concientrarion oi F DI flo.:do. not mean f FO)1, the. receive almost no porfolio lo' s s hipter that FDI onhs benctits the largt r counrri<1 s io the 0 3. A 1cncemrarino F! lows is oten.observ wi i des eloping countries ith the higltrt n Di tt SP> coinr.jes as well. For exampie, nearl> 90-prnthiii are relai,ch snul-scale econni.' Fta regions, almntifllFiti.tln - FDI to some f Ilie larger recipien hia .ee boosted Mexio wete absor.bed in central stares and,thiie r¶dr by good polbcies. The largest FD! r.pichts ha'euaner- the United States (iJNCTlD 2u01 uhile mr India te op age \Xorld Bank polies rating ofi 4 , compark d ith 3.3 five recipiem state lMaharashta. X mil rna ik tor o rher de' eloping c.unrries Prhaps nikn.rnprr Andhra [h nd F9elhl:KX r tor dJ,rccrmining F 1)1 1iin s. hi:.u e\er, is [be..aneii'oi ia id apo, s~nt t:Aä~4?~3ttpl ce,. Counries that have undergone in improkeinent m ?4apfar r I a 'rlm nt ohdsrtlti ntni the miivesment .lmate mat see'i ige inmlo'. FDI unil FDtilnesiidu.iDolar.!a and \ngist.e Y01 1. 39 GLOBAL DEVELOPMENT FINANCE of Mexico's major airlines), whereas some foreign Figure 2.5 FDI as ratio to GDP, 1991-2001 investors have withdrawn large-scale offers to Percentacquire stakes in private companies (including two PercentBrazilian telecommunications companies). FDI 5 flows to Eastern Europe remained stable; while Top 10 large-scale privatization programs in banking and 4 telecommunications neared completion, the region received an increase in greenfield investment. Net 5 FDI flows to Middle East and North Africa re- All developing mained at about the level of the past few years. 2 The De Beers sale boosted flows to Sub-Saharan Africa. FDI to East Asia and Pacific declined de- spite higher FDI to China, because of slow growth - All developing countries - Top 10 recipients All eveopig Contres op 1 reipintsin several regional economies, low demand for 0 1 1 1 1 1 funds in the high-tech industries, and reduced 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 M&A transactions in the East Asian crisis coun- Source: World Bank, Global Development Finance: Country Tables and sources tries (figure 2.6). cited therein, various years; World Bank, World Development Indicators, various years; World Bank staff estimates for 2001. source of FD1 While the data are incomplete, it appears that de- veloping countries have become a major source of financial group to Citigroup for $12.5 billion FDI flows to other developing countries. Out of boosted FDI flows, and in South Africa, a foreign $185 billion FDI inflows to developing countries in firm took over De Beers mining company by ac- 1999, only $72 billion are identified by the Organ- quiring shares worth $20 billion. In Poland lower isation for Economic Co-operation and Develop- FDI flows signaled the completion of major privati- ment (OECD) as coming from the industrial coun- zation transactions. In other countries changes in tries. Developing countries also receive about $40 FDI flows reflected changes in the overall economic billion in FDI flows from other high-income coun- environment rather than the impact of a few trans- tries.9 If these statistics are accurate, the remainder actions. Examples include Brazil, where economic of developing countries' FDI inflows (about one- uncertainty restrained greenfield FDI; Argentina, third or $70 billion) would have to be from other where lower FDI flows reflected a slowdown in pri- developing countries (figure 2.7). South-South FD1 vate sector M&A transactions with the increasing may also have contributed to the resiliency of FDI economic difficulties; Korea, where the process of flows during the financial crisis. By these calcula- corporate and financial restructuring has slowed;8 tions, South-South FDI flows continued to rise in and China, where FDI boomed with the anticipa- 1998 and 1999 despite the financial crises, during tion of accession to the World Trade Organization. which total FDI flows from high-income OECD The extent to which FDI inflows in China represent countries declined. additional resources to the country remains open South-South FDI has increased at the same time to question, because a significant portion of regis- as South-South trade was rising (intra-developing tered FDI to China may have originated in the countries imports rose from 30 percent of their total country (box 2.3). imports in 1990 to 36 percent in 1999). Thus, the These major changes largely determined the production and ownership structures of developing regional trends. FDI continued to fall in Latin countries seem to have become more integrated America, the largest recipient region, because through FDI, not only with the industrial countries, cross-border M&A activity in the region dropped but also with other developing countries. In addi- by around 5 percent. Several privatization plans tion, major developing-country exporters who face have been postponed or delayed (examples include quota restrictions in industrial countries may have Copel, Brazil's electricity generation and transmis- invested abroad in order to export from countries sion company, and Cintra, the holding company that are less affected by such trade barriers. 40 P R I V A T E CAPITAL FLOWS TO EMERG ING MARK E T S Box 2.3 Round-tripping of capital -flows betveen China and Hong Kong F DI intloku to Chiiia urgcd in the 19n of narket retormns and the intrn-dii - nier. Ilie FDI inflo s trum Hong Kong land the \ irigm tion of incernukes tor FDI, ncldti,iig cunCessions on rax,. Islndl appear ro be highly c'rrelated wrh >tlows Iron leasing,o land and property, g,' erninnt guarantees tor Chi in the fori ul -ther isestmem .n (rmo%i investmncis, and spccul arrangenirn r rep rnn runri .i bank lcpusirsl hlvd abroad bi Chiese residenr,, and er- and repitriitin (it treign esch.inge. Prrerernces for tur- ror, :md omssionz n Chna's halance of paymentirs (se. tig eign capr.l are helie% ed tc, hase encouraged Chncese in- urr hel'u i. H,iig kong, in irs turn, reports large aron i vestors :o moi e mona offshore and ihen bring it back to of E! .nflows troni mm uiuland China, and from otishrerv-f- China disguscd ae; foreign imse;nen[ iSicular 19R. An-¯ iciil cénrers such as Bermuda aid rhe \irgin Islands. other rn sn,rion tor -round-trippmg,' or irecyclmg, the concern rhär the go. ernmemn mas inpttse exchange le- strictiors on residents, as curreJ ir [Liv 193 ' (Adanism 1993; Guinter 1496. Sorme carh studie, esimaited flut C roundt-IppIng accounrted ftor nearIv a quarier ot foreigi' inflows tt China iii 192 1 ardv 199, p. lir-; Harrold - and Lai; 19 , 3. . 2-i. The extent of recs cling rna have H1.xi K-io iin, i i) 1 3. increased in re,tn i ears i bu tigire. ri. 13 tJi U.K i kl . Thrrightout the 1 9w0,, FDI intitst4 tto China orig; p- 4 lö l nated mttstls itide the imdusral ounträe., noribly lirt ni 1 - Hong Kong K¯hiu. For c\aiple, FDI intliews trirm Hong -,i e b1 : Kong cotnrturted nearlv half of totr il FDI flows to China ä) räm'.r S pk'. -irir H t,> 1 3 in 1996 Hung Kong's share has declined ncc 1997 , to, below 41 purceni by 2010l bee rahle belo% . This deline erI. has been oftser be, i comparable nert ie mn FD! inflns reported from the \irein Islands. husever, wlä.li ugges[s hr - Round-lripping al capital flows: China and Hong Kong (China), 1986-1999 FDI inflows from Hong Kong (China) and Macao j -rrir1 1r o 1.rr- rr0-15 1986 1987 1988 1969 1 19% 1991 1992 1993 1914 1995 1996 199" 1998 1999 i , \orld K icn l. ,jet LIn tät4 41 GLOBAL DEVELOPMENT FINANCE Figure 2.6 Regional trends of FDI flows, 1991-2001 Billions of dollars 100 90 80 -4-- East Asia 70 _Lf - A and Pacific -0- Europe and 60 Central Asia 50- - Latin America 40 _and the Caribbean 304... Middle East and North Africa 20 ri- South Asia 10 **-. *--w-- Sub-Saharan Africa 0 10 I I 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Source: World Bank, Global Development Finance: Country Tables and sources cited therein, various years; World Bank staff estimates for 2001. Figure 2.7 North-South and South-South FDI, 1991-99 FDI flows to developing countries North-South and South-South FDI shares Billions of dollars Percent 120 100 North-South --North-South/total FDI 100 - South-South 80 - South-South/total FDl 80 60 60 4040 20 20 0 0 -20 -20, 1991 1992 1993 1994 1995 1996 1997 1998 1999 1991 1992 1993 1994 1995 1996 1997 1998 1999 Source: World Bank staff estimates. The data given above calculate South-South many of which have capital controls, exchange FDI by comparing developing countries' FDI in- controls, and high taxes on investment incomes, flows with recorded outflows from other regions. combined with weak accounting rules and tax ad- This is probably more reliable than basing the cal- ministration. Nevertheless, the trend of increasing culation on identified outflows from developing outflows of FDI from developing countries is also countries. The problem of under-reporting FDI evident from the data on identified outflows re- outflows is acute tn the developing countries, ported in the country pages of the IMF balance of 42 PRIVATE CAPITAL FLOWS TO EMERGING MARKETS payments statistics. However, reported outflows turn after 1998." As a result, public sector debt re- from developing countries, which reached only mained high (at 50 percent of GDP in mid-2001), $12 billion by 1998, are much smaller than the es- and maturities shortened. timate given above, due to under-reporting of out- The steady appreciation of the dollar in the flows by source countries. second half of the 1990s and the sharp Brazilian devaluation led to a 15 percent real exchange rate appreciation berween January 1997 and mid-2001, further constraining growth. Most importantly of Emerging market financial crises all, deflation persisted throughout the economy in 2001 (consumer prices have fallen by a cumulative 3 per- T he past year has seen a continuation of the se- cent over the past three years), and the real economy vere economic crises of the 1990s that afflicted remained stuck in recession, leading to a further rise major middle-income emerging markets (Mexico in in an already intolerably high unemployment rate, 1994-95, East Asia in 1997-98, the Russian Feder- With nominal incomes across the economy falling ation in 1998, and Brazil in 1998-99). The causes sharply during 2001, there was little realistic chance of each crisis differed in important respects, but in for the authorities to meet the tax revenue projec- all of them shortcomings in external financial man- tions that were the backbone to a planned "zero agement and defects in corporate and financial sec- deficit" budget strategy. Market awareness of the tor governance played an important role. The past sizeable dollar liabilities of both the public and pri- year's problems in Argentina and Turkey shared vate sectors completed a picture that made creditors many features with these earlier crises. leery of maintaining, let alone adding to, exposures A critical difference, however, is that conta- as the end of the year approached. gion effects to other emerging markets, and other Public disturbances-in part a reaction to lim- debt markets, have been limited (box 2.4). This is its imposed on cash withdrawals from the banks- especially noteworthy since Argentina's crisis de- led to the resignation of the Argentine president in veloped into a full-blown sovereign default. The December 2001. Soon after, the government for- only recent instance of such an extreme outcome mally defaulted on its debts and the currency was by a major debtor was the Russian Federation in devalued. A floating exchange rate system was in- August 1998; that situation produced severe dislo- troduced in mid-February. It remains to be seen cation across global financial markets. who will bear the considerable losses from the de- The crisis in Argentina has its roots in the valuation, but given all these dislocations, a phase buildup of vulnerabilities after the highly success- of renewed output declines and rising unemploy- ful exchange rate-based stabilization of the early ment seems inevitable. The only issue now is how 1990s. After a long history of inflation (including long this situation will persist. a period of hyperinflation) and failed efforts to Turkey also faced a severe crisis in 2001, stabilize, the adoption of a dollar-based currency which was marked by efforts to control a large board in 1991 stopped the country's inflation in its public sector deficit (12 percent of gross national tracks.10 The country experienced a post-stabiliza- product [GNP] in 2000), high levels of public sec- tion boom on the order of 7 percent growth in tor debt (in the range of 90 percent of GNP by GDP, while the reduction in interest rates toward end-2001), and difficulties in rolling over short- world levels stimulated domestic demand. term debt (100 percent of reserves). Adoption of a However, substantial vulnerabilities remained, crawling peg in 1999 was- aimed at reducing high and were increasingly exposed during the second levels of inflation. Fixing the exchange rate en- half of the 1990s. Despite strong export growth, couraged large capital inflows with a substantial foreign exchange revenues were insufficient to fi- buildup of foreign exchange liabilities of the bank- nance buoyant import demands, rendering the ing system. In February 2001, the government was country dependent on capital inflows. Fiscal policy compelled to abandon the crawling peg, which led was not only too loose on average, but was also to a 26 percent real devaluation (year-on-year) unhelpfully procyclical-too expansionary in the by the end of 2001 and large losses in the bank- recovery phase of 1996-97, leaving the authorities ing sector that the government is now cleaning with no scope but to tighten policy into the down- up. There are a number of reasons, however, why 43 GLOBAL DEVELOPM ENT F I.N A N C E Box 2.4 Finanicial niarket contagion f -i Argentine crisis here is little evidence that xnvestrs ha ererd Car e d r catd cidsa d -od 0 from most other emerging n arkers becau,e of the e ri- - Eg -6: sis in Argentina. The.correlationbetsken secondars niar. b ;Y;Äm kets bond spreads between Argetin i nd 15 emerging ? n iu; markets rose from 0.27 in the inths bernre the e\icer- 4 bation of Argentina's. difficulties i1Ocr i lr 20 0.1 - J . If-H * from October 2000 to August21.2 How% c er, this period coincided with the global growh slohd n that was associated with a general rise-in spread, imd i tile F ir . n d vri, 1. h u r . volatility of spreads (and measVie' ,-rrclaipns ind to C A ghis a n >.: rise with increases in volatilityX soi is diticuli toe isel.ite ä the impact of the two crises. Brazil dovs a ppca r th' been affected by the crisis in its neighbor to the ;outh, 7 perhaps because they compete in the sane markeis.' dcVluäin Ii rssh hkh w r consid The correlation berween Braziliah'a-id ·\rgentine spre id, er ible-,ur- 8 th b le ms increased from 0.6 in mid-2000 to bKrv eein 0.8 anJ 0.9 n Argenria rör sne in- ust in .Jtor, måv il- in each of the three-month periodsfrom Crober 2000 to read hai e riken Dhire ,r lpMi t nr' . Lessjrin iib August 2001. However;late'in the ear mirker nt iitn sorbig i hr lse ,nArgn_iiie bonds. Nforover many toward Brazil improved, and spreads närrou ed despite the im estors r . l ks~, ioeragted this timv irnunUfflrjn-däring increasing probleis in Argentina,,, - the A,iin ri,', i partIArulrl after tih deb&ae thit hbly Looking at specific crisis episoces jOctuber 2010, le.erag2ed spLculators u red ut h the-Rti 5iaii delj lua. March/April 2001, July 200,1;andiccimblr 20ii1, we noni., hich me.ins ihat theic is a rtdiccd räc.d t.) liquidate( can see some rise in the pieäd,s' on,ofher emerging market acros-rhe-boar i[ rncmer nargiri ealls. At the same tume, bonds. However, the rise in spreads during the ,risps eri. dc%pcpimg oiirr, ire less ulicralu rh.ån they kere a ods varied, and spreads tended to return ro frncr Ieels tek sears ago. Currenty, verv few måior ernérgng markets relatively quickly.The,index,of erne-rgog iwirker spreads ha e pegged e.change rares, k hich proved'yu be particu- was at almost the same-leve1.in;dcirber 2001 1s in lariv -. uliierable to contagion froin the collapse of rther October 2000. Overall; spreads in nierging mrkets v.- pegged exchange rjtes. I.evels of reser es have risen uille cluding the two crisis countries app(ar to h.i\e ben alurk shorr-tcrm debr leiels have tallen, improkrieza ke>indi- affected by the crisis in Argentina,nd we're srahle ulii citlr rf , ulncibihty. Se%eral of the Asian countrics are the September 11 terrorist attåsc pre.senl rurning curreni ä,count ->urpluse ind are There are various'reasons Whyjhe Argentine crisis less dependeni on internatinnil :,r l rnakts.- Fiall, has generated such limited contagån uftets so tjr, in l i mternanonal imrt res er1rt.S.ed c '.ern l,finaeing marked contrast to the East Asan c s, and ihe Rssinii press.ircs on heak ti idebrå enign imfrkets. Turkey's difficulties have been less severe than • Turkey's debt is higher than Argentina's (rela- Argentina's: tive to output), but a greater share is owed to domestic residents, which helped facilitate ef- • Despite the crisis, Turkey is making significant forts at restructuring. progress in improving thé fiscal accounts: the • A larger and more diversified export sector primary balance of the consolidated public means that exchange rate depreciation can have sector shifted from a deficit equivalent to 2 a greater and more rapid impact on production. percent of GNP in 1999 to (an estimated) sur- • Turkey's strong ties to Europe and its impor- plus of 5.7 percent of GNP in 2001. tance as a front-line state following the Sep- • The exchange rate regime was less rigid and tember 11 attacks have helped to facilitate thus provided for an easier (albeit still very substantial financial support. However, the at- messy) exit mechanism. tacks also severely damaged Turkey's foreign 44 PRIVATE CAPITAL FLOWS TO EMERGING MARKETS exchange receipts, due to the drop in revenues tine authorities structured their economic from tourism and slower export growth. A system around the inviolability of the one- new IMF standby arrangement to help Turkey for-one exchange rate peg against the dollar. absorb this additional external shock and sus- However, this structure encouraged investors tain its reform program is expected to be in to incur mounting dollar liabilities, in the be- place in February 2002. lief that the government would maintain the peg. The size of dollar-denominated debt then Lessons of the turmoil in Argentina greatly increased the economic costs when the The situation in Argentina is difficult, and the role peso was devalued. of clear-sighted economic policy is critical. The The success of multilateral rescue packages de- challenge for the Argentine authorities now is to pends critically on strong adjustment by recip- adopt appropriate measures to allow the economy ient countries. Crises can be successfully re- to take advantage of the newfound flexibility of a solved only when policy implementation is floating exchange rate, while also addressing some strong; government commitment to taking dif- the key structural problems that have been ex- ficult adjustment measures is critical. Multilat- posed and worsened by recent developments. It eral financing is designed to support, not sub- is worth noting that-in the cases of Mexico in stitute for, adjustment. The size of potential early 1995, Thailand and Korea in the winter of outflows dwarfs the resources available to the 1997-98, the Russian Federation in the fall of multilaterals. Moreover, greatly increasing the 1998, and Brazil in early 1999-the early stages in size of rescue packages could encourage exces- the move to a free float were very difficult and it sive risk taking by private investors, although took time for signs of successful stabilization to be so far the evidence that rescue packages have visible. The Argentine crisis is especially complex, generally contributed to risk taking is incon- since it combines large private sector foreign ex- clusive (box 2.S). change exposure and public sector default. There is more work to be done on private sec- It is not too early to draw important lessons tor involvement in crisis prevention and resolu- from the developments in Argentina. Most of tion. Recent experience has underscored the these lessons reinforce those that became evident importance of clearer definition of the limits during the East Asian and Russian crises of on official resources and of the rules and re- 1997-98. Five stand out: sponsibilities of the official sector, debtor coun- tries, and their private creditors. Contingent * Fixed exchange rate regimes are vulnerable to credit lines can provide for new money in case asymmetric shocks. The success of fixed ex- of crisis. But the government's counter-parties change rate regimes requires that the countries can avoid increasing their exposure during a involved are affected similarly by shocks. crisis by selling other holdings of government Events of the past few years, including the de- bonds, thus undermining confidence. In the cline in commodity prices and the Brazilian de- case of Argentina voluntary debt exchanges valuation, required a devaluation in Argentina were relatively easy to organize, but they did to restore external balance. But at the same little to ease the country's financing difficulties. time the dollar was appreciating, responding These challenges point to the need to consider to a very different set of economic factors. The more ambitious proposals for facilitating or- resulting appreciation of the peso depressed derly workouts of problematic private sector output, particularly given rigidity in labor debts, and the recent proposal by the IMF to markets which impeded real wage adjustment. provide for a standstill of debt payments in The resulting recession in turn undermined order to allow time for an orderly restructuring support for the program. will, no doubt, be debated in the year ahead. * There are severe costs associated with hanging Contagion can be contained through prudent on to a pegged, overvalued exchange rate. In external financial management. Most coun- Mexico (December 1994) and Thailand (third tries in Latin America and Asia that are depen- quarter of 1997), failed defenses of currency dent on private capital flows have strength- pegs led to country credit crises. The Argen- ened their ability to withstand shocks through 45 GLOBAL DEVELOPMENT FINANCE Box 2.5 Moral hazard and rescue packages C onsiderable concern -has bee risd that the expecta provisions that would facilitate the private sector absorbing tion of multilateral suPport forcrisis-hit countries losses. A modification to collective actiorlauses could per- may encourage excessive risk taking by.investors in emerg- mit the restructuring of bond instruments by majority vote, ing market debt (Meltzer 2000; Caloniris 2000).'tis of the ceditors rather than unanimity. This would reduce difficult to evaluate what might, have'happened in fhi ab- the ability of small creditors to force repayment of their sence of rescue packages, and so far-the evidence that res- debts as the price of agreement to restructure and greatly cue packages have encouraged excessive risk taking is in- ease the'complexity involved in restructuring bonds. The conclusive. Zhang (1999) finds that spreads on emergiig implications of such modifications to collective action market bonds in the seven quarters following recovery clauses are difficult to determine. Eichengreen and Mody from the Mexican crisis were n"b lower-than-precrisis - (2000) found that collective action clause$ with this provi- levels, after controlling for other determinants of sprads. sion tend to reduce the borrowing costs of-more credit- Lane and Phillips (2000) find no evidence that IMF-related worthy borrowers and raise them for less creditworthy news and announcements of rescue packages had, anim- ones, which would strengthen market discipline. However, mediate impact on spreads. By contrast, Eichengreen and Becker and others (2001) found no evidence that such col- Mody (1998) find that, by 1996', spreads on emerging lective action clauses increase yields for either higher- or market bonds had fallen to levels that failed to adequately, lower-ated issuers. compensate for the risk of 16ndirng and srdd fell further - An6ther, complementary, approach is to provide for of- in 1997. ficially sanctioned standstills that would impose a cooling- Concern that some investors have escaped the losses off period to avoid investor panic (Eichengreen and Mody associated with financial crises las b6oiedconcern over ' 2001); still another approach under some conditions is to- moral hazard. It is difficult to estimate creditor losses from use IMF facilities to continue lending to countries when bor- recent emerging market crises although losses are less - tx-owers,are in arrears (Goldstein 1998; Fischer and Citrin than they would have been in theabsence of official sup---- 2-000). The Bank of Canada and Bank of England (2001) port. International equity investors may have lost $16 bil-' have recmmended adoption of an officially sanctioned lion during the Asian crisis (International Council'of Securi- standstill to provide a "time-out" during which governments ties Agencies 1999) and in"eq:aonal lanks $60 billion can demonstrate their commitment to reform, and hence en- (UNCTAD 2001). L osses during the Asian and Russian courage investors to return. Kaufrrian and-Litan (1998)i6 crises may have totaled $350 billion (Institute of interna- propose that multilateral support be contingent on changes tional Finance, various years).ee'rtheless, tde provision of >in boriowing country livis that iiplement automatic write- multilateral funds uidaobtedly facilitated the repayment of - 'downs on foreign currency denominated interbank loans. international banks during the Mexic an and'Asian crises. All of these proposals face difficIulties: Prequalification Authorities had to balance th6 erosion if matket discipline, requirements could precipitate crises for countries that fail. with the consequences of a complete collapse, which could, Banks' prior commitments to rollover loans during a crisis have had severe effectson many emergingimarkers. can come at the cost of a sell-off of other assets, because While the evidence of mrl hazird-idiced"exessive banks attenpt 'to limit their total ekposuri to the crisis . lending is inconclusive, giver!hhe uncertainties involved it is country; It is-difficult to define before the crisis what partic- prudent to explore means of,reducing the potential impact , ular institutional arrangements would be most desirable to of multilateral support,on m6ralhiard: 'Ofthe.15-:18gest I-- il-in private investors. This may depend, in part, on - emerging market borrowersRi I1997Ivhich t' ether ac- whether a liquidity or solNency crisis is involved. Standstills count for 80 percent of capital market fl6ws to developing and write-down'requirements could have a chilling effect countries), 8 had been the subject of rescue packages,4y'- on the pfovision of finance to emerging markets (although 2001. Some-of themreceived seve'ral iidividial loans..' majority based collective action clauses could support mar- Some proposals have.focused on limiting the itexibility of ket discipline). Nevertheless; there is a growing recognition multilateral instituti6ns by allowing:rescue packages only' that greater attention to private sector participation in re- - for solvent borrowers who p equalify for loans (Meltzer solving crises is warranted. For example, the recent IMF 2000). Other proposals have emihasiied prioi actions loan to Argentina provided that the disbursement of some that force-private creditors to recognize losses opr6vide - committed resources could be brought forward to support resources during a crisis. For example, eligibility for rmifti a voluntary and market-based operation to increase the via- national assistance during i future crisis could be bondif bility of Argentina's debt profile. A-review of international tioned on the government's obtaining lprior commitmeni by arrangements for crisis support that provided for greater the private sector to roll over malturing claims or to pro ide' private s6ctor recognition of losses could -help limit the new money. Still other proposals have focused on ex ante potential for moral hazard-in future lending. 46 PRIVATE CAPITAL FLOWS TO EMERGING MARKETS flexible exchange rate regimes, disciplined do- Table 2.6 Projected capital market flows to mestic monetary policies and, most important developing countries of all, limited short-term external liabilities (billions of dollars) and near-term refinancing needs. These mea- -2O2 2 00, 2I04 sures have helped limit the spread of problems from Argentina to other emerging markets Bonds 9 66 over the past year. Equity 0 32 Loans 93 73 f9'1 East Asia and Pacific 'I S4 59 8 Latin America The prospects for capital market Other SS 44~ 7' flows and FDI Note: These projections for 2002-04 are based on 53 separate vector autoregression (VAR) models (see annex 2.1 for a descrip- Capital market flows are expected to contract tion) for bond, equity and bank lending flows to 21 emerging further in 2002- market economies (ranked according to the size of gross flows in 2001 starting with the top recipient country): Brazil, Mexico, Capital market commitments, after dropping from Korea, Turkey, South Africa, Argentina, China, Poland, Malaysia, $228 billion in 2000 to only $171 billion in 2001, the Repblica Bolivariana deVenezuela, Colombia, the Philippines, .5 Russia, Lebanon, Hungary, Egypt, India, Thailand, Indonesia, may moderate further to some $160 billion in Lithuania, Morocco. The flows covered in these models accounted 2002 (see table 2.6), which is the lowest level for 81 percent of gross capital market flows to developing countries in 2001. The projected flows were then scaled up using 2001 actual since 1994. Investors are likely to remain cautious flow numbers, to arrive at the total for all developing countries. about emerging markets in early 2002, because the synchronized economic slowdown in all major in- dustrial countries limits demand for developing countries' exports, affecting the latter's ability to ing flows are expected to rise by nearly a third by service external debt. Risk appetite remains low 2004, compared to the level in 2002, while equity and financing constraints on banks and other in- flows are expected to recover rapidly from the ex- vestors remain tight in the industrial countries, so tremely low level of 2001. the demand for developing-country assets (espe- The pace of recovery in gross flows will also cially subsovereign assets) is likely to remain low vary depending on creditworthiness and demand during the first half of 2002, at least. These influ- conditions in recipient countries. The trends in the ences are likely to outweigh the reduction in inter- forecast are driven by East Asia and Latin America, est rates and increase in liquidity with the easing which accounted for over two-thirds of total capital of monetary policy in the United States (and, to a market commitments in 2001. Flows to East Asia lesser extent, in Europe) over the past year.i7 will increase relatively rapidly, largely because of China's forecast strong growth, low level of short- -but a rebound is anticipated for 2003 term debt, and high level of international reserves. The recovery in industrial countries that is antici- By contrast, the recovery in flows to some of the pated to begin in the second half of 2002 should East Asian crisis countries may be slower because set the stage for a rise in capital market commit- excess capacity continues to depress the demand for ments, to $179 billion in 2003 and $216 billion in finance. In Latin America and the Caribbean flows 2004. Capital flows should recover because eco- will recover more slowly, in part because Argentina nomic growth in most of the major emerging mar- is likely to see impaired access to the capital mar- ket economies is expected to improve and inter- kets in the wake of its restructuring of outstanding national interest rates are expected to remain low. debt. Also, commodity exporters in the region will The recovery in flows will also be supported by see only a limited rise in export revenues (and thus the low levels of short-term debt and high levels of market access), because non-oil commodity prices reserves in many emerging markets after the expe- are expected to rise by only 8 percent in 2003, and rience of the financial crises in the late 1990s. For remain 25 percent below the level of 1997, and oil 25 major emerging markets, the ratio of short- prices are expected to fall through 2003. By con- term debt to reserves fell from about one in 1997 trast, Mexico is expected to benefit from the recov- to two-thirds by June 2001. Bond and bank lend- ery in the United States, and is likely to see a sharp 47 GLOBAL DEVELOPMENT FINANCE rise in flows due to improved economic conditions. and a decline in global FDI flows were accom- Flows to the other regions will also rise, and they panied by rough stability of FDI flows to develop- generally maintain their share of total capital flows ing countries. This resiliency of FDI to developing during the forecast period. countries in the face of adverse global economic conditions reflects the importance of domestic de- Any rebound depends on developments terminants of FDI flows (see section above on FDI in Argentina trends in 2001). In addition, some of the major re- The crisis in Argentina is a major risk to this fore- cipients of FDI flows, in particular China, are ex- cast. Before the events of the past year Argentina ac- pected to continue to achieve robust growth despite counted for 16 percent of emerging markets' bonds the global slowdown. outstanding on the international capital markets. While FDI flows are expected to remain re- Proposals to restructure Argentina's bonds could silient, the projected 4 percent per year increase reduce investors' willingness to take on emerging from 2001-04 (2 percent in real terms) is less than market assets, particularly if negotiations are lengthy half the rate experienced over the 1990s. We antic- and marked by confrontation.i However, there are ipate that the same forces that drove FDI in the several reasons why the contagion effects of the cri- 1990s-globalization in production due to techno- sis could be limited. Over the past year the Argen- logical innovations in communications and trans- tine crisis has had only a limited and fleeting impact port, coupled with better policies in developing on the demand for the debt of other emerging mar- countries-will continue over the next few years. kets (see box 2.4). The crisis in Argentina has been However, the stock of FD1 in developing countries long anticipated, which has tended to mute the im- is much larger now than 10 years ago, and exports, pact on investors in comparison with the crises in an important driver of FDI, are expected to grow East Asia and the Russian Federation, which were at a much lower pace over the next few years (less major surprises. Secondary market prices on Argen- than 3 percent more rapidly than GDP, compared tine bonds have already fallen substantially, and re- with 6 percent during the 1990s). Moreover, M&A flect relatively low recovery rates. Many current activity by multinationals, an important source of bondholders are likely to have bought the bonds at FDI flows, is declining after its peak in 2000. Al- low prices, or to already have adjusted their portfo- though recent surveys indicate that multinationals' lios to account for losses, so they may not react sig- investment plans were relatively unaffected by the nificantly to a debt restructuring. In fact, a speedy September 11th terrorist attacks, the full impact of settlement with creditors that involves a debt re- the economic slowdown on multinationals' invest- structuring sufficient to enable Argentina to make ments remains uncertain.20 Thus it is unlikely that regular repayments could improve market senti- FDI flows would rise as rapidly over the next few ment and increase secondary market prices of Ar- years as they did over the last decade. Neverthe- gentine debt. The forecasts assume that any debt less, by 2004 FDI flows would remain the largest renegotiation will be settled quickly; although Ar- source of finance for developing countries. gentina (and Turkey) receive little in the way of new The bulk of FDI inflows are forecast to con- commitments over the forecast period, these crises tinue to go to countries with relatively large mar- have a relatively limited impact on investors' will- ket size and reasonably good policies. Brazil, ingness to lend to other emerging markets. China, and Mexico attract more than half of flows to the sample countries. The growth rate of FDI is FDI is expected to rise steadily high to countries with good policies and rapid ex- FDI flows to developing countries are expected to pansion of trade. FDI in East Asian economies is be much less sensitive to cyclical developments than expected to rise by over 10 percent per year, due capital market flows." In 2002 FDI to developing to robust increases in flows to China, where the countries is forecast at $160 billion, a slight decline new commitments are already rising significantly, from the estimated $168 billion in 2001, consistent as well as to Korea and Thailand, where strong with slow growth in global output and little in- recovery in GDP and exports is expected. The an- crease in world trade. The same resiliency of FDI ticipated economic growth is likely to boost FD1 flows was seen in 2001, when the recession in in- flows in South Asia, largely driven by India. On dustrial countries, near stagnation in world trade, the other hand, Latin America's share of FDI to 48 PRIVATE CAPITAL FLOWS TO EMERGING MARKETS developing countries will decline, because privati- from January 1990 to December 2001.22 The flow zation transactions (which made up a substantial forecasts are then summed up, and a scaling factor share of FDI to Latin America in the 1990s-see (equal to actual flows to all developing countries World Bank, GDF, 2001: appendix 4) is likely to divided by the model-generated flows in 2001) Is play a less significant role in attracting FDI. used to compute flows for all developing countries as a group. The 21 countries included in this round ac- counted for 81 percent of gross capital market Annex 2.1: Forecasts of Private Flows flows in 2001 (85 percent of bond flows, 96 per- to Developing Countries cent of equity flows, and 75 percent of bank lend- ing). The coverage of these countries in various Capital market flows types of flows as well as in different regions is The econometric framework used for generating summarized in table 2A.1. Also in 2001, the coun- the forecasts for capital market flows to developing tries covered in these forecasting exercise ac- countries follows Taylor and Sarno 1997, which counted for 99 percent of all flows to East Asia, 81 extended the framework developed by Fernandez- percent of flows to Latin America, 73 percent for Arias and Montiel 1996. In this framework equi- Europe and Central Asia, 83 percent for South librium, or "desired" level, of capital flows to a de- Asia, 57 percent for Sub-Saharan Africa, and 58 veloping country is affected by both global factors percent for the Middle East and North Africa. and country-specific factors. Changes in current Forecasts generated by these VAR models idi- capital flows are then determined partly by the dif- cate that industrial-country growth had a positive ference between desired and actual capital flows in impact on the supply of capital flows to developing the previous period and partly by the changes in countries. Increases in interest rates reduced capital the factors determining the desired level of capital flows, while increases in U.S. high-yield spreads flows. were positively associated with increases in EMBI Global factors include growth in the industrial spreads, which in turn had a negative effect on countries (proxied by the U.S. GDP), global liquid- capital flows. In simulations with the model for ity (indicated by the U.S. interest rates), risk aver- last year's Global Development Finance (World sion on the part of international investors (proxied Bank 2001) changes in industrial-country growth by U.S. high-yield spread and Emerging Market had a significantly larger impact on capital flows Bond Index [EMBI] spread), and the prices of than changes in interest rates. Indeed, changes in oil and non-oil commodities. Developing country- U.S. interest rates and the U.S. high-yield spread specific variables include domestic economic caused only a slight deviation in capital flows from growth (proxied by the index of industrial produc- their original trends, and flows soon began to re- tion), domestic consumer price index, domestic vert to their original values (Mody and others credit, domestic interest rates, the level of inter- 2001). The effects of oil and non-oi commodity national reserves relative to short-term debt, and prices varied depending on whether a country was (separately) relative to imports, and the stock price index.21 The global variables are assumed to evolve exogenously, without being influenced by develop- Table 2A.1 How representative Is the forecasting ing-country variables. The latter variables, how- model? ever, are jointly determined along with capital flows, since they affect and are in turn affected by capital flows. The econometric framework uses Bond total the vector autoregression (VAR) technique that de- Equity total Loan total termines country-specific variables endogenously East Asia and Pacific on the basis of their lagged values, taking the global Latin Arerica and the Caribbean Europe and Central Asia variables as exogenous. The model is estimated separately for bonds, Sub-Saharan Africa equity, and loans for each of the 21 major develop- Middle East and North Africa ing countries, using monthly data for the period T 49 GLOBAL DEVELOPMENT FINANCE Table 2A.2 Comparison of forecasts with actual mance. GDP growth has been found to be as- capital market flows to developing countries sociated with larger FDI inflows in several (billions of dollars) studies (Root and Ahmed 1979; Nigh 1985). Year ir i 2. The ratio of exports to GDP represents export- 1990 orientation, which should increase a country's 1991 attractiveness to multinationals by providing 1992 greater access to export markets (Caves, Pot- bothter, and Spence 1980; Saunders 1982). A third 1995 of world trade is accounted for by intrafirm 1996 transactions i a production,who also sk i. 1997 vndeere1mretinaaednahsiherop ital the bulk of FDI flows. 1998 ~~ 1999 3. The GDP growth rate of the top seven indus- 2000 4" I trial countries is used to account for a change in the relative attractiveness of emerging mar- kets to international investors. Thus higher a net exporter or importer of oil and non-oil com- industrial-country growth is associated with modities in a given year. lower FDI inflows to developing countries. Domestic economic factors also played a criti- 4. A better investment climate, in terms of sound cal role in determining capital flows to developing macroeconomic policies, open regimes toward countries. However, these domestic factors are also FD1, and nondiscriminatory frameworks for treated as endogenous in the model, so that they business facilitation, is likely to induce FDI in- both affect, and are affected by, capital flows. A flows to the recipient economy (see chapter 3; decline in capital flows was generally associated UNCTAD 1998). with decreases in the level of domestic credit, do- mestic industrial production, and stock prices. In- The model is estimated for the panel data from creases in reserves were associated with higher cap- 198 1-2000, which covers 30 developing countries ital inflows, while increases in short-term debt that account for more than 80 percent of FD1 reduced flows. A moderate increase in the price flows to developing countries.24 GDP growth in level was positively associated with capital inflows, developing countries, GDP growth in industrial whereas a strong upsurge in prices tended to dis- countries, and exports are lagged under the as- courage capital flows (Mody and others 2001). sumption that FDI is determined largely on the Table 2A.2 compares the flows estimated using basis of long-term commitments by multinationals the methodology outlined above with their histori- (World Bank 1999). Note that this approach to cal trend. Evidently, the model performs fairly well. estimating FD flows does not take into account cyclical effects, as was done with the forecasts of FDI capital market flows. Such effects are probably of The forecast of FDI included in the text is based on less importance to FDI, which typically is based on an econometric model of the determinants of FDI, the prospects for growth over a longer time hor- expressed as a share of developing countries' GDP. zon than for capital market flows. Large and growing markets can accommodate The constant variable [&,) (i=1,..,30) and coef- more suppliers and help them achieve scale and ficients {pkj (k=,..,5) are estimated from the equa- scope economies (UNCTAD 1998), and the size of tion below, and applied to the set of expected val- the recipient country's internal market as measured ues for the independent variables to forecast FDI by GDP is one of the most frequently applied vato- flows for 2001m04 .2 ables in the past research on determinants of FDIi.23 The determinants of FDI include: FDI,& + 1 (GGDP,) + 12 (EX) + b F3 (G7,) + 134 (IC + l35 (T) 1. The average growth rate of GDP over three years prior to the current period is a proxy for FDI, GGDP, EX, IC, G 7, and T represent, respec- investors' view of future economic perfor- tively, FD as ratio to GDP, average growth rate of 50 PRIVATE CAPITAL FLOWS TO EMERGING MARKETS Table 2A.3 Statistics for the forecast of FDI volume I of GDF. The World Bank also presents a Indepnden varabicnarrow measure of net flows on debt for individual countries in volume II of GDF. GDP growth rate ExportsThe data on net aggregate resource flows pre- G-7b.04 sented in GDF reflect liability transactions only Investment (gross disbursements minus repayments). Capital Timeoutflows (such as net lending by developing-country Adjusted R2 residents abroad), short-term flows, and net use of a. Dnots sinifcane atthe1 prcen leel.IMF credit are not included. This results in a sub- a. Denotes significance at the 1 percent level. b. Group of Seven: Canada, France, Germany, Italy, Japan, the stantial difference between net long-term flows as United Kingdom, and the United States. shown in GDF and net external finance as shown c Denotes significance at the 5 percent level. Source: World Bank, Global Development Finance: Country Tables in the balance of payments. and sources cited therein, various years; World Bank, World Devel- These data are available only on an annual opinent Indicators, various years; and World Bank staff estimates. example loan commitments and bond issues) are available at higher frequency. The analysis of capi- GDP over three years, export volume as ratio to tal flows in this chapter depends heavily on this GDP, investment climate index, annual growth higher-frequency data. The quality of the most re- rate of GDP of the G-7 countries, and time trend. cent year estimates varies depending on the lending category. Reasonably accurate information is avail- able from market sources on gross disbursements from bond markets and commercial banks. Debt Annex 2.2: Measuring resource flows repayments are calculated from information on to developing countries terms, although actual payments may vary. Data on International organizations that collect and re- portfolio equity flows are particularly difficult to port data on international financial transactions estimate: while data on international equity issues use different approaches to measuring the move- are readily available, estimates of direct foreign ment of financial resources to and from developing purchases in developing-country stock markets are countries. The IMF's World Economic Outlook re- based on reports from exchanges that differ in ac- ports flows in a balance of payments framework. curacy and coverage. An alternative approach is to aggregate from more specialized systems that independently compile sta- tistics for different types of flows: the World Bank takes a recipient country or debtor perspective and Notes operates the Debtor Reporting System. The OECD 1. Moody's Investors Service classifies Barbados, takes a donor or creditor country perspective: its Botswana, Chile, China, Croatia, the Czech Republic, El data are derived from information on aid activities Salvador, Estonia, Hungary, the Republic of Korea, Lithua- nia, Malaysia, Mexico, Mauritius, Oman, Poland, Saudi reported to the Development Assistance Commit- Arabia, the Slovak Republic, Thailand, Trinidad and To- tee and on export credits reported through the bago, Tunisia, Uruguay, and South Africa as investment- Creditor Reporting System. The Bank for Interna- grade countries. tional Settlements also takes a creditor perspective 2. In part, this reflects dollar appreciation. In Special and compiles information on a quarterly and on a Drawing Rights (SDRs), developing countries' export rev- semi-annual basis on the claims of its reporting enues increased by 2.6 percent. repoting3. This calculation reflects the fall in European and banks on developing countries. U.S. interest rates, the share of floating rate debt, and the In Global Development Finance (GDF) the share of euro- and dollar-denominated debt. It is a lower World Bank uses a broad concept of net aggregate bound of the impact of lower interest rates, since countries resource flows: equal to net disbursements on long- could switch to dollar-denominated debt to take advantage term loans, direct investment, portfolio equity of the larger decline in U.S. interest rates. term quity4. The largest rise in speculative-grade spreads re- flows, and official and private grants. These data flected, in part, the problems of telecommunications and are presented in the text and summary tables of other technology firms. However, the increase was wide- 51 GLOBAL DEVELOPMENT FINANCE spread (only 5 out of 15 high-yield sectors saw a decline in 17. This forecast for capital market flows is based on spreads in 2001). an econometric model that takes into account global macro- 5. The global volume of credit to investment-grade economic developments (such as industrial-country growth borrowers rose by 4 percent in 2001, while credit to specu- and interest rates) that are largely exogenous to individual lative-grade borrowers fell by 23 percent. developing countries, as well as domestic macroeconomic 6. The most recently issued Treasury securities tend to developments in individual countries (see annex 2.1). be more frequently traded, and hence more liquid, than se- 18. The debt workout process may be difficult. Some curities that were issued earlier. Since both recently issued recent events have made it more attractive for holdout in- and off-the-run Treasury securities have the same risk-free vestors (that is, those who do not agree to a bond restruc- return, the spread between the two is used by some ob- turing). See the case of the Elliott Associates vs. Peru as dis- servers as an indicator of liquidity preference. However, this cussed in World Bank 2001. spread may also reflect technical market factors (Duffie 19. This forecast is based on an econometric model 1996). (estimated from panel data for a sample of 30 countries that 7. The top 10 developing country FDI recipients (in account for 80 percent of FDI flows to developing coun- order of the size of flows) are China, Brazil, Mexico, Ar- tries), where the major determinants of FDI are the level of gentina, Poland, Chile, Malaysia, Korea, Thailand, and the GDP, the past growth rate of GDP, growth in industrial Repiblica Bolivariana de Venezuela. countries, the share of exports in GDP, and the policy envi- 8. A number of planned sales of domestic firms have ronment (see annex). been delayed or called off, including a long-standing acqui- 20. A. T Kearney 2001; UNCTAD 2002. sition plan of Daewoo Motors by General Motors and the 21. See World Bank 2001, chapter 2, for more on the cancellation of a plan by Deutsche Bank's subsidiary to pur- explanation of the choice of variables. chase Seoul Bank. 22. We did not estimate a VAR model for an individual 9. About $25 billion of this amount represents flows type of commitment (hank lending, bond issues, or portfolio through Hong Kong (China) that may have originated in equity flow) if it constituted less than 5 percent of total China. flows received by the country. 10. In the face of capital mobility, fixing the exchange 23. Literature includes Root and Ahmed 1979; Schnei- rate limits the ability of the central bank to print money. The der and Frey 1985; Papanastassiou and Pearce 1990; and exchange rate-induced stabilization of import prices also en- Wheeler and Mody 1992. See also UNCTAD 1998 for de- hances credibility by showing evidence that inflation is com- tailed discussions. ing down. Agreement to forgo further wage and price in- 24. Some adjustments were made to FDI data for select creases requires a metric against which mark-ups and countries where a small number of large-scale privatization contracts can be gauged; a pegged exchange rate provides just transactions distorted the trend, or the major privatization such a measure. In contrast, other approaches to stabiliza- programs have reached completion, or both. tion-keying on reductions in the rate of money growth or on 25. The set of constant variables represents fixed ef- the central bank's inflation target-are harder to verify and fects across countries. therefore less credibility-enhancing. Fischer (2001a) observes that few if any countries have successfully brought down high inflations without first stabilizing the exchange rate. 11. Fiscal policy was tightened by 1.7 percent of GDP References in 1999, 1 percent in 2000, and 1.3 percent in 2001, accord- The word processed describes informally reproduced works ing to J. P. Morgan estimates (Werling 2001). that may not he commonly available through libraries. 12. Similarly, the correlation of spreads on Turkish bonds with other emerging markets rose from 0.12 before Adams, A. H. 1993. "Hong Kong's Charms." The China the crisis to 0.39 afterwards. Business Review (November-December). 13. Twenty-six percent of Argentine exports go to A. T Kearney. 2001. "FD1 Confidence Index-Flash Sur- Brazil and 11 percent of Brazilian exports are to Argentina. vey." Presented at OECD Global Forum on Interna- Moreover, each country's top 10 markets (which for Ar- tional Investment-New Horizons and Policy Chal- gentina and Brazil cover 57 percent and 64 percent of ex- lenges for Foreign Direct Investment in the 21st ports, respectively) are also the top 10 for the other country, Century, November 26-27, Mexico City. with the exception of Mexico (for Argentina) and Uruguay Bank of Canada and Bank of England. 2001. "Resolution of (for Brazil). International Financial Crises." February. Processed. 14. The evidence of contagion effects is even weaker if Becker, Torbjorn, Anthony Richards, and Yungyong we look at stock market prices. There is almost no evidence Thaicharoen. 2001. "Bond Restructuring and Moral from stock market prices that the Argentine or Turkish Hazard: Are Collective Action Clauses Costly?" IMF crises affected other emerging markets, again with the ex- Working Paper 01192. International Monetary Fund, ception of the impact on Brazil. Washington, D.C. 15. There is also concern that rescue packages may en- Bosworth and Collins. 1999. "Capital Flows to Developing courage borrowers to pursue unsustainable policies in antic- Economies: Implications for Saving and Investment." ipation of being bailed out. This is unlikely, considering the IMF Seminar Series. No. 1999-21, pp. 1-44. economic costs to countries hit by the crises and the loss of Calomiris, Charles. 2000. "When Will Economics Guide power of politicians who governed in the run-up to crises. IMF and World Bank Reforms?" Cato Journal 20 16. Cited in Helfer 1998. Spring/Summer. 52 PRIVATE CAPITAL FLOWS TO EMERGING MARKETS Caves, R. E., M. E. Porter, and A. M. Spence. 1980. "Com- Lardy, N. 1995. "The Role of Foreign Trade and Investment petition in the Open Economy." Harvard University in China's Economic Transformation." China Quar- Press, Cambridge. terly (U.K.) 144:1065-82. Clark, E. 1998. "Risk Aversion, Wealth and International Meltzei, A. 2000. "Report of the International Financial In- Capital Flows." Review of International Economics stitutions Advisory Commission." U.S. Congress, (U.K.) 6: 507-15. Washington, D. C. Dollar, David, Giuseppe larossi, and Taye Mengistae. 2001. Mody, Ashoka, Mark P. Taylor, and Jung Yeon Kim. 2001. "Investment Climate and Economic Performance: "Modeling Economic Fundamentals for Forecasting Some Firm Level Evidence from India." Prepared for Capital Flows to Emerging Markets." World Bank, Economists Forum, May. World Bank, Washington, Washington, D.C. Processed. D.C. Nigh, D. 1985. "The Effect of Political Events on U.S. Di- Duffie, Darrell. 1996. "Special Repo Rates." Journal of Fi- rect Foreign Investment: A Pooled Time-Series Cross- nance. June. Sectional Analysis." Journal of International Business Eichengreen, Barry, and Ashoka Mody. 1998. "What Ex- Studies 16: 1-17. plains Changing Spreads on Emerging Market Debt: Papanastassiou, M., and R. D. Pearce. 1990. "Host Coun- Fundamentals or Market Sentiment?" NBER Working try Characteristics and the Sourcing Behaviour of U.K. Paper W6408, Cambridge, Mass. February. Manufacturing Industry." Discussion Papers in Inter- - 2000. "Would Collective Action Clauses Raise Bor- national Investment and Business Studies, Series B, rowing Costs?" World Bank Policy Research Working Vol. 2 (140), Department of Economics, University of Paper 2363. May. Washington, D.C. Reading. United Kingdom. Fernandez-Arias, Eduardo, and Peter J. Montiel. 1996. Root, F. R., and A. A. Ahmed. 1979. "Empirical Determ- "The Surge in Capital Inflows to Developing Coun- nants of Manufacturing Direct Foreign Investment in tries: An Analytical Overview." World Bank Economic Developing Countries." Economic Development and Review 10: 51-77. Cultural Change 27: 751-67. Fischer, Stanley. 2001a. "Exchange Rate Regimes: Is the Bipo- Saunders, R. S. 1982. "The Determinants of Inter-Industry lar View Correct?" Finance and Development June. Variation of Foreign Ownership in Canadian Manu- Fischer, Stanley, and D. Citrin. 2000. "Strengthening the In- facturing." Canadian Journal of Economics 15: 77-84. ternational Financial System: Key Issues." World De- Schneider, F, and B. S. Frey. 1985. "Economic and Political velopment: 1133-42. Goldstein, Morris. 1998. The Asian Financial Crisis: Determnt of Foeg De Ie n" r Causes, Cures, and Systemic Implications. Institute for Devlopen T3 (2): 161-75. International Economics: Washington, D.C. ica T 1998 Cia and For In mn Guay, W. 1999. "The Sensitivity of CEO Wealth to Equity Two Tae f C a R .Wd. Risk: An Analysis of the Magnitude and Determi-(UK)2:5-62 "Jokurn nalyi of inancagcnoicsd (ndeternds Taylor, Mark P., and Lucio Sarno. 1997. "Capital Flows to nants." Dveloping Countries: Long- and Short-Term Deteri- 53 (1): 43-71. Gunter, Frank R. 1996. "Capital Flight from The People's NAn od an Conomice 11. Republic of China: 1984-94." China Economic Re- UCA (nted Nations Cnfereneo aend d view 7 (1): 77-96. Harrold, P., and R. Lall. 1993. "China, Reform and Devel- Determinants. Geneva. opment in 1992-93." World Bank Discussion Papers . 2001. World Investment Report 2001: Promoting 215, Washington, D.C. Hausmann, Ricardo, and Eduarto Fernandez-Arias. 2000. - 2002. "FDI Downturn in 2001 Touches Almost All "What's Wrong with International Financial Mar- Regions." Press Release TAD/INF[PR36, January 21, kets?" Inter-American Development Bank, Research Geneva. Department Working Paper 429, Washington, D.C. Werling, Vladimir. 2001. "Argentine Confidence Crisis: Fac- Helfer, R. 1998. "Rethinking IMF Rescues." Brookings In- ing a Policy Dilemma." Economic Research, Morgan stitution Conference Report #1. http://vww.brookings. Guaranty Trust Company. August 10. org/palconferencereport/crl/crl.htm. Wheeler, David; and Ashoka Mody. 1992. "International In- IMF (International Monetary Fund). Various years. World vestment Location Decisions: The Case of U.S. Firms." Economic Outlook. Washington, D.C. Journal of International Economics 33: 57-76. International Council of Securities Agencies. 1999. "Private World Bank. 1999. Glohal Development Finance. Washing- Burden Sharing: A Voluntary Approach." http://www. ton, D.C.: World Bank. sia.com/international/html/burden.html. . 2000a. Glohal Economic Prospects. Washington, Institute of International Finance. Various years. "Capital D.C.: World Bank. Flows to Emerging Market Economies." . 2000b. Glohal Development Finance. Washington, Kumar, Manmohan S., and Avinash Persaud. 2001. "Pure D.C.: World Bank. Contagion and Investors' Shifting Risk Appetite: Ana- - 2001. Glohal Development Finance. Washington, lytical Issues and Empirical Evidence." IMF Working D.C.: World Bank. Paper 01/134, Washington, D.C. Zhang, Xiaoming Alan. 1999. "Testing of Moral Hazard in Lane, T., and S. Phillips. 2000. "Does IMF Financing Result Emerging Market Lending." Institute of International Fi- in Moral Hazard?" IMF Working Paper 00/168. nance Research Papers, 99-1. August. Washington, .C. 53  3 The Poor Countries' International Financial Transactions Poor countries have benefited from ment climates. Those countries that established the the growth of global capital flows stable macroeconomic policies and effective regula- T he globalization of production and financial tory regimes necessary to attract foreign bank par- services has provided the opportunity for poor ticipation increased the access of domestic banks countries to increase their reliance on private sector to trained personnel and technological advances, international financial transactions.' Poor countries while rising competition from foreign banks helped lack access to capital markets and official flows reduce the costs of financial intermediation. Poor have fallen, while total aid has declined along with countries' greater openness to capital flows means the share of the poor countries. However, foreign di- that they have to cope with the macroeconomic rect investment (FDI) flows have risen substantially: effects of capital mobility. Sustainable macroeco- while the poor countries remain dependent on offi- nomic policies marked by low inflation and debt cial external finance, they now receive the same levels are essential to limit capital outflows, and amount of FDI as other developing countries, in re- sharp changes in outflows (or capital repatriation) lation to the size of their economies (table 3.1). FDI can complicate efforts at stabilization. flows to the poor countries have become more di- versified: the share of the mineral- and oil-exporting countries in total FDI to the poor countries fell from almost half in 1991 to 20 percent in 1997. Poor Financial integration in the countries have participated in the global expansion poor countries of commercial banks: foreign banks' assets now ac- count for 40 percent of total bank assets in the poor Financial integration has increased since countries, twice as high as in 1995. Despite capital the 1980s controls, poor countries' residents have placed sig- The poor countries' private international financial nificant amounts of capital abroad: the stock of transactions increased substantially during the capital outflows from the poor countries were larger 1990s. Official flows have fallen with the decline relative to cumulated domestic savings and the in total aid and the fall in the poor countries' share stock of reserves, and only slightly smaller relative of aid (see chapter 4), while capital market flows to gross domestic product (GDP), than outflows (bank lending, bond issues, and portfolio equity) from other developing countries. have remained relatively small. By contrast, FD1 As in middle-income countries, the quality of has risen seven-fold, and now represents over 40 the investment climate determines the extent of percent of all long-term resource flows (table 3.2).2 poor countries' access to capital and the extent to Nevertheless, the poor countries' reliance on pri- which foreign capital benefits the domestic econ- vate flows remains somewhat below that of other omy. Countries with sound investment climates developing countries, where private flows averaged tend to attract more FDI, limit capital outflows, about 4 percent of GDP in the late 1990s. and enjoy greater productivity of both foreign and One indicator of the extent of integration domestic capital than countries with weak invest- with the rest of the world is the correlation be- 55 GLOBAL DEVELOPMENT FINANCE Table 3.1 Net external financial flows to of 1.4 percent of GDP. Albuquerque (2001) has developing countries, 1999 noted that countries with worse international credit (percent of GDP) ratings tend to have greater difficulties in attract- i al.ing capital market flows than in attracting FDI. ___________g~~gw OD~rflws This dependence on FDI ratner than cptlmar- ket flows reflects a range of higher risks associated Poor countries -JS 5,f 15 Poor ountres ~with investing in poor countries, notably less stable Other developing countries ~areonomic conditions, weaker institutions, countries 8m,acro7ec13/ S and a less favorable environment for private sec- a. Includes bonds, portfolio equity, and bank lending. b. Official development assistance. Source: World Bank Debtor Reporting System (DRS) and staff countries are relatively undiversified. For example, estimates. estimtes.primary commodities account for 70 percent of ex- ports from Sub-Saharan Africa. The poor countries tween investment and savings.3 Countries that are are thus more prone to exogenous shocks, such as tightly integrated into global financial markets changes in the terms of trade and, in the case of should exhibit a low correlation between domestic agricultural products, adverse weather conditions. savings and gross investment. For example, if a Higher risk leads to a bias toward equity finance, in natural disaster reduces domestic savings but does part because FDI typically includes management not affect the return on new investment, firms in expertise and branding, which help to compensate well-integrated economies can rely on interna- for greater risk. Perhaps more important, banks tional capital markets to maintain investment lev- face difficulties in raising interest rates sufficiently els. At the extremes, in an autarkic economy say- to compensate for risk, owing to adverse selection. ings and investment are identical (the correlation Different entrepreneurs have different (and unob- is one), while in a perfectly integrated economy the servable) probability of repaying loans. The more correlation would in theory be zero.4 In the poor risky entrepreneurs are willing to pay a higher in- countries, the correlation between savings and in- terest rate, so banks limit risk by rationing credit vestment declined sharply in 1995-99, after a steep through quantity limits, rather than through rise from the late 1980s to the mid-1990s (figure changes in interest rates. 3.1). The variability in the series over time makes it difficult to say whether the recent decline will be -including asymmetric information sustained over the medium term. Again, the cor- International investors often have little informa- relation in the poor countries remains above that tion on poor-country borrowers. Most poor coun- of other developing countries, although the differ- tries often have relatively small markets, little ence has narrowed since the mid-1980s. coverage in the international media, and signifi- cant geographic and cultural distance from high- The preference for FDI reflects high risks- income countries. Thus external investors are par- While FDI to the poor countries has surged since ticularly subject to asymmetric information with the mid-1980s, net capital market flows to the poor respect to opportunities in poor countries: that is, countries has remained near zero. In other develop- the owners of firms tend to have much more infor- ing countries these resources represent an average marion on the firms' profitability than lenders or Table 3.2 Net long-term capital flows to poor countries, 1986-99 ktcia flows refect a. rag3f.ihrrik 9sscae Foreign direct investment o a M4 1.1 2.7 Source: World Bank rIRS. 56 THE POOR COUNTRIES' INTERNATIONAL FINANCIAL TRANSACTIONS Figure 3.1 Five-year rolling correlation between savings and investment, 1974-1999 Correlation coefficient 0.8 0.7 --- Other developing -49- Poor 0.6 0.5 W 04 02 01 0 0 1 I l l IIII l I l I l l l i 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 Source: World Bank data. outside investors, particularly foreign ones. High 2001). On the debt side, the laws and infrastruc- risk in the presence of asymmetric information ture necessary to collect on collateral in the case of leads to quantity constraints on loans (Stiglitz and loan defaults are often inadequate, so that banks Weiss 1981), and debt contracting may not be fea- are often unwilling to lend.6 While increased secu- sible or desirable (Trester 1998). Lending to poor ritization of loans is a potential approach to im- countries is thus severely constrained, and much of proving access to debt flows, the cost and com- the bank lending that occurs must be guaranteed plexity of arranging such deals, and the risks (see chapter 4). By contrast, when foreign firms involved in reducing the flexibility of foreign ex- take an ownership stake through FDI they can change management and taking on large debts at exert more control over local managers, and thus market rates, limit the use of securitization by the obtain better access to information (compared with poor countries (box 3.1). banks) about a project's current and potential profitability (Razin, Sadka, and Yuen 1997). Trade credit is often an attractive financing option The preference for FDI also reflects Another means of increasing credit to risky coun- institutional weakness in debt tries in the presence of asymmetric information is and capital markets to borrow from suppliers rather than banks. Trade The institutional and legal structures required to credit, a financial agreement under which an ex- reliably enforce contracts in the debt and capital porter (or supplier) extends credit to finance the markets are often lacking in poor countries. Pro- purchase by an importing firm, offers a good alter- tection of minority shareholders is often limited, native for firms that lack access to banks. Suppli- disclosure standards are inadequate, and the ad- ers are often better placed than banks to lend to ministrative processes necessary to buy and sell firms in developing countries because suppliers shares impose high costs and delays, so issuance on have considerable information on the firm and its the capital markets is discouraged. Stock markets markets, and thus are less affected by asymmetric tend to be very small in the poor countries. For ex- information. Suppliers can impose greater sanc- ample, of the 19 African stock markets, almost tions in the case of default by cutting off access to half have market capitalization of less than $1 bil- supplies and repossessing goods against which lion, compared to the $220 billion capitalization credit has been granted. Suppliers have an advan- of the Johannesburg exchange (Oxford Analytica tage over financial intermediaries in selling repos- 57 GLOBAL DEVELOPMENT FINANCE Box 3.1 Improvinig market access through future-flow,securitization Securitization-the conversion into tradable securitieF'-of Future4low securltization, 1987-99 future hard-currency receivibles isap tential heaiinsof. Bons fdollars improving the access of ppor countries to mternational capi- tal markets. At the same time, scuritization in the poor 92 countries must be handled cautiously, due to the limits im- posed on government's access rdforeigri exchange and the risks of incurring debt at marketaites- ; . In a typical future-flow transaction, the-botrower pledges the future revenues fromsales of a product (for a3 example, oil) as collateral. By a 'jgal ar,rangement betwee 4 the borrower and major international customers,,payments 20 1s U * for the future product are-directly deposited in an ffshoxe 2 collection account managed byfairustee?The debt-is seticed o.s 04U0U from this account, and excess collectionsare forwarded to the 0 r- I borrowing entity in the developing country. This transaction 1987-9o 1991 1992 t993 1994 1995 1996 1997 1998 1999 structure reduces the ability ofthe governmient toInterfere -Soure: Fitch, Moody's, Standard & Poor's. with debt servicing, while the irket,risk arisiniidfro price and volume volatility is mitigatedby setting the amount of collateral higher thanthe dlebt svice liibility So far, tee vronrment involved in these transactions may identify priori- have been no debt defaulto rted futuifflow assetaled 'ties for reform. Public policy to facilitate future-flow-backed securities issued by developing-country borrowers, even dir securitiztions could focus on clarifying bankruptcy laws, re- ing crises. For example, in he ecomli ations transac- ducing transaction costs by facilitating the pooling of receiv- tion mentioned below,Pkistan cotmueff service this debt ables generated by several issuers, and educating policymak- even in the face of selective defalt o ofi'isovereign debt. ers and potential issuers about the benefits and risks involved. Future-flow securitizition in developn "countries ice A number of factors, however, constrain the growth of future- the first important future-flow secriitiidd transaction in a, flow transactions in the poor countries, including the high developing country (by Mexic Isemex in 1987,15(O preparation costs and l,ong lead times involved, and the lack future-flow securitizations (that' ere rated,by major ratng of legal larity on bankruptcy procedures in many countries. agencies) have raised mre than $36 billion. The issuance'of Securitized lending also presents some risks to poor- future-flow receivable-backed securities increased especially country governments. Securitized arrangements that commit after the Mexican crisis in,19949 (seerfigure). About 45 ' substantial share of a country's foreign exchange resources percent of rated future-flow transactions in U.S. dollar terms inay also reduce the attractiveness of nonsecuritized debt. A (and one-sixth in terms of inumber of-deals) are backed by-oil ,country's securitizations may violate negative pledge commit- and gas export receivables. Hardcirrency futt." e -ecivabiles ients to multilateral lenders. Escrow accounts reduce the au- such as credit card and telephn recei ables, and worker t iflexibility in mobilizing and managing foreign ex- remittances, and even export receivables to be generated in' change. For example, escrow account arrangements made by the future by new investrient projects?a e also beensect"i- a public sector companymay make it impossible for a gov- tized. In Argentina, soie-provimes have securitized portions ernment to draw on the company's foreign exchange receipts of their future tax receivables frorh the,federal government. to support imports during a temporary decline in the terms of Future-flow securitization. rureflj securitizationhas trade, thus imposing a costly and perhaps unnecessary adjust- been used rarely in the poor countries. One example i the ' ment. Committing a large share of the public sector's foreign 1997 transaction in which Pakistan Telecommunications - exchange receipts to securitized arrangements can signifi- Company Limited, a state-owned, onpahy, raised2$250ii-, :cantly increase the economic contraction required due to a lion in bonds backed by future telphone settlenet receiV wihdrawal of flight capital. There is also a danger of prolif- ables from international telephori comparnies. This issue was eration: governments that agree frequently to the use of such rated investment grade, four notches higher than the sovereign arrangements may see creditors insist on them in most cases. rating. Given their revenes from commodities, tourism, and This concern may be more muted in the case of a private remittances, poor countries could potentially raise as much as company, although even here governments with foreign ex- $11 billion by securitizing exports(using a conservative-5if change suirreader requireients may see their access to foreign overcollateralization ratio o 6i ,9iaieeivables),'ifaddiion exchangedecline.The major issue is that poor-country gov- to the potential for securitizationof telephone receivables. ernments, and in particular heavily indebted governments, Securitized lending may be useful at teimargin to in- - must remain cautious about contracting debt at market rates. crease access to finance and to gaiA ntry to capitamarkets.- Securitized arrangementsmay facilitate access to capital mar- There may also be positive externalities associatedlwith secu- kets, but they do not necessarily make it prudent for poor ritization: the close scrutiny of the'legal a sd stitutional ent counries to borrow on hard terms. 58 THE POOR COUNTRIES INTERNATIONAL FINANCIAL TRANSACTIONS sessed goods, since usually the supplier already has a network for selling its goods, especially if they Figure 3.2 FDI-to-GDP ratios, 1991-2000 have not been transformed by the buyer. By con- trast, a bank's threat to cut off future finance may have little influence on the buyer's immediate op- - Poor - Nompoo erations (Petersen and Rajan 1994). Moreover, the 30 prospect of a close and continuing trade relation- 25 ship with the supplier reduces the likelihood that a solvent buyer would default, as the cost of obtain- 20 ing goods from a single firm can be lower than 15 purchasing them through separate transactions (Mian and Smith 1994).F199 05 0 0 FD1 o te por cunties1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 FDI to the poor countries Source: World Bank, Global Development Finance: Country Tables and sources Poor countries benefit from a global surge in FDI flows- The surge in FDI reflects both the increase in global FDI flows and improvements in the invest- confirmed the importance of the investment cli- ment climate in the poor countries. Global FDI mate in determining the level and efficiency of do- flows increased by 24 percent per year during mestic investment (box 3.2). 1991-2000 as reduced trade barriers and tech- The poor countries have made significant nological innovations encouraged the growth of progress in improving the investment climate. The globally integrated supply networks (World Bank median inflation rate in the poor countries fell to 2001a). Developing countries as a group saw FDI under 5 percent by the late 1990s, compared with flows rise 20 percent at constant prices, and the almost 8 percent early in the decade. The poor rise in FDI as a share of GDP during the 1990s was countries' average fiscal deficit fell from 7 percent virtually identical in the poor and other developing of GDP in the early 1990s to 4 percent in the late countries (figure 3.2), although the share of the 1990s. Almost half of a sample of 44 poor coun- poor countries in total FDI to developing countries tries (the choice of countries was based on data declined during the 1990s. FDI flows to the poor availability) reduced their fiscal deficit by more countries increased to almost 3 percent of GDP than 2 percent of GDP, and only 12 saw a deterio- and 15 percent of domestic investment, about the ration in the fiscal deficit. Some countries achieved same ratios as in other developing countries. broader reforms to encourage private sector activ- ity. Restrictions on foreign entry and ownership -and improvements in their investment were either eased or removed, and export process- climates ing zones (EPZs) and various tax and duty reduc- The rise in FDI flows to the poor countries over tions were introduced. Twenty-two out of a sample the 1990s in part reflects significant progress in of 24 poor countries either introduced EPZs or improving the investment climate, a term which provided other forms of tax- or duty-exemption refers to the numerous ways in which government for imports, or reduced taxes on imports over the policies affect the productivity of investment by 1990s. Several countries eased rules on foreign cur- fostering openness to trade and FDI, macroeco- rency transactions, at least as far as the current ac- nomic stability, fair and efficient public sector count is concerned (see below). The poor countries administration, low corruption and effective law also have made some progress in health and educa- enforcement, strong financial institutions, the pro- tion indicators that reflect improvements in human vision of effective infrastructure, sound regulation, capital, a critical component of a strong invest- and measures to ensure the health and education ment climate. For example, the adult illiteracy rate of the work force. Several empirical studies have declined from 45 percent in 1990 to 37 percent in 59 GLOBAL DEVELOPMENT FINANCE Box 3.2 The investment climate and domestic investment T he economic literature provides considerable empiri- Brempong and Traynor(1999) also ide evidence on cal evidence regarding the impacr of, the investment-. the negative effect of political instability,on investment climate on the level and productivity:6f private invest :for a cross-section of 39 Sub-Saharan African countries ment. The elements of the investment climate covered in during,1975-88. Studies on individual countries in Africa empirical studies include macroeconomic-policy, the legal have,provided similar evidence (Thomas 1994 for Tanza- framework, political instability, infrastructure, and health nia, and Jenkins 1998 for Zimbabwe). In a study of 18 and education services. Both the policy framework and Latin American countries over the period 1970 to 1981, uncertainty concerning its administration are important. Gyimah-Brempong and Mufloz de Camacho (1998) show Poor macroeconomic policies have a negative impact that political instabilityreduces investment in both human on the level of investment. Pfeffermann and Kisunko, and physical capital. Using a sample of 40 countries, (1999) list inflation among the major deterrents to invest- Bubnova (2000) points out that political disorder aggra- ment worldwide. Ndikumana (2000) shows that inflation yates risk and therefore reduces private infrastructure has had a negative effect on investnenfin Sub-Saharan irvestment' Africa, while Oshikoya (1994) gets the same results'for Thelack of adequate infrastructure "nd human capital, a sample of low-income countries. Other authors have- has been found to reduce private investment. Pfeffermann found that uncertainty about macroeconomic policies and Kisunko (1999) report that inadequate infrastructure reduces investment (Alesina and Tabellini 1989). Sev&ral constitutes one of the major obstacles to doing business. authors have shown that real exchange rate volatility,'a Reinikka'and Sevensson (1999) identify the role of unreli- proxy for uncertainty, is negatively related to private- i- able and inadequate power supply in reducing investment vestment (Aizenman and Marion 199S ;Serv6n 1996anA in,Uganda,,despite considerable progress in establishing 1998; Serv6n'and Solimano 1993; PrBhetti and Weder macroeconomic stability and structural reform. Cshikoya 1998; Hausmann and Gavin 1996). (1994) finds a positive relationship between the infrastruc- An appropriate legal framework, ad its fair enforce- ture component of public sector investment and private ment have an importanit impact or investment. Uncer- investment in low-income countries. A study on Pakistan tainty in property rights enforcement- (Kn~ack ahd'Kc -' 17how588. Scomplementar nffectuof pblic-infrastructure 1995) and corrupiion-(Mauro1995)have significanthneg- av investment on private sector investment (Sakr 1993). Like- ative effects on investment.' Brunetti and Weder (1998),in wise, a study of the Caribbean region (Clements and Levy a cross-sectional study of 60 countries, find that the lack' 1994) shows that public education investment have signifi- of rule of law and a high-evel of corruption are especially cant effects on private investment. detrimental to investment. Analyses baedt on surveys (Pfi Analysiesofsubnational ipedinmentsto Thvetment effermann and Kisunko 1999),addpanel data (Bubnov& ' have aho&e m'phasized the importance of the investment 2000) emphasize corruption, crime, and,unpredictable' climate. In a study of Indian states Dollar, arossi, and public administration as deterrents to,investment. Individ- Mengisae (2001) find that after controlling for establish- ual country studies also provide evidence of the impact of, ment size and industry type, the variation in factor produc- the policy environment'on investment6'Africa. For exam'- tivity across the states can in part be attributed to the van- pie, Devarajan, Easterly, ind&Nck (201), findthat inap-?. ation in-regulatory burden.'The study also shows that the propriate public policiesseverelyreduced the productivity,' average anual fixed capital formation isfour times higher of the Tanzanian manufacturing sector in states with better investment climates (based on business Empirical studies also have found that political managers' rankings) than in others. A survey of percep- instability has a significant negative effect, on investment ions of business environment in five regions of Russia (see studies of large cross-couIntry-data(siets by Baro identified inflation, lack of access to financing, poorly [1991] and Alesina and Perotti [199.6])., survey, of for- -functioening judiciary systems, and administrative barriers eign-owned firms in 24 'African counttries(found political 9to investment (that is, high tax rates, tax regulations, and and policy stability to be the most important factors af- corruption in the public sector) as the most serious obsta- fecting their investmentmdecisions (Sievets2001). Gyimah- ces to investment (Coolidge, Kisunko, and Rahman 2001). 60 THE POOR COUNTRIES' INTERNATIONAL FINANCIAL TRANSACTIONS 1999, and the infant mortality rate dropped from Table 3.4 FDI as ratio to GDP and policy performance 85 per 1,000 live births in 1990 to 73 in 1999. index in poor countries Nevertheless, the investment climate in most io, X poor countries remains less attractive than in many middle-income countries. The average fiscal deficit Middle 4. 3.0 is one percentage point of GDP higher in the poor countries than in the other developing countries. Note: This excludes oil and mineral exporters. The policy perfor- Health sector indicators are worse, despite the pro- mance index is measured in 1995. FDI as ratio to GDP is an average outlnedabov, Fr exmpl, lie epectncy during the 1996-99 period. The sample for this figure consists of gress outlined above. For example, life expectancy countries. at birth remains 13 years below the level in other Source: World Bank, Global Development Finance: Country Tables developing counties, and the adult illiteracy rate is and sources cited therein, various years, World Bank, World Devel- moretha twce s hgh.Growh i th por cun- opment Indicators, various years; and World Bank staff estimates. more than twice as high. Growth in the poor coun- tries has been slower: per capita GDP rose by only 0.3 percent per year in the 1990s, compared with 1.9 percent in other developing countries.' achieved the greatest improvement in the invest- ment climate for a sample of 23 African countries during 1992-97 (World Economic Forum 1998), and the ratio of FDI to GDP rose by 81 percent in Improved investment climate is Uganda, 35 percent in Tanzania, and 33 percent in associated with rapid growth of FDI Mozambique.'O Armenia pushed ahead with open- P oor countries that made progress in improving ing sectors to foreign investors and promoting pri- the investment climate during the 1990s at- vatization, which led to an 80 percent upsurge in tracted large FDI increases. In the countries where FDI as ratio to GDP over the past decade. Priva- policy and institutional performance improved tization transactions accounted for a significant most, FDI as a ratio to GDP increased by 25 per- share of FDI inflows in some of these countries (15 cent per year, while in the countries whose policies percent in Uganda from 1992-97, and 25 percent improved least, the FDI-to-GDP ratio increased by in Bolivia from 1995-99). less than 6 percent annually (table 3.3). The coun- tries that showed relatively good policy and insti- Policy measures that attract FDI- tutional performance in 1995 received more FDI In addition to overall improvements in the invest- as a ratio to GDP during 1996-99 (table 3.4). ment climate, policy measures that are specifically The relationship between improvements in the designed to ensure equal treatment of foreign and investment climate and increases in FDI flows can domestic investors have been important in attract- also be seen in the experience of individual poor ing FD1 to the poor countries. New laws on for- countries. Uganda, Tanzania, and Mozambique eign investment have been formed to permit profit repatriation since the early 1990s, while accessions to international agreements and institutions as Table 3.3 Annual change in policy performance well as conclusions of bilateral investment treaties and FDI as ratio to GDP, 1991-99 and double taxation treaties have accelerated (percent) (UNCTAD 2001a). According to a survey con- ~ ducted by UNCTAD in 1997, 26 of the 32 least 10% developed countries in Africa in the survey had a Improvement in policy liberal or relatively liberal regime toward the repa- performance index triation of capital. Table in3 FDI as ratio to GDP a2n5.d l p Note: Highest and lowest groups of countries are hased on the order -and factors that discourage it of improvement in the policy performance index during the period Some of the poor countries have not achieved the of 1991-99. Policy performance is measured py the Bank's Country Policy Performance Rating. improvements in the investment climate necessary Source: World Bank, Global Development Finance: Country Tables higher FDI flows. Civil strife, which and sources cited therein, various years; World Bank, World Devel- noure opment Indicators, various years; World Bank staff estimates affected 13 poor countries during the 1990s, can 61 GLOBAL DEVELOPMENT FINANCE depress foreign investment (although some of the eliminate an important constraint on the market countries affected by conflict have continued to access of African firms. receive foreign investment in protected natural re- source projects). Some countries continue to im- -but only if the investment climate is sound pose restrictions on foreign entry and ownership Nevertheless, estimates of the average impact of and foreign exchange transactions, as well as dis- FDI on growth in poor countries are mixed, in criminatory tax provisions. In Kenya, where for- contrast to comparable estimates for developing eign investors face multiple licensing requirements countries as a group, which often show a positive and high withholding taxes on royalties, FDI re- impact of FDI on growth." Kumar and Pradhan mained less than 0.2 percent of GDP during (2001) find that a 1 percent rise in the ratio of FDI 1991-99 (Pigato 2001). Similarly, in Yemen, where to GDP in the poor countries is associated with an sizable outflows of FDI have been recorded since increase in GDP growth of about 0.18 percent, the mid-1990s, licensing requirements discouraged compared with a rise of 0.12 percent in the case new investments, despite incentives such as tax hol- of domestic investment.'4 By contrast, Blomsrm, idays and customs exemptions. Pakistan has seen a Lipsey, and Zejan (1994) found the impact of FDI steady decline in FDI inflows since 1996 due to in- on growth of the lower-income countries to be vestor concerns over political developments. positive but not statistically significant. These mixed results reflect weak investment FDI can boost investment and productivity- climates in some countries. Even if FDI is strongly Recent empirical work indicates a strong link be- linked to higher investment, increased investment tween the volume of FDI and domestic investment. may generate limited benefits for growth if the in- Bosworth and Collins (1999) and Mody and Mur- vestment climate is poor. Devarajan, Rajkumar, shid (2001) find that a dollar of FDI results in an and Swaroop (1999) present some cross-country almost one-dollar increase in investment. By con- evidence for Africa in which neither public nor trast, international portfolio flows and bank loans private investment is correlated with growth due have a much smaller impact on investment. In ad- to low capacity utilization and a distorted policy dition to the impact of FDI on the volume of environment.l1 Bhagwati (1978) and Balasubra- investment, the presence of foreign firms can gen- manyam, Salisu, and Sapsford (1996) find that the erate important benefits for domestic firms by in- effect of FDI on growth is stronger in countries creasing their knowledge of-and access to-ad- that pursue export-oriented trade policies than in vanced technology, by improving the overall skills those adopting inward-oriented policies. Even in of the work force, and by increasing demand for poor countries with sound macroeconomic poli- domestic firms' products and the supply of in- cies and limited public sector interventions in com- puts." These "spillover" benefits of FDI are great- petitive markets, low levels of education and skills est in countries with sound investment climates may limit the benefits of FDI. Borensztein, De Gre- marked by well-developed human capital, efficient gorio, and Lee (1995) and UNCTAD (1999b) find infrastructure services, sound governance, and that the interaction between FDI and an indica- strong institutions. tor of human capital in cross-country regressions The presence of foreign firms also can be im- has a significant impact on growth in developing portant in the poor countries by improving local countries, but that FDI alone does not. 16 firms' access to international markets. The role of The size of the technological gap between do- foreign firms as export catalysts has been examined mestic and foreign firms may limit the benefits of for some 2000 Mexican manufacturing plants for FDI to poor countries. FDI can be highly growth- the period 1986-90. Controlling for factor costs, enhancing when FDI and domestic investment are output prices, and other variables, Aitken, Hanson, closer substitutes, which is more likely in techno- and Harrison (1994) found that the presence of logically advanced countries than in developing foreign affiliates significantly increases the proba- countries (de Mello 1999). If local firms have in- bility that domestic firms export. To the extent that sufficient capacity to absorb technology and skills growth in Sub-Saharan Africa is reduced by foreign from foreign affiliates, then the poor-country firms investors' lack of information (Collier and Gun- might lose out in the face of competition from ning 1999), exposure to foreign firms may help foreign firms (Kokko 1994; Kokko, Tansini, and 62 THE POOR COUNTRIES' INTERNATIONAL FINANCIAL TRANSACTIONS Zejan 1996; Kathuria 1998; Fry 1992; Agosin and international markets (Harrison 1994; Levin- and Mayer 2000). In addition, resource- or labor- sohn 1993; Tybout 2000; and Hoekman, Kee, and seeking FDI-which is the most common form of Olarreaga 2001). Economies with more active poli- FDI in the poor countries-is likely to generate cies toward fighting monopoly power tend to grow fewer backward or forward linkages for domestic faster, even after controlling for the height of trade enterprises compared to FDI in intermediate or barriers (Hayri and Dutz 1999). capital goods industries-the type more common Research on the impact of foreign entry on in middle-income countries (Ozawa 1992; Porter market concentration in competitive markets is lim- 1990). ited. Several studies have found little evidence of Even when the short-term impact of FDI is anticompetitive practices, including studies in the limited by a poor investment climate, the medium- Republic of Korea after the opening to FD1 in 1998 term impact on growth may be positive. Initially (Yun 2000), in Mexico on the competitive effects of domestic firms may see an erosion of their market foreign acquisitions of domestic firms (Mexico Fed- share due to the entry of foreign firms with supe- eral Commission on Competition 1997), and in the rior technology. Subsequently, however, domestic Czech Republic on the impact of sales of domestic firms may regain market share as they absorb firms to foreigners on market concentration in spillovers of technology and skills through verti- manufacturing (Zemplinerova and Jarolim 2000). cal-backward and forward-linkages of foreign The second area where foreign entry may act firms with domestic enterprises (Marksun and to stifle competition is in natural monopolies that Venables 1997). In a study of 55 poor countries for are subject to economies of scale and have limited the 1980-99 period, a 1 percent increase in FDI potential for cross-border provision of services as ratio to GDP in the current period reduces do- (such as telecommunications and power). For ex- mestic investment as ratio to GDP by 0.8 percent. ample, the privatization of state-owned monopo- However, a 1 percent increase in the FDI-to-GDP lies, without either removing barriers to entry or ratio in the previous period results in 0.7 per cent establishing an effective regulatory framework to increase in the domestic investment ratio of the maintain competitive prices, can lead to a private current period (Kumar and Pradhan 2001). monopoly. Here efforts to maintain efficient mar- kets are more difficult than in competitive markets such as manufacturing, as poor countries often lack the institutional capacity required to effectively reg- Effective competition policies ulate natural monopolies. Thus building adequate are critical rules and institutions to regulate natural monopo- In the absence of effective competition policies, lies may be necessary before privatization. How- FDI also can have a negative impact on the do- ever, once the decision is made to privatize, fear of mestic economy by establishing a local monopoly natural monopolies is not a reason to bar foreign and reducing production to maintain high prices, participation in bidding for privatized firms. thus generating rents for foreign investors. There are two types of situations where firms might be FDI in the mining sector has risen with able to keep prices higher than competitive levels policy reform over a considerable length of time. The first is in The investment climate is not the only determinant competitive markets in small economies where the of the allocation of FDI among the poor countries. government maintains barriers to entry, for exam- Some countries receive significant levels of FD1 sim- ple through high trade barriers or by limiting for- ply because they have natural resources that are not eign entry to particular firms. Here the obvious widely available. The rents associated with the ex- remedy is to reduce trade barriers and establish an ploitation of these resources may be so high as to open regime for FDI. As many of the poor coun- compensate for weaknesses in the overall investment tries have small markets that could be dominated climate. In some cases, investment in natural re- by a few firms, ensuring low barriers to entry is a source sectors can be isolated by imposing special high priority. Opening the economy to import com- regulations, building dedicated infrastructure, or petition tends to lower domestic market concentra- even providing special security in regions affected by tion and reduce price differentials between the local conflict. Nevertheless, with improvements in the in- 63 GLOBAL DEVELOPMENT FINANCE vestment climate in non-natural-resource-exporting countries and the increase in privatization programs, Figure 3.3 Foreign direct investment in the share of oil- and mineral-exporting countries in mining exploration and government the poor countries' FDI flows fell from almost 50 policies percent in 1991 to 20 percent in 1997. Millions of dollars Even in mineral-exporting countries, the qual- ity of the investment climate is an important deter- 250 minant of access to FDI. Global surveys indicate 2 that efficient and stable policies, liberal and trans- parent mining legislation, and accountable and 150 nondiscriminatory tax regimes play a key role in the international mining companies' investment 100 decision making, second only to geological condi- A tions (Naito and others 1998; Clark and Naito 50 1 1997; Otto 1992; Johnson 1990). According to a 0 0 A 1997 survey of 35 countries, long-term success in attracting FDI in mining exploration depends on -50 the quality of the legal, fiscal, and institutional 0 0.2 04 06 0.8 1 framework, in addition to the existence of mining Government reform index (0-1) resources and a favorable geographic location. Note: Triales represent poor countries, while circles Eigh recivedthe igh- represenr or er developing countries. Eight of the 10 countries that receivedSource: Naito, Remy, and van der Veen 2001. est FDI in exploration in 1997 had better-than- average policies, as measured by an index of re- forms in the mining sector (Naito and Remy 2001).17 One major obstacle facing the poor coun- counting for over 40 percent of total exports in tries in increasing minerals production is the poor 1999. Mining sector reform has typically addressed quality of policies in many countries. Of the 13 the establishment of an appropriate legal frame- poor countries in the survey, 10 scored less than work for private sector activities, including the fiscal 0.4 on the reform index (indicating worse-than- regime; modernization of government institutional average policies) and only three scored more than arrangements in the mining sector; public enterprise 0.7. In middle-income countries, by contrast, 8 reform and privatization; and establishment of a scored below 0.5 and 13 above (figure 3.3). sound environmental management system. Nevertheless, some poor countries have under- taken significant reforms of their mining sectors during the 1990s in order to attract foreign invest- ment in mineral resource development (World Bank The participation of foreign banks in 1992 and 1996; Otto 1995; Smith and Naito 1998; poor countries' financial systems Naito, Remy, and Williams 2001). According to re- cent forecasts by World Bank staff, some countries Foreign bank presence in the poor counties that have launched substantial reform programs are increased in the 1990s- expected to achieve significant increases in explo- In addition to capital flows, poor countries are ration investment and-subsequently-increases in tied to the international financial system through the value of the minerals produced and exported foreign banks. During the 1990s, the liberalization (table 3.5)." For example, Mali had historically at- of financial markets in combination with rapid tracted very little foreign investment in mining. In trade growth (which increased banks' ties with the 1990s the country undertook a reform of the exporters from developing countries) spurred the rules governing mining and strengthened govern- global expansion of banks. Cross-border mergers ment oversight and service institutions. As a result, and takeovers of banks rose from 320 over the new investment started to flow in, leading to two course of the 1980s to about 2,000 in the 1990s. new operating mines, and gold has become the The middle-income countries of Latin America and largest contributor to Mali's export earnings, ac- East Asia and the transition economies experienced 64 THE POOR COUNTRIES' INTERNATIONAL FINANCIAL TRANSACTIONS Table 3.5 Mining sector performance in three bank activity, down from almost half in 1995. countries, before and after reforms Foreign banks represent 38 percent of the total (millions of dollars) number of banks in the poor countries, up from EXploration Production 13 percent in 1995 (figure 3.4). Foreign banks' as- eIVT A, i- A,t, , sets account for more than 40 percent of total bank assets inthe poor countries, twice as high as Mal l . 42 1 in 1995. It is possible, however, that the sizeable Tanzarna 1 W'.'sii losses incurred by foreign banks in the Argentine n.a. ot aplicble.crisis may discourage a continued expansion of n.a. Not applicable. Sources: Naito, Remy, and van der Veen 2001 and sources cited foreign banks in developing countries, at least in therein. Staff projections based on ongoing projects and price the near term.21 forecasts. forecsts.Some poor countries have had significant for- eign bank presence for a long time (beginning with colonial domination of local banking systems), a rapid increase in the number of foreign banks.19 and colonial ties remain an important determinant These recipients accounted for the biggest share of of the home country of foreign banks. U.K. banks banks going to the developing world. However, account for about one-third of all foreign bank the poor countries have also seen a substantial rise capital in English-speaking Africa, and French in foreign bank entry, as the failure of state- banks enjoy a similar presence in French-speaking directed financial systems led to the privatization Africa. In low-income transition economies, the of many financial institutions and the removal of home countries of the foreign banks are more di- obstacles to the establishment of new banks in many verse, reflecting weaker cultural or colonial ties, countries. For example, in Africa cross-border mer- although geographic proximity is an important gers between financial institutions in the 1990s determinant of foreign bank presence. For exam- surged to 96, up from only seven in the 1980s pIe, Turkish banks are important in a number of (Buch and Delong 2001).20 In 2000 only 15 of the Central Asian countries, Arab banks are present 58 low-income countries had no reported foreign in the Republic of Yemen and Pakistan, and banks Figure 3.4a Foreign bank presence in poor Figure 3.4b Foreign bank presence in countries Africa Average across countnies (percent) Foreign bank assets as a percentage of total bank assets 45 70 40 Foreign banks as a (figur 3 Fe bn a percentage of total 60 l i d Foreign bank assetsosse0 incurredbyforeignbank 0min percentage of total h s ia f bank assets 25 _ 40 20 ______2 15_ _ _ 10~ .~ 5 -. 1995 2000 Franco Luso Anglo- Asia and Transition phone phone phone Pacitic economies Atrica Atrica Atrica Note: Data include only low-income countries that allow foreign bank presence and have not witnessed open conflict from 199d rhrough 2000. Source: World Bank staff calculations based on Bankscope. 65 GLOBAL DEVELOPMENT FINANCE from middle-income East Asian countries have es- the weak regulatory framework and the frequent tablished subsidiaries in low-income East Asia. policy reversals in the financial sector-including nationalizations of foreign banks-increase the reg- -but regulatory barriers limit opportunities ulatory risk perceived by investors, while the effec- Despite the rise in the presence of foreign banks in tive subsidy to loss-making state banks distorts many poor countries, regulatory barriers and the competition and creates an additional entry barrier. limited opportunities in poor countries' financial systems continue to constrain foreign bank partici- Foreign bank presence is associated with pation. Regulatory barriers are higher in poor higher efficiency of banking systems in the countries than in other developing countries. On an poor countries index that ranges from 0 (closed) to 1 (fully open), The presence of foreign banks is associated with middle- and high-income countries scored, on aver- improvements in the efficiency of banking systems age, 0.77-well above the average (0.54) for all in the poor countries. Increased competition from countries.22 The main determinants of differences foreign banks may reduce intermediation costs by in commitments made to the World Trade Organi- eroding excess profits that domestic banks can zation concerning the liberalization of financial ser- enjoy due to the small size of the financial systems vices were found to be income level, openness to of many poor countries (see World Bank 2001b). trade, and the depth and competitiveness of the fi- In poor countries where foreign bank presence is nancial sector (Qian 2000; Sorsa 1997). On these greater than average, financial intermediation indicators, poor countries generally score worse costs tend to be lower, as reflected in domestic than middle-income countries. Many poor coun- banks' lower net margins and noninterest income. tries also have limited scope for the provision of fi- At the same time, domestic banks' overhead costs nancial services, owing to the small scale of trading, are lower in countries with substantial foreign the low level of savings, and competition from tra- bank presence, perhaps indicating improved prac- ditional and informal methods of savings collection tices learned from the foreign banks. On balance, (such as rotating savings and credit associations). domestic banks' pretax profitability in high-foreign- The high cost of doing business-despite low entry markets is much lower than in markets with wages-is an additional obstacle, reflecting poor low foreign bank presence (figure 3.5). business infrastructure, and greater difficulties in Differences in domestic bank performance evaluating loans in low-income countries. Finally, across markets with varying levels of foreign bank entry are also likely to reflect other factors, apart from the presence of foreign banks-for example, differences in macroeconomic conditions that af- FIgure 3.5 Effect of greater foreign bank presence on fect bank profitability. Taking into account differ- intermediation costs and domestic bank profitability ences in country circumstances and the financial Percentage of domestic bank assets (average, 1995-2000) characteristics of individual banks, econometric 7 results confirm that stronger foreign bank pres- 6 E Low foreign ence is associated with significantly lower domes- bank entry tic bank net interest margins, non interest income, 5 High foreign and overhead costs (see annex 3.1). The net im- 4 - bank entry 2 2 inadmesnc bakcprfitailiy,vatiroontrllinffo the influence of other factors This decline is a 2 partial influence, which may be offset in the long term to the extent that foreign bank entry is asso- ciated with lower financial intermediation costs, awhich could improve credit provision to the pri- Nei Noninterest Overhead Loanloss Pretax vate sector and thus foster higher growth and margin income provision proit bank profitability (Levine 1996). Source: World Bank. Foreign bank entry can help improve the qual- ity of domestic bank staff by training staff that 66 THE POOR COUNTRIES' INTERNATIONAL FINANCIAL TRANSACTIONS then move to domestic banks. For example, countries with higher foreign bank presence. The Citibank is said to have trained an estimated 5,000 benefits perceived by SMEs may reflect, first, the bankers in developing countries. In Pakistan, the lower interest margins spurred by foreign bank government hired personnel from Citibank, Bank entry, which may help expand the amount of lend- of America, Soci6t6 G6n6rale, and ABN-AMRO to ing to SMEs even if the share of lending to them de- help rehabilitate its national commercial banks, dines. Second, foreign bank competition for large starting in 1997. French and British banks that customers may displace some domestic banks, fore- have long been active in Africa have also con- ing them to more actively seek new market niches. tributed to training of banking personnel there. This could potentially improve credit access for Foreign banks also can facilitate the provision of small borrowers in the medium term. On the certain financial services, such as international whole, based on a sample of 59 countries, Barth, syndications, letters of credit confirmations for ex- Caprio, and Levine (2001b) concluded that limita- ports to third countries, treasury products for tions on foreign bank entry (captured by a cross- commodity hedgers, depositary receipts, and inter- country comparable survey of national regulatory national mergers and acquisitions possibilities for agencies) tend to be associated with a smaller share local corporate customers. of bank credit to the private sector in GDP. Foreign banks have also contributed to the Greater foreign bank presence may also help soundness of domestic banking systems by partici- attract foreign bank lending to poor countries, al- pating in the privatization of failed state banks. though the evidence is limited. Increased foreign For example, the sale of Tanzania's National Bank bank presence can facilitate project selection and of Commerce (NBC) to ABSA, a South African screening of borrowers, thus improving foreign bank, led to a sharp acceleration in the pace of re- banks' access to information, a critical input to structuring and in loan recovery efforts. When lending decisions. Poor countries with high foreign ABSA took over NBC in March 2001 it launched bank presence attracted nearly 50 percent more in- an aggressive loan recovery effort that generated ternational bank lending as a share of their GDP immediate results. Whereas previously NBC had than countries with no foreign banks (figure 3.6). been continually thwarted in its collection efforts by court injunctions and other avoidance tactics, ABSA successfully overcame many of these obsta- Figure 3.6 Effect of greater foreign bank cles, thereby establishing its credibility and elicit- presence on international bank lending ing more constructive behavior from borrowers.24 to poor countries Despite the improvements in efficiency brought Percentage of debtor countres'GDP (average across about by greater foreign bank penetration, policy- countries, 1995-2000) makers in developing countries are often concerned 10 that access to credit may be impaired for some sectors of the economy-in particular small and 8 medium enterprises (SMEs)-because foreign banks tend to serve primarily large customers com- 6 pared with domestic banks. However, evidence from a survey of over 4,000 enterprises in 38 devel- 4 oping and transition economies-including 8 poor countries-suggests that, though large enterprises seem to take better advantage of foreign bank pres- ence, benefits appear to also accrue to SMEs No foreign a Below 25% Above 25%a (Clarke, Cull, and Soledad Martinez Peria 2001). bank presence In countries with high foreign bank penetration, Note: Total claims of BIS reporting banks on poor SMEs tended to rate interest rate costs and access countries. to long-term loans as lesser constraints than in a. Foreign bank assets as a percentage of total bank assets in poor countries, 1995-2000 (average). countries with low foreign bank entry. Medium- Source: World Bank, based on Batk for International size enterprises also appear to finance a larger share Settlements data. of investment through commercial bank loans in 67 GLOBAL DEVELOPMENT FINANCE Of course, this relationship may be due to other vision, a history of high nonperforming loans, and factors. For example, countries with high foreign government pressure for unprofitable lending to bank presence may also have better investment cli- loss-making state enterprises. On the other hand, mates, which would explain the higher level of for- foreign bank presence may have a positive impact eign loans. Countries with low foreign bank pres- on financial stability, because it helps introduce ence may also restrict private borrowing from better risk management practices, while foreign abroad, thus limiting the amount of outstanding banks are likely to be better supervised by home international bank claims. country regulators. One approach to investigating the impact of Foreign bank entry does not appear to be foreign banks on stability is to examine whether associated with greater risk taking by the domestic banks' portfolio and performance domestic banks- characteristics that have been shown to affect the While the fall in domestic bank profitability that is chances of a financial crisis differ significantly in associated with foreign bank entry may signal re- "low" and "high" foreign bank entry environments duced financial intermediation costs for bank (Demirgil -Kunt and Detragiache 2000; Goldstein, clients, it may also engender instability: banks that Kaminsky, and Reinhart 2000).25 Analysis suggests see a decline in their franchise value may have that poor-country banking systems with high for- an incentive to take on greater risks (Hellmann, eign bank presence had, on average, a smaller share Murdock, and Stiglitz 2000). Pressure on domestic of nonperforming loans in the late 1990s (figure banks may also increase if foreign banks capture 3.7). Provisions for nonperforming loans are also the most lucrative segments of the market (such higher in countries with large foreign bank pres- as loans to export-oriented manufacturing), thus ence. While domestic banks in low-entry countries leaving domestic banks more exposed to the low- provision less than 100 percent for each nonper- end, less profitable segments. This problem could forming loan, banks in high-entry markets provi- be particularly severe in many poor countries, sion, on average, 150 percent. To be sure, lower where domestic banks may lack the expertise to nonperforming loans and better provisioning may compete effectively with foreign banks and domes- partly reflect better prudential requirements and tic banks may already be weakened by poor super- supervision in countries that are more attractive to foreign banks. On balance, domestic banks in poor countries with high foreign bank presence do not Figure 3.7 Effect of greater foreign bank presence on nonperforming loans nonprforing oans-but a banking system that is more Nonperforming loans as a percentage Loan loss provisions as a of domestic bank assets percentage of nonperforming loans competitive and open to foreign entry (average, 1995-2000) (average, 1995-2000) can increase risks 7 180 While on average foreign bank presence is not as- 6 160 sociated with collateral damage to domestic banks, 140 on occasion foreign banks have increased domestic 5 10 financial instability by pulling out of host countries -120 4 100 or by contagion from problems in the home coun- 80 try. A foreign bank affiliate may be forced to cut 3 80 back on its local asset portfolio, in response to a 2 60deterioration of the parent bank's balance sheet. 40 The impact of a decline in lending by a foreign 1 20 bank may be particularly great in poor countries, 0 0 where the number of banks is limited and foreign Low foreign bank entry High foreign bank entry banks are often major players. For example, Kent- Nonperforming loans - Loan loss provisions to nonperforming bank of Turkey, which had purchased the National Commercial Bank of Albania in 1999 (with 60 Source: World Bank; Claessens and Lee 2001. percent market share in deposits and loans), had to be taken over by the Turkish Deposit Insurance 68 THE POOR COUNTRIES' INTERNATIONAL FINANCIAL TRANSACTIONS Fund. Fears over instability were calmed, however, lished in 1990 has been an important step toward when the Turkish Fund lent $10 million to the Al- ensuring uniform and more efficient supervision of banian bank. In Romania, rumors that the Turkish financial institutions in the eight member countries shareholder in Banco Turco (24 percent market of the West African Economic and Monetary Union share) was directing the funds of Banco Turco Ro- (IMF 2001a). mano back to Turkey led to a run on the bank in 2000. The run was stopped when the Turkish gov- ernment persuaded Vakifbank of Turkey, a bank partially owned by the government, to support the Capital outflows bank. The sale of Uganda Commercial bank, the ost poor countries have de facto open finan- main state bank, to a Malaysian industrial and real M cial systems, in the sense that residents are estate company had to be cancelled when the par- able to place assets abroad-although these trans- ent bank got into difficulties. actions, referred to as capital outflows, are not al- These events point to the potential transmis- ways legal. Since most capital outflows are not sion of instability from foreign banks, particularly recorded, they are measured by inference, as the those from countries subject to substantial insta- difference between recorded capital inflows and bility and without strong regulation and supervi- the sum of the current account deficit and in- sion. Diversification of the home countries of for- creases in international reserves. This measurement eign banks is particularly important to reduce is inevitably imprecise.27 Despite these difficulties, exposure to financial contagion. However, to min- there is no doubt that outflows are large relative to imize risks of contagion, the host country regula- economic activity in many, if not most, of the poor tors also should be careful in screening entrants on countries, which has important implications for the basis of two criteria: the quality of the foreign the volume of domestic investment and the con- bank's domestic supervisory framework and the duct of macroeconomic policy. This section dis- foreign bank's reputational risk exposure (to pro- cusses the determinants of capital outflows and tect its reputation, a large international banking their implications for the domestic economies of group is more likely to recapitalize a subsidiary the poor countries. than to let it fail). With increased presence of foreign banks, Capital outflows are high relative to domestic maintaining effective cross-border supervision has savings for the poor countries become important to reduce the risk of conta- The poor countries have experienced substantial gion.6 However, enforcing effective cross-border capital outflows over the past two decades. Never- supervision raises difficult policy challenges for theless, capital outflows remain smaller than in- poor countries, as it requires a regular exchange of flows, and in most poor countries net external fi- high-quality financial information between the nance makes a positive contribution to domestic home and host country regulators. The host super- investment. Cumulated outflows totaled $62 bil- visors should also be ready to permit on-site in- lion, equivalent to 17 percent of GDP, almost 12 spections by the home country supervisors. Many percent of cumulated savings for 1980-99, nearly poor countries lack the resources and capabilities a fifth of cumulated official flows during 1980-99, to effectively align their prudential regulation with and nearly two-and-a-half times international re- best practice and comply with cross-border super- serves in 1999 (table 3.6).28 Capital outflows from vision guidelines. Moreover, almost all poor coun- the poor countries were larger relative to domestic tries have relatively small financial systems, so that savings and reserves, and only slightly smaller rel- the fixed cost of establishing effective supervision ative to GDP, than outflows from other developing can be high. Regional cooperation among poor countries (which generally are viewed as more fi- countries could help, by upgrading and harmoniz- nancially integrated with the rest of the world). ing standards of prudential regulation in financial Capital outflows are extremely volatile, how- services, pooling resources and expertise, and inten- ever and these aggregate data conceal considerable sifying information exchange. For example, despite variation over time and across countries. Since the need to further reinforce the regulatory frame- 1985, capital outflows from the poor countries work, the West African Banking Commission estab- have varied from less than 3 percent of GDP to just 69 GLOBAL DEVELOPMENT FINANCE Table 3.6 Cumulated outflows during 1980-99 Poor countries 62 17 11.5 8.1 19 242 Other developing countries 1,182 20 6.5 6.6 278, 175 a. This ratio is high because aid flows to middle-income countries are very small. Sources: IMF Balance of Payments; World Bank staff estimates. over zero (meaning capital repatriation) (figure Table 3.7 Volatility of capital flows, 1990-99 3.8). Moreover, the cross-sectional standard devia- OutflowsasshareofGDP tion of the ratio of capital outflows to GDP is (coefficent of variation) greater than the average over the period. Another Poor countries 3.6 way of gauging cross-sectional variability is that Other developing capital outflows averaged $8 billion a year during 1995-99, but 20 countries have outflows that total Note: For each country group, the mean is estimated by dividing the over $10 billion, while 6 countries account for sum of flows by the sum of GDP for each year, and tben taking the mean over the decade. Standard deviation is computed using tbe an- more than $2 billion of reverse outflows (repatria- nual averages for the decade. Coefficient of variation is the ratio of tion of residents' capital). standard deviation to the mean. Resource flows include short-term debt flows and are taken from GDE. Outflows are estimated using Indeed, capital outflows from the poor coun- IMF BOR tries are more volatile than outflows from the Source: World Bank staff estimates. middle-income countries, while inflows are less volatile (presumably because the poor countries re- macroeconomic stabilization as the middle-income ceive little of the more volatile capital market flows) countries. Moreover, at lower levels of income, (table 3.7). This highlights an important point: volatility is likely to be more costly in terms of wel- many poor countries face the same issues surround- fare (a decline in income can push more people to ing capital flows volatility and the implications for subsistence levels or below). Poor countries typi- cally lack the range of instruments (for example, an efficient government bond market) available to Figue 38 Cpitl ouf lws rommiddle-income countries to deal with macroeco- Figure 3.8 Capital outflows from nomic volatility, and they are also more Subject to developing countries, 1985-99 volatility from the external sector due to their de- Percentage of GDP Percntae ofGDPpendence on primary commodities. The average 1 volatility of the poor countries' terms of trade (as A measured by the coefficient of variation) in 1990- Porcunre 99 was about 40 percent higher than in other de- Poorveloping countries. Thus the poor countries face -1 higher levels of volatility, volatility is more costly for them, and they are less equipped to deal with it, -2V compared with middle-income countries. -3 A poor investment climate encourages Middle-income countries capital outflows -4-7r7-- I I i i iiii it The quality of the investment climate in the poor 1985 1987 1989 1991 1993 1995 1997 1999 countries is the main determinant of the level of Source: IMF Balance of Payments (BOP); World Bank capital outflows. War and civil conflict, corrup- Stion, macroeconomic instability, uncertainty over property rights, high tax rates, weak governmental 70 THE POOR COUNTRIES' INTERNATIONAL FINANCIAL TRAN S ACTIONS institutions, financial sector repression, and un- versals of outflows were associated with macro- necessary constraints on private sector economic economic stabilization and structural adjustment activity encourage outflows by limiting the oppor- programs. A high debt-to-GDP ratio raises the risk tunities for profitable domestic investment (box of future taxation, and also the risk of default on 3.2) and increasing the risk of confiscation or cap- sovereign liabilities to residents. Cumulative capi- ital losses on funds held domestically (Tornell and tal outflows averaged 39 percent of GDP in poor Velasco 1992).29 Several authors have mentioned countries with higher than average debt-to-GDP that capital flight is driven by the desire to safe- ratios, but only 5 percent of GDP in countries with guard incomes derived from corruption and crime lower than average debt ratios. (see Varman-Schneider 1995 in the case of India, In some countries, preferential treatment of and Loungani and Mauro 2000 in the case of the foreign capital versus domestic capital also boosted Russian Federation). In poor countries with better outflows in the form of round tripping (see exam- than average economic policies (as measured by pie of round-tripping in China in chapter 2). Pref- the Bank's Country Policy Performance Rating), erential treatment for foreigners may include tax the stock of capital outflows totaled only 6 percent breaks, preferential access to prime land and other of GDP, compared with 30 percent of GDP in inputs, and exemption from exchange controls countries with worse than average policies (table faced by residents (Dooley 1986; Khan and UI- 3.8). Sheets (1996) found that inflation, budget Haque 1985; Eaton 1987; Ize and Ortiz 1987).31 deficits, and low interest rates were associated Such discriminatory treatment of resident capital with increased capital flight. Schineller (1997, relative to nonresident capital may encourage in- 1999) also found that the fiscal deficit was an im- vestors to deposit their wealth in a foreign bank, portant determinant of capital outflows, and re- and then raise debt financing from the same bank for their domestic investments (Lessard and Wil- liamson 1987). Table 3.8 Cumulated outflows as a share of GDP, just as a poor investment climate encourages 1999 outflows, improvements in the investment climate (percent) can encourage capital repatriation. Ajayi (1997) Pnor OIIVF~.. ~ describes how improvements in macroeconomic stability and better governance encouraged the re- Investment climate versal of capital flight in C6te d1voire, Central Policy environment, African Republic, Uganda, Ghana, and Kenya dur- GDP growth ighng the 1980s and 1990s. Olopoenia (2000) esti- r- mated that capital flight from Uganda rose during Debt/GDP periods of political instability (1971-74, 1976-79, M2/GDP and 1981-87), but there was a "reflow" of flight ccapital following a return to peace and economic Trade/GDP liberalization (including exchange rate unification and lifting of exchange controls) during the 1990s. Incohe effects In Kenya, Tanzania, and Uganda, high Treasury Performianc aing.m bill rates offered by governments have attracted Gmi i funds from returning emigrants (Bhinda, Griffith- Jones, and Martin 1999). Tax amnesty programs have been used as another means of attracting in- resident capital flows (see Ng'eno 2000 for the example of Kenya). Exchange premium Pi- However, such programs can only provide one-off, short-term effects (Das-Gupta and Mookherjee Note: Outflows cumulated over the 1980-99 period High and ow 1995), and are effective only if accompanied by usually refer to above and below median of the concerned variable, measures to reduce the distortions that encouraged The numbers reported are sum of cumulated outflows for countries above median (say) divided by sum of GDP of the same countries, outflows in the first place. If repeated, tax amnesty a. Policy environment is neasured by World Banks country policy programs increase incentives for evasion as tax- perforieance rating. Source: World Bank staff estimates, payers wait for the next amnesty. 71 G LOB AL DE VE LO PM EN T FINA N C E Outflows are also associated with increased wealth that increased the demand for international wealth and globalization portfolio diversification (box 3.3). By contrast, the Capital outflows do not always signal a poor in- poor countries with higher than average per capita vestment climate. In many middie-income coun- incomes (for the poor-country group) experienced tries, the rise in capital outflows before the East smaller outflows (table 3.8), perhaps because Asian crisis appeared to be tied to increases in wealth levels, while higher than those of the aver- Box 3.3 0apaa esteews r t countries C apital outfluw from the middle-inc*ne eeun 197-98 " and the R7sian F0deraj ion .signi:- have a diffeentto than ouflows r th an poon of aia w msy als erse.ot r. poor countries, and t ominant motivations a .e dif- t- in. For eame, he ex -erie o he eråss may ferent as weil. Mani dineome eoun.ies beeame al hav ene aged mesi yestors to r benefit more integrated into a ecenomy over te courts from epi it and m et gua.antees en eFeign d -b.. of the 1990s. In th f he decade, th official Te differen mtvat.ns o pial outfo fen-m data showed sharise in private aital inflows, but t id -ineF -eun.e av mean that som of this was substani0 fse byan rn -rease flows, asincrea véed i traderansaeriens boose peer unties de nt hel. o exampe mr d the desire for 'rsification (Gerdo d 1neem* enrs wth -etr pelie an with higher Levine 1988). ne er of-apital.outflows from per eai ineem ha eper.ined almost h same middle-income Cuntri h for o 'reig diret 1eel o eu ulativ ea..ital . té. as middleineme or portfoho inmestment.see 7 gures). Thus, in he ey cis vl r p tr lal fto mys Trhe 1990s, growing apil u ws from many iddle- g eavi. ' e i ome - ts income countri ensistn with eernomic oges ful iddle-ie_m eonri s a faeiLated gå:. while in the p~ä @l ä) s apital outflows ofen re- ile st ao r e k p flected a poor c4mt~ for investment an slow growth. In e ds ( the second half,o, 19-9s capital outlows by reidenQts· eeunrs w d' oGD increased from oufies affeeted b.y erises, for exame n'ess to t-ad, a aen- ,ne -ua ha e delily Mexico in 19 I nesia Korea, and Taad in. EU.es. CodmP4sto-O 6 mulated uflw fro -Noijh DsmrL6Gim9m m idd ncom onIsdri ng 1980-99 . -E. Oßffaum ílg) FD, -.4 3 Eur'.f~* N rr. - . n r h r . 72 THE POOR COUNTRIES' INTERNATIONAL FINANCIAL TRANSACTIONS age poor country, had not reached levels where substantial international diversification was neces- Figure 3.9 Cumulated outflows and minerals exports sary. Higher trade openness may also encourage Outflows (1980-99) asa percentage of GDP outflows as residents have more contact with in- ternational markets, there is a rising incentive to hold foreign exchange as a hedge against changes 60n in the exchange rate, and the scope for misinvoic- O0 ing of exports and imports increases. Capital out- flows from poor countries with higher than aver- 40 age ratios of trade to GDP equaled 41 percent of 30 GDP, compared with 8 percent in countries with lower than average trade-to-GDP ratios.32 20 Income inequality also can have an important 10 impact on outflows. Cumulated outflows from poor countries with high inequality, as measured by 0 Less than 20%a Between 20% and 40%a Greater than the Gini coefficient, averaged 50 percent of GDP, a. percenrage of ores and metals exports in merchandise exports compared with 7 percent for poor countries with Source: World Bank, World Development Indicators, rarious Years; World Bank low inequality. A high concentration of wealth may staff mean that some residents have large individual portfolios that make them more likely to diversify their assets and more able to pay the implicit and growth, capital outflows, capital inflows, the real explicit transaction costs associated with capital exchange rate, and fiscal deficits) and controls for outflows. High income inequality may also be as- the role of other influences (Such as degree of in- sociated with greater sociopolitical risks, which equality and structure of trade) can improve our would in turn encourage outflows. The size of out- understanding of the forces at work. This analysis flows is positively related to large mineral resources uses panel vector autoregression (explained in (such as oil, gold, and diamonds [figure 3.9]), and more detail in annex 3.1), in which each of the en- countries with large natural resource endowments dogenous variables is related to lagged values of also tend to have higher income inequality (Goreux the other endogenous variables. 2001). For example, the largest source of capital The results for all developing Countries idi- outflows from Sub-Saharan Africa is Nigeria, cate a two-way relationship between capital out- where outflows seem to be highly correlated with flows and the government's track record in foster- oil exports (Ajayi 2000). ing growth and maintaining economic stability. It is difficult to determine whether simple Higher growth rates are associated with reduced comparisons of the investment climate and capital capital outflows in the next period, while higher outflows, as shown in table 3.8, reflect causality capital outflows appear to contribute to reduced (and in which direction) or the influence of some growth rates in the next period. Similarly, a higher third factor that determines both indicators. For fiscal surplus is associated with smaller capital out- example, large capital outflows may be associated flows in the next period. Capital outflows are also with high debt ratios because residents place funds significantly related to capital inflows, which may abroad in order to escape the potential for higher either reflect round-tripping or the tendency for fi- taxes to service the debt. Alternatively, high capital nancially integrated economies to engage in both outflows may reduce growth, thus increasing debt- external borrowing and lending. Thus there is to-GDP ratios. Or, high levels of corruption may strong support for the existence of virtuous (and vi- mean that large inflows of official finance end up cious) cycles, in which, for example, a fall in capital in private hands and are then transferred abroad- outflows increases the domestic resources available thus increasing both external public debt and pri- for growth, which in turn lowers outflows. The vate outflows. An analysis of the relationship be- qualitative results for poor countries follow a simi- tween capital outflows and other macroeconomic lar pattern, although the statistical significance of variables that takes into account the mutual inter- the coefficients is found to be weaker than the re- actions among endogenous variables (such as suIts for all developing countries.33 73 GLOBAL DEVELOPMENT FINANCE Most poor countries have controls on capital By contrast, the share of poor countries re- account transactions- porting capital account restrictions has remained While many poor countries have achieved a signifi- at about 90 percent since the mid-1970s, with a cant reduction in restrictions on current account slight rise during the mid-1980s and a slight de- transactions since the 1980s, most continue to im- dine in the mid-1990s when a few countries liber- pose restrictions on capital account transactions. alized capital account transactions (figure 3.10). In Four indicators that have often been used to mea- addition, there has been almost no change in the sure trends in foreign exchange restrictions over share of poor countries reporting various capital time are: (a) existence of multiple exchange rates, account restrictions in the more detailed format (b) export earnings surrender requirements, (c) con- used since 1995. While it is impossible to make trols on current account transactions, and (d) con- a precise comparison of the late 1990s with ear- trols on capital account transactions.34 The first ler years, the broad conclusion is that most poor two of these indicators are available over a long countries have maintained capital account restric- time series through the most recent year, while the tions over the course of the last 30 years. The latter two are available on a comparable basis only share of other developing countries reporting capi- through 1995.3S tal account restrictions also has changed little The poor countries have made progress in re- since the early 1970s, but it remains well below ducing current account restrictions. The share of the share of poor countries imposing capital ac- reporting poor countries that imposed current ac- count restrictions. count restrictions fell from 75 percent in 1985 to 44 percent in 1995. It appears that the trend to- -but capital controls are porous ward liberalization of current account restrictions Controls often have only a limited impact on capi- continued in the second half of the 1990s: the share tal outflows in the context of a weak investment of reporting poor countries that require exporters climate, where domestic investment opportunities to surrender foreign exchange earnings to the gov- are limited and fears of confiscation or reduction ernment dropped from 64 percent in 1995 to 52 in the value of assets give residents considerable percent in 2000. Moreover, the share of reporting incentive to put their money abroad. Controls poor countries with multiple exchange rates fell have had some success in the middle-income coun- from 29 percent in 1995 to only 10 percent in tries when they are limited in time or in purpose 2000.36 (see box 3.4). But they have had particularly lim- Figure 3.10 Capital account restrictions Percent 1001 90 80 70 60 Poor (including transition) Nonpoor (including transition) 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 Source: IMF Exchange Arrangements and Exchange Restrictions. 74 THE POOR COUNTRIES' INTERNATIONAL FINANCIAL TRANSACTIONS Box 3.4 Narrowly focused capital controls in emerging markets T hailand'; and Mala.%sia's experiences with capital fon- actions including trade flo%v. FDI, and portfolio trols on outflos, and Chile's experience %iith capital The initial control were modified on -ci eral occa;ions, in- controls on inflows, pro% ido -Lme ct idence that controls cluding both lo,sening in response ro changing econoic can be elfectie if narrow I focused and adusted in re- conditions as o eI as tightening to close loopholes that the spOnse to artempts at circumcnotionprv ae sector had begin to exploit Measured against the In 1991 the Chilcan governmenr imposed controls on objecti% o -purihing';speClation by limiting ifttchre infloss to lengthen the maturinr of inllows and increase liqHdiV. the controls %%ere at least partiill% sucleSSf0l, as the c.apacit of the central bank to conduct an independent thev contributed to . %%ide and persistent gap betvcen on- moneiar policy The controls consisted of unremuncrated ;hor and atfhc sisap rates OINIF 20ui. reser;c requliremerts (i1RRs) that (initiallk i mandated that Capital controls %%ere adopted in Alaiwia in Septeni- 20 percent -f the deposit remain iII a non-interest-piving ber 1998, when tie e,chnnge rate had already depreciated act'iint for the duration of Ehr credit. "Nlinimum stay" sharpl%, making sizablc further 000ittLinlikely. More- rejuirmenr-, of three ears %erc placed ont FDI and port- 0%er, as in Thailand, the a s control %,crt: eIecrive folio flr% s \\ hile ubsequent chinges were made in the n naturL, designed to curtail if not elimiiitc I the possibil- spetific' o the controls (change- in the URR pertenmage, i of spe,ulation aganst tie riiggit \% hile IL.a% rng ordinar rediCtions ri the minimuni sta, extension - or exeniptions trade and FDI 0 ,naftected. The contol. .cre in-nedi- froi coveraize, the underl ing restrictions remained in atelv effecti. The prohibition on iieraccouit tr.nsictions plale untd I 99. The control, cheited a rug-of-war be- irrualls lialtedoffhire riiggit trading, while the ind.l. tween the authoriti s and the private sector. in wthich per- r' 12-month holding period on pirtolio repatriation odi, LsLies' b% the priL ate s-ccir in diluting the eltIs - shut dcn But in retrospect it is alsol cler that- ne,5 ot the controls led to efforts bs government ti cose the Nlalayian controls were imposed after the virst Of the. the loopholes. E. idence 'uggest that there was some criSi tad pased. so that their in.ior oitribution '%> one lengthening of the maturity t inflo%s with little impact c. safeguarding an ue L cC rathcr than li-- on the aggregate %alue of inflos . In addition, domestic ing the direct impact of the crisis itself see also DornbUsch interest rates \%ere marginally -delinked" trom interna- 2001: and Kaplan and Rodrik 20011. The control s% stem tional markets, prem- iding the authorities iw ith an increased relied hea%i on comprehensive regulation and bureau- space for policy maneuver IDe Gregonrio. Edwards, and cratic intervennon, and active adjUstment and fine-tuning L Valdes 20001. The benefits must be balanced off against the controls b% the authorittes occurred 'n response ti pri- the costs, though. which included raising the cost of bor- vate sector etforts to evade the impact iHood 2111). roxt ing for domestic firms (especiall% those withour access What lessons can be drawn from these cxperience, of to international markersi. capital controls? First the success of controls depends in Both Thailand and Malaysia resorted to controls on part on defining a sufficiently narrow oblecte. BOth capital outiflo%s as part of their response to the Asian cri- Malaysia and Thailand had some success in limiting spb.- sis. In Thaulnd. the controls were first adopted early in lation through offshore markets. Second, the control s% the crisis, in an effort to limit offshore speculation against rem must rematn dynamic: the private sector will i itably the baht. The controls were intended to be narro%%. and Strive to minimize or avoid the impact of controls, necessi- did not apply to legitimate commercial and financial trans- tating taminIsErase responses to fine-tune the regulations. ited success in the poor countries, where controls in an effort to bypass foreign exchange surren- are typically imposed over an extended period, so der requirements. Similarly, imports may be that individuals and firms have ample opportunity overinvoiced to gain access to larger amounts to find means of getting around them. of foreign exchange. Residents also may falsify Means of circumventing capital controls in- import letters of credit and customs declara- h rude: tions to bypass exchange controls. Smuggling. Goods may be smuggled and the Trade mnisinvoicing. A portion of the export proceeds deposited in banks. Sometimes, earnings may not be reported to the authorities barter may be arranged for trading Contra- 75 GLOBAL DEVELOPMENT FINANCE band (for example, diamonds for arms in Examples include Chile, Colombia, and Egypt in Sierra Leone [see Goreux 20011). the early 1990s (Schadler and others 1993). Changes in transfer prices and leading and As one motivation for capital outflows is to lagging of intracompany transfers are used guard against a real devaluation of the domestic for shifting funds abroad (Mathieson and currency, several middle-income countries have al- Rojas-Suarez 1993). lowed local deposits denominated in foreign cur- rencies to reduce capital flight and induce nonresi- A common method of effecting fund transfers dent inflows (for example, India, Mexico, Uruguay, in the presence of exchange controls is hawala and Turkey [see Rojas-Suarez 1990]). Moves to- (meaning "trust" in Hindi), also known as hundi ward capital account liberalization such as allow- in Pakistan, or fei ch'ien (literally "flying money") ing foreign currency deposits may reduce distor- in China. In a hawala transaction, a developing- tions and corruption that studies find to be country resident who wants to transfer funds to a associated with capital controls (Edwards 1999; transferee abroad deposits local currency with an Loungani and Mauro 2000), and can increase the agent and obtains a "chit." The agent instructs his supply of capital to help governments manage diffi- colleague in a foreign country to pay an equivalent cult times. In Turkey, for example, worker remit- amount of foreign currency to the transferee upon tances doubled between 1988 and 1989 in re- presentation of the chit (or simply a code). It is be- sponse to such a policy. Remittances also doubled lieved that the net amount outstanding at the end between 1992 and 1994 in India when nonresident of a long period of time is settled through smug- workers were allowed to hold foreign currency de- gling. Thus hawala is not a distinct means of evad- posits onshore. ing capital controls, but rather a means of effect- Some of the poor countries have also moved ing international payments transactions when toward liberalizing controls on inflows. In the desired, with ultimate settlement done by the 1990s liberalization of exchange regulations led means of capital outflows outlined above. This to rapid growth of foreign currency accounts in a method (believed to have originated in China dur- few countries in Sub-Saharan Africa (for example, ing the T'ang dynasty) is fairly common in South Ghana, Tanzania, and Uganda), and a significant Asia, the Middle East, Sub-Saharan Africa, and part of these funds reflected the return of flight Southeast Asia.17 capital (Bhinda, Griffith-Jones, and Martin 1999). Controls on capital outflows not only have According to Stryker (1997), foreign currency de- limited success over the medium term, they may posits held by residents onshore in Ghana increased also discourage capital inflows. Foreigners will be significantly over the early 1990s, to make up a unwilling to invest where there is significant uncer- third of total deposits by the end of 1996. Private tainty regarding their legal ability to repatriate transfers to Uganda increased from $80 million in profits and ultimately liquidate the investment. 1991 to $415 million in 1996, following capital ac- The presence of capital controls, even if they are count liberalization that permitted residents to widely evaded, will create such uncertainty, be- open foreign exchange denominated accounts; de- cause foreigners are typically less knowledgeable posits in such accounts accounted for one-quarter about the feasibility and risks involved in commit- of broad money in Uganda in April 2000 (Kasek- ting technical violations of the law. Also, multina- ende 2000). In Kenya, the legalization of foreign tionals are usually unwilling to undertake illegal currency deposits in the early 1990s in the context transactions because of the harm to their reputa- of high real interest rates attracted large short-term tions and the likelihood of being made an example flows: the level of international reserves rose from if enforcement of controls is tightened in the fu- $81 million at the beginning of the second quarter ture. Conversely, removing capital controls can en- of 1993 to $685 million a year later. courage inflows (Laban and Larrain 1997). Several Liberalization of the capital account, however, countries have eased controls on outflows when can prove costly, especially when combined with faced with large inflows (to limit currency appre- interest rate liberalization in the context of a weak ciation and loss of export competitiveness, see macroeconomic policy environment and underde- Calvo, Leiderman, and Reinhart 1993), but the lib- veloped financial markets. Capital account liberal- eralization actually resulted in increased inflows. ization (including allowing local foreign currency 76 THE POOR COUNTRIES' INTERNATIONAL FINANCIAL TRANSACTIONS accounts) has to be complemented by sound and Martin (1999) found that increased private cap- macroeconomic policy and prudential banking reg- ital inflows contributed to real effective exchange ulations, but poor-country governments often lack rate (REER) appreciation in Tanzania, Uganda, the resources to obtain the information required Zambia, and Zimbabwe during 1990-97.39 The do- for effective supervision, and corporate governance mestic liquidity expansion that resulted from capital and accountability can be weak. If liberalization in- inflows may also have been a factor behind the im- duces a large repatriation of flight capital by resi- prudent lending and borrowing behavior by banks dents, or attracts significant nonresident inflows, in these countries. In Uganda, despite prudent fiscal the currency may appreciate and, at the same time, policy and attempts to supervise banks and regulate domestic liquidity may expand, generating infla- corporate borrowings (the Financial Institutions tionary pressures. Liquidity management in such a Statute of 1993), two banks had to be taken over situation may not be easy, especially since many for restructuring in 1995. The accumulation of poor countries do not have sufficient instruments short-term foreign liabilities was a source of distress of monetary policy to conduct sterilization. (Steril- in these problem banks (Kasekende 2000). In ization may also prove to be very expensive, as in Kenya, nonperforming loans as a share of total the case of Indonesia before the crisis in 1997.) In- loans rose from 20 percent in 1994 to over 30 per- creased dollarization of domestic liabilities through cent in 1997 (Ngugi 2000; Brownbridge 1998)-the allowing foreign currency accounts may also com- resulting banking crisis may have been related to the plicate monetary and exchange rate management." surge in repatriated outflows (from $177 million in Moreover, allowing unrestricted capital flows 1994 to $682 million in 1997). can increase the risks assumed by domestic banks Moreover, most of the poor countries are and corporations, as happened in East Asia before small economies with heavy dependence on pri- the 1997 crisis (Corsetti, Pesenti, and Roubini 1998; mary commodities (and are thus subject to severe Krugman 1998). In the presence of a pegged ex- terms-of-trade shocks, as noted above), and they change rate and relatively high domestic interest have relatively shallow capital markets. A com- rates, capital account liberalization can encourage pletely open capital account could magnify the im- residents to take unhedged foreign currency expo- pact of external shocks. For example, a sharp fall sure (if the pegged exchange rate is expected to be in the price of a major export commodity could maintained, borrowers can take low interest rate lead to large capital outflows in anticipation of a foreign loans and place the funds in high-yielding devaluation, potentially leading to overshooting of domestic accounts). This can result in significant the exchange rate. The same process would occur currency mismatches on banks' balance sheets, with capital controls, but to a lesser degree. In ad- which in turn can lead to huge losses if a fall in con- dition, short-term controls that exempt FDl trans- fidence triggers capital outflows (or if devaluation actions may be an attractive option for poor coun- of the currency is required for any reason) (Eichen- tries that lack market access and hence do not green and others 1999; World Bank 1999a). Even have to take into account the impact of controls in with a floating exchange rate (so that the incentive discouraging portfolio inflows. for unhedged exposures is reduced), sharp changes Thus the poor countries need to move cau- in the exchange rate can introduce considerable tiously toward liberalizing capital account transac- volatility in the balance sheets of banks with large tons. Countries that have already opened the capi- foreign currency exposure. Middle-income coun- tal account, established sustainable macroeconomic tries have suffered very severe consequences from policies, and made the difficult adjustments re- capital account liberalization combined with weak quired to maintain stability in the face of capital in- financial institutions and insufficient supervision. flows (particularly establishment of strong corpo- Poor countries with even greater financial sector rate and financial sector institutions and effective weaknesses could confront serious difficulties with supervision) should not backtrack by imposing con- open capital accounts. trols. Many poor countries continue to confront There is some evidence that the liberalization of weak financial sector institutions and difficult chal- capital inflows in Sub-Saharan African countries lenges in achieving strong governance and sustain- was associated with both macroeconomic and fi- able macroeconomic policies. Liberalizing capital nancial sector difficulties. Bhinda, Griffith-Jones, inflows under such conditions can lead to excessive 77 GLOBAL DEVELOPMENT FINANCE risk taking and exacerbate macroeconomic instabil- variables are measured as a share of total domestic ity. Poor countries need to take into account the de- bank assets. gree of volatility of their economies, and be confi- Apart from foreign bank presence, the regres- dent in the quality of their policies and institutions, sions relate the domestic banks' performance idi- before undertaking the risks involved in capital ac- cators to the financial characteristics of individual count liberalization. banks (such as equity capital and other earning as- sets) and their apparent cost-efficiency (as measured by the overhead expense ratio). The regressions also control for the impact of the macroeconomic envi- Annex 3.1: Econometric analysis of ronment on bank performance. Macroeconomic foreign bank participation factors that may affect interest margins, profitabil- T he effects of foreign bank presence on the op- ity, and provisioning for bad loans include the ate eration of domestic banks can be more com- of GDP growth, inflation, and the real interest rate. pletely examined by formal econometric evidence. In addition to the observed share of foreign banks, The regressions in table 3A.1 investigate how for- an attempt is made to capture the contestability of eign bank presence affects five performance indica- the domestic market, as measured by the country tors of domestic banks: (a) net margin, (b) nonin- commitments on commercial presence in banking terest income, (c) before-tax profits, (d) overhead under the General Agreement on Trade in Services expenses, and (e) loan loss provisions. All of these (GATS) financial services agreement of 1997. Re- Table 3A.1 Foreign bank presence and domestic bank performance 7,(3) ) caos2) Before tax (4) Loan loss Net margin/ta Nont, income/ta profits/ta Overhead/ta prov./ta Foreign bank share ba76 pr0f128o -0.320' -0.124 M io.026) (0,021) - (0.063) (0.020) (0.065) Index on degree of entry t.150 o0b046s h 0.008 -0.097 f -0.037, (0.010) (0.010) (0.023) (0.010) (0.020) Equity/ta 0.9'0.0373 0.365' -0.025c -0. 210' (0.0a31) (0.011) (0100) (0.014) t .079) Other earning. assets/ta m k0e01t3b 0.096 a mea e -0.081 (0.010) (0.007) (0.022) (0.006) (0.021) Cust. & short-term funding/ta 0i040g b0.001 0.020 0.004 0.010 (1.020) (0.012) (0.058) (0.009) (0.048) Overhead/ta 008r n/ 0.444 -0.168 0.222 (0.084) (0.059) (0.247) (0.273) Growth rate of GP/cap 063 -0.049 0.670 -0.150 -0.690 (0.019) (0.010) (0.155) (0.029) (0.02) Inflation rate y/0.27 0.007 0.060 0.008 -0.031 (0.009) (0.007) (0.011) (0.008) (0.009) Real interest rate t0.069 0.010 0.136 -0.0291' -0.073' (0.01) (0.012) (0.032) (0.012) (0.025) Constaist -0.040 0.0413 -0.075 0.1374 0.0841 0.023) (0.012) (0.060) (0.009) (0.050) Adusted R 0.368 0.429 0.503 0.233 0.423 No. of obs. 134/p. .0149 1342 1362 1213 Note: Regressions are estimated using weighted least squares pooling bank level data across 36 countries for the 1994-2000 period. Only domestic hank observations were used. The number of domestic banks in each period is used to weight the observatioiis. Heteroskedasticity- corrected standard eriors are given in parentheses. a. Significance level of I percent. b. Significance level of 5 percent. c. Significance level 06l 10 percent. Source: Claessens and Lee 2001. 78 THE POOR COUNTRIES' INTERNATIONAL FINANCIAL TRANSACTIONS gressions thus also include a "liberalization index"- Foreign bank share = Number of foreign banks to first created by Sorsa (1997) for the 1995 financial total number of banks. A bank is defined as a services negotiations, and adapted by Qian (2000) foreign bank if it has at least 50 percent for- for the 1997 GATS negotiations. The index runs eign ownership. from 0 to 1. GDPIcap = Real GDP per capita in thousands of The estimated regression is as follows: U.S. dollars. Inflation = Annual increase of the GDP deflator. Ill = ao + 3 FSi, + 13, B,,,+ 13, X,, + 8, Sit + s, Liberalization Index = Degree of commercial pres- ence in banking as allowed in the financial I,,, is the dependent variable (for example, be- services negotiations of 1997 and as reported fore tax profits/total assets) for domestic bank i in in Qian 2000. country j at time t. FS,, is the number of foreign banks in country j at time t as a share of the total All individual bank-level variables are oh- number of banks. B,,t are financial variables for rained from the Bankscope database of IBCA; ad- domestic bank i in country j at time t. X,, are coun- ditional data are obtained from various sources. try variables for country j at time t, and S,, is the All macro data are from the World Bank. "liberalization index." Further, o, is a constant, and 1, 13, 13, and 14, are coefficients, while e,,, is an Economeic analysis of capital outflows error term. Capital outflows can be both the cause and the ef- Estimating a regression in levels-as opposed fect of macroeconomic variables. While a macro- to differences-can be a correct approach provided economic variable (such as growth or fiscal deficit) it is the presence, rather than entry, that causes the may cause outflows, it may also be affected by out- local banking systems to behave differently. More- flows. This relationship would, of course, depend over, the foreign bank presence at time t should be on the extent to which capital outflows are offset determined by entry incentives as of period t-1. If by capital inflows. In turn, inflows may cause out- the foreign bank share is only endogenous to flows and vice versa.41 lagged bank variables, the regression can be esti- The presence of such interactions among van- mated separately using cross-country time-series ables would violate the standard ordinary least data (see further Claessens and others 1998).40 squares assumption that the explanatory variables are exogenous (that is, not correlated with the error Variable definitions and sources term). This endogeneity problem can be partially Net margin/ta = Interest income minus interest addressed by using instrumental variable regressors, expense over total assets. hut single-equation models cannot fully capture the Noninterest income/ta = Other operating income dynamic interactions among several endogenous such as trading costs, advisory fees, and so on variables. A popular method that can capture such over total assets. interactions is the vector-autoregression (VAR) tech- Before-tax profits/ta = Before-tax profits over total nique. For our purpose, we applied the dynamic assets. panel-VAR technique that combines the advantages Overhead/ta = Personnel expenses and other non- of the VAR model with the advantages of panel interest expenses over total assets. data analysis that can admit observable and un- Other expenses/ta = Nonoverhead, noninterest, observable country fixed effects. Such fixed effects other expenses over total assets. would include variables that vary a great deal Equity/ta = Book value of equity (assets minus lia- across countries but remain relatively "fixed" over bilities) over total assets. time for each country-for example, financial de- Other earning assets/ta = Assets other than loans velopment, or demographic patterns. and non-interest-earning assets such as cash We estimate a panel-VAR model with five van- and non-interest-earning deposits at other ables in the following order: capital inflows; capital banks, over total assets. outflows (negative = capital repatriation); the Customer and short-term funding/ta = All short- REER (an increase implies erosion of export com- term and long-term deposits plus other nonde- petitiveness); growth; and the fiscal balance (posi- posit short-term funding over total assets. tive = surplus, negative = deficit). This ordering im- 79 GLOBAL DEVELOPMENT FINANCE plies that the capital flow variables can affect the Table 3A.4 Results of panel-VAR regression for poor macroeconomic variables without restriction (con- countries temporaneously or lagged as the data dictate) but that the macroeconomic variables are restricted to affecting the capital flows variables only through REER Growth lance a lag. Results We ran a panel-VAR regression for all (137) devel- Fiscal balance -0.414 0.0 oping countries for 1980-99 (546 observations), Note: An asterisk indicates significance at 5 percent level or higher. and a separate regression for the poor countries Source: World Bank staff estimates. (142 observations) for the same period. The re- gression coefficients of the five equations are sum- marized in table 3A.2 for all developing countries and in table 3A.4 for the poor countries. The im- Table 3A.5 Summary of Impulse response functions, pulse response functions are summarized in table poor countries 3A.3 for all developing countries and in table 3A.5 for the poor countries. (The impulse responses il- lustrate the effect of a one standard deviation In REER Growth shock to each variable on all the other variables, Inflos taking into account the knock-on effects through Outflows the system over time.) This summary details any REER significant effect over several years at the 5 percent Fiscal balance leve andthe ign f tht efect.Source: World B ank staff estimates. - Table 3A.2 Panel-VAR results for all developing The results for all developing countries pro- countries vide support for the existence of virtuous (and vi- beedn vaiale cious) cycles among the five variables under con- Inflws utflws EER rowh Fical sideration (for example, outflows lead to lower lowREE Growthowt growth which in turn causes further outflows). Innosf0l62w005 Inflws ~ iii9 .ai)~' li~The qualitative results for poor countries follow a OutflowsOuflows0195*-0. similar pattern, although the statistical signifi- Grwt 112) 1, 1. 1 Oi. 11fl, Grn ow th e 1.094*io 0.371*ns ndi puser Fiscal balance -0.41 0.028* Fisal .'i'11 sponses is found to be weaker than in the case of Note: An asterisk indicates significance at 5 percent level or higher. all developing countries.42 Source: World Bank staff estimates However, these results from the panel-VAR exercise tend to be sensitive to the choice of time period or the presence of outliers. The data on ma- croeconomic variables and, in particular, on capi- Table 3A.3 Summary of impulse response functions, tal flows, display considerable Volatility over time all developing countries and also suffer from substantial cross-sectional DepFndeat vaNbe variation. The volatility is even worse in the case of Iflw & utIFlowME REER Growih balance poor countries. InflowsS InflowsMeasuring capital outflows from OOutflows developing countries Growth + +Measuring capital otlw isinherentlydifcl Growthoutflows isdffcl FiFisca balance l e and imprecise. Typically, outflows are measured in- Source: World Bank staff estimates directly, as the residual of "sources of funds" over 80 THE POOR COUNTRIES' INTERNATIONAL FINANCIAL TRANSACTIONS the "uses of funds" from the balance of payments basis of available data (Lessard and Williamson (World Bank 1985; Morgan Guaranty 1986; Cline 1987; Collier and others 2001; Varman-Schneider 1985). This is the procedure adopted here. The 1991). sources of funds include all identified inflows and Finally, estimates of the stock of outflows used credit items in the capital account of the balance of in this chapter are calculated simply by cumulating payments, while uses of funds are the current ac- annual flows over time. This is thc lower bound count deficit and increase in international reserves. for an estimate of the stock of outflows, as the cal- By the balance of payments identity, this residual es- culation ignores interest earnings. Some authors timate yields identical estimates to capital outflows assume that all interest earnings on flight capital calculated directly as the sum of FDI outflows, debt are reinvested abroad, and use the U.S. Treasury outflows, portfolio equity outflows, other outflows, bill rates for estimating interest earnings (Collier, and debit items on the capital account. All data are Hoeffler, and Pattillo 2001). This may provide taken from the International Monetary Fund Bal- some further information on the stock of outstand- ance of Payments Statistics database. ing assets. However, for the purposes of this chap- One of the shortcomings of the residual mea- ter we prefer to emphasize the size of flows leaving sure is that it treats all errors and omissions in the the economy over time (rather than residents' cur- balance of payments as capital outflows. In reality, rent holdings), and therefore do not adjust the cu- errors and omissions may reflect unrecorded cur- iulative stock for any estimate of earnings. rent account transactions as well (Chang, Claes- sens, and Cumby 1997), and also measurement and recording errors and lagged registration (Egger- Notes stedt, Hall, and van Wijnbergen 1995). Another 1. See the overview for a definition of poor Countries shortcoming is that this measure ignores outflows 2. Even so, private capital flows remain well below the takig pacethrugh xpot udernvoiingor rn- average of 5 percent of GDP achieved during the late 1 970s. taking place through export underinvoicing or iand port overinvoicing (Chang, Claessens, and Cumby investment/GDP across countries, in each year. 1997). It is hard to estimate capital flight through 4. In reality even in the highly integrated industrial trade misinvoicing. Even if estimates of over- and economies the correlation hetween investment and saving is underinvoicing were accurate, not all misinvoicingand Horioka 1980). representscng wee accat all mirinvamcin 5. Data weaknesses (particularly on savings in devel- reprsens fnds sedforcapial ligt. Fr eam- oping countries) mean that these figures can provide only a ple, exports may be overinvoiced to take advantage general indication of trends in integration. Also, note that of export subsidies, and imports may be underin- the correlation hetween savings and investment in the mid- voiced to reduce import tariffs (Eggerstedt, Hall, dle-income countries does not decline over the 1990s, de- and van Wijnbergen 1995; Ajayi 1997). spite the massive rise in capital inflows. in part this is due to The residual approach is less restrictive than the fact that a large portion of these inflows were Used to in- crease reserves or capital outflows, and thus had only a lim- other measures that are defined according to the ited role in supporting domestic investment. motives behind capital flight. For example, the 6. Fleisig (1996) Outlines how lack of appropriate laws "hot money measure" suggested by Cuddington and institutions constrains hank lending in developing coun- (1986) attempts to separate the "speculative" or tries. Weak institutions likely make these problems most se- short-term components of capital outflows from vere in the poor countries. 7. The overcollateralization ratio of 5:1 is taken from "normal" outflows. Dooley's method measures Ketkar and Ratha 2000. only that part of capital outflows that does not 8. Knack and Keefer (1995) use a large cross-country generate a corresponding investment income re- time-series dataset; and Mauro's (1995) cross-country ported to the domestic authorities (Dooley 1986). dataset covers 58 countries. Interestingly, Claessens and Naud6 (1993) showpoor Contries results in part Inteestngl, Clessns nd aud~(193) how from declines InI Output inl countries affected by conflict. that the World Bank residual and Dooley methods However, even excluding the conflict countries, the poor actually produce similar estimates of capital flight. countries' per capita oitput rose by only 0.6 percent per We have not attempted to measure the magnitude year in the 1990s. of capital outflows according to motives (for ex- 10. UNCTAD (1999a) confirms that the three African ampl, secuatie reson, tx easio, o siply countries that were most successful in attracting F DI flows ample, speculative reasons, tax evasion, or simplysignificant portfolio diversification) given that motives are ductions in inflation rates and the govcrnment deficit (as a highly subjective and difficult to distinguish on the ratio to GDP). 81 GLOBAL DEVELOPMENT FINANCE 11. See World Bank 1999b, chapter 3; and UNCTAD growth improves bank profitability, but also makes banks 2001b, chapter 4 for detailed discussions of spillover effects less conservative in their provisioning policies. Inflation is in developing countries. associated with higher net interest margins, profitability, 12. See World Bank 2001a. and overheads, consistent with the notion that high infla- 13. The positive impact on growth in developing coun- non requires higher bank margins and profitability to main- tries in general is discussed in World Bank 2001a. tai real bank capital, and that the cost of operating in those 14. This resull is based on a study of 55 poor countries environments is also higher. during 1980-99 based on a Solow-type production that 24. World Bank staff. makes output a function of stocks of capital, labor, human 25. Levine (1999)-building on earlier work by capital, and productivity (see Mankiw, Romer, and Weil Demirgq-Kunt and Deragiache (1998) that controls for 1992; Benhabib and Spiegel 1994). the effects of other factors that are likely to produce bank- 15. Private investment is only correlated with growth if ing crises-has found that the probability that a crisis would Botswana is included in the sample. occur is lower in countries with a higher share of foreign 16. Borensztein, De Gregorio, and Lee (1995) include bank participation. Moreover, Barth, Caprio, and Levine 69 developing countries for 1970-89. UNCTAD (1999b) (2001a) have estimated that the likelihood of a major bank- analyzes the lagged impact of FDI inflows on the average ing crisis is higher in countries with greater limitations on growth rates of about 100 developing countries for five 5- foreign bank presence. year periods over 1970-95. 26. The Basel Committee on Banking Supervision 17. Underachievers in attracting FDI among the cou- (1996) has elaborated guidelines for supervision of cross-bor- tries with a high reform index can be explained by limited der banking that make the solvency of foreign subsidiaries availability of geological and technical information, made- the joint responsibility of home and host supervisor), authori- quate supporting services and infrastructure, and inconve- ties (see also IMF 2000b). Under these guidelines, the home nient geographical location of major mines. country Supervisor is responsible for the consolidated super- 18. Mineral resources are finite, so an accurate mea- vision of the bank on a global basis, while the host countries surement of the benefit of minerals exploitation would sub- are responsible for maintaining the liquidity of foreign tract from these production data the change in asset values branches and subsidiaries, based on their better knowledge of associated with the depletion of the stock of minerals in the local market conditions. ground (see estimates of "genuine savings" in World Bank 27. The problems involved with this and other ap- 2001d, p. 183). Thus the data on production overstate the proaches to measuring capital outflows are discussed in true benefits to the economy of minerals exploitation. 19. The share of bank assets controlled by foreign 28. Ti banks in the Czech Republic, Poland, and Hungary rose dns ases helarod he stock is cu d m- from 12 percent in 1994 to 57 percent in 1999. Similarly, in Latin America, by the end of the decade, foreign banks con- lating over the 1980-99 period, which ignores the stock of trolled more than half of the banking systems of several capital outflows as of 1980 because of lack of data. The cal- countries (Argentina, Chile, Mexico, and the Repiblica culation also excludes interest earned on outflows held Bolivariana de Venezuela), up from between 10 and 20 per- abroad as well as any outflows through underinvoicing of cent in 1994 (Mathieson and Roldos 2001). exports and overinvoicing of imports (see annex 3.1). 20. These numbers refer to mergers where at least one 29. See Collier, Hoeffler, and Pattillo 2000; Cudding- partner is a commercial bank, and thus include cases where ton 1986; Dornbusch 1985; Dooley 1988; Rojas-Suarez a foreign bank acquires a nonbank financial institution. The 1990; Meyer and Bastos-Marquez 1990; Sheets 1996; data cover only those banks reporting to Bankscope, which Lessard and Williamson 1987. includes only locally incorporated foreign-owned banks, not 30. If foreigners are exempt from exchange controls, the branches of foreign banks. then residents may have an incentive, for example, to place 21. To cushion domestic debtors from the currency de- receipts from trade flows abroad by under- or overinvoic- valuation, the government originally sought to convert dol- ing, and to then use a foreign front to invest these funds do- lar debts under $100,000 into pesos, while pledging to re- mestically. In this way the resident investor gains greater fund dollar-denominated deposits in dollars. According to control over the use of profits without forgoing domestic in- estimates, the cost of the currency mismatch for banks could vestment opportunities. well exceed their total equity-coming on top of losses due 31. Indonesia does not record a net outflow in 1998, to borrowers defaulting. Most of these losses are being in- but net inflows were strongly negative. curred by Spanish banks, which had gained a prominent po- 32. This is despite the fact that trade misinvoicing is sition in Argentina since the liberalization of the country's not included in these estimates of outflows (see annex 3.1). banking system in the early 1990s. 33. The results from the panel-VAR exercise should be 22. The index is calculated by: (a) assigning a number treated with some caution, as the data display considerable to a qualitative judgment of the nature of World Trade Or- volatility over time and also suffer from substantial cross- ganization commitments in three areas (cross-border supply, sectional variation. As a result, the results tend to be sensi- consumption abroad, and commercial presence); and (b) tive to the choice of time period or the presence of outliers. taking the average of these numbers (Qian 2000). 34. See IMF 2001b. Examples of controls on current 23. Among other control variables, overhead costs account transactions include restrictions on the repatriation tend to be passed on to customers, in the form of higher of capital and limits to the amount of foreign exchange that margins and fees. In terms of country characteristics, GDP can be obtained for travel. 82 THE POOR COUNTRIES' INTERNATIONAL FINANCIAL TRANSACTIONS 35. Beginning in 1996, the classification system used to of Sub-Saharan Africa." Working Paper 97/68. Inter- characterize current and capital account restrictions was national Monetary Fund, Washington, D.C. changed, with the single "yes/no" variable replaced by a - 2000. "Capital Flight and External Debt in Nige- more disaggregated assessment that is not comparable to ra." In S. Ibi Ajayi and Mohsin S. Khan, eds., External the earlier measures. Debt and Capital Flight in Sub-Saharan Africa. Wash- 36. Multiple exchange rates are typically used either to ington, D.C.: International Monetary Fund. impose different prices for current versus capital account Albuquerque, Rui. 2001. "The Composition of Interna- transactions, or to discriminate among different types of tional Capital Flows: Risk-Sharing through Foreign Di- current transactions. rect Investment." University of Rochester. Processed. 37. For more information on this and other "alterna- Alesina, Alberto, and Roberto Perotti. 1996. "Income Dis- tive remittance systems," see Financial Action Task Force tribution, Political Instability, and Investment." Euro- 2000; and United Nations 1998. pean Economic Review 40(6): 1203-28. 38. Indeed, the presence of extensive dollarization of li- Alesina, Alberto, and Guido Tabellii. 1989. "External abilities has been advanced as a principal reason why some Debt, Capital Flight and Political Risk." Journal of In- countries that on paper have exchange rate flexibility appear ternational Economics 27(3/4): 199-220. not to use that flexibility in practice (the "fear of floating" in Balasubramanyam, V. N., M. Salisu, and D. Sapsford. 1996. the language of Calvo and Reinhart 2000). Balifio, Bennett, "Foreign Direct Investment and Growth in EP and IS and Borensztein (1999) review the additional complications Countries." The Economic journal, 106(434): 92-105. of monetary management in dollarized economies. Balifo, Tomns, Adam Bennett, and Eduardo Borensztein. 39. In Tanzania, after controlling for the effects of 1999. "Monetary Policy in Dollarized Economies." terms of trade, a 1 percent increase in net capital inflows is Occasional Paper 171. International Monetary Fund, estimated to lead to an appreciation of 4 percent in the Washington, D.C. REER (Kimei and others 1997). Barro, Robert. 1991. "World Interest Rates and Invest- 40. Should these assumptions be false, two equations ment." Working Paper Series (U.S.). No. 3849. Na- should be estimated simultaneously-one explaining the tional Bureau of Economic Research, Cambridge, entry decision, and the other explaining the impact of entry Mass. on contemporaneous local banking profits (Claessens and Barth, James, Gerard Caprio Jr., and Ross Levine. 2001a. Lee 2001). "Banking System around the Globe: Do Regulations 41. For example, the proceeds from the sale of a com- and Ownership Affect Performance and Stability?" In pany to nonresidents may be deposited offshore by the resi- Frederic S. Mishkin, ed. Prudential Supervision: What dent seller; or residents may indulge in round-tripping of Works and What Doesn't. Chicago. University of flows, so that outflows are brought back as inflows. Chicago. 42. The coefficient of the real exchange rate in the out- - 2001b. "Bank Regulation and Supervision: What flows equation has a negative sign, implying that an appreci- Works Best?" Policy Research Working Paper 2725, ation of the currency reduces outflows with a lag. This result The World Bank, Washington, D.C. November. is counter-intuitive, and may reflect the use of the official ex- Basel Committee on Banking Supervision. 1996. 'Report on change rate, rather than a market rate, to calculate the real International Developments in Banking Supervision." exchange rate. Many of the countries in the sample had ex- Report number 10. Basel. change controls and substantial differences between market Benhabib, Jess, and Mark M. Spiegel. 1994. "The Role of and official rates, especially during the 1980s. Human Capital in Economic Development: Evidence from Aggregate Cross-country Data." Journal of Mon- etary Economics 34 (October): 143-73. Bhagwati, J. N. 1978. Anatomy and Consequences of Ex- change Control Regimes. Vol. 1. Studies in Interna- References tional Economic Relations, No. 11. Cambridge, Mass.: The word processed describes informally reproduced works National Bureau of Economic Research. that may not be commonly available through libraries. Bhinda Nils, Stephany Griffith-Jones, and Matthew Martin. 1999. "Perception and the Causes of Flows." In Nils Agosin, M. R., and Ricardo Mayer. 2000. "Foreign Invest- Bhinda Leape, Matthew Martin, and Stephany Grif- ment in Developing Countries: Does It Crowd in Do- fith-Jones, eds. Private Capital Flows to Africa: Per- mestic Investment?" Discussion Paper 146. United Na- ception and Reality. The Hague: Forum on Debt and tions Conference on Trade and Development, Geneva. Development. Aitken, Brian, Gordon H. Hanson, and Ann E. Harrison. Blomstran, Magnus, Robert E. Lipsey, and Mario Zejan. 1994. "Spillovers, Foreign Investment, and Export Be- 1994. "What Explains the Growth of Developing havior." Working Paper 4967. National Bureau of Countries?" In William J. Baumol, Richard R. Nelson, Economic Research, Cambridge, Mass. and Edward N. Wolff, eds. Convergence of Productiv- Aizenman, Joshua, and Nancy Marion. 1995. "Volatility, ity. New York: Oxford University Press. Investment and Disappointment Aversion." Working Borensztein, Eduardo, Jos6 De Gregorio, and Jong-Wha Paper 5386. National Bureau of Economic Research, Lee. 1995. "How Does Foreign Direct Investment Af- Cambridge, Mass. feet Economic Growth?" Working Paper S057. Na- Ajayi, Ibi. 1997. "An Analysis of External Debt and Capital tional Bureau of Economic Research, Cambridge, Flight in the Severely Indebted Low Income Countries Mass. 83 GLOBAL DEVELOPMENT FINANCE Bosworth Barry, and Susan Collins. 1999. "Capital Flows to Collier, Paul, Anke Hoeffler, and Catherine Pattillo. 2000. Developing Economies: Implications for Saving and In- "Flight Capital as a Portfolio Choice." World Bank vestment." Brookings Papers on Economic Activity. Economic Review 15 (1): 55-80. The Brookings Institute, Washington, D.C. Coolidge, Jacqueline, Gregory Kisunko, and Aminur Rah- Brownbridge, M. 1998. "The Causes of Financial Distress man. 2001. "Administrative Barriers to Investment in in Local Banks in Africa and Implications for Pruden- the Russian Federation: Inter-Regional Comparisons." tial Policy." United Nations Conference on Trade and Prepared for Economists' Forum 2001. World Bank, Development. Development Discussion Papers (Inter- Washington, D.C. national). No. 132:1-24. Corsetti, Giancarlo, Paolo Pesenti, and Nouriel Roubii. Brunetti, A., and B. Weder. 1998. "Investment and Institu- 1998. "Fundamental Determinants of the Asian Crisis: tional Uncertainty: A Comparative Study of Different A Preliminary Empirical Assessment." Paper prepared Uncertainty Measures." Weltwirtschaftliches Archiv for the conference on Perspectives on the Financial Cri- 134(3): 513-33. sis in Asia, sponsored by journal of International Bubnova, N. B. 2000. "Governance Impact on Private In- Money and Finance and Fordham Univetsity, New vestment: Evidence from the International Patterns of York. Infrastructure Bond Risk Pricing." Technical Paper Cuddington, John T. 1986. Capital Flight: Estimates, Issues 488. World Bank, Washington, D.C. and Explanations." Princeton Studies in International Buch, Claudia, and Gayle Delong. 2001. "Cross-Border Finance 58. Princeton University, Princeton, N.J. Bank Mergers: What Lures the Rare Animal?" Baruch Das-Gupta, Arindam, and Dilip Mookherjee. 1995. "Tax College, New York. Processed. Amnesties in India: An Empirical Evaluation." Discus- Calvo, Guillermo, and Carmen Reinhart. 2000. "Fear of sion Paper S3. Boston University, Institute for Eco- Floating." Working Paper Series (U.S.) 7993. National nomic Development, Boston, Mass. Bureau of Economic Research, Cambridge, Mass. De Gregorio, Jos6, Sebastian Edwards, and Rodrigo 0. Calvo G., L. Leiderman, and C. Reinhart. 1993. "Capital Valdes. 2000. "Controls on Capital Inflows: Do they Inflows and Real Exchange Rate Appreciation in Latin Work?" Working Paper 7645. National Bureau of Eco- America: The Role of External Factors." IMF Staff Pa- nomi Rer Cambrig Mass pers 40:108-51. Washington, D.C. d eow Luid Foregnri es tentLe Chang, K. S. Claessens, and R. Cumby. 1997. "Conceptual rw Evicer Time Siad nD . and methodological issues in the measurement of capi- fo talflght" ntenaionl ouralof Finance and Eco- Demirguq-Kunt, A., and E. Detragiache. 1998. "The Deter- tal (Ugh." ra 1 our1al minants of Banking Crises in Developing and Devel- nomics (U.K.) 2: 101-19. oe onre. M tf aes 5Mrh 119 Claessens, Stijn, and Jong Kun Lee. 2001. "Foreign Banks Wahi on D.C. in Low-Income Countries: Recent Developments and g Impacts." University of Amsterdam, and World Bank, - 2000. "Monitoring Banking Sector Fragility: A Imachnts. DC Multivariate Logit Approach." World Bank Economic Washington, D.C. Claessens, S., and D. Naud6. 1993. "Recent Estimates of Review14(2): 287-307. Capial ligt." orl Bak PoicyResarchWoring Devarajan, S., W. R. Easterly, and H. Pack. 2001. " Is In- Capital Flight." World Bank Policy Research WorkingToo High? Macro and Paper 1186. Washington, D.C. Micro Evidence." Policy Research Working Paper Claessens, Stijn, Ash Demirguq-Kunt and Harry Huizinga. 2519. World Bank, Washington, D.C. 1998. "How Does Foreign Entry Affect the Domestic Devarajan, S., Andrew Sunil Rajkumar, and Vinaya Swa- Banking Market?" Policy Research Paper 1918. World roop. 1999. What Does Aid to Africa Finance?" Bank, Washington, D.C. World Bank Policy Research Working Paper 2092. Clark, A. and K. Naito. 1997. "Structural Reform of the Washington, D.C. Mining Industry in Asia and the Pacific Region." Asian Dollai, David, Giuseppe Iarossi, and Taye Mengistae. 2001. Journal of Mining July/August: 28-42. "Investment Climate and Economic Performance: Some Clarke, George R., Robert Cull, and Maria Soledad Mar- Firm Level Evidence from India." Prepared for Econo- tinez Peria. 2001. "Does Foreign Bank Penetration Re- mists' Forum 2001. Wurld Bank, Washington, D.C. duce Access to Credit in Developing Countries? Evi- Dooley, M. 1986. "Country Specific Risk Premiums, Capi- dence from Asking Borrowers." Policy Research tal Flight and Net Investment Income Payments in Se- Working Paper 2716. World Bank, Washington, D.C. lected Developing Countries." International Monetary November. Fund, Research Department, Washington, D.C. Clements, B., and J. V. Levy. 1994. "Public Education and - 1988. "Capital Flight: A Response to Differences in Other Determinants of Private Investment in the Financial Risks." IMF Staff Papers 35 (September): Caribbean." Working Paper 94/122. International 422-36. Washington, D.C. Monetary Fund, Washington, D.C. Dornbusch, Rudiger. 1985. "External Debt, Budget deficits Cline, William R., ed. 1995. International Debt Reexam- and Disequilibrium Exchange Rates." In Gordon W. med. Washington, D.C.: Institute for International Smith and John T. Cuddington, eds., International Economics. Debt and the Developing Countries. Washington, Collier, Paul and Jan Willem Gunning. 1999. "Why Has D.C.: World Bank. Africa Grown Slowly?" Journal of Economic Perspec- - 2001. "Malaysia's Crisis: Was it Different?" Pre- tives 13(3): 3--22. pared for conference on Preventing Currency Crises in 84 THE POOR COUNTRIES' INTERNATIONAL FINANCIAL TRANSACTIONS Emerging Markets, National Bureau of Economic Re- H. Reisen, eds., Securing Stability and Grozvth in Latin search, January 11-13. Forthcoming in Sebastian Ed- America: Policy Issues and Prospects. Paris: Organisa- wards and Jeffrey Frankel, eds. Preventing Currency ton for Economic Co-operation and Development. Crises in Emerging Markets. Chicago: University of Hayr, Aydin, and Mark Dutz. 1999. "Does More Intense Chicago Press. Competition Lead to Higher Growth?" Discussion Eaton. 1987. "Public Debt Guarantees and Private Capital paper 2249. Centre for Economic Policy Research, Flight." Working Paper Series. (U.S.) 2172 (March): London. 1-33. National Bureau of Economic Research, Cam- Hellmann, Thomas E, Kevin C. Murdock, and Joseph E. bridge, Mass. Stiglitz. 2000. "Liberalization, Moral Hazard in Bank- Edwards, Sebastian. 1999. "How Effective Are Capital ing, and Prudential Regulation: Are Capital Require- Controls?" Journal of Economic Perspectives 13(4): ments Enough?" American Economic Reviei 90(l): 65-84. 147-65. Eggerstedt, H., R. Brideau Hall, and S. van Wijnbergen. Hoekman, Bernard, Hiau Looi Kee, and Marcelo Olar- 1995. "Measuring Capital Flight: A Case Study of reaga. 2001. "Markups, Entry Regulation and Trade: Mexico." Policy Research Working Paper 1121. World Does Country Size Matter?" World Bank. Washington, Bank, Washington, D.C. D.C. Processed. Eichengreen, Barry, Michael Mussa, Giovanni Dell'Ariccia, Hood, Ron. 2001. "Malaysian Capital Controls." Policy Enrica Detragiache, Gian Maria Milesi-Ferrettir, and Research Working Paper 2536. World Bank, Washing- Andrew Tweeedie. 1999. "Liberalizing Capital Move- ton, D.C. ments: Some Analytical Issues." Economic Issues 17. IMF (International Monetary Fund). 2000a. "Country Ex- International Monetary Fund, Washington, D.C. periences with the Use and Liberalization of Capital Feldstein, Martin, and Charles Horioka. 1980. "Domestic Controls." Washington, D.C. Savings and International Capital Flows." Working - 2000b. The Role of Foreign Banks in Emerging Mar- Paper 0310 (NBER Reprints 0085). National Bureau kets." In International Capital Markets: Development, of Economic Research, Cambridge, Mass. Prospects, and Key Policy Issues. Washington, D.C. Financial Action Task Force on Money Laundering. 2000. - 2001a. "West African Economic and Monetary "Report on Money Laundering Typologies, 1999- Union: Recent Economic Developments and Regional 2000." (Available at www.oecd.org/fatf). Policy Issues in 2000." IMF Country Report. Washing- Fleisig, Heywood. 1996. "Secured Transactions: the Power ton, D.C. of Collateral." Finance and Development 33 (June). - 2001b. "Annual Report on Exchange Arrange- International Monetary Fund, Washington, D.C. ments and Exchange Restrictions." Washington, D.C. Fry, Maxwell J. 1992. "Foreign Direct Investment in a Ize, Alain, and Guillermo Ortiz. 1987. "Fiscal Rigidities, Macroeconomic Framework: Finance, Efficiency, In- Public Debt and Capital Flight." IMF Staff Papers 34 centives and Distortions." PRE Working Paper. World (June): 311-32. Bank, Washington, D.C. Jenkins, Carolyn. 1998. "Determinants of private invest- Goldstein, M., G. Kaminsky, C. Reinhart. 2000. Assessing ment in Zimbabwe." Journal of African Economies 7 Financial Vulnerability: An Early Warning System for (March):34-61. Emerging Market. Washington, D.C.: Institute for In- Johnson, C. 1990. "Ranking countries for mineral explo- ternational Economics. ration." Natural Resources Forum 14 (3): 178-86. Gordon, David, and Ross Levine. 1988. "The Capital Flight Kaplan, Ethan, and Dani Rodrik. 2001. "Did the Malaysian Problem." International Finance Discussion Paper Capital Controls Work?" Prepared for conference on 320. U.S. Federal Reserve Board, Washington, D.C. Preventing Currency Crises in Emerging Markets, Na- Goreux, Louis. 2001. "Conflict Diamonds." Africa Region tional Bureau of Economic Research, January 11-13. Working Paper 13. World Bank, Washington, D.C. Forthcoming in Sebastian Edwards and Jeffrey Gyimah-Brempong, K., and S. Mufioz de Camacho. 1998. Frankel, eds. Preventing Currency Crises in Emerging "Political Instability, Human Capital and Economic Markets. Chicago: University of Chicago Press. Growth in Latin America." Journal of Developing Kasekende, Louis A. 2000. "Capital Account Liberaliza- Areas 32(4): 449-65. ton: The Ugandan Experience." Paper presented at the Gyimah-Brempong, K., and T. L. Traynor. 1999. "Political Overseas Development Institute, London, June 21. Instability, Investment and Economic Growth in Sub- Kathuria, V 1998. "Technology Transfer and Spillovers for Saharan Africa." Journal of African Economies 8 (1): Indian Manufacturing Firms." Development Policy 52-86. Review 16 (1): 73-91. Harrison, Ann. 1993. "Are there Positive Spillovers from Ketkar, SIihas, and Dilip Ratha. 2000. "Development Fi- Foreign Direct Investment? Evidence from Panel Data nancing During a Crisis: Securitization of Future Re- for Morocco." Journal of Development Economics. ceivables." Policy Research Working Paper 2582. (Netherlands) 42 (October): 51-74. World Bank, Washington, D.C. - 1994. "Productivity, Imperfect Competition, and Khan, Mohsi, and Nadeem UI-Haque. 1985. "Foreign Trade Reform: Theory and Evidence." Journal of In- Borrowing and Capital Flight: A Formal Analysis." Fi- ternational Economics 36: 53-73. nance and Development 32(4): 606-28 Hausmann, R., and M. Gavin. 1996. "Securing Stability Kimei, C., G. D. Mierna, B. Tarimo, and A. Msutze (with and Growth in a Shock-Prone Region: The Policy M. Martin and N. Bhinda). 1997. "Tanzania: Capital Challenge for Latin America." In R. Hausmant and Flows Study." Paper presented to EFA workshop on 85 GLOBAL DEVELOPMENT FINANCE Private Capital Flows and Macroeconomic Policy in Mian, S., and C. Smith. 1994. "Getting Help with Credit Sub-Saharan Africa, Cape Town, July 1997. Administration." Corporate Finance 11. Knack, S., and P. Keefer. 1995. "Institutions and Economic Mody, Ashoka, and Antu Panin Murshid. 2001. "Growing Performance: Cross-Country Tests Using Alternative up with Capital Flows." International Monetary Fund, Institutional Measures." Economics and Politics 7 Washington, D.C. Processed. (November): 207-27. Morgan Guaranty Trust Company. 1986. "LDC Capital Kokko, A. 1994. "Technology, Market Characteristics, and Flight." World Financial Markets. March, 13-15. Spillovers." Journal of Development Economics 43 Naito, Koh, Felix Remy, and Peter van der Veen. 2001. (2): 279-93. "Mining Sector Reform and FDL" World Bank, Wash- Kokko, A., Ruben Tansini, and Mario C. Zejan. 1996. "Local ington, D.C. Processed. Technological Capability and Productivity Spillovers Naito, K., Felix Remy, and J. P. Williams. 2001. "A Coi- from FDI in the Uruguayan Manufacturing Sector." parative Revicw of Legal and Fiscal Frameworks for Journal of Development Studies 32 (4): 602-11. Exploration and Mining: Best Practices." MingJour- Krugman, Paul. 1998. What Happened to Asia? (available nal Limited, p. 177. at http://web.mit.edulkrugman/www/disinter.html). Naito, K., H. Myoi, J. Otto, D. Smith, and M. Kamitani, Kumar, Nagesh, and Jaya Prakash Pradhan. 2001. "Foreign 1998. "Mineral Projects in Asian Countries: Geology, Direct Investment, Externalities and Economic Growth: Regulations, Fiscal Regimes and the Environment." Some Empirical Explorations for the Poor Countries." Resources Policy 24 (2): 87-93. A background paper for GDF 2002. World Bank, Ndikumana, L. 2000. "Financial Determinants of Domestic Washington, D.C. Investment in Sub-Saharan Africa: Evidence from Panel Laban, Raul, and Felipe Larrain. 1997. "Can a Liberalization Data." World Development 28 (2): 381-400. of Capital Outflows Increase Net Capital Inflows?" Jour- Ng'eno, N.K. 2000. "Capital Flight in Kenya." In S. Ibi nal of International Money and Finance 16(3): 415-31. Ajayi and Mohsin S. Khan, eds., External Debt and Lessard, Donald, and John Williamson. 1987. Capital Flight Capial Flight in Sub-Saharan Africa. Washington, and Third World Debt. Washington, D.C.: Institute for D.C.: International Monetary Fund. International Economics. Ngugi, R. 2000. "Financial Reform Process in Kenya." Levine, Ross. 1996. "Foreign Banks, Financial Development African Development Review12(1): 52-77. African and Economic Growth." In Claude E. Barfield, ed. In- Development Bank (International). ternational Financial Markets: Harmonization versus OECD (Organisation for Economic Co-operation and De- Competition. Washington, D.C.: AEl Press. velopment). 2001. Cited in figure 4.1. - 1999. "Foreign Bank Entry and Capital Control Olopoenia, Razaq A. 2000. "Capital Flight from Uganda." Liberalization: Effects on Growth and Stability." Uni- in Ajoyi and Kahn, eds. External Debt and Capital versity of Minnesota. Processed. Flight in Sub-Sabaran Africa. Washington, D.C.: Inter- Levinsohn, James. 1993. "Testing the Imports-as-Market- national Monetary Fund. Discipline Hypothesis." Journal of International Eco- Oshikoya, T 1994. "Macroeconomic Determinants of Do- nomics 35(1/2): 1-22. mestic Private Investment in Africa: An Empirical Loungani, Prakash, and Paolo Mauro. 2000. "Capital Analysis." Economic Developient and Cultural Flight from Russia." IMF Policy Discussion Paper Change 42(3): 573-96. 00/6. International Monetary Fund, Washington, D.C. Otto, J. 1992. "A Global Survey of Mineral Company In- Mankiw, N. Gregory, David Romer, and David N. Weil. 1992. vestment Preferences." In J. Otto and T. Waelde, eds., "A Contribution to the Empirics of Economic Growth." Mineral Investment Conditions in Selected Countries Quarterly journal of Economics 107(2): 407-37 of the Asia-Pacific Region. Publication ST/ESCAP/ Marksun, J. R., and Anthony J. Venables. 1997. "Foreign 1197. New York: United Nations. Direct Investment as a Catalyst For Industrial Devel- - 1995. "Legal Risk Analysis for Mining Projects." opment." Working Paper 624. National Bureau of Mineral Industry International No. 1025. Institute for Economic Research, Cambridge, Mass. Mining and Metallurgy, London. Mathieson, Donald J., and Liliana Rojas-Suarez. 1993. Oxford Analytica. 2001. "Africa: Regional Stock Markets." "Liberalization of the Capital Account: Experiences September 4. and Issues." Occasional Paper 103. International Ozawa, Teruro. 1992. "Foreign Direct Investment and Monetary Fund, Washington D.C. Economic Development." Transnational Corporations Mathieson, Donald J., and Jorge Roldos. 2001. "The Role 1(1): 27-54. of Foreign Banks in Emerging Markets." Paper pre- Petersen, M. A., and R. G. Rajan. 1994. "The Benefits of pared for the IMF-World Bank-Brookings Institution Lending Relationships: Evidence from Small Business conference on Financial Markets and Development, Data." Journal of Monetary Economics (1): 3-38. Washington, D.C. - 1997. "Trade Credit: Theories and Evidence." Re- Mauro, P. 1995. "Corruption and Growth." Quarterly view of Financial Studies. U.S. (10): 661-91. Journal of Economics 1 10(April): 681-712. Pfeffermann, G., and G. Kisunko. 1999. "Perceived Obstacles Mexico Federal Commission on Competition. 1997. Annual to Doing Business: Worldwide Survey Results." Interna- Economic Competition Report. Mexico City. tonal Finance Corporation, Washington, D.C. Processed. Meyer and Bastos-Marquez. 1990. "A fuga de capital no .2001. "The Foreign Direct Investment Environment Brasil: 1975/88." Pesquisa e Planelamento Economico in Africa." Africa Region Working Paper 15. World (Brazil) 0:4945. Bank, Washington, D.C. 86 THE POOR COUNTRIES' INTERNATIONAL FINANCIAL TRANSACTIONS Porter, M. E. 1990. The Competitive Advantage of Nations. Stiglitz, Joseph E., and Andrew Weiss. 1981. "Credit Ra- New York: Free Press. tioning in Markets with Imperfect Information." Qian, Ying. 2000. "The Process of China's Market Transi- American Economic Review 71: 393-410. non, 1978-1998: The Evolutionary, Historical, and Stryker, Dirck. 1997. "Dollarization and its Implications for Comparative Perspectives." Journal of Institutional and Ghana." Associates for International Resources and Theoretical Economics (Germany). 156(1): 151-79. Development, Cambridge, Mass. Razin, Asaf, Efraim Sadka, and Chi-Wa Yuen. 1997. Thomas, R. H. 1994. "Trading and Investing in Tanzania: The "Channeling Domestic Savings into Productive Invest- Legal Framework." Development South Africa 11(3). ment under Asymmetric Information: the Essential Tornell, Aaron, and Andres Velasco. 1992. "The Tragedy of Role of Foreign Direct Investment." Working Paper the Commons and Economic Growth: Why Does Cap- 6338. National Bureau of Economic Research, Cam- ital Flow from Poor to Rich Countries?" Journal of bridge, Mass. Political Economy 100(6): 1208-31. Reinikka, R., and J. Sevensson. 1999. "How Inadequate Trester, Jeffrey. 1998. "Venture Capital Contracting tinder Provision of Public Infrastructure and Services Affects Asymmetric Information." Journal of Banking and Fi- Private Investment." Policy Research Working Papers nance 22 (August): 675-99. 2262, World Bank, Washington, D.C. Tybout, James. 2000. "Manufacturing Firms in Developing Rojas-Suarez, Liliana. 1990. "Risk and Capital Flight in De- Countries: How Well Do They Do, and Why?" journal veloping Countries." Working Paper 90/64. Interna- of Economic Literature 38(l): 11-44. tional Monetary Fund, Washington, D.C. United Nations. 1998. U.N. General Assembly Special Session Sakr, K.1993. "Determinants of Private Investment in Pak- on the World Drug Problem. 8-10 June. www.un.org istan." Working Paper 93/30. International Monetary UNCTAD (United Nations Conference on Trade and Devel- Fund, Washington, D.C. opment). 1999a. "Foreign Direct Investment in Africa: Schadler, S., M. Carkovic, A. Bennett, and R. Kahn. 1993. Performance and Potential." Geneva. "Recent Experiences with Surges in Capital Inflows." 1999b. World Investment Report. Geneva. Occasional Paper 108. International Monetary Fund, 2001a. "Foreign Direct Investment in Least Devel- Washington, D.C. oped Countries at a Glance." Geneva. Schineller, Lisa M. 1997. An Econometric Model of Capital _ 2001b. World Investment Report 2001. Cited fig- Flight from Developing Countries. International Fi- nance Discussion Paper 579. U.S. Federal Reserve oe24 nanc Dicusson aper579 U.S FeeralResrve Varman-Schneider, Benu. 1991. Capital Flight from Devel- Board, Washington, D.C. oping Countries. Boulder: Westview Press. . 1999. A Nonlinear Econometric Analysis of Capital World Bank. 1985. World Development Report Interna- Flight." International Finance Discussion Paper 594. tional Capital and Economic Development. New York: U.S. Federal Reserve Board, Washington, D.C. p p Serven, L. 1996. "Does Public Capital Crowd out Private Oxford.Univrsty Press. Capital?: Evidence from India." Policy Research P 19. "Srateg f anMing ." h Working Papers 1613. World Bank, Washington, D.C. P 18. Wr an ing tn D.C. - 1998. "Macroeconomic Uncertainty and Private In- 196 " Mining S aer Lati Amric and vestment in Developing Countries: An Empirical Inves- th ibben.Ch tigation." Policy Research Working Papers 2035. World Bank, Washington, D.C. . 1999a. Global Economic Prospects. Washington, Serven, L., and A. Solimano. 1993. "Debt Crises, Adjust- D.C. ment Policies and Capital Formation in Developing . 1999b. Global Development Finance 1999. Countries: Where Do We Stand?" World Development Washington, D.C. 21: 127-40. - 2001a. Glohal Development Finance 2001. Sheets, Nathan. 1996. "Capital Flight from the Countries in Washington, D.C. Transition: Some Theory and Empirical Evidence." In- . 2001b. "Finance for Growth: Policy Choices in a ternational Finance Discussion Paper 514. U.S. Federal Volatile World." Washington, D.C. Reserve Board, Washington, D.C. - 2001d. World Development Indicators 2001. Sievers, Sarah E. 2001. "Competitiveness and Foreign Di- Washington, D.C. rect Investment in Africa." In Augustin Kwasi Fosu, World Economic Forum. 1998. "The Africa Competitive- Saleh M. Nsouli, and Aristomene Varoudakis, eds., ness Report 1998." Geneva. Policies to Promote Competitiveness in Manufacturing Yun, Mikyung. 2000. "Cross-Border M&As and their Im- in Sub-Saharan Africa. Organisation for Economic Co- pact on the Korean Economy." Paper presented at the operation and Development. UNCTAD Seminar on Cross-border M&As and Sus- Smith, D., and K. Naito. 1998. "Asian Mining Legislation: tained Competitiveness in Asia: Trends, Impacts and Policy Issues and Recent Developments." Resources Policy Implications, Bangkok, March 9-10. Processed. Policy 24: 125-32. Zemplinerova, A., and M. Jarolim. 2000. "FDI through Sorsa, Piritta. 1997. "The GATS Agreement on Financial M&A vs. Greenfield FDI: The Case of the Czech Re- Services: A Modest Start to Multilateral Liberaliza- public." Paper presented at the UNCTAD and OeNB tion." Working Paper 97/55. International Monetary Seminar on FDI and Privatization in Central and East- Fund, Washington, D.C. erD Europe, Vienna, March 2-3. Processed. 87  41 Strengthening Official Financial Support for Developing Countries Mixed results from aid have led to a true that even in many countries with good poli- fall in aid cies, lack of administrative capacity lowers the S low progress in poverty reduction during the marginal productivity of aid as aid levels rise. 1990s outside Asia increased concerns about the However, recent research indicates that aid levels to effectiveness of aid.' Many countries have achieved most countries with strong economic programs are impressive growth rates with the support of aid well below the threshold where aid becomes in- flows, and since 1990 the share of people living effective. This analysis supports the view that a in extreme poverty in developing countries has doubling of aid could make an effective contribu- dropped from 29 percent to 23 percent, led by rapid tion toward reaching the Millennium Development progress in China and India. Nevertheless, growth Goals, provided that the aid is allocated wisely. has been slow in many of the poorest aid recipients Donors also have made progress in improving (see chapter 3), and in Sub-Saharan Africa the share the design and administration of aid programs, al- of the population living on less than a dollar a day though much more remains to be done. Greater stagnated during the 1990s, contributing to a grow- efforts are directed at ensuring that policy condi- ing perception that aid flows have failed to support tions in adjustment assistance reflect a program development. This perception, in conjunction with that has the full support of the government and fiscal pressures in donor countries and the declin- other domestic stakeholders. This new emphasis ing strategic value of aid (from the perspective of involves greater selectivity in aid disbursements. donors) with the end of the Cold War, led to a sharp The administrative burden of aid is less because fall in aid over the 1990s. the share of tied aid is reduced, and the govern- Mixed progress in poverty reduction also led ment is assuming more leadership in promoting to a reevaluation of aid policies, and to a growing aid coordination. consensus on donor policies required to increase aid effectiveness. Perhaps most importantly, the allocation of aid is increasing to those countries The policy framework with good policies. Despite high levels of aid, most countries with good policies can continue to ab- Providing a policy environment conducive to sorb additional aid resources without seriously im- growth and development- pairing the effectiveness of that aid. High aid levels The growing consensus on how to improve donor to countries with good policies should not raise policies has its roots in the mixed success of efforts fears of excessive dependence. Over time, strong to help developing countries recover from the fail- growth should generate the increase in tax rev- ure of many economic policies of the 1970s and enues required for a decline in aid. Aid does not, in 1980s. Growth in many developing countries was general, increase the volatility of government re- depressed by unsustainable macroeconomic poli- sources, and appropriate policies can ensure that cies, financial repression, high trade barriers, perva- aid does not contribute to inflationary pressures sive state interventions in competitive markets, and or cause excessive exchange rate appreciation. It is complex administrative constraints on entrepre- 89 GLOBAL DEVELOPMENT FINANCE neurial activity. Donor programs during the late Recognition of the failure of aid to boost growth 1980s and throughout the 1990s thus increasingly in many heavily indebted poor countries (HIPCs)- focused on supporting efforts at providing an eco- increases the legitimacy of focusing resources on nomic policy environment conducive to growth and debt relief. Ultimately, improved policies in develop- development. Improvements in economic policies ing countries and a more effective approach to aid during the 1990s did help many developing coun- should strengthen donor support for increasing aid tries to achieve substantial increases in growth rates resources. These messages underscore the important over the "lost decade" of the 1980s. However, many themes emerging from the United Nations (U.N.)s of the poorest countries continued to be left behind, Financing for Development (FfD) process (see box and it became clear that weak institutions and poor 4.2). Unfortunately, recent aid trends have been dis- governance were at least as significant constraints appointing, and there appears to be little likelihood on development as inflation and price controls, that a rise in aid will be significant and sustained. -with a reform of donor policies- At the same rime, some instruments that donors used to support developing countries' economic Trends in aid programs proved inadequate. Compliance with conditionality under adjustment lending was mixed. A widening gap between the availability of Official lending and guarantees coupled with poor aid and the needs of recipients- policies contributed to debt burdens. Aid programs Aid flows dropped sharply over the last decade in increased the administrative burden in many coun- real terms, and by 2000 stood more than 10 per- tries where capacity was a principal constraint on cent below the 1990 level. Expressed as a share growth. Recognition of these problems catalyzed of donors' gross national product (GNP), aid fell efforts to strengthen the framework for adjustment from 0.33 percent In 1990 to 0.22 percent in 2000. assistance, provide debt relief, and reduce the ad- Only five donor countries reached (or surpassed) the ministrative burden of aid by improving donor U.N.'s target of 0.7 percent of GNP which was en- coordination. These efforts do not represent an dorsed by Group of Seven (G-7) countries at entirely new departure: aid coordination, capacity the Earth Summit in Rio in 1992. At the same time, constraints, and adjustment assistance have been a the need for aid continues to grow. Developing focus of analysis for some time. Nevertheless, in the countries' population rose by 17 percent during the past few years concerted efforts have been made to 1990s, and the number of people (outside China) adjust donor policies in the context of recent expe- living on less than $1 a day has remained roughly rience. At the Bretton Woods institutions, this shift the same. Some 60 million people in developing in assistance to low-income countries is being im- countries are infected with the human immunode- plemented through the Poverty Reduction Strategy ficiency virus. The Millennium Development Goals Paper (PRSP) approach (see box 4.1). cannot be met without increased aid. For example, preliminary calculations indicate that a doubling of -to increase the effectiveness of aid aid, appropriately allocated, will be necessary to These two debates over development policy-that a halve poverty by 2015. Estimates of the aid (above deepening of reform programs must address critical current levels) required to meet the goals for edu- institutional and governance issues that constrain cation, health, and the environment (see box 4.2) growth, and that donor policies must support coun- range from $35 billion to $76 billion.2 Vigorous try ownership, reduce the administrative demands steps to increase the availability of aid resources, in of aid programs, and focus on development re- conjunction with improved donor policies to sup- suits-are intricately related. A greater focus on de- port increased aid effectiveness, should be the top velopmenr outcomes may be useful in determining priorities for the international community. the overall allocation of funds by donors and as a basis for monitoring and evaluation of reform pro- -particularly over the last two years- grams. The recognition that institutional capacity is After a modest recovery in aid flows beginning in a major constraint on growth underlines the impor- 1998, the past two years have seen a further de- tance of easing the administrative burden of aid. dine. Concessional aid flows are measured in two 90 STRENGTHENING OFFICIAL FINANCIAL SUPPORT FOR DEVELOPING COUNTRIES Box 4.1 The PRSPs In December 1999 the Boards of the International t-More col,crept partnerships aid aid coo'dion. Monetary Fund'(liMF) and the World Bank approved aPRPs re itendcd to be instrIMeIts by hich go%- new approach to the challenge of reducing poverty Ilow -"eriments c-in achieve better aid ioirdination. It is income countries based on country-owned poverty reduc- good practice lor the PRSP process to be inclus.ve ot tion strategies that would serve as a framework for devel don.rs,. and miost countrie, are in fact doing tli.. in- opment assistance. Much has been accomplished during cluding, for example, through the representation of the past two years---niine countries have completed their donors on PRP workmg groups first full PRSP "rd three courntries have completed theii Jlan.uz:. and simi-ing donor proceditres. along- first annual PRSP implementation progress reports. Some -side a greater locius on devcpment re;idr. as opposed' 41 countries have also completed their interim poverty re to monioring and eforts to contrul inputs. Each PRSP duction strategies'(I-PRSPs) and eight countries,have sub- is expected to include inermediate and longer-rerm sequently submitted their PRSP preparation status repIrts indicatur, on poierti outcome,, (o enable regular mon- for consideratio-i bthe BoardI itoring ot progresb, upon which governments would The central message of the forthcoming Review of anniinll report. It is hoped that this % ill uncourage the PRSP Approach4 is a substantial affirmation by low- go.eriime1Fs and their txternal partners to focus on income countries-as well as development partners andcivi the same et of targets and indicators o%er - sufficientlY society organizations of the value of the PRSP approach, long period, so is to reduce the costs associired with and the importance of country ownership as a guiding niilnple reporting reqtirIements, durin whiich time it principle, and a corresponding recognition of the need for 'wolJ be po4sible to measurc results and to, aJdlu.t Jo-. flexibility to allow for different country starting points. mestic strategics and cmtrnal assistance wccotdinglk. It is widely recognized that aligning donor programs with the PRSP iscrucial to sustaining this approach; li -n In the longer term it i< , -ptcred that the PRSP will part the PRSP'appr-oach has been designed to over60e'rne acilitate greaer aid allocations to counries xk ah good long-standing problems of poor donor coordination, weak policy environm,ncs. To the extent that PRbPs reveal v. hat country ownership of donor-financed programs, andthe 'a counL% is truly prepared to do on rei ins of polich and fragmentation ofgovernmental'programs and institutions institutional relorms and expcndittire allocations), the caused by multiple,'and often inconsistent, donor iitervenl Aould pro% ide reliable indicator tor donors io allocate tions. Donor -aligniient is needed at various levels, bothj funds on the ba-is of policies. Over time a country's pert or- substantive (inensuring'that donors respect country prior fance with respect to its PRSP objecie iboth policv ities) and in terms of processes (to reduce the transaction ieasur% and development resulIS could help impro%e costs associated ivith aid). donors' judgments concerning the allocation of aid. Key challengeof'the PRSPfor development paaAners,' As reported in the upcoming Review%, early e Idence include: ' aiithe PRSP process is posi-te. and subtaial invest- tnmnts ,are being made by lo%%-income countries and de%el- Pursuing new approaches to support government opment partners in making ihis approach v.ork. While the ownership. Governments prepare their own povert qualit o the early full PR5Ps has varied (for example, in reduction strategies through a participatory process termsf participation, data collection, the,reaism of- onwf designed to build broad'ownership at te national terim goals5nd institutional capacity tb mtor exptmdi1-of level. Medium-term reform programs supported by tures and the linkto poverty reductio6), the process has Poverty Reduction Support Credits (PRSCs) will be helped promote ownership, encouraged abetter a principally drawn from, or will elaborate on, policy within countries, broadened the understanding of p measures contained in the PRSPS.5 ment ISSueS, and helped improve d9por coordinain ways: aid recorded as received by developing coun- countries (see the data annex at the end of this tries and aid recorded as provided by donor coun- chapter). Aid flows received by developing coun- tries. The two measures are different because in any tries (excluding technical cooperation grants) fell by given year the concessional funding provided by 3.8 percent in 2000 to $40.7 billion and they are es- donors to multilateral institutions is not the same timated to have declined by a further 3.4 percent in as those institutions' disbursements to developing 2001 (see table 4.1). Much of this decline was due 91 GLOBAL DEVELOPMENT FINANCE Box 4.2 The Financing for Development (FfD) process T he analysis presented ini his documn supports ti h- apacity building; developing countries must cooper- agenda of the FfD conference that will take-placein atethrough strengthening their infrastructure to sup- March 2002, in Monterrey, Mexico' The FfD process em- port trade and lowering their own trade barriers. The phasizes the importance of a comprehensive approach to launch of a "development round" following the Doha the mobilization of resources for development and of the meeting of the World Trade Organization will involve flexibility and partnerships required to nsire that the idegotiations of market access issues covering agricul- needs and opportunities of different countries are taken - ture, services, and manufactures, as well as rules gov- into account in the support provided by the international, erning dispute settlement, disciplines on regional inte- community. The purpose ofFfD is to assure the means"bo gration, environment, and trade-related intellectual reduce poverty and reach theAlillenniumDevelopment property rights In addition, negotiations may also Goals as well as other internationally agreed-on develop- rake place regarding investment, competition, trade ment targets. facilitation, and transparency in government procure- The FfD agenda recognizes that-the means of reaching ment. This approach should enable progress to be these goals must be defined broadly. Policy reforms in de-' made in improving market access for developing veloping countries-are required to boost growth and re- countries (assumingthey are willing to negotiate on duce poverty. At the same time, industrial,tountries deed the basis of reducing their own barriers to trade), to open their markets to provide sufficient opportunities, which is the main priority for the trade agenda. for developing countries to berefit from the world trading, * Resources. Poor countries with good policies will need system, to help shape improvefments in the international,"- increased assistance to meet the development goals financial architecture, and to'boost the aid resources re- articulated in the U.N.-sponsored Millennium Decla- quired to help countries meet the developi-nent goals.The - ration.,These goals include halving- extreme poverty, main messages of Global Detelopnient Finance 2002 can achieving universaf primary education, eliminating be viewed under this paradigm: - - gender disparity in education, reducing infant and - child mortality and maternal mortality, ensuring ac- * Policies. The discussion f,counitrypolicies at the Ff cessto reproductive health services, nd implementing conference will focus or idiiproying ftlinvement-d-' a national strategy for sustainabled'evelopment in mate in developing countries. In particular, policies every country. Progress since 1990 has been too slow focused on maintaining macroeconomic stability,ii- to achieve most of the goals, and a stepped-up effort creasing openness to trade and-foreign direct invest-- by developing countries, industrial countries, and mul- ment (FDI), improving governance, and strengthening tilateral institutionsis required,to have any chance of financial sector instittiti6ns wili help developing coun- meeting them.6 This effort should include a doubling tries benefit from greater-financial integration while of aid to achieve the poverty goal, provided that these reducing the potential costs'-This document,shows resources are allocated to countries with good policies how a strong investmenit'ciimate,in the-poor countriesa- (where aid will be most effective) and with many poor can boost the effectivenesscf 'aid, increase. domestic people. Some of the funding needs required to meet investment by limiting capip] outfl-owsand attracting- - the health and education goals are the same as those more FDI inflows, an dimprove the productivity of ' required to halve poverty, but some will require dedi- investment. At the samerime,Ais document outlines'-" cared funding, such as the need to address communi- ongoing improvements -in -oifor policiesto.s sengthen ncable diseases or to promote "Education for All." A administration, of aid programs, increase the effective portionof these resources should be used to finance ness of policy-conditiotiaiyasa- meanso ofnhancing - g- ' gtob,al'public goods, such as the creation of new vac- government credibility'and'c commnitmn e'sure that cites, and thus would not be channeled through ndi- debt relief is directed at countries with kood policies,- vidual developing-ountry governments. and ensure that guaranteed lending does not con-thoh tribute to unsustainabled ebt -burdens. ' h In corutries with poor policies, even very large * Opportunity., All countiiies -n" edrto cop:erat e in - - amounts of'aid are likely-rb achieve only a limited and integrating the developing duntries into the worldt short-lived impact on poverty. There is,- therefore, an in- trading system. Industrial countries must cooperate evipable tension between allocating aid to achieve the max- through opening their markets (particularlyincagri-n tmum global progress toward the goals-and allocating aid culture and textiles) and providing resourcesfor so that eachcountry or region has a chance of meeting 92 STRENGTHENING OFFIC IA L FINANCIAL SUPPORT FOR DEVELOPING COUNTRIES Box 4.2 wi(covrn' those goals. To resolve thiSf attention rh,l bl aidtio sl in fctbianghe mriasxed be focu, d (in iniprovi g pohi esgli6n LUft.ie 7,.hcre I h %, rr n% muiddl-1101i 0 FmrN4" are weak rt Iha e Finalk. ch n itn fo I communitv taces a dilemma in ver .00 countries with good policies Nvrthcl in supporting progre to ward the goaisin mlddle'-income - It 5 1portant or donors to consider how otadre h countrie; %ith poor region; . a' not be adei;able Io scere poverty isiues in some middle-nme .onire provide largc arnount. ol ad to countr.es Ihat hate sub- one recommendation would be tr f'und relaelm stantia! tinancial re-,ou.rCe bae.n,t made progress il project aimed at demonstrating effective approache a alleviating po,.err% in m-)mt.r'tn~'r~ .oe isd to a drop in Japanese aid to East Asia, because dis- fall in aid from Japan, and the removal of coun bursements against the large commitments made at tries from the list of those eligible to receive ODA the time of the Asian crisis fell. Preliminary esti- because their per capita incomes now exceed the mates suggest a continued increase in aid to Eastern cutoff for flows to be counted as aid.8 Adjusting Europe and Central Asia, both due to stepped-up for the change in the DAC list, ODA fell by 0.2 assistance to the Balkans and support for the efforts percent in real terms in 2000. The decline was due of the advanced Eastern European countries to join to the fact that in the G-7 countries.aid fell by 4.8 the European Union (EU). Aid flows have declined percent in real terms; aid from non-G-7 countries to Sub-Saharan Africa due to delays in implementa- increased by 8.3 percent in real terms. tion of reform in some countries; aid flows have de- clined to a lesser extent to South Asia despite a rise -and little sign of a reversal of this trend in in humanitarian assistance to India following the the medium term- devastating earthquake in 2001. The prospects for a rise in aid over the medium The amount of official development assistance term are mixed. Several donors, in particular the (ODA) provided by donors fell by 1.6 percent in United Kingdom and several of the non-G-7 coun- real terms in 2000 to $53.1 billion, or 0.22 per- tries, have been able to set and meet medium-term cent of Development Assistance Committee (DAC) targets for substantial increases in aid flows. How- members' GNP (data on aid flows from donors for ever, there is little sign of substantial increases in 2001 are not yet available). This decline, which re- aid from the four largest donors-France, Ger- versed the upward trend that commenced in 1998, many, Japan, and the United States-which to- was due to two special factors: the above-noted gether account for almost two-thirds of all aid. In Tablec4.1 Net official aid to developing countries, by type and source, 1990-2001 (billions of dollars) Aid i4K I i ODA and official aid .to the fac t t i th - 4i fb Grants (ex tech coop) h0I 31 3'5 2. 2- 35 3 2 -)'~~i Bilateral percen Xi s mi Multilateral 6 458 81 ,~4 tj ,~ Concessional loans In I term. Bilateral . .and 3tl s i Multilateralt iu Memo item Techbcoop grants . . .....Th pro s p.e. o r . . . . a r .i ai e t i Note: Data are based on the OECD DAC definition of aid as measured by donors. These data differ from concessional flows reported in volume 2, which are primarily based on information collected through the World Bank Debtor Reporting System Source: OECD DAC; World Bank Debtor Reporting System; staff estimates. 93 GLOBAL DEVELOPMENT FINANCE Figure 4.1 ODA from donor countries in relation to their GNP, 1990-2000 Billions of dollars Percent of GNP 70 - 0.4 Percentage of GNP of DAC countries (right axis) L 60 - Others -0.3 Netherlands 50 - 5LUnited States 40 - United Kingdom 0 2 Japan 30 - Germany O] France 20 - 10 - 0 00 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000a Net ODA flows (left axis) a. Provisional dara. Source: OECD April 2001. part, slow growth or declines in aid flows as re- and East Timor following the end of each conflict, corded by the DAC reflect the removal of a few in Central America following Cyclone Mitch, and countries from the list of countries eligible for aid. in Turkey and India following earthquakes (in For example, recorded aid flows from France were 1999 and 2001, respectively). Aid flows also rise affected by the removal of French Polynesia and sharply in times of global conflict-for instance, New Caledonia, the largest beneficiaries of French by 20 percent during the Gulf War of 1991. While assistance, from the list of ODA recipients. In the these flows are an important element in maintain- United States the country's largest aid recipient, Is- ing uninterrupted trade flows and mitigating rael, was removed from the list of aid recipients in human suffering, they are temporary in nature and 1997, while the general skepticism about the value specific in their objectives. As worthy as these ob- of aid has limited the ability to rebuild the U.S. aid jectives are, they are unlikely to have a significant program. Germany's aid budget fell by 7.5 percent impact on long-term development goals. in 2000, and the integration of the former East The global war on terrorism is also likely to re- Germany continues to put pressure on the German sult in a temporary increase in aid as donors move federal budget. Japan, which is running a large fis- to offset the economic and humanitarian needs in cal deficit aimed at boosting domestic demand, has countries at the center of conflict. A total of $5 bil- announced a 10 percent cut in the aid budget for lion was pledged for Afghanistan in January 2002, fiscal 2002. although the bulk of this is expected to come from existing aid budgets. Commitments to Afghanistan -although the terrorist attacks on September in 2002 are expected to be almost $2 billion. How- 11 may translate into a short-term increase ever, absorptive capacity is limited and the actual The conflict stemming from the tragic events of inflow to Afghanistan, including the $350 million September 11 is likely to spur a rise in aid in the in emergency assistance already delivered since near term. Donors typically respond rapidly and September 11, is expected to be on the order of $1 generously to disaster-for example in Kosovo billion by the end of 2002. 94 STRENGTHENING OFFICIAL FINANCIAL SUPPORT FOR DEVELOPING COUNTRIES Aid is not always focused on context the end of the Cold War may have im- poverty reduction- proved the opportunities for allocating aid accord- Aid has the greatest impact on poverty reduction ing to poverty alleviation, rather than to strategic when it is provided to countries with good policies criteria. and many poor people (World Bank 1998). All donors made a formal commitment to poverty re- -and the share of aid going to low-income duction by endorsing the international develop- countries is falling- mental targets set out in the Organisation for Eco- The multiplicity of motivations for aid is neither nomic Co-operation and Development (OECD) surprising nor necessarily unfortunate. The use of DAC's Shaping the 21st Century. Most donors aid to further other interests increases popular have policy statements that cite poverty reduction support for aid in donor countries, and may be en- as the, or one of the, overarching goals of their aid tirely consistent with making progress in develop- programs. Trumbull and Wall (1994) estimate that ment. For example, the United States provided ODA allocations are responsive to the needs substantial aid to the Republic of Korea and Tai- of recipient countries, as represented by high levels wan (China) during the 1950s and 1960s, most of infant mortality (as well as issues surrounding likely for strategic reasons. But these countries political-civil rights). Nevertheless, donors have were spectacularly successful in reducing poverty, several motivations for aid that are not always as well. However, the many motivations that un- consistent with allocating aid for the greatest derlie aid allocations may also have some role in poverty impact. Aid may be used to support coun- impairing aid allocation from the standpoint of tries with which the donor has strong historical poverty reduction. The share of aid going to low- connections. For example, Alesina and Dollar income countries has fallen from 61 percent in the (2000) find that aid allocation is greatly influenced early 1980s to 56 percent in the late 1990s. Con- by whether a recipient was a former colony. Aid siderable aid still goes to countries that have ready may be directed at solidifying regional ties; Japan's access to private capital flows, and countries that largest aid program is to countries in Asia. Aid also graduate from aid recipients to Part 11 of the DAC is used to pursue strategic interests: Alesina and list of recipient countries do not always experience Dollar (2000) find that recipients who vote with a reduction of aid flows (an estimated $9 billion donors in the U.N. tend to get more aid, Maizels was given to high-income countries or those on the and Nissanke (1984) relate aid to arms transfers Part 11 list in 2000). Moreover, aid to low-income from the major donors, and Boschini and Olofs- countries with good policies equaled only 1.2 per- gard (2001) explain the decline in aid during the cent of their GDP (see table 4.2), slightly below 1990s as being a byproduct of the end of the Cold the average for other low-income countries. This War. Thus some of the disaffection with the impact ratio has declined sharply since the early 1990s, of aid on poverty reduction does not reflect the in- which reflects the fall in overall aid and rapid eco- trinsic ineffectiveness of aid, but rather the large nomic growth in countries with good policies (as share of aid that is allocated on the basis of their share of aid has been stable). Thus substan- "strategic" criteria, instead of on the basis of the tial progress still is required to ensure that aid is quality of policies and the number of poor. In this directed to countries with good policies. Table 4.2 Trends In aid allocation (percent) ing tooato povert aleiain rahrta 1ttaei Sadth hare of aidgon to low-incomeconre Thcen mutplct ofta moaaiosfrdi)s ete Aidai to furhe othere inteest increase populargpliie Note: Policy performance is measured by Country Portfolio Performance Review prepared by the World Bank. Source: World Bank; OECD. 95 GLOBAL DEVELOPMENT FINANCE The macroeconomic impact of aid risen, but the real resources being purchased have S trengthening aid effectiveness will require con- remained unchanged. In this case aid would be en- tinued progress in allocating aid to countries tirely inflationary. At the other extreme, the central with good policies. But will increasing aid levels to bank may sell the foreign exchange, but the gov- countries with good policies in itself erode the ef- ement does not spend its domestic currency fectiveness of aid? In the poor countries aid levels holding. The extra supply of foreign exchange is an are often large enough to have important macro- infusion of additional real resources to the econ- economic repercussions. Will the marginal produc- omy (as purchasers of foreign exchange use it to tivity of aid (in terms of raising growth rates) de- buy imports); more goods are available, but nomi- cline as the share of aid in economic activity nal demand is unchanged. In this scenario the price increases? Is aid likely to increase inflation, lead level will fall-and aid would be deflationary. Fi- to excessive exchange rate appreciation, or erode nally, if the two decisions are perfectly coordinated the efficiency of government administration? And (the central bank sells all the foreign exchange, and if the answer to any of these questions is yes, then the government spends all the domestic currency), should aid be reduced, or could changes in policy the net effect is to slightly reduce the price level. increase a country's ability to absorb aid produc- This is because the sale of dollars precisely offsets tively? This section concludes that most poor coun- the initial increase in the nominal money supply, tries with good policies should be able to maintain so that the nominal money supply is unaltered. Yet aid effectiveness while absorbing further increases real economic activity is now greater and so the in aid. There is no rationale for constraining aid to demand for real money balances will have risen. countries with good policies because they receive This will be satisfied by a decline in the price level. "too much" aid. Usually the only circumstance in which aid be- comes inflationary is if there is a coordination fail- Aid and the sustainability of fiscal policy ure.9 However, coordination of the two decisions is in the short term simple: expenditures of counterpart funds need to With appropriate economic management, large be matched with sales of reserves. amounts of aid do not increase inflation. Under- standing the potential impact of aid on inflation -which is facilitated by an appropriate requires an appreciation of how aid enters the gov- definition of the government deficit ernment budget. Aid is received by the government It is important to the credibility of government pol- as foreign exchange. The government then, in ef- icy that the definition of the deficit used in discus- fect, sells this foreign exchange to its own central sions of macroeconomic policy reflect the noninfla- bank, which in turn credits the government's ac- tionary impact of aid. Because grants are essentially count in domestic currency (sometimes referred to equivalent to revenue for the purposes of evaluat- as "counterpart funds"). Thus, the central bank ing the inflationary impact of fiscal policy,'0 the ap- now owns the foreign exchange, which it initially propriate definition for the fiscal deficit consistent holds in its reserves; at the same time, the govern- with macroeconomic stability is the deficit after ac- ment now owns the domestic currency, which it counting for aid flows. In the case of concessional initially holds in its account at the central bank. loans, ideally it is the grant component that should be treated as revenue."1 In countries with large aid Aid is not inflationary with policy inflows, different treatments of aid in the fiscal ac- coordination- counts can have a significant impact on the re- If decisions by the central bank and the govern- ported size of the budget deficit. For example, in ment are not coordinated, it is possible for aid to the late 1990s, Ethiopia had a deficit of 8 percent increase inflationary pressures. For example, if the of GDP-if aid were treated as a financing item. government spends the domestic currency (thus in- Recalculated to treat grants and the grant compo- creasing the demand for goods and services in the nent of concessional loans as part of revenue, the economy), but the central bank does not spend deficit was only 0.8 percent of GDP. By contrast, the foreign exchange, then the domestic price level Zimbabwe in the late 1990s received very little aid rises; in other words, nominal expenditures have and had a deficit of 5 percent of GDP. Using the 96 STRENGTHENING OFFICIAL FINANCIAL SUPPORT FOR DEVELOPING COUNTRIE S definition of the deficit that treats aid as a financ- Aid may compensate for other sources of ing item would indicate that Ethiopia's fiscal policy volatility. Guillaumont and Chauvet (2001) find was more inflationary than Zimbabwe's, yet clearly that the effectiveness of aid rises as it is provided to the exact opposite was the case. Much of the countries that are prone to external shocks. There framework for public discussion of fiscal policy is some evidence that multilateral flows to poor comes from ideas articulated by economists and countries help cushion against external shocks by policymakers in industrial countries, where the compensating for withdrawals of private flows (see problem of interpreting the impact of aid on the box 4.3). Collier and Dehn (2001) analyze the ef- fiscal accounts does not arise. Therefore, the defi- fect of aid on growth during periods of negative nition of the budget deficit used in aid-recipient shocks in the context of the aid-growth model de- countries should be such that a level of deficit veloped by Burnside and Dollar (2000). They find deemed to be problematic in OECD countries should that an additional dollar of aid during an extreme be similar to that which signals a policy problem negative shock period raises the growth rate by in aid-recipient countries. Regional groupings of substantially more than in normal periods. By off- African countries are indeed starting to adopt their setting the initial income loss, the aid avoids the own norms analogous to the EU's stability pact, multiplier contraction in output. The magnitude of and it is essential that these norms be based on a these multiplier effects suggests that the rate of re- definition of a deficit that corresponds to economic turn on aid during extreme negative shocks is re- rationality and that produces figures that are well markably high. Aid would be used most effectively understood by the public. in compensating for shocks if care is taken to dis- tinguish between temporary shocks (that should be Volatile aid flows need not translate into financed) and permanent declines in income (that volatile government resources should be adjusted to).13 The international com- Large amounts of aid to the poor countries with munity increasingly recognizes the importance of good policies are unlikely to increase the volatility aid in cushioning external shocks. For example, to of government resources or lead to excessive re- offset the impact of external shocks expected in the liance on aid flows. Lensink and Morrissey (2000) aftermath of September 11, the estimates of low- find that instability of aid resources can have a neg- income countries' possible resource requirements ative effect on growth. Pallage and Robe (1998) during the 13th Replenishment of the International find that aid has been more volatile than recipient Development Association (IDA-13) have been re- countries' output, and aid has been pro-cyclical. vised upward by about $2 billion. However, other empirical work indicates that aid Though aid does not usually increase the does not generally increase the volatility of govern- volatility of resources, it is possible rhat heavy re- ment resources. Since the alternative to receiving aid liance on aid could impose adjustment costs if aid is to finance expenditures through taxation, the ap- were suddenly to decline. There are three circum- propriate benchmark for the volatility of aid is the stances that may cause aid flows to decrease.i4 volatility of revenues. In a sample of 36 African aid First, per capita income in a recipient country can recipients, Collier (1999) found that the coefficient rise sufficiently so that the country is no longer eli- of variation on aid was slightly lower than for rev- gible for aid. There is no need to be cautious of de- enue. Bulir and Hamann (2001), in a global sample pendence on aid while the economy is poor, just be- of aid recipients, find that aid is more volatile than cause one day it will be sufficiently rich that it will tax revenues (with both expressed in U.S. dollars), no longer need any aid. Moreover, higher income is but the difference was not statistically significant.12 associated with a greater ability to finance expendi- If aid and tax revenue are almost equally volatile tures from taxes; in 1998 current revenue equaled (for example, in U.S. dollars) then unless aid and 14 percent of GDP in low-income countries, 19 per- tax are perfectly correlated, aid must reduce overall cent in middle-income countries, and 29 percent in volatility. Collier (1999) found a slight negative cor- high-income countries. Second, aid may be cut off relation between aid flows and revenues. In that because economic policy deteriorates substantially; case the addition of aid to revenues actually reduces however, this is not a reason for a country with the volatility of overall resources. good policies to refuse aid. Finally, donors may 97 GLOBAL DEVELOPMENT FINANCE Box 4.3 The relationship bRi multilateral flows in poor co ulrlareral fl(o% v to poor .ountries appear to have .I i gunre bll. en ever ir the kng trm AL i1 lo" in%erst relationship to pri art: lowi. There are ,.iri ri v le toLI an Im vn in h prhg 1 A L ouN InterprLtrions iii the econtmic ltEraturc of this rel - 9:nstitu'ionl en,Ir7nmnrt of a counr. w:hilh w i d n- tionship in the context ot Ill de.-eloping countries. Das- : ourage greater private flows. Several rh gupra anJ Ratha i20i)(0i argue that multilateral lending l empirival upport for - eaal.vric ef mu pla% a stabilizing role during periods of credit ratioriing. ateral floi on r.t ate flows. Iharas and Shishido (199'I LErrick i I1 sees this relationship as e% idence that o 1und tha durin 194-35. b alleviaring Mrdir rationing multilateral tlows cro% d out pr'.ate flow%s. Easterk . and imp 0 ing 0reJironithinLs by inreasng int-rna3on.l (1 4'"i and Senion i( argue that multrlatera led reser. es, for eX mplb ing programs creare incenmes ior horro%%iing 1oZernmL pdlover etteets that a Ita]wd rivate w,s. See also to dela% economic reforns, so that pri%ate lenders with- kruegr 198 Summers 1999; and heckI'rnd Stirn 20'0 dra%% in reaction to increased multilateral loans. This relativnshi is borne o t by stathi.ual tePt. PaneI The intrse relationship beteen multilateral and pri daa analses for lw-inKome ioun mes (Ifor he period vate flos. hotieer, nced not npl "cruwJmg-out of 19 0-- 1 indicate inegarie relationship betweei multi- pri ate tlows to de%eloping countries. Indecd an ierse I rral and priate flo in the same period t a id.te lctronship in the short term mrnav be consrent % itb a con rlarionZ3u a si.x-ea g contrat. bIateral 110. plenientar. relationship over the long rcrm. With respect to' Iinn.rants5 sLenm tl e aa signitan andnpo N e short-term c%clical ariables 1for example, an increaise m ef atofpr.ivat flows in the nourrent period, but a ne GL)P grl:,%%rh or an interc,t r.u,e hike in the industrial cooni- i ect with a lag. T.h0 result ma etle-r th o r tries), priiate 110ws tend to behave prochcrally Wrld tstra gi' and none5nomicoonsiderauon' Inad Bank 201:ii1a)' herva.; uif-,cial tlo%% s ire e' pected to react a 1I7-.,1:1Et iWarerl I ono'rs Alesinad J 011ar, -0)i1. sharply reduce levels of aid for reasons unrelated to and Reichel 1995).15 However, Burnside and Dol- the recipients, for example because donors con- lar (2000), Collier and Dollar (2001a), and Dur- front widespread fiscal difficulties. Changes in aid harry, Gemmell, and Greenaway (1998) show that flows tend to be implemented slowly, and it is un- aid makes an effective contribution to growth in likely that any such reduction in aid would present countries with good economic policies.16 The ex- very sharp adjustment costs to individual develop- tent of the impact on growth can be seen by look- ing countries. Nevertheless, this concern does un- ing at IDA, which is well targeted on low-income derscore the importance of donors providing for countries with reasonable policies. At the margin, stable aid flows over time. an additional billion dollars of IDA funds raises the growth rate sufficiently to lift around 434,000 Aid has a positive impact on growth in people out of poverty.17 Collier and Dollar (2001b) countries with good policies- find that in good policy environments aid raises in- So far we have shown that there is little reason to vestment by almost double the value of the aid; worry about the adverse impact of aid on the sus- Collier and Dollar (2001c) also find that in good tainability of economic policies in countries whose policy environments a $1 billion injection of aid economic policies are sound. We now turn to the raises FDI by $600 million. question of whether increases in aid are likely to continue to have a positive impact on growth. -although approptiate policies may be There is growing evidence that aid has a positive necessary to limit "Dutch disease" effects- impact on growth in countries with good policies. The finding that on average aid has had a positive Earlier empirical studies had consistently found a impact on growth in good policy environments weak relationship between aid and investment and does not imply that aid levels can rise forever showed little impact of aid on growth (see, for ex- without a resulting adverse effect on growth. In- ample, Griffin 1971; Snyder 1990; Boone 1994; creasing levels of aid may erode growth by causing 98 STRENGTHENING OFFICIAL FINANCIAL SUPPORT FOR DEVELOPING COUNTRIES "Dutch disease." Since aid is foreign exchange, it -and access to large nontax resources may only directly augments the supply of those goods erode government accountability that are internationally tradable. It will thus lower The productivity of aid may decline due to reasons their equilibrium price relative to those goods that other than the Dutch disease. It may be possible can only be traded domestically (nontradables). for governments to have more resources than are This relative price change induces a resource shift good for their societies. Access to very large non- in the economy from tradables to nontradables. tax resources can erode the accountability of gov- Among the tradables are exports, so that aid will ement. Indeed, the history of accountable gov- tend, all things being equal, to reduce exports. In ernments in the now-developed societies dates fact, other things are not equal. The aid may en- from the need for governments to raise tax revenue able governments to lower taxes on exports, (see, for example, Hoffman and Norberg 1994). which in the poor countries is typically the most Similarly, Sachs and Warner (2000) establish that heavily taxed sector. Additionally, aid might fi- governance is worse in countries where the govern- nance infrastructure expenditures that facilitate ment has access to large rents from natural me- exports, such as roads and ports. However, it sources. Consistent with this theory, Knack (2000) seems reasonable to expect that in most circum- finds that aid tends to be associated with increased stances aid will indeed reduce exports. Van Wijn- corruption. On the other hand, Burnside and Dol- bergen (1986) found that increases in aid were as- lar (2000) and Dollar and Svensson (2000) found sociated with an appreciation of the real exchange that aid neither improved nor worsened policies. rate in African countries. Several empirical studies This is disappointing because it implies that aid present evidence of the adverse impact of the may not induce reform; on the other hand, it ndi- Dutch disease on exports (see, for example, La- Cates that aid does not appear to cause a general- plagne, Treadgold, and Baldry 2001; Soderling ized deterioration in economic policies. 2000; and Sekkat and Varoudakis 2000). Collier A more likely reason for diminishing returns and Hoeffler (2000) show that, controlling for the to aid is administrative and managerial conges- level of economic policy as measured by the World ton. If the really scarce resource in the public sec- Bank's Country Portfolio Performance Review, a tot is competent and motivated civil servants, then rise in aid is associated with a decline in the share each additional aid project, in competing for the of primary commodity exports in GDP. Since for same limited pool of skills, inflicts negative exter- Africa these exports still make up around 70 per- nalities on other projects. Beyond a point, these cent of all merchandise exports, it is likely that aid congestion effects can fully offset the direct bene- in Africa reduces total exports.18 fits of the project. Similarly, Taslim and Weliwita The question remains, is a decline in exports (2000) argue that both public and private invest- caused by aid-induced real exchange rate apprecia- ments in developing countries are limited by the tion undesirable? It should be recognized that the stock of entrepreneurial skills, so that increased Dutch disease is more of a problem if the aid flow aid is reflected in reduced saving. is short-lived, so that adjustment costs are in- curred when aid flows in and when it ceases. But The marginal productivity of aid depends aid to the poor countries is rarely a matter of a few upon the quality of policies- years, and thus the value of aid will be greater Aid is likely to be subject to diminishing returns." than any distortionary effects due to real exchange The Collier and Dollar (2001a) results indicate, rate appreciation. The reallocation of resources howevei that the level of aid where the marginal out of tradables could be undesirable if either ex- productivity is zero depends on the quality of poli- ports are initially too low because of taxation, or cies, and this level is quite high for countries with because exports raise growth through learning and good policies. Countries with the highest score on competition effects that enhance productivity; the World Bank's Country Portfolio Performance Kraay (1999) finds some evidence of this for Review (CPPR) continue to enjoy aid's positive im- China, and Bigsten and others (1999) for Africa. pact on growth at levels of aid up to 30 percent of However, a more rational response to these prob- gross domestic product (GDP). Durbarry, Gem- lems would be to use aid to reduce taxation or to mell, and Greenaway (1998) find that aid contin- finance infrastructure facilities that help exporters. ues to make a significant contribution to growth 99 GLOBAL DEVELOPMENI FINANCE up to 40 percent of GDP in countries with a stable poverty by 2015 (20 percent of the population in macroeconomic policy environment.20 The median poverty in 1999). CPPR among poor countries is 3.2, at which level These estimates of the impact of aid are con- (by the Collier and Dollar estimations) the impact servative. They assume that donors have no im- of aid on growth would remain positive up to 19 pact on the quality of policies or the elasticity of percent of GDP, while aid averages 8 percent of poverty reduction with respect to growth. It may GDP for poor countries with better than average be true that donors have had only a limited impact ratings. By these calculations, 28 out of the 34 on policies, and that aid is often fungible (so that poor countries with better than average policies the kind of projects financed would not affect the could continue to absorb increasing amounts of poverty elasticity). However, a recent study of aid aid before the marginal productivity of aid drops and reform in Africa concludes that donors could to zero.21 have a more systematic impact on policy if they in- Recent calculations indicate that a doubling of creased aid as policies improved (World Bank aid will be necessary to reach the goal of halving 2001c), which is the allocation rule used in this the share of the developing-country population simulation. Further, if the improvement in policies that lives on less than $1 a day by 2015 (World is reflected in better provision of public services Bank 2001b). But improvements in the allocation that benefit the poor, then countries with good of aid are also critical to achieving the poverty policies will have higher elasticities of poverty re- goal. Collier and Dollar (2001a) develop a model duction with respect to growth. Thus the impact for allocating aid that reflects the view that the im- on poverty of a doubling of aid, allocated accord- pact of aid on poverty depends on the quality of ing to policies and the extent of poverty, is likely policies. to be larger than assumed in this simulation. A doubling of aid that is distributed according to quality of policies and the level of poverty im- -so aid efficiency can be improved plies significant changes in aid allocation. South Thus recent econometric evidence indeed suggests Asia would receive an increase in the share of total that countries can receive too much aid. The most aid from 11 percent in 1999 to 45 percent.22 The likely explanation for this is neither the Dutch dis- largely middle-income regions of Europe and Cen- ease, nor the deterioration of governance, but the tral Asia, Latin America and the Caribbean, and high congestion costs incurred by attempting to im- the Middle East and North Africa would together plement many aid projects through a bureaucracy receive only 4 percent of total aid, compared with with limited capacity. If this analysis is correct, it about a third in 1999. The share of East Asia and has five important implications: First, in countries Pacific would decline slightly, because the middle- with good policies, actual aid inflows are unlikely income countries receive much less aid, but aid to be near the point where the marginal productiv- would expand sharply to Vietnam and the Philip- ity of aid is zero (the saturation point). Second, in pines due to their relative poverty and good poli- those poor countries that currently are close to or cies. Finally, the share of aid going to Sub-Saharan beyond their saturation points, the key task is to Africa would change very little, because some of raise the saturation point by improving policy. the better performers would receive significant in- Third, aid programs should aim to reduce conges- creases but other countries with very poor policies tion costs. Switching more aid from projects to would experience an actual decline in aid flows. programs would almost certainly raise absorptive The increases in aid-to-GDP levels are modest for capacity. Fourth, to the extent that the capacity most countries, and for all of the countries with constraint is due to a lack of competent and mot- good policies aid remains well below the level vated civil servants, incentive systems in the public where the marginal productivity of aid falls to sector may need revision. Fifth, if the public sector zero. In Sub-Saharan Africa, the region with the faces real constraints upon its capacity to spend highest level of aid relative to GDP, the average marginal resources effectively, it should reduce tax ratio of aid to GDP would rise only slightly. Fi- receipts relative to aid. While aid augments the re- nally, the doubling of aid would lift an estimated sources available to the economy, taxation reduces 15 million people permanently out of poverty each them by introducing distortions (for example, in- year, for a total decline of 225 million people in creased income taxes may reduce the incentive to 100 STRENGTHENING OFFICIAL FINANCIAL SUPPORT FOR DEVELOPING COUNTRIES work). A sensible growth strategy for a very low- income economy with a dysfunctional civil service Figure 4.2 Compliance with conditionality would be for rising aid inflows to be used partly for and economic performance reducing the share of tax revenue in GDP. Countries showing improvement (percent) 80 70 -_____ Conditionality and adjustment - lending 50 Strengthening the use of policy conditionality in adjustment lending is an important element in efforts to improve the effectiveness of aid (see World Bank 2001d). Policy conditionality refers to 20 the practice of basing the disbursement of donor 10 funds on the implementation of specific policies. 0 Policy conditionality can support the effectiveness Growth Inflation GDI of adjustment assistance by helping to avoid dis- Performance indicator bursements to governments with inappropriate N Strong ompliers X Weak complers policies. For recipients, agreement on specific con- 0 Poor compliers ditions for disbursement (as opposed to basing dis- bursement on a general evaluation of the govern- Sourre: World Bank 1997. ment's program) can improve the transparency of donor decisions and the reliability of aid disburse- ments (Mosley 1999). By increasing the cost of backtracking on policies (in terms of worsening re- there is no consensus for reform." Conditions at- lations with donors or losing disbursements), com- tached to adjustment lending may not contribute mitments to donors can enhance the government's to successful outcomes in cases where donors lack credibility in sticking to policies that face opposi- adequate information (on local conditions, govern- tion from special interests or that have short-term ment capacity, and the extent of government com- costs but long-term benefits. Case studies of the mitment) or the interests of donors and recipients strong reform programs in Ghana and Uganda diverge. Conditionality is the outcome of a bargain- suggest that conditionality was successful at facili- ing process that can be subject to failures of coordi- tating clear decisions from political leadership and nation and unintended outcomes.25 To the extent publicly signaling the government's commitment that this process leads to a reform program that is (World Bank 2001c). In turn, enhancing credibility not fully owned by the government, the success of can encourage more rapid adjustments to new the program can be severely undermined. Domestic policies by the private sector and hence reduce the political support is critical for the adjustment pro- short-term employment and output costs of ad- gram (Rodrik 1996; World Bank 1998; Dollar and justment. Greater compliance with conditionality Easterly 1998; Dollar and Svensson 2000). Both under World Bank loans was significantly related cross-country reviews and individual case studies to improved economic performance (figure 4.2).23 have confirmed the critical importance of strong country ownership of the adjustment program to Country ownership is key to success- the successful use of conditionality in adjustment A country's commitment and capacity to imple- lending (McClearly 1991; Berg 1991). Johnson and ment the reforms supported by adjustment lending Wasty (1993) find that strong ownership was a are key to effective adjustment and sustained devel- major reason for success in 75 percent of adjust- opment. Research on aid effectiveness indicates ment programs with good results. The Interna- that when a country's commitment or implementa- tional Monetary Fund (IMF 1998) attributed poor tion capacity is weak, conditionality is unlikely to implementation of IMF programs in Zambia be effective. In other words, conditionality by itself (1978-91) and Uganda (late 1980s) to lack of own- cannot lead to the adoption of better policies when ership; these are in contrast to successes in Bolivia, 101 GLOBAL DEVELOPMENT FINANCE Uganda in the 1990s, and C6te d'Ivoire, where ule.21 World Bank (1997) found that out of 35 ad- ownership was strong. justment operations in Sub-Saharan Africa, com- pliance was rated as strong in 10 countries, and as Conditionality and World Bank adjustment weak or poor in 25 countries. Indeed, the perfor- lending have evolved- mance of World Bank adjustment lending im- Conditionality was originally directed largely at proved sharply throughout the 1990s. Operations achieving macroeconomic stability and reducing Evaluation Department outcome scores increased market distortions, and adjustment assistance was from 60 percent satisfactory in the 1980s to 68 conceived as a financing vehicle for short-term bal- percent satisfactory in fiscal 1990-94, and to 86 ance of payments support. Over the years, the percent satisfactory in fiscal 1999-20002 The policies covered by conditionality and the goals of World Bank's Quality Assurance Group found that adjustment lending have evolved in tandem with the great majority of a sample of adjustment loans countries' broader reform agendas, and have be- in 1999 were satisfactory or better regarding van- come increasingly focused on long-run, structural, ous dimensions of program design (World Bank social, and institutional issues. The 1980s' narrow 2000b). Bilateral aid evaluations also typically focus on short-term stabilization and addressing find satisfactory outcomes for a high proportion distortions gave way in the 1990s to greater atten- of adjustment programs (see, for example, USAID tion to poverty reduction, institutions, and complex 2001; SIDA 1999). social and structural reforms. This shift included an It is of course difficult to attribute improved explicit focus on good governance, with strong sup- compliance wholly to improvements in the design port for public sector management reforms. of conditionality. There are several reasons why Reflecting in part the growing long-run struc- adjustment programs were more successful during tural and institutional focus of countries' reform the 1990s, including a more favorable interna- agendas, Bank-supported adjustment programs tional economic environment (at least in some have grown more complex, even while the average years), greater selectivity on the part of the donors, number of conditions in adjustment loans has fallen and greater recognition of the importance of gov- significantly, from 61 conditions in the late 1980s ement ownership in crafting an effective adjust- to 33 conditions in fiscal 2000. The number of con- ment program. It is likely that changes in the ditions tends to be higher and complexity tends to process of adjustment lending, including greater be a greater challenge in countries with weak per- selectivity and encouraging ownership through a formance and capacity, where adjustment lending is less intrusive approach to the design of reform less successful (World Bank 2001d). This highlights programs, was at least as important as the change the ineffectiveness of attempts to address perfor- in the focus of conditionality to address underly- mance deficiencies and capacity limitations through ing structural, social, and institutional issues. What a larger number of more complex and detailed con- is clear is that changes in the overall approach to ditions, and confirms the importance of continuing adjustment assistance have contributed to more to focus adjustment support in countries with good successful reform programs. policy and institutional environments. Aid coordination and the administrative -and the quality of Bank adjustment lending burden of aid has improved The idea that donors could increase the effective- The record of policy conditionality in promoting ness of aid by improving the coordination of their the objectives of adjustment programs, as reflected activities is not new (Pearson 1969). Donors have in the degree of compliance with agreed-on condi- made extensive efforts to consult on their aid opera- tions, has improved in recent years. The problems tions and thus avoid the imposition of conflicting or affecting conditionality in the 1980s have been duplicative administrative requirements, and they well documented.26 Some of these problems may have improved the quality and consistency of policy have persisted into the early 1990s. Killick, Guna- advice, most notably through consultative group tilaka, and Marr (1998) find that only 25 percent meetings, round tables sponsored by the United Na- of World Bank adjustment operations from tions Development Programme, aid meetings under 1989-90 to 1993-94 were completed on sched- the auspices of the OECD DAC, the U.N.'s Devel- 102 STRENGTHENING OFFICIAL FINANCIAL SUPPORT FOR DEVELOPING COUNTRIES opment Assistance Framework (which harmonizes tying requirements limit competition, increase ad- U.N. agencies' activities) and the Strategic Partner- ministrative burdens, and lead to countries pur- ship for Africa. Successive IDA replenishment re- chasing goods with an inappropriate technology ports during the 1990s urged greater efforts at coor- with greater than desired capital intensity. The ad- dination. Considerable work remains to strengthen ditional cost imposed by tying aid has been esti- aid coordination, which is particularly important in mated in the range of 10-30 percent (OECD 2001; the poorest aid recipients that receive very signifi- Morrissey and White 1994; and Jepma 1991). cant levels of aid relative to domestic resources. There are also significant indirect costs, including suspension of standard procurement procedures Reducing administrative burdens- and higher cost maintenance due to dependence on Aid often imposes a substantial administrative bur- imported parts that may not be readily available. den on recipient governments (Van Arkadie 1986; Considerable progress has been made to re- Lister and Stevens 1992). Van de Walle and John- duce tied aid requirements, and the share of bilat- ston (1996) report that Kenya had 600 projects eral aid that is tied has dropped from 65 percent in from 60 different donors during the mid-1980s, 1990 to 38 percent in 2000, though there is con- while Zambia had 614 projects from 69 donors. siderable variation across donors. The share of In Tanzania there were even more-over 2,000 tied aid to the least developed countries is about projects from 40 donors. Administratively, 600 50 percent, higher than the average for all devel- projects could translate over the course of a year oping countries primarily because these countries into as many as 2,400 quarterly reports for various receive more of the type of aid that is still subject oversight agencies and perhaps 1,000 missions re- to tying (for example, food aid and technical assis- questing meetings with key officials and comments tance). The DAC High Level meeting in May 2001 on their reports. Disch (2000, p. 39) describes the agreed on a recommendation to untie ODA to the multiplicity of import support programs in least developed countries to the extent that is pos- Mozambique in the late 1980s, each with different sible. By January 2002 many important compo- procedures and time delays that typically took six nents of ODA to the least developed countries will to nine months for importers to navigate. The re- be untied, including balance of payments support sult: skyrocketing import costs. Donors have com- and debt forgiveness. The OECD estimates that peted with each other and with the government to this will raise the level of untied aid to the least de- recruit scarce local experts for projects, thus under- veloped countries to 70 percent. mining the government's capacity (Eisenblatter 1999). Lancaster (1999, p. 501) notes the implica- Changes in process can strengthen aid tions for budget management of uncoordinated coordination and reduce administrative donor projects negotiated with individual min- burdens istries, each demanding counterpart funding for re- Procedurally, a number of different strategies for current costs. In addition to administrative bur- improving coordination have been advanced, in- dens, failures in aid coordination can result in cluding sectorwide approaches, greater donor spe- donors pressing inconsistent policy advice on gov- cialization, more support for capacity building, ernments. For example, in the mid-1980s the and greater flexibility in some donor requirements. World Bank and the United States Agency for In- Sectorwide approaches can facilitate country own- ternational Development urged the Kenyan govern- ership by reducing micromanagement by donors ment to reduce the role of the National Cereals and and by eliciting longer-term commitments from Produce Board at the same time as another donor both sides to help build genuine partnerships. was financing a major expansion of its facilities For donors, sectorwide approaches offer a realistic (Mosley 1986). compromise between detailed micromanagement and provision of general budget support, since re- -and shifting away from tied aid- sponsible ministries may be held accountable for One of the better-known impediments to aid effec- results. Sectorwide approaches are most appropri- tiveness is tied aid, which often reflects donors' ate when both macro and sector reform processes commercial interests rather than recipients' devel- are in place and when governments have a clear vi- opment needs. Various studies have found that sion and ownership of objectives. In Uganda, for 103 GLOBAL DEVELOPMENT FINANCE example, strong government motivation, active dening recipient governments. Greater delegation participation by civil society in program monitor- of decisionmaking authority to the field would also ing, and a credible medium-term budget process facilitate better coordination (World Bank 2001f). made the Universal Primary Education project a success (Brown and others 2001). However, sector Above all, government leadership is the key finance is likely to be ineffective if either sector Strong leadership from the recipient government is policies or macroeconomic and budget management essential for successful aid coordination (Eisenblat- are weak. In addition, sectorwide approaches may ter 1999). For example, Botswana, the fastest- limit government's ability to reallocate funds across growing country in Africa for some time (and in sectors, compared with disbursing aid through bud- many years the fastest-growing country in the get support programs. world) has had the vision and capacity to manage Greater donor specialization is needed. The the aid process (Brautigam and Botchwey 1995). In difficulties of aid coordination increase sharply as Botswana the government maintains effective con- more donors become involved in any one area, so trol of aid with strong institutions backing up a co- specialization along geographic or functional lines herent vision. Donors are encouraged to specialize according to comparative advantage is desirable. in specific sectors to build up their expertise and Yet the trend has been toward increasing diffusion minimize administrative burdens (van de Walle and of donor activities, and World Bank (1999) found Johnston 1996). Likewise, the governments of few examples of aid coordination efforts that led Ghana and Uganda, two of the more successful re- to greater specialization (see also World Bank formers in Africa, have played an active role in co- 2001e). Reviewing aid to Ghana in the first half of ordinating donor activities. the 1990s, Eriksson (2001) found a steady increase in the number of sectors for each bilateral donor and a decline in bilateral commitments per sector. Capacity building is one key to progress. Lim- Aid and debt relief ited capacity and institutional weakness impede the formulation of country-owned strategies, and Strengthening the effectiveness of aid through undermine the trust donors need to allow coun- debt relief- tries to take responsibility for detailed financial The increase in concessional debt relief, and ef- and project management. Yet capacity building forts to tie debt relief to effective reform programs, has been one of the least effective areas of donor have been important components of efforts to activity, and in many of the world's poorest coun- strengthen the effectiveness of aid. Debt reduction tries the quality of public administration has sys- in the form of concessional rescheduling of guaran- tematically deteriorated (Lancaster 1999). Some teed commercial claims began in 1988 with the in- donor practices may have even contributed to the troduction of Toronto terms by the Paris Club, problem through insistence on special project which allowed for a reduction of one-third for management units that draw government officials eligible claims. The level of debt forgiveness has from their regular duties, and recruitment of expa- subsequently been raised progressively, to 50 per- triate technical assistance personnel whose terms cent reduction (in net present value [NPVI terms) of reference are to substitute for local capacity in 1991 (London or enhanced Toronto terms), rather than to build it. Regular civil service staff and 67 percent NPV reduction in 1994 (Naples assigned to projects may be expected to give prior- terms).29 Donors forgave bilateral ODA claims, fi- ity to project work even if there is a conflict with nanced debt swaps, contributed to the buyback of their normal responsibilities (Lancaster 1999; van commercial debt through the IDA debt reduction de Walle and Johnston 1996). facility program, and supported programs to help More flexibility by some donor agencies is debtor countries meet multilateral debt service needed to transfer responsibility and accountabil- obligations. Efforts to deepen debt relief for poor ity to recipients. Incompatible procedures for re- countries suffering from unsustainable debt but- porting, accounting for disbursements, and pro- dens culminated in the HIPC Initiative. All in all, curement raise transaction costs and inhibit closer DAC donors have forgiven about $29 billion in coordination among donors, while severely bur- debt over the past 30 years. Of this total, forgive- 104 STRENGTHENING OFFICIAL FINANCIAL SUPPORT FOR DEVELOPING COUNTRIES Table 4.3 Forgiveness of ODA claims, 1970-2000 (millions of dollars) HIPCs .PS.i i, 5I 4i iC,f Other developing coun. Source: OECD DAC, national aid agencies, and staff estimates. ness of ODA loans by DAC donors has amounted ness. The pace of delivery of debt relief increased to almost $20 billion (see table 4.3), and donors in 2001. All countries that reached their decision have claimed credit in their aid budgets for the for- points by the end of 2000 are now receiving in- giveness of $8.5 billion in non-ODA claims, and terim relief, and their aggregate level of debt is have provided almost $400 million in grants in forecast to fall from 60 percent of GDP in 1999 to support of the IDA debt reduction facility. How- 28 percent after debt relief. Current plans call for ever, the figures recorded by the DAC probably a reduction in debt service obligations by one- underestimate the full extent of the debt relief, be- third ($1.1 billion) during 2001-03,11 for an aver- cause they do not include irrevocable commitments age savings of close to $50 million per Country per to forgive future ODA claims, while for non-ODA year. Debt service as a percentage of exports for claims the reporting norms are complex and have the 24 countries is expected to decrease from 16.8 taken time to be fully integrated into the statistical percent in 1998-99 to 8.2 percent in 2001-03. systems of the export credit agencies. -while 4 of these countries have reached the -as 24 countries have reached the decision completion point point under the HIPC Initiative- As of December 2001 four countries (Bolivia, Mo- The HIPC Initiative, launched in 1996, aims to in- zambique, Tanzania, and Uganda) had reached the crease the effectiveness of aid by helping poor completion point, where the remainder of the countries achieve sustainable levels of debt while committed debt relief is delivered. For example, strengthening the link between debt relief and Mozambique reached its completion point in Sep- strong policy performance. Forty-two countries, primarily from the Sub-Saharan Africa region, are identified as eligible to receive debt relief under this Figure 4.3 NPV of external debt of the 24 countries that initiative. In 1999 the scope of the initiative was reached their HIPC decision point widened to accelerate and deepen the provision of Billions of dollars (decision point terms) debt relief. As of December 2001, 24 countries 60 have reached the decision point30 (the point where debt relief is approved by the Executive Boards of 50 the IMF and the World Bank, and interim relief be- gins). These countries are now receiving debt ser- 40 vice relief which will amount to about $36 billion _ over time, a $21 billion reduction in the NPV of 21 their outstanding debt stock (see figure 4.3). 20 -resulting in a halving of the NPV of their 10 external debt- 0 The 24 countries that have reached their decision NPV befote NPV after NPV after NPV before points have experienced a halving of their external traditional relief tradiional relief HIPC relief additional bilateral stock of debt in NPV terms. When combined with debt forgiveness other debt reduction mechanisms, this implies a Source: World Bank two-thirds reduction in their external indebted- 105 GLOBAL DEVELOPMENT FINANCE Table 4.4 Impact of HIPC Initiative in 24 decision-point cases NPV of total ex..--1 J, hr Debt as a percent .. .1 I' Average debt ser ,c i L1r.u pr . . r r . Average debt set - r I Average debt set ,c i . .n i,r F r.,' Average social s J.. . .. ..r. r C.. il Average social se....... r .-.7 7. .-..-iu Source: World Bank. tember 2001, and will receive debt relief amount- since 1996. It should be noted, however, that ODA ing to $4.3 billion, which will cut its debt by 72 to the better performers excluding debt relief has percent (in NPV terms). As a result, Mozambique's declined by 2 percent since 1996. The HIPC Initia- annual debt service payments will be reduced to an tive has been essential to place beneficiary coun- average of 6 percent of export earnings and 10 tries on a path to long-term debt sustainability12 percent of government revenue over 2000-10, as and has resulted in increased resources, as shown compared with .20 and 23 percent, respectively, in by the decline in actual debt service payments rela- 1998. Another dozen countries could reach their tive to earlier years. Even countries with significant completion point in 2002. payments arrears received an important, if more modest, increase in new financial resources, while HIPC has helped provide a more effective the HIPC Initiative also will help normalize their environment for aid- relations with creditors. Nevertheless, it is of criti- The HIPC Initiative, in addition to increasing re- cal importance that donors maintain their ODA ef- sources for debt relief, has helped to support pol- fort in the form of new money as well as debt relief, icy improvements and thereby contributed to aid particularly as the expected supply response to effectiveness. Debt relief under the HIPC Initiative lowering debt levels may take some time to occur. is intended for countries that are pursuing effective poverty reduction strategies, and increased social However, creditors need to continue to deliver expenditures is a critical element. For the countries on HIPC that have reached decision points under the HIPC Full participation by all creditors is essential to en- Initiative, social expenditures are projected to in- sure that the 24 countries already at decision points crease about 1.1 percent of GDP compared with reach sustainable external debt levels and, more 1998-99 (table 4.4). broadly, to ensure that the HIPC Initiative achieves its objectives in full. While most bilateral credi- -which is reflected in ODA flows tors-including all Paris Club creditors-and the There is some evidence that ODA flows to the majority of multilateral and commercial creditors HIPCs are being allocated to the better performers, have already been delivering on their commitments a prerequisite for aid effectiveness. Countries that to provide relief to HIPCs, a number of creditors have either reached a decision point (indicating have not. In particular, some of the non-Paris Club general agreement with donors on the economic official bilateral and commercial creditors (repre- program) or have sustainable levels of debt (indi- senting about 10 percent of the debt relief to be cating that their policies were adequate to achieve delivered) along with a few multilateral creditors sustainable debt levels with traditional debt relief have not yet agreed to provide relief to the coun- mechanisms) observed an increase of 3 percent in tries that have reached their decision points under gross ODA flows since the initiation of the pro- the Initiative. Indeed, a small number of creditors gram in 1996. This is in marked contrast to ODA have resorted to litigation as a means of recovering flows to countries with unsustainable debt levels assets; of those, there are a few cases where claims that have not yet reached a decision point; in those of official bilateral or commercial creditors have countries, gross ODA has fallen by more than half been bought on the secondary market at a discount 106 STRENGTHENING OFFICIAL FINANCIAL SUPPORT FOR DEVELOPING COUNTRIES in order to maximize recovery through litigation.33 cies' new commitments to developing countries Given the relatively small number of creditors in- rose to an average $75 billion a year in the first half volved, these problems will not likely undo the of the 1990s (mirroring the steep rise in private achievements of the HIPC Initiative. However, the flows), and then declined in the wake of the Asian litigation alone could prove to be very costly for in- crisis.34 Nevertheless, new commitments remained dividual HIPCs in terms of legal representation and at $50 billion in 2000, or 40 percent of all commit- the implications of adverse judgments. ments from private creditors, excluding bonds. Export credit agencies have become increas- Postconflict countries present a special ingly more involved in investment insurance.35 The challenge Berne Union member agencies extended $13 bil- The most important challenge for the HIPC Initia- lion of insurance against FDI projects in develop- tive in the year ahead is to bring the remaining eli- ing countries in 2000 (five times more than in gible countries to their decision points as quickly 1990), and the total investment under cover by as possible, so that these countries can begin to re- member agencies (the outstanding exposure or ceive debt relief. This challenge presents special dif- stock) rose to $58 billion at end-2000, compared ficulties since most of these countries have recently with $9 billion in 1990. This strong growth in in- emerged from, or are still engaged in, armed con- vestment insurance mirrors the surge in direct in- flict, and many of them are struggling with gover- vestment flows (investment insurance by Berne nance issues. At the same time, these countries Union members has covered on average around 12 have a particularly acute need for debt relief be- percent of the FDI flows to developing countries) cause of their major reconstruction requirements and has been important in privatization and pri- and the urgent need for speedy and effective action vate sector involvement in the provision of infra- to help break the cycle of violence, low growth, structure services. and severe poverty. The framework of the HIPC Initiative has the Multilateral institutions are expanding their flexibility to front-load assistance to countries af- guarantee activities fected by conflict, and a relatively large share of Multilateral institutions also expanded their guar- debt relief can be made available at an early stage, antee activities during the 1990s. The guarantee taking into account the profile of debt service pay- programs of the World Bank Group, which are in- ments due and the absorptive capacity of the coun- tended to serve as a catalyst for private sector activ- try. To ensure progress toward sustainable growth, ities in developing countries, supported $18 billion the structural and social triggers for the comple- in flows in the second half of the 1990s, double tion point will be customized to reflect the particu- the level of guarantees extended in the period lar set of priorities and needs of the postconflict 1990-95. Moreover, the financing leveraged by countries. For example, improvement in fiscal these guarantees is substantial: International Bank management and demobilization of excombatants for Reconstruction and Development partial credit were part of the completion point conditions for and partial risk guarantees of $2 billion helped Guinea-Bissau. galvanize almost $20 billion in total project costs. In poor counties, partial risk guarantees from IDA help insure private lenders against country risks that are beyond the control of investors. To date, Strengthening the effectiveness of three countries-Bangladesh, C6te d1voire, and official guarantees Uganda-have benefited from an IDA partial risk In addition to aid flows, official agencies channelguarantee for a power project. The three guaran- resources to developing countries through guar- tees total $206 million, and the aggregate project antees of private sector loans and investments. Ex- costs are $1 billion. The Multilateral Investment port credit agencies' total exposure to developing Guarantee Agency (MIGA) is in the forefront of ef- countries reached an estimated $500 billion at the forts to facilitate investment in poor countries and end of 2000-one-quarter of developing countries' to ensure that projects have a significant develop- total long-term external debt. Export credit agen- mental impact. Since 1988, MIGA has issued 550 107 GLOBAL DEVELOPMENT FINANCE guarantees for projects in 79 developing countries. by export credit agencies in 1990, while today only Total coverage issued exceeds $9 billion, bringing 25 percent is guaranteed. Similar trends are evident the estimated amount of foreign direct investment for Latin American borrowers such as Chile and facilitated since inception to more than $42 billion. Brazil, and for Malaysia and Thailand prior to the Poor countries accounted for over 20 percent of 1997 crisis.37 MIGA's gross portfolio on June 20, 2001, spread across 26 countries. The regional development -but these facilities have also increased poor banks, including the Inter-American Development countries' debt Bank, the Asian Development Bank, the European While export credit agencies have made an impor- Bank for Reconstruction and Development, and tant contribution to boosting the teal resources some of the smaller regional banks, have also de- available to poor countries, access to guaranteed fi- veloped wide-ranging guarantee programs. nance also has contributed to unsustainable debt burdens. During the past decade, the HIPC coun- Poor countries rely on guarantees for large tries have received almost $20 billion in loan external financial commitments- commitments guaranteed by export credit agencies, Official guarantees have supported a limited vol- and export credit commitments to HIPCs averaged ume of finance to the poor countries, compared $1.8 billion per year from 1990-96, when the HIPC with other developing countries. The export credit Initiative began. Since then, steps have been taken agencies' total exposure to the poor countries to ensure that the debt reduction under the HIPC equals $40 billion at end-2000, or only 8 percent Initiative is associated with efforts to avoid incur- of the agencies' total exposure to developing coun- ring additional debt on nonconcessional terms. The tries. Most poor countries are not able to support HIPC Initiative framework provides that new exter- large inflows of guaranteed finance, which is typi- nal finance for these countries should be predomi- cally provided at nonconcessional terms. Neverthe- nantly in the form of grants or loans on highly con- less, export credit agencies are important for the cessional terms. The injunction on nonconcessional poor countries: the agencies account for some 16 borrowing was reinforced by the communiqu6 of percent of the poor countries' long-term debt.36 the Development Committee in April 1999 and New commitments to the poor countries from ex- more recently by U.S. legislation that governs U.S. port credit agencies were $2.4 billion in 2000, or contributions to the HIPC Trust Fund." The IMF 80 percent of gross capital market financing from also agrees with HIPC governments regarding limits private sources. Officially supported export credits on nonconcessional borrowing within the context can provide financing that would not otherwise be of the Fund's concessional facility. These limits are available from private sources, or that would be established on a case-by-case basis, after an evalua- available only at prohibitive terms. In poor coun- ton of the impact of new borrowing on the sustain- tries, official guarantees are nearly always required ability of the debt burden. to access external finance for large projects; every major bank commitment over $20 million over the Some HIs are reducing their reliance on past five years has had some official guarantee. Of- guaranteed loans ficial investment insurance also has helped facili- HIPCs that have reached a decision point, and tate investment flows to more than one-third of the hence have a policy framework in place that is poor countries, and it provided for about 30 per- agreed-on with the international community, have cent of all FDI in poor countries. seen a reduction in export credit commitments Guarantee arrangements have played a particu- from $0.9 billion per year in 1990-96 to $0.5 bil- larly important role in facilitating greater private lion from 1997-2000. Moreover, in these coun- sector participation in infrastructure and in mining tries very little by way of new export credits are projects that require large investors (see box 4.4 on going to public sector borrowers, with the bulk of the Mozambique Mozal project). Access to offi- the finance absorbed by the private sector. Coun- cially supported export credits also may help build tries within the HIPC group that have continued a reputation that facilitates access to nonguaranteed to attract significant export credit financing in- finance in the future. For example, in China two- dude those with sustainable debt burdens and im- thirds of all private source finance was guaranteed portant oil producers (for example Angola) or off- 108 STRENGTHENING OFFICIAL FINANCIAL SUPPORT FOR DEVELOPING COUNTRIES Box 4.4 Official guarantees and the Mozal project O fticial guaraintec h.ae helped attract external tinance proftts, and the income ot.torvigii %orkers. C o' Otir the Mozal aluminLm smelter, the single largest repitriation it a11 di idencs; and a first calloh e3rngs pri e seccor in%estment c .cr undertaken in Nl1'zambi4ue tor debt service pavilenr- Such inltites are availale to and oic of the largest projects It, be developed in i lini all exporting indusrrif, in Nlozambiqie. Ihe most -f ited recourse basis in Sub-Saharan Afri,a The tir,r phase erg' was an importnt factor, and favorable ra', r of the project ($2.3 billion for the aluminum smeltcrj i a1 negiitired with dIe Soth African power urlit. The g read% completed. and the 'econd phase, which %%ill double eminent %%ill re,el. I ptr,Lnr orlth gross inc frin capacirN. is under construcnon Parialk as a result .f - saes. Mozal'; success, pri ate sector pro1ects worth another The Ntozal plant. which is a iread; in prodtition %-ill $6.5 billion are in the pipeline.)' Fort% percent ct the ir- dlutible the country', total export- and add an e-iniared nancing requirements w%ere ie 1% equiie pro.- ided by the pt rcent to DM although the contrijeL6on to eplo mit spon%irs, the Billit-in Group,' Mitsubishi Corporation of is iinited the project added 5,mi terporar) workers dur Japan, the Industrial De%elopment Corporation ot Suth ig the tonsrruction phaS but only 800 full-Em'pern Afr:ca. and b- the go%ernnient (A Mozimbique. Loan fi nent obsi As other planned projects delop expott nancinM w%a met b% officialli supportcd export credits, should ri,e. bv nearl i., perceni of GDP in 2010. and loan, and guaraires romi the European Intesiment though tii, will be partill ottsct by high1cr import t Bank and the International Finance Corporation (11C) rav. materials, de bf ;ers ice on loans, and remittance 0t and sc,ral de.vlopnient tinmnce agencies, including ones profits and wage, of foreign v orkers. The net impact (1 from Germans, South Africi, and France. The perces%cd th balante rit paviis in 201) is eqimaned it less i- political and comninercial risks invoked in the project were 3 ptrcent of GDI. Ocher benefi,hide infra,tructure high. and the participation ot IFC and official guarantors developmentr indostrilition. and the ptomnioh ot were an v,seitial iatalIs to dra., in funding from private regional inegration creditors. These benefits Must hc balanced against thicrikfrom The success of securing financing l i argely due no the proict's contributtion to higher private sctor deb. a well-structured project w% ith leaing international spon- Borrovin I% the private sector has already risen from *n sors. -upported bs Mozambiques nipressi%e reform pro- average of $36 million berween I990-9$ to $340 million gram and rapid rec:s er5 front the %%ar. The country's In I 9QY-2UL1i), and it is expe,tcd to a%erage well over proxinui to ith Atri-i and the retuirii to operation o1 S41ip million mer the net fur to ive o rears. Pris ate sector the (1_ahora Bass;a hl, droelectric pow%er dlamn ha'%L also en- debt service is projectred to rise ito 20 percent ot ex%ports abled Niozarbique to becomet one o.f the t%Hles "C coun- o%cr the next five sears, assuming all the proposed proects tries to hai%te attracd suhst.intial private sector inmest. are realized. While the projects promise o generate sufti- ment Iroin c x[crnal 'ources. In addit,iin. the Prcoiect has cient, returns to- coiver debt eCiit pa% rnent-s. thle expected been, upported b. a package ot in,cntivc s. inClUding ex- u_iA pii:he_privae secitor's debt and veb erx ice p&it eniptions; from taxes; (oil inip..rted materials, corporate tt?foiI'igilanr mniitoaring b\ the authorine,. shore marinn financing centers (Liberia) that can for example the Export Credit Guarantee Depart- pledge assets as collateral, mentr in collaboration with the Department for International Development, has instituted a produc- Rethinking the costs and benefits tive expenditure screening process that applies to all of guarantees- IDA-only countries. Public sector projects in poor Export credit agencies are also taking steps to en- countries are reviewed to ensure that the project sure that the activities they support (including guar- supports the borrowing country's public expendi- antees and insurance) produce real economic and ture priorities. For private sector projects the em- social benefits that are worth the buildup of debt. phasis is on meeting environmental and social stan- Several export credit agencies employ processes that dards and examining the risks of the debt being screen projects for their effectiveness and are look- assumed by the public sector or compromising the ing beyond standard issues such as environment borrowing country's overall debt management stra- and gender screening to include debt sustainability tegy. Export credit agencies are also taking steps to and development impact. In the United Kingdom, implement common anticorruption measures, to re- 109 GLOBAL DEVELOPMENT FINANCE Table 4.5 Export credit commitments to HIPCs, and territories by donor governments and their 1990-2000 agencies that are developmental in intent and de- (annual averages in brilions of dollars) (annal aeraes i bilion ofdollrs)signed to promote economic welfare. ODA and 990- 1. 7 000 OA loans are provided on concessional financial HIPCs at decision point terms, with at least a 25 percent grant element (cal- HIPCs with sustainable debt culated as the NPV of the future payment stream Others discounted at 10 percent). Total Source: OECD. tries can be measured in two ways: when aid per- formance by DAC donors is measured, ODA in- cludes bilateral disbursements of concessional financing to developing countries plus the pro- voke insurance cover if corrupt practices are identi- vision by bilateral donors of concessional financ- fied, and to blacklist corrupt companies. ing to multilateral institutions (for example, IDA). When resource receipts by developing countries -and limiting tied aid are measured, ODA (and, where relevant, OA) in- Export credit agencies also are making progress to dude disbursements of concessional financing reduce the practice of attaching tied aid to export from bilateral agencies and multilateral sources. credit programs. In the past, export credit agencies The two measures will not be the same because the have combined their own financing with official concessional funding received from donor sources aid to create financing packages referred to as by multilateral institutions does not match those "mixed credits" or "parallel financing," where at institutions' disbursements to developing countries least some part of the package is tied to the pro- in any given year. curement of goods and services from specific coun- Aid and debt forgiveness. The directives for tries. The practice of tied aid can impair the effec- reporting aid statistics are agreed-on within the tiveness of donor support for developing countries OECD DAC, and these include specific guidelines by increasing project costs, making procurement on the measurement of debt forgiveness. The im- procedures more complex, and skewing decisions pact on aid volumes varies depending on whether on technology and capital intensity. Under the the claim being forgiven is an official development terms of the OECD Arrangement on Guidelines for loan that was originally disbursed from the aid Officially Supported Export Credits, projects that budget or a commercial loan extended or guaran- are deemed to be financially viable with commer- teed by an official export credit agency. The for- cial loans will not receive any tied aid. giveness of an ODA loan does not give rise to any new net disbursement of aid. Statistically the bene- fit is reflected in the fact that because the cancelled or "forgiven" repayments will not take place, net Annex 4.1 ODA disbursements will not be reduced. The for- giveness of a non-ODA claim has an impact on net Aid definition and measurement ODA. Such forgiveness can be counted by donors Defining aid. The international forum for defining as part of their overall aid effort at the time the aid is the OECD DAC.41 There are two categories claim is forgiven. Statistically forgiveness of a non- of aid provided by DAC donors: ODA and official ODA claim does give rise to a new disbursement aid (OA). The DAC list of aid recipients is divided of aid and net ODA disbursements will increase. into Part I and Part II recipients. Only countries on Official development finance. The concept of Part I receive ODA; those on Part II (which includes official development finance is broader than that of several countries in Eastern and Central Europe, aid. It measures all receipts from official creditors. and Israel) receive OA. Only ODA may be counted It includes (a) ODA and OA from bilateral sources, by DAC countries as part of their "aid effort." (b) grants and concessional and nonconcessional ODA and OA are defined the same way: both development lending by multilateral agencies, and consist of loans or grants to developing countries (c) other official bilateral flows that are considered 110 STRENGTHENING OFFICIAL FINANCIAL SUPPORT FOR DEVELOPING COUNTRIES to be developmental in intent but for which the credits (representing a risk of futute claims) and ar- grant element is too low to qualify as ODA or OA. rears and unrecovered claims (resulting from non- payment and claims payments by agencies). Export credits: data sources and coverage A limitation of the Berne Union data is that Data on export credits need to be interpreted with they are not readily comparable with other types of care. Export credit agencies typically provide insur- debt statistics, and they do not accurately reflect ance cover for repayment of both principal and in- trends in new disbursements. Some agencies do not terest; data provided to the Berne Union and to the report export credit activity by the government OECD are based on agencies' exposure, including (which may undertake export credit finance sepa- future interest payments. Also, agencies typically rately from the export credit agency). Most agen- report the full value of contracts, including undis- cies include the insurance of certain transactions bursed amounts. It is therefore difficult to relate that are not exports; for example, insurance commitment data to actual disbursements. Specific against exchange rate movements or insurance of complications arise when nonpayment by the preshipment risks, which do not involve export debtor gives rise to arrears and rescheduling. Most credits. Data presented in the annual reports of agencies include arrears and rescheduled claims, in- some export credit agencies refer to the full value cluding capitalized interest, in their reports to the of the exports supported, a measure that includes Berne Union and the OECD, but interest accrued down payments by the buyer as well as self partici- on arrears is not recorded as an increase in claims pation by the exporter in the credit. by the export credit agency. Similarly when unre- The OECD compiles two types of data on ex- covered claims are regularized through a Paris port credits. The Statistics on External Indebted- Club rescheduling agreement, agencies do not re- ness reports the stock of export credits on a basis cord an increase in exposure in their reports to the broadly consistent with other external debt data: Berne Union or the OECD despite the fact that the this is covering outstanding disbursed principal longer repayment periods on rescheduled claims in- only. However, since this does not reflect the way creases the future interest at risk. The recording of most export credit agencies keep their accounts, rescheduling arrangements on concessional terms estimation by either the reporting country or the (that is with an element of debt reduction) also staff of the OECD is required. The second set of varies across agencies making the data for debtor data from the OECD is compiled by the Secre- countries experiencing debt servicing problems tariat of the Export Credit Group, which records particularly hard to interpret. the flow of new commitments of export credits The data provided by the export credit agen- with initial maturities of over one year, and initial cies are collected by both the Berne Union and the maturities of over five years, as well as the stock of OECD. The Berne Union quarterly survey of mem- officially supported short-term credits. ber agencies includes data for about 60 developing countries and economies in transition on outstand- ing commitments, unrecovered claims, outstanding Notes offers, and new commitments. The most attractive 1. Aid is defined as grants plus concessional loans. element of the Berne Union survey is that data are 2. Of course, aid devoted to reducing poverty will also collected in the way most agencies actually keep have an impact on education, health, and the environment. their books; the concept commitment encompasses Thus these calculations are not entirely additional to the fore- insured principal and, in most cases, interest on cast of aid requited to halve poverty. See World Bank 2001i. undibured s wll a dibured redis. hisfacll- 3. These include Bolivia, Burkina Faso, Honduras, Mau- undisbursed as well as disbursed credits. This facili-Mozambique, Nicaragua, Tanzania, and Uganda. tates consistency in reporting and avoids errors that 4. See http://www.worldbank.org/poverty/strategies/ can arise when agencies are asked to make esti- review/index.htm mates of statistical concepts for which they have no 5. See Interim Guidelines for PRSCs, available at hard numbers. The Berne Union data are available www.worldbank.org. 6. In some cases, progress is not fast enough, while in with a substantially shorter time lag than data from others there has even been a deterioration (for instance, 14 other sources. The data also provide a breakdown countries saw increases in child mortality hetween 1990 and of total exposure into commitments on outstanding 1999). rear an unrcovred lais (rsuling romnon GLOBAL DEVELOPMENT FINANCE 7. See World Bank 2001a. 21. The Collier and Dollar results are based on GDP 8. Ten countries or territories were removed from the valued at purchasing power parity, which provide a standard list of ODA recipients on January 1, 2000: Aruba, French measure allowing comparison of real price levels between Polynesia, Gibraltar, the Republic of Korea, Libya, Macao, countries (see World Bank 20011), while this calculation uses the Netherlands Antilles, New Caledonia, Northern Mari- GDP valued at dollars. Since the GDP of a developing coun- anas, and the Virgin Islands. try valued at purchasing power parity is typically larger than 9. In a few high-inflation countries where the domestic GDP valued in dollars, this calculation understates the num- retail market is substantially dollarized, foreign exchange ber of poor countries where increased levels of aid will con- sold by the central bank could be used to buy nontradable tinue to have a positive impact on growth. goods, and thus contribute to inflationary pressures. 22. The increase in aid to India, which has about one- 10. While aid should be treated as revenue in the fiscal third of the world's extreme poor but only gets about 5 per- accounts, it is not equivalent to revenue generated by taxes: cent of total aid, is constrained to $11 billion. Absent this (a) aid augments the resources available to the economy adjustment, the framework would imply massive and unre- whereas taxation merely transfers them from the private alistic increases in aid to India. sector to the government; (b) unlike taxation, aid does not 23. Based on case studies of African countries. See distort relative prices; and (c) aid has radically lower costs also Mercer-Blackman and Unigovskaya 2000, and Jayara- of administration than taxes. jah and Branson 1995. In some cases, the complexity of 11. For example, for the International Development conditions contributed to compliance failure. Association (IDA), the grant element is roughly 70 percent. 24. See World Bank 1998 and 2001b. Thus 70 percent of an IDA loan should be viewed as rev- 25. The relationship between donors and recipients enue, and 30 percent as a commercial loan. This approach has been modeled both as the outcome of a bargaining game does face some practical difficulties, in part because the ex (Mosley, Harrigan, and Toye 1991) and in a framework ante calculation of the grant element depends on expecta- where recipients are viewed as agents, implementing condi- tions regarding future exchange rates and interest rates, and tuns desired by donors (Killick 1997; White and Morrissey in part because it could introduce inconsistencies between 1997; Svensson 2000). fiscal and external accounts. 26. See Mosley, Harrigan, and Toye 1991; and Adam 12. Appropriately, both studies measured the volatility 1995. of aid in constant dollars, which provides an indication of 27. However, measuring the extent of implementation the real value of aid resources available to the economy. of structural adjustment programs is problematic, because Bulir and Hamann (2001) find that aid is significantly more volatile than revenues if both variables are expressed as a pror re iteded t e leibe an aremouti od- share of GDP, or if only the relatively aid-dependent coun- 28. Weighted by disbursements, the scores for out- tries are considered. 13. See World Bank 2000c, chapter comes increased from 73 percent satisfactory in fiscal 14. One major aid program, ID, has explicit alloca- 1990-94 to 97 percent in fiscal 1999-2000. 14. ne ajoraidprogam,IDA hasexpicitallca- 29. The NPV refers to the discounted value of future tion criteria, and the bilateral donors also follow criteria debt service payments, where the discount rate is some mar- that are well understood (Alesina and Dollar 2000), so it is possible to define the conditions under which aid may fall. ket rate. This concept was introduced to measure the impact 15. A few of the earlier studies did find a positive im- on the debt burden of different terms on rescheduling. It pact of aid on growth (Dowling and Hiemenz 1983; Levy also provides a comparable measure of the debt burden 1988; and Hadjimichael and others 1995). among countries when a substantial share of outstanding 16. Hansen and Tarp (2001) criticize the Burnside and claims is at concessional rates. Dollar result that policies enhance aid effectiveness as non- 30. The 24 countries that have reached a decision robust to choice of sample. However, Collier and Dehn point are Benin, Bolivia, Burkina Faso, Cameroon, Chad, (2001) show that even on the Hansen-Tarp sample the Ethiopia, Gambia, Guinea, Guinea-Bissau, Guyana, Hon- Burnside-Dollar result holds up, once terms-of-trade shocks duras, Madagascar, Malawi, Mali, Mauritania, Mozam- are included in the specification. bique, Nicaragua, Niger, Rwanda, Senegal, Sdo Tom6 and 17. A one-off expenditure of $1 billion would result in Principe, Tanzania, Uganda, and Zambia. a temporary phase of higher growth, but this temporary 31. Compared to actual debt service paid prior to growth would take the economy to a permanently higher HIPC assistance in 1998-99. level of income. Thus, the poverty reduction produced even 32. See Claessens and others (1996) on the importance by a one-off injection of IDA funds is permanent. of removing the debt overhang facing the HIPCs, and World 18. Africa is probably the only region in which the Bank (2001g) for key aspects of maintaining external debt Dutch disease effects of aid need to be considered, since aid sustainability. as a share of both GDP and exports is much higher than in 33. See World Bank (2001h) for a more detailed dis- any other region. cussion of the status of creditor participation and for exam- 19. With growth as the dependent variable, Collier and pIes of litigation by commercial creditors against HIPCs. Dollar (2001a) find that the coefficient on the square of aid is 34. New commitments include the value of new busi- significant and negative, indicating diminishing returns to aid. ness insured, new lending facilities, and guarantees for new 20. However, Lensink and White (1999) found that aid FDI (but excluding trade finance with maturities of less than in excess of 40 percent of GDP lowers the growth rate. one year). 112 STRENGTHENING OFFICIAL FINANCIAL SUPPORT FOR DEVELOPING COUNTRIES 35. Investment insurance by export credit agencies ex- proaches." Working Paper 142. Overseas Develop- cludes commercial risks: it is normally limited to coverage ment Institute, London. of nationalization or expropriation without compensation, Bolr, A., and J. Hamann. 2001. "How Volatile and Pre- losses on investment due to war or civil unrest, and inability dictable are Aid Flows: and What Are the Policy Inipi- to convert and transfer or remit profits and dividends. canons." International Monetary Fund, Washington, 36. Differences in the definitions used in data from the D.C. Processed. export credit agencies and the private markets may distort Burnside, C., and D. Dollar. 2000. "Aid, Policies and this comparison. Growth." American Economic Review 90(4): 846-68. 37. Berne Union statistics. Checki, Terence J., and Ernest Stern. 2000. "Financial Crises 38. This legislation stipulates that a HIPC country in the Emerging Markets: The Roles of the Public and must commit to not borrow on nonconcessional terms for at Private Sectors." Economics and Finance 6 (13). least two years from any multilateral development bank Claessens, Stijn, Enrica Detragiache, Ravi Kanbur, and Peter benefiting from the U.S. contributions. Wickham. 1996. "Analytical Aspects of the Debt Prob- 39. These include a factory to produce iron slabs, a gas lems of Heavily Indebted Poor Countries." World pipeline, mining and processing of titaniferous mineral Bank Policy Research Working Paper 1618. World sands, and the expansion of the Mozal smelter. Bank, Washington, D.C. 40. Billiton, formerly a South African company but Collier, P. 1999. "Aid 'Dependency': A Critique." journal of now listed on the London Stock Exchange, is the major African Economies 9 (4): 528-45. shareholder in Mozal with a 47 percent stake. Billiton has Colliei, P, and J. Dehn. 2001. "Aid, Shocks and Growth." substantial positions in the markets for aluminum, coal, Policy Research Working Paper 2688, World Bank, nickel, ferroalloys, and industrial minerals. Washington, D.C. 41. Aid is also provided by a few countries that are not Collier, P., and D. Dollar. 2001a. "Aid Allocation and members of the OECD DAC, including the Republic of Poverty Reduction." Policy Research Working Paper Korea, Turkey, and several oil-exporting countries in the 2041. The World Bank, Washington, D.C. Middle East. _ 2001b. "Development Effectiveness: What Have We Learnt?" Development Research Group, World Bank, Washington, D.C. Processed. ____ 2001Ic. "Aid, Risk, and the Special Concerns of References Small States." In Siall States in the Global Econ- The word processed describes informally reproduced works omy, D. Peretz. Economic Paper 44, Commonwealth that may not be commonly available through libraries. Secretariat. Collier, P., and A. Hoeffler. 2000. "Aid, Policy and Peace: Adam, Christopher. 1995. "Adjustment in Africa: Reforms, Reducing the Risks of Civil Conflict." World Bank, Results, and the Road Ahead." World Economy (U.K.) Washington, D.C. Processed. 18 (September): 729-35. Dasgupta, Dipak, and Dilip Ratha. 2000. "What Factors Alesina, Alberto, and David Dollar. 2000. "Who Gives For- Appear to Drive Private Capital Flows to Developing eign Aid to Whom and Why?" journal of Economic Countries? And How Does Official Lending Re- Growth 51(1): 33-63. spond?" Policy Research Working Paper 2392. World Berg, Elliot. 1991. "African Adjustment Programs: False At- Bank, Washington, D.C. July. tacks and True Dilemmas." Paper presented at the Disch, Arne. 2000. "Aid Coordination and Aid Effective- Conference on Structural Adjustment: Retrospect and ness." Report to the Norwegian Ministry of Foreign Prospect. The American University, Department of Affairs. Processed. http://odin.dep.no/archive/ud- Economics, Washington, D.C. bilder/Ol/Ol/Rappo0l5.pdf. Bigsten, A., P. Collier, S. Dercon, M. Fafchamps, B. Gau- Dollar, David, and William Easterly. 1998. "The Search for thier, J.W. Gunning, J. Habarurema, A. Oduro, R. the Key: Aid, Investment, and Policies in Africa." Oostendorp, C. Pattillo, M. Soderdom, F. Teal, and World Bank, Development Research Group, Washing- A. Zeufack. 1999. "Exports and Firm-level Efficiency ton D.C. in African Manufacturing." Development Research Dollar, David, and J. Svensson. 2000. "What Explains the Group, World Bank, Washington, D.C. Processed. Success or Failure of Structural Adjustment Pro- Boone, P. 1994. "The Impact of Foreign Aid on Savings and grams." Economicjournal 110(466): 894-917. Growth." London School of Economics. Processed. Dowling, J. M., and U. Hiemenz. 1983. "Aid, Savings and Boschini, Anne, and Anders Olofsgard. 2001. "Foreign Aid: Growth in the Asian Region." The Developing an Instrument for Fighting Poverty or Communism?" Economies 21 (1): 1-13. Processed. Durbarry, Ramesh, Normal Gemmell, and David Greenaway. Brautigam, Deborah, and Kwesi Botchwey. 1995. "The In- 1998. New Evidence on the Ipact of Foreign Aid on stitutional Impact of Aid Dependence on Recipients Economic Growth. Centre for Research in Economic in Africa." Chr. Michelsen Institute Working Paper, Development and International Trade Research Paper Norway. 1998/8. University of Nottingham, United Kingdom. Brown, Adrienne, Mick Foster, Andy Norton, and Felix Easterly, William. 1999. "How Did Highly Indebted Poor Naschold. 2001. "The Status of Sectorwide Ap- Countries Become Highly Indebted? Reviewing Two 113 GLOBAL DEVELOPMENT FINANCE Decades of Debt Relief." Working Paper No. 2225, Laplagne, Patrick, Malcolm Treadgold, and Jonathan Baldry. World Bank, Washington, D.C. 2001. "A Model of Aid Impact in some South Pacific Eisenblatter, Bernd. 1999. "Globalization and aid policies. Microstates." World Development 29 (2): 365-383. New Forms of Donor Cooperation and Coordina- Lensink, Robert, and Oliver Morrissey. 2000. "Aid instabil- tion-Effects and Experiences." Quarterly Journal of ity as a measure of uncertainty and the positive impact International Agriculture 38 (4): 375-394. of aid on growth." journal of Development Studies Eriksson, John. 2001. "The Drive to Partnership: Aid Coor- February. London. dination and the World Bank." World Bank, Washing- Lensink, Robert, and Howard White. 1999. "is There an ton, D.C. Aid Laffer Curve?" University of Nottingham (U.K.), Griffin, K. 1971. 'Foreign Capital: A Reply to Some Crit- Centre for Research in Economic Development and In- ics." Bulletin of the Oxford University Institute of ternational Trade. Credit Research Paper 9916. Economics and Statistics 32 (2): 156-61. Lerrick, Adam. 1999. "Has the World Bank Lost Its Way?" Guillaumont, P., and L. Chauvet. 2001. "Aid and Perfor- Euromoney December. mance: A Reassessment." Journal of Development Levy, V. 1988. "Aid and Growth in Sub-Saharan Africa: The Studies 37 (6). Recent Experience." European Economic Review 32: Hadjimichael, M. T., D. Ghura, M. Muhleisen, R. Nord, and 1777-1795. E. M. Ucer. 1995. "Sub-Saharan Africa Growth, Say- Lister, Stephen, and Mike Stevens. 1992. Aid Coordination ings, and Investment, 1986-93." International Mone- and Management. Washington, D.C.: World Bank. tary Fund Occasional Paper no. 118. Washington, D.C. Processed. Hansen, H., and F. Tarp. 2001. "Aid and Growth Regres- Maizels, A., and M. K. Nissanke. 1984. "Motivations for sions." Journal of Developing Economies 64 (2): Aid to Developing Countries." World Development 12 547-70. (9). Hoffman, P. T., and K. Norberg, eds. 1994. Fiscal Crises, McCleary, William. 1991. "The Design and Implementation Liberty, and Representative Government, 1450-1789. of Conditionality." In Thomas, Vinod, Ajay Chhibber, Stanford, Calif.: Stanford University Press. Mansoor Dailam, and Jaime de Melo, eds. Restruc- IMF (International Monetary Fund). 1998. "External Eval- turing Economies in Distress: Policy Reform and uation of the ESAF: Report by a Group of Independent the World Bank. New York: Oxford University Press, Experts." Washington, D.C. pp.197-215. Jayarajah, C., and W. Branson. 1995. "Structural and Sec- Mercer-Blackman, Valerie, and Anna Unigovskaya. 2000. toral Adjustment: World Bank Experience, 1980-92." "Compliance with IMF program indicators and Operations Evaluation Department, World Bank, growth in transition economies." IMF Working Paper Washington, D.C. WP/00/47, Washington, D.C. Jepma, Catrinus. 1991. The Tying of Aid. Paris: Develop- Morrissey, Oliver, and Howard White. 1994. "Evaluating ment Center of the OECD. the Concessionality of Tied Aid." Institute of Social Johnson, J., and S. Wasty. 1993. "Borrower Ownership of Studies Working Paper: Subseries on Money, Finance, Adjustment Programs and the Political Economy of and Development (Netherlands) No. 53/January: 1-18. Reform." Discussion Paper 199. World Bank, Wash- Mosley, Paul. 1986. "Politics of Economic Liberalization: ington, D.C. USAID and the World Bank in Kenya 1980-84." Kharas, Homi, and Hisanobu Shishido. 1991. "The Transi- African Affairs (U.K.) 85 (January): 107-19. tion from Aid to Private Capital Flows." In Uma Lele - 1999. "Recent changes in aid technology: is the and Ijaz Nabi, eds., Transitions in Development: The White Paper an adequate response?" Public Adminis- Role of Aid and Commercial Flows. San Francisco: tration and Development 19 (29). ICS Press. Mosley, Paul, J. Harrigan, and J. Toye. 1991. Aid and Killick, Tony. 1997, "Principals, Agents, and the Failings of Power: The World Bank and Policy Lending, Vol. 1. Conditionality." Journal of International Development London: Routledge. 9: 483-95. OECD (Organisation for Economic Co-operation and De- Killick, Tony, with Ramani Gunatilaka, and Ana Marr. velopment). 2001. "Untying Aid to the Least Devel- 1998. Aid and the Political Economy of Policy oped Countries." OECD Observer Policy Brief, Paris. Change. London: Routledge. Pallage, Stephane, and Michel Robe. 1998. Foreign Aid and Knack, Stephen. 2000. "Aid Dependence and the Quality of the Business Cycle. Center for Research on Economic Governance: Cross-Country Empirical Tests." Policy Fluctuations and Employment Working Paper no. 63. Research Working Paper 2396. World Bank, Washing- University of Quebec at Montreal. ton, D.C. Pearson, Lester B. 1969. "Partners in Development: Report Kraay, Art. 1999. "Exportations et performances economi- of the Commission on International Development." ques : etude d'un panel de enterprises chinoises." Revue New York: Praeger. d'Economie du Developpement (France) 7 (1-2): Reichel, Richard. 1995. "Development Aid, Savings and 183-207. Growth in the 1980s-A Cross-Section Analysis." Say- Krueger, Anne 0. 1998. "Whither the World Bank and the ings and Development 19 (3): 279-96. IMF?" Journal of Economic Literature 36: 1983-2020. Rodrik, Dani. 1996. "Understanding economic policy re- Lancaster, Carol. 1999. Aid to Africa. University of Chicago form." Journal of Economic Literature 34 (March): Press. 9-41. 114 STRENGTHENING OFFICIAL FINANCIAL SUPPORT FOR DEVELOPING COUNTRIES Sachs, J., and M. Warner. 2000. "Natural Resource Abun- Rate Misalignment." Journal of International Eco- dance and Economic Growth'. In Leading Issues in nomics 21: 82-116. Economic Development, 7th ed. G.M. Meier and J.E. White, Howard, and Oliver Morrissey. 1997 "Conditional- Rauch, Oxford University Press. ity When Donor and Recipient Preferences Vary." Sekkat, Khalid, and Aristomene Varoudakis. 2000. "Ex- Journal of International Development 9: 497-505. change Rate Management and Manufactured Exports World Bank. 1997. "Adjustment lending in SLb-Saharan in Sub-Saharan Africa." Journal of Development Eco- Africa: an update." Operations Evaluation Division, nomics 61: 237-253. Washington, D.C. SIDA (Swedish International Development Agency). 1999. - 1998. "Assessing Aid: What Works, What Doesn't, "Dollars, Dialogue and Development" An Evaluation and Why?" World Bank Policy Research Report, of Swedish Programme Aid http://www.sida.se/Sida/ Washington, D.C. www.worldbank.org/research/aid jsp/Crosslink.jsp?d=41 0 . 1999. "The Drive to Partnership: Aid Coordination Snyder, Donald. 1990. "Foreign Aid and Domestic Savings: and the World Bank." Operations Evaluation Depart- A Spurious Correlation?" Economic Development and ment, Washington, D.C. Cultural Change 39 (October): 175-81. . 2000a. Global Development Finance. Washington, Soderling, Ludvig. 2000. "Dynamics of Export Perfor- D.C. mance, Productivity and Real Effective Exchange Rate - 2000b. "Quality at Entry in CY99: A QAG assess- in Manufacturing: the Case of Cameroon." journal of ment." Quality Assurance Group, Washington, D.C. African Economies 9 (4): 411-29. . 2000c. Global Economic Prospects. Washington, Summers, Lawrence. 1999. "A New Framework for Multi- D.C. lateral Development Policy." Remarks to the Council - 2001a. "Report of the Task Force on the World on Foreign Relations, New York, March 20, www. Bank Group and the Middle-Income Countries." treas.gov/press/releases/ps477.htm. See also his testi- (SecM2001-0204). March 27. mony before the House Banking Committee on March - 2001b. "Financing for Development." (SecM2001). 20, 1999, www.treas.gov/press/releases/ps480.htm. August 28. Svensson, Jakob. 2000. "When Is Foreign Aid Policy Credi- - 2001c. Aid and Reform in Africa. Washington, ble? Aid Dependence and Conditionality." Journal of D.C. Development Economics 61: 61-84. . 2001d. Adjustment Lending Retrospective. Opera- Taslim, M. A., and A. Weliwita. 2000. "The Inverse Relation- tions Policy and Country Services Department. Wash- ship between Saving and Aid: An Alternative Explana- ington, D.C. tion. Journal of Economic Development 25 (1/June). - 2001e. Global Development Finance. Washington, Trumbull, William, and Howard Wall. 1994. "Estimating D.C. Aid-Allocation Criteria with Panel Data." The Eco- - 2001f. "Review of Aid Coordination in an Era of nomicjournal 104 (July). Poverty Reduction Strategies. IDA 10-12." Operations USAID (United States Agency for International Develop- Evaluation Department, Washington, D.C. ment). 2001. "FY 2000 performance overview." Wash- - 2001g. The Challenge of Maintaining Long-Term ington, D.C. http://www.dec.org/partners/2000_po/ External Debt Sustainability. (http://www.world- pdf_docs/fy2000po.pdf bank.org/hipc/hipc-review/hipc-review.html) van Arkadie, Brian. 1986. "Some Realities of Adjustment: - 2001h. Heavily Indebted Poor Countries (HIPC) An Introduction," Development and Change 0012- Initiative: Status of Implementation. (IDA/SecM2001- 155X 17(3): 371-86. London. 0543/1) van de Walle, Nicolas, and Timothy Johnston. 1996. "Im- - 20011. "Goals for Development: History, Prospects, proving aid to Africa." Policy Essay No. 21. Overseas and Costs." A background paper for the September Development Council, Washington, D.C. 2001 Development Committee Meeting. Washington, van Wijnbergen, Sweder. 1986. "Macroeconomic Aspects of D.C. Effectiveness of Foreign Aid: On the Two-Gap Model, - 2001j. World Development Indicators. Washing- Home-Goods, Disequilibrium, and Real Exchange ton, D.C. 115  Appendixes Appendix 1 Debt Burden Indicators and Country Classifications Appendix 2 Commercial Debt Restructuring Appendix 3 Official Debt Restructuring Appendix 4 Regional Economic Developments and Prospects Appendix 5 Global Commodity Price Prospects  Appendix 1 Debt Burden Indicators and Country Classifications Country classifications for 2002 the present value of debt service to GNI and 132 LOBALpercent for the present value of debt service to ex- L OBL DVELOMEN FIANCECLASI- ports), the country is classified as moderately in- fies indebtedness based on two ratios: the debted. If both ratios are less than three-fifths of ratio of the present value of total debt ser- the critical value, the country is classified as less in- vice to gross national income (GNI) and the ratio debted. Countries are further classified as low-in- of the present value of total debt service to exports. come if 2000 GNI per capita was $755 or less and These ratios cast a country's indebtedness in terms as middle-income if 2000 GNI per capita was of two important aspects of its potential capacity $756 or more but less than $9,265. Combining to service the debt: exports (because they provide these criteria leads to the identification of severely foreign exchange to service debt) and GNI (be- indebted low-income countries (SILICs), severely cause it is the broadest measure of income genera- indebted middle-income countries (SIMICs), mod- tion in an economy). For the 136 countries that re- erately indebted low-income countries (MILICs), port to the World Bank's Debtor Reporting System moderately indebted middle-income countries (DRS) the debt data are drawn from this source. (MIMICs), less indebted low-income countries GNI and export data are from World Bank files, as (LILICs), and less indebted middle-income coun- shown in the Country Tables volume of Global tries (LIMICs; table A1.1). Development Finance. Export figures are earnings The use of critical values to define the bound- from goods and services, including worker remit- aries between indebtedness categories implies that tances. Data on official grants are not included, al- though they may be a stable source of foreign ex- preted with caution. If a country has an indicator change in some countries. that is close to the critical value, a small change in The two indebtedness ratios in Global Devel- the indicator may trigger a change in indebtedness opment Finance 2002 are calculated as follows: classification even if economic fundamentals have * The ratio of the present value of total debt changed significantly. " Th raio o th preentvalu oftota det .Moreover, these indicators do not represent an service in 2000 to average GNI in 1998, 1999, exhaustive set of useful indicators of external debt. and 2000. * The ratio of the present value of total debt service in 2000 to average exports (including debt servicing capacity of countries in which gov- workerernment budget constraints are key to debt service workr rmitancs) n 198,199, ad 200. difficulties. Countries (such as the franc zone If either ratio exceeds a critical value-80 percent countries in Africa) that allow the use or free con- for the debt service to GNI ratio and 220 percent version of a foreign currency can face government for the debt service to exports ratio-the country is budget difficulties that are related to servicing ex- classified as severely indebted. If the critical value ternal public debt, but that are not necessarily re- is not exceeded but either ratio is three-fifths or flected in balance of payments data. In other coun- more of the critical value (that is, 48 percent for tries, the servicing of domestic public debt may be 119 GLOBAL DEVELOPMENT FINANCE Table A1.1 Income and indebtedness classification criteria Indebtedness classification PV/XGS less than 220 percent VCs than S ige ta but higher than 132 percent or peren or i hi beePVIGNI less than 80 percent IP :tIss Income classification but higher than 48 percent (li) Low-income: GNI per capita Se ted Moderately indebted ePindebe* less than $755 onm lie low-income countries 1Wi-nlome MUur6 Middle-income: GNI per capita erl 1 Moderately indebted between $756 and $9,265 m u middle-income countries Note: PV/XGS is present value of debt service to exports of goods and services. PV/GNI is present value of debt service to GNL Source: World Bank. a source of fiscal strain that is not reflected in bal- discounted to compute its present value. For vari- ance of payments data. Moreover, rising external able-rare loans, for which the future debt service debt may not necessarily imply payment difficul- payment cannot be precisely determined, debt ser- ties, especially if there is a commensurate increase vice is calculated using the rate at the end of 2000 in the country's debt servicing capacity. Thus, these for the base specified for the loan. indicators should be used in the broader context of a country-specific analysis of debt sustainability. Classification of low-income In the context of the Heavily Indebted Poor countries Countries (HIPC) Initiative, countries are classified Thirty-three countries are classified as SILICs, 16 based on the ratio of the present value of public and publicly guaranteed debt to exports of goods as iLics and 12 a TjiC sta ave The and services, excluding worker remittances. ened and they joined the severely indebted group those countries for which a joint debt sustainabil- ity analysis has been undertaken, the indicators are calculated on this basis and shown in italics in ta- Uganda have improved, and they are now classi- bles A1.4 and A1.5. fied in the moderately indebted group of low-in- The discount rates used to calculate present come countries. Debt indicators for Armenia, value are interest rates charged by the Organisa- Georgia, and Vietnam have also im- tion for Economic Co-operation and Development proved, and they are now classified in the less in- (OECD) countries for officially supported export debted group of low-income countries. credits. They represent, on average, the most fa- vorable terms for fixed-rate nonconcessional debt Classification of middle-income that countries are able to contract in international countries loan markets. The rates are specified for 19 curren- In the middle-income group, 8 countries are classi- cies, including Group-of-Seven currencies-British fied as SIMICs, 27 as MIMICs, and 40 as LIMICs. pounds, Canadian dollars, French francs, German The debt indicators for Ecuador have declined, marks, Italian lire, Japanese yen, and U.S. dollars. and it has moved into the group of severely in- International Bank for Reconstruction and Devel- debted middle-income countries. Also, Croatia has opment currency-pool loans, International Devel- joined the moderately indebted middle-income opment Association credits, and International countries. By contrast, debt indicators for Bolivia, Monetary Fund loans are discounted at the special Bosnia and Herzegovina, and Bulgaria have im- drawing rights lending rate. For debt denominated proved, and they have now joined the moderately in other currencies, discount rates are the average indebted group of middle-income countries. Mo- of interest rates on export credits charged by other rocco's debt indicator has also improved, and it is OECD countries. now classified in the less indebted group of middle- In present value calculations, debt service on income countries. fixed-rate loans is determined and each payment is 120 DEBT BURDEN INDICATORS AND COUNTRY CLASSIFICATIONS Table A1.2 Classification of DRS economies Severely ind Severely indebted M i b Moderately indebted Less ih%fftd Less indebted w-inco middle-income Iwincom m middle-income it'w middle-income Argerrnnr BUr., A r en.i brnBr ru1 fmll,r r bs srdiil-:r, Ecurdibn, j TIrc Bohr'ia' 13 an'g Iade' jVr , 1 Burua Eriuuaddor1 Ganr..:-ni.i.n (is . nnaad(p cM lord.r ire', r Bulgmii" croahR.ubb ku.i . Peru I, tzrch RePLAit m Rep .:.r 5r, ArA' Reu-hi., P, T . v.ne..j. Rtp ofCrma d I;..r,d, D.-..c.n Rep.lc.c f%ihm.m". INr Republ.c Rep eri . I 1 ck hi,.Rep ..I tR u riFederano.:r. Lar 13 jm.-- 3~L iirhua3n ia 1 . 1. -; .. T m r1 Pr jod Xle rd.~N jd Tr [corocol .r .. ..d IIs la, < l c Ar Luo, h*o a ii]-. Irm nrrid3d arid Tohvo crzgatina =.h.c in . H r ilu i.Ftd Rep f 'P4y .l ur. . .u ..r a - u-.r-. -< Notes: Tables AI1.2 and A1. classify all DRS and 12 non-DRS economies. Economies are divided among income groups according to 2000 GNI per capita, calculated using the World Bank Atlas method. Income groups are low-income, $755 or less; lower-middle-income, $756-$2,995; upper-middle-income, $2,996-$9,265; and high-income, $9,266 or more. a. Countries whose indebtedness classification has worsened. i. Countries whose indebtedness classification has improved. c. Countries whose income classification has chaoged. Source: World Bank. 121 GLOBAL DEVELOPMENT FINANCE Table A1.3 Classification of non-DRS economies n k . e & .'ilm!R ur.A '.T' - - --,r,n-d Table A1.4 Major economic indicators, 2000 (millions of dollars) Country 'I.N Albania Algerta - ~ I ei. Angola 4 . 3104, " Argentina , "'j.~r! Armemia44 . Azerbaijan 4 - '! Bangladesh .n Belarus I 'rii4 Belize 'I-I Bemn in- Bhutan -- Bolivia' - - ¯ s Bosnia and Herzcg', .r 2 - 4 Botswana 4 Brazil Bulgaria Ii - ,,r ... 5 i 9~ Burkina Faso i.23 2·w ~ Burundi . -i--~ Cambodia . Cameroon ¯'4 "J2 Cape Verde : 'r i Central African R.ritric - 4 b '2 Chad n -. -, Chile . - Chma 144 s .-2 Colombia -ss i 15 Comoros _r Congo, Dem. Rep ..i I 1,4 ai Congo, Rep. of 4 4, Costa Rica - 4 4,,. Côte dIvoire Il - i 1.3 i,- ii Croatia ¯ir .1j4 4, 1I 2i Czech Republic 21," 21 l -1,4 i. .li 3 Diboun -i2 13 - St S Dominica Dominican RepL .5 4. i I I Ecuador I 2,I I 41 I Ih1 i 2 , ii Egypt, Arab Rep .. i lFr El Salvador 4 q-, 41 -n I , Equatorial Gum: 3.¯ i Eritrea 'fiI i'r Estonia . '. 44 122 DEBT BURDEN INDICATORS AND COUNTRY CLASSIFICATIONS Table A1.4 Major economic indicators, 2000 (continued) (millions of dollars) Country PV W 1iD INT GNI Ethiopia 54 3,264 39 530 6,331 Fiji I 129 ) 7 1,8 1435 Gabon ,$ 3,879 4 216 113 4,240 Gambia, The 8 265 19 7 27 415 Georgia 33 1,271 11, 49 Un@2 3,042 Ghana 5#803,921 4 157 2,5 5,037 Grenada 207 166 12 4 2 377 Guatemala 4 4,326 418 245 4 18,743 Guinea ,2S 2,341 113 57 870 2 931 Guinea-Bissau 7 700 6 3 7 203 Guyana 14 842 11 48 74 660 Haiti 1,16 691 20 M2 4,064 Honduras 40M8 3,110 57 192 991 5,794 Hungary 2)9 27,841 7 6 1,502 3' 44,061 India 00, 70,886 9' fl 3,902 77, 1 470,480 Indonesia 41480 134,996 18 7,476 74,2 142,657 Iran, Islamic Rep. of ,9S 7,421 3 604 0 105,286 Jamaica 4 28 4,341 64 246 6,993 Jordan ,22 7,623 6 301 5 7 8,313 Kazakhstan 66' 6,689 ,8400 448 0,9 17,052 Kenya 4,680 486 131 ,8 10,223 Korea, Rep. of 1'4$151 2 128,445 24,0 5,802 2]1 i3 455,019 Kyrgyz Republic 1,869 1,405 173 76 5 1,223 Lao PDR 249 1,208 2 0 51 1,670 Latvia , 3,332 6' 183 ,57 7,176 Lebanon 10,1 * 10,547 1,21 650 54 17,420 Lesotho 515 6 24 4 1,132 Liberia 032 . 2,053 1 0 - - Lithuania 485 4,772 90 277 5,27 11,120 Macedonia, FYR of 4 5 1,252 1 53 74 3,524 Madagascar 41 2,986 3 38 40 3,804 Malawi 543 .. 1,487 5 20 00 1,660 Malaysia 4 797 42,872 9 2,289 18,2'1 82,392 Maldives 07 145 2 6 46 526 Mali 0 - 1,340 9 28 04 2,260 Mauritania 1,218 00 33 38 909 Mauritius 2,358 55 149 , 4,349 Mexico 157,038 S,2 13,722 1 '28l1 559,765 Moldova 1,138 5 5581 1,357 Mongolia $2 566 11 (25 954 Morocco 11987 9955 32,457 Mozambique 1,162 31 76 3,603 Myanmar @JI(i1 4,341 8 16 84 7,337 Nepal 5) 2 1,558 I01 29 ,529 5,687 Nicaragua 5,545 I@ 109 r 1,34 2,110 Niger .,Do P 1,041 MI 11 3 1,842 Nigeria K641 27,207 i 429 ER OB 36,726 Oman 5,958 385 11 98 19,853 Pakistan - %M008 26,607 985 8(40,W 59,620 Panama 7TM5 7,285 456 9)15'4 9,368 Papua New Guinea &g0 2,195 90 22 3,651 Paraguay I) 2,950 131 3 88 7,554 Peru 28,411 4W 1,681 1 {55 51,925 Philippines 50,766 f 2,724 49,3. 79,317 Poland O1 58,144 I$9M 2,573 4918 156,819 Romania 1(996R 10,006 &W1Y94 588 4 36,381 Russian Federation I601,4'010 148,076 ENM%7" 5,601 S 239,952 Rwanda 91W,5 722 4 1,775 Samoa 129 9 4 236 Sao Tome and Principe F 0 M l 196 2 14 44 Senegal 2,416 8 , 74 "15S 4,286 (Table continues on next page) 123 GLOBAL DEVELOPMENT FINANCE Table A1.4 Major economic indicators, 2000 (continued) (millions of dollars) Country §EDT PV INT GNI Seychelles 154 7 6 579 Sierra Leone 792 43f. 10 616 Slovak Republic QA9,071 5575, 1W0O55718,767 Solomon Islands 5 99 9 3 285 Somalia 2,274 0 0 - South Africa 24,8 23,774 3860 1,251 122,643 Sri Lanka 6,981 7I 218 7,670 16,368 St. Kitts and Nevis 140 116 e- 2 7 - I - 275 St. Lucia 221 e 19 669 St. Vincent and the Grenadines 198 159 RS 7 312 Sudan 15,71 14,824 6 3 9,754 Swaziland 218 10 $ag 1,515 Syrian Arab Republic 2.1 20,842 344- 222 15,965 Tapkistan 940 31 936 Tanzania 4,487 61 , 8,984 Thailand 76,555 14 17 4,850 120,544 Togo 1,018 30 10 1,195 Tonga 35 1 155 Trinidad and Tobago 47 2,553 500 168 6,701 Tunisia 0,610 10,522 1,900 553 ** 18,572 Turkey 114,603 21,136 6,857 201,517 Turkmenistan - 4,227 Uganda 984 1 9 47 6,156 Ukraine 1 16 11,500 3,66 709 30,849 Uruguay 129 8,204 1,313 637 19,383 Uzbekistan 4,340 4,245 899 280 - Vanuatu ( 35 2 1 224 Venezuela, R.B. de 1496 38,744 5,846 2,553 'Q 119,321 Vietnam 178' 11,137 1,303 350 179 31,344 Yemen, Republic of .,616 4,238 2'24 96 0.5,870 7,386 Yugoslavia, Fed. Rep. of 12,007 177 104 8,448 Zambia 6 5,002 186 88 99 2,791 Zimbabwe i 3,599 471 165 7,142 - Not available. Note: For definition of indicators, see Sources and Definitions section. Numbers in italics are from debt sustainability analyses undertaken in the context of the HIPC Initiative. Present value estimates for these countries are for public and publicly guaranteed debt only, and export fig- ures exclude worker remittances. a. Debt numbers do not include foreign currency private sector (nonguaranteed) external debt. Source: World Bank DRS; staff estimates. 124 DE BT B UR D E N I NDICATORS AND COUNT RY CL AS SIF IC AT I O NS Table' A1.5 Key indebtedness ratios, 1998-2000 (percent) Country S PV/XGS ED/NIf PV/GNI TDi/>" S INT/XGS Albania 4 46 2 14 1 Algeria 156 3 54 7 Angola 170 83 270 2 Argentina 425 2 55 30 Armenia 117 31 3 Azerbaijan 7 64 26 21 2 3 Bangladesh 15 120 21 2 Belarus 11 3 3 Belize ]">9 125 7l 69 6 Benin 36 253 9 203 Bhutan 111 36 1 Bolwiaa 3 200 32 11 Bosma and Herzegovina í 145 49 24 9 Botswana 10 7 0 Brazil 346 39 36 9 22 Bulgaria 147 8 80 8 7 Burkina Faso 210 25 5 Burundi ,6 84 985 4 86 33 14 Cambodia 158 66 1 Cameroon 02 257 70 11 Cape Verde 128 47 2 Central African Republic 7356 54 4 Chad 40 222 7 40 0 f 4 Chile J 3 163 51 9 Chm 54 133 China3 Colombia 209 206 39 32 12 Comoros 14 296 74 2 Congo, Dem. Rep. of 848 797 128 2 Congo, Rep. of So 235 7 254 2 Costa Rica 7 57 32 3 Côte d'Ivoire 226 3 117 12 Croatia 1 130 61 4 7 Czech Republic 5 60 40 I 4 Djibouti 71 3 Dominica 53 34 2 Dommican Republie 44 26 2 Ecuador 2 202 82 8 Egypt, Arab Rep. of 115 25 3 El Salvador 77 31 4 Equatorial Guinea 13 43 0 Eritrea 75 25 Estonia 69 7 62 3 Ethiopia 343 51 5 Fiji 12 89 Gabon 142 96 7 Gambia, The 217 62 2 Georgia 124 39 4 Ghana 6157 59 1 6 Grenada 69 59 48 51 Guatemala 102 23 5 Guinea 286 9 69 I 7 Guinea-Bissau 7 1,321 3 321 12 4 Guyana 121 1 120 7 Haiti 132 29 17 8 4 Honduras 126 S5 6 Hungary 9 94 65 62 5 India 105 16 5 Indonesia 212 111 10 Iran, Islamic Rep. of 33 2 Jamaica 99 63 5 Jordan 136 103 95 2 5 (Table continues on next page) 125 G L O B A L D E VEL OP M EN T F INA NC E Table A1.5 Key indebtedness ratios, 1998-2000 (continued) (percent) Country EDT/XGS PV/XGS EDT/GNI PV/GNI TDS/XGS INT/XGS Kazakhstan 80 81 36 36 22 4 Kenya 225 167 59 44 17 5 Korea, Rep. of 74 70 35 33 13 3 Kyrgyz Republic 315 242 138 106 30 13 Lao PDR 504 243 173 83 8 2 Latvia 102 100 51 50 17 5 Lebanon 8 190 60 61 33 12 Lesotho 127 91 61 44 12 4 Liberia Lithuania 98 96 48- 47 18 5 Macedoma, FYR of 92" 78 41 35 10 3 Madagascar 498 333 118 79 10 3 Malawi 536 314 144 84 12 4 Malaysia 42 43 55 57 6 2 Maldives 46 32 45 32 4 1 Mali 462 209 110 50 15 3 Maurintama 508 319 201 126 26 8 Mauritius 89- 89 57 56 21 6 Mexico 92 96 32 33 36 7 Moldova 4 138 86 79 16 7 Mongolia 149 99 91 60 .5 2 Morocco 142 127 53 47 26 7 Mozambique 825 187 129 29 14 4 Myanmar 345 248 97 69 5.- Nepal 204 113 54 30 7 2 Nicaragua 68 640 340 283 35 8 Niger 510 345 77 52 9 3 Nigeria 203 162 105 84 6 2 Oman '73 69 38 36 10 3 Pakistan 298 247 55 45 27 9 Panama 77 79 78 80 10 5 Papua New Guinea 119 100 73 61 14 4 Paraguay 80 77 39 37 9 4 Peru 306 304 54 54 46 17 Philippines 107 109 66 67 14 6 Poland 140 128 41 37 23 5 Romania 9 93 27 27 22 5 Russian Federation 166 154 70- 64 12 5 Rwanda 1,070 628 65 38 30 8 Samoa 177 115 85 55 8 5 São Tomé and Principe 027 1,307 647 417 29 - is Senegal 22,1 151 73 50 14 5 Seychelles 33 28 26 4 1 Sierra Leone 20" 800 181 121 43 11 Slovak Republic -'71 68 48 46 19 4 Solomon Islands 83 53 50 32 5 2 Somalia - - ~ -- South Africa - ,68 65 20 19 11 3 SnLanka 127 98 58 44 10 3 St. Kitts and Nevis 93 77 52 43 13 4 St. Lucia 58 54 38 35 10 5 St. Vincent and the Grenadines 1 91 63,- 52 9 4 Sudan 1400 1,319 171 161 5 0 Swaziland 23 19 -9 15 2 1 Syrian Arab Republic 357 344 141 136 6 3 Tapkistan 169 136 109 88 13 4 Tanzania 544 395 70 51 19 5 Thailand 104 100 69 66 18 6 Togo 281 199 108 77 6 2 Tonga 140 84 37 22 10 2 Trinidad and Tobago 65 67 39 41 13 3 Tumsia 112 111 55 55 20 6 126 DEBT BURDEN INDICATORS AND COUNTRY CLASSIFICATIONS Table A1.5 Key indebtedness ratios, 1998-2000 (continued) (percent) Country % I 1%\G r 0 '1 GNI INT \GV Turkey lI- I Turkmenistan -. Uganda 4 - Ukraine 3 - Uruguay 1h2 II X Uzbekistan I '- Vanuatu I 5 I I Venezuela, R.B. de 134 3- Vietnam Yemen, Republic of 14 . Yugoslavia, Fed. Rep. of - li-4 Zambia j Zimbabwe l I.3 - Not available. Note: For definition of indicators, see Sources and Definitions section. In the estimated ratios, the numerator refers to the 2000 data and the denominator is an average of 1998 to 2000 data. Numbers in italics are from debt sustainability analyses undertaken in the context of the HIPC Initiative. Present value estimates for these countries are for public and publicly guaranteed debt only, and export figures exclude worker remittances. a. Debt numbers do not include foreign currency private sector (nonguaranteed) external debt. Sources: World Bank DRS; staff estimates. 127 GLOBAL DEVELOPMENT FINANCE Table A1.6 Classification of economies by income group and region, January 2002 Sub-Saharan:Africa Asia ,Europe and Central Asia Middle East and North Africa East and Eastern Income Southern East Asia Europe and Rest of Middle North group Subgroup Africa West Afnca and Pacific South Asia Cential Asia Europe East Africa Americas Low- Angola , Benin Cambodia Afghanistan Armenia. Yemen, Haiti income Burundi Burkina Faso Indonesia Bangladesh Azerbaijan Republic of Nicaragua Comoros Cameroon Korea, Dem. Bhutan Georgia Congo, Dem. Central People's India Kyrg'y Rep. of African Republic of Nepal Republic Eritrea , Republic Lao PDR Pakistan Moldova Ethiopia - Chad Mongolia Tajikistan Kenya , Congo, Rep of Myanmar Ukraine" Lesotho Côte d'Ivoire Solomon Uzbekistan Madagascar Gambia, The Islands Malawi Ghana 'Vietnam Mozambique Guimea Rwanda Guinea-Bissau Somalia Liberia Sudan Mal Tanzania Mauritaa tganda Niger Zambia Nigeria Zimbabwe Sao Tomé and Primcipe Senegal Sierra Leone ITogo~ ~> Middle- Lower Namibia Cape Verde China Maldives Albani' Iran, Islamic Algeria Belize income Swaziland Equatorial Fiji Sri Lanka Belarus Rep. of Djibouti Bolivia Gumea Kiribati 'B6snia'ånd Iraq Egypt, Arab Colombia Marshall Herzegovia Jordan Rep. of Cuba Islands Bulgaria Syrian Arab Morocco Dominican Micronesia, Kazakhstan Republic Tunisia Republic Federated Latvi.' West Bank Ecuador States of Lithuania and Gaza El Salvador Papua New Macédoîia Guatemala Guinea FYRobf Guyana Philippines Romaia Honduras Samoa Russia6 Jamaica Thailand Federation Paraguay Tonga Turkmeniitan Peru Vanuatu Yugoslavi St. Vincent Fed.;Re of and the Grenadines Suriname- Upper Botswana Gabon American Croatia Gibralter Bahrain Libya Antigua and Miuritis Samoa €zech Isle of Man Lebanon -Barbuda Mayotte Korea, Republic Turkey Oman Argentina Seychelles Rep. of Estonta - Saudi Arabia Brazil - South Afria Malaysia Hungary - Chile Palau Poland Costa Rica Slo-vak Dominica Repubic Grenada Mexico Panama Puerto Rico St. Kitts and Nevis - St. Lucia Trinidad and Tobago Uruguay Venezuela,' R.B. de Subtotal: 156 25. 23 , 23 8 3 10 6 32 128 DEBT BURDEN INDICATORS AND COUNTRY CLASSIFICATIONS Table A1.6 Classification of economies by income group and region, January 2002 (continued) I Sub-Sa.aa A%-a W hurope NndCenL1aRA%ia , Midlie Ea,i ind N.irjF, Airic, Income E.a%t Asa Reof Middi Norih group Subgroup and Paci, So.th 4sia CerI Aa Europe East AIFICa Amerma. High- OECD Ausiralia Csfiada income Japan Belgium United icrie; New Zealrd Denmar Fin1land FrancT (.~uaiTGermanyd Icserelands rri ~Cam~ I~ II Ma Ireland Italy Uxembourg Net erlins Nora Sweden, SwItzerlan Sinnk prni Non- Brune2 0lona Andorri Itael Kt3lfa Arba OECD French i annel ons ad l s bahram , The Polynesia ilandI Qatar Barbadox; Guaams Coprus Lited Ar)b Bermuda -iHng kar on a Faeroe imi Em.riet C ma n -Chma3, Grienland 1t3rdl. Miacan, Chna' Lechtenosteines Netherlnd , -New Mon3cc, Arale Ciledlon..3 Sir. larinn.:ug. N. Marianc Ilands Islands FLo a i hinar Total: 209 JP . 2-3 35 814 4 1 For operational and analytical purposes, the World Bank's main criterion for classifymng economies is GNI per capita. Every economy is classified as low mncome, mid- dbe income (subdivided into lower middle and upper middle), or high income. Other analytical groups, based on geographic regions and levels of external debt, are also used. Low-income and middle-mncome economies are sometimes referred to as developing economies. The use of the term is convenient; it is not mntended to imply that all economies in the group are experiencmng similar development or that other economies have reached a preferred or final stage of development. Classification by income does not necessarily reflect development status. This table classifies all World Bank member economies, and all other economies with populations of more than 30,000. Economies are divided among mncome groups according to 2000 GNI per capita, calculated using the World Bank Atlas method. The groups are low-income, $755 or less; lower-middle-mncome, $756-$2,995; upper-middle-income, $2,996-$9,265; and high-income, $9,266 or more. a. Former Yugoslav Republic of Macedonia. b. The French overseas departments French Guiana, Guadeloupe, Martinique, and R6union are mecluded mn France. c. On July 1, 1997, China resumed its exercise of sovereignty over Hong Kong. d. On December 20, 1999, China resumed its exercise of sovereignty over Macao. Source: World Bank data. 129 GLOBAL DEVELOPMENT FINANCE Table A1.7 Classification of economies by income group and indebtedness, January 2002 Income group Subgroup Severe'y indebted Moderately indebted Less indebted Nor classified by indebtedness Low- Afghanistan Madagascar Burkina Faso Armenia income Angola Malawi Cambodia Azerbaijan Benin Mauritania Gambia, The Bangladesh Burundi Myanmar Ghana Bhutan Cameroon Nicaragua Haiti Eritrea Central Africen Niger Kenya Georgia Republic Nigeria Mali India Chad Pakistan Moldova Korea, Democratic Pople's Comoros Rwanda Mongolia Republic of Congo, Dem. S5o Tom Mozambique Lesotho Republic and Principe Senegal Nepal Congo, Rep. of Sierra Leone Togo Solomon C6te d'lvoire Somalia Uganda Islands Ethiopia Sudan Uzbekistan Ukraine Guinea Tajikistan Yemen, Republic Vietnam Guinea-Bissau Tanzania Zimbabwe Indonesia Zambia Kyrgyz Republic Lao PDR Liberia Middle- Lower Cuba Algeria Albania Macedonia, Marshall Islands income Ecuador Belize Belarus FYR oP Micronesia, Fed. Sts. Guyana Bolivia Cape Verde Maldives West Bank and Gaza Iraq Bosnia and China Morocco Jordan Herzegovina Djibouti Namibia Peru Bulgaria Dominican Paraguay Syrian Arab Colombia Republic Romania Republic Honduras Egypt, Arab Sri Lanka Jamaica Rep. of Suriname Papua New Guinea El Salvador Swaziland Philippines Equatorial Tonga Russian Federation Guinea Vanuatu Samoa Fiji Yugoslavia, St. Vincent and Guatemala Fed. Rep. of the Grenadines Iran, Islamic Thailand Rep. of Tunisia Kazakhstan Turkmenistan Kiribati Latvia Lithuania Upper Argentina Chile Antigua and Seychelles American Samoa Brazil Croatia Barbuda Slovak Republic Isle of Man Gabon Estonia Bahrain South Africa Mayotte Gibraltar Botswana St. Kitts and Palau Hungary Costa Rica Nevis Puerto Rico Lebanon Czech Republic St. Lucia Malaysia Dominica Trinidad and Mauritius Grenada Tobago Panama Korea, Rep. of Turkey Libya Uruguay Mexico Venezuela, R.B. de Oman Poland Saudi Arabia 130 DEBT BURDEN INDICATORS AND COUNTRY CLASSIFICATIONS Table A1.7 Classification of economies by income group and indebtedness, January 2002 (continued) Income group Subgroup ISvry ld Moderately indebted I btiZ ed Nor classified by indebtedness High- OECD Australia Japan income Austria Luxembourg .. Belgium Netherlands S Canada New Zealand Denmark Norway Finland Portugal Franceb Spain Germany Sweden Greece Switzerland Iceland United Ireland Kingdom Italy United States Non- Andorra Kuwait OECD *Aruba Liechtenstein Bahamas, The Macao, Chinac Barbados Malta Bermuda Monaco Bruneti Netherlands Cayman Antilles Islands New Caledonia Channel N. Mariana Islands Islands Cyprus Qatar Faeroe Islands San Marino French Singapore Polynesia Slovenia Greenland Taiwan, China Guam United Arab Hong Kong, Emirates Chinad Virgin Israel Islands (U.S.) Total: 209 4 44 61 This table classifies all World Bank member economies, and all other economies with populations of more than 30,000. Economics are divided among income groups according to 2000 GNI per capita, calculated using the World Bank Atlas method. The groups are low-income, $755 or less; lower-middle-income, $756-$2,995; upper-middle-income, $2,996-$9,265; and high-income, $9,266 or more. Standard World Bank definitions of severe and moderate indebtedness are used to classify economies in this table. Severely indebted means either present value of debt service to GNI exceeds 80 percent or present value of debt service to exports exceeds 220 percent. Moderately indebted means either of the two key ratios exceeds 60 percent of, but does not reach, the critical levels. For economies that do not report detailed debt statistics to the World Bank DRS, present-value calculation is not pos- sible. Instead, the following methodology is used to classify the non-DRS economies. Severely indebted means three of four key ratios (averaged over 1998-2000) are above critical levels: debt to GNI (50 percent); debt to exports (275 percent); debt service to exports (30 percent); and interest to exports (20 percent). Moderately in- debted means three of the four key ratios exceed 60 percent of, but do not reach, the critical levels. All other classified low- and middle-income economies are listed as less indebted. a. Former Yugoslav Republic of Macedonia. b. The French overseas departments French Guiana, Guadeloupe, Martinique, and R6union are included in France. c. On December 20, 1999, China resumed its exercise of sovereignty over Macao. d. On July 1, 1997, China resumed its exercise of sovereignty over Hong Kong. Source: World Bank data. 131  Appendix 2 Commercial Debt Restructuring Developments in 2001 amount. The total operation cost came to about In 2001, 26 debt-restructuring agreements be- $2.4 million, of which $0.4 million was disbursed tween debtor countries and their commercial from IDA and $2 million was financed by partici- creditors were completed, restructuring about pating bilateral donors. The acceptance rate of the $104.9 billion of outstanding debt.1 Among low- creditors in the operation was 90 percent and thus income countries, Honduras, Tanzania, and the exceeded the required minimum threshold level of Republic of Yemen bought back $452 million of 75 percent. principal debt through the International Develop- Tanzania. Sponsored by the IDA Debt Reduc- ment Association (IDA) Debt Reduction Facility tion Facility and the governments of Germany and (see table A2.1).2 Ukraine exchanged $21.5 Switzerland, Tanzania completed the first phase of million of its external debt to retire the remaining a debt buyback operation in April. The operation amount outstanding from last year's bond ex- extinguished $76.6 million of eligible principal change operation. Among middle-income coun- debt and about $79.2 million of associated inter- tries, Argentina restructured $85.7 billion of est. The debt was bought back at a price of 12 domestic and external debts through a series of cents to a dollar of principal with a 5 percent for- debt exchange operations to manage the near-term eign exchange risk margin. About $10.1 million debt-service profile. Brazil, Jordan, and Mexico was paid to the eligible creditors, and the creditors retired about $8.6 billion of their collateralized contributed about $65 thousand to Tanzanian Brady bonds through Brady-Eurobond exchange nongovernmental organizations under the Debt and straight-buyback operations. Colombia and for Development Clause. Turkey restructured $10.5 billion of domestic The Republic of Yemen. Funded by the IDA debt through Treasury notes exchange and an Debt Reduction Facility and the government of auction process. Also, Panama bought back $160 Norway, Switzerland, and the Netherlands, in million of a global bond with cash payment and June the Republic of Yemen completed a debt buy- warrants. back operation to retire $362 million of principal and $245 million of associated interest. This oper- IDA-sponsored debt buybacks ation included debt owed to about 50 creditors- in low-income countries suppliers' credit was about 85 percent of the prin- Honduras. In August, Honduras completed the cipal debt, and commercial debt amounted to first phase of a debt buyback operation, which about 15 percent. The buyback price for the oper- was funded by the IDA Debt Reduction Facility ation was set at 2.94 cents per dollar of the princi- and co-financed by the governments of the pal debt.3 Total operation cost amounted to about Netherlands, Norway, and Switzerland. This oper- $11.4 million, of which $7.6 million came from ation extinguished $13 million (principal only) of IDA and $3.8 million came from participating hi- the total $14.5 million of eligible debt. The pur- lateral donors. The acceptance rate of the creditors chase price was 18 cents per dollar of the principal in the operation was about 91 percent. 133 GLOBAL DEVELOPMENT FINANCE Table A2.1 Completed operations financed by the IDA Debt Reduction Facility as of December 2001 (millions of dollars) Percentage - of eligible Principal - principal fll -f Country at extinguished - d i extinguished Albania July 1 5 371.3 AI . 99.0 97.4 Bolivia May 1993 170.0 - 94.0 2'7.3 C6te d'lvoire arch 19 8 724.5 100.0 39 Ethiopia January 1 9 226.0 _ 0 80.0 18.8 Guinea pril 20 0 62.2 IMI3 - 75.0 Guyana Novembe 992 69.2 t._ 800 100.0 0.2 Guyana August1 99 34.4 _ 62.0 3.4 Honduras August:200' 13.0 " 18. 89.8 2.4 Mauritania August.996 53.0 E 100 98.0 5.9 Mozambique Iecembe 19 123.8 1064.0 3.4 Nicaragua lcembe 995 1,099.4 10 . 81.0 89.2 Niger arch 1991 107.0 -- DA$ 99.0 19. Sio Tom6 and Principe August 1994 10.1 A RM 87.0 1. Senegal becember 1996 71.0 - N 96.0 5 Sierra Leone September 1995 234.7 -- 8 - 73.0 .5 Tanzania Apri 200 76.6 $20 71.8 10. Togoecembe 9' 44.9 85 99.0 6.1 UgandFebrua 99 153.0 1 & 89.0 2F2.6 VietnMarch 98 20.4 460 6.6 .0 Yemen, Republic of Pebruary20 362.4d .- 91.4 11.4 ZambiaSe er 1994 199.7 78.0 4.9 a. Of original face value of principal. b. Includes technical assistance grants. c. 16 cents for cash buyback and 20 cents for long-term bonds. d. Excludes $40.7 million of debt owed to the Czech Republic and the Slovak Republic. e. The effective buyback price reflecting a previous debt reduction of 80 percent on the Russian debt. Source: World Bank staff estimates. Other debt restructuring Debt buyback and swaps in low-income countries in middle-income countries Ukraine. In March, following the successful com- Argentina. The government of Argentina success- pletion of last year's bond exchange, the Ukrainian fully completed the first debt swap for the year in government conducted a new bond exchange op- February, swapping $4.8 billion of its short- and eration with the objective of swapping the remain- medium-term peso and dollar-denominated debt for ing bonds from a minuscule 0.8 percent of in- a new $2.6 billion Bonte (Bonos del Tesoro, or Trea- vestors who did not take part in the last year's sury bond) due 2006 and a new $1.6 billion global exchange. Under this operation, about $21.5 mil- bond due 2012. This transaction converted $2.9 bil- lion of the external commercial debt was ex- lion of bonds into a new 5-year Bonte and $1.9 bil- changed with the remaining amount of the nonex- lion of bonds into a new 11-year global bond. The changed eligible debt at approximately $540 new global bond was priced at par with a coupon of thousand (less than 0.075 percent of total eligible 12.375 percent while the new Bonte came at par debt). Participating investors received a six-year with a coupon of 11.75 percent. Of the 17 bonds el- Eurobond, denominated in either euro at an inter- igible for the exchange, 12 bonds were exchanged at est rate of 10 percent or U.S. dollar at an interest a discount, and S bonds were swapped at a pre- rate of 11 percent. Bonds eligible for the swap mium. The net present value saving of about $18.5 were a deutsche mark 16 percent Eurobond due in billion resulted from a net reduction in interest pay- February, euro 10 percent amortizing notes due in ments. March 2007, U.S. dollar 11 percent amortizing In June 2002, Argentina carried out one of the notes due in March 2007, and U.S. dollar 11 per- largest debt swap operations ever conducted by a de- cent amortizing notes due in March 2007. veloping country, including almost a quarter of its 134 COMMERCIAL DEBT RESTRUCTURING outstanding peso- and dollar-denominated bonds. was funded out of the collateral liberated from This operation replaced about $29.5 billion of do- swapped Brady bonds. Brady bonds exchanged for mestic and external bonds with five new securities the new global bond included $682 million of par with a face value of $30.4 billion. Foreign investors bonds, $1.16 billion of discount bonds, $212 ril- exchanged between $8 billion and $9 billion in this lion of C-bonds, and $47 million debt conversion deal. Local banks and pension funds swapped the re- bonds (DCBs).6 mainder of tender. The swap extended the average Colombia. Colombia completed a massive $2.5 maturity of the government's debt by 2.78 years and billion (Ps5.6 trillion) swap of domestic debts in the average yield of the new bonds was close to 15 June 2001 to reduce financing needs for next year. percent. The deal reduced the debt servicing cost by The government split the swap into three transac- $16.5 billion through the end of 2005 on account of tions through which participating investors could the 5-year grace period. However, it increased Ar- exchange their old bonds for 3-year, 5-year, and 10- gentina's overall debt by about $2.25 billion. year Titulos de Tesoreria notes (Treasury notes). In August the government arranged a voluntary About 54 percent of the old debt was exchanged for swap of Treasury securities maturing the second half a 10-year note with a 7.8 percent coupon (Ps3 tril- of 2001 with major private domestic banks for a lion). A further 32 percent of debt was swapped for total of $1.4 billion. Participating banks were given 3-year notes with a 15 percent coupon (PsI.8 tril- the option of choosing between new 1-year and 3- lion) and the remaining 14 percent was placed in 5- year bonds. In December, Argentina successfully year notes (PsO.8 trillion). The swap lengthened the completed a massive local debt swap with domestic average maturity of government bonds to 4.5 years banks and pension funds. This operation aimed to from 3.4 years, and reduced the amortization costs extend bond maturities and reduce near-term inter- by an average of 32 percent through to 2003.7 est and amortization payments on local bonds. Mexico. In 2001, the Mexican government car- About $50 billion of domestic bonds-$41 billion in red out three swap operations to buy back $4.74 federal debt and $9 billion in provincial debt-was billion of Brady bonds, and conducted five straight exchanged for new lower coupon loans guaranteed buybacks to retire about $1.025 billion of Brady by tax collection4 (see table A2.2). The new loans bonds through open market repurchase.8 In the first extended maturities of all debt due before 2010 by exchange, in March, Mexico issued a $3.3 billion three years and lowered the interest rates by at least 18-year global bond with a yield of 9.3 percent in 30 percent compared with the original bond, with a exchange for Brady bonds (pars and discounts) with cap of 7 percent for fixed rate bonds and Libor+ 300 the same nominal value. The deal achieved roughly basis points for floating bonds. This operation al- $115 million of net present value savings and freed lowed the federal government to reduce amortiza- up about $1.5 billion in collateral. In May Mexico tion payments by $2.5 billion in 2002, $3 billion in swapped $1.05 billion of outstanding Brady bonds 2003, and $2.4 million in 2004. It also saved $3.6 for $1 billion of a 10-year global bond through a re- billion of interest payments in 2002, $2.4 billion for opening of its 10-year; $1.5 billion bond issued in the federal government and $1.2 billion for the January 2001. The reopening was priced to yield provinces. 8.426 percent to maturity, the lowest ever for a 10- Brazil.s In March, Brazil completed a par-for- year global issue by Mexico. This transaction re- par exchange of Brady bonds for a new global bond leased $475 million of underlying collateral and re- due 2024. The new $2.15 billion global issue carried alized net present value savings of around $60 a relatively low coupon of 8.875 percent and was million. In August, Mexico completed a $1.5 billion, priced at a deep discount, around 72 cents on the 30-year global bond issuance involving a $1.06 bil- dollar. Unlike the typical debt swap where the old lion cash sale and a $440 million exchange for out- bonds are exchanged for a smaller number of new standing Brady bonds. The new issue was priced at a instruments, this operation traded old bonds for new deeply discounted $92.58, yielding 9.02 percent and bonds at par value but with a low coupon. In order a spread of 335 basis points over U.S. Treasuries to compensate for the low coupon, the government with an 8.3 percent coupon. By retiring the swapped offered participating investors a cash premium to Brady bonds, Mexico realized about $10 million of make up any difference in the market value of Brady net present value savings and liberated $200 million bonds and the new global bond. The cash payment of collateral. 135 GLOBAL DEVELOPMENT FINANCE Between April and November 2001, Mexico cent for debt due in 20029, as well as $2.45 billion carried out five buyback operations to fully extin- of five-year U.S. dollar-indexed bond priced to guish about $1.03 billion of its outstanding Brady yield a 50 basis points premium to the three-year debts by exercising call options and private buy- dollar indexed bond10. They also received $1.5 backs. By exercising the embedded call options, billion of one-year FRNs (floating-rate notes) and Brazil fully retired around $10 million (NLG31.3 $1.1 billion of two-year FRNs. This operation ex- million) of Dutch guilder-denominated discount tended the average maturity of domestic debt from bonds in April; $15 million (JPY1.73 billion) of 5.3 months to 37.2 months. Debt service costs Japanese yen-denominated discount bonds in were estimated to be reduced by around $1.6 bil- May; $300 million of U.S. dollar-denominated lion. discount bonds in October and $400 million (FRF3 billion) of French franc-denominated par Notes bonds in December. In October 2001, Mexico also 1. Transaction information was based on World Bank bought back around $100 million (CAD162 mil- staff research and Goldman Sachs' debt swap data, which lion) of Canadian dollar-denominated discount was provided by David Cohen. bonds through private open-market repurchase. 2. About $324 million of associated interest was also Jordan. In 2001, according to the Jordanian extinguished. Ministry of Finance, Jordan conducted six buy- 3. The buyback price reflects an initial debt reduction back operations to retire $44 million of its out- of the former Soviet Union commercial debt at zero cost. Ex- standing Brady bonds. Under these operations, cluding a previous debt reduction, the buyback price was set Jordan bought back its Brady par bonds at an av-debt. Jordn bughtbac itsBray pa bods a anav-4. The central bank is planning to use the financial erage price of 78.8 cents per dollar of face value. transaction tax to fulfill guarantees. Total operation cost, including the arrangement 5. There were six other Brady bond exchanges by the fees of $100.5 thousand, came to about $36.2 mil- Brazilian government in 2001, swapping about $3 billion of lion. These repurchase operations were financed outstanding Bradys, including $106.8 million of par bonds by the proceeds from privatization and the sale of in June, $371 million and $108 million of Els(eligible inter- theest bonds) in June and July, $2.29 billion and $220 million the eleaed cllatral.of discount bonds in June and July, $7.5 million of new Panama. In July 2001, Panama bought back money bonds in June, and $18.6 million of FLIRBs (front- $160 million of its global bond due in 2002 using loaded interest reduction bond) in June. The ministry of fi- the proceeds from a reopening of a $750 million nance exchanged its Brady bonds for the domestic securities global bond issued in February 2001. This opera- with the central bank and some other domestic banks, turn- tion represented the new approach to management ing dollar-denominated debt into real-denominated debt. of the country's debt service profile. The govern- These transactions were not included in the total debt re- ment offered bondholders a cash payment of structured, since they were between the government agen- $1,039 per $1,000 of face value to compensate for 6. Brady bonds eligible for the exchange included U.S. the loss of a coupon payment, as well as two war- dollar par series Z-L bonds due 2024 (par bonds), U.S. dol- rants allowing investors to exercise an option to lar discount series Z-L bonds due 2024 (discount bonds), buy new 10-year global bonds with a 9.625 per- U.S. dollar front-loaded interest reduction with capitaliza- cent coupon to be issued in January 2002 for $990 ton series L bonds due 2014 (C bonds) and U.S. dollar debt in cash. The buyback generated estimated savings conversion series L bonds due 2012. of san 7. Amortization payments were cut by 22.5 percent for of $.6 illon i tems f dbt-srvie cst.2001 to Ps41 trillion from Ps3.2 trillion, 34.1 percent for Turkey. In June, the Turkish government con- 2002 to Ps5.4 trillion from Ps8.2 trillion, and 40 percent for ducted a voluntary debt swap of short-term lira- 2003 to Ps1.8 trillion from Ps3trillion. denominated debts into longer-term dollar-linked 8. Mexico has also likely bought back some other debts and floating-rate lira paper through a Dutch style through open market purchase, but in undisclosed amounts. auction. About $7.3 billion of debt was swapped 9. At the supplemental auction, three-year dollar-in- in the initial auction and $730 million was ex- dexed bond was priced to yield 14.39 percent for debt ma- changedturing in 2001 and 13.95 percent for debt maturing in 2002. chaned n th folowng spplmenary ucton.10. The yield of five-year dollar-indexed bond is calcu- Participating debt holders received $2.45 billion lated by using the interest rate, which is determined by the of three-year U.S. dollar-indexed bond yielding acceptance rate for 3-year dollar-indexed bond plus 50 basis 14.89 percent for debt due 2001 and 14.45 per- points spreads. 136 C O M M E R C I A L D E BT RESTRUCTU RING Table A2.2 Multilateral debt relief agreements with commercial banks, January 1980-December 2001 Amount restructured Other assistance Repayment terms Consolidation period (millions of U.S. dollars) (millirns of US dollars) (conisolidation portion onlv) New Short term , Maturity Grace Country and Length log-term credit (years/ (years! Interest date of agreement Start date (montbn) Deferment Rescheduling money mantenance months) months) (margin) Albania july 1995 Debt buyback (see notes) Algeria February 1992 Ste nlotes 1,500 5-8/0 3/0 1/11% june 1995 March 1994 3,200 12/6- 16 6/6 % Argentina January 1983 l january 1983 12 1,300 1/2 0/7 1Y August 1983 500 4/6 3/0 2Y August 1985 1 January 1982 48 9,777 3,593 3,100 10/0 3/0 im August 1987 See notes 24,286 1 253 3,500 19/0 7/0 % April 1993 DDSR agreement (see notes) September 1997 Voluntary debt swap (see notes) Mar/Sept. 1998 Debt buyback (see notes) March 1999 Voluntary debt swap and debt buyback (see notes) Feb/june 2000 Voluntary debt swap (see n,otes) FebJJune/Aug. 2001 P..nu 24 Oct. 00 11 Mj Incrrn 12 "1Ililt.knu In, Bolivia 18 July 86 31 De i5 I lui. .t 12 1.' 44" 14 Nov. 88 31 Dec 5 1 'Ir %, I . 1.' 22" u 15 Mar. 90 31 D.c 5 IUn '2 i' 2~M 24Jan.92 31 Dc I II ll2 5 \eun ile.u 24 Mar. 95b I De S5 I I .i 3re I. 4 lr'.u 14 Dec. 95 jI Dec. l E. K sN-. ',i %. I ..u Ni, 30 Oct. 98 11 D,: Sr I1.. ire.. I" So \ku Mc, 10July01 1 D . [ 5 I il. Ij.. " i N uu NL,.L u Bosnia and Herzegovina 30 Oct. 98 2 Uec i Itu S9 ...uMn 7July 00 21 c1 I··I .i, '. 1 .,, Brazil 23Nov.83 II bP 3 \u,: .13 -n Sn 21Jan.87 -M r i hu , ·' . -o' . :r éu 29 July 88 3l ti I \uc 55 2 -n 4.4' - Nue 26 Feb. 92 Ma bI c3 I I 2 2' - ,I 100. 13 4 1. 1" Bulgaria 17 Apr. 91 1 lr' iA r j 12 - Ar I'i 'All. 14 Dec. 92 1 1 , I L CC lI" 251 i, l r. 4 13 Apr. 94 <4 I -n. - it r 4 1' -100 200 e 511 Burkina Fasot 15 Mar. 91 1 91 I Mat i 1 lui I Nenu \rn, 7 May 93 bn 91 icr "3 lv'0 36 Menu ler.. 20June 96 ln 91 2 lun, t10 4 \ler.. Mn 24 Oct. 00 iie«r. rebel I - i Nleru .en. Cambodia 26Jan. 95 31 Dec SJ I i)n A 100 249 Nltnu Menu (Table contmnues ont next page) 155 GLOBAL DEVELOPMENT FINANCE Table A3.2 Multilateral debt relief agreements with official creditors, January 1980-December 2001 (continued) Consolidation period Consolidation includes Repayment terms2 for current maturities Share Amount Contract Previously of debt consolidated Maturity Grace Country and date cutoff Start Length rescheduled consolidated (milfions of (years/ (years/ of agreement date date (months) Arrears debt (percent) dollars) months) months) Cameroon 24 May 89 31 Dec. 88 1 Apr. 89 12 y 100 535 9/6 6/0 23 Jan. 92 31 Dec. 88 1 Jan. 92 9 y 100 1,080 19/5, 14/8 9/11, 8/2 24 Mar. 94 31 Dec. 88 1 Apr. 94 18 y y 100 1,259 Menu Menu 16 Nov. 95 31 Dec. 88 1 Oct. 95 12 y 100 1,129 Menu Menu 24 Oct. 97 31 Dec. 88 1 Oct. 97 36 y 100 1,270 Menu Menu 24 Jan. 01 31 Dec. 88 1 Jan. 01 35 y 100 1,300 Menu Menu Central African Republic 12 June 81 1 Jan. 81 1 Jan. 81 12 y 85 72 8/6 4/0 9 July 83 1 Jan. 83 1 Jan. 83 12 y 90 13 9/6 5/0 22 Nov. 85 1 Jan. 83 1 July 85 18 y 90 14 9/3 4/9 14 Dec. 88 1 Jan. 83 1 Jan. 89 18 y 100 28 Menu Menu 15 June 90 1 Jan. 83 1 Jan. 90 12 y. y 100 4 Menu Menu 12 Apr. 94 1 Jan. 83 1 Apr. 94 12 y y 100 32 Menu Menu 25 Sept. 98 1 Jan. 83 1 Sept. 98 34 y y 100 23 Menu Menu Chad 24 Oct. 89 30 June 89 I Oct. 89 15 y 100 24 Menu Menu 28 Feb. 95 30 June 89 1 Apr. 95 12 y 100 24 Menu Menu 14 June 96 30 June 89 1 Jan. 96 32 y y - 100 12 Menu Menu 13 June 01 30 June 89 1 May 01 24 y y 100 15 Menu Menu Chile 17 July 85 1 Jan. 85 1 July 85 18 65 146 6/3 2/9 2 Apr. 87 1 Jan. 85 15 Apr. 87 21 85 157 6/2 2/7 Congo, Democratic Republic of 9 July 81 1 Jan. 79 1 Jan. 81 24 90 500 9/6 4/0 20 Dec. 83 30 June 83 1 Jan. 84 12 y y 95 1,497 10/6 5/0 18 Sept. 85 30 June 83 1 Jan. 85 15 y 95 408 9/5 4/11 15 May 86 30 June 83 1 Apr. 86 12 y 100 429 9/6 4/0 18 May 87 30 June-83 I Apr. 87 13 y y . 100 671 14/6 6/0 23 June 89 30 June 83 1 June 89 13 y y 100 1,530 Menu Menu Congo, Republic of 18 July 86 1 Jan. 86 1 Aug. 86 20 y 95 756 9/2 3/8 13 Sept. 90c 1 Jan. 86 1 Sept. 90 21 y y 100 1,052 14/3 5/9 30 June 94c - Jan. 86 1 July 94 11 y 100 1,175 19/7,14/7 10/1,5/1 16 July 96 ._1Jan. 86 30 June 96 36 y y 100 1,758 - Menu Menu Costa Rica 11 Jan. 83 1 July 82 1 July 82 18 v 85 136 8/3 3/9 22 Apr. 85 1 July 82 1 Jan. 85 15 y 90 166 9/5 4/11 26 May 89 1 July 82 1 Apr. 89 14 y 100 182 9/5 4/11 17 July 91 1July 82 1 July 91 9 y - y 100 139 9/7 5/1 22 June 93 1 July 82 Arrears as of 31 June 93 y 100 58 6/6 2/0 C6te d'Ivoire 4 May 84 1July 83 1 Dec. 83 13 100 230 8/6 4/0 25 June 85 1 July 83 1 Jan. 85 12 100 213 8/6 4/0 27 June 86 1 July 83 1 Jan. 86 36 Variable 370 8/7 4/1 18 Dec. 87 1 July 83 1 Jan. 88 16 y y 100 567 9/4 5/10 18 Dec. 89 1 July 83 1 Jan. 90 16 y y 100 934 13/4 5/10 20 Nov. 91 1 July 83 1 Ocr. 91 12 y y 100 806 14/6 8/0 22 Mar. 94 1 July 83 1 Mar. 94 37 y y 100 1,849 Menu Menu 24 Apr 98 1 July 83 1 Ap. 98 36 y y 100 1,402 Menu Menu 156 OFFICIAL DEBT RESTRUCTURING Table A3.2 Multilateral debt relief agreements with official creditors, January 1980-December 2001 (continued) -l Cc-n,eI.J,.ir r,a .On ld1o shiJCs Rcep.nmcni ut,msa ..nra-_Pre n , , fldbi h-lJjJA ur,1, Grace CouIer,n -r- InJu J m. Jjiid .dIj[.- -m "ar: 1cears/ of...''-r..ni';I,Li,.-. Jh'u' ddueL---S' "U,r[..' *'ionths) 21 'ih '5 - 2 ln "5 12 -I"' 'i 2/1 Cubi 1 Abr i 3/11 21 b .h l I Im " s 2 l'."' 5/0 .12 ..ur y5 [lk .- lU >r 17 June 98 lad ai 3 - -'lu S c ,2 n . 24 Oct. 00 a iplreite > tMu Melaii- Sierra Leone 8 Nov. 80 I li 1 il 4r. 41 - 8 Feb. 84 a 13- 4 12 ¾ luri %.u 19 Nov. 86 I 1 l, Nr.$I lul I "4 4. ju 20 Nov. 92 l I N- "2 1l 00 N-krLMer.u Men,. 20 July 94 I Aug r . 10 . kna Mlenu 25Apr.96 I Jap 4n 24 1l,tMu lua,u 16 Oct. 01 I k\ t ijl 3,, 0 Myn u Somalia 6 Mar. 85 1 |an xi 12 t5 'o S:f 22 July 87 II n 24 al1a Sudan 18 M ar. 82 I j|la .,I I lul, '< I y3y 4Feb.83 1 an 1 ian *s3 12 2 May 84 in1 . 4-r (Table contrmnues on next page) 161 GLOBAL DEVELOPMENT FINANCE Table A3.2 Multilateral debt relief agreements with official creditors, January 1980-December 2001 (continued) Consolidation period Consolidation includes Repayment terms" for current maturities Share Amount Contract Previously of debt consolidated Maturity Grace Country and date cutoff Start Length rescheduled consoldated (millions of (years/ (years/ of agreement date date (months) Ariars debt (percent) dollars) months) months) Tanzania 18 Sept. 86 30 June 86 1 Oct. 86 12 y 100 1,046 9/6 5/0 13 Dec. 88 30 June 86 1 Jan. 89 6 y y 100 377 Menu Menu 16 Mar. 90 30 June 86 1 Jan. 90 12 y y 100 199 Menu Menu 21 Jan. 92 30June 86 I Jan. 92 30 y ,y' 100 691 Menu Menu 21 Jan. 97 30 June 86 1 Dec. 96 36 y y 100 1,608 Menu Menu 14 Apr. 00 30 June 86 31 Mar. 00 36 y y 100 714 Menu Menu Togo 20 Feb. 81 1 July 80 1 Jan. 81 24 85 232 8/6 4/0 12 Apr. 83 1.Jan. 83 1 Jan. 83 12 y y 90 300 9/6 5/0 6 June 84 1 Jan. 83 1 Jan. 84 16 y 95 75 9/4 4/10 24 June 85 1 Jan. 83, 1 May 85 12 95 27 10/6 5/0 22 Mar. 88 1 Jan. 83 1 Jan. 88 15 y 100 139 15/5 7/11 20 June 89 1 Jan. 83 16 Apr. 89 14 y 100 76 Menu Menu 9 July 90 1 Jan. 83 1 July 90 24 y 100 88 Menu Menu 19 June 92d 1 Jan. 83 1 July 92 24 y, 100 52 Menu Menu 23 Feb. 95 1 Jan. 83 1 Feb. 95 33 y y 100 237 Menu Menu Trnidad and Tobago 25 Jan. 89 1 Sept. 88 1 Jan. 89 14 y 100 209 9/5 4/11 27 Apr. 90 1 Sept. 88 1 Mar. 90 13 100 110 8/4 3/10 Turkey 23 July 80* 30June80 1 July 80 36 y 90 3,000 9/0 4/6 Uganda 18 Nov. 81 1 July 81 1 July 81 12 y 90 30 9/0 4/6 1 Dec. 82 fJuly 81 1 July 82 12 90 19 9/0 4/6 19 June 87 1July 81 1 July 87 12 y y 100 170 14/6 6/0 26 Jan. 89 1-July 81 1 Jan. 89 18 y y 100 89 Menu Menu 17 June 92 1July 81 1 July 92 17 y y 100 39 Menu Menu 20 Feb. 95 1 July 81 1 Feb. 95 Stock y y 100 110 Menu Menu 24 Apr. 98 1 July 81 1 Apr. 98 Stock 100 148 Menu Menu 12 Sept. 00 1July 81 1 Sept. 00 Stock y 100 150 Menu Menu Ukraine 13 July 01 31,Dec. 98 19 Dec. 00 22 100 580 12/0 3/0 Vietnam 14 Dec. 93 1 Jan. 90 Arrears as of 31 Dec. 93 y 100 791 Menu Menu Yemen, Rep. of 24 Sept. 96 1 Jan. 93 1 Sept. 96 10 y 100 113 Menu Menu 20 Nov. 97 1Jan. 93 1 Nov. 97 36 y 100 1,444 Menu Menu 14 June 01 1 Jan. 93 31 Dec. 00 Stock y 100 420 Menu Menu Yugoslavia, Fed. Rep. (Serbia/ Montenegro) 22 May 84* 2 Dec. 82- 1 Jan. 84 12 100 500 6/6 4/0 24 May 85* 2 Dec. 82 1 Jan. 85 16 90 812 8/4 3/10 13 May 86* 2 Dec. 82 16 May 86 23 85 9011 8/6 4/0 13 July 88* 2 Dec. 82 1 Apr. 88 is y 100 1,291 9/5 5/11 16 Nov. 01 1 Apr. 88 20 Dec. 00 Stock y 100 4,500 Menu Menu 162 OFFICIAL DEBT RESTRUCTURING Table A3.2 Multilateral debt relief agreements with official creditors, January 1980-December 2001 (continued) Consohdaton period nsolidation includes Repayment terms, for current maturities Share n Previously of debt Maturity Grace Country and date Sta Length I consolidated of agreement date (months) r t (percent) months) months) Zambia 16 May 83 jA1 Jan 83 12 90 9/6 5/0 20 July 84 81 Jan 84 12 100 9/6 510 4 Mar. 86 n1 Jan 86 12 100 9/6 510 12 July 90 .1July 90 18 100 Menu Menu 23 July 92 1Jn 83 1 July 92 33 100 Menu Menu 27 Feb. 96 1 Jan 96 36 yent1005nen. 16 Apr. 99 1an. 83 1 Apr.-99 36 ~ 'y100 100 Menu 1Menn Nut applicable. The rescheduling was concluded outside of formal Parts Club auspices. "Menu" terms refer to the options agreed to at the 1988 Toronto economic summit meeting. Note. Tbe figures in this table are commitment values (amounts of agreed-on debt relief). They corre spond to the disbursement figures (minus debt forgiveneqs, wshlen applicable) for debt restructuring shown in the Global Development Finance: Country Tables. All agreements shown in this table, except those indicated witnh son as- terisk, were negotiated through the Paris Club. a. Maturity is measured here from the end of the consolidation period to the date of the final amortization paymenst; the grace period is the time between the end of the consolidation period and the date of the first amortization payment. The secretariat of the Paris Club measures grace and maturity froim the midpoint of the con- solidation period. b. The agreement signed in March 1995 covered a 36-month period, hutsa new agreement wsas signed in December 1995 covering the stock of debt, starting 12 months after the beginning of the consolidation period of the previous agreement. c. Agreement with a Paris Club-designated lower-middle-income country with heavy official debt. These agreements also allows for debt conversions, subject to the limit for each creditor country (for non-ODA debt) of US$10 million or 10 percent of the debt outstanding as of the beginning of the consolidation period, whichever is higher. Where two sets of figures for repayment terms (maturity and grace) are given, the first set represents ODA debt and the second non-ODA debt. d. The agreement was canceled. e. The agreement was implemented in 1991 because of the agreement's conditionality on an IMF program that took place in 199 1. I. Agreement follows the deferral signed in January 1992 by the former Soviet republics. Sources: World Bank DRS; IMF; Paris Club Secretariat. 163  Appendix 4 Regional Economic Developments and Prospects East Asia and Pacific Europe and Central Asia Latin America and the Caribbean Middle East and North Africa South Asia Sub-Saharan Africa East Asia and Pacific GNI per capita, 2000: $1,060 Recent developments Across the region fiscal balances Growth in the East Asia and Pa- deteriorated, but the impact of this GDPgroth n Est sia 199-201cific region slowed sharply in stimulus on domestic inflation and ex- GDP2001, from 7.4 percent in change rate levels was limited. For Year-over-year percentage change 2000 to 4.6 percent. Excluding China, Thailand, Korea, and the Philippines, 12 \Korea, Rep. of growth slowed more, from 7 percent currency depreciation vis-a-vis the dol- in 2000 to 2.3 percent in 2001. The lar ranged between 2 and 6.5 percent a ~downturn in growth was concentrated during 2001. The 9 percent deprecia- / Chna in the high-tech exporting countries, tion of the rupiah, despite a 400 basis outheast Asia which suffered disproportionately point increase in policy interest rates, from the global recession in high-tech made Indonesia an exception in the re- o - sectors.1 Chinese growth was main- gion. ta-ed at rates above 7 percent Despite a decisive reduction in C5~ C gp 'b 0 through substantial fiscal stimulus, current account surpluses, the region 20 /P 1 , &6b d' o 000 IF c~ And Vietnam grew by 5.5 percent by saw a $50 billion increase in reserves Souce.Word Bnkfurther increasing its share in export during the year. The improved exter- Souemarkets. But, partly related to its ex- nal positions since the East Asian crisis port success, the country suffered sig- made an increase in reserves possible nificant terms-of-trade losses, despite a fall in export revenues and Industrial production in East Asia, 1999-2002 Regional merchandise export vol- domestic stimulus, and a decline in Three-month moving average, year-over-year percentage change umes showed little to no growth in capital flows. Indonesia, struggling 30 0*Korea, Re f2001, following an advance of 23 per- with a difficult political and social en- cent in 2000. And dollar export rev- vironment, was one of the few coun- 20 ha enues fell (3.7 percent) for the first tries where reserves declined. atime since 1998. Moreover, loss of Recrn revenue was significant in the 0 wake of September 11, especially for Capital market flows decline Southeast AsiG Thailand and the Philippines. Worker Gross flows from the international fremittances in the latter country were capital markets fell dramatically dur- Jane e 2000 2000 200J July Jan down sharply, a direct effect of the ing 2001. The aggregate of bond and slowdown in world trade and tourism, equity issuance and bank lending Source: iataherhaiga-decrlexBanktstaffoestriete because a significant portion of remit- dropped by $25.8 billion in the year tances come from Philippine seamen (almost 40 percent) to reach a level of on cargo and cruise vessels. $40.9 billion-only moderately above Export in dollars, growth for East Asia, Many countries in the region were the outturns of 1998, at the peak of 1999-2001 able to use countercyclical monetary the East Asian crisis. In contrast with Three-month moving average, year-over-year and fiscal policies to limit the adverse conditions at that time-large-scale percenta4e0cane V effects on growth caused by worsening withdrawal of short-term banking 30external conditions. Rapidly falling in- debt-the 2001 downturn in flows re- ternational interest rates facilitated flects in part a decrease in demand for ardomestic interest rate declines in sev- funds in the region, as well as limited to eral countries. Over the course of investor supply of capital for high-tech 0Suheas 2001, policy rates fell 140 basis points ventures. The falloff in flows to East A0csiaa in the Republic of Korea and 550 Asia amounted to one-half of the de- Korea, Rep of basis points in the Philippines. In both dine in flows to all emerging markets March Sep. March Sep March Sep countries unemployment stabilized in for the year. 1eee 1999 2000 2000 2001 2001 the first half of 2001 and decreased in Banking flows, which tradition- Source: Datastream. the second half of the year. ally account for about 50 percent of 166 REGIONAL ECONOMIC DEVELOPMENTS AND PROSPECTS market-based financing for the region, the market, and issuance dropped by Export and import volume growth in East Asia, 1998-2001 dropped by $10.7 billion, or 36 per- $200 million in the former and $625 cent, with the decline in international million in the latter country during the Three-month moving average, year-over-year percentage change 40Chn bank credit broadly based across year. countries. This development likely re- Bond market conditions were flects the confluence of several factors: variable over the course of 2001. 10 increasing availability of preferred Spreads on East Asian secondary mar- .5 long-term capital through the interna- ket debt increased moderately from an tional bond markets; decreasing de- average 295 basis points in 2 to mand for funds from a balance of pay- 350 basis points in 2001-but these -30 ments perspective, against the figures tend to mask the dynamics of March Sept March Sept March Sept March Sept background of continued (albeit di- financial market developments in the 1998 1998 1999 1999 2000 2000 2001 2001 minishing) current account surpluses year. A narrowing of the regional Note EA-5comprsesIndonesia,theRepublicofKorea, Malaysia, tePhilippines, adTaln and record high reserve levels; and spread by 50 basis points over the first Source World Bank Economic Policy and Prospects Group falling requirements for working capi- half of 2001 was reversed by October tal and trade finance, with steep reces- as concerns regarding the Argentine sion in high-tech manufacturing sec- situation and the health of the global tors across developing East Asia. economy mounted. By early January International equity placement fell 2002, however, spreads fell quickly by from a record $23 billion in 2000 to 100 basis points, with the Philippines, Gross capital market flows to East Asia, $7.2 billion. China, which is the pre- a focus of some market concern, expe- 19912001 mier source of equity issuance among riencing an easing of more than 200 Billions of dollars emerging markets ($21.9 billion basis points. placed in 2000, or 62 percent of total 100 equity during that year), found inter- 80 national market conditions unfavor- Foreign direct investment 60 able for issuing larger share volumes Net inflows of foreign direct invest- 40 (particularly for high-tech firms). And ment (FDI) to the region were $48.5 investor appetite for emerging market billion in 2001, a decline of $3.6 bil- exposure was diminished during most linor69pcetfom20.A 0 . of the year. Despite the less opportune first glance, this amount must be con- 1991 1993 1995 1997 1999 2001 environment, Korea and Thailand sidered a robust outturn in the context Source Euromoney stepped up international placements, of global and East Asian regional re- the former more substantially, from $1 cessions. Once more, such aggregates billion in 2000 to $3.7 billion in 2001. tend to mask a shifting distribution of The shift from bank-sourced fi- investment flows across countries of nancing to the bond markets was of the region. note during the year. Although bond The bulk of FD1 flows continues Capital market flows by country, 2000 and 2001 issuance in East Asia advanced by to be directed to China, at $44 billion Billions of dollars about $1 billion over 2000 levels to in 2001, a sharp $6 billion advance Korea, 1 $14.7 billion, a number of countries over the $38 billion recorded in 2000, Rep o- - I began to participate more fully in the returning flows to the record levels of China - market. China increased issuance by 1997. But, with the exception of Viet- Malaysia more than $1 billion to $2.5 billion, nam, FDI flows into other large East Malaysia by $750 million to $2.2 bil- Asian countries declined-in some lion, and Thailand to $280 million. In cases precipitously-during the year. Thailand contrast, Indonesia and the Philip- As an extreme case, for the fourth year Indonesia pines, under the weight of structural in succession, FDI flowed out of In- 20 25 30 35 economic difficulties at present, were donesia at an accelerated pace of $6 limited in their ability to expand use of billion in 2001.2 FDI dropped by $6 Source WorldBanhEconomicPolicyandProspectsGroup 167 GLOBAL DEVELOPMENT FINANCE East Asian bond spreads, 2001 billion, or two-thirds, in Korea; by verse surprises, the momentum under- Basis points above US Treasuries $660 million, or 40 percent, in lying the current upswing could pro- 1.100 - Malaysia; by 32 percent in Thailand; duce regional growth near 7 percent 1.000 All emergin markets an by 0 percent in the Philippines. by 2003, growth similar to the robust 900 800 The decline in FDI flows to these outturns of 2000. 700 countries likely has its roots in the cur- As most countries in the region 600 - rent downturn in the semiconductor have generally kept real exchange rate 500 - and computer markets, where for sev- levels below their precrisis averages, 300 eral years overseas investment in new they are expected to benefit fully from 300 East Asia 200.......................aacity had been buoyant. FD1 flows the turnaround in global trade, with bilin or tw-hrs in Koea by o 6 to East Asia, excluding China, accumu- export growth of nearly 10 percent in alated to $50 billion from 1998 to 2000, 2003 and 2004. Source J P Morgan Chase through Bloomberg a good proportion of the accumulation Yet, downside risks to this more occurring in the technology sectors. encouraging view remain large. The recovery in global high-tech sectors may not be as robust as has been the Prospects and risks case in past cycles. In particular, the Buoyed by the anticipated rebound in unprecedented bursting of the high- global high-tech markets (signs of tech bubble in financial markets may Selected East Asian bond spreads, 2001 which are now emerging in the United exert a larger drag on the availability Basis points above US Treasuries States), East Asia is expected to lead of funds for high-tech firms than is as- 700 Ph,nsrecovery in the developing world. sumed in the baseline. In broader Phi6ippines0~ During the fourth quarter of 2001, in- scope, near-term revival of world de- 500 -~. dustrIal production in high-tech East mand continues to hinge on the spend- 400 Asia, (Korea and Malaysia among ing behavior of the U.S. consumer. 300 Korea, Rep oS Malaysia middle-income countries, Sinsa)ore And abrupt changes in consumer con- rand Taiwan [China] among the newly fidence could serve to delay or protract industrialized economies) has turned recovery for some rime. China the corner decisively, in line with an Within East Asia, the economic 01 //J 4Z\F// 0- A/ upturn in Asian semiconductor sales. and political situation in Indonesia re- 4:0 However, recovery in GDP growth is mains difficult, potentially dampening likely to appear muted in annual fig- the recovery foreseen in the baseline. ures for 2002, as regional output is And the "war on terrorism" has in- expected to register 5.2 percent creased global uncertainty, with sev- growth in the year, up from 4.6 per- eral countries in the region potentially cent in 2001. But in the absence of ad- directly affected in future. Foreign direct investment, 1991-2001 Billions of dollars 200 All meveloping countries 160 120 80 40 ---- Source World Bank 168 REGIONAL ECONOMIC DEVELOPMENTS AND PROSPECTS Notes Foreign direct investment by country, 1. The most adversely affected were Korea and the newly industrialized economies Billions of dollars (NIEs)-Hong Kong (China), Singapore, and China Taiwan (China), which are not included in the Korea, aggregate for developing East Asia. Rep of 2. Negative net FDI in Indonesia is the Thailand result of repayments of intrafirm loans from foreign subsidiaries to their parents abroad. Vietnamc 2000 and2200 Philippines 2001 Malaysia Indonesia -10 0 10 20 30 40 so Source World Bank Economic Policy and Prospects Group Real effective exchange rates, 1997-2002 East Asia and Pacific forecast summary Index, January 1997= 100 120 (percent per year) China Baseliic forecast 11 Growth rates/raios 1999 2000 200 2002 2003 2004 100 I Korea, Real GDP growth 7.0 7.4 5.2 6.9 6.5 90 of Consumption per capita 6.3 6.4 5.5 5.6 5.9 80 GDP per capita 0 5.9 6.4 4.2 6.0 5.6 ou Population R 1.1 1.0 0.9 0.9 0.8 70 Gross domestic outheast Asia investment/GDPa 303 28-9 29.9 31.4 31.2 31.0 Inflationb 0.1 3.4 6.8 6.4 6.3 A . !b oe C0 0 0 o C,v Central government budget balance/GDP 1 -2.5 -2.7 -2.9 -2.9 -2.5 Export market growthc 7.3 14.2 2.6 7.8 7.6 Export volumed 1-2 7.5 22.0 2. 7.1 9.4 10.7 Source J P Morgan Chase Terms of trade/GDP' 03 0.1 -0.9 -1.2 0.2 -0.5 -0.3 Current account/GDP 0.5 4.6 3.5 .9 2.3 1.8 1.2 Memo items GDP growth: East Asia High-tech emerging Asia:' Semiconductor excluding China 37.0 ' 23 3.5 5.9 5 dollar sales and industrial production, a. Fixed investment, measured in real terms 1997-2001 b. Local currency GDP deflator, median. Percent change, three-month/three-month, c. Weighted average growth of import demand in export markets. seasonally adjusted annual rate d. Goods and nonfactor services 100 Semiconductor sales -40 e. Change in terms of trade, measured as a proportion of GDP (percent). (left axis) Source: World Bank baseline forecast, February 2002. 75 9 so 0 20 ato 0Industnial A t -25 production ty25 (right axis) V 1 June June June June June 1997 1998 1999 2000 2001 Note Through November 2001 a Republic of Korea, Malaysia, Singapore, and Taiwan (China) Source Datastream, World Bank staff estimates 169 Europe and Central Asia GNI per capita, 2000: $2,010 GN ercpia 00:$200Recent developments policy through much of 2001; the for- Output growth in Europe and mer Yugoslav Republic of Macedonia, Central Asia slowed markedly where civil war disrupted economic GDPto an estimated 2.2 percent in activity; and Slovenia, which wit- GDPgroth ate, 170-0012001 from 6.4 percent in 2000. A nessed a contraction in investment Annual percentage change Anu15 renaecag sharp deceleration in export markets that was due in part to an increase in 15 Comonwealth contributed to this s do- interest rates. exports to 10 ofIndependent statesmestic factors were more significant- Western Europe contributed to soften- 5 \Eastern EuropeA with a particularly pronounced down- ing growth, given the Central and _Vturn in Turkey and a notable Eastern European economies' high ex- -5 deceleration of growth in Poland. posure to Euro Area import demand. *urope and p10e a erformance varied across For example, Hungary, Poland, and countries, the general trend was an the Slovak Republic's export volume -1 5. ,,,. easing of growth during 2001. Most growth decelerated markedly in 2001. .~5.~b . ~European and Central Asian countries However, import compression in a Source. World Bank. experienced declining inflation-par- number of Central and Eastern Euro- ticularly in the Commonwealth of In- pean countries, including Turkey and dependent States (CIS), where the me- Poland, mitigated weaker export per- Industrial production, 1997-2001 dian rate fell from 17.3 percent in formance from a balance of payments Three-month moving average, year-over-year percentage change 25 Hungary falling interest rates, reflecting a drop The countries seeking accession to 20 Poland in import prices and international in- the European Union (EU) received a 15 reterst rates. At the same time, a num- boost in November 2001, when the 1O7 ~ / er of European and Central Asian European Commission released its a J111 ; countries (the Czech Republic, Kaza- regular annual reports on the 10 Euro- Russan achkhstan, Poland, Ukraine, and Uzbek- pean and Central Asian candidates- Fde a on Republic istan, among others) witnessed some all Central and Eastern European Turkey deterioration in fiscal balances, gener- countries.1 The commission reported Jan July Jan July Jan. July Jan July Jan. July ally because of the adoption of more that all candidates made substantial 1997 1997 1998 1998 1999 1999 2000 2000 2001 2001 expansionary fiscal policies, progress in adopting the aqcuis com- Source. Datastream The Central and Eastern Euro- munautaire (body of European Coi- pean countries (including the Baltics munity law) over the year and that and Turkey) experienced a contraction most of the countries (excluding Bul- Russian imports and partner exports, Rusin9 mprs7ndpaterexors in output of 0.8 percent in 2001, fol- garia and Romania) are expected to be 1997-2001 Year-on-year percentage change lowing a robust 5 percent expansion ready to join in the near term. Some 60 Other CIS exports in 2000. Pulling down the outturn for hurdles remain, as EU members have central this group, the severe banking and yet to formalize agreements in a num- 20 Europe exports . currency crises in Turkey in February bet of important areas, mainly per- i,i 2001 ushered in a sharp contraction of tamning to agriculture and budgetary 2'1 its economy over the year. Growth in issues. Turkey exports the Central and Eastern Europe, ex- The CIS achieved strong 5.5 per- -4 Rusa edrto cluding Turkey, declined to 2.9 percent cent growth in 2001, though this is a Russian Federation imports (bars) in 2001 from 3.8 percent in 2000. significant moderation from the ex- This moderate deceleration reflected a ceptionally robust 7.9 percent output 6,rivate creditors 0.2 3.8 10.1 8.5 8.7 4 Bonds 0.0 0.0 0.1 0.9 0.5 3 Irivate nonguaranteed 0.0 0.1 0.2 1.3 1.1 2 Bonds 0.0 0.0 0.0 0.1 0.1 qET FLOWS ON DEBT 0.5 7.1 -0.7 -4.3 -0.1 lublic and publicly guaranteed 0.5 7.0 -0.7 -4.4 0.0 )fficial creditors 0.2 4.9 1.4 -3.4 -1.7 Multilateral 0.0 0.5 1.1 -0.3 -0.2 -2 Bilateral 0.2 4.4 0.3 -3.1 -1.5 -3 Irivate creditors 0.3 2.1 -2.0 -1.0 1.7 Bonds 0.0 0.0 -0.1 0.5 4.3 'rivate nonguaranteed 0.0 0.1 0.0 0.1 -0.1 -5 Bonds 0.0 0.0 0.0 0.1 0.1 1994 1995 1996 1997 1998 1999 2000 2001 M Private a Official )eutsche mark 7.7 6.4 6.6 4.1 .Net flows on long-term debt by borrower -rench franc 18.6 9.3 11.1 8.3 apanese yen 0.0 5.7 6.3 8.1 Government and public enterprises ound sterling 4.1 1.3 1.4 0.8 .. J.S. dollars 33.0 46.4 38.2 38.8 . 0 4ultiple currency 3.1 4.7 7.5 3.4 .. 2000 01 other currencies 32.5 12.4 12.5 12.4 . 9 otal change in debt stocks . -5.7 -12.4 .. 1997 Net flows on debt 0.5 8.7 0.6 -7.5 -3.7 1 Net change in interest arrears .. .. -1.5 0.1 Interest capitalized .. .. 0.2 0.1 1995 Debt forgiveness or reduction .. .. -10.6 -0.2 1994 Cross-currency valuation .. .. 5.3 -4.0 ..------ Residual .. .. 0.2 -0.9 .. -2.0 -10 0.0 10 20 30 40 50 60 US$ billion lLL CREDITORS Private sector Interest (%) 4.6 6.4 7.4 5.8 Maturity (years) 18.6 18 I 13.5 11.9 2001 Grant element (%) 33.5 24.1 15.8 20.1 2000 )fficial creditors Interest (%) 3.7 4.7 5.5 4.2 .. 1999 Maturity (years) 23.6 24.1 21 4 19.6 .. Grant element (%) 43.6 38.8 32.4 38.4 . 'rivate creditors 1997 Interest (%) 6.3 8.6 8.8 6.7 .. 1996 Maturity (years) 9.4 10.7 7.6 8.0 Grant element (%) 14.7 5.7 3.4 10.5 19 4emo: 1994 :ommitments 1.2 11.5 15.3 7.0 .. Official creditors 0.8 6.4 6.6 2.4 -15 .0 0 01.0 Copoito ofelong-term debt, b1999. Privte cediors .4 51 88 4. ______________________________Pnvate__ 232 SOUTH ASIA (US$ billion) 1970 1980 1990 2000 2001 US$ billion External debt P SUMMARY DEBT DATA 180 -- - ---- - - - TOTAL DEBT STOCKS (EDT) 12.3 37.8 129.5 165.7 159.3 160 Long-term debt (LDOD) 11.7 32.9 112.6 157.7 151.9 Public and publicly guaranteed 11.6 32.5 1108 146.4 140.6 140 Private nonguaranteed 0.1 0.4 1.7 11.4 11.4 Use of IMF credit 0.1 2.5 4.5 1.9 1.7 120 Short-term debt 0.4 2.5 12.4 6.0 5.6 of which interest arrears on LDOD 0.0 0.0 0.0 0.1 0.1 100 Memo: IBRD 0.9 1.2 9.6 10.2 9.6 80 IDA 1.3 7.1 21.1 33.9 32.5 60 TOTAL FLOWS ON DEBT Disbursements 1.6 6.3 12.1 14.4 9.4 40 Long-term debt 1.6 4.6 10.3 14.2 9.2 20 Public and publicly guaranteed 1.5 4.3 10.0 13.7 8.3 Private nonguaranteed 0.0 0.3 0.3 0.5 0.9 IMF purchases 0.0 1.6 1.9 0.2 0.1 1994 1995 1996 1997 1998 1999 2000 2001 Memo: N official 0 Pnvate IBRD 0.1 0.2 1.6 0.8 0.7 IDA 0.1 09 1.6 1.7 2.0 Principal repayments 0.8 1.5 5.3 9.1 8.1 Long-term debt 0.5 1.2 4.1 8.4 7.8 Public and publicly guaranteed 0.5 1.1 3.8 7.7 6.9 - bln g atersr lo Private nonguaranteed 0.0 0.1 0.4 0.7 0.9 IMF repurchases 0.3 0.3 1.2 0.7 0.3 16 Memo: IBRD 0.1 0.1 0.6 1.2 0.9 14 IDA 0.0 0.0 0.2 0.6 0.6 12 Net flows on debt 0.8 5.8 8.4 4.3 0.9 of which short-term debt 0.0 1.0 1.6 -1.0 -0.4 10 Interest payments (INT) 0.3 1.2 6.2 5.6 5.7 Long-term debt 0.3 0.9 4.7 5.3 5.4 Net transfers on debt 0.5 4.6 2.3 -1.3 -4.9 6 Total debt service (TDS) 1.1 2.8 11.4 14.7 13.8 4- AGGREGATENET ESOURCEFLOWS ANDNETR -S 1=1 GO 2- NET RESOURCE FLOWS 1.4 6.4 9.1 13.3 8.9 Net flow of long-term debt (ex. IMF) 1.0 3.4 6.1 5.8 1.4 0, Foreign direct investment (net) 0.1 0.2 0.5 3.1 4.2 1994 1995 1996 1997 1998 1999 2000 2001 Portfolio equity flows 0.0 0.0 0.1 2.1 0.9 Pdvate non-debt flows 0Pnvate debt flows Official flows Grants (excluding technical coop.) 0.3 2.8 2.4 2.2 2.4 1 NET TRANSFERS 1.1 5.5 4.3 7.7 2.6 Interest on long-term debt 0.3 0.9 4.7 5.3 5.4 Profit remittances on FDI 0.0 0.0 0.1 0.3 0.9 Debt Indicators _ I Percent MAJOR ECONOMIC INDICTORS Gross national income (GNI) 82.1 234.0 399.7 608.4 651.0 Exports of goods & services (XGS) 4.3 23.0 39.9 106.9 112.6 250.0 of which workers' remittances .. 4.6 5.0 15.8 16.2 Imports of goods & services (MGS) 6.1 30.4 54.7 116.4 121.7 International reserves (RES) 15.4 8.9 47.3 55.1 200.0 Current account balance -6.3 -14.2 -6.6 -9.9 ~ 1'. - DET INItA10R .~150.0 EDT / XGS (%) 287.5 164.4 324.7 154.9 141.5 100.0 EDT / GNI (%) 15.0 16.2 32.4 27.2 24.5 TDS / XGS (%) 25.2 12.0 28.7 13.8 12.3 INT / XGS (%) 6.8 5.3 15.5 5.3 5.1 50.0 INT / GNI (%) 0.4 0.5 1.5 0.9 0.9 RES / MGS (months) .. 6.1 2.0 4.9 5.4 0.0 Short-term / EDT (%) 3.5 6.5 9.6 3.6 3.5 1994 1995 1996 1997 1998 1999 2000 2001 Concessional / EDT (%) 75.5 74.1 56.3 52.4 52.4 - EDT/GNP Multilateral / EDT (%) 18.0 24.7 29.5 38.1 38.4e 233 SOUTH ASIA (US$ billion) ___________________________ 1970 1980 1990 2000 2001 Composition of long-term debt, 1999 DEBT OUTSTANDING (LDOD) 11.7 32.9 112.6 157.7 151.9 Private Public and publicly guaranteed 11.6 32.5 110.8 146.4 140.6 29% Official creditors 10.9 30.1 86.5 112.2 108.3 Multilateral 2.2 9.3 38.2 63.0 61.2 Bilateral 8.7 20.8 48.3 49.1 47.1 Private creditors 0 7 2.4 24.3 34.2 32.2 Bilateral Bonds 0.0 0.0 2.6 11.3 11.0 31% Private nonguaranteed 0.1 0.4 1.7 11.4 11.4 Bonds 0.0 0.0 0.0 2.9 2.5 DISBURSEMENTS 1.6 4.6 10.3 14.2 9.2 Public and publicly guaranteed 1.5 4.3 10.0 13.7 8.3 Official creditors 1.4 3.3 6.9 7.1 6.7 Multilateral 0.2 1.4 4.4 4.3 4 1 Bilateral 1.2 1.9 2.5 2.8 2.6 Multilateral Private creditors 0.1 1.0 3.1 6.5 1.6 40% Bonds 0.0 0.0 0.4 5.5 0.0 Private nonguaranteed 0.0 0.3 0.3 0.5 0.9 Bonds 0.0 0.0 0.0 0.0 0.1 PRINCIPAL REPAYMENTS 0.5 1.2 4.1 8.4 7.8 Composition of net flows Public and publicly guaranteed 0.5 1.1 3.8 7.7 6.9 US$ billion on long-term debt Official creditors 0.4 0.9 2.4 5.4 4.5 5-- Multilateral 0.1 0.1 1.0 2.9 2.4 Bilateral 0.3 0.8 1.4 2.5 2.1 4 Private creditors 0.1 0.2 1.4 2.3 2.4 Bonds 0.0 0.0 0.3 0.6 0.3 3 Private nonguaranteed 0.0 0.1 0.4 0.7 0.9 2 Bonds 0.0 0.0 0.0 0.0 0.4 NET FLOWS ON DEBT 1.0 3.4 6.1 5.8 1.4 4 Public and publicly guaranteed 1.0 3.2 6.2 6.0 1.4 Official creditors 1.0 2.3 4.5 1.8 2.2 Multilateral 0.1 1.3 3.4 1.5 1.8 Bilateral 0.9 1.1 1.2 0.3 0.4 Private creditors 0.0 0.9 1.7 4.2 -0.8 .2 Bonds 0.0 0.0 0.1 4.9 -0.3 Private nonguaranteed 0.0 0.2 -0.1 -0.2 0.0 Bonds 0.0 0.0 0.0 0.0 -0.3 1994 1995 1996 1997 1998 1999 2000 2001 U Private U Official Deutsche mark 9.3 8.4 5.9 3.7 .. Net flows on long-term debt by borrower French franc 1.4 2.3 1.6 1.3 Japanese yen 5.3 8.9 11.9 14 9 . Government and public enterprises Pound sterling 21.7 17.8 4.8 2.5 .. 1 - - U.S. dollars 41.1 41.3 51.6 55.4 2001 Multiple currency 9.2 7.4 13.7 15.0 .. 2000 All other currencies 11.5 10.0 4.5 2.0 1999 BEB i~§-FL0W Rt~N 1L ~N1998 Total change in debt stocks .. .. 13.1 -1.6 .. 1997 Net flows on debt 0.8 5.8 8.4 4.3 0.9 1996 Net change in interest arrears .. . 0.0 0.0 Interest capitalized .. .. 0.0 02 .. 1995 Debt forgiveness or reduction .. .. 0.0 0.0 1994 Cross-currency valuation .. .. 1.6 -5.7 .. Residual .. .. 3.1 -0.5 .. 0.0 2.0 4.0 60 80 100 US$ billon ALL CREDITORS Private sector Interest (%) 2.5 4.7 4.6 6.6 Maturity (years) 32.6 32.6 24.7 13.6 2001 Grant element (%) 45.1 27.1 34.2 20.6 .. Official creditors Interest (%) 2.2 2.2 3.6 5.2 .. 1999 Maturity (years) 34.9 39.2 29.0 23.1 . 9 Grant element (%) 48.7 34.5 40.8 35.0 . Private creditors 1997 Interest (%) 5.9 12.8 6.7 7.7 1996 Maturity (years) 11.7 10.9 15.3 5.4 . 199 Grant element (%) 12.5 2.6 19.7 8.2 1995 Memo: 1994 Commitments 2.1 8.4 13.5 10.8 10 Official creditors 1.9 6.5 9.3 5.0 S-15 -10 05 00 05 10 1l5 20$25 Private creditors 0.2 2.0 4.2 5 8 ..__I 234 SUB-SAHARAN AFRICA (US$ billion) 1970 1980 1990 2000 2001 US$ billion External debt SUMMARYDEBT DATA, 250 ------ --------- -- -- - ---- - TOTAL DEBT STOCKS (EDT) 6.9 60.9 176.9 215.8 208.9 Long-term debt (LDOD) 6.1 46.7 149.4 175.8 168.7 Public and publicly guaranteed 5.8 42.1 144.1 163.8 156 5 2o Private nonguaranteed 0.3 4.6 5.3 12.0 12.3 Use ofIMF credit 0.1 3.0 6.6 6.7 6.3 Short-term debt 0.8 11.2 20.9 33.3 33.8 150 of which interest arrears on LDOD 0.0 1.7 9.3 14.3 14.2 Memo: IBRD 0.6 2.5 9.2 3.5 2.9 100- IDA 0.2 2.6 15.8 33.6 34.5 TOTAL FLOWS ON DEBT Disbursements 1.3 11.8 10.4 8.7 10.9 50 - Long-term debt 1.2 10.6 9.7 8.2 10.4 Public and publicly guaranteed 1.1 9.4 9.1 7.3 8.4 Private nonguaranteed 0 1 1.2 0.6 0.8 1.9 0 IMF purchases 0.0 1.2 0.7 0.5 0.6 1994 1995 1996 1997 1998 1999 2000 2001 Memo: a Official 0 Private IBRD 0.1 0.4 0.8 0.1 0.1 IDA 0.1 0.4 2.0 2.2 2.4 Principal repayments 0.5 3.2 5.6 8.1 9.8 Long-term debt 0.4 2.8 4.6 7.5 9.0 Public and publicly guaranteed 0.3 2.2 4.2 6.2 7.6 lo Private nonguaranteed 0.1 0.6 0.5 1.4 1.4 IMF repurchases 0.1 0.4 1.0 0.5 0.8 Memo: 25 IBRD 0.0 0.1 0.7 0.6 0.5 IDA 0.0 0.0 0.1 0.3 0.3 Net flows on debt 0.9 10.1 7.1 -0.5 1.7 of which short-term debt 0.1 1.5 2.3 -1.1 05 Interest payments (INT) 0.2 3.5 5.3 4.3 4.7 Long-term debt 0.2 2.4 4.4 3.1 3.8 Net transfers on debt 0.7 6.6 1.7 -4.8 -3.1 Total debt service (TDS) 0.6 6.7 10.9 12.3 14.5 AGGREGATE NET RESOURCE4 LOWS-AND NETv TRANSFERS (LONT ERM, NET RESOURCE FLOWS 1.7 11.4 17.9 18.5 25.5 Net flow of long-term debt (ex. IMF) 0.9 7.8 5.1 0.6 1.3 0-. Foreign direct investment (net) 0.4 0.0 0.8 6.7 13.6 1994 1995 1996 1997 1998 1999 2000 2001 Portfolio equity flows 0.0 0.0 0.0 0.9 0.7 E Private non-debt flows 0 Pnvate debt flows 0 Official flows Grants (excluding technical coop.) 0.4 3.6 12.0 10.3 9.9 NET TRANSFERS 0.7 6.1 11.8 10.6 15.8 Interest on long-term debt 0.2 2.4 4.4 3.1 3.8 Profit remittances on FDI 0.7 2.9 1.7 4.8 5.9 Debt indicators Percent MAJOR ECONOMIC INDICATORS Gross national income (GNI) 60.6 258.8 280.9 302.6 294.4 Exports of goods & services (XGS) .. 92.7 84.9 121.3 116.8 250.0 of which workers' remittances .. 0.8 0.8 0.7 1.1 Imports of goods & services (MGS) .. 93.5 91.7 127.8 128.8 International reserves (RES) 3.1 22.9 15.4 26.3 26.6 2000 Current account balance . 0.1 -1.8 1.4 -0.6 1500- DFj3T, IDCATOkS-' EDT / XGS (%) .. 65.7 208.3 177.9 178.8 1o00 EDT / GNI(%) 11.4 23.5 63.0 71.3 70.9 TDS /XGS (%) .. 7.2 12.8 10.2 12.4 INT /XGS(%) .. 3.8 63 3.5 4.1 500 INT /GNI (%) 0.3 1.4 1.9 1.4 1.6 RES /MGS (months) .. 3.0 2.0 2.5 2.5 Short-term / EDT (%) 10.9 18.4 11.8 15.4 16.2 1994 1995 1996 1997 1998 1999 2000 2001 Concessional / EDT (%) 46.3 27.0 33.1 38.7 38.8 ETXS- D/N Multilateral U EDT (%) 12.6 12.5 21.6 25.3 25.5f 235 SUB-SAHARAN AFRICA (US$ billion) 1970 1980 1990 2000 20 1970 1980 1990 2000 2001Composition of long-term debt, 1999 L06-.EMDEBI Private DEBT OUTSTANDING (LDOD) 6.1 46.7 149.4 175.8 168.7 22% Public and publicly guaranteed 5.8 42.1 144.1 163.8 156.5 Official creditors 4.2 25.7 108.7 138.0 131.5 Multilateral 0.9 7.6 38.2 54.5 53.3 Bilateral 3.3 18.1 70.6 83.5 78.2 Pnvate creditors 1.6 16.3 35.4 25.8 24.9 Bonds 0.4 0.6 0.3 9.1 9.1 Private nonguaranteed 0.3 4.6 5.3 12.0 12.3 Bonds 0.0 0.0 0.0 2.0 3.2 DISBURSEMENTS 1.2 10.6 9.7 8.2 10.4 Multilateral Bilateral Public and publicly guaranteed 1.1 9.4 9.1 7.3 8.4 32% 46% Official creditors 0.7 4.2 7.2 3.8 4.6 Multilateral 0.2 1.7 4.5 3.1 3.7 Bilateral 0.5 2.6 2.6 0.7 0.9 Private creditors 0.5 5.1 1.9 3.5 3.9 Bonds 0.0 0.1 0.0 0.8 0.9 Private nonguaranteed 0.1 1.2 0.6 0.8 1.9 Bonds 0.0 0.0 0.0 0.7 13 PRINCIPAL REPAYMENTS 0.4 2.8 4.6 7.5 9.0 Composition of net flows Public and publicly guaranteed 0.3 2.2 4.2 6.2 7.6 US$ billion on long-term debt Official creditors 0.2 0.7 2.6 2.7 3.4 Multilateral 0.0 0.2 1 3 1.7 1.9 Bilateral 0.1 0.5 1.2 1.0 1.5 Private creditors 0.2 1.5 1.6 3.5 4.2 Bonds 0.0 0.0 0.0 0.5 1.0 Private nonguaranteed 0.1 0.6 0.5 1.4 1.4 Bonds 0.0 0.0 0.0 0.0 0.0 NET FLOWS ON DEBT 0.9 7.8 5.1 0.6 1.3 Public and publicly guaranteed 0.8 7.2 4.9 1.1 0.8 Official creditors 0.5 3.6 4.6 1.1 Multilateral 0.1 1.5 3.2 1.4 1 8 Bilateral 0.4 2.1 1.4 -0.3 -0.6 Private creditors 0.3 3.6 0.3 0.0 -0.4 Bonds 0.0 0.0 0.0 0.3 -0.1 Private nonguaranteed 0.0 0.6 0.2 -0.5 0.6 -2 Bonds 0.0 0.0 0.0 0.7 1.2 1994 1995 1996 1997 1998 1999 2000 2001 N Private m Off icial Deutsche mark 6.6 7.0 6.4 2.6 Net flows on long-term debt by borrower French franc 14.4 13.8 14.1 7.6 Japanese yen 0.1 5.4 4.0 4.2 Government and public enterrises Pound sterling 22.5 5.7 5.4 2.0 2 U.S. dollars 21.4 35.4 36.5 61.6 lultiple currency 11.2 8.8 10.4 7.0 2002000 All other currencies 23.4 22.2 20.5 12.5 199 Total change in debt stocks .. .. 19.8 -0.5 . . 19 Net flows on debt 0.9 10.1 7.1 -0.5 1.7 1996 Net change in interest arrears .. .. 2.2 -5.3 Interest capitalized .. .. 1.7 9.0 . . 1995 Debt forgiveness or reduction .. .. -2.9 -0.7 1994 Cross-currency valuation .. .. 7.7 -6.6 Residual .. .. 3.9 3.5 00 10 2.0 30 40 50 60 70 US$ billion ALL CREDITORS Private sector Interest (%) 3.7 7.0 4.3 3.2 Maturity (years) 23.9 17.3 32.6 24.8 2001 [- Grant element(%) 47.5 21.8 43.5 48.7 Official creditors Interest (%) 2.0 4.1 3.5 0.9 1999 Maturity (years) 32.0 25.1 37.1 39.1 Grant element (%) 67.4 42.8 51.2 77.5 Private creditors 1997 Interest (%) 6.6 10.0 8.1 5.9 1996 Maturity (years) 10.2 9.4 11.7 8.1 Grant element (%) 13.3 0.2 7.7 15.0 Memo: 1994 2ommitments 1.9 13.3 11.5 6.0 21 51 Official creditors 1.2 6.7 9.5 3.2 -20 - 1000 5 0 Private creditors 0.7 6.6 2.0 2.8P 236 SEVERELY INDEBTED LOW-INCOME COUNTRIES (US$ billion) 1970 1980 1990 1999 2000 RUS$ bilon External debt SUMMRY DET AAA 400 ------------------- -------- TOTAL DEBT STOCKS (EDT) 13.1 84.3 252.9 353.1 342.6 350 Long-term debt (LDOD) 11.7 67.8 213.2 286.6 279.7 Public and publicly guaranteed 11.0 60.5 198.8 233.5 235.1 Private nonguaranteed 0.6 7.2 14.3 53.1 44.7 Use of IMF credit 0.3 3.0 5.7 17.4 17.6 Short-term debt 1.2 13.5 34.0 49.0 45.2 250 of which interest arrears on LDOD 0.0 1.7 10.6 23.9 19.5 Memo: 200 IBRD 0.8 3.4 18.8 18.1 17.6 IDA 0.4 3.4 13.6 27.4 27.7 150 TOTAL FLOWS ON DEBT 1oo Disbursements 2.3 14.7 19.5 16.5 12.7 Long-term debt 2.3 13.5 19.3 14.1 110 50- Public and publicly guaranteed 2.0 11.7 13.9 11.5 8.2 Private nonguaranteed 0.3 1.8 5.4 2.6 2.9 , IMF purchases 0.1 1.2 0.3 2.4 1.7 1993 1994 1995 1996 1997 1998 1999 2000 Memo: N Official 1l Private IBRD 0.1 0.7 2.1 2.0 1.2 1 IDA 0.1 0.4 1.3 1.6 1.7 Principal repayments 0.7 4.5 10.9 17.5 16.8 Long-term debt 0.6 4.0 10.1 17.0 16.3 Public and publicly guaranteed 0.5 2.8 8.5 9.6 7.9 ne Private nonguaranteed 0.1 1.2 1.6 7.4 8.3 IMF repurchases 0.1 0.5 0.8 0.5 0.6 T Memo: IBRD 0.0 0.1 1.1 1.5 1.4 25 IDA 0.0 0.0 0.1 0.2 0.3 Net flows on debt 1.7 12.9 13.8 -3.0 -3.5 of which short-term debt 0.1 2.7 5.2 -2.0 0.6 15 Interest payments (INT) 0.3 4.8 9.1 9.9 10.9 Long-term debt 0.3 3.5 7.6 7.9 8.7 10 Net transfers on debt 1.4 8.1 4.7 -12.8 -14.4 Total debt service (TDS) 1.0 9.3 20.0 27.3 27.7 AFGREGATE NET RES URCE FLOWSAND NET;S T -SERS 1)NG+1Rf .5 NET RESOURCE FLOWS 2.3 12.8 19.7 9.4 2.9 Net flow of long-term debt (ex. IMF) 1.7 9.5 9.2 -2.9 -5.2 10 Foreign direct investment (net) 0.1 0.0 2.0 3.4 0.4 1993 1994 1995 1996 1997 1998 1999 2000 Portfolio equity flows 0.0 0.0 0.3 1.3 0.4 0 Private non-debt flows 0 Private debt flows E Official flows Grants (excluding technical coop.) 0.5 3.4 8.2 7.7 7.3 NET TRANSFERS 1.2 3.9 9.0 -3.5 -10.4 Interest on long-term debt 0.3 3.5 7.6 7.9 8.7 Profit remittances on FDI 0.8 5.5 3.0 4.9 4.6 Debt indicators Percet Gross national income (GNI) 59.0 261.8 287.2 319.4 341.5 350.0 Exports of goods & services (XGS) 76.4 114.2 144.0 of which workers' remittances 2.2 2.4 3.9 2.6 300.0 Imports of goods & services (MGS) 87.2 125.0 142.8 International reserves (RES) 2.0 22.5 16.9 35.3 38.3 2500 Current account balance -2.7 -7.4 -4.2 8.5 -, --~- --.200.0 DEBT INDICATORS EDT / XGS (%) .. .. 331.0 309.2 237.9 EDT / GNI(%) 22.2 32.2 88.1 110.5 100.3 100.0 TDS / XGS (%) .. .. 26.2 23.9 19.3 INT / XGS(%) .. .. 11.9 8.6 7.6 500 INT / GNI (%) 0.5 1.8 3.2 3.1 3.2 RES / MGS (months) .. .. 2.3 3.4 3.2 0.0 Short-term / EDT (%) 119 16.0 13.5 13.9 13.2 1993 1994 1995 1996 1997 1998 1999 2000 Concessional / EDT (%) 57.1 36.1 34.3 33.8 33.7 ED/G- DTN Multilateral/UEDT (%) 1e.1 11.5 19.8 21.4 21.5s 237 SEVERELY INDEBTED LOW-INCOME COUNTRIES (US$ billion) 1970 1980 1990 1999 2000 Le= -FR . DE.. B Composition of long-term debt, 1999 DEBT OUTSTANDING (LDOD) 11.7 67.8 213.2 286.6 279.7 Prvate Public and publicly guaranteed 11.0 60.5 198.8 233.5 235.1 28 Official creditors 9.2 41.0 150.4 195.1 202.5 Multilateral 1.3 9.7 50.0 75.6 73.8 Bilateral 7.9 31.3 100.4 119.5 128.7 Private creditors 1.8 19.5 48.4 38.3 32.6 Bonds 0.1 0.2 0.7 6.0 6.0 Private nonguaranteed 0.6 7.2 14.3 53.1 44.7 Bonds 0.0 0.0 0.1 8.5 6.4 DISBURSEMENTS 2.3 13.5 19.3 14.1 11.0 Bilateral Public and publicly guaranteed 2.0 11.7 13.9 11.5 8.2 Multilateral 46% Official creditors 1.4 5.7 11.3 10.2 7.3 26% Multilateral 0.2 1.9 6.0 6.1 4.9 Bilateral 1.2 3.8 5.3 4.1 2.4 Private creditors 0.6 6.0 2.6 1.3 0.8 Bonds 0.0 0.0 0.0 0.0 0.0 Private nonguaranteed 0.3 1.8 5.4 2.6 2.9 Bonds 0.0 0.0 0.1 0.0 0.4 PRINCIPAL REPAYMENTS 0.6 4.0 10.1 17.0 16.3 Composition of net flows Public and publicly guaranteed 0.5 2.8 8.5 9.6 7.9 US$ billion on long-term debt Official creditors 0.3 1.1 4.2 5.2 4.3 10 Multilateral 0.0 0.3 1.9 29 3.1 Bilateral 0.2 0.9 2.3 2.3 1.2 8 Private creditors 0.3 1.7 4.3 4.4 3 6 Bonds 0.0 0.0 0.1 0.3 00 Private nonguaranteed 0.1 1.2 1.6 7.4 8.3 Bonds 0.0 0.0 0.0 1.2 24 NET FLOWS ON DEBT 1.7 9.5 9.2 -2.9 52 Public and publicly guaranteed 1.5 8.9 5.4 1.9 0. -2- Dfficial creditors 1.1 4.6 7.1 5.0 30 Multilateral 0.2 1.7 4.1 3.1 Bilateral 1.0 2.9 3.1 1.9 1 2 Private creditors 0.4 4.3 -1.8 -3.1 2 8 Bonds 0.0 0.0 -0.1 -0.3 00 Private nonguaranteed 0.2 0.6 3.8 -4.8 -5.5 Bonds 0.0 0.0 0.1 -1.2 -2.1 1993 1994 1995 1996 1997 1998 1999 2000 2 7 Private 0 Official D32. Deutsche mark 6.8 7.3 6.4 4.2 2.9 Net flows on long-term debt by borrower French franc 8.6 8.4 8.8 6.1 4.8 1apanese yen 4.3 10.7 13.0 17.7 14.5 Government and public enterprises Pound sterling 9.7 2.9 3.6 2.7 1.4 .S. dollars 38.9 40.8 33.3 45.7 55 2000 1 - "4ultiple currency 9.8 8.4 14.9 9.5 8.8 1999 kl1l other currencies 21.5 18.6 15.4 9.7 8.8 1 9981 DEBJT T K-L WRE§NI TIN19 rotal change in debt stocks 29.8 -7.2 -10.4 1996 Net f1ows on debt 1.7 12.9 13.8 -3.0 C3o5 of1ne Net change in interest arrears . .. 2.5 2.8 -4.4 19 Interest capitalized . .. 1.5 1.1 9.1 1 994 Debt forgiveness or reduction . .. -1.4 -4.7 -0.5 1993 Cross-currency valuation . .. 11.0 -2.0 -11.3 Residual . .. 2.3 -1.5 0.2 0.0 2.0 4.0 60 80 100 12.0 US billion 4,LL CREDITORS Pnivate sector Interest (%) 3.3 7.2 5.3 3.6 3.2 Maturity (years) 27.4 18.6 22.4 19.8 25.1 2000 Grant element (%) 52.3 22.0 34.5 42.1 50.9 - Dfficial creditors 199 1992 Inteest %) 22 4. 4.7 3.2 0.5 199 Matunty (years) 33.5 26.7 24.5 21.4 30.2 1- Grant element (%) 66.0 44.3 39.7 45.9 61.0 Private creditors 1996 Interest ()6.5 10.5 8.0 7.1 5.9 19 Maturity (years) 10.1 9.8 12.8 6.1 6.8 Grant element ()13.9 -2.5 10.5 9.8 14.7 19 vlemo: 1993 ommitments 3.3 16.8 18.0 9. 1 5.5 -lo Official creditors -2.3 -80 -6.0 1490 -20 00 20 4.0 60 80 100 c.6 8.4 84.8 6.1 4.8US$ billion Privatecreditors 0.9 8.0 3.2 1.0 1.2 238 SEVERELY INDEBTED MIDDLE-INCOME COUNTRIES (US$ billion) 1970 1980 1990 1999 2000 US$ billion External debt SUMARY DEBT DATA 600 - ---- ---- TOTAL DEBT STOCKS (EDT) 15.7 140.5 298.7 512.7 502.3 Long-term debt (LDOD) 13.6 109.6 234.8 412.8 411.7 500 Public and publicly guaranteed 6.8 84.6 225.8 267.4 267.2 Private nonguaranteed 6.8 24.9 9.0 145.4 144.5 Use of IMF credit 0.0 0.6 6.3 14.8 8.2 400 Short-term debt 2.1 30.4 57.7 85.1 82.5 of which interest arrears on LDOD 0.0 0.0 22.4 3.1 3.0 Memo: 300- | IBRD 0.6 3.2 14.2 19.4 20.5 IDA 0.0 0.2 0.2 0.3 0.3 200- TOTAL FLOWS ON DEBT Disbursements 3.3 21.6 9.8 75.6 74.0 100 Long-term debt 3.2 21.4 9.2 69.4 71.7 Public and publicly guaranteed 1.7 16.0 8.3 32.9 34.1 Private nonguaranteed 1 6 5.4 0.9 36.5 37.6 0 IMF purchases 0.0 0.2 0.6 6.2 2.3 1993 1994 1995 1996 1997 1998 1999 2000 Memo: * Official 13 Private IBRD 0.1 0.7 1.4 3.7 3.1 MOfficial _DPrivate IDA 0.0 0.0 0.1 0.0 0.0 Principal repayments 1.7 11.1 14.4 73.7 68.1 Long-term debt 1.6 10.9 12.7 70.6 59.9 Public and publicly guaranteed 0.8 6.9 11.6 27.1 29.0 US$ billion 100 -- Private nonguaranteed 0.9 4.0 1.1 434 30.9 IMF repurchases 0.1 0.2 1.8 3.1 8.2 90 Memo: IBRD 0.0 0.2 1.6 1.7 1.7 80 IDA 0.0 0.0 0.0 0.0 0.0 Net flows on debt 1.5 20.4 -0.3 -1.1 4.2 60 )f which short-term debt 0.0 10.0 4.3 -3.0 -1.7 [nterest payments (INT) 0.7 12.7 8.0 29.6 30.6 50 Long-term debt 0.7 9.4 6.2 24.9 26.3 40 Net transfers on debt 0.8 7.8 -8.3 -30.7 -26.4 Fotal debt service (TDS) 2.5 23.8 22.5 103.3 98.6 30 AGGREGATE NET RS RCE FLOWS AND'NET T20 SFElS (LON( 10- 4ET RESOURCE FLOWS 2.2 16.6 1.8 58.5 65.7 14et flow of long-term debt (ex. INF) 1.6 10.5 -3.5 -1.2 11.8 0-. 7oreign direct investment (net) 05 2.8 3.2 55.8 46.7 1993 1994 1995 1996 1997 1998 1999 2000 'ortfolio equity flows 0.0 0.0 0.0 2.7 5.7 * Private non-debt flows 0 Private debt flows 0 Official flows 3rants (excluding technical coop.) 0.1 3.4 2.1 1.3 1.5 qET TRANSFERS 0.4 5.0 -7.1 26.1 30.2 nterest on long-term debt 0.7 9.4 6.2 24.9 26.3 'rofit remittances on FDI 1.1 2.2 2.8 7.6 9.3 Debt indicators Percent OR ECONOMIC INDICATORS 4500 3ross national income (GNI) 85.6 393.5 675.1 923.2 985.9 400.0 xports of goods & services (XGS) .. 50.5 73.0 124.5 141.8 >f which workers' remittances .. 0.7 1.2 4.6 5.0 350.0 mports of goods & services (MGS) .. 72.2 76.3 167.5 180.2 nternational reserves (RES) 3.1 22.9 19.8 75.9 72.0 1 urrent account balance . -17.6 0.5 -37.4 -33.7 250.0 - DEBTINDICATORS 200.0 -DT/ XGS(%) .. 278.4 409.3 412.0 354.3 150.0 DT / GN(%) 18.3 35.7 44.3 555 51.0 DS / XGS (%) . 47.2 30.8 83.0 69.6 100.0 NT / XGS(%) .. 25.1 11.0 23.8 21.6 NT / GNI(%) 0.9 3.2 1.2 3.2 3.1 ZES / MGS (months) .. 3.8 3.1 5.4 4.8 0.0 ;hort-term / EDT (%) 13.1 21.6 19.3 16.6 16.4 1993 1994 1995 1996 1997 1998 1999 2000 oncessional / EDT (%) 12.1 14.8 16.2 6.5 6.2 vlultilateral / EDT (%) 5.7 4.1 7.8 9.0 9.4 O/GEDGN 239 SEVERELY INDEBTED MIDDLE-INCOME COUNTRIES $ billion) ___________________________ 1970 1980 1990 1999 20 bilin 190 190 190 99 20 Composition of long-term debt, 1999 BT OUTSTANDING (LDOD) 13.6 109.6 234.8 412.8 411.7 lic and publicly guaranteed 6.8 84.6 225.8 267.4 267.2 icial creditors 3.2 34.0 104.3 122.8 122.9 AIultilateral 0.9 5.7 23.4 46.2 47.0 18% lilateral 2.3 28.3 80.9 76.6 76.0 iate creditors 3.5 50.6 121.5 1446 144.3 londs 0.4 4.1 14.1 125.6 125.5 vate nonguaranteed 6.8 24.9 9.0 145.4 1445 londs 0.0 0.0 0.0 29.1 28.1 iBURSEMENTS 3.2 21.4 9.2 69.4 71.7 Mie )lic and publicly guaranteed 1.7 16.0 8.3 32.9 34.1 11% icial creditors 0.5 4.6 4.5 10.7 9.5 Iultilateral 0.2 1.2 2.4 9.1 7.8 lilateral 0.3 3.4 2.1 1.6 1.6 'ate creditors 1.1 11.4 3.8 22.2 247 londs 0.1 0.5 0.2 21.1 21.1 vate nonguaranteed 1.6 5.4 0.9 36.5 37.6 londs 0.0 0.0 0.0 2.7 4.3 [NCIPAL REPAYMENTS 1.6 10.9 12.7 70.6 59.9 Composition of net flows )lic and publicly guaranteed 0.8 6.9 11.6 27.1 29.0 US$ billion on long-term debt icial creditors 0.2 1.3 4.9 9.0 10.8 40 4ultilateral 0.1 0.3 2.3 4.1 6.2 lilateral 0.2 1.0 2.6 4.9 4.6 35 ,ate creditors 0.5 5.6 6.7 18.1 182 30 onds 0.0 0.2 0.9 9.7 15.4 vate nonguaranteed 0.9 4.0 1.1 43.4 30.9 25 onds 0.0 0.0 0.0 3.5 5 20 1 FLOWS ON DEBT 1.6 10.5 -3.5 -1.2 11 Plic and publicly guaranteed 0.9 9.0 -3.3 5.7 5.1 cial creditors 0.3 3.2 -0.3 1.7 -13 10 lultilateral 0.1 0.9 0.1 5.0 ilateral 0.2 2.4 -0.4 -3.3 3.0 5 ,ate creditors 0.6 5.8 -3.0 4.1 65 0 onds 0.1 0.3 -0.7 11.4 57 vate nonguaranteed 0.7 1.4 -0.2 -6.9 6.7 .5 onds 0.0 0.0 0.0 -0.8 -0.8 1993 1994 1995 1996 1997 1998 1999 2000 E Private N Official tsche mark 10.1 6.0 6.6 4.7 4.6 Net flows on long-term debt by borrower ich franc 1.6 2.4 4.4 2.0 1.7 mese yen 0.1 5.8 5.3 6.5 6.5 Government and public enterprises nd sterling 6.2 1.9 1.5 1.0 0.9 20 dollars 60.1 51.2 47.9 61.6 60.8 tiple currency 12.8 6.8 11.9 4.9 4.3 1999 other currencies 7.4 4.3 3.5 7.9 9.8 1998 DEB ~ 7 ST0 -OWR1 1 997 il change in debt stocks .. .. 8.8 4.4 -10.4 1 et flows on debt 1.5 20.4 -0.3 -1.1 4.2 et change in interest arrears .. .. 8.8 -1.1 -0.1 kterest capitalized .. .. 0.9 0.1 0.5 199 iebt forgiveness or reduction .. .. -6.5 -1.6 -3.4 1993 ross-currency valuation .. .. 6.0 -4.2 -6. esidual -0.1 123 -5.2 00 5.0 10.0 150 20.0 250 US$ billion L CREDITORS Private sector iterest (%) 6.9 11.2 6.7 8.7 9.5 laturity (years) 13.9 11.8 15.9 9.2 15.1 rant element 16.8 -4.1 21.0 5.0 0.2 icial creditors 1999 iterest (%) 6.0 7.4 5.7 5.9 6.3 1998 laturity (years) 22.6 17.0 19.8 12.2 16.6 19971 ,rant element M% 26.0 17.8 29.5 20.6 21.7 'ate creditors 1996 iterest (%) 7.4 12.9 8.7 10.2 10.1 1995 no: 1993 imitments 2.2 18.1 5.8 35.0 31.6------- ftca ceior . .4 39 127 50 -10.0 -5.0 0.0 50 100 150 200 250 300 rivate creditors 1.5 12.7 1.9 22.4 26.6USbllo 240 MODERATELY INDEBTED LOW-INCOME COUNTRIES (US$ billion) 1970 1980 1990 1999 2000 US$ billion External debt SUEMMAR) EL .ATA 60- TOTAL DEBT STOCKS (EDT) 2.0 12.1 39.2 54.3 52.6 Long-term debt (LDOD) 1.8 9.8 32.6 46.5 45.6 5o Public and publicly guaranteed 1.7 9.4 31.4 43.5 42.6 Private nonguaranteed 0.1 0.5 1.2 3.0 3.0 Use of IMF credit 0.1 0.8 2.2 3.0 2.6 40 Short-term debt 0.1 1.5 4.5 4.8 4.4 of which interest arrears on LDOD 0.0 0.0 0.8 0.5 0.8 Memo: 30 IBRD 0.1 0.5 1.5 1.0 0.9 IDA 0.1 0.9 6.7 14.5 14.5 20- TOTAL FLOWS ON DEBT Disbursements 0.2 2.7 3.4 4.3 3.0 10 Long-term debt 0.2 2.3 3.0 3.9 2.8 Public and publicly guaranteed 0.2 2.2 2.7 3.5 2.6 Private nonguaranteed 0.0 0.1 0.2 0.4 0.2 0 IMF purchases 0.0 0.4 0.4 0.4 0.2 1993 1994 1995 1996 1997 1998 1999 2000 Memo: N Official N Private IBRD 0.0 0.1 0.0 0.1 0.1 IDA 0.0 0.2 1.0 1.0 1.0 Principal repayments 0.1 0.7 1.6 2.5 2.3 Long-term debt 0.1 0.6 1.2 2.0 1.8 Aggregate net resource flows Public and publicly guaranteed 0.1 0.5 1.1 1.8 1.6 v bon Private nonguaranteed 0.0 0.1 0.1 0.2 0.3 IMF repurchases 0.0 0.1 0.4 0.5 0.5 8 Memo: IBRD 0.0 0.0 0.2 0.1 0.1 7 IDA 0.0 0.0 0.0 0.1 0.1 6 Net flows on debt 0.1 2.2 2.7 2.8 0.0 of which short-term debt 0.0 0.2 0.9 1.0 -0.7 5-1 Interest payments (INT) 0.0 0.5 1.1 1.2 1.1 Long-term debt 0.0 0.3 0.7 0.9 0.9 4 Net transfers on debt 0.1 1.7 1.6 1.6 -1.1 , Total debt service (TDS) 0.1 1.2 2.6 3.7 3.5 2+ NET RESOURCE FLOWS 0.3 3.4 6.6 8.0 6.8 Net flow of long-term debt (ex. IMF) 0.1 1.7 1.8 1.9 0.9 0 Foreign direct investment (net) 0.1 0.2 0.0 1.1 1.1 1993 1994 1995 1996 1997 1998 1999 2000 Portfolio equity flows 0.0 0.0 0.0 0.0 0.0 N Pnvate non-debt flows E Private debt flows * Official flows Grants (excluding technical coop.) 0.1 1.5 4.8 4.9 4.8 NET TRANSFERS 0.2 2.8 5.6 6.7 5.6 Interest on long-term debt 0.0 0.3 0.7 0.9 0.9 Profit remittances on FDI 0.1 0.3 0.3 0.4 0.4 Debt indicators Percent MA (OR E4(ON(OMI INN ATObRS 3000--------- - - - - - Gross national income (GNI) .. 47.3 79.1 69.0 66.8 Exports of goods & services (XGS) .. .. 18.9 25.5 28.1 250.0 of which workers' remttances .. 0.4 1.9 1.6 1.6 Imports of goods & services (MGS) .. .. 22.5 30.7 31.7 International reserves (RES) .. .. 2.5 8.2 8.4 2000 Current account balance .. -2.3 -2.4 -1.9 0.2 150.0 EDT / XGS (%) .. .. 208.1 213.2 187.0 1oo.o EDT GNI (%) 25.7 49.6 78.7 78.8 TDS / XGS (%) . .. 14.0 14.4 12.3 INT / XGS (%) .. .. 5.7 4.6 4.1 50.0 INT / GNI (%) .. 1.1 1.4 1.7 1.7 RES / MGS (months) .. .. 1.3 3.2 3.2 0.0 Short-term / EDT (%) 4.3 12.7 11.4 8.9 8.4 1993 1994 1995 1996 1997 1998 1999 2000 Concessional / EDT (%) 50.6 40.2 51.8 62.4 63.7 ETXSD/G Multilateral / EDT (%) 8.8 17.7 30.1 40.9 41.5 241 MODERATELY INDEBTED LOW-INCOME COUNTRIES S$ billion) 1970 1980 1990 1999 2000 Composition of long-term debt, 1999 Private -BT OUTSTANDING (LDOD) 1.8 9.8 32.6 46.5 45.6 14% iblic and publicly guaranteed 1.7 9.4 31.4 43.5 42.6 ficial creditors 1.2 6.4 26.2 40.1 39.2 Multilateral 0.2 2.1 11.8 22.2 21.8 Bilateral 1.1 4.2 14.4 17.9 17.3 ivate creditors 0.5 3.0 5.2 3.4 3.5 Bonds 0.2 0.6 0.3 0.1 0.1 ivate nonguaranteed 0.1 0.5 1.2 3.0 3.0 Bilateral Bonds 0.0 0.0 0.0 0.3 0.3 38% .SBURSEMENTS 0.2 2.3 3.0 3.9 2.8 Multilateral iblic and publicly guaranteed 0.2 2.2 2.7 3.5 2.6 48% ficial creditors 0.1 1.5 2.1 2.5 2.2 Multilateral 0.0 0.5 1.4 1.6 1.5 Bilateral 0.1 1.0 0.7 0.8 0.7 ivate creditors 0.0 0.7 0.6 1.0 0.4 Bonds 0.0 0.1 0.0 0.0 0.0 ivate nonguaranteed 0.0 0.1 0.2 0.4 0.2 Bonds 0.0 0.0 0.0 0.0 0.0 tlNCIPAL REPAYMENTS 0.1 0.6 1.2 2.0 1.8 Composition of net flows iblic and publicly guaranteed 0.1 0.5 1.1 1.8 1.6 US$bdhon on long-term debt ficial creditors 0.0 0.2 0.7 1.2 1.1 bl Multilateral 0.0 0.0 0.4 0.6 0.6 Bilateral 0.0 0.1 0.3 0.6 0.5 2 vate creditors 0.0 0.3 0.4 0.6 0.5 Bonds 0.0 0.0 0.0 0.0 0.0 2 ivate nonguaranteed 0.0 0.1 0.1 0.2 0.3 Bonds 0.0 0.0 0.0 0.0 00 1- ,T FLOWS ON DEBT 0.1 1.7 1.8 1.9 0.9 iblic and publicly guaranteed 0.1 1.7 1.7 1.6 1.0 ficial creditors 0.1 1.3 1.5 1.3 1.1 Multilateral 0.0 0.5 1.0 1.0 1.0 0 Bilateral 0.1 0.9 0.4 0.2 0.1 vate creditors 0.0 0.4 0.2 0.4 -0.1 -1 Bonds 0.0 0.0 0.0 0.0 0.0 ivate nonguaranteed 0.0 0.0 0.1 0.2 -0.1 1 Bonds 0.0 0.0 0.0 0.0 0.0 1993 1994 1995 1996 1997 1998 1999 2000 I Private U Official utsche mark 9.1 7.0 2 5 3.2 3.0 Net flows on long-term debt by borrower -nch tranc 5.6 12.4 10.1 4.9 4.3 >anese yen 0.1 1.2 3.4 9.3 8.9 Government and public enterprises und sterling 45.6 10.9 3.9 2.1 1.8 S. dollars 14.6 28.1 35.7 51.9 54.1 2000 jltiple currency 6.0 11.3 11.4 9.9 9.4 1999 I other currencies 18.3 26.1 30.6 5.7 6.1 1998 DEBTS Of~K- ~W R~0 %1UAT40N1997 tal change in debt stocks .. .. 4.9 -0.1 -1.7 1996 Net flows on debt 0.1 2.2 2.7 2.8 0.0 1995 Net change in interest arrears .. .. 0.1 -0.1 0.3 Interest capitalized .. .. 0.2 0.1 0.0 1994 Debt forgiveness or reduction .. .. -1.5 -0.7 -0.3 1993 Cross-currency valuation .. .. 1.3 -1.8 -2.3 Residual .. .. 2.1 -0.4 0.4 0.0 0.5 10 15 2.0 25 30 US$ billion ,L CREDITORS Private sector Interest (%) 1.8 4.2 3.0 2.4 1.5 Maturity (years) 28.4 24.2 51.9 27.8 36.2 2000. . Grant element (%) 63.7 42.8 54.9 56.9 70.5 2 1rcial creditors 1999 Interest (%) 1.4 3.5 2.5 1.5 1.3 1998 Maturity (years) 30.2 27.8 57.7 33.1 37.4 Grant element (%) 68.3 49.9 61.1 68.4 73.0 ivate creditors 1998 Interest (%) 6.1 6.7 6.9 5.2 7.2 1995 Maturity (years) 7.8 11.5 10.9 11.1 5.3 1994 Grant element (%) 12.0 18.0 10.8 20.8 6.9 emo: 1993 mmitments 0.2 3.1 3.6 3.4 1.9 - 0 3 Official creditors 0.2 2.4 3.2 2.6 1.9 -0.2 01 00 0.1 02 03 04 Pnvate creditors 0.0 0.7 0.5 0.8 0.1 242 MODERATELY INDEBTED MIDDLE-INCOME COUNTRIES (US$ billion) 1970 1980 1990 1999 2000 US$ bilbon External debt S YE T -A --- TOTAL DEBT STOCKS (EDT) 17.2 149.9 351.3 722.0 703.8 700 - Long-term debt (LDOD) 13.6 104.7 288.5 595.5 583.4 Public and publicly guaranteed 10.1 89.4 266.5 413.6 403.0 600 Private nonguaranteed 3.5 15.2 22.1 181.9 180.5 Use of IMF credit 0.2 3.2 7.4 26.5 25.2 Short-term debt 3.4 42.1 55.3 99.9 95.1 500 of which interest arrears on LDOD 0.0 0.0 6.0 7.0 4.8 Memo: 400 IBRD 1.3 5.9 27.2 28.6 28.7 IDA 0.2 0.6 1.1 2.9 2.9 300 TOTAL FLOWS ON DEBT 200 Disbursements 3.5 30.3 53.1 82.4 74.7 Long-term debt 3.3 28.9 50.9 79.6 70.6 oo Public and publicly guaranteed 2.2 21.8 43.8 46.9 41.9 Private nonguaranteed 1.1 7.2 7.1 32.7 28.6 0 IMF purchases 0.2 1.3 2.2 2.8 4.2 1993 1994 1995 1996 1997 1998 1999 2000 Memo: M Official 0 Private IBRD 0.2 1.3 3.8 3.4 3.5 IDA 0.0 0.0 0.1 0.4 0.2 Principal repayments 1.6 11.9 38.3 61.8 65.4 Long-term debt 1.4 11.5 36.6 56.1 61.3 Public and publicly guaranteed 0.8 8.3 34.1 29.3 32.5 US$ billion Private nonguaranteed 0.6 3.2 2.5 26.7 28.8 IMF repurchases 0.2 0.4 1.8 5.7 4.1 Memo: IBRD 0.1 0.3 2.5 2.8 2.7 80-, IDA 0.0 0.0 0.0 0.0 0.0 Net flows on debt 1.9 25.6 10.3 8.0 7.7 60 of which short-term debt 0.0 7.3 -4.5 -12.6 -1.6 Interest payments (INT) 0.5 12.9 25.3 36.4 37.5 Long-term debt 0.5 7.9 20.7 30.5 31.1 Net transfers on debt 1.4 12.8 -15.0 -28.4 -29.8 Total debt service (TDS) 2.1 24.8 63.6 98.2 102.9 30 AGGREGATE NET RESlJRCgE FLOWS AND NETTRANSFFERS (LONGTERM) 20- NET RESOURCE FLOWS 2.6 21.2 27.3 69.6 48.9 Net flow of long-term debt (ex. IMF) 1.9 17.5 14.4 23.5 9.3 0 Foreign direct investment (net) 0.3 2.4 8.2 34.8 29.2 1993 1994 1995 1996 1997 1998 1999 2000 Portfolio equity flows 0.0 0.0 1.2 6.2 5.8 U Private non-debt flows E Private debt flows 0 Official flows Grants (excluding technical coop.) 0.4 1.3 3.5 5.1 4.5 NET TRANSFERS 0.8 10.0 1.7 25.9 2.9 Interest on long-term debt 0.5 7.9 20.7 30.5 31.1 Profit remittances on FDI 1.3 3.3 4.9 13.2 14.9 Debt indicators Percent MAjORECONOMICINDICATORS 1600 - Gross national income (GNI) .. 842.3 1,204.1 1,126.2 1,223.5 140.0 Exports of goods & services (XGS) .. 159.8 258.6 521.5 625.3 of which workers' remittances .. 3.1 7.0 9.3 9.1 1200 Imports of goods & services (MGS) .. 164.5 266.6 481.1 561.1 International reserves (RES) .. 58.7 71.8 202.2 220.5 1000 Current account balance .. -6.2 -5.8 43.1 53.2 80.0 DEBT INDICATORS EDT / XGS (%) .. 93.8 135.8 138.5 112.5 EDT/GNI(%) .. 17.8 29.2 64.1 57.5 TDS / XGS (%) .. 15.5 24.6 18.8 16.5 INT / XGS (%) .. 8.0 9.8 7 0 6.0 20.0 INT / GNI(%) .. 1.5 2.1 3.2 3.1 RES / MGS (months) .. 4.3 3.2 5.0 4.7 Short-term / EDT (%) 19.8 28.1 15.8 13.8 13.5 1993 1994 1995 1996 1997 1998 1999 2000 Concessional / EDT (%) 25.6 8.8 9.0 7.0 6.9 ED/G- DTN Multilateral / EDT (%) 11.5 6.6 13.7 8.6 8.7f 243 MODERATELY INDEBTED MIDDLE-INCOME COUNTRIES '$ billion) 1970 1980 1990 1999 2000 Composition of long-term debt, 1999 :BT OUTSTANDING (LDOD) 13.6 104.7 288.5 595.5 583.4 blic and publicly guaranteed 10.1 89.4 266.5 413.6 403.0 icial creditors 6.8 30.5 96.9 198.9 193.3 Vultilateral 2.0 9.9 48.1 62.3 61.4 Bilateral 31lateral 4.9 20.7 48.8 136.6 131.9 23% vate creditors 3.2 58.9 169.5 214.8 209.7 3onds 0.5 4.1 39.8 111.9 139.5 vate nonguaranteed 3.5 15.2 22.1 181.9 180.5 Private 3onds 0.0 0.0 0.1 35.7 34.0 66% 3BURSEMENTS 3.3 28.9 50.9 79.6 70.6 blic and publicly guaranteed 2.2 21.8 43.8 46.9 41.9 Mtilateral icial creditors 1.1 5.5 15.9 14.9 13.2 Vultilateral 0.4 2.1 7.4 8.0 6.9 3ilateral 0.7 3.4 8.5 7.0 6.3 vate creditors 1.1 16.2 27.9 32.0 28.7 3onds 0.0 0.5 3.3 18.2 17.2 vate nonguaranteed 1.1 7.2 7.1 32.7 28.6 3onds 0.0 0.0 0.0 4.0 1.8 INCIPAL REPAYMENTS 1.4 11.5 36.6 56.1 61.3 Composition of net flows blic and publicly guaranteed 0.8 8.3 34.1 29.3 32.5 US$ billion on long-term debt icial creditors 0.3 1.6 7.3 10.3 11.8 4 dultilateral 0.2 0.4 3.8 5.0 5.7 3ilateral 02 1.1 3.5 53 6.1 35 vate creditors 0.4 6.7 26.7 19.0 20.6 30 3onds 0.1 0.1 2.1 7.7 7.8 vate nonguaranteed 0.6 3.2 2.5 26.7 28.8 25 3onds 0.0 0.0 0.0 1 4 3.3 20 T FLOWS ON DEBT 1.9 17.5 14.4 23.5 9.3 15 blic and publicly guaranteed 1.4 13.5 9.7 17.6 9.5 icial creditors 0.7 4.0 8.6 4.6 1.4 10-- Iultilateral 0.2 1.7 3.6 3.0 1.2 Ilateral 0.5 2.3 5.0 1.7 0.2 5 vate creditors 0.7 9.5 1.2 130 8.1 londs 0.0 0.4 1.2 104 9.5 vate nonguaranteed 0.4 4.0 4.6 6.0 -0.2 - 3onds 0.0 0.0 0.0 2.5 -1.5 1993 1994 1995 1996 1997 1998 N Private 0 Official itsche mark 7.3 7.5 14.9 11.7 10.0 Net flows on long-term debt by borrower nch franc 5.9 8.4 4.3 1.7 1.8 anese yen 0.5 8.0 13.6 13.8 13.0 Govement and public enterprises ind sterling 6.2 1.9 1.4 0.7 0.6 ;. dollars 56.3 47.9 39.0 59.5 61.1 2000 Itiple currency 15.7 16.4 15.8 6.9 6.4 1999 other currencies 6.9 7.4 7.4 4.2 5.7 1998 9 u noW B M Ud 1997 al change in debt stocks .. .. 27.1 8.3 -18.3 199 qet flows on debt 1.9 25.6 10.3 8.0 7.7 995 et change in interest arrears .. .. 4.5 1.6 -2.2 nterest capitalized .. .. 0.4 5.3 4.4 1994 )ebt forgiveness or reduction .. .. -4.1 -0.5 -11.7 1 ross-currency valuation .. .. 14.0 -6.0 -14.6 Zesidual .. .. 2.0 -0.2 -1.9 00 5.0 10.0 150 20.0 250 300 350 400 - US$ billion .L CREDITORS Prvate sector nterest (%) 5.7 9.7 7.9 6.0 7.1 vlaturity (years) 16.8 15.5 13.4 14.4 12.1 2000 3rant element (%) 25.0 3.5 12.2 23.1 13.7 ricial creditors 1999 nterest (%) 4.7 7.0 6.9 3.9 5.8 1998 vlaturity (years) 22.9 23.7 17.6 19.9 19.1 1997 3rant element (%) 35.3 21.1 21.0 39.2 27.8 vate creditors 1996 nterest (%) 6.9 11.4 8.6 7.6 7.6 1995 1 vlatunty (years) 9.5 10.5 10.7 10.3 9.1 ]rant element (%) 12.7 -7.2 6.6 11.2 7.4 1994 mo: 1993 nmitments 2.6 25.1 49.0 48.7 45.0 )fficial creditors 1.4 9.5 19.1 20.8 13.7 0.0 500 100 2 Private creditors 1.2 15.6 29.9 27.9 31.26 244 OTHER DEVELOPING COUNTRIES (US$ billion) 1970 1980 1990 1999 2000 US$ billion External debt SUMMARY DEBT DATA 1,000------------------ TOTAL DEBT STOCKS (EDT) 24.8 222.6 516.3 923.6 890.6 900 Long-term debt (LDOD) 21.9 159.8 410.2 751.5 727.3 Public and publicly guaranteed 17.7 137.1 391.3 584.0 542.6 800 Private nonguaranteed 4.3 22.7 18.9 167.5 184.7 700 Use of IMF credit 0.2 4.6 13.2 17.2 10.6 Short-term debt 2.7 58.1 92.9 154.9 152.8 600 of which interest arrears on LDOD 0.0 0.7 12.9 5.8 5.2 Memo: 500- IBRD 1.6 9.2 33.8 52.6 52.5 400 IDA 1.1 6.8 23.4 41.5 43.5 300- TOTAL FLOWS ON DEBT Disbursements 4.2 44.6 50.8 107.3 110.3 200- Long-term debt 4.1 41.6 46.0 104.4 108.6 Public and publicly guaranteed 2.9 34.7 40.9 62.1 56.7 100 Private nonguaranteed 1.2 7.0 5.2 42.4 51.9 0 IMF purchases 0.0 3.0 4.8 2.8 1.7 1993 1994 1995 1996 1997 1998 1999 2000 Memo: E Official 0 Pnvate IBRD 0.2 1.8 6.2 5.9 5.5 IDA 0.1 0.9 2.0 2.4 2.4 Principal repayments 2.7 16.3 28.2 120.4 119.5 Long-term debt 2.4 15.5 24.8 102.9 111.8 Public and publicly guaranteed 1.5 12.2 20.7 64.8 70.1 US$ billion Private nonguaranteed 0.8 3.3 4.1 38.2 41.6 180- IMF repurchases 0.3 0.8 3.4 17.5 7.8 160-T Memo: IBRD 0.1 0.5 3.1 4.0 4.2 140 IDA 0.0 0.0 0.1 0.5 0.6 120 Net flows on debt 1.4 41.2 33.3 -18.5 -9.2 of which short-term debt 0.0 12.9 10.6 -5.4 0.0 100 Interest payments (INT) 0.8 18.0 26.9 45.0 46.6 80 Long-term debt 0.8 11.8 19.4 35.3 37.4 Net transfers on debt 0.7 23.2 6.4 -63.4 -55.9 60- Total debt service (TDS) 3.5 34.3 55.1 165.4 166.2 AGGREGATE NET RES)URCELivWSAND NET T1ANS'( lSG14E1M) 0 20- NET RESOURCE FLOWS 3.8 28.8 42.7 126.3 136.8 Net flow of long-term debt (ex. IMF) 1.7 26.2 21.2 1.5 -3.2 0 1. Foreign direct investment (net) 1.2 -1.0 10.8 89.2 89.2 1993 1994 1995 1996 1997 1998 1999 2000 Portfolio equity flows 0 0 0.0 1.2 24.3 39.0 U Private non-debt flows U Private debt flows DOffical flows Grants (excluding technical coop.) 0.8 3.6 9.5 11.2 11.8 NET TRANSFERS 1.0 5.7 19.6 76.7 83.1 Interest on long-term debt 0.8 11.8 19.4 35.3 37.4 Profit remittances on FDI 2.0 11.4 3.7 14.3 16.2 Debt indicators Percent CMOR EONOMICINDICATOD I ( A 120.0 Gross national income (GNI) 324.4 1,346.6 2,014.2 3,424.0 3,744.9 Exports of goods & services (XGS) .. 380.4 480.0 1,031.5 1,260.7 of which workers' remittances .. 7.6 13.0 31.2 34.1 Imports of goods & services (MGS) .. 339.4 484.2 1,025.8 1,242.0 International reserves (RES) .. .. 112.2 442.3 491.6 80.0 Current account balance .. 20.9 -10.0 10.7 20.8 IDEAT INDICATORS EDT / XGS (%) .. 58.5 107.6 89.5 70.7 400 EDT / GNI(%) 7.6 16.5 25.6 27.0 23.8 TDS / XGS(%) .. 9.0 11.5 16.0 13.2 INT / XGS (%) .. 4.7 5.6 4.4 3.7 200 INT / GNI (%) 02 1.3 1.3 1.3 1.2 RES / MGS (months) .. .. 2.8 5.2 4.8 0 Short-term / EDT (%) 10.8 26.1 18.0 16.8 17.2 1993 1994 1995 1996 1997 1998 1999 2000 Concessional / EDT (%) - 40.1 18.6 24.5 17.5 17.3 Multilateral / EDT (%) 11.9 9.6 14.4 15.1 16.0e 245 OTHER DEVELOPING COUNTRIES JS$ billion) 1970 1980 1990 1999 2000 Composition of long-term debt, 1999 lr6 1EMUE lEBT OUTSTANDING (LDOD) 21.9 159.8 410.2 751.5 727.3 ublic and publicly guaranteed 17.7 137.1 391.3 584.0 542.6 iflicial creditors 13.1 66.3 226.3 308.6 286.0 Bilateral Multilateral 2.9 21.3 74.1 139.3 142.6 20% Bilateral 10.2 45.0 152.3 169.3 143.4 rivate creditors 4.5 70.8 165.0 275.4 256.6 Bonds 0.5 4.1 52.4 123.4 121.3 Private i rivate nonguaranteed 4.3 22.7 18.9 167.5 184.7 60% Bonds 0.0 0.0 0.6 49.9 55.7 ISBURSEMENTS 4.1 41.6 46.0 104.4 108.6 ublic and publicly guaranteed 2.9 34.7 40.9 62.1 56.7 ifficial creditors 1.9 11.8 19.0 24.4 21.9 Multilateral Multilateral 0.4 3.6 10.4 13.3 12.4 20% Bilateral 1.5 8.3 8.6 11.1 9.5 rivate creditors 1.0 22.8 21.8 37.7 34.8 Bonds 0.0 0.5 2.6 14.2 17.1 rivate nonguaranteed 1.2 7.0 5.2 42.4 51.9 Bonds 0.0 0.0 0.6 8.5 13.0 RINCIPAL REPAYMENTS 2.4 15.5 24.8 102.9 111.8 Composition of net flows ublic and publicly guaranteed 1.5 12.2 20.7 64.8 70.1 US$billion on long-term debt fficial creditors 0.7 3.1 8.5 19.9 20.8 40 -b -o Multilateral 0.1 0.6 4.1 6.7 7.6 Bilateral 0.6 2.5 4.3 13.2 13.3 35 rivate creditors 0.8 9.1 12.2 44.9 49.3 30 Bonds 00 0.2 2.5 7.9 17.6 25 rivate nonguaranteed 0.8 3.3 4.1 38.2 41.6 Bonds 0.0 0.0 0.0 7.2 6.4 20 ET FLOWS ON DEBT 1.7 26.2 21.2 1.5 -3.2 15 ublic and publicly guaranteed 1.4 22.5 20.1 -2.7 -13.4 10 fficial creditors 1.2 8.8 10.6 4.5 1.1 Multilateral 0.3 2.9 6.3 6.6 4.9 5 Bilateral 0.9 5.8 4.3 -2.1 -3.7 0 ivate creditors 0.2 13.7 9.6 -7.2 -14.5 -5 Bonds 0.0 0.4 0.2 6.3 -0.5 rivate nonguaranteed 0.4 3.7 1.1 4.2 10.2 -l0 Bonds 0.0 0.0 0.6 1.3 6.6 1993 1994 1995 1996 1997 1998 1999 2000 N Private N Official eutsche mark 9.7 6.1 7.1 4.1 3.5 Net flows on long-term debt by borrower -ench franc 4.0 3.6 5.4 3.0 2.8 panese yen 3.1 5.4 10.7 12.1 11.3 Government and public enterprises und sterling 12.8 5.1 2.4 1.1 1.0 .S. dollars 45.8 55.5 43.4 57.9 61.6 2000 lultiple currency 10.2 10.8 15.6 8.7 7.9 1999 11 other currencies 13.4 6.1 10.6 6.5 5.3 1998 D~~~~~~~~ BTS 6lKF O E N8 L A I ~ 997 Dtal change in debt stocks .. .. 32.9 13.4 -33.0 1996 Net flows on debt 1.4 41.2 33.3 -18.5 -9.2 1995 Net change in interest arrears .. .. -0.5 0.9 -0.6 Interest capitalized .. .. 2.8 0.6 0.0 1994 Debt forgiveness or reduction .. .. -20.9 0.0 -9.4 1993 Cross-currency valuation .. .. 15.2 -1.9 -16.4 Residual 3.1 32.3 2.6 0.0 5.0 100 150 200 250 300 350 US$ billion LL CREDITORS Private sector Interest (%) 5.1 9.1 7.0 5.4 6.9 Maturity (years) 22.2 17.3 17.8 17.2 11.5 2000 Grant element (%) 26.3 6.7 18.2 22.6 15.3 iticial creditors 1999 Interest (%) 3.9 4.8 5.4 3.5 5.7 1998 Maturity (years) 27.9 27.3 23.8 17.8 21.4 1997 Grant element (%) 34.3 27.3 30.6 36.6 29.6 rivate creditors 19 Interest (%) 7.7 12.6 8.5 7.0 7.4 1995 1 Maturity (years) 9.9 9.3 12.5 16.8 7.5 1994 __ Grant element (%) 9.5 -9.7 7.0 10.8 9.5 I lemo: 1993 ommitments 3.6 35.5 47.0 56.1 40.9 Official creditors 2.5 15.7 22.2 25.7 11.6 00 20 40 60 80 100 120 14.0 160 Private creditors 1.2 19.8 24.8 30.3 29.2 246 LOW-INCOME COUNTRIES (US$ billion) 1970 1980 1990 1999 2000 70uS$ billion External debt S M IIBDT 700 TOTAL DEBT STOCKS (EDT) 23.5 125.2 418.6 574.0 550.5 Long-term debt (LDOD) 21.7 102.9 357.3 487.5 471.1 600 Public and publicly guaranteed 20.8 94.9 340.3 422.4 412.8 Private nonguaranteed 0.9 8.0 17.0 65.2 58.2 500- Use of IMF credit 0.3 5.3 11.3 24.8 23.6 Short-term debt 1.5 17.0 50.0 61.7 55.9 400 - of which interest arrears on LDOD 0.0 1.7 12.9 26.1 20.6 Memo: IBRD 1.4 4.8 28.1 28.9 27.6 300 - IDA 1.6 10.5 38.6 70.5 72.9 TOTAL FLOWS ON DEBT 200 Disbursements 3.5 21.8 33.0 31.3 30.8 Long-term debt 3.4 19.0 30.5 27.6 28.6 100 Public and publicly guaranteed 3.1 16.8 24.7 23.5 24.2 Private nonguaranteed 0.3 2.2 5.8 4.1 4.5 0 : IMF purchases 0.1 2.8 2.5 3.7 2.2 1993 1994 1995 1996 1997 1998 1999 2000 Memo: 1 Official * Private IBRD 0.2 0.9 3.4 3.1 2.1 IDA 0.1 1.5 3.6 4.2 4.4 Principal repayments 1.3 6.1 16.8 30.3 30.1 Long-term debt 1.0 5.4 14.6 28.4 27.8 Public and publicly guaranteed 0.9 4.0 12.6 19.1 18.1 60 bn Private nonguaranteed 0.1 1.4 2.0 9.3 9.7 IMF repurchases 0.3 0.7 2.2 2.0 2.3 Memo: 0 IBRD 0.1 0.2 1.8 2.5 2.5 IDA 0.0 00 0.2 0.8 0.9 Net flows on debt 2.2 19.3 23.7 0.0 0.5 of which short-term debt 0.1 3.6 7.5 -0.9 -0.3 Interest payments (INT) 0.5 6.0 15.4 16.5 17.7 30 Long-term debt 0.5 4.3 12.4 13.9 14.8 Net transfers on debt 1.7 13.3 8.3 -16.5 -17.2 Total debt service (TDS) 1.9 12.1 32.2 46.9 47.8 NET RESOURCE FLOWS 3.7 21.0 33.5 26.9 25.2 Net flow of long-term debt (ex. IMF) 2.4 13.6 15.9 -0.8 0.9 0, Foreign direct investment (net) 0.3 0.2 2.2 9.7 6.6 1993 1994 1995 1996 1997 1998 1999 2000 Portfolio equity flows 0.0 0.0 0.4 2.6 2.5 U Private non-debt flows M Private debt flows * Official flows Grants (excluding technical coop.) 1.0 7.1 14.9 15.3 15.2 NET TRANSFERS 2.3 10.8 17.7 7.1 4.9 Interest on long-term debt 0.5 4.3 12.4 13.9 14.8 Profit remittances on FDI 0.9 5.8 3.4 5.7 5.5 Debt indicators Percent j;§RE @N@Xfl ICTO@RS n,350.0 --- -- - ---. -- - Gross national income (GNI) 165.8 595.4 851.7 983.3 1,031.1 Exports of goods & services (XGS) .. .. 140 1 251.0 303.3 of which workers' remittances .. 5.3 7.0 18.5 18.2 Imports of goods & services (MGS) .. .. 172.3 278.2 315.7 250.0 International reserves (RES) 36.7 26.2 88.5 97.0 Current account balance -9.5 -22.0 -10.7 6.3 200.0 &B INDICA R15 EDT / XGS (%) .. .. 298.8 228.7 181.5 EDT / GNI(%) 14.2 21.0 49.2 58.4 53.4 1000 TDS / XGS(%) .. .. 23.0 18.7 15.8 INT / XGS (%) .. .. 11.0 6.6 5.8 500 INT / GNI(%) 0.3 1.0 1.8 1.7 1.7 RES / MGS (months) .. .. 1.8 3.8 3.7 0.0 1 i Short-term / EDT (%) 6.5 13.6 12.0 10.8 10.2 1993 1994 1995 1996 1997 1998 1999 2000 Concessional / EDT (%) 64.7 45.5 43.6 39.6 40.1 ED/S- TG Multilateral / EDT (%) 13.0 15.5 21.9 25.9 26.8f 247 LOW-INCOME COUNTRIES IS$ billion) 1970 1980 1990 1999 2000 W"_ Composition of long-term debt, 1999 EBT OUTSTANDING (LDOD) 21.7 102.9 357.3 487.5 471.1 Private ablic and publicly guaranteed 20.8 94.9 340.3 422.4 412.8 28% fficial creditors 18.2 70.2 262.1 344.8 337.2 Multilateral 3.1 19.4 91.8 148.8 147.7 Bilateral 15.1 50.8 170.3 196.0 189.5 ivate creditors 2.6 24.7 78.2 77.6 75.7 Bonds 0.3 0.8 3.6 14.7 19.6 rivate nonguaranteed 0.9 8.0 17.0 65.2 58.2 Bonds 0.0 0.0 0.1 11.6 9.5 Bilateral 41% ISBURSEMENTS 3.4 19.0 30.5 27.6 28.6 ublic and publicly guaranteed 3.1 16.8 24.7 23.5 24.2 fficial creditors 2.4 9.4 18.6 19.4 16.3 Multilateral 0.4 3.6 10.6 11.7 10.2 Multilateral Bilateral 2.0 5.7 8.0 7.7 6.1 ivate creditors 0.7 7.5 6.1 4.1 7.8 Bonds 0.0 0.1 0.4 0.0 5.5 rivate nonguaranteed 0.3 2.2 5.8 4.1 4.5 Bonds 0.0 0.0 0.1 0.0 0.4 IUNCIPAL REPAYMENTS 1.0 5.4 14.6 28.4 27.8 Composition of net flows iblic and publicly guaranteed 0.9 4.0 12.6 19.1 18.1 US$ billion on long-term debt fficial creditors 0.6 1.9 6.7 10.0 10.9 15 Multilateral 0.1 0.4 3.0 5.2 5.7 Bilateral 0.5 1.5 3.7 4.8 5.2 ivate creditors 0.3 2.1 5.9 9.1 7.2 10 Bonds 0.0 0.0 0.4 0.8 0.6 ivate nonguaranteed 0.1 1.4 2.0 9.3 9.7 5 Bonds 0.0 0.0 0.0 2.0 2.4 ET FLOWS ON DEBT 2.4 13.6 15.9 -0.8 0.9 0 - f- - iblic and publicly guaranteed 2.2 12.8 12.1 4.3 6.1 ficial creditors 1.8 7.5 11.9 9.3 5.4 -5 Multilateral 0.3 3.3 7.6 6.5 4.5 Bilateral 1.6 4.2 4.3 2.8 0.9 Ivate creditors 0.4 5.3 0.2 -5.0 0.7 Bonds 0.0 0.1 0.0 -0.8 4 9 ivate nonguaranteed 0.2 0.8 3.8 -5.1 -5.2 -15 Bonds 0.0 0.0 0.1 -2.0 -2.1 1993 1994 1995 1996 1997 1998 1999 2000 M Private U Official :utsche mark 8.0 7.2 5.4 4.0 3.1 Net flows on long-term debt by borrower -nch franc 5.5 7.0 6.5 4 3 3.6 panese yen 4.3 9.1 10.9 15.4 13.5 Government and public enterprises und sterling 18.2 8.7 4.0 2.6 1.8 S. dollars 37.4 37.7 37.7 48.1 55.5 2000 ultiple currency 8.1 8.1 12.8 10.3 9.6 1999 l other currencies 18.1 16.5 18.5 9.6 7.3 1998 EMIL-.BTS KF WR.ECI.Oi6NtW1LA.T161 1997 tal change in debt stocks .. .. 47.7 -2.8 -23.5 1996 Net flows on debt 2.2 19.3 23.7 0.0 0.5 1995 Net change in interest arrears .?3.0 2.9 -5.5 Interest capitalized .. .. 1.7 1.8 9.1 1994 Debt forgiveness or reduction .. .. -2.8 -5.2 -9.5 1993 Cross-currency valuation .. .. 14.3 -2.4 -18.0 Residual .. .. 7.9 0.1 0.0 0.0 50 100 150 200 US$ billion E VERA16KG 1 ~1MsTET L CREDITORS Private sector Interest (%) 3.0 6.3 4.8 3.1 4.8 Maturity (years) 28.9 22.9 26.5 23.8 20.2 2000 Grant element (%) 47.7 24.2 35 8 49.3 37.0 icial creditors 1999 Interest (%) 2.1 3.5 4.0 2.7 3.2 1998 Maturity (years) 33.8 30.9 30.5 26.0 29.7 1997 Grant element (%) 56.9 39.4 42.4 53.9 54.8 ivate creditors 1996 Interest (%) 6.4 10.8 7.2 5.6 7.4 1995 Maturity (years) 10.3 10.0 14.1 9.4 5.5 Grant element (%) 12.9 -0.3 15.6 19.4 9.2 tmo: 1993 mmitments 4.5 26.5 31.5 17.4 17.6 Official creditors 3.6 16.4 23.7 15.0 10.7 -10.0 -50 00 50 100 il0 Private creditors 0.9 10.1 7.8 2.3 6.9 - on 248 MIDDLE-INCOME COUNTRIES (US$ billion) 1970 1980 1990 1999 2000 US$ billion External debt 1) =S1IMA f!.ET D%,AW 2,500 - -- -- --. ----------- . -- TOTAL DEBT STOCKS (EDT) 49.2 484.2 1,039.8 1,991.8 1,941.4 Long-term debt (LDOD) 41.0 348.7 822.0 1,605.5 1,576.6 Public and publicly guaranteed 26.5 2862 773.5 1,119.7 1,077.6 Private nonguaranteed 14.5 62.5 48.5 485.8 499.0 Use of IMF credit 0.4 7.0 23.4 54.1 40.7 Short-term debt 7.8 128.5 194.4 332.2 324.1 1,500 of which interest arrears on LDOD 0.0 0.8 39.8 14.1 12.7 Memo: IBRD 3.0 17.4 67.5 90.7 92.5 1,000 IDA 0.3 1.4 6.5 16.2 16.1 TOTAL FLOWS ON DEBT Disbursements 10.0 92.0 103.6 254.8 244.0 500 Long-term debt 9.7 88.8 97.9 243.8 236.0 Public and publicly guaranteed 5.9 69.5 84.8 1334 1194 Private nonguaranteed 3.8 19.3 13.0 110.4 116.6 IMF purchases 0.3 3.2 5.8 11.0 8.0 1993 1994 1995 1996 1997 1998 1999 2000 Memo: * Official 13 Pnvate IBRD 0.5 3.6 10.2 12.0 11.3 IDA 0.0 0.1 0.8 1.2 0.8 Principal repayments 5.5 38.4 76.7 245.5 242.1 Long-term debt 5.0 37.1 70.7 220.2 223.3e Public and publicly guaranteed 2.7 26.7 63.4 113.5 123.0 3S50 l~n Ageae e eouc lw 23500 .-. - --- Private nonguaranteed 2.3 10.4 7.3 106.7 100.31 IMF repurchases 0.4 1.3 6.0 25.3 18.8 Memo: 300- LBRD 0.2 0.8 6.7 7.5 7.5 IDA 0.0 00 0.0 0.1 0.2 250 Net flows on debt 4.5 83.1 36.0 -11.8 - 1.3 20 of which short-term debt 0.0 29.5 9.0 -21.1 -3.2 Interest payments (INT) 1.8 42.8 54.9 105.5 109.0 Long-term debt 1.8 28.5 42.1 85.7 89.5 150 1 R_ Z O Off4cial in P n at NeDtAnfrondb 2.6 40.3 .8 -1.3 10.3 L-AGREGTE N11 4ES V4@11W A1&0ET S' ",.ONG iffiE, NET RESOURCE FLOWS 7.5 61.9 64.7 245.0 236.0 Net flow of long-term debt (ex. IMF) 4.6 I .7 27.2 23.6 12.7 0 Foreign direct investment (net) 1.9 4.2 21.9 174.6 160.1 1993 1994 1995 1996 1997 1998 1999 2000 Portfolio equity flows 0.0 0.0 2.3 31.8 48.3 Private non-debt flows 0 Private debt flows 0 Official flows Grants (excluding technical coop.) 0.9 6.0 13.2 14.9 14.7 NET TRANSFERS 1.3 16.5 11.1 124.7 106.6 Interest on long-term debt 1.8 28.5 42.1 85.7 89.5 Profit remittances on FDI 4.4 16.9 11.4 34.6 39.8 Debt indicators Percent IDA 00 0 0 0.010100. Gross national income (GNI) 504.3 2,316.3 3,428.3 4,892.6 5,346.2 140.0 Exports of goods & services (XGS) .. 566.8 766.9 1,565.3 1,895.9 of which workers' remittances . 8.8 18.4 32.0 34.2 1200 Imports of goods & services (MGS) .. 544.9 764.7 1,552.4 1,843.1 International reserves (RES) . 175.4 196.9 675.5 . 1000 Current account balance .6 1.6 -2.9 210 426 r ~80.0- 600 EDT / XGSe D .3 85.4 135.6 127.3 102.4 ED /GN () .8 20.9 30.3 40.7 36.3 40 TDS / XGS (% . 14.3 17.2 22.4 18.5 INT / XGS O C W .5 7.6 7.2 6.7 5.8 20.0. INT / GNI (%) 0.4 1.9 1.6 2.2 2.0 RES/M S (months) e n n3.9 3.1 5.2 4.8 00 Short-term/EDT (%) 15.9 26.6 18.7 16.7 16.7 1993 1994 1995 1996 1997 1998 1999 2000 Concessional / EDT (G) 19.3 11.1 12.6 8.6 8.3 S- TG Multilateral / EDT (%) 8.7 6.1 11.1 9.9 10.2 249 MIDDLE-INCOME COUNTRIES 'US$ billion) 1970 1980 1990 1999 2000 Composition of long-term debt, 1999 DEBT OUTSTANDING (LDOD) 41.0 348.7 822.0 1,605.5 1,576.6 Public and publicly guaranteed 26.5 286.2 773.5 1,119.7 1,077.6 Official creditors 15.5 108.1 342.1 520.7 506.7 Bilateral Multilateral 4.3 29.4 115.6 196.9 198.9 20% Bilateral 11.2 786 226.5 323.9 307.8 Private creditors 11.0 178.1 431.4 598.9 570.9 Bonds 1.5 12.3 103.7 352.2 3728 Private nonguaranteed 14.5 62.5 48 5 485.8 499.0 Private Bonds 0.0 0.0 0.7 111.8 115.1 67% DISBURSEMENTS 9.7 88.8 97.9 243.8 236.0 Public and publicly guaranteed 5.9 69.5 84.8 133.4 119.4 Multilateral Official creditors 2.6 19.8 34.3 43.3 37.8 13% Multilateral 0.9 5.7 16.9 26.3 23.4 Bilateral 1.7 14.1 17.4 17.0 14.5 Private creditors 3.3 49.8 50.5 90.1 81.6 Bonds 0.1 1.6 5.7 53.4 49.9 Private nonguaranteed 3.8 19.3 13.0 110.4 116.6 Bonds 0.0 0.0 0.6 15.1 19.1 PRINCIPAL REPAYMENTS 5.0 37.1 70.7 220.2 223.3 Composition of net flows Public and publicly guaranteed 2.7 26.7 63.4 113.5 123.0 US$ billion on long-term debt )fficial creditors 1.0 5.3 18.8 35.5 37.9 1oo Multilateral 0.3 1.3 9.6 14.1 17.4 Bilateral 0.7 4.0 9.3 21.4 20.5 "rivate creditors 1.7 21.4 44.5 78.0 85.1 80 Bonds 0.1 0.5 5.2 24.9 40.2 rivate nonguaranteed 2.3 10.4 7.3 106.7 100.3 60 Bonds 0.0 0.0 0.0 11.4 14.8 qET FLOWS ON DEBT 4.6 51.7 27.2 23.6 12.7 40 "ublic and publicly guaranteed 3.1 42.8 21.5 19.9 -3.6 )fficial creditors 1.6 14.4 15.5 78 -0 1 20 Multilateral 0.6 4.4 7.4 12.2 6.0 Bilateral 1.0 10.0 8.1 -4.4 -6.1 01 - Irivate creditors 1.5 28.4 6.0 12.1 -3.6 0- Bonds 0.0 1.1 0.5 28.6 9.8 "rivate nonguaranteed 1.5 8.9 5.7 3.7 16.3 -20 Bonds 0.0 0.0 0.6 37 4.3 1993 1994 1995 1996 1997 1998 1999 2000 U Private E Official )eutsche mark 9.0 6.4 10.1 7.1 6.2 Net flows on long-term debt by borrower rench franc 4.9 4.9 5.3 2.5 2.4 apanese yen 0.7 6.1 10.3 11.2 10.5 Government and public enterpnses 'ound sterling 5.1 1.7 1.5 0.8 0.6 J.S. dollars 55.2 53.7 42.8 60.3 61.7 2000 1 vIultiple currency 14.2 11.8 15.5 6.7 6.1 1999 UI other currencies 9.4 5.8 6.0 5.5 6.6 1998 1 otal change in debt stocks . .. 557 21 6 -50.3 1996 Net flows on debt 4.5 83.1 36.0 -11.8 -1.3 1995 Net change in interest arrears .. .. 12.5 1.2 -1.4 Interest capitalized .. .. 4.1 5.4 4.9 1994 Debt forgiveness or reduction .. .. -31.6 -2.2 -15.8 1993 Cross-currency valuation .. .. 33.1 -13.5 -32.9 Residual 1.6 42.5 -3.9 0.0 100 200 300 400 500 600 700 800 US$ billion lLL CREDITORS Private sector Interest (%) 6.2 10.3 7.7 6.6 7.8 Maturity (years) 16.3 13.8 14.6 13.7 12.5 2000 Grant element (%) 22.8 1.6 13.8 17.0 94 )ficial creditors 1999 Interest (%) 5.0 6.7 6.4 4.3 6.1 1998 Maturity (years) 23.1 21.3 19.4 16.3 18.5 1997 Grant element (%) 34.1 23.2 24.6 32.0 25.1 Irivate creditors 1996 Interest (%) 7.4 12.2 8.7 8.2 8.3 1995 Maturity (years) 9.6 9.7 11.1 11.9 10.6 Grant element (%) 11.8 -10.3 5.7 6.6 45 1994 4emo: 1993 .ommitments 7.6 72.1 92.0 135.0 107.3 Official creditors 3.7 25.5 39.5 55.0 25.9 00 100 200 300 40.0 500 600 Private creditors 3.8 46.6 52.6 80.1 81.5 US$ billion 250 SPECIAL PROGRAM OF ASSISTANCE (US$ billion) 1970 1980 1990 1999 2000 US$ billion External debt SUMMARYDEB [ BlATA 140 --- ------ -- ------------------ - TOTAL DEBT STOCKS (EDT) 4.7 37.1 95.6 105.8 102.0 Long-term debt (LDOD) 4.2 29.0 79.8 91.4 88.5 120 Public and publicly guaranteed 4.0 25.9 75.7 87.1 84.3 Private nonguaranteed 0.2 3 1 4.1 4.4 4.2 oo- Use of IMF credit 0.1 2.0 4.5 5.1 4.9 Short-term debt 0.4 6.1 11.4 9.3 8.6 80 of which interest arrears on LDOD 0.0 1.6 3.3 2.8 3.1 Memo: IBRD 0.2 1.7 4.9 1.2 1.0 60 IDA 0.2 2.0 12.8 28.9 29.1 40 TOTAL FLOWS ON DEBT Disbursements 1.0 7.7 7.1 3.9 3.9 Long-term debt 1.0 6.9 6.5 3.4 3.4 20 Public and publicly guaranteed 1.0 6.3 6.0 3.2 3.4 Private nonguaranteed 0.1 0.6 0.4 0.2 0.0 0 . 4 - - IMF purchases 0.0 0.7 0.6 0.5 0.5 1993 1994 1995 1996 1997 1998 1999 2000 Memo: a Official N Private IBRD 0.0 0.3 0.3 0.0 0.0 IDA 0.1 0.3 1.8 1.8 2.1 Principal repayments 0.3 2.1 3.4 3.5 2.6 Long-term debt 0.2 1.9 2.7 3.0 2.2 US$bilkon Aggregate net resource flows Public and publicly guaranteed 0.2 1.5 2.3 2.5 2.0 14 Private nonguaranteed 0.0 0.4 0.4 0.5 0.3 IMF repurchases 0.0 0.2 0.7 0.5 0.4 Memo: 12 IBRD 0.0 0.I 0.4 0.3 0.2 IDA 0.0 0.0 0.0 0.2 0.3 1o Net flows on debt 0.8 6.1 5.2 0.5 0.2 of which short-term debt 0.0 0.5 1.6 0.1 -1.1 Interest payments (INT) 0.1 1.9 2.4 2.0 1.8 Long-term debt 0.1 1.4 1.7 1.7 1.4 6 Net transfers on debt 0.6 4.2 2.9 -1.5 -1.6 Total debt service (TDS) 0.4 4.0 5.8 5.5 4.4 4 2 NET RESOURCE FLOWS 0.9 7.8 13.2 11.1 11.4 Net flow of long-term debt (ex. IMF) 0.8 5.0 3.7 0.4 1.2 o Foreign direct investment (net) -0.1 0.6 0.4 2.0 1.7 1993 1994 1995 1996 1997 1998 1999 2000 Portfolio equity flows 0.0 0.0 0.0 0.0 0.0 0 Private non-debt flows I Private debt flows E3 Official flows Grants (excluding technical coop.) 0.2 2.2 9.1 8.6 8.4 NET TRANSFERS 0.5 5.7 11.0 8.9 9.4 Interest on long-term debt 0.1 1.4 1.7 1.7 1.4 Profit remittances on FDI 0.3 0.7 0.5 0.5 0.5 Debt indicators Percent AJ%1J E0 N0 CR WNNNUCA 60S00 0 Gross national income (GNI) 18.5 70.7 90.5 99.3 92.4 Exports of goods & services (XGS) .. 19.7 21.9 27.5 28.5 500.0 of which workers' remittances 0.4 0.6 0.5 0.6 Imports of goods & services (MGS) 28.3 32.6 39.2 38.8 International reserves (RES) 1.5 2.7 3.3 8.0 8.5 4000 Current account balance -7.2 -6.6 -5 8 -4 2 3000 EDT / XGS (%) .. 188.3 436.3 384.5 357.5 200.0 EDT / GNI (%) 25.4 52.5 105.6 106.5 110.3 TDS / XGS (%) .. 20.4 26.5 20.1 15.6 INT / XGS (%) .. 9.6 10.8 7.4 6.3 100.0 INT / GNI (%) 0.7 2.7 2.6 2.1 1.9 RES / MGS (months) .. 1.2 1.2 2.4 2.6 0.01 Short-term / EDT (%) 8.0 16.5 11.9 8.8 84 1993 1994 1995 1996 1997 1998 1999 2000 Concessional / EDT (%) 51.7 32.9 46.5 62 7 63 8 EDT/XGS - EDT/GNP Multilateral / EDT (%) 10.0 14.6 28.6 40.4 41.4 251 SPECIAL PROGRAM OF ASSISTANCE (US$ billion) 1970 1980 1990 1999 2000 Composition of long-term debt, 1999 LE - DE-B I Private DEBT OUTSTANDING (LDOD) 4.2 29.0 79.8 91.4 88.5 10% Public and publicly guaranteed 4.0 25.9 75.7 87.1 84.3 Official creditors 2.9 16.9 65.3 81.8 79.3 Multilateral 0.5 5.4 27.3 42.7 42.2 Bilateral 2.5 11.5 380 39.0 37 1 Private creditors 1.1 9.0 10.4 5.3 5.1 Bonds 0.2 0.0 0.0 2.4 2.3 Private nonguaranteed 0.2 3.1 4.1 4.4 4.2 Bonds 0.0 0.0 0 0 0.3 0.3 Multilateral Bilateral DISBURSEMENTS 1.0 6.9 6.5 3.4 3.4 8% 42% Public and publicly guaranteed 1.0 6.3 6.0 3.2 3.4 Official creditors 0.5 3.0 5.3 3.0 3.2 Multilateral 0.1 1.2 3.3 2.6 2.8 Bilateral 0.4 1.7 2.0 0.4 0.5 Private creditors 0.4 3.3 0.7 0.2 0.1 Bonds 0.0 0.0 0.0 0.0 0.0 Private nonguaranteed 0.1 0.6 0.4 0.2 0.0 Bonds 0.0 0.0 0.0 0.0 0.0 PRINCIPAL REPAYMENTS 0.2 1.9 2.7 3.0 2.2 Composition of net flows Public and publicly guaranteed 0.2 1.5 2.3 2.5 2.0 US$ billion on long-term debt Official creditors 0.1 0.4 1.4 2.1 1.7 4 Multilateral 0.0 0.1 0.9 1.1 1.0 Bilateral 0.1 0.3 0.5 1.0 0.7 Private creditors 0.1 1.1 0.9 0.4 0.3 3 Bonds 0.0 0.0 0.0 0.0 00 3 Private nonguaranteed 0.0 0.4 0.4 0.5 0.3 Bonds 0.0 0.0 0.0 0.0 0.0 2 NET FLOWS ON DEBT 0.8 5.0 3.7 0.4 1.2 2 Public and publicly guaranteed 0.7 4.8 3.7 0.7 1.4 1 Official creditors 0.4 2.5 3 9 0.9 1 5 Multilateral 0.1 1.1 2.4 1.5 1.8 Bilateral 0.3 1.5 1.5 -0.6 -0.2 0 Pnvate creditors 0.3 2.2 -0.2 -0.2 -0.1 1 Bonds 0.0 00 0.0 0.0 0.0 Private nonguaranteed 0.0 0.2 0.0 -0.3 -0.2 1 Bonds 0.0 0.0 0.0 0.0 0.0 1993 1994 1995 1996 1997 1998 1999 2000 0 Private s Official Deutsche mark 6.8 5.9 3.9 3.3 3 1 Net flows on long-term debt by borrower French franc 19.3 17.7 17.4 11.5 10.7 Japanese yen 0.0 7.7 3.2 6.1 5.5 Government and public enterprises Pound sterling 21.6 5.4 3.7 2.6 2.3 U.S. dollars 21.3 28.4 30.4 50.1 52.3 2000 g Multiple currency 60 8.8 12.7 8.6 8 9 1999 All other currencies 24.6 23.7 26.4 15.3 14.7 1998 MAW=S dMM M C"DNW=I IA Mill 1997 I Total change in debt stocks .. . 11.1 -88 -3.8 1996 Net flows on debt 0.8 6.1 5.2 0.5 0.2 1995 Net change in interest arrears .. .. 0.2 -0.4 0.3 Interest capitalized . .. 1.3 0.5 0.3 1994 Debt forgiveness or reduction .. .. -2.5 -5.0 -0.7 1993 Cross-currency valuation .. .. 4.0 -4.3 -4.1 : i Residual 2.8 0.0 0.1 00 10 2.0 30 40 50 US$ billion AWERE 54( IB W(& MM1MiS ALL CREDITORS Private sector Interest (%) 3.3 6.5 3.3 1.4 0.9 Maturity (years) 27.1 18.4 39.4 35.7 39.4 2000 Grant element (%) 53.2 25.0 52.7 71.4 77.7 Official creditors 1999 Interest (%) 1.8 3.7 2.7 1.3 0.9 1998 Maturity (years) 34.6 26.9 42.8 37.1 39.6 Grant element (%) 71.0 47.2 58.6 74.0 78.0 Private creditors 1996 Interest (%) 6.5 9.4 8.5 3.1 3.5 1995 Maturity (years) 10.3 9.6 120 19.0 20.6 Grant element (%) 13.0 2.2 5.1 40.6 42.7 1994 Memo: 1993 Commitments 1.3 8.9 7.7 3.3 3.1 - Official creditors 0.9 4.5 6.8 3.0 3.1 -04 -03 -02 -01 00 01 02 03 04 US$ billion Private creditors 0.4 4.4 0.8 0.3 0.0 ______________________     T he integration of developing countries in the global economy increased sharply in the 1990s with improvements in these countries' economic policies and the massive expansion of global trade and finance driven by technological innovations in communucations, transport, and data management, coupled with some lowering of barriers to trade and financial transactions. Despite limited access to capital markets, many of the low-income developing countries also witnessed a sharp increase in their international financial ties during the 199os. Relative to the size of their economies, the poor countries receive almost the same amount of foreign direct investment-and their residents place about as much money abroad-as other developing countries. These are some of the conclusions reached by Global Development Finance 2002, Analysis and Summary Tables volume, a report that also highlights recent trends and prospects for the global economy, capital flows, and improvements in policies surrounding aid flows. Global Development Finance 2002: *Predicts that a vigorous recovery will begin later this year * Finds that the growth slowdown reduced capital market flows, but not foreign direct investment * Underlines the importance of policy performance for attracting and benefiting from capital flows, both private and official * Concludes that improved policies in both recipient and donor countries are helping to strengthen the effectiveness of aid flows. As in previous years, the Country Tables volume of Global Development Finance 2002 includes a comprehensive set of tables with statistical data for 136 countries that report debt under the World Bank Debtor Reporting System, as well as summary data for regions and income groups. It contains data on total external debt stocks and flows, aggregates, and key debt ratios, and provides a detailed, country-by-country picture of debt. Global Development Finance 2002 debt data are also available on CD-ROM, with more than 200 historical time series from 1970 to 2000, and country group estimates for 2001. Global Development Finance 2002 is an indispensable resource for governments, economists, investors, financial consultants, acadeuics, bankers, and the entire development community. For more information on the analysis, please see www.worldbank.org/prospects; further detail about the Country Tables can be found at www.worldbank.org/data. For general and ordering information, please visit the World Bank's publications website at www.worldbank.org/publications, or call 703-661-1580; within the U.S., please call 1 (800) 645-7247. THE WORLD BANK 1818 H Street, NW Washington, DC 20433 USA Telephone: 202 477-1234 Facsimile: 202 477-6391 Internet: www.worldbank.org E-mail: feedback@worldbank.org 15085 ISBN-0-8213-5085-4